Skip to main content

Full text of "USPTO Patents Application 09870801"

See other formats

IP Workshop: Table of Contents 

Page 1 of 4 

Sponsorec by: 



Coalition for Networked Information 

About CNI 

Task Force 





Site Map 

Search our site 

Proceedings: Technological 
Strategies for Protecting 
Intellectual Property in the 
Networked Multimedia 

Coalition for Networked Information 

Interactive Multimedia Association 

John F. Kennedy School of 

Science, Technology & Public Policy Program 

Massachusetts Institute of 

Program on Digital Open High-Resolution 

Copyright (c)1994 Interactive Multimedia Association. 
Permission to copy without fee all or part of this material is 
granted provided that the copies are not made or distributed 
for direct commercial advantage and the IMA copyright 
notice appears. If the majority of the document is copied or 
redistributed, it must be distributed verbatim, without 
repagination or reformatting. To copy otherwise requires 
specific permission. 

All brand names and product names are trademarks or 
registered trademarks of their respective companies. Rather 


IP Workshop: Table of Contents 

Page 2 of 4 

than put a trademark symbol in every occurrence of other 
trademarked names, we state that we are using the names 
only in an editorial fashion, and to the benefit of the 
trademark owner, with no intention of infringement of the 

Published by: 

Interactive Multimedia Association 
Intellectual Property Project 
3 Church Circle 
Suite 800 

Annapolis, MD 21401-1933 
Phone: (410) 626-1380 
FAX: (410) 263-0590 

Table of Contents 

The Strategic Environment for Protecting Multimedia 
Brian Kahin 

Copyright and Information Services in the Context of 

National Research and Education Network 

RJ. (Jerry) Linn 

Response to Dr. Linn's Paper 

Joseph L Ebersole 

Permission Headers and Contract Law 

Henry H. Perritt, Jr. 

Protect Revenues, Not Bits: Identify Your Intellectual 

Branko Gerovac and Richard J. Solomon 

Intellectual Property Header Descriptors: A Dynamic 

Luella Upthegrove and Tom Roberts 

Internet Billing Service Design and Prototype 

Marvin A Sirbu 


IP Workshop: Table of Contents 

Page 3 of 4 

Metering and Licensing of Resources: Kala's General 
Purpose Approach 

Sergiu S. Simmel and Ivan Godard 

Deposit, Registration and Recordation in an Electronic 
Copyright Management System 

Robert E. Kahn 

Dyad: A System for Using Physically Secure 

J.D. Tygarand Bennet Yee 

Intellectual Preservation and Electronic Intellectual 

Peter S. Graham 

A Method for Protecting Copyright on Networks 
Gary N. Griswold 

Digital Images Multiresolution Encryption 

Benoit Macq and Jean-Jacques Quisquater 

Video-Steganography: How to Secretly Embed a 
Signature in a Picture 

Kineo Matsui and Kiyoshi Tanaka 

Need-Based Intellectual Property Protection and 
Networked University Press Publishing 

Michael Jensen 

The Operating Dynamics Behind ASCAP, BMI and 
SESAC, The U.S. Performing Rights Societies 

Barry M. Massarsky 

Meta-lnformation, The Network of the Future and 
Intellectual Property Protection 

Prof. Kenneth L Phillips 


IP Workshop: Table of Contents 

Page 4 of 4 

Protocols and Services (Version 1): An Architectural 

Consortium for University Printing and Information 
Distribution (CUPID) 

A Publishing and Royalty Model for Networked 

Theodor Holm Nelson 

Acronyms List 

© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Monday, July 2, 2001 . 


IP Workshop - Kahin: Strategic Environment for Protecting Multimedia 

Page 1 of 10 

Sponsorec by: 

About CNI 

Task Force 





Site Map 

Search our she 


I Coalition for Networked Information 

The Strategic Environment for 
Protecting Multimedia 

by Brian Kahin 

The advent of distributed computing over high-bandwidth wide- 
area networks looks like a worst-case scenario for intellectual 
property. Owners of content -- text, images, music, motion 
pictures -- are understandably fearful of releasing proprietary 
information into an environment which is lacking in security and 
has no accepted means of accounting for use and copying. The 
variety of formats and the variety of proprietary interests involved 
complicate the problem and attempts at solutions. 

On April 2 and 3, 1993, four organizations involved in networking 
and multimedia issues sponsored a two-day workshop at 
Harvard's John F. Kennedy School of Government to address the 
problem. These organizations - the Coalition for Networked 
Information, the Interactive Multimedia Association, the MIT 
Program on Digital Open High Resolution Systems, and the 
Information Infrastructure Project in the Kennedy School's 
Science, Technology and Public Policy Program - represented a 
set of different perspectives on what all saw as a broad common 
problem. The workshop was designed to: 

• map the territory between secure systems and the need for 
practical, user-friendly systems for marketing information 
resources and services; 

• survey the technological landscape, evaluate the potential 
benefits and risks of different mechanisms, define a 
research agenda, and frame related implementation and 
policy issues; 

• consider how and where within the overall infrastructure 
different technologies are best implemented; and 

• present and analyze models for explaining protection 


IP Workshop - Kahin: Strategic Environment for Protecting Multimedia 

Page 2 of 10 

systems and strategies. 

Speakers were invited to address these issues along with the 
potential roles of particular technologies and mechanisms: billing 
servers; type-of-service identifiers; header descriptors; labeling 
and tagging; fingerprinting; digital signatures; contracting 
mechanisms; EDI (electronic data interchange); copy protection; 
serial copy management; authentication servers; software 
envelopes; encryption; display-only systems; concurrent use 
limitations; and structured charging. 

In part the workshop responded to the continued dramatic growth 
of the global Internet and the planned National Research and 
Education Network (NREN), the follow-on to the federally funded 
portion of the domestic Internet. The Internet offers the beginning 
of a switched, multifunctional, multimedia environment for sharing 
resources and for marketing information products and services 
in short, for applications and practices that will shape the 
broadband information infrastructure of the future. Complex 
network-accessible library systems have been designed and 
developed for disseminating nonproprietary information, but until 
there are adequate mechanisms and safeguards for handling 
proprietary information, investment will be inhibited. 

At the urging of the Association of American Publishers and the 
Information Industry Association, Congress included in the High 
Performance Computing Act of 1991 provisions that appeared to 
address this problem. The National Research and Education 
Network was to: 

(1) be developed and deployed with the computer, 
telecommunications , and information industries.... 

(5) be designed and operated so as to ensure the 
continued application of laws that provide network 
and information resources security measures, including 
those that protect copyright and other intellectual 
property rights. ... 

(6) have accounting mechanisms which allow users or 
groups of users to be charged for their usage of 
copyrighted materials available over the Network.... 

[15 USC 5512 (c) ] 

The Act also required the Director of the Office of Science and 
Technology Policy to report to Congress by the anniversary of 
the Act (i.e., December 9, 1992) on "how to protect the copyrights 
of material distributed over the Network...." [15 USC 5512(g)(5)]. 
H.R. 1757, the proposed "National Information Infrastructure Act 
of 1993" which has just passed the House, rewrites the 
provisions in the 1991 Act, preserving the mandate on copyright 


BP Workshop - Kahin: Strategic Environment for Protecting Multimedia 

Page 3 of 10 

in the 1991 Act and adding a requirement for research on 
copyright protection. 

However, federal agencies have yet to address these issues in 
depth. Many agency personnel, as well as many within academia 
and the private sector, believe that the protection of intellectual 
property on the NREN, as on any network, needs to be 
addressed at the an applications level, not within the design of 
the network. (Jerry Linn's paper in this volume takes this 
perspective.) Many also believe that the problem should be 
addressed first by the private sector. After all, since there is a 
market for networked information, there should be a market for 
technologies that protect intellectual property. Shouldn't the 
government focus its scarce resources on enabling resource- 
sharing within the research community, where there is relatively 
little need to protect intellectual property? 

However, while the Bush Administration saw the NREN program 
as focused on scientific research, the Clinton/Gore Administration 
envisions the NREN program, and more generally, the Internet, 
as part of a broad strategy to drive the development of a 
commercial information infrastructure which encompasses mass- 
market publishing and entertainment. If this broader goal is 
legitimate grounds for public investment, then arguably the 
government should be involved in supporting mechanisms to 
protect intellectual property. 

Certainly the benefits (new network-accessible resources, etc.) 
that could be generated by the availability of billing servers on the 
Internet could justify public investment. But is the federal 
government, which typically disseminates its own information for 
free or cost, a knowledgeable and careful enough sponsor to 
avoid skewing or prejudicing the playing field for private 
investment? If promulgation of standards would encourage 
private investment, might not private sector organizations 
proceeding through RFTs (requests for technology) do a better 
job leveraging the market? If the government is to be involved in 
standards development, what role should it play? There are 
many different models for government involvement, and broad 
industry support for standards, but little discussion of where or 
how federal support should be implemented. 


Owners of rights to music, images, and other forms of content 
view the emerging network environment as the latest 
evolutionary stage to threaten the stability and security of the 
distribution chain. First there was the transformation of analog 
copying through xerography and electromagnetic recording 


IP Workshop - Kahin: Strategic Environment for Protecting Multimedia 

Page 4 of 10 

(cassette recorders and VCRs). This was followed by the 
digitization of information and the development of the personal 
computer as a general purpose authoring and publishing 
machine of constantly increasing capacity and capability, able to 
manipulate not just text, but sound, images, and finally video. 
The final stage in this evolutionary path is switched broadband 
networking, which allows computer users to publish all over the 
world with great efficiency - a development already in evidence 
within well- networked research communities. Mindful of the free 
and promiscuous behavior of information in this increasingly 
functional and capacious environment, content owners have 
been understandably reluctant to license their property. 

However, the evolution toward a user-enabling broadband 
environment actually brings with it an increased number of legal 
tools for protecting intellectual property (see Figure 1). True, 
there is some uncertainty about the application of these tools, but 
they offer important hooks that can be combined with other 
elements of a property protection strategy. Indeed, from the 
multimedia developer/producer's perspective, these tools may 
add to difficulties in licensing content, because of the need to 
clear additional rights. 

Advancing technology also offers new prospects for securing 
proprietary information so that it cannot be copied casually, 
mediating access so that users can locate and use information 
easily, and assessing charges for access and use in a 
reasonable and comprehensible manner. There is a tension here 
between mechanisms that protect and control, on the one hand, 
and features and characteristics that foster interoperation and 
usability. Limiting technologies may directly inconvenience and 
frustrate users or add to the complexity of a product, increasing 
the likelihood of bugs - problems which have contributed to the 
failure of technological protections in the past. 

environment amM^m^tcmaitfm&n 



electronic media 

© reproduction, pubtk performance 


® rtpro&i&mpubHcpetfomvance, 

patent: manufacture, sale, use 


© reproduction* public performance, 

adaptation, public display 
patent manufacture, sate, me 

Figure t Intellectual pioperiy tiiols in increasingly soph&icated environment 


IP Workshop - Kahin: Strategic Environment for Protecting Multimedia 

Page 5 of 10 

Software copy protection, which was commonplace in the mid- 
1980s, has been all but abandoned. This was partly because the 
Copyright Act allowed users to make backup copies, which 
legitimized the marketing and distribution of software that allowed 
minimally motivated users to unlock copy-protected software. 
Copy protection mechanisms thus proved ineffective for 
determined copiers while they remained awkward and frustrating 
for unsophisticated new users, the very people to whom software 
publishers were looking to expand the customer base. Copy 
protection also imposed unanticipated burdens on the support 
services that software publishers provided to their customers. 

In 1984, ADAPSO (now the Information Technology Association 
of America) proposed an outboard hardware lock as an industry 
standard for copy protection. While this approach appeared more 
effective than software-based solutions, it also raised questions 
of who would pay to implement it, as well as possible antitrust 
problems. Hence, in place of copy protection, the software 
publishing industry has come to rely on the threat of lawsuits in 
the vulnerable corporate environment as a means of copyright 

The problems faced by the ADAPSO proposal can be addressed 
by legislation. In fact, in 1992 Congress amended the Copyright 
Act to mandate a closed hardware-secured environment 
incorporating serial copy management for next-generation digital 
audio recording technology (DART). This elaborate legislation 
included provisions for fees to be levied on hardware and 
recording media to compensate the owners of rights in music and 
sound recordings. However, the computer industry took care to 
ensure that the complex DART regime was strictly limited to 
consumer audio technology and did not affect the nascent 
multimedia industry. 

The Copyright Act of 1976 was carefully designed to be 
technology-neutral. With the exception of the provisions on cable 
retransmission, it is an elegant piece of legislation in which 
general principles are applied with remarkable uniformity to many 
different kinds of works. But the practicalities of enforcing 
copyright protection reveal critical differences among types of 
information. Whether the work is text, images, sound recording, 
video, or computer program makes a big difference - as does 
whether it is analog or digital, or whether it is mass- market or 
niche-market. The one-size-fits-all vision has been eroded by the 
need to address special problems within particular industries. So 
legislation has addressed these issues case by case, as in the 
1980 amendments concerning computer software (codified as 
Section 117), the Record Rental Amendment Act of 1984, and 
the Computer Software Rental Amendments Act of 1990 


IP Workshop - Kahin: Strategic Environment for Protecting Multimedia 

Page 6 of 10 

The DART legislation is the latest example, and it foreshadows 
similar issues presented by the advent of digital video 


The nature of the threat is important in assessing the need for 
special protection. There are three distinct possibilities. First, 
there is true piracy, the making of unauthorized copies for sale 
(or selling unauthorized access to transmissions); second is 
unauthorized copying in a business environment; third is erosion 
of the consumer market by copying and redistribution among 
family and friends. 

Protection against piracy is facilitated by the fact that the bigger 
and more successful the operation, the more visible and 
vulnerable it becomes. Criminal penalties are available under the 
Copyright Act, which means that copyright owners can expect 
help directly from the government in such situations. But today 
the big piracy problems are concentrated in particular countries. 
Protection from foreign piracy ends up as a political issue: How 
much pressure is the U.S. willing to place on certain 
governments to crack down on pirate operations within their 
borders? Typically, this pressure is applied in the process of 
trade negotiations. 

The second area, protecting against unauthorized copying within 
businesses, is an issue principally for software publishers. The 
Software Publishers Association (SPA) has developed a very 
effective program to combat the problem by advertising a hotline 
and relying on disaffected former employees to report improper 
copying. In this case, the threat of liability and attendant bad 
publicity appears to have had significant impact on software 
management practices, at least within the U.S. 

The third area, erosion of the consumer market through 
consumer copying, is perhaps the most problematic. It is 
impractical, if not impossible, to control through litigation. Indeed, 
to some degree, consumer copying is a common, socially 
accepted practice. This is especially true for the copying of 
audiocassettes and CDs and for the videotaping of broadcast 
and cable television. The DART provision for serial copy 
protection is relatively weak in that it does not preclude making 
multiple copies from the original purchased product; it only 
precludes making copies from the copies. SPA opposed the 
DART legislation because it legitimized personal copying, 
thereby strengthening attitudes that might carry over to computer 
software. Ironically, unauthorized copying of software may, in 
fact, enhance opportunities to market new versions, as recent 


IP Workshop - Kahin: Strategic Environment for Protecting Multimedia 

Page 7 of 10 

promotional offers of free financial management software have 

Furthermore, the ability to make copies increases perceived 
value. For example, the licensing of movies to cable, including 
"pay-per-view," undoubtedly results in considerable home 
copying and retention of such copies by consumers. But the fact 
that consumers can get relatively high-quality copies in this 
manner (at least compared to copying from a videocassette) 
increases their willingness to pay for premium cable services and 
pay-per-view cable. This in turn is presumably reflected in the 
licensing fees that cable services are willing to pay movie studios 
for their product. Similarly, the fact that CDs can master better 
cassette copies than cassettes undoubtedly helps sustain higher 
retail prices for CDs. 

There are also editorial and marketing strategies to minimize 
consumer erosion. In general, a product that is part of a series or 
a larger whole is less susceptible than a standalone product. 
Examples include the versions of software, the sound recordings 
of a particular artist or group, and subscriptions to a series. 

While consumer copying of videocassettes, sound recordings, 
and computer software has been widespread, it is not clear that 
still images will be copied and circulated to the same degree. 
There is simply is not the same kind of substantial, specific 
demand for individual photographs that there is for popular 
songs, recent movies, and software. Images are generally 
marketed in collections, and indeed there may be a market for 
electronic image collections analogous to coffee table books or 
home videos. Such collections, like other CD-ROM-based 
multimedia products, would be difficult to duplicate for the 
foreseeable future, and extracted images may have little value in 

It should be relatively easy for multimedia publishers to license 
works, and especially fragments of works, that have little value in 
isolation. Although content owners may well be concerned about 
context, a clip from a song or a movie may stimulate demand for 
the original. A run-time version of a software program may elicit 
interest in the fully functional original. Abstracts of journal articles 
can elicit interest in the full text. 

These observations highlight the critical distinction between 
technology used to limit access and technology as a facilitator. 
The former includes restrictive technologies such as encryption, 
user authentication, and copy protection. Facilitative technology 
aims to provide a seamless interface to information which 
enables the user to navigate and synthesize the information as 


IP Workshop - Kahin: Strategic Environment for Protecting Multimedia 

Page 8 of 10 

transparently as possible. This can add enormous value by 
putting information in a rich and useful context. The availability of 
functionally and contextually enriched information diminishes the 
value of the same information in flat and isolated form and 
therefore reduces incentives to extract and redistribute content. 
Of course, systems can combine restrictive and facilitative 

Online systems can also enable continuing contractual 
relationships between publishers and end-users. Contracts can 
supplement copyright protection and are especially important for 
databases of factual material, where copyright protection may not 
be available for individual records. By contrast, contracts are very 
difficult to establish in a retail sales environment, notwithstanding 
the ambitious claims in shrink-wrap licenses. 

There are practical limits to technology-mediated access. Online 
vendors have pioneered the use of complex pricing algorithms in 
which users pay for connect time, searches, hits, and volume - 
all of which relate to cost or value. But most users, especially 
inexpert users, prefer the flat-rate pricing associated with CD- 
ROM databases, which is easy to budget for and encourages 
experimentation and use. Most consumer online services now 
mix a flat rate for basic services with metering for premium 
services. Flat-rate pricing is the norm for most information 
transactions: books, cable television, multimedia products, 
videocassettes, CDs, newspapers, videogames, computer 

Flat-rate pricing is not necessarily per-copy. Software, for 
example, may be licensed on a per-copy, per-user, per-machine, 
per-site, concurrent-use basis, or some combination thereof. 
Licensing the software for use only on a particular computer may 
have made sense for mainframes, but it fits less well in a 
distributed computing environment in which users may have 
access to several computers at different times. There is growing 
acceptance of concurrent licensing (with software lockout when 
the authorized number of users is reached) as a fair method of 
licensing programs for use over a local area network. Per-copy 
licensing remains easy to enforce under copyright law and, in 
fact, provides the basis for SPA's auditing and enforcement 
program. However, few individual users are inclined to uninstall 
software from one computer just so they can use it temporarily on 

Pricing and licensing strategy can be viewed as a kind of soft 
intellectual property protection. If users feel that prices are fair 
and reasonably related to use, they will be less inclined to look 
outside legitimate distribution channels or to make copies for 


IP Workshop - Kahin: Strategic Environment for Protecting Multimedia 

Page 9 of 10 


Labeling is another soft strategy that can take on a wide variety 
of forms: copyright notices on every page; "FBI warnings" on 
videocassettes; personalized sign-on screens; appeals to the 
user's sense of fair play and appreciation for the product or 
service. Labeling can usually be embedded in the content, so 
that it cannot easily be removed. It thereby diminishes the 
experiential value of the content (which is therefore less likely to 
be redistributed) or makes it clear that copies are derived 
improperly from the original context. Alternatively, labeling can be 
made invisible so that it becomes a "fingerprint," which, when 
properly decoded, reveals the original source of pirate copies. 

Figure 2 illustrates strategic options for network publishing along 
two dimensions. The vertical dimension extends from inclusive 
strategies to facilitate use and expand the market to exclusive 
strategies which maintain the market by excluding nonpaying 
users. The horizontal dimension shows the spectrum of strategic 
tools that extend from marketing and legal tools on the left to 
purely technological tools on the right. 

The diagram shows the importance of expanding the network of 
users as well as the need to limit that network. At the policy end, 
one former objective is typically assigned to the marketing 
department, the latter to the legal department. These divisions 
embody different cultures and sometimes do not communicate 
well with each other. However, the technology end of the diagram 
is entirely in the hands of designers and engineers. The exclusive 
mechanisms and the inclusive mechanisms, like the designers 
and the engineers, must work well together to co-exist in the 
same product. 






WAiS, gopher 














riguie 2, Strategic dimensions for network publishing 


IP Workshop - Kahin: Strategic Environment for Protecting Multimedia 

Page 10 of 10 

In the end, corporate strategy must integrate tools for identifying 
and controlling intellectual property with a broad understanding of 
marketplace realities and the legal framework for licensing 
distribution and use. While there remains great uncertainty about 
how multimedia information will be stored, processed, and 
delivered, and uncertainty about the scope and characteristics of 
the market, it is clear that the options are many and that 
navigating the networked multimedia environment demands 
unprecedented thought and skill. 


Brian Kahin is Director of the Information Infrastructure Project in 
the Science, Technology and Public Policy Program at Harvard's 
John F. Kennedy School of Government and General Counsel for 
the Interactive Multimedia Association. He recently edited 
Building Information Infrastructure (McGraw-Hill, 1992), a 
collection on papers on issues in the development of the National 
Research and Education Network. 


©2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


IP Workshop - Linn: Copyright and Information Services 

Page 1 of 14 

Sponsorec by: 


About CNI 

Task Force 






Site Map 

Search our she 


Coalition for Networked information 

Copyright and Information 
Services in the Context of the 
National Research and Educatic 

Network W 

by R.J. (Jerry) Linn 


The High Performance Computing Act (HPCA) of 1991 (P.L. 102-19 
places unenforceable requirements to protect copyrights and intellec 
property rights on the National Research and Education Network 
(NREN). This paper discusses the roles and responsibilities of the N 
and associated information services; technical approaches to 
authentication, redistribution and authorization of use of electronic 
documents over the NREN; and an amendment to the High Perform 
Computing Act. 


It is clear that when the High Performance Computing Act of 1991 w 
written the notion of digital libraries was a consideration of the autho 
is also clear that the Congress intended that copyrighted materials b 
distributed over the Network. The legislative history of the Act affirm; 
position. The Act, as drafted in the 100th Congress (S.1067 and H.F 
3131 , 1990), included provisions for authorization of appropriations 1 
the National Science Foundation to establish digital libraries. Other I 
introduced into Congress which provide for similar authorization of 
appropriations include S.2937, introduced in 1992, and S.4, introduc 
1993. Indeed, prior to 1991, digital libraries were integral to the think 
related to "information services" and were the stimulus for language 
incorporated into the HPC Act of 1991 with respect to protection of 
copyright.[2] The language employed in the Act of 1991 assumes 
information services are embedded in the Network, as part of a netw 
infrastructure. However, the term "network" has very specific and na 
connotations when used by professionals in the computer and 


IP Workshop - Linn: Copyright and Information Services 

Page 2 of 14 

communications communities versus the broad definition of the terrr 
the Act. Furthermore, recent papers and reports from a workshop 
focused on the National Research and Education Network (NREN) r 
the common understanding that the NREN is only an access mediur 
application services.[3] Therein lies the weakness of the legislation: 
definition of the "Network" is too broad to assign responsibility for 
protection of copyright and intellectual property rights. Furthermore, 
professional community to whom the courts would turn for expert 
witnesses to aid in interpretation of the law is not likely to agree with 
reasonableness of the requirements that the Act places on operator; 
the Network or the ability to enforce its provisions except in the 
computers attached to the Network which offer information services. 

Specifically, the Act defines the "Network" in Sec. 4 as follows: 

(4) 'Network 1 means a computer network referred to as 
the National Research and Education Network established 
under section 102; and Sec. 102 (c> "Network 
Characteristics" states: 

The Network shall - 

(5) be designed and operated so as to ensure the 
continued application of laws that provide network and 
information resources security measures, including those 
that protect copyright and other intellectual property 
rights, and those that control access to data bases and 
protect national security; < 

(6) have accounting mechanisms which allow users or 
groups of users to be charged for their usage of 
copyrighted materials available over the Network and, 
where appropriate and technically feasible, for their 
usage of the Network; 

There are several important things to note because they become "fir 
premises" for a discussion. First, the NREN is a concept (the Act ne 1 
defines who owns and operates it; the Act authorizes appropriation c 
Federal funding to agencies to implement the concept). Second, the 
NREN is a logical entity derived from a network of networks (an intei 
And third, the NREN is a part of the /ntemef-that network of networl 
whose span is global and whose common denominator is a shared r 
and address space. 

The Network established under Sec. 102(a) does not imply that the 
Federal government installs or owns the physical assets of the NRE 
(e.g., optical fiber cables, routers) nor does it preclude the NREN fro 
being derived from commercial, private sector sources and services. 
ambiguity is important. The definition and ownership of the NREN ar 
cast in concrete (like highways); this omission allows the NREN (or | 


EP Workshop - Linn: Copyright and Information Services 

Page 3 of 14 

of it) to transition from government provided and/or subsidized servit 
to commercial for-profit services, or an evolving combination of both 
Evolving Federal policy supports transition to commercial services a 
required services become commercial commodities. 

Which networks comprise the NREN and who owns/subsidizes then 
not as important as understanding that "ownership" of subnetworks, 
levels of subsidy and recipients of subsidy are all subject to change 
time. Therefore, defensible answers for issues related to copyright, 
intellectual property rights and the NREN must take into account the 
diversity of the technology base in component subnetworks, of 
ownership, of agency missions and goals, and of those services 
accessed by the NREN versus common services provided by 
subnetworks comprising the NREN/lnternet This complexity sugges 
that it will be beneficial to partition the problem into smaller compone 
for analysis and discussion. 

Subsequent subsections present the "Network" as a set of services, 
establish both technical and pragmatic reasons for doing so, and dis 
protection mechanisms appropriate for the decomposed services. 
Specific technical mechanisms are outlined which may be employed 
distribute and protect copyrighted materials by an information servio 
Finally, an argument is presented that the HPC Act of 1991 should fc 
amended such that the protection of copyrights and intellectual prop 
is properly the responsibility of information service providers and us« 
An amendment is offered which would realize the position presentee 


A delineation of network services aligned with widely recognized 
technical boundaries and terms will aid in a dialogue because functit 
and responsibilities can be discussed within an established framewc 
Professionals familiar with network architectures associate specific 
functions and services with well-known named layers of a network 
architecture. The terms and concepts used below are recognized by 
international community of computer and communications professioi 
[4] Thus, it is unnecessary to define new terms and concepts in orde 
establish a framework to discuss issues. 

The functions associated with the two lowest layers of a network 
architecture are physical, point-to-point connectivity and signaling, a 
data transmission via data links which interconnect computers or rot 
Next in the hierarchy are network-layer functions which select routes 
relay data packets enroute to their destination. These functions are 1 
least common denominator of a "computer network" and are often 
implemented by routers which comprise or interconnect wide area 

the transport layer establishes end-to-end connectivity and may pro 


IP Workshop - Linn: Copyright and Information Services 

Page 4 of 14 

for retransmission of data packets lost or corrupted by lower layers. 
Thus, the transport layer provides a reliable end-to-end communicat 
medium for application programs and services. Note that the public 
switched network may also be used to provide an end-to-end 
communications path between computers; however, end-to-end 
communications is achieved by different technical means. 

Information services are provided by application-layer programs and 
supporting protocols at the end points of a communications path. 
Examples of application-layer services are electronic mail and file 
transfer, which are implemented by application-layer protocols (e.g., 
Simple Mail Transfer Protocol (SMTP) and X.400 are electronic mail 

Connectivity of subnetworks in the NREN/lnternet functions at the 
network layer (see Figure 1). Each subnetwork serves as a switching 
fabric for a set of computers; i.e., the network layer software receive 
relays packets of data from one node in the network to another node 
based only on its destination address. Note that the routing and rela 
(switching) functions assigned to the network layer are the least conr 
denominator of the Internet (NREN). Specifically, a subnetwork (e.g. 
college campus, midlevel network or the NSFnet of the National Sci< 
Foundation) may use one set of technologies and another subnetwo 
may use another. However, the "glue" that interconnects them is a 
common, minimal set of protocols necessary to provide the roi 
and relay functions. Any additional set of functions is optional in th* 
network layer and is only likely to be incorporated if actually requirec 
given environment (e.g., security, network management). Therefore, 
assumption that the "Network" is a uniform, ubiquitous environment 
erroneous-particularly when the NREN is viewed as a set of 
interconnected autonomous subnetworks. 

This is a greatly simplified sketch of a multi-layered network architec 
The sketch highlights crucial networking design concepts; i.e., specr 
functions are assigned to layers in a network to accommodate an ar 
lower-layer communications technologies and for design and 
maintenance purposes. However, we have sufficient information anc 
set of terms which is rich enough to pose questions about how and 
where the requirements of the Act might be implemented and to exp 
why they might or might not be reasonable requirements in the first 
place. We are also prepared to identify and discuss conflicting objec 
if proper design and engineering principles are not followed. 


IP Workshop - Linn: Copyright and Information Services 

PageS of 14 

Thus, when the Act states: 

the Network shall — 

(5) be designed and operated so as to ensure the 
continued application of laws that provide network 
and information resources security measures, including 
those that protect copyright and other intellectual 
property rights, and those that control access to data 
bases and protect national security; 

we can ask: "What does this mean? What protection is required? He 
may required protection be achieved in the context of existing netwo 
architectures? And, who should be responsible?" 

Clearly, the routing and relay functions of the network layer will not 
protect copyrighted materials. In fact, they do not even assure delive 
data packets. Therefore, the "Network" described in the Act requires 
more functions than those described for the network layer. So it is 
appropriate to ask: "What protection is inherent in a network; what 
additional protection is required; and where is it most appropriately 

Under normal circumstances, network-layer software does not inspe 
the contents of data packets. There are at least two good reasons m 
do so. First, inspection of packets for any purpose introduces 
unnecessary overhead and degrades the throughput of the network 
serious consideration in high-speed networks). Second, inspection c 
packets (or streams of data) jeopardizes the privacy of the informatk 
being transmitted. Also, recall that the network-layer software was 
described earlier as "least common denominator," with the implicatic 
that any additional functions were optional. Consequently, the netwc 
layer is not a viable candidate for uniform protection of copyrighted 

Data integrity protection against accidental changes is assured if sp< 


IP Workshop - Linn: Copyright and Information Services 

Page 6 of 14 

transport protocols are employed at the end points of a connection. 
Specifically, the network can protect against accidental loss or corru 
of data during transmission from one point to another. This is true fo 
Transmission Control Protocol (TCP) and the Organization for 
International Standardization (ISO) Transport Class 4 (TP4); both de 
and retransmit lost and corrupted data. However, the transport proto 
cannot protect against redistribution of materials obtained from a 
legitimate source, nor can they assure the authenticity of the materia 
transmitted over the network. The means to assure authenticity of 
materials and achieve protection from deliberate abuse by end usen 
implement the required protection mechanisms in computer systems 
part of the application programs which deliver services to users. 


New protective services can be created for information disseminatio 
which can also be applied to those materials that have a copyright 
However, requirements for protection must be defined before descril 
how protection might be achieved. Below is a set of requirements wl 
serve as a starting point for a discussion. 

Protections and Features Required 

Authentication: A mechanism is required to certify that any material 
received is a bona fide copy of the original (data authentication) and 
possibly who it came from (origin authentication). If the copy is not 
authentic, then this fact should be detectable and the copy discarde< 
Recall that the transport layer may provide for integrity protection ag 
accidental changes, but authentication provides a means for protect 
against both accidental and intentional changes. 

Limited redistribution: Publishers want to control distribution to those 
have paid a fee for the use of copyrighted materials. Mechanisms sY 
be implemented to restrict the number of copies printed to those pai< 
and to the individual who paid for them. 

Protection against plagiarism and change: Authors and publishers d- 
want their materials used without appropriate attribution, nor do they 
want the materials excised, edited, or modified such that authenticity 
jeopardized. Information should be stored in a form which makes it 
difficult, if not impossible, to remove the copyright mark, or excise or 
modify text. 

Object form: Information should be stored and exchanged in 
standardized but device-independent forms. Processing software 
employed by a user should display or print the materials in an 
appropriate form given the constraints of the user's video display an< 


IP Workshop - Linn: Copyright and Information Services 

Page 7 of 14 

To discourage plagiarism, excising parts of the text and other 
unauthorized uses of the information, an object could be put in a "se 
envelope' 1 and distributed in one of several forms which are not easi 
read and modified by humans. These forms could include SGML, G' 
and PostScript or other useful forms. SGML denotes the Standard 
Graphics Markup Language. SGML text would require processing ol 
input text to render meaningful output on either a video display or pr 
G4Fax denotes Group 4 Facsimile which is a compressed bit streanr 
using an international standard for scanning and compressing facsin 
images. It may be displayed or printed on raster scan output devices 
(video display or printer). G4Fax could readily be used for interlibran 
exchange to avoid document handling and scanning. PostScript den 
the form used by PostScript printers. It is a page description languac 
that is widely implemented, is useful for printing purposes only, and 
would not require significant processing if directed to a printer. 

Appropriate remuneration: Remuneration could take the form of a 
subscription fee, license fee, contract, or fee for services rendered, i 
appropriate. Dissemination may be by an author, original publisher, 
information service, or library (hereafter called an authorized distribi 

It is assumed that interlibrary loan and electronic redistribution of sin 
copies of papers to individuals by libraries who have a subscription, 
license or contract with a publisher constitutes "fair use." It is also 
assumed that fees for services will be established (commercial, for-f 
and not-for-profit) and public access could be via public libraries. 
Specifically, an individual could ask for and get a copy of a paper or 
article as easily as he or she can reproduce it on a copier in a library 
at a comparable price). Remuneration by an individual patron could 
the time the material was obtained, if there was a fee. 


A set of mechanisms may be combined to address the requirements 
outlined above. For discussion purposes, we consider a body of mat 
(information) as an "object" with certain components and attributes. ( 
attribute is an electronic "copyright" mark; the object forms noted eai 
are another attribute. Object-oriented technology associates process 
of objects with their attributes. For simplicity, however, we describe < 
object as an envelope and its contents. The information on the enve 
is visible and the contents hidden and sealed with a digital signature 
Examples of information on the envelope could include title, author^ 
abstract, keywords (e.g. full bibliographic record) and attributes 
describing the form of the object, a digital signature, copyright status 
(yes/no), and date and timestamp associated with an authorized cop 
Visibility of information on the envelope has other obvious advantagi 
related to search and retrieval of information stored in digital librarie; 
they are outside the scope of this paper. Figure 2 presents a graphic 


IP Workshop - Linn: Copyright and Information Services 

perspective of the concepts. 

Page 8 of 14 


For our purposes we assume: 

• an object is processed by standardized software (hereafter cal 
rendering software)] 

• creating an original information object, file transfer over a netw 
and rendering of the information on a video display or printer a 
built-in functions of the rendering software; 

• the rendering software is 

o inexpensive or free because it is in the interest of the put 
and of publishers and authors to protect their intellectual 
property, and 

o widely available; e.g., distributed by publishers, informal 
service providers, computer manufacturers; 

• copies of objects are exchanged using the rendering software- 
copy is obtained from an authorized distribution source (may b 
individual if there is no fee for use); and 

• the structure and exchange formats of objects are standardize* 
(either de facto or de jure). 

Active Protection Mechanisms 

Two active mechanisms implemented in the rendering software will 
achieve the requirements for protection outlined in the previous sect 

Authentication: Confirmation of authenticity of the source and conter 
the envelope can be achieved by use of a public key, digital signatui 


IP Workshop - Linn: Copyright and Information Services 

Page 9 of 14 

algorithm. The public key is provided by the author or publisher and 
written on the envelope. The public key is used to verify the digital 
signature of the information written on the envelope and its contents 
either is changed, the digital signature verification algorithm detects 
reports failure. If verification failure is detected when an object is bei 
obtained from an information service, its retransmission should be 
requested. (This might occur if data were lost or corrupted.) If a failu 
detected when displaying or printing an object, further processing sh 
be inhibited. This might indicate a bootleg copy, or a mismatch of us 
identification with that on the envelope, or it might indicate that the 
authorized number of copies have been printed. Optionally, the obje 
could be destroyed by the rendering software when verification fails. 

Limited redistribution: Identifying the holder of the copy on the envel- 
(e.g., user identification) and a copy counter can be employed to lim 
electronic redistribution. The user identification and initial value of th 
copy counter stored on the envelope are established when a copy is 
obtained from an authorized distribution source. The number of prinl 
copies allowed is a function of the fee paid. The copy counter is use- 
restrict the number of copies rendered on a printer. As the copy coui 
is decremented, a residual copy count and new digital signature is 
computed and affixed to the envelope to prevent an unlimited numb* 
copies from being printed. 

Note that sending a copy of an object via electronic mail, redistributh 
a bulletin board and other simple copying mechanisms will not upda* 
contents of the envelope which contains the date and timestamp of < 
authorized copy. If the date and timestamp in the directory entry for ; 
containing an object do not match those in the envelope of the objec 
rendering software considers the copy to be unauthorized. Consequ 
the information contained in the envelope will not be presented to a 
by the rendering software and unauthorized copies are useless. 

Materials may be displayed on a video display an unlimited number 
times by the user identified on the envelope. Other users are prohibi 
from displaying an object with the "copyright" attribute. However, 
unlimited rendering and redistribution is permitted if an authorized 
distribution source omits the "copyright" attribute on the envelope, 01 
enters "unrestricted" in either the user identification or copies author 

Passive Protection Mechanisms 

Object form: The object forms described above are not human 
interpretable forms (SGML, G4Fax, PostScript). Furthermore, an obj 
stored in a form which may not be displayed or printed without the 
rendering software unless it is extracted from within its envelope. 
Although this is a passive protection mechanism, significant technic? 
information and expertise are required to defeat it. 


IP Workshop - Linn: Copyright and Information Services 

Page 10 of 14 

Note that all the forms described above prevent easy redistribution t 
simply making a copy and mailing or printing it with utility software 
because the rendering software is required to display or print an obj( 
These forms also inhibit using a simple editor to "cut and paste" text 
another document because no form is human readable, and direct u 
access into the contents of the envelope is not allowed by the rende 

Write protection: Write protection is the first line of defense required 
protect the authenticity of information disseminated by an informatio 
service. It restricts the privileges to create or modify stored informati' 
the rightful owner(s); these are called "write privileges" associated w 
file. Restrictions are essential for any information service and must t 
implemented within the computer system offering the information se 
Write protection is not a function of the "Network" but is a responsibi 
the parties operating an information service. 

In summary, two active forms of protection are proposed for intellect 
property: authentication and limited redistribution. Two complements 
passive mechanisms are also identified, but are inadequate on their 
(object form and write protection). All the mechanisms suggested an 
implementable on computers accessed by a network, and are comp 
independent of the networking technology used to access an inform; 
service. All mechanisms are applicable to any information distributee 
over a computer network whether or not the information carries a 
copyright mark. 


Separation of the roles and responsibilities of the "Network" and 
"information service providers" provides a logical and pragmatic 
framework for disentangling and discussing the legal and technical 
issues related to the NREN and copyright. 

First, the NREN is a concept (or logical entity) rather than something 
physical with fixed boundaries. The present and future NREN will be 
of the global Internet. As such, its owners are both public and private 
entities, and it is not uniform in the underlying technology deployed. 
Pragmatically, it is impossible to require any owner of part of the Inte 
(a subnetwork) to add new, optional network functions which do not 
serve the owners immediate needs. Consequently, the Network as ; 
whole can only provide the "least common denominator" services wi 
respect to networking functions. These common functions are selecl 
of routes and forwarding packets enroute to their destination; this is 
called "packet switching." Often, technical people think of the "Netwc 
in terms of these limited functions; e.g., NSFnet provides the packet 
switching and routing functions to interconnect other networks. 

Second, the language of Sec. 102 (c)(5) implies that operators of 

http ://www . cni . org/docs/ima. ip- workshop/Linn.html 


IP Workshop - Linn: Copyright and Information Services 

Page 11 of 14 

subnetworks which are part of the Network could be liable for the ille 
actions of both the providers of information services accessed via th 
Network and the users of these information services; i.e., "must be 
designed and operated to ensure ... including those that protect 
copyright ..." 

These requirements to protect copyrights and intellectual property rij 
are at odds with established protection for common carriers who alsi 
provide networks capable of providing access to information service 
which distribute copyrighted materials. Carriers are not liable for the 
illegal activities of their users. Surely, a telephone company would n< 
held legally liable if an information service used facsimile machines 1 
illegally sell and distribute journal articles. Note that it is technically 
feasible for the NREN to become integrated with the public switched 
network in the near future (e.g., narrowband ISDN services (Integrat 
Services Digital Network) could be used to access the Internet). Usii 
this situation as an example, there could be a dichotomy in terms of 
requirements and liabilities related to operators of subnetworks with 
respect to a single illegal act; e.g., if part of the access path was via 
public switched network and part via a midlevel network. 

Third, consider that the "operator of the Network" is responsible for 
collecting and redistributing fees to the "appropriate entity" for use o 
copyrighted materials (c.f. Sec. 102 (c)(6) in the introduction). Is it lih 
that private sector providers of information services (e.g., a publishe 
want an intermediary (Uncle Sam/Federal agencies) to collect and 
redistribute funds for services rendered? Even if an information serv 
did want this service, which "network operator" is responsible (or wo 
accept the responsibility)? Federal agencies operating a subnetwork 
not want the responsibility of collecting and redistributing fees for pri 
sector parties. Note that definitions of "operator of the Network" and 
"appropriate entity" (author, publisher, ...) are open questions. 
Particularly, when user access is granted via a sequence of subnets 
who is the network operator? Is it the "operator" who provides the "u 
access to the network, the operator who connects the information 
service, both, or some more complex combination? 

Finally, a number of network-independent mechanisms may be emp 
by information service providers to limit redistribution and assure tha 
copies remain unmodified. These include data compression, authori; 
use meter (copy counter), and public-key, digital-signature technique 
Digital signature can be employed as a tool to "seal an envelope" an 
verify the authenticity of copyrighted materials distributed over the 
Network. These mechanisms can be implemented to protect copyrig 
and the interests of publishers and authors completely independent 
the network technology used to access the materials. 

Definition of standardized technical practices to achieve the desired 
results and inexpensive software to distribute, protect and render 


IP Workshop - Linn: Copyright and Information Services 

Page 12 of 14 

copyrighted materials are all that are needed to protect the interests 
publishers and achieve the intent of the High Performance Computir 


Sections 102 (c)(5) and (6) of the Act place unrealistic and unenforo 
requirements on the "Network" and its operators (Federal, State or 
private sector parties) to (1) protect copyrights and intellectual prope 
rights; and (2) account for use, collect fees and remunerate copyrigh 
holders. These should be the responsibility of the information service 
providers and users of information services. These are unrealistic 
burdens to place on Federal agencies or private sector operators of 
subnetworks which are part of the NREN (Internet). 

While it is impossible to assure complete protection against maliciou 
individuals, the appropriate remedy is to develop and deploy technic 
protections in the appropriate places, and apply the law in the same 
manner it is used to prevent bootleg copies of paper documents beir 
reproduced on copiers. 

The rationale developed in this paper could be used to interpret the 
existing law and develop regulations and rules aligned with the prop- 
amendment. If regulations and rules with the same intent were writt€ 
they would not clarify the intent of Congress[6J and would be more 
readily challenged in the courts. An amendment would clarify the int< 
Congress and make the law enforceable. The author believes that c 
on this issue is in the public interest as well as that of authors and 
publishers. To this end, an amendment is proposed as an appendix. 


Proposed Amendment to the HPC Act of 1991 

Insert the following definition at the end of Sec. 4. 

"(6) "Information Service Provider" means an entity 
or individual who disseminates information, data, or 
copyrighted materials to others, for free or for fee 
as appropriate." 

(Note that this definition is broad enough to include libraries, for-prof 
publishers, or individuals who want to participate in an "electronic pr 
-and is not restricted to the dissemination of copyrighted materials). 

Substitute the following for Sec. 102 (c)(5) and (6): 

The Network shall — 



IP Workshop - Linn: Copyright and Information Services 

Page 13 of 14 

"(5) be designed and operated so as to enable the 
continued application of laws, regulations, directives 
and standards that prescribe security measures for networ 
and information resources and those that control access t 
data bases and protect national security; 

"(6) have accounting mechanisms which allow users or grou 
of users to be charged for their usage of the Network, wh 
appropriate; " 

and insert after Sec. 102 (e) - 

"(f) Information services which distribute copyrighted 
information shall be designed and operated so as to enabl 
the continued application of laws which protect copyright 
and other intellectual property rights, including appropr 
remuneration of copyright holders, while allowing for the 
s fair use 1 provisions of the copyright law." 

and renumber Sec. 102 "(f)" and "(g)" as "(g)" and "(h)". 

1. This paper is a contribution of the National Institute of Standards < 
Technology. As such, it is not subject to copyright. The opinions 
expressed in this paper have not been endorsed by the Federal 
Networking Council, or any other federal working group. 

2. These provisions were first specified in a draft of H.R. 3131 , Title 
"Information Services," Sec. 302, "Copyrighted Materials," 1990. 

3. Proceedings of the NREN Workshop, Monterey, CA, Sept. 16-18, 
1992, EDUCOM. 

4. The terminology employed is based upon the "Open Systems 
Interconnection-Basic Reference Model," published by the 
Organization for International Standardization in 1984. A similar 
delineation of functions and terminology is used in the Internet 
architecture defined by the Internet Architecture Board/Internet 
Engineering Task Force. 

5. The techniques described in this paper are equally applicable to c 
media other than video displays and printers. Thus, "object form" is 
intended to denote some machine-processable form of digital inform 
which requires "rendering software" to present the content of the 
information in a human interpretable form-video, audio, printed text 
some combination thereof. 

6. In a conversation with the author, Mike Nelson ,who was on then 

http : //www . cni . org/docs/ima. ip - workshop/Linn .html 


IP Workshop - Linn: Copyright and Information Services 

Page 14 of 14 

Senator Gore's staff and now is in the Office of Science and Techno 
Policy in the White House, said/'Yes, we knew headers were require 
but protection of copyright by the * Network' is essential. Thus, the la 
reflects the intent of Congress." 


RJ. (Jerry) Linn, a computer scientist, is Associate Director of the 
Computer Systems Laboratory at the National Institute of Standards 
Technology (NIST) in Maryland. As a Commerce-Science Fellow in ' 
U.S. House of Representatives, he worked on the High Performance 
Computing Act of 1990. His research activities include formal protoc 
design, specification and testing. 

R.J. Linn 

Associate Director for Program Implementation 
Computer Systems Laboratory 
B164 Technology Bldg. 

National Institute of Standards and Technology 
Gaithersburg, MD 20899 
linnr j@osi . ncsl . nist . gov 


© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 1 of 27 

Sponsor ec by: ^H%TMfc T| 

a tni 

p , 

ARL | Coalition for Networked Information 

About CNI 

Task Force 





Site Map 

Search our she 

Permission Headers and 
Contract Law 

by Henry H. Perritt, Jr. 

Law and technology must work together to minimize free riding on 
the intellectual contributions of authors and publishers. Contract-law 
and evidence-law doctrines can protect contributors in well-designec 
digital library systems, but unduly relying on encryption and other 
sophisticated technologies to protect intellectual property frustrates 
the vision of .an open architecture for electronic publishing because 
they impose transaction costs disproportionate to the risk. 


One of the ways to protect intellectual property on the NREN is 
through a digital library concept. Under this concept, a work would 
have attached to it a "permissions header" defining the terms under 
which the copyright owner makes the work available. The digital 
library infrastructure, implemented on the NREN, would match 
request messages from users with the permissions headers. If the 
request message and the permissions header match, the user wouk 
obtain access to the work. This concept encompasses major aspect 
of electronic contracting, which is already in wide use employing 
electronic data interchange (EDI) standards developed by ANSI 
Committee X12.[1] 

This paper explains the relationship between the digital library 
concept and EDI practice, synthesizing appropriate solutions for 
contract law, evidence, and agency issues that arise in electronic 
contracting. The question of how electronic signatures should work 
to be legally effective is an important part of this inquiry. The paper 
also defines particular types of service identifiers, header 
descriptors, and other forms of labeling and tagging appropriate to 
allow copyright owners to give different levels of permission, 
including outright transfer of the copyright interest, use permission, 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 2 of 27 

copying permission, distribution permission, display permission, and 
permission to prepare derivative works. The paper considers how 
payment authorization procedures should work in conjunction with a 
permissions header and digital library concept in order to integrate 
the proposed copyright licensing procedures with existing and 
anticipated electronic payment authorization systems. The paper 
necessarily considers whether existing standards approaches 
related to SGML (Standard Graphics Markup Language) and X12 
are sufficient or whether some new standards development efforts 
will be necessary for implementation of the concepts. The paper 
considers- the relationship between technology and law in enforcing 
intellectual property, and emphasizes that the traditional adaptation 
of legal requirements to levels of risk is appropriate as the law is 
applied to new technologies. 

There are certain common issues between the intellectual property 
question and other applications of wide area digital network 
technology. The question of signatures and writings to reflect the 
establishment of duties and permissions and the transfer of rights is 
common to the intellectual property inquiry and to electronic 
commerce using EDI techniques. There also are common questions 
involving rights to use certain information channels: First 
Amendment privileges, and tort liability. These are common not only 
to technological means of protecting intellectual property but to all 
forms of wide area networking. 


The law recognizes intellectual property because information 
technology permits one person to get a free ride on another person's 
investment in creating information value. Creative activity involving 
information usually is addressed by copyright, although patent has a 
role to play in protecting innovative means of processing information 
[2] The concept of intellectual property arose in the context of 
letterpress printing technology. Newer technologies like xerography 
and more recently small computer technology and associated word 
processing and networking have increased the potential for free 
rides and accordingly increased the pressure on intellectual 

The concern about free ride potential is especially great when 
people envision putting creative works on electronic publishing 
servers connected to wide area networks intending to permit 
consumers of information products to access these objects, 
frequently combining them and generally facilitating "publishing on 
demand" rather than the well known publishing just in case, typified 
by guessing how many copies of a work will sell, printing those in 
advance, and then putting them in inventory until someone wants 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 3 of 27 

The concern is that it will be too easy to copy an entire work without 
detection and without paying for it. Worse, it will be easy to copy an 
entire work and resell it either by itself or as a part of a new 
derivative work or collection. 

But technology is capable of protecting investment in new ways as 
well as offering the potential for a free ride. Computer networks 
make it possible to restrict access and to determine when access 
occurs. Depending on how new networks are designed, they may 
actually reduce the potential for a free ride. The digital library is one 
way of realizing that potential. Professor Pamela Samuelson has 
observed that the digital library model replaces intellectual property 
with a system of technological controls.[3] 


Basic Concepts 

A digital library is a set of information resources ("information 
objects") distributed throughout an electronic network. The objects 
reside on servers (computers with associated disk drives connected 
to the network). They can be retrieved remotely by users using 
"client" workstations. 

Origin of Concepts 

The phrase "digital library" and the basic concept were first 
articulated in a 1989 report growing out of a workshop sponsored by 
the Corporation for National Research lnitiatives.[4] From its 
inception, the digital library concept envisioned retrieval of complete 
information resources and not merely bibliographic information. [5] 

The technologies for remote retrieval of complete information object: 
using electronic technologies are in wide use through the 
WESTLAW, Dialog, LEXIS, NEXIS, and National Library of Medicine 
databases. These remotely accessible databases, however, unlike 
the digital library, involve a single host on which most of the data 
resides. The digital library concept envisions a multiplicity of hosts 

Recent Developments 

The remotely accessible database host concept is converging with 
the digital library concept as more of the electronic database 
vendors provide gateways to information objects actually residing or 
other computers. This now is commonplace with WESTLAW access 
to Dialog, and Dialog's gateways to other information providers. 

The most explicit implementation of the digital library concept is the 


EP Workshop - Perritt: Permission Headers and Contract Law 

Page 4 of 27 

Wide Area Information Service (WAIS), which implements ANSI 
standard Z.39.50.[6J WAIS permits a remote user to formulate a 
query that is applied to a multiplicity of WAIS servers, each of which 
may contain information responsive to the query. The WAIS 
architecture permits search engines of varying degrees of 
sophistication, resident on WAIS information servers, to apply the 
query against their own information objects, reporting matches back 
to the user.[71 Future implementations of WAIS will permit automatic 
refinement of searches according to statistical matching techniques. 

The Corporation for National Research Initiatives (CNRI) has 
proposed a test bed for an electronic copyright management system 
[8] The proposed system would include four major elements: 
automated copyright recording and registration, automated on-line 
clearance of rights, private electronic mail, and digital signatures to 
provide security. It would include three subsystems: a registration 
and recording system (RRS), a digital library system (DLS), and a 
rights management system (RMS). The RRS would provide the 
functions enumerated above and would be operated by the Library 
of Congress. It would provide "change of title" information.[9] The 
RMS would be an interactive distributed system capable of granting 
rights on line and permitting the use of copyrighted material in the 
digital library system. The test bed architecture would involve 
computers connected to the Internet performing the RRS and RMS 

Digital signatures would link an electronic bibliographic record (EBR 
with the contents of the work, ensuring against alteration after 
deposit.[101 Multiple RMS servers would be attached to the Internet. 
A user wishing to obtain rights to an electronically published work 
would interact electronically with the appropriate RMS. When 
copyright ownership is transferred, a message could be sent from 
the RMS to the RRS,[11] creating an electronic marketplace for 
copyrighted material. 

The EBR submitted with a new work would "identify the rights holdei 
and any terms and conditions oh the use of the document or a 
pointer to a designated contact for rights and permissions."[12] The 
EBR, thus, is apparently equivalent to the permissions header 
discussed in this paper. Security in the transfer of rights would be 
provided by digital signatures using public key encryption, discussec 
further, infra in the section on encryption. 

Basic Architectural Concepts 

The digital library concept in general contemplates three basic 
architectural elements: a query, also called a "knowbot" in some 
descriptions; a permissions header attached to each information 
object; and a procedure for matching the query with the permissions 

http ://www . cni . org/docs/ima. ip-workshop/Perri tt .html 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 5 of 27 


Two kinds of information are involved in all three architectural 
elements: information about the content of information objects 
desired and existing, and information about the economic terms on 
which an information object is made available. For example, a query 
desiring court opinions involving the enforcement of foreign 
judgments, evidencing a desire to download the full text of such 
judicial opinions and to pay up to $1 .00 per minute of search and 
downloading time, would require that the knowbot appropriately 
represent the subject matter "enforcement of foreign judgments." It 
also requires that the knowbot appropriately represent the terms on 
which the user is willing to deal: downloading and the maximum 
price. The permissions header similarly must express the same two 
kinds of information. If the information object to which the 
permissions header is attached is a short story rather than a judicial 
opinion, the permissions header must so indicate. Or, if the 
information object is a judicial opinion and it is about enforcement of 
foreign judgments, the permission header may indicate that only a 
summary is available for downloading at a price of $10.00 per 
minute. The searching, matching, and retrieval procedure in the 
digital library system must be capable of determining whether there 
is a match on both subject matter and economic terms, also copying 
and transmitting the information object if there is a match. 

Comparison to EDI 

Electronic Data Interchange (EDI) is a practice involving computer- 
to-computer commercial dealing without human intervention. In the 
most widespread implementations, computers are programmed to 
issue purchase orders to trading partners, and the receiving 
computer is programmed to evaluate the terms of the purchase 
order and to take appropriate action, either accepting it and causing 
goods to be manufactured or shipped, or rejecting it and sending an 
appropriate message. EDI is in wide use in American and foreign 
commerce, using industry-specific standards for discrete commercia 
documents like purchase orders, invoices, and payment orders, 
developed through the American National Standards Institute. 

There obviously are similarities between the three architectural 
elements of the digital library concept and EDI. There is a structured 
way of expressing an offer or instruction, and a process for 
determining whether there is a match between what the recipient is 
willing to do and what the sender requests. 

There is also, however, an important difference. In the digital library 
concept, a match results in actual delivery of the desired goods and 
services in electronic form. In EDI practice, the performance of the 
contractual arrangement usually involves physical goods or 

http : //www . cni . org/docs/ima. ip - workshop/Perritt .html 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 6 of 27 

performance of nonelectronic services. 

Nevertheless, the digital library and EDI architectures are sufficiently 
similar and, it turns out, the legal issues associated with both are 
sufficiently similar to make analogies appropriate. 

Elements of Data Structure 

For purposes of this paper, the interesting parts of the data structure 
are those elements that pertain to permission, more than those 
elements that pertain to content of the information object to which 
the header is attached. Accordingly, this section will focus only on 
permissions-related elements, after noting in passing that the 
content part of the header well might be a pointer to an inverted file 
to permit full text searching and matching. 

The starting point conceptually for identifying the elements of the 
permissions header are the rights exclusively reserved to the 
copyright owner by [[section]] 106 of the copyright statute. But these 
exclusive rights need not be tracked directly because the owner of 
an information object is free to impose contractual restrictions as 
well as to enjoy rights granted by the Copyright Act. Accordingly, it 
seems that the following kinds of privileges in the requester should 
be addressed in the permissions header: 

• outright transfer of all rights 

• use privilege, either unrestricted or subject to restrictions 

• copying, either unlimited or subject to restrictions like 
quantitative limits 

• distribution, either unlimited or subject to restrictions, like 
geographic ones or limits on the markets to which distribution 
can occur 

• preparation of derivative works. 

Display and presentation rights, separately identified in [[section]] 
106, would be subsumed into the use element, because they are 
particular uses. 

The simplest implementation would allow only binary values for eacf 
of these elements. But a binary approach does not permit the 
permissions header to express restrictions, like those suggested in 
the enumerated list. Elements could be defined to accept the most 
common kinds of restrictions on use, and quantitative limits on 
copying, but it would be much more difficult to define in advance the 
kinds of geographic or market-definition restrictions that an owner 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 7 of 27 

might wish to impose with respect to distribution. 

In addition to these discrete privileges, the permissions header musl 
express pricing information. The most sensible way of doing this is t< 
have a price associated with each type of privilege. In the event that 
different levels of use, copying, or distribution privilege are identified 
the data structure should allow a price to be associated with each 

A complicating factor in defining elements for price is the likelihood 
that different suppliers would want to price differently. For example, 
some would prefer to impose a flat fee for the grant of a particular 
privilege. Others might wish to impose a volume-based fee, and still 
others might wish to impose a usage or connect-time based fee. Th« 
data structure for pricing terms must be flexible enough to 
accommodate at least these three different approaches to pricing. 

Finally, the data structure must allow for a specification of 
acceptable payment terms and have some kind of trigger for a 
payment approval procedure. For example, the permissions header 
might require presentation of a credit card number and then trigger i 
process that would communicate with the appropriate credit card 
database to obtain authorization. Only if the authorization was 
obtained would the knowbot and the permissions header "match." 

There is a relationship between the data structures and legal 
concepts. The knowbot is a solicitation of offers. The permissions 
header is an offer. The matching of the two constitutes an 
acceptance. The "envelope" discussed elsewhere in these 
proceedings could be the "contract." 

There are certain aspects of the data structure design that are not 
obvious. One is how to link price with specific levels of permission. 
Another is how to describe particular levels of permission. This 
representation problem may benefit from the use of some deontic 
logic, possibly in the form of a grammar developed for intellectual 
property permissions. Finally, it is not clear what the acceptance 
should look like. Conceptually, the acceptance occurs when the 
knowbot matches with a permissions header, but it is unclear how 
this legally significant event should be represented. 


The CNRI test bed proposal envisions the use of public key 
encryption to ensure the integrity of digital signatures and to ensure 
the authenticity of information objects. Public key encryption permits 
a person to encrypt a message - like a signature - using a secret 
key, one known only to the sender, while permitting anyone with 
access to a public key to decrypt it. Use of public key cryptography 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 8 of 27 

in this fashion permits any user to authenticate a message, ensuring 
that it came from the purported sender.[13] A related technology 
called "hashing" permits an encrypted digital signature to be linked 
to the content of a message. The message can be sent in plain text 
(unencrypted) form, but if any part of it is changed, it will not match 
the digital signature. The digital signature and hashing technologies 
thus permit not only the origin but also the content integrity of a 
message of arbitrary length to be authenticated without necessitatis 
encryption of the content of the message. This technology has the 
advantage, among others, that it is usable by someone lacking 
technological access to public key encryption. An unsophisticated 
user not wishing to incur the costs of signature verification 
nevertheless can use the content of the signed information object. 

It is well recognized that encryption provides higher levels of security 
than other approaches. But security through encryption comes at a 
price. Private key encryption systems require preestablished 
relationships and exchange of private keys in advance of any 
encrypted communication. The burdens of this approach have led 
most proponents of electronic commerce to explore public key 
encryption instead. But public key systems require the establishmen 
and policing of a new set of institutions. An important infrastructure 
requirement for practicable public key cryptography is the 
.establishment and maintenance of certifying entities that maintain 
the public keys and ensure that they are genuine ones rather than 
bogus ones inserted by forgers. A rough analogy can be drawn 
between the public key certifying entities and notaries public. Both 
kinds of institutions verify the authenticity of signatures. Both kinds 
require some level of licensing by governmental entities. Otherwise 
the word of the "electronic notary" (certifying entity) is no better than 
an uncertified, unencrypted signature. In a political and legal 
environment in which the limitations of regulatory programs have 
been recognized and have led to deregulation of major industries, it 
is not clear that a major new regulatory arrangement for public key 
encryption is practicable. Nevertheless, experimentation with the 
concept in support of digital library demonstration programs can helf 
generate more empirical data as to the cost and benefits of public 
key encryption to reinforce electronic signatures. 

On the other hand, it is not desirable to pursue approaches requiring 
encryption of content. No need to encrypt the contents is apparent ir 
a network environment. Database access controls are sufficient to 
prevent access to the content if the permissions header terms are 
not matched by the knowbot. On the other hand, if the electronic 
publishing is effected through CD-ROMs or other physical media 
possessed by a user, then encryption might be appropriate to 
prevent the user from avoiding the permissions header and going 
directly to the content. 

Encrypted content affords greater security to the owner of 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 9 of 27 

copyrighted material, because someone who has not paid the price 
to the copyright owner must incur a much higher cost to steal the 
material. But the problem is everyone must pay a higher price to use 
the material. One of the dramatic lessons of the desktop computer 
revolution was the clear rejection of copyright protection in personal 
computer software. The reasons that copy protection did not survive 
in the marketplace militate against embracing encryption for content 
Encryption interferes with the realization of electronic markets, 
because producer and consumer must have the same encryption 
and description protocols. Encryption burdens the processing of 
electronic information objects because it adds another layer. Some 
specific implementations of encryption require additional hardware a 
appreciable costs. 

Digital libraries cannot become a reality until consumers perceive 
that the benefits of electronic formats outweigh the costs, compared 
to paper formats. Encryption interferes with electronic formats' 
traditional advantages of density, reusability, editability, and 
computer search ability; also, by impairing open architectures, they 
may perpetuate some of paper's advantages with respect to 

The need for encryption of any kind depends upon whether security 
is available without it. That depends, in turn, on the kinds of free 
rides that may be obtainable and the legal status of various kinds of 
electronics transactions in the digital library system. 


Copyright: What legal effect is intended? 

The design of the permissions header and the values in the 
elements of the header must be unambiguous as to whether an 
outright transfer of a copyright interest is intended or whether only a 
license is intended. If an outright transfer[15] is intended, then the 
present copyright statute requires a writing signed by the owner of 
the rights conveyed. [16] Recordation of the transfer with the 
copyright office is not required, but provides advantages in enforcing 
transferee rights.[17] On the other hand, non-exclusive licenses 
need not be in writing nor registered. If the electronic transaction 
transfers the copyright in its entirety, then the rights of the transferor 
are extinguished, and the rights of the transferee are determined by 
the copyright statute. The only significant legal question is whether 
the conveyance was effective. 

On the other hand, when the copyright is not transferred outright but 
only certain permissions are granted or certain rights conveyed, the 
legal questions become more varied. Then, the rights of the 
transferor and the obligations of the transferee are matters of 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 10 of 27 

contract law. It is important to understand the degree to which the 
contract is enforceable and how it is to be interpreted in the event of 
subsequent disputes. The following sections consider briefly the first 
sale doctrine as a potential public policy obstacle to enforcing 
contractual restrictions different from those imposed by the copyrigh 
statute and then explore in greater depth whether electronic 
techniques satisfy the formalities traditionally required for making a 
contract, whether they adequately ensure against repudiation, and 
whether they provide sufficient information to permit predictable 
interpretation of contractual obligations and privileges. 

First Sale Doctrine 

The first sale doctrine may invalidate restrictions on use. It is 
impermissible for the holder of a patent to impose restrictions on the 
use of a patented product after the product has been sold. 
Restrictions may be imposed, however, on persons who merely 
license the product.[18] The rationale for this limit on the power of 
the owner of the intellectual property interest is that to allow 
limitations on use of the product would interfere with competition 
beyond what the Congress - and arguably the drafters of the 
Constitution - intended in setting up the patent system. 

The first sale doctrine applies to copyright owners.[19] Indeed, 
because of the First Amendment's protection of informational 
activity, the argument against restrictions after the first sale may be 
even stronger in the copyright arena than in the patent arena. 

The first sale doctrine is potentially important because it may 
invalidate restrictions imposed on the use of information beyond 
what is authorized by the Copyright Act and by common law on tradi 
secrets. Thus, there may be serious questions about the legal 
efficacy of use restrictions, although such restrictions are common ir 
remote database service agreements. The vendors could argue that 
the limitations pertain to the contractual terms for delivery of a 
service rather than use of information as such. The characterization 
avoids the overlap with copyright and thus may also avoid the 
conflict between federal policy and contract enforcement [20] 

Contract Formation Issues 

The law does not enforce every promise. Instead, it focuses its 
power only on promises surrounded with certain formalities to make 
it likely that the person making the promise (the "promisor") and the 
person receiving the promise (the "promisee") understood that their 
communication had legal consequences. A threshold question for 
the digital library system is whether the traditional formalities for 
making a contract are present when the contract is made through 
electronic means. The digital library system considered in this paper 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 11 of 27 

clearly contemplates that a contract is formed when the knowbot an< 
the permissions header achieve a match. In this respect, the digital 
library concept converges with EDI, where trading parties 
contemplate that a contract to perform services or deliver goods is 
formed when a match occurs either upon the receipt of a purchase 
order or upon the transmission of a purchase order 

It is not altogether clear, however, whether the match between 
values and computer data structures meets contract formation 
requirements, particularly those expressed in various statutes of 
frauds. Statutes of frauds require "writings" and "signatures" for 
certain kinds of contracts - basically those contemplating 
performance extending beyond a period of one year. [21] 

In many instances, the digital library contract will be fully performed 
almost instantaneously upon delivery of the information object after 
the knowbot and the permissions header match. In such a case, the 
statute of frauds is not a problem and its requirements need not be 
satisfied. In other cases, however, as when the intent of the owner o 
the information object is to grant a license to do things that will 
extend beyond one year, the statute of frauds writing and signature 
requirements must be met. 

Historical application of statutes of frauds by the courts clearly 
indicates that there is flexibility in the meaning of "writing" and 
"signature." A signature is any mark made with the intent that it be a 
signature. [22] Thus an illiterate person signs by making an "X," and 
the signature is legally effective. Another person may sign a 
document by using a signature stamp. Someone else may authorize 
an agent to sign his name or to use the signature stamp. In all three 
cases the signature is legally effective. There may of course be 
arguments about who made the X, or whether the person applying 
the signature stamp was the signer or his authorized agent, but 
these are evidentiary and agency questions, not arguments about 
hard and fast contract-law requirements. 

Under the generally accepted legal definition of a signature, there is 
no legal reason why the "mark" may not be made by a computer 
printer, or for that matter by the write head on a computer disk drive 
or the data bus in a computer random access memory. The 
authorization to the computer agent to make the mark may be given 
by entering a PIN (personal identification number) on a keyboard. Tc 
extend the logic, there is no conceptual reason to doubt the legal 
efficacy of authority to make a mark if the signer writes a computer 
program authorizing the application of a PIN upon the existence of 
certain conditions that can be tested by the program. The resulting 
authority is analogous to a signature pen that can be operated only 
with a mechanical key attached to somebody's key ring, coupled witl 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 12 of 27 

instructions to the possessor of the key. 

Which of these various methods should be selected for particular 
types of transactions must not depend on what the law requires, 
because the law permits any of these methods. Rather, it must 
depend on the underlying purposes of the legal requirement and 
which method best serves those purposes. 

The real issue is how to prove that a particular party made the mark. 
In other words, the contingency to be concerned about is 
repudiation, not absence of formalities. Repudiation should be dealt 
with through the usual evidentiary and fact finding processes rather 
than artificial distinctions between signed and unsigned documents. 

Authority is skimpier on how flexible the "writing" requirement is. The 
best approach is to borrow the fixation idea from the copyright 
statute and conclude that a writing is "embodiment in a copy . . . 
sufficiently permanent or stable to permit it to be perceived, 
reproduced, or otherwise communicated for a period of more than 
transitory duration."[23] 

The most important thing conceptually is to understand the purpose 
of the writing and signature requirements. They have two purposes: 
awareness or formality, and reliability of evidence. Signature 
requirements, like requirements for writings and for original 
documents, have an essentially evidentiary purpose. If there is a 
dispute later, they specify what kind of evidence is probative of 
certain disputed issues, like "who made this statement and for what 
purpose?" The legal requirements set a threshold of probativeness. 
Surely the values in a knowbot as well as the values in a 
permissions header constitute a "mark," and someone who 
knowingly sets up potential transactions in a digital library scheme 
can have the intent that the mark be a signature. 

When a contract is made through a signed writing, it is more likely 
that the parties to the contract understand what they are doing. The} 
are aware of the legal effect of their conduct because the writing in 
the signature involves a greater degree of formality than a simple 

The awareness/formality purpose can be served by computerized 
contracting systems. This is so not so much because the computers 
are "aware" of the affect of their "conduct." Rather, it is true because 
the computers are agents of human principals. The programming of 
the computer to accept certain contract terms is the granting of 
authority to the computer agent to enter into a contract. The fact thai 
a principal acts through an agent engaging in conduct at a later poin 
in time never has been thought to defeat contract formation in the 
traditional evolution of agency and contract law. Nor should it when 


EP Workshop - Perritt: Permission Headers and Contract Law 

Page 13 of 27 

the agent is a computer. 

Fulfillment of the evidentiary purpose depends on the reliability of thi 
information retained by the computer systems making up the digital 
library. Such systems must be designed to permit the proponent of 
contract formation to establish the following propositions if the other 
party to the purported contract attempts to repudiate it. 

1. It came from computer X. 

2. It accurately represents what is in computer X 
[24] now. [25] 

3. What is in computer X now is what was in computer 
X at the time of the transaction. 

4. What was in computer X at the time of the transaction 
is what was received from the telecommunications 
channel . [26] 

5. What was received from the telecommunications channel 
is what was (a) sent, (b) by computer Y. 

Two other questions relate to matters other than the authenticity of 
the message: 

6. Computer Y was the agent of B. 

7. The message content expresses the content of the 
contract (or more narrowly, the offer or the 
acceptance) . [27 ] 

Factual propositions 1-4 can be established by testimony as to how 
information is written to and from telecommunications channel 
processors, primary storage, and secondary storage. Factual 
proposition 5 requires testimony as to the accuracy of the 
telecommunications channel and characteristics of the message tha 
associate it with computer Y. Only the last proposition (number 5) 
relates to signatures, because signature requirements associate the 
message with its source.[28] The other propositions necessitate 
testimony as to how the basic message and database management 
system works. It is instructive to compare these propositions with the 
kinds of propositions that must be established under the business 
records exception to the hearsay rule when it is applied to computer 
information. tt.html 


EP Workshop - Perritt: Permission Headers and Contract Law 

Page 14 of 27 

Those propositions may be supported with non-technical evidence, 
presented by non-programmers. A witness can lay a foundation for 
admission of computer records simply by testifying that the records 
are generated automatically and routinely in the ordinary course of 
business. The more inflexible the routine, and the less human 
intervention in the details of the computer's management of the 
database, the better the evidence.[29] 

The ultimate question is trustworthiness, and if the computer 
methods are apparently reliable, the information should be admitted 
unless the opponent of admissibility can raise some reasonable 
factual question undercutting trustworthiness. [30] 


Assuming that the permissions header and knowbot constitute 
sufficient writings to permit a contract to be formed and that the 
signature requirement also is met, through digital signature 
technology or otherwise, there still are difficult contract interpretation 
questions. Contract interpretation questions arise not only after 
contractual relationships are formed, but also in connection with 
deciding whether there has been offer and acceptance, the 
prerequisites to contract formation. [31] Contract interpretation 
always seeks to draw inferences about what the parties intended. 
When contract interpretation issues arise at the contract formation 
stage, the questions are what the offeror intended the content of the 
offer to be and what the offeree intended the content of the 
purported acceptance to be. The proposed digital library system 
envisions extremely cryptic expressions of offer and acceptance - 
by means of codes. The codes have no intrinsic meaning. Rather, 
extrinsic reference must be made to some kind of table, standard, oi 
convention associating particular codes with the concepts they 
represent. Extrinsic evidence is available to resolve contract 
interpretation questions when the language of the contract itself is 
ambiguous, and perhaps at other times as well. [32] The codes in the 
permissions header and knowbots certainly are ambiguous and 
become unambiguous only when extrinsic evidence is considered. 
So there is no problem in getting a standard or cable into evidence. 
The problem is whether the parties meant to assent to this standard. 

In current EDI practice, this question is resolved by having parties 
who expect to have EDI transactions with each other sign a paper 
trading partner agreement, in which the meaning of values or codes 
in the transaction sets is established. [33] But requiring each pair of 
suppliers and users of information in a digital library to have written 
contracts with each other in advance would defeat much of the utility 
of the digital library. Thus the challenge is to establish some ground 
rules for the meaning of permissions header and knowbot values to 
which all participants are bound. There are analogous situations. 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 15 of 27 

One is a standard credit card agreement that establishes contractua 
terms among credit card issuer, credit card subscriber, and 
merchant who accepts the credit card. The intermediary -- the credit 
card company - unilaterally establishes contract terms to which the 
trading partners assent by using and accepting the credit card.[34] 
Also, it is widely recognized that members of a private association 
can, through their constitution and bylaws, establish contractual 
relationships that bind all of the members in dealing with each other. 
[35] In the digital library system, similar legal arrangements can 
establish the standards by which electronic transactions between 
permissions header and knowbots will bind the transferor and the 
transferee of information. 

Third Party Liability 

It is not enough merely to ensure that the licensee is contractually 
bound. Trading partners also must ensure that the participants in 
funds transfers have enforceable obligations. For example, if the 
digital library system envisions that the information object would not 
be released to the purchaser without simultaneous release of a 
payment order, the supplier may be interested in enforcing the 
obligations of financial intermediaries who handle the payment ordei 
This implicates the federal Electronic Funds Transfer Act, and Artick 
4A of the Uniform Commercial Code, regulating wire transfers. 


Satisfy the Business Records Exception to the Hearsay Rule 

The discussion of contract formalities earlier in this paper concluded 
that legally enforceable contracts can be formed through electronic 
means and that the significant legal questions relate to reliability of 
proof and intent of the parties to be bound by using the electronic 
techniques. This section considers the reliability of proof further. 
Traditional evidence law permits computer records to be introduced 
in evidence when they satisfy the requirements of the business 
records exception: basically that they are made in the ordinary 
course of business, that they are relied on for the performance of 
regular business activities, and that there is no independent reason 
for questioning their reliability.[36] 

The business records exception shares with the authentication 
concept, the statute of frauds, and the parol- evidence rule a commoi 
concern with reliability.[37] The same procedural guarantees and 
established practices that ensure reliability for hearsay purposes 
also ensure reliability for the other purposes. Under the business 
records exception, the proponent must identify the source of a 
record, through testimony by one familiar with a signature on the 
record, or circumstantially.[38] The steps in qualifying a business 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 16 of 27 

record under the common law, which since have been relaxed, [39] 

• proving that the record is an original entry made in the routine 
course of business 

• proving that the entries were made upon the personal 
knowledge of the proponent/witness or someone reporting to 

• proving that the entries were made at or near the time of the 

• proving that the recorder and his informant are unavailable.[40 

These specific requirements are easier to understand and to adapt 
to electronic permissions and obligations formed in a digital library 
system by understanding the rationale for the business records 
exception. The hearsay rule excludes out-of-court statements 
because they are inherently unreliable, primarily because the maker 
of the statement's demeanor cannot be observed by the jury and 
because the maker of the statement is not subject to cross 
examination. On the other hand, there are some out-of-court 
statements that have other guarantees of reliability. Business 
records are one example. If a continuing enterprise finds the records 
sufficiently reliable to use them in the ordinary course of business, 
they should be reliable enough for a court. The criteria for the 
business records exception all aim at ensuring that the records realh 
are relied upon by the business to conduct its ordinary affairs. 

The Manual for Multidistrict Litigation suggests steps for qualifying 
computer information under the business records exception: 

1 . The document is a business record. 

2. The document has probative value. 

3. The computer equipment used is reliable. 

4. Reliable data processing techniques were used. [41] 

Key to adapting the business records exception to electronic 
permissions in a digital library system are points 3 and 4. 
Establishing these propositions and the evidentiary propositions set 
forth elsewhere in this paper requires expert testimony. Any designe 
of a digital library system must consult with counsel and understand 
what testimony an expert would give to establish these propositions. 
Going through that exercise will influence system design. 


Page 17 of 27 

Reinforce the Evidentiary Reliability by Using 
Trusted Third Parties 

The evidentiary purpose of contract formation requirements can be 
satisfied by using a trusted third party as an intermediary, when the 
third party maintains archival records of the transactions. The third 
party lacks any incentive for tampering with the records and when 
the third party's archiving system is properly designed, it can provide 
evidence sufficient to establish all of the propositions. 

This third party intermediary concept is somewhat different from the 
concept of a certifying agent in digital signature systems. To be sure 
the custodian of transaction records envisioned by this section could 
be the same as the certifying entity for public and key encryption, bu 
the custodian role can be played in the absence of any encryption. 
Indeed, the digital library itself is a good candidate for the custodian 
role. The library has no incentive to manipulate its records in favor o 
either the producers of information value or the consumers. In order 
to carry out its affairs, it must use these transactional records in the 
ordinary course of business, thereby making it likely that digital 
library records would qualify under the business records exception. 


Obviously, the digital library concept depends upon the possibility of 
an automated comparison between the knowbot and the 
permissions header. This means that potential requesters of 
information and suppliers of information must know in advance the 
data structures for representing the elements of the permissions 
header and the knowbot. This requires compatibility. Compatibility 
requires standardization. Standardization does not, however, 
necessarily require "Standards" in the sense that they are developec 
by some bureaucratic body like ANSI. It may simply imply market 
acceptance of a particular vendor's approach. Indeed, each digital 
library might use different data structures. All that is necessary is 
that the structure of the knowbot and the structure of the permission: 
header be compatible within any one digital library system. Also, as 
demands emerge for separate digital libraries to communicate with 
each other, there can be proprietary translation to assure 
compatibility between systems much as common word processing 
programs translate to and from other common formats and much as 
printers and word processing software communicate with each othei 
through appropriate printer drivers. In neither of these cases has am 
independent standards organization developed a standard that is at 
all relevant in the marketplace. 

Standardizing the elements of knowbot and permissions headers 
involves content standardization, which generally is more 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 18 of 27 

challenging than format standardization.[42] A permissions 
header/knowbot standard is a system for representing legal 
concepts and for defining legal relations. As such, the standard is 
basically a grammar for a rule-based substantive system in a very 
narrow domain.[43] The data elements must correspond to legally 
meaningful relational attributes. The allowable values must 
correspond to legally allowable rights, obligations, privileges and 
powers. In other words, the standard setter must meet many of the 
challenges that a legal expert system designer working with 
Hohfeldian frameworks must meet.[44] This adds a constraint to the 
standards setting process. Unlike setting format standards, where 
the participants are free to agree on an arbitrary way of expressing 
format attributes, participants in setting a content standard must 
remain within the universe of permissible content. The set of 
permissible values is determined by the law rather than being 
determined only by the imagination of format creators. 


One of the many profound observations by Ithiel de Sola Pool[45] 
was that copyright always has depended upon technological 
bottlenecks for its enforceability. The printing press was the original 
enforcement bottleneck. Now, a combination of the printing press 
and the practical need to inventory physical artifacts representing th< 
work constitute the enforcement bottlenecks. As technologies 
change, old bottlenecks disappear and enforceability requires a 
search for new bottlenecks. When there are single hosts, like 
WESTLAW, Dialog, LEXIS, and CompuServe, access to that host is 
the bottleneck. The problem with distributed publishing on an open 
architecture internet is that there is no bottleneck in the middle of the 
distribution chain corresponding to the printer, the warehouse or the 
single host. 

If new bottlenecks are to be found, they almost surely will be found 
at the origin and at the point of consumption. Encryption and 
decryption techniques discussed elsewhere in this volume 
concentrate on those bottlenecks as points of control. It also is 
possible that rendering software could become the new bottleneck. 

Even with those approaches, however, a serious problem remains ir 
that the new technologies make it difficult or impossible to 
distinguish between mere use and copying. Thus the seller cannot 
distinguish between an end user[46] and a potential competitor. On 
the other hand, the new technologies permit a much better audit trai 
potentially producing better evidence for enforcement adjudication. 

If network architectures for electronic publishing evolve in the way 
that Ted Nelson suggests with his Xanadu concept[47], the real 
value will be in the network and the pointers, not in the raw content. 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 19 of 27 

Thus, the creative and productive effort that the law should reward is 
the creation and production and delivery of pointers, presentation, 
distribution, and duplication value. If this is so, then technological 
means will be particularly important, foreclosing access by those 
lacking passwords and other keys and limiting through contract wha 
a consumer may do with the information. 

In such an architecture, the law either will be relatively unimportant 
because technology can be counted on to prevent free riding, or the 
law will need to focus not on prohibiting copying or use without 
permission, but on preventing circumvention of the technological 
protections. Thus, legal approaches like that used to prevent the 
sale of decryption devices for television broadcasts and legal issues 
associated with contract enforcement may be more important than 
traditional intellectual property categories. 


The law generally imposes sensible levels of transaction costs. 
Usually, transaction costs are proportional to the risk. Figure 1 
shows a continuum of risk and transaction cost in traditional and nev 
technologies. A real estate closing involves significant risks if there i; 
some dispute later about the transaction. Therefore, the law affords 
much protection, including a constitutional officer called a registrar o 
deeds who is the custodian of records associated with the 
transaction. The risk level analogous to this in electronic publishing 
might be access to an entire library including access software as 
well as contents. Next on the continuum is a transaction involving a 
will or power of attorney. There, the risk is substantial because the 
maker of the instrument is not around to help interpret it. The law 
requires relatively high levels of assurance here, though not as grea 
as those for real estate transactions. The law requires witnesses an< 
attestation by a commissioned minor official called a notary public. 
The electronic publishing analogy of this level of risk might be the 
contents of an entire CD-ROM. 

Next in level of risk is the purchase of a large consumer durable like 
an automobile. The law requires somewhat less, but still significant, 
protections for this kind of transaction: providing for the filing and 
enforcement of financing statements under the Uniform Commercial 
Code. The electronic publishing analogy might be the transfer of 
copyright to a complete work. Next along the risk continuum is the 
purchase of a smaller consumer durable like a television set. Here, 
the law typically is reflected in written agreements of sale, but no 
special third party custodial mechanisms. The electronic publishing 
analogy might be use permission for a complete work. 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 20 of 27 

At the other end of the continuum is the purchase of a relatively 
small consumer item, say a box of diskettes. Neither the law nor 
commercial practice involves much more than the exchange of the 
product for payment, with no written agreement or anything else to 
perform channeling, cautionary, evidentiary, or protective functions. 
The electronic publishing analogy might be use permission for part 
of a work. 

Realizing the potential of electronic publishing in distributed 
information networks requires sensitivity to the transaction costs of 
too much security. Requiring $10 boxes of diskettes to be sold like 
real property would impose unacceptable transaction costs. 
Similarly, an encrypted object combined with rendering software is 
probably inconsistent with an open architecture. Because of the 
difficulty of setting standards for such technologies, this approach to 
intellectual property protection probably would be effectuated by 
proprietary approaches, thus frustrating the vision of an open marke 
for electronic publishing. 


Realization of the digital library vision requires a method for 
collecting money and granting permission to use works protected by 
intellectual property. The concept of a knowbot and a permissions 
header attached to the work is the right way to think about such a 
billing and collection system. Standards for the data structures 
involved must be agreed to, and systems must be designed to 
satisfy legal formalities aimed at ensuring awareness of the legal 
significance of transactions and reliable proof of the terms of the 

In the long run, not only must these technological issues be 
resolved, with appropriate attention to levels of risk and protections 
available under traditional legal doctrines, but also further concepts 
development must be undertaken. Proponents of electronic 
publishing over wide area networks need to think about the 
appropriate metaphors: whether it is a library or a bookstore, if a 
library whether with or without Xerox machines, if a bookstore 
whether it is a retail bookstore or a mail order operation. Then, 
thought must be given to how standards will be set. Finally, and 
most important, much more needs to be understood about the need 
for third party institutions. There is a good deal of enthusiasm for 
public key encryption. Yet the vulnerability of public key encryption 
systems is in the integrity of the key authority. In traditional legal 
protections, the third party custodians or authenticating agents like 
notaries public and registrars of deeds receive state sanction and 
approval, and in the case of registrars of deeds, public funding. We 
must be clearer as to whether a similar infrastructure must be 
developed to protect against substantial risks and the use of EDI an< 


EP Workshop - Perritt: Permission Headers and Contract Law 

Page 21 of 27 

electronic publishing technologies. 

Finally, and perhaps most importantly, we must be thoughtful about 
what legal obligations, imposed on whom, are appropriate. The 
suggested paragraphs 102(e) and (f) in the High Performance 
Computing Act look very much like King James I's licensing of 
printing presses. They also look like the FBI's proposal to prohibit th» 
introduction of new technologies until certain conformity with past 
legal concepts is assured. Such approaches make the law a hurdle 
to new technology - an uncomfortable position for both law and 


1 . The use of EDI techniques to meter usage and determine charge; 
for use of intellectual property is an example of billing and collection 
value in a typology of different types of value that can be produced ii 
electronic marketplaces for information. See Henry H. Perritt, Jr., 
Market Structures for Electronic Publishing and Electronic 
Contracting in Brian Kahin, ed., BUILDING INFORMATION 
University and McGraw-Hill 1992) (developing typology for different 
types of value and explaining how market structures differ for the 
different types); Henry H. Perritt, Jr., Tort Liability, the First 
Amendment, and Equal Access to Electronic Networks, 5 
Harv.J.Law & Tech. 65 (1992) (using typology often types of value 
to analyze access by competing producers of value). 

2. See, e.g., U.S. Pat. No. 5,016,009, Data compression apparatus 
and method (May 14, 1991); U.S. Pat. No. 4,996,690, Write operato 
with gating capability (Feb. 26, 1991); U.S. Pat. No. 4,701,745, Data 
compression system (Oct. 20, 1987); Multi Tech Systems, Inc. v. 
Hayes Microcomputer Products, Inc., 800 F. Supp. 825 (D. Minn. 
1992) (denying summary judgment on claim that patent for modem 
escape sequence is invalid). 

3. Comments on the 8/21 draft of "Knowbots in the Real World" fronr 
the intellectual property workshop participants, page 6 (author 
unknown, source unknown). Professor Samuelson also observed 
that the workshop, despite its title, actually did not focus much on 
intellectual property issues. See fn. 4, below. 

4. Corporation for National Research Initiatives, Workshop on the 
Protection of Intellectual Property Rights in a Digital Library System: 
Knowbots in the Real World-May 18-19, 1989 (describing digital 
library system). 

5. See generally Clifford A. Lynch, Visions of Electronic Libraries 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 22 of 27 

(libraries of future can follow acquisition-on-demand model rather 
than acquiring in advance of use; Z39.50 protocol will facilitate 
realization of that possibility, citing Robert E. Kahn & Vinton G. Serf, 
An Open Architecture for a Digital Library System and a Plan for Its 
Development The Digital Library Project, volume 1: The World of 
Knowbots (draft) (Washington D.C.: Corporation for National 
Research Initiatives; 1988)). 

6. Clifford A. Lynch, The Z39.50 Information Retrieval Protocol: An 
Overview and Status Report, ACM Sigcomm Computer 
Communication Review at 58 (describing Z39.50 as an OSI 
application layer protocol that relieves clients from having to know 
the structure of data objects to be queried, and specifies a 
framework for transmitting and managing queries and results and 
syntax for formulating queries). 

7. Brewster Kahle, Wide Area Information Server Concepts (Nov. 3, 
1989 working copy; updates available from Brewster 
(describing WAIS as "open protocol for connecting user interfaces 
on workstations and server computers") (describing information 
servers as including bulletin board services, shared databases, text 
searching and automatic indexing and computers containing current 
newspapers and periodicals, movie and television schedules with 
reviews, bulletin boards and chat lines, library catalogues, Usenet 

8. Robert E. Kahn, Deposit, Registration, Recordation in an 
Electronic Copyright Management System (August 1 992) 
(Corporation for National Research Initiatives, Reston, Virginia). 
CNRI claims a trademark in "knowbot" and "digital library". 

9. Kahn 1992 at 4. 

10. Kahn 1992 at 6. 

11. Kahn 1992 at 10. 

12. Kahn 1992 at 12. 

13. Kahn 1992 at 15. 

14. Browsability through techniques like the collapsible outliner 
function in Microsoft Word for Windows and competing products 
requires more chunking and tagging value in the form of style and 
text element codes. Handling this additional formatting information 
through encryption and description processes is problematic. 

15. "A 'transfer of copyright ownership 1 is an assignment, mortgage, 
exclusive license, or any other conveyance, alienation, or 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 23 of 27 

hypothecation of a copyright or of any of the exclusive rights 
comprised in a copyright, whether or not it is limited in time or place 
of effect, but not including a non-exclusive license " 17 U.S.C. 
[[section]] 101 (1988). 

16. 17 U.S.C. [[section]] 204(a) (1988); Valente-Kritzer Video v. 
Pinckney, 881 F.2d 772, 774 (9th Cir. 1989) (affirming summary 
judgment for author; oral agreement unenforceable under Copyright 
Act); Library Publications, Inc. v. Medical Economics Co., 548 F. 
Supp. 1231, 1233 (E.D. Pa. 1982) (granting summary judgment 
against trade book publisher who sought enforcement of oral 
exclusive distribution agreement; transfer of exclusive rights, no 
matter how narrow, must be in writing), affd mem., 714 F.2d 123 (3c 
Cir. 1983). 

17. 17 U.S.C. [[section]] 205 (1988) provides constructive notice of 
the contents of the recorded document, determining priority as 
between conflicting transfers, and determines priority as between 
recorded transfer and non-exclusive license. The former requiremen 
for transfers to be recorded in order for the transferee to maintain ar 
infringement, 17 U.S.C. [[section]] 205(d), was repealed by the 
Berne Act Amendments [[section]] 5. 

. 1 8. Under Adams v. Burke, 84 U.S. (1 7 Wall.) 453 (1 873), a 
patentee must not attempt to exert control past the first sale. In 
general, use restrictions may be placed only on licensees, consisten 
with General Talking Pictures v. Western Elec., 304 U.S. 175 (1938) 
See generally Baldwin-Lima-Hamilton Corp. v. Tatnall, 169 F. Supp. 
1 (E.D. Pa. 1958) (applying no control after purchase rule). 

19. See Red-Baron-Franklin Park, Inc. v. Taito Corp., 883 F.2d 275, 
278 (4th Cir. 1989) (purchase of video game circuit boards did not 
create privilege to perform video game under first sale doctrine); 
United States v. Moore, 604 F.2d 1228, 1232 (9th Cir. 1979) (piratec 
sound recording not within first sale doctrine in criminal copyright 
infringement prosecution). But see Mirage Editions, Inc. v. 
Albuquerque A.RT. Co., 856 F.2d 1341, 1344 (9th Cir. 1988) (first 
sale doctrine did not create privilege to prepare derivative work by 
transferring art in book to ceramic tiles). 

20. The way in which the first sale doctrine would impact the 
electronically imposed use restrictions is by frustrating a breach-of- 
contract lawsuit by the licensor against a licensee who exceeds the 
use restrictions. The licensee exceeding the use restrictions would 
argue that it violates public policy to enforce the restrictions and 
therefore that state contract law may not impose liability for their 
violation. See generally Restatement (second) of Contracts 
[[section]] 178 (1981) (stating general rule for determining when 
contract term is unenforceable on grounds of public policy). 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 24 of 27 

21 . In addition, the Copyright Act itself requires signed writings for 
transfers of copyright interests. 17 U.S.C. [[section]] 204(a). (1988). 

22. Michael S. Baum & Henry H. Perritt, Jr., ELECTRONIC 
(contract, evidence and agency issues) [hereinafter "Baum & 
Perritt"]. Accord, Signature Requirements Under EDGAR, 
Memorandum from D. Goelzer, Office of the General Counsel, SEC 
to Kenneth A. Fogash, Deputy Executive Director, SEC (Jan. 13, 
1986) (statutory and non-statutory requirements for "signatures" ma; 
be satisfied by means other than manual writing on paper in the 
hand of the signatory ..." In fact, the electronic transmission of an 
individual's name may legally serve as that person's signature, 
providing it is transmitted with the present intention to 

23. 17 U.S.C. [[section]] 101 (1988). For copyright purposes, a work 
is created, and therefore capable of protection, when it is fixed for 
the first time. 17 U.S.C. [[section]] 101 (1988). n [l]t makes no 
difference what the form, manner, or medium of fixation may be~ 
whether it is in words, numbers, notes, sounds, pictures, or any 
other graphic or symbolic indicia, whether embodied in a physical 
object in written, printed, photographic, sculptural, punched, 
magnetic, or any other stable form, and whether it is capable of 
perception directly or by means of any machine or device 'now 
known or later developed. 1 " 1976 U.S. Code Cong. & Admin. News 
5659, 5665. The legislative history further says that, "the definition o 
'fixation 1 would exclude from the concept [representations] purely of 
an evanescent or transitory nature-reproductions such as those 
projected briefly on a screen shown electronically on a television or 
other video display, or captured momentarily in the 'memory 1 of a 
computer." 17 U.S.C. [[section]] 102 note (excerpting from House 
Report 94-1476). 

24. Or, more likely, what is on the computer medium read by 
computer X, such as a magnetic cartridge used for archival records. 
Further references in the textual discussion to "what is in computer ) 
now" should be understood to include such computer-readable 

25. Cf .R. Peritz, Computer Data and Reliability: A Call for 
Authentication of Business Records Under the Federal Rules of 
Evidence, 80 Nw.U.LRev. 956, 980 (1986) (proof that a printout 
accurately reflects what is in the computer is too limited a basis for 
authentication of computer records). 

26. In some cases, the electronic transaction will be accomplished 
by means of a physical transfer of computer readable media. In suet 
a case, this step in the proof would involve proving what was 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 25 of 27 

received physically. 

27. See generally Peritz, 80 Nw.U.LRev. at 979 (citing as examples 
of authentication Ford Motor Credit Co. v. Swarens, 447 S.W.2d 53 
(Ky. 1969) (authentication by establishing relationship between 
computer-generated monthly summary of account activity and the 
customer reported on); Ed Guth Realty, Inc. v. Gingold, 34 N.Y.2d 
440, 315 N.E.2d 441, 358 N.Y.S.2d 367 (1974) (authentication of 
summary of taxpayer liability and the taxpayer)). 

28. Of course, a paper document signed at the end also is probative 
of the fact that no alterations have been made. In this sense, a 
signature requirement telescopes several steps in the inquiry 
outlined in the text. 

29. United States v. Linn, 880 F.2d 209, 216 (9th Cir. 1989) 
(computer printout showing time of hotel room telephone call 
admissible in narcotics prosecution). See also United States v. 
Miller, 771 F.2d 1219, 1237 (9th Cir. 1985) (computer-generated toll 
and billing records in price-fixing prosecution based on testimony by 
billing supervisor although he had no technical knowledge of system 
which operated from another office; no need for programmer to 
testify; sufficient because witness testified that he was familiar with 
the methods by which the computer system records information). 

30. See United States v. Hutson, 821 F.2d 1015, 1020 (5th Cir. 
1987) (remanding embezzlement conviction, although computer 
records were admissible under business records exception, despite 
trustworthiness challenged based on fact that defendant embezzled 
by altering computer files; access to files offered in evidence was 
restricted by special code). 

31. Restatement (Second) of Contracts [[section]] [[section]] 17, 24, 
35 (1981). 

32. John E. Murray, Jr., Murray on Contracts [[section]] 89, 3rd ed. 
(Charlottesville, VA: Michie, 1990). 

33. See Baum & Perritt [[section]] 2.6; The Electronic Messaging 
Services Task Force, The Commercial Use of Electronic Data 
Interchange-A Report and Model Trading Partner Agreement, 45 
Bus.Law. 1645 (1990); Jeffrey B. Ritter, Scope of the Uniform 
Commercial Code: Computer Contracting Cases and Electronic 
Commercial Practices, 45 Bus.Law. 2533 (1990); Note, Legal 
Responses to Commercial Transactions Employing Novel 
Communications Media, 90 Mich.LRev. 1 145 (1992). 

34. Garber v. Harris Trust & Savings Bank, 432 N.E.2d 1309, 1311- 
1312 (III. App. 1982) ("each use of the credit card constitutes a 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 26 of 27 

separate contract between the parties;" citing cases). It is not quite 
this simple, because both merchant and credit card customer have 
separate written contracts with the credit card issuer. But there is no 
reason that a supplier of information to a Digital Library System and 
all customers of that system might not have their own contracts with 
the Digital Library System in the same fashion. 

35. Rowland v. Union Hills Country Club, 757 P.2d 105 (Ariz. 1988) 
(reversing summary judgment for country club officers because of 
factual question whether club followed bylaws in expelling 
members); Straub v. American Bowling Congress, 353 N.W.2d 1 1 
(Neb. 1984) (rule of judicial deference to private associations, and 
compliance with association requirements, counseled affirmance of 
summary judgment against member of bowling league who 
complained his achievements were not recognized). But see Wells \ 
Mobile County Board of Realtors, Inc., 387 So.2d 140 (Ala. 1980) 
(claim of expulsion of realtor from private association was justifiable 
and bylaws, rules and regulations requiring arbitration were void as 
against public policy; reversing declaratory judgment for defendant 

36. F.R.E. 803(6) (excluding business records from inadmissibility a: 
hearsay); 28 U.S.C. [[section]] 1732 ("Business Records Act" 
permitting destruction of paper copies of government information 
reliably recorded by any means and allowing admission of remaining 
reliable record). 

37. See Peritz, 80 Nw.U.L.Rev at 978-80, 984-85 (noting body of 
commentator opinion saying that business records exception and 
authentication are parallel ways of establishing reliability). 

38. See F.R.E. 901(b)(4) (appearance, contents, substance, internal 
patterns, as examples of allowable authentication techniques). 

39. Peritz, 80 Nw.U.L.Rev. at 963-64 (identifying steps and trend 
resulting in F.R.E.). 

40. Peritz, 80 Nw.U.L.Rev. at 963. 

41 . Peritz, 80 Nw.U.L.Rev. at 974 (reporting four requirements of 
Manual, and endorsing their use generally). 

42. See Henry H. Perritt, Jr., Format and Content Standards for the 
Electronic Exchange of Legal Information, 33 Jurimetrics 265 (1993) 
(distinguishing between format and content standardization). 

43. Marc Lauritsen, Senior Research Associate at Harvard Law 
School, has written about the relationship between substantive legal 
systems and the field of artificial intelligence. 


IP Workshop - Perritt: Permission Headers and Contract Law 

Page 27 of 27 

44. Anne Gardner, An Artificial Intelligence to Legal Reasoning 
(Cambridge, MA: MIT Press, 1987); Kevin Ashley, Modeling Legal 
Argument (Cambridge, MA: MIT Press, 1990). 

45. Ithiel de Sola Pool, Technologies of Freedom (Cambridge, MA: 
Harvard University Press, 1983), p. 16-17, 249. 

46. It may not be particularly important to limit competition by 
consumers, because the consumers will never have the pointers an< 
the rest of the network infrastructure. 

47. This concept was discussed at the conference. 

Henry H. Perritt, Jr., Professor of Law at Villanova University School 
of Law, was Deputy Undersecretary of Labor in the Ford 
Administration, and worked on telecommunications on President 
Clinton's Transition Team. He holds a B.S. and S. M. from MIT. an< 
a J.D. from Georgetown, and has written ten books and 30 articles. 

Henry H. Perritt, Jr. 
Villanova University School of Law 
Villanova, PA 19085 
(215) 645-7078 
FAX (215) 896-1723 


© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


EP Workshop - Gerovac/Solomon: Protect Revenues, Not Bits 

Pagel of 15 

Sponsorec by: 


| Coalition for Networked Information 

About CNI 

Task Force 



Protect Revenues, Not Bits: Idei 
Your Intellectual Property 

by Branko Gerovac and Richard J. Solomon 


Site Map 

Search our she 


Work-in-progress by the television and motion picture community foi 
header is described. This is not a Utopian solution for all media and ; 
such a header would aid in tracking, auditing, and particularly identif 
bitstreams, an important parameter for intellectual property protectio 

"The New World Order is no longer about bullets, 

but about bits," — the villain in Sneakers. ((c)1992. 

Universal Pictures Inc.) 

The Sneakers villain missed the whole point of the digital revolution: 
money - lots of it -- not by stealing bits, or by altering them, or by ju* 
around. That's penny ante stuff. Big money is made by adding value 
Until we understand the concept of value-added, toiling in the minuti 
information protection can be fairly unproductive. 

The stored program computer has changed the basic concepts of wl 
property is about. Combined with high-speed digital telecommunicat 
performance computing permits an ease of storage, retrieval, manip 
transmission -- and replication completely at odds with our notions 
based on 400 years of the printing press. The idea of a machine tha 
itself to hide its tracks, copy ad infinitum and effortlessly without intrc 
bit error, and send its output around the world at the speed of light u 
the original is still something that the average person finds difficult tc 
(despite widespread and growing use of tens of millions of these de\ 

While other distributive mechanisms have made it more difficult to pi 
rights over the centuries, computer/communications poses a unique 
because of the increasing scale, speed, and power of digitally-basec 
technologies. But before we discuss the subtleties of computer techi 
potential tools for controlling the flow of digitized works, it is importar 
goals of information owners, distributors, vendors and packagers arc 
the implementations and applications. There is no total "solution," te 


IP Workshop - Gerovac/Solomon: Protect Revenues, Not Bits 

Page 2 of 15 

otherwise, to the intellectual property "problem." Yet, by narrowing o 
devise some useful strategies which would facilitate electronic distrit 
maintain sufficient income streams to compensate artists, investors, 

If we are concerned with money flows that is, with whether the int€ 
owner gets paid or not - then, as in all successful businesses, the to 
what generates revenue and what is trivial. However, if we are cona 
rights, privacy rights, cultural heritage, or accuracy in presentation, t 
may be different and the technical fixes may also be different. 

In this paper we describe work in progress by the television and mot 
community for a digital header. This is not a Utopian solution for all n 
applications, but such a header would aid in tracking, auditing, and f 
identifying imaging bitstreams. 

Not that the other goals are not important. It is recognized that digita 
flows not only change the concept of intellectual property, but chang 
global society works and interacts. For example, it is obvious that ar 
tracking the flow of intellectual property, or the reverse flows of com| 
creates audit trails of metain formation. This is an inherent offshoot o 
added bitstreams, and it has intrinsic value itself. That is, metainforn 
more valuable (to whomever controls its use) than the information b< 

Kenneth Phillips deals with the intersection of intellectual property pi 
metainformation in another paper at this workshop. 


The concept of copyright is rooted in the technology of print.[2] Until 
other distributive mechanisms were merely mild disturbances to the 
rights could be physically controlled by control of the press, for the o 
had similar features to the press in that the device, transmitter, recoi 
film duplicating machinery, etc., could be located in space, time, and 
Computer/communications changes all that. 

The press pre-dated the idea of a "copy right," and it is not irrelevanl 
the recognition that there could be a property right in text, and a prat 
roya/ties[3] emerged when the printing press reached a stage of dev 
it threatened the sovereign's hegemony. King Phillip and Queen Mai 
1557, in an effort to stop seditious and heretical ideas from being cir 
realm, limited the right of printing to members of the Stationers 1 Corr 
Company was given the right to search for and seize anything printe 
statute or proclamation with draconian measures prescribed for viols 
resistors. [4] By 1565, the Company created a system of copy rights 
members, thereby both privatizing the state function of censorship ($ 
efficient with a profit motive) and simultaneously creating a novel mc 
business practice. 


IP Workshop - Gerovac/Solomon: Protect Revenues, Not Bits 

Page 3 of 15 

Copyright attorneys tend to dismiss the historical background as a d 
misses the critical point of copyrights as an intellectual property gate 
determination of the locus of control using a technology which serve 
entry to the marketplace. Because numerous copies were made in c 
huge and relatively expensive presses - it was feasible to identify th 
number, and often the destination of printed materials by human ove 
printshop, at that time in history, was the practical point to apply con 
profit or against heresy, or both. Yet, even then "the increasing numl 
made it impossible to acknowledge every sermon, almanac, and bal 
modes of reproduction where such an easy locus did not exist, the c 
copyright was not applied under common law. Until quite recently, c< 
applied to conversation, speeches, jokes, or singing of songs, wheth 
public. (Perhaps we may consider, along with the technologies of m; 
that the associated technologies of selective audio and image captu 
extended the locus of control.) Copyright, until the modern age, rem; 
protection applied to (or with) a specific technology; though in demoi 
nations it was rarely used successfully for censorship, but instead to 
and other rights. 


The technology which permitted audit and control was not so much 1 
but the mechanism of pressing ink to paper media. We have seen th 
technology increasingly strain the use of copyright laws as a societa 
money back to the property owners, as well as effective protection a 
alteration, etc. First came cheaper presses, then photographic devic 
typesetting and the steam-driven press, audiographs, mimeographs : 
television, xerography, and then broadband appliances of all types. I 
loci, the economics, and the manipulation of media, making infringer 
police and even harder to define. 

With the computer and its appliances, if not now, soon we will be ab 
make a perfect duplicate of anything, from the Gutenberg Bible to th 
without consuming the original. Property rights used to deal with tan; 
physical incarnation of a bit may be in tangible form at some instant 
are volatile, elusive, and the process that manipulates bits inherently 
tracks as it copies the bits from one register to another, from one pa 
another, from one end of a network to another - that is just the way 
machine[6] works. In any open network of millions upon billions of pi 
logic devices, an audit path not only is a messy concept, but it is me 
for very tiny slices of time. There is no such thing as an "original" be< 
are originals, as well. 

The machine is at once a series of processes, concepts, and synthe 
(and maybe machine) intelligence - so mixed that it is difficult to sep 
from the whole. And since the Turing definition of this stored-prograr 
holds, the machine can change its own instructions, redefining itself, 
new machine in a twinkling. A network of such machines permitting i 
transfer of digitized information is not what the Stationers 1 architects 



IP Workshop - Gero vac/Solomon: Protect Revenues, Not Bits 

Page 4 of 15 

when they privatized heresy control. The idea of the nationwide - or 
"machine room," whereby for some small slice of time in the middle < 
unused processor, PC, and switch on the Gigabit network performs : 
virtual process, begs the questions of digital copy rights, much less 1 
locus for auditing. 

What are we to make of this confusion? Basically that we are upon i 
way to add value, not subtract, steal, or transfer value. 


The law of geometric increase is the critical publishing idiosyncrasy 
environment. If you electronically mail a copyright article to two corrc 
then they each send a copy to two others, ... and if each recipient re 
article, say, every 15 minutes (only a stroke of a key on a computer * 
others, how long before the whole world sees it?[7J Two to the thirty- 
4.29 billion (not coincidentally, the same as the address space for a 

How then can we convert the added value of the bitstream into a re\ 
The tool that needs to be developed to identify proper use of the rev 
be the same tool used to uncover improper rights infringements. Sm 
identifying blocks of data in a designated bitstream may work in a m; 
environment if only because to sell, distribute, and create a demand 
productions you cannot hide. The SMPTE[8] header/descriptor is su 


Early in the FCC's advanced television selection process there was 
"HDTV is not just about Television. "[10] This eventually prompted a 
harmonization effort to encourage future ATV/HDTV[11] standards t« 
with computer and telecommunication practices. The issues that we 
the protection of intellectual property, that in digital systems "bits are 
distinguishing one kind of data from another is key to use of the date 
parallels in the HDTV process as well. 

It is now accepted by the FCC Advisory Committee on Advanced Te 
(ACATS) that in order to share high-resolution image data across sy 
industry boundaries a universal header/descriptor is required. A hea 
must support digital transmission of video sideband information (clos 
and secondary audio programming) as well as potential new types o 
(image coding parameters and digital copyright signatures). That is, 
be "extensible" for future needs not anticipated today. This is easily < 
designing a structure without limiting the contents of the header/desi 

In 1990, the FCC formally adopted "interoperability" as a ATV select 
Planning Subcommittee Working Party 4 (PS/WP4) was directed to • 
interoperability criteria and to evaluate HDTV proponent systems ac< 
parallel with this formal governmental process, two Society of Motior 


IP Workshop - Gerovac/Solomon: Protect Revenues, Not Bits 

Page 5 of 15 

Television Engineers (SMPTE) task forces were formed: the Header 
force to investigate issues and solutions related to identification and 
digital video streams, and the SMPTE Hierarchy task force to investi 
and other broader architectural issues. Membership of the SMPTE t; 
open to anyone in the world, and numerous trans-oceanic teleconfei 
working meetings were held in this regard. Electronic mail via the wc 
has been heavily used to keep all interested parties informed, includ 
of the task forces. Fax distribution lists supplemented the e-mail; the 
electronics contributed to speeding up the process. 

The SMPTE task forces provided input to PS/WP4 as well as set the 
SMPTE standards and recommended practices. The Header/Descri| 
finished its work and produced a final report on January 3rd, 1992, v 
the June issue of the SMPTE Journal. In April, a SMPTE working gr< 
to take the task force report and produce a standard. 

The FCC working party has considered a family of standards with th 
common to all environments. The header plays a central role coordii 
transmission methods, and application types. 

Headers (and descriptors) are fundamentally data representation ob 
enable exchange and proper interpretation of data in heterogeneous 
The proposed SMPTE structure is but a small part of a greater archi 
framework for advanced, digital information systems and networks. " 
structure under study provide guidance for the identification of digita 
property, as well. 

The definitions and structures discussed are derived from current pr 
telecommunications and computer industry. When television standai 
the early 1940s and 1950s, the need for such practice was not prev< 
broadcasting environments, however the introduction of sophisticate 
digitized capture and storage, and the use of video in many other an 
over-the-air broadcasting has created a set of ad hoc standards that 
evolutionary path to true interoperability. So things like the SMPTE t 
be seen as an early element of header architecture. 


The following objectives form the basis for the Header's design critei 
definition. Design ramifications and implications are sometimes the r 
interaction among two or more objectives. Thus, the descriptions he 
as a whole. 

Unambiguous Self Identification 

The header should uniquely identify the encoding employed for the t 
message and thereby indicate how the data is to be interpreted. 



IP Workshop - Gerovac/Solomon: Protect Revenues, Not Bits 

Page 6 of 15 

Uniqueness. Identifiers must be unique both internationally ar 
industries. When an identifier is extracted from the data strean 
no ambiguity as to meaning. 

Completeness. The header format must be "fully defined", to ; 
Were the format not fully defined, it would not be possible to gi 
extraction of the identifier. 

Sufficiency. Only the identifier should be "necessary and suffi 
determine how to proceed with interpretation of the payload. A 
machine may need additional information (or programming) to 
payload, but the identifier should fully determine how to proce€ 


All video (and associated) data streams should incorporate the head 

Compliant Low Cost Receivers. The desire for broad use of 
translates into the need to minimize cost to the user and thus t 
equipment. Low cost receivers (especially in the near term) ms 
their ability to handle the full scope and flexibility that the head* 
provide. Universality requires that: (a) low cost implementation 
considered in the header design; (b) all compliant receiver imp 
must recognize the header, and properly interpret those fields 
operation; and (c) all compliant data streams must incorporate 

The minimum requirements for a "header compliant receiver" c 

o a minimal implementation must recognize the header len 
interpret it to determine message length — a minimal imp 
recognize the universal identifier field (and subfields) anc 
determine whether the data is appropriate to its operatior 

o all operations must be specified/specifiable using the hes 
i.e., no out-of-band data streams no implementation wi 
interpret header fields 

Cost/Performance Effectiveness. The minimum header shoi 
straightforward to decode so that low cost equipment (as well < 
performance, high quality equipment) can be implemented tha 
properly interpret the header for their operation. Though all imf 
recognize the header, some may choose not to decode all pos 
streams in order to achieve lower cost. 

Compactness. Use of the header should incur a relatively sm; 
the underlying data stream. Compactness is a relative requirer 


IP Workshop - Gerovac/Solomon: Protect Revenues, Not Bits 

Page 7 of 15 

Compliant Data Stream. A minimally compliant data stream ir 
should include a properly and fully encoded minimal header d€ 
message length and identification. 

Sovereignty. Though it is certainly desirable for universality ai 
that there be a small number of standards spanning across na 
economic communities, and trade agreements, it is only realisl 
there will be political desires for sovereignty in standards desig 
this is not a technical issue (per se), and cannot be guaranteec 
design, the structure of the identifier field and its relationship tc 
standards organizations need to be carefully considered and a 
allowances made. 

Standards Compliance. Universality is enhanced by recogniz 
existing work of standards bodies in relation to the design of th 
existing standards and practices are applicable to meeting des 
they should be considered. 

Interactive (two-way) and Broadcast (one-way) Communic 

header and protocol should support both interactive (two-way) 
(one-way) communication. The major implication here is the nc 
protocol to manage information exchange in both kinds of envi 

High Bandwidth/Low Latency Application. The header and 
support full motion (e.g., high bandwidth), live action (e.g., low 
as off-line (e.g., post production and storage) applications. 


The header should be designed to last for a long time. SMPTE sugg 
based on the apparent lifetime of today's TV systems, but if we cons 
documents used daily in law, religion, literature and general culture < 
hundreds, even thousands of years, we might want to consider the ii 
designing a digital structure that could be at least decoded by our he 
generations hence. 


Forward Looking Specification Spaces. Longevity has ramil 
header length and identifier fields. Both need to be consistent i 
needs, yet have large enough "specification spaces" to be app 

Maximum Length. Typically, a header will be associated with 
of frames. (Occasionally, much shorter messages will be used 
situations for general control and information.) The "length fielc 
specify message size appropriate for current uses, yet accomr 
advances in technology. The major factors determining messa 
number of frames, raster size, pixel size, and compression fac 


IP Workshop - Gerovac/Solomon: Protect Revenues, Not Bits 

Page 8 of 15 

message sizes for imagery are on order of 1 MB - some appln 
smaller, some larger. To support resolutions that match high q 
resolution wall-size displays may require on order of 1 GB. 

Number of Unique Identifiers. The number of potential codin 
specified by the IDENTIFIER FIELD is harder to gauge. Only a 
number of coding schemes are often envisioned, perhaps a fe> 
hopefully, the number of standards will be small. However, tec 
permit coding schemes to proliferate. Further, a structured idei 
(rather than a simple numerical ordering) may be helpful in adr 
interpreting identifier values. 

Identifier Immutability. The value of an identifier and its refen 
once assigned, must be "immutable." Practically, it is not possi 
unambiguously the meaning of an identifier across hundreds o 
machines during a transition. Mutability also entails considerat 
universality, longevity, and low cost implementations. 

Identifier Registry. To achieve effective harmonization, one o 
organizations need to be the central authority to allocate and c 
identifiers, and/or a well-defined registration process needs to 
the intellectual property organizations (WIPO, UNESCO) could 
with the ITU and ISO.[12] Registration procedures are largely < 
of design of the header. However, universality suggests that al 
design be given to international standards, standards bodies, i 
Longevity and unique identification objectives, in combination, 
identifier and its specified encoding once assigned and registe 
reassigned or redefined. 

Experimental and Pre-Standardization Uses. It is desirable 
defined method to use the header structure without preregister 
identifier, and thereby, without needlessly littering the identifier 
without delaying experimental/research activities due to registr 
Thus, experimental systems would use special identifiers, and 
in a closed environment for which the particular identifier has n 
open environment, experimental identifiers could contain embe 
interpretation information, or an identified source (instead of a 
queried for instruction on how to interpret the payload. 


The header should permit optimal sharing of data streams across ge 
and equipment technologies and services. 


Well-Formed Public Definition. A header that permits interop 
defined and publicly available. Only then can equipment and a 
producers comply with header requirements. And only then ca 


IP Workshop - Gerovac/Solomon: Protect Revenues, Not Bits 

Page 9 of 15 

assured that equipment and material from a variety of sources 

Varied Requirements. Different applications place different re 
video data stream. For example, some applications will require 
resolution, others not; some applications will require special ot 
spaces; others will simply need to appear nice; some applicatic 
multichannel high-quality audio, others will be silent; etc. Such 
requirements are a major factor in needing to support a large r 
standards identifiers. 

Alternate Standards. Several standards setting bodies are de 
evaluating imaging standards. 

Historical Conventions. Several uses/conventions are histori 
and represent a body of existing material and experience, e.g. : 
per second) film, NTSC/PAL/SECAM video production, synthe 
computer graphics, animation, special effects, simulation, etc. 
design does not address these specifically, the header should 
accommodate existing usage. 

Transcoding. Given the variety of potential encodings/standai 
necessary to translate from one encoding to another. 


The header should be able to incorporate future unforeseen technol* 
algorithmic advances and improvements in quality, performance, an< 
without obsoleting existing components and infrastructure. 


Large Numbering Scheme. To enable longevity, the header c 
provide a large enough numbering scheme to incorporate futui 
alternatives, improvements, and advances in quality, performa 
functionality: 1) the specification space of the header (the leng 
fields) should be big enough to accommodate future expansior 
extensibility should not obsolete existing compliant equipment 

Flexibility. Given the variety of applications and transmissions 
envisioned, the header must be flexible both in its design and i 


At a given time, uniform generation, transmission, and display chara 
support a range of quality and cost. Though more a property of the p 
encoding, the header format should permit scalable encodings. 



IP Workshop - Gerovac/Solomon: Protect Revenues, Not Bits 

Page 10 of 15 

The SMPTE Header is derived from an existing ISO/ITU(CCITT) sta 
use within the computer and telecommunications industries called A 
Notation 1 (ASN.1). ASN.1 is a comprehensive and extensible tool fi 
interchange in heterogeneous transmission and storage environmer 
features is that ASN.1 does not exclude other standards, but acknov 
of alternative methods and provides a mechanism with which to ider 
any data, whether defined in ASN.1 or not. [13] 

It is much like a programming language, such as C, Pascal, or PostJ 
of software tools and utilities to support ASN.1 has been developed, 
types include primitives (integer, Boolean, string, etc.), and construe 
choice, etc.) that can be used to build arbitrarily complex data structi 
process is recursive: types can be constructed from other constructe 
arbitrarily complex structures and substructures may be defined. Fui 
components of constructed types may be optional, allowing for even 

ASN.1 supports the notion of embedding, which allows one or more 
be contained within another. Thus, a sequence of frames can be em 
outer header (or envelope) that labels a program segment. This can 
coarser granularity -- shots, scenes, programs, etc. Similarly, it can I 
granularity to embed audio tracks, closed captioning, descriptors, et< 

Two valuable features of ASN.1 include: 

1 . Separation of data description (Abstract Syntax) and data encc 
Syntax or Encoding Rules). Data structures are described in a 
syntax and automatically translated into bits and bytes for trani 

2. Deployed ASN.1 compliant systems may interpret new structui 
hardware modification. 

The following excerpt is extracted from a tutorial prepared for the SK 

The SMPTE Header is a compatible subset of the ASN.1 EXTI 
ASN.1 compliant protocol interpreters can extract and interprel 
without ambiguity. Its definition is quite flexible, but some comf 
optional allowing for a minimal header that is simple, compact, 
recognized. The ASN.1 notation for EXTERNAL is formally del 

EXTERNAL is a constructed type, meaning a sequence of prin 
types, and is encoded with the same basic tag, length, value fc 

A tag value of 40 decimal (or 28 hexadecimal) identifies the E> 
and indicates the value is defined outside the current ASN.1 cc 
identifier component indicates what standard to apply in decod 
Length indicates the number of octets (octet is the ASN.1 term 

http://\\ Solomon.html 


IP Workshop - Gerovac/Solomon: Protect Revenues, Not Bits 

Page 11 of 15 

bit byte) occupying the remaining message (total message size 
for EXTERNAL tag and length) and is encoded in the usual ms 
described above. 

Thus all SMPTE headers start with the EXTERNAL type tag ar 
The EXTERNAL tag and length fields for a 1010 octet EXTERI 
1000 octet payload prepended with 10 octets of identifier) wou 
octets with the following hexadecimal representation: 

EXTERNAL tag indicates beginning of the header 

| EXTERNAL length occupies the next 2 octets (s) 

| | EXTERNAL value occupies the next 1010 oc 

| | | EXTERNAL value 

I I I I 

28 82 03F2 xx ... xx 

The EXTERNAL value is a sequence of three fields: direct-reference 
reference, and payload. Each field is a primitive ASN.1 type, and is < 
the usual tag, length, and value format (see section 3.3). The direct- 
indirect- reference fields are optional; A header may contain one or 1 
Tag fields are used to indicate the inclusion of optional fields: 

[ tag= 28 ] [ length ] [ direct ref ] 
[ indirect ref ] [ payload ] 

direct-reference. The direct-reference option contains the unr 
is of type OBJECT IDENTIFIER and it uniquely identifies the p 
Identifiers are registered and administered internationally by IS 
their constituent organizations. Identifier administration is desc 

indirect-reference. The indirect-reference option is an integer 
length identifier. It is a more compact and efficient method of ic 
frequently transmitted data. Identifier-to-integer mappings are < 
time of transmission, either though bi-directional negotiation or 
assignment by including both direct- and indirect-reference opl 
the same header on a periodic basis. 

payload. The payload field is an octet string encoded accordir 
identified in the direct- or indirect-reference fields. The payload 
aligned option defined in the formal ASN.1 EXTERNAL type s[ 


The header's direct-reference field contains a universal identifier ind 
payload is encoded. Identifier values are assigned, registered, and z 
either (1) by CCITT and ISO, or (even WIPO, UNESCO, etc.) in the 
standards development; or (2) by delegated member bodies, compa 
organizations (such as SMPTE, IEEE, etc.), who assume responsibi 
administering a portion of the identifier space. Solomon.html 


IP Workshop - Gerovac/Solomon: Protect Revenues, Not Bits 

Page 12 of 15 

Identifier Hierarchy 

Identifiers are organized in a hierarchy. The root (prefixes) of the ide 



|- recommendation [0] 
|- question [1] 
|- administration [2] 
|- network operator [3] 


|- standard[0] 

|- registration authority [1] 

I - member body [2] 

|- identified organization [ 3] 



joint ISO CCITT[2] 

CCITT committees 
CCITT Study Groups 
country PTTs (count 
X.121 organizations 

ISO standards 
ISO authorities 
member bodies (coun 

delegated to SMPTE 

delegated to ANSI 

A few prefixes are of particular interest, iso.standard registers all ISC 
ccitt.administration and iso.memberbody are assigned to sovereign I 
by their international telephone country code). Portions of iso.organi; 
delegated by ISO to organizations and companies so that the individ 
can manage the assignment of their own portion of the identifier spa 

Header Examples 

The following examples show two commonly used header configura 
example shows a header for a message containing 1000 octets of P 
imagery. Only the direct-reference form of identification is used and (Px64) as shown above. This example puts together man> 
shown in previous sections. The complete header is encoded in 14 < 
of the total message) and has the following hexadecimal representa 

EXTERNAL tag indicates the header is a constructed ASN.l 

| EXTERNAL length occupies the next 2 octets (s) 

I | EXTERNAL value occupies the next 1010 octets 

I | | OBJECT IDENTIFIER tag indicates use of direct - 

I I || OBJECT IDENTIFIER occupied the next 4 octet 

II III I payload tag 

II III II payload length in next 2 oct 

II III III payload occupies next 1 


28 82 03F2 06 04 00088205 81 82 03E8 

The next example shows a header for a 100-octet long payload with 
reference option used, value (1), could be an alias for a copyright de 
complete header is encoded in 7 octets (about 7% of the total mess; 


IP Workshop - Gero vac/Solomon: Protect Revenues, Not Bits 

Page 13 of 15 

following hexadecimal representation: 

EXTERNAL tag indicates the header is a constructed ASN.l 

I EXTERNAL value occupied next 105 octets 

I | INTEGER tag indicates use of indirect-reference opt 

I | | indirect-reference value in next octet 

I I I I indirect-reference value is 1 

I I I I I payload tag 

I I I I I I payload occupies next 100 octets 
I I I I I I I 
28 69 02 01 01 81 64 


Following is a trivial example of how one might describe a copyright 
It serves only to elicit formal definition of a universal digital copyright 
by a joint body of intellectual property, communications, and comput 

Copyright : := SEQUENCE 

version INTEGER > 

version-0 .1(0) 


years SEQUENCE OF NumericString, 
bylines SEQUENCE OF PrintableText , 

all -rights -reserved ( 0 ) 


permission PrintableText OPTIONAL, 
disclaimer PrintableText OPTIONAL, 

payment-method ElectronicPaymentStandard OPTIONAL 



1. These concepts were outlined in detail for the Congressional Offic 
Assessment by one of the authors (Solomon) as a contractor on the 
intellectual property, published in 1985. See R. Solomon, "Computei 
Concept of Intellectual Property," in Martin Greenberger, ed., Electrc 
Plus, Knowledge Industry, 1985. Also see R. Solomon & Jane Yurov 
Electronic Technologies and International Intellectual Property Issue 
Technology Assessment , U.S. Congress, May 1985; and, R. Solom 
Property and the New Computer-Based Media," Office of Technolog 
U.S. Congress, August 1984. 

2. 'The right only began to assume importance when the invention o 
the multiplication of 'copies' of a work infinitely quicker and cheaper 
painstaking products of monkish scribes, as well as appreciably mor 
the compositions of most professional scriveners." I. Parsons, "Copv 
Society," in A. Briggs, ed., Essays in the History of Publishing, 1974, Solomon.html 


IP Workshop - Gerovac/Solomon: Protect Revenues, Not Bits 

Page 14 of 15 

3. Payments to the Crown for the privilege of publishing via print. 

4. Grossman, Bernard A. "Cycles in Copyright," New York Law Schc 
22:2&3 (1977). G. Blagden, The Stationers' Company, 1960. 

5. Grossman, op. cit, p. 263. 

6. Or a Type 4 Turing machine, depending on whom you wish to giv< 

7. R. Solomon, in a paper co-authored with the late Ithiel de Sola Po 
this in a different context for the OECD in a discussion of transborde 
"Intellectual Property and Transborder Data Flows", Stanford Journa 
Law, Summer 1980; and The Regulation of Transborder Data Flows 
Telecommunications Policy, September 1979. 

We leave it to the reader to calculate how long it takes to reach the \ 
every 15 minutes the number of recipients doubles. 

8. Society of Motion Picture and Television Engineers (U.S.). 

9. The details of the header structure is still work-in-progress which i 
to this forum for comment and suggestions. The precise formats, fiel 
etc., are subject to substantive change as the standardization procei 
invite comments and suggestions. 

10. Generally ascribed to Prof. William F. Schreiber of MIT, circa 191 

11. Advanced Television and High-Definition Television. 

12. World Intellectual Property Organization, United Nations Econon 
Commission, International Telecommunication [NO 'S'] Union, Interr 
Standards Organization. 

13. ASN.1 is derived from work at Xerox Palo Alto Research Center 
Courier in the late 1970s. A 1984 version was used in the first draft ( 
X.400 series of recommendations on message handling systems. IS 
jointly developed ASN.1 in 1988 for the presentation layer of the Op< 
Interconnect model. 

ASN.1 is now widely used in a range of international standards activ 
the CCITT X.500 directory service, and both OSI and Internet netwo 
protocols, the Common Management Information Protocol and Simp 
Management Protocol respectively. This suggests the possibility tha 
systems may be networked devices that may fit into a common netw 

14. For a formal description of ASN.1 refer to ISO 8824/8825 and C< 
A more accessible description can be found in: Marshall T. Rose, 77 Solomon.html 


IP Workshop - Gerovac/Solomon: Protect Revenues, Not Bits 

Page 15 of 15 

Practical Perspective on OS/, Prentice Hall, 1990. 

Branko Gerovac is with the Digital Equipment Corp., Maynard, Mass 
associate of the MIT Program on Digital Open High Resolution Syst< 

Richard J. Solomon is Associate Director of the MIT DOHRS Progra 
for Technology, Policy and Industrial Development, Cambridge, Mas 

© 2002 Coalition for Networked Information. All Rights Reserve 
Last updated Wednesday, July 3, 2002. 


IP Workshop - Upthegrove/Roberts: Intellectual Property Header Descriptors 

Page 1 of 5 

Sponsorec by: 


About CUt 

Task Force 





Site Map 

Search our she 


[Coalition for Networked Infor mation 

Intellectual Property Header 
Descriptors: A Dynamic 

, by Luella Upthegrove and Tom Roberts 


The global electronic library will need standards that facilitate 
the controlled distribution and protection of digitized 
intellectual properties, and that encourage library expansion 
and access. This paper describes a system based on 
intellectual property distribution and protection that is 
currently being tested at Case Western Reserve University, 
and defines a global header descriptor applicable to the 
electronic distribution of intellectual properties. 


The mission of the Library Collections Services Project (LCS) 
at Case Western Reserve University (CWRU) is to establish 
an online multimedia repository to serve the academic and 
research needs of the CWRU community. To this end, LCS 
has created a number of prototype applications that 
demonstrate the potential of a networked multimedia 
repository. These prototypes address the interests of the 
providers and consumers of intellectual property (IP) resident 
in the repository. 

Early on, the LCS project team recognized the responsibility 
it had to maintain and protect the electronic IP. The team 
collected IP management requirements by meeting with 
members of the publishing and legal communities, 
reproduction rights organizations, librarians, online 
information service providers, and academicians. 

The issues that resulted from these meetings fall under the 


IP Workshop - Upthegrove/Roberts: Intellectual Property Header Descriptors 

Page 2 of 5 

general headings of: IP protection, IP use management, and 
royalty compensation. 


The LCS team began its system design by defining end-to- 
end system components based on the requirements 
gathered. The requirements fell logically into the broad 
categories of: Ownership, Compensation, Permissioning, 
User Access, Privacy/Confidentiality, and Permitted Uses. 
Prior to building the prototype repository and applications, 
appropriate permissions were obtained from participating 

Applications were designed that verify user authorization, 
access the IP, and manage IP use before and during 
repository access. These applications compare user 
information and usage data coupled to the IP to determine 
access and use. Comparison of user information and usage 
data satisfies the protection and use requirements detailed in 
license agreements negotiated with IP rightsholders. 

These applications, all of which adhere to a set of data and 
protocol standards, are called compliant applications. Only 
compliant applications can access the IP in the repository. 


To expand this model, consider that all IP consumers are 
part of the local environment. They access repositories on 
which they have registered accounts, and information 
passed between user and repository is managed at the local 
level. Users maintain the ability to query information 
contained in remote repositories; however, the request for 
the IP transacts between the user's local repository and the 
remote repository in the global environment 

The LCS model can be expanded to this global environment. 
To accomplish this the following assumptions, significant 
issues in themselves, are made: 

1 . Permissions for storage, access, use, and 
compensation have been negotiated and agreed upon. 

2. Compliant applications are resident and in use on all 
participating systems. 

3. Economic structures for billing and compensation have 
been established.^ 


EP Workshop - Upthegrove/Roberts: Intellectual Property Header Descriptors 

Page 3 of 5 

4. Technical strategies for locating IP are in place. 

A typical global transaction may develop as follows. 

A user locates and requests IP on a remote repository. The 
request is routed through their local repository to the remote 
repository where the requested IP resides. The protocol of 
this communication contains a standard request that includes 
information identifying the IP, the requesting repository, the 
user environment specifications, and the intended use. The 
remote repository verifies the request, constructs a header 
descriptor based on that request, and replies to the 
requesting repository. This header descriptor is in the form of 
a standardized global header descriptor. 

Using the local environment presented earlier as a base, 
data common across the global environment can be 
identified. The following elements are proposed for inclusion 
in a global header descriptor. 


to include rightsholder identification and contact information 
for use in compensation, special permissioning, and 
copyright code compliance. 

Permitted Uses 

to include uses as negotiated with the rightsholder detailing 
authorized users, display resolutions, print capability, etc. 

Royalty Compensation 

to include the compensation framework as it relates to the 
permitted uses. 

IP Attributes 

to include the physical attributes, and component parts 
comprising the IP. 

To accommodate this information, each descriptor element 
would contain a variable length data string preceded by a 
standard ID. These elements would be mapped to the local 
repository for use by functionally compliant applications. 

In the form of a dynamically generated global header 
descriptor, information common, particular, and primary to IP 
providers and purveyors can be developed to enable global 


IP Workshop - Upthegrove/Roberts: Intellectual Property Header Descriptors 

Page 4 of 5 

and local protection and use management. 


The challenge of developing standards for the global 
electronic library may seem overwhelming; however, inaction 
will render the vision vain. Opportunities are afforded to 
those who begin now to define the framework of the new 

This paper presents a prototype system designed for 
intellectual property protection and use management in a 
local electronic environment. It then begins to describe a 
global header descriptor based on the two premises that: the 
electronic environment is comprised of local users connected 
to primary global repositories, and that intellectual property 
access is mediated by applications compliant to established 
protection and use monitoring requirements. 

To this end, it is proposed that a global header descriptor 
contain a set of data elements that identify intellectual 
property: Ownership, Permitted Uses, Royalty 
Compensation, and IP Attributes. 

Local and global standards must cooperate to provide 
access and use controls such that IP providers, purveyors, 
and consumers are confident that their interests are 
protected. Properly designed standards will enable 
repositories to fulfill their responsibilities, and encourage the 
use and expansion of the global electronic library. 


Thomas Roberts, DBA Communication Media and 
Documentation Services, is consulting with CWRLTs LCS 
Project. CM&D specializes in knowledge transfer using 
established and emerging technologies. With CWRU, CM&D 
is identifying and analyzing copyright, permissioning, and 
royalty compensation issues as they apply to electronic 
intellectual property distribution. 

Tom Roberts 

Library Collections Services 
Case Western Reserve University 
10900 Euclid Ave., Baker 6 
Cleveland, OH 44106-7033 
t f r4 @po . cwru . edu 

Luella Upthegrove, Database Administrator for CWRLTs LCS 
project, helped design and develop the prototype electronic 


IP Workshop - Upthegrove/Roberts: Intellectual Property Header Descriptors 

Page 5 of 5 

library, and is currently involved in planning for the second 
version of the system. 

Luella Upthegrove 
Library Collections Services 
Case Western Reserve University 
10900 Euclid Avenue 
Cleveland, OH 44106-7033 
(216) 368-8921 
FAX: (216) 368-8880 

© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002.^ 


IP Workshop - Sirbu: Internet Billing Service 

Page 1 of 15 

Sponsorec by: 

A a 


About CM 

Task Force 





Site Map 

Search our she 


| Coalition for Networked Information 

Internet Billing Service 
Design and Prototype 

by Marvin A. Sirbu 


A group of students in the M.S. program in Information 
Networking at Carnegie Mellon University have designed and 
implemented a prototype of an Internet Billing Service--an 
electronic credit card service for the Internet environment. * 
The service provides account management, authentication, 
access control, credit verification, management reporting, 
billing, and collection services to network-based service 


A worldwide data networking infrastructure is gradually falling 
into place which will allow consumers and service providers 
to interact in a vast electronic marketplace. In France some 
9,000 services are available over the Minitel network. In the 
U.S., access to electronic databases generates billions of 
dollars a year in business. As networked computers 
proliferate at home and in the workplace, more and more 
consumers are in a position to shop in the electronic 

Already the Internet, a loose confederation of independently 
managed networks, links eight million users and some one 
million computers on 10,000 separate subnetworks in more 
than 40 countries. Once used only by universities and 
research organizations, the Internet is used today by 
individuals and corporations for a wide range of commercial 
purposes including disseminating information, searching 
remote databases, and providing access to specialized 


IP Workshop - Sirbu: Internet Billing Service 

Page 2 of 15 

computer resources. Major information service providers 
such as Dialog and BRS can now be reached via the 

While it is relatively simple for an entrepreneur to set up a 
small computer capable of providing information to the 
worldwide Internet community, it is much more difficult to 
arrange a mechanism to charge users for the services 
rendered and to collect payments. Current billing 
mechanisms for electronic services are costly and 
inconvenient for both service providers and end users. In the 
absence of a centralized billing service, users must initiate a 
service agreement with each service provider before using 
its services, and must keep track of its access point, 
password, and bills. Also, there is no central directory of 
service providers. It is uneconomical for small service 
providers to advertise, check credit, authenticate users, 
control access, bill and collect payments, maintain audit 
trails, and keep usage statistics. 

Both service providers and end users need a reliable, easily 
accessible, fast and inexpensive intermediary, a billing 
service. The billing service could be compared to an 
electronic credit card for services on a network. It would 
allow small service providers to concentrate on providing 
services by contracting out the functions of billing users and 
collecting payments. 

This document describes the design and implementation of a 
prototype of a computer-based Internet Billing Server (IBS) 
developed by a project team from the Information Networking 
Institute (INI) at Carnegie Mellon University. The project 
team had two major tasks: (1) specify functional 
requirements for a full-scale billing server, and (2) design 
and develop a prototype which, while satisfying only a subset 
of these requirements, demonstrates the feasibility of the 
concept. Specification of the full set of requirements brings 
out important issues and ensures that the prototype design 
has no major flaws which limit its extensibility to a full-scale 
system. We will summarize the full set of requirements, 
noting in passing where the actual prototype differs. Our 
design demonstrates how an Internet Billing Server could 
facilitate the emergence of a real electronic marketplace. 


The design of a network billing server is difficult because 
markets for electronic services are different from markets for 


IP Workshop - Sirbu: Internet Billing Service 

Page 3 of 15 

physical goods and non-electronic services. A network billing 
server must be designed to take these differences into 
account, in order to avoid fraud and disputes. These 
differences are outlined below. 

First, in a network, the users, the service providers, and the 
billing service are geographically separated. Credit cards are 
designed for situations where the users physically present 
their credit cards at the time of purchase so that merchants 
may validate their signature. Although credit cards are used 
today for placing orders over the phone, such methods are 
highly insecure; ordering over a network makes it difficult to 
verify the identity of the parties. Indeed, to reduce fraudulent 
charges, many merchants will only ship goods ordered over 
the phone to the billing address of the credit card holder. 
Secure network authentication protocols, such as Kerberos, 
may be part of a solution but the legal liability and 
responsibility of participants in an electronic market is not 
well defined. 

Second, given the high processing speeds of electronic 
services, a user can accidentally run up a huge bill within a 
matter of seconds without having an opportunity to cancel it. 
In contrast, if there is a mistake in ordering a physical 
product, jt can be corrected before the product is shipped or 
the product can be returned after delivery. Even though a 
non-electronic serwce cannot be returned, the user can still 
cancel it while it is being performed: one can leave a hotel if 
its service is not satisfactory or is too expensive. Providing a 
similar capability for halting the provision of network-based 
services in midstream is complex. 

Third, in contrast to physical goods, it is difficult to determine 
the price for an electronic service in advance. For example, if 
the price of a database query were based on the number of 
bytes of information it generates, it would be difficult to 
determine the query's price without searching the database. 
It is infeasible to let the user have a look at the information to 
assess its value; it is also infeasible to price information 
solely upon objective measures such as its size in bytes. 
This difficulty in judging the quality of information may give 
rise to disputes which are difficult to resolve. This makes it 
important to carefully define the legal role of the billing 

Fourth, electronic information can be easily duplicated and 
redistributed. This makes product "returns" meaningless 
because a user can copy an electronic file before returning it. 
It is also much easier to copy and redistribute an electronic 


IP Workshop - Sirbu: Internet Billing Service 

Page 4 of 15 

version of a book than copy and redistribute a printed version 
of the same book. 

The INI Internet Billing Server is able to address some, but 
not all, of these issues. Security is provided using passwords 
and encryption. The IBS also provides a capability for setting 
and enforcing spending limits. The billing server provides a 
flexible mechanism for charging for network services, and for 
price negotiation, but it does not pretend to resolve the 
problem of determining a service's value. Nor does it 
address the issue of illegal copying and redistribution of 
purchased information. 


A billing server plays the same role vis-a-vis end users and 
service providers as a credit card company does vis-a-vis 
cardholders and merchants. Consider a credit card holder 
going to rent a car. He begins by identifying himself to the 
rental car company by presenting his credit card and driver's 
license. He negotiates with the car company for the service 
he desires, the cost per day and the maximum number of 
days he expects to keep the car. The merchant then verifies 
the customer's credit with the credit card issuer and places a 
hold on the customer's credit for the maximum amount of the 
rental. When the customer returns the car, the transaction is 
complete, and the car rental agency sends a final invoice to 
the credit card company, with a copy to the consumer. At the 
end of the billing cycle, the credit card company sends a bill 
for all purchases charged to the card, including the car 
rental, and the customer sends back his payment. The credit 
card company pays the merchant after deducting its fees. 

Our model of transactions in the network marketplace is 
similar to the car rental scenario: a customer or end user- 
through his computer-interacts over a network with two 
other computer systems: the service provider, and the 
Internet Billing Server, as illustrated in Figure 1 . 


All Qf the steps described above for renting a car have their 
counterparts in the network marketplace. However, instead 
of face-to-face communications as in the car rental scenario, 
the end user's computer interacts with the service provider's 
computer over the network. Each step in the interaction 
forms part of the Internet Billing Protocol (IBP), a proposed 
standardized method of interaction among an end user's 
computer, the service provider's computer, and the Internet 


IP Workshop - Sirbu: Internet Billing Service 

Page5 of 15 

Billing Server computer. 

The Internet Billing Server is more than a computer and a set 
of standardized protocols, however; it is a model for a 
business which provides valuable services to network 
marketplace entrepreneurs. The Internet Billing Service acts 
as a factor for the service provider, providing prompt 
payment while taking over all aspects of billing and 
collections. To be successful as a business, the Internet 
Billing Service must satisfy two sets of customers. It must 
attract merchants by giving them easy access to a large 
group of customers, and providing them a cost-effective way 
to receive payment for services provided. It must make it as 
easy as possible for service providers to make use of the 
Internet Billing Server, working with the providers to modify 
both client and server software to implement the Internet 
Billing Protocol. It must attract end users by providing a 
powerful and flexible capability for managing end-user 
accounts and by giving end users access to a large number 
of service providers. 

While we have described the Internet Billing Service as an 
independent business, large organizations often have a need 
to create an internal equivalent of an Internet Billing Server. 
For example, the central administration at a university such 
as CMU could operate a billing server as a single 
mechanism to bill for online library access, computing 
services, printing services, and electronic mail. While we 
recognize this as another potential application for the billing 
server software developed in this project, the focus has been 
on the design and implementation of a public, third-party 
billing server. 

As designed by the project team, the INI Internet Billing 
Server provides the following functions to service providers 
and end users: 

# Account Management The IBS enables end users to 
establish an account relationship with the Billing Server 
which will permit them access to any number of service 
providers. Service providers establish accounts which 
enable them to use the IBS to bill their clients for 
services rendered. 

• Authentication: The IBS provides a service for 
authenticating both end users and service providers 
prior to any transaction and for secure communications 
between the parties. 


IP Workshop - Sirbu: Internet Billing Service 

Page 6 of 15 

• Access Control: The IBS provides access control for 
both end users and service providers. Information 
associated with an end user account can specifically 
designate a list of services that may be accessed, or a 
list of services that specifically may not be accessed. 

• Price Negotiation: Using the Internet Billing Protocol, 
the end user may determine the services available from 
the service provider and the posted prices. The Internet 
Billing Server can record the mutual agreement of the 
end user and the service provider on a set of prices, 
and the maximum amount which the end user has 
authorized for this set of transactions. 

• Credit Verification: The Internet Billing Service will 
verify to the service provider that the customer has 
sufficient credit to pay for the proposed transaction up 
to the agreed cap. 

• Final Invoice: At the conclusion of the transaction, the 
service provider can send a final invoice to the Billing 
Server using the IBP. The Internet Billing Server will 
send an authenticated copy of the invoice to the end 

• Periodic Billing: The Internet Billing Server will generate 
periodic billing statements to customers detailing all of 
their transactions and the sums owed. 

• Collections: The Internet Billing Service will collect 
funds from end users and make payments from these 
funds to service providers. 

• Directory Services: The Internet Billing Service 
provides a "white pages" and "yellow pages" service for 
identifying service providers. 

• Help Service: The Internet Billing Server provides an 
online help manual service. 

• Software Libraries. In the client server model, every 
service provider must make available to end users 
client software capable of accessing the service- 
provider's service. A successful Internet Billing Service 
must provide a set of library routines which make it 
simple to upgrade both client and server software to 
support the Internet Billing Protocol. These modules 
are shown logically in Figure 2. 


IP Workshop - Sirbu: Internet Billing Service 

Page 7 of 15 



In developing the Internet Billing Server and the Internet 
Billing Protocol we were guided by several fundamental 

• The Internet Billing Server will operate in a transaction- 
oriented environment. All communications between the 
parties will be based on a remote procedure call 
communications paradigm. This is in sharp contrast to 
most current network-based information services. 
These typically have been provided via large 
timeshared computers. Users log in over low-speed 
networks from dumb terminals-or PCs emulating dumb 
terminals-and are charged by connect time. However, 
as desktop computers have replaced dumb terminals 
and networks have increased in speed, a new 
information access paradigm has emerged: client- 
server. In this paradigm, powerful desktop computers 
running user-friendly client software interact with 
remote servers on a transaction basis. In a few 
seconds large files of information can be requested and 
transferred from servers to clients. File Transfer 
Protocol, Gopher, and Wide Area Information Service 
(WAIS) are but a few of the client-server protocols used 
by numerous clients and servers on the Internet. In a 
client-server environment there is no notion of connect 
time. Accordingly, services must be billed on a per- 
transaction basis. 

• The billing server should have high availability since, in 
its absence, the service providers will not be able to 
offer their services to end users. 

• The billing server should be scalable. It is difficult to 
predict the initial number of customers and the growth 
pattern, even though the number of potential customers 
is large. These latter two points suggest that the billing 
server should be designed to run on replicated, 
distributed computers, thus providing modular 
scalability and high availability through redundancy. 

• Communications between the parties-end user, 
service provider, and billing server-should be based on 
widely available telecommunications standards to 
ensure the largest market for the services. 


DP Workshop - Sirbu: Internet Billing Service 

Page 8 of 15 

• Secure authentication and encryption are critical 
because all three parties will be connected via insecure 
public networks. Without a secure authentication 
mechanism there is a substantial potential for fraud. 

• Before using a service, the end user must understand 
and agree to the prices and terms of the exchange. 
The transaction protocol in the billing system must 
support an initial price negotiation between the end 
user and the service provider. The billing server should 
be informed about the outcome of this negotiation by 
both the service provider and the end user. To avoid 
disputes the billing server should make sure that the 
user and the service provider have the same version of 
the agreement. 

• The users should be able to limit their financial 
exposure on a transaction by specifying a spending 
cap. If the cost of an ongoing transaction exceeds this 
spending cap then the end user should be able to 
choose whether to abort the transaction, or continue it 
by raising the spending cap. 

• The billing server should not become a bottleneck 
slowing the speed of interaction between the end user 
and the service provider. In particular, the billing server 
should not be a gateway for communication between 
the end user and the service provider. Interactions with 
the billing server should be as few and as simple as 

• The billing server software should help users in their 
account management. It should support hierarchical 
accounts so that corporate users can get bills 
aggregated by organizational units such as 
departments, regions or divisions. Similarly, a provider 
of multiple services may use an account hierarchy to 
organize information on the use of each service. 

With these general requirements in mind, the project team 
prepared a detailed requirements document specifying all of 
the capabilities required of the Internet Billing Server. 


Our prototype Internet Billing Service was implemented using 
widely available technology. The Billing Server prototype is 
designed to run on a Digital Equipment Corporation 
workstation class machine running the Ultrix operating 


IP Workshop - Sirbu: Internet Billing Service 

Page 9 of 15 

system. It was written in C and uses the Ingres Database 
Management System. It also uses Transarc Corporation's 
Base Development Environment (BDE) to provide 
multithreading, which allows the prototype to process 
concurrent requests efficiently. For communication between 
the billing server, end users, and service providers, the 
prototype uses the remote procedure call (RPC) portion of 
the Distributed Computing Environment (DCE) provided by 
the Open Software Foundation and the Transmission Control 
Protocol/Internet Protocol (TCP/IP) protocol suite. 
Implementations of DCE are available for the OS2/2.X and 
Microsoft's Windows NT operating systems as well as Unix. 

Authentication is implemented using the Kerberos protocol 
developed at M.l.T. All communications between the parties 
are encrypted for security using the Data Encryption 
Standard (DES) encryption method. 

Code libraries enabling rapid modification of client and server 
software to support the Internet Billing Protocol were written 
in C. As a test of the complete system, we modified versions 
of File Transfer Protocol (FTP) client and server software to 
make use of the Internet Billing Server. Using these software 
packages, a service provider could distribute information 
using FTP and bill for it using the Internet Billing Server. 


Figure 3 illustrates the sequence of steps involved in the use 
of the Internet Billing Server. 

Step 0 - Establishing an Account 

Prior to engaging in a network-based transaction, and end 
user must first establish an account with the Internet Billing 
Server. In our design, any number of accounts may be 
organized in a hierarchical fashion by allowing each account 
to have sub-accounfs, each of which is also an account. See 
Figure 4 for an example of a single hierarchical account 

The hierarchical structure represents authority over 
accounts; the end user of a parent account has authority 
over the end user of a sub-account. Every hierarchy has an 


IP Workshop - Sirbu: Internet Billing Service 

Page 10 of 15 

Account Administrator, who is able to view financial 
information, and modify certain account characteristics for all 
the accounts in the hierarchy. 

Billing and usage information can be aggregated by various 
branches of the hierarchical structure, or detailed information 
for each node in the structure can be supplied. An 
organization should have the ability to give managers 
privileges to modify some of the information for the accounts 
of their subordinates. This hierarchical structure allows the 
account environment within the Internet Billing Server to 
mirror the environment within organizations. 

Account hierarchies may be individually or collectively 
billable. In the first case each account is fully billable, i.e., it 
contains all the financial information such as balance due, 
adjustments, payments and usage information; the hierarchy 
is used merely to provide aggregate billing information for 
management and control. This model may be more 
appropriate for decentralized organizations. In the second 
type of hierarchy only the parent or root of the hierarchy is 
fully billable, i.e., only the root account contains the full billing 
information for the hierarchy whereas for other accounts only 
the usage information is listed. The prototype only supports 
hierarchies where each account is a fully billable account. 

Service providers which offer multiple services may want to 
order their accounts in a hierarchy to help maintain valuable 
marketing and usage information. Since each distinct service 
requires a unique Kerberos identifier and an account will not 
provide multiple Kerberos identifiers, each service must be 
given a separate account within the Internet Billing Server. 
By allowing these accounts to be placed into a hierarchical 
structure, the Internet Billing Server can make one 
aggregated payment to the service provider instead of a 
separate payment for each service. In addition, hierarchical 
accounts make it easier to supply the service providers with 
one statement containing usage information for all of their 

Step 1 - User Authentication and Access Control 

Since a network is a mutually suspicious environment, the 
service providers, the end users, and the billing server must 
authenticate each other prior to any transaction. This step 
may be compared to credit card users showing their driver's 
license to prove their identity while using a credit card. As 
noted, we use a Kerberos-based authentication system for 
secure communication between end users, service 


IP Workshop - Sirbu: Internet Billing Service 

Page 11 of 15 

providers, and the billing server. Cross-realm Kerberos 
authentication may be required in the full-scale system if 
large users authenticate end users within their organization 
and then ask the billing server to accept their authentication. 
However, our prototype does not need cross-realm Kerberos 
authentication because it functions within a small group of 
end users and service providers. All communication is 
encrypted using the Data Encryption Standard for security. 

After authentication, the end users may directly request 
access to a specific service provider, or may search through 
an index of service providers classified by service categories 
to select the service they want. The prototype does not 
support the directory service. The billing server checks the 
access control lists of both the end user and the service 
provider to ensure that the end user is allowed to access the 
requested service. In the full set of requirements, the end- 
user and service-provider accounts may have two types of 
access control lists: (1) two positive access control lists-one 
listing specific service providers/end users and the other 
listing categories of service providers/end-users; and (2) 
similarly, two negative access control lists. End users' lists 
specify which service providers they can (positive lists) or 
cannot (negative lists) access; similarly, service providers' 
lists specify which end users are allowed or not allowed 
access to them. 

The negative access control lists override the positive lists, 
and determine which specific services or service categories 
cannot be accessed from the account. Corporate users could 
use positive access lists to allow access only to company- 
approved service providers. Parents could use negative 
access lists, analogous to 900 telephone service blocking, to 
prevent their children from accessing frivolous or high-cost 
services. We have implemented only negative access lists 
for end users and only positive access lists for service 

If access is allowed, the billing server issues the end user a 
Kerberos ticket for the service provider; that authenticates 
the end user to the service provider. As mentioned before, 
the end user must have the client software specific to the 
service provider (for example FTP or Internet Gopher 
interface software) in order to access the service provider. 

Steps 2 and 3 - Price Negotiation and Spending Cap 

After getting a Kerberos ticket from the billing server, the end 
user and the service provider negotiate a price for the 


IP Workshop - Sirbu: Internet Billing Service 

Page 12 of 15 

requested service and a spending cap for the transaction. 
This is called an agreement. Note that the end user is 
communicating with the service provider's computer, not with 
a human representative of the service provider. 

The end user sends a copy of the agreement, encrypted with 
his private key, to the service provider who forwards the end 
user's copy to the billing server, along with his own copy of 
the agreement. This prevents an unscrupulous service 
provider from changing the agreement before sending it to 
the billing server. It also reduces the communication load on 
the billing server, since it receives only one combined 
message rather than two separate messages from the 
service provider and the end user. 

The full-scale billing server allows renegotiation of spending 
caps if the initial spending cap proves to be insufficient. 
However this capability was not implemented in the 

Step 4 - Verifying Spending Cap and Credit 

The billing server decrypts the two copies of the agreement 
and compares the end user's version with the service 
provider's version. If the two copies match, then the billing 
server checks if the end user has sufficient funds to pay for 
the transaction and places a hold on the end user's funds in 
the amount of the spending cap. It then sends an 
authorization to the service provider. 

In the full-scale server, the end users can specify their 
preferred payment method; this could be historical billing, 
advance deposit or credit card. With historical billing the user 
receives a bill for the services that they used at the end of a 
specified period of time. Advance payment means that the 
user deposits funds with the billing server before using 
services, and receives a periodic statement of the services 
used and the funds remaining. With credit card billing, their 
credit card is billed when accumulated charges reach a 
specified limit. In addition to these three options, corporate 
users can use purchase orders, a form of historical billing, for 
making payments. The prototype allows deposit in advance 
as the only payment method. 

Step 5 - Performing the Service 

Messages are exchanged between the client and the service 
provider to perform the service-e.g. retrieve information, 
perform calculations, or spool a print file. 


IP Workshop - Sirbu: Internet Billing Service 

Page 13 of 15 

Step 6 - Generating an Invoice 

After the service provider has rendered the service, it sends 
the billing server an invoice detailing the services performed 
and the actual amounts to be charged. The billing server 
checks whether the price information on the invoice is 
identical to the price information received earlier during the 
price negotiation stage. This protects the users from 
unscrupulous service providers. The billing server then 
forwards this invoice to the user. Since the identity and credit 
capacity of the end users were previously checked by the 
billing server, the service providers have assured payment 
for their services. The service providers are required to 
maintain an audit trail to handle customer inquiries which 
cannot be resolved by the billing server. 


End user accounts can go through various states, as 
illustrated in Figure 5. To open an account, the end user 
account administrator sends a request to the billing server. 
Once all of the information required for the creation of an 
account is entered, the account enters the "new" state. An 
account cannot begin to access services until the billing 
server's account administrator verifies and approves the 
account characteristics and the billing server's financial 
administrator verifies and approves the financial information. 
Once these verifications are complete, the account is 
activated. An account which has been activated enters the 
"active" state and is then allowed to accumulate charges for 
services it accesses through the billing server. 


An account goes into the "deactive" state if it has an overdue 
balance for an unreasonable period of time. It goes back to 
the "active" state if the balance is paid. An account can also 
enter the "closed" state by end user request or if payment is 
not received while it is in the "deactivated" state. 

Accounts may enter the "paid," "written off," or "referred to 
agency" states after being in the "closed", state. An account 
enters the "paid" state if the final balance due is paid in full. 
An account enters the "written off' state if the billing server 
financial administrator determines that payment for the 
balance due will not be received. An account enters the 
"referred to agency" state if the billing server's financial 
administrator refers the account to a collection agency. 


IP Workshop - Sirbu: Internet Billing Service 

Page 14 of 15 

Since the prototype handles only debit model accounts 
where users pay in advance, there is no need for an 
approval process. The "new" state is not needed. Again 
because of payment in advance and the credit check 
performed during transactions, end users 1 accounts cannot 
owe money to the billing server. Therefore, the "deactive" 
"paid," "written off," and "referred to agency" states are also 
not needed. In the prototype, when an account is "closed" it 
is removed from the database. Therefore account states are 
not supported by the prototype. 

Users can access their own account information at the billing 
server through an interface that allows them to view financial 
information, and to modify certain account characteristics. 
The full-scale billing server also provides on-line help to its 
users. The help, which can be accessed through a keyword 
search, consists of text screens describing how to perform 
basic operations. Since on-line help is not central to the 
billing server, it is not implemented in the prototype. 


What distinguishes the Carnegie Mellon project from other 
piecemeal or service-specific solutions is its comprehensive 
analysis of the network services billing problem. The project 
has made two contributions: (1) it has highlighted the 
complex and challenging issues involved in the design of the 
Internet Billing Server; and (2) it has demonstrated the 
feasibility of its proposed solution through the successful 
design and implementation of the prototype. Even though the 
prototype implements only a subset of the full requirements 
and may have to be significantly modified, it is an important 
first step. A commercial service based on the concepts in the 
INI Internet Billing Server could be the key to the rapid 
growth of entrepreneurial service providers in the Internet 

For further information, please contact The Information 
Networking Institute at Carnegie Mellon University, 
Pittsburgh, Pennsylvania 15213. Tel: (412)-268-7195. 


1. John Quarterman. "In Depth. (What can businesses get 
out of the Internet?)" Computerworld, February 22, 1993, p. 

2. For a complete report of the project including the full 
Requirements Specification and Prototype Design 


DP Workshop - Sirbu: Internet Billing Service 

Page 15 of 15 

Documents, contact the Information Networking Institute, 
Carnegie Mellon University, Pittsburgh, Pennsylvania 15213- 

3. Jennifer Steiner, Clifford Neuman and Jefferey Schiller. 
"Kerberos: An authentication service for open network 
systems." USENIX Winter Conference, 9-12 February 1988, 
Dallas, Texas. 

4. Richard Batelaan, etal. An Internet Billing Server: System 
Requirements. Carnegie Mellon University Information 
Networking Institute, August 1992. 

* This report represents the collective work of 13 students in 
CMLTs Master of Science Program in Information Networking 
and is derived from their final project report: Richard 
Batelaan, Richard Butler, Chun Yi Chan, Tie Ju Chen, 
Michael Evehchick, Paul Hughes, Tao Jen, John Jeng, Jon 
Millett, Michael Riccio, Ed Skoudis, Chris Starace, and Peter 
Stoddard. The project was directed by William Arms, John 
Leong and Dennis Smith of CMU. Dhananjay Gode served 
as Teaching Assistant and prepared the first draft of this 

© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


IP Workshop - Simmel/Godard: Metering and Licensing 

Pagel of 35 

Sponsores by: 


| Coalition for Networked Information 

About CNI 

Task Force 





Site Map 

Search our she 

Metering and Licensing of 
Resources: Kala's General 
Purpose Approach 

by Sergiu S. Simmel and Ivan Godard 

This paper describes the licensing and metering capabilities of 
Kala[1], a persistent data server. Kala offers a suite of low-level 
primitives for constructing both simple and sophisticated licensing 
(pay-per-user) and metering (pay-per-use) models. Kala allows 
the licensing and/or metering of access to any software facilities, 
both data (passive resources) and executable code and 
associated services (active resources). A few examples model 
concrete business needs, bridging the gap between 
technologically motivated mechanism and business motivated 


This section motivates the work on economic grounds. It also 
introduces several terms used throughout the paper. 

Components, Subassemblies and Applications 

It is natural to think of software as being an assemblage of 
components. Software structure starts with small-grain 
components such as functions, classes, and data values. These 
are grouped in subassemblies, such as a library or a subsystem. 
Successive composition finally yields what we all think of as 
applications - a conventional packaging of functionality. 

The basic distinction between components, subassemblies and 
applications is that of granularity. In electronic hardware the 
analogous entities are electronic components (such as integrated 
circuits and passive components), boards (usually ready to be 



IP Workshop - Simmel/Godard: Metering and Licensing 

Page 2 of 35 

plugged into bus connectors), and systems (such as personal 

For simplicity, we will use the term component to mean both 
components and subassemblies. We will use the term 
subassembly only if there is need to distinguish them from small- 
grain components. 

Both program components and data components exist. The former 
includes such examples as operating system subsystems, runtime 
libraries, specialized classes, etc. The latter includes font 
collections, clip-art sheets, economic indicators, and so on. 

Components, subassemblies, and applications go through an 
economic cycle which includes two mechanisms relevant to our 
discussion: the distribution mechanism and the revenue collection 

Distribution is the mechanism by which a component is made 
available to another component or final consumer (end-user). 
Several distribution techniques are conventionally used throughout 
the industry: embedding (static linking), runtime linking, and 
runtime loading. 

Revenue Collection is the mechanism by which payment for a 
component is made to reach the producing vendor, regardless of 
the context in which the component is actually used. 
Conventionally, revenue collection is done at the time the 
component is distributed. 

The Economics of Components 

Markets for software applications have been established for some 
time. There is a market for subassemblies as well, although 
substantially smaller and with much less potential for growth and 
diversification. The market for small-grain components is very 
small, although a great many are given away free or bundled with 
larger units. 

Many analysts attribute the relative lack of an open market for 
components to the tight coupling between delivery and revenue 
collection. The reasons include, among others: 

• Hard to Control Scarcity. Distribution cannot be effectively 
controlled because information is easy to duplicate. This 
leads to large percentages of fraudulent use.[2J Since the 
software industry needs to collect enough revenue to pay its 
own bills, the result is higher prices paid by a small fraction 
of users, with a secondary effect of diminishing market sizes. 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 3 of 35 

• Hard to Select. It is difficult and inconvenient to "test-drive" 
software components and applications, because users are 
asked to pay the high price of acquisition before being able 
to determine whether the software is of adequate quality or 
even needed. 

• Hard to Get Fair Revenue. It is difficult to predict actual 
distribution volumes at the time redistribution arrangements 
are made. For simplicity, many such arrangements are 
based on a flat fee. Inevitably, many end up being unfair to 
one of the parties, with negative effects on the entire 

An unprofitable component market deprives the software industry 
at large of: 

• Well Crafted Components. Components should be 
produced by the best specialists in the relevant technology. 
However, there is no incentive for them to enter this market, 
since there is little chance that their efforts would pay off. 

• Higher Quality Infrastructure. Most of the industry's focus 
is on producing applications - the only merchandise one can 
make real money on. This strong bias has negative long- 
term effects on the quality of the software infrastructure and 
the specialized components out of which the industry builds 
these applications. 

• Slower development of technology. Fewer quality 
components leads to less reuse and more reinventing the 

The natural solution to all of these problems is to decouple 
delivery and revenue collection: make the delivery mechanism 
direct, easy, and largely free, and then use a separate mechanism 
that insures collection and proper allocation of revenue. The only 
charges at distribution time should be to cover media 
manufacturing (including printed documentation, if any), and 
physical transportation, if any. These charges can also be 
eliminated in many (but not all) instances through delivery via 
high-speed networks. 

A simple solution, indeed. Why has it not happened? Reasons are 
multiple, including: 

• Cultural Barriers. We've done business in one way for over 
40 years now. The basic methods were established in an era 
where mainframe software was the predominant product. 
Additional techniques were added with the advent of 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 4 of 35 

standalone personal computers. While most of the software 
(by revenue) to be sold over the next decade will not fall into 
either of these categories, the culture that generated the 
practices is still strong. 

• Opposition by Monopolies. The large monopolies perceive 
change that may establish a fairer way of compensating 
value, emphasizing quality, and encouraging the small but 
highly skilled producer to be not in their best immediate 
interest. While one could argue that this is just a matter of 
perception, and that in fact change is in everybody's best 
long-term interest, the perceptions remain and strongly 
influence the attitude of many of the larger participants in this 

• Absence of Supporting Infrastructure. To actually effect 
such a change, a simple, secure, flexible, and general 
purpose infrastructure must exist and support the new 
manner of collecting revenue well. In its absence, the 
discussion remains academic. 

While we recognize the importance and seriousness of the first 
two barriers listed above, we are not addressing them here.[3] 
This paper is about the third barrier. Our thesis is that such a 
technological infrastructure now exists. We are presenting both a 
model and a concrete, industrial-quality, commercial 
implementation of it, as part of the Kala technology and persistent 
data server product. 

Revenue Collection 

To be useful in practice, the revenue collection mechanism must 
satisfy several requirements: 

a. Be safe. The mechanism must address safety very 
seriously, so that it seen as truly solving the revenue loss 
problem, as opposed to replacing it with another variant of it. 
Thus, the mechanism must be foolproof against major fraud. 

b. Be recursive. The mechanism must not only allow the 
collection of revenue, but also the allocation of some portion 
of it to subsidiary suppliers. This reflects the components- 

■ made-out-of-components structure of software. 

c. Be flexible. The mechanism must be just that: a 
mechanism. It should allow the component vendor the 
maximum flexibility to establish policies and reflect business 
arrangements and special cases. 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 5 of 35 

d. Be efficient. The mechanism must introduce very little, if 
any, overhead to the normal functions of the components 
and applications it supports. 

e. Be invisible. The mechanism must be largely 
transparent, and simplify customers 1 lives, rather than 
become a nuisance like dongle chains. 

There are two principal arrangements possible between the 
producer and the consumer of a component: 

• Pay-per-use. This is an arrangement whereby the consumer 
pays the producer for as much of the producer's software as 
the consumer's software actually uses. The measurement of 
use is specified as part of the arrangement. This technique is 
often referred to as metering. 

• Pay-per-user. This is an arrangement whereby the 
consumer's software is permitted to use the producer's 
software for a fixed period of time in return for a fixed fee, 
independent of whether or how much the consumer's 
software actually uses the producer's software. This 
technique is usually known as licensing, and sometimes 
referred to as pay-per-copy. 

The pay-per-user technique is certainly the dominant one in 
today's software industry, but the pay-per-use technique is hardly 
unknown. Pay-per-use is used extensively by utilities (your gas 
and electricity consumption is metered, and so is most of your 
telephone use), postal services (through the now widespread 
postal meters), city services (parking meters), music industry 
(jukeboxes), etc. 

In the information industries, pay-per-use is employed extensively 
by information database providers (e.g., reference searches, legal 
databases, medical databases), bulletin board operations, and 
consumer information providers (e.g., CompuServe, Prodigy). 

For component distribution and revenue collection purposes, we 
submit the following additional requirement: 

f. Provide Dual Technique. Both pay-per-use and pay-per- 
user arrangements must be supported; metering becomes 
the default mechanism in the absence of any pre-paid 

We motivate this requirement with an example involving end-users 
and an application. Similar examples can be constructed to 
involve software components only. 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 6 of 35 

Suppose that you are the manager of a medium-size software 
development group, say 30 people. You have 25 software 
engineers and 5 technical writers. Your platform is a network of 
Unix workstations. You must purchase an authoring system to be 
the technical writers' main tool, and also to be used occasionally 
by the software engineers. Typically, you'll be presented with 
products based on floating licenses. 

You know that your writers will use the system every day, 
intensively. You also know that your engineers will rarely use it, 
but once a month they will all want to use it simultaneously to write 
their status reports to the management. The question is: how 
many floating licenses should you buy? 

If you buy 5, then either no engineer will be able to use it, or they 
will constantly fight over licenses with the writers. This decision will 
waste you time, energy, and perhaps even people! 

If you buy 30, then everybody will be happy, except for your CFO: 
you will end up with 25 licenses sitting around unused for most of 
the time. This decision will waste you quite a bit of money! 

If you buy any number between 5 and 30, you'll still waste 
purchasing money and still have to force people to come to work 
at odd hours to abide by their licenses. 

You don't have to worry about any of the above if the application 
offers both pay-per-user and pay-per-use. You buy 5 licenses to 
satisfy the predictable, steady use by the writers. You also pre-pay 
some amount of use, and have the application run off the meter 
any time more than 5 people try to use it. If you run out of meter, 
you call your vendor with your credit card and buy more. Problem 
solved: save both headaches and money at the same time with a 
simple but flexible approach. 


This section introduces the resource management conceptual 
model of metering and licensing. We introduce several new 
notions, including resource, vendor, account, etc. We also explain 
in detail the concept of acquiring a resource, and the resource 
acquisition algorithm. 

Resources and Sub-resources 

A resource is an abstraction representing access to a software 
component, such as: 

• software subsystem (e.g., a math library, a persistence 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 7 of 35 

library, etc.), or 

• data (e.g., a font family, an encyclopedia entry, a stored 
movie, etc.), or 

• generally, an object or cluster of objects. 

The relationship between a resource and the related software 
component is such that one can only access the component via 
the associated resource. 

Accounts, Vendors and End-Users 

Each resource has an account which contains the current 
balance, measured in meter units. If the account has a negative 
balance, the resource owes units to one or more resources. If the 
account has a positive balance, the resource is owed units by one 
or more resources. 

Meter units are the currency in which all transactions between 
resources take place. Meter units are converted to cash when 
revenue is distributed to or collected from vendors. 

Each component and the corresponding resource is owned by a 
vendor. The vendor can be the manufacturer of that component, 
an agent for the manufacturer, or anyone who has acquired the 
legal right to sell access to that component. 

For accounting purposes, end-users are also represented by 
resources. Thus, each end-user's resource has an account. This 
account is debited any time the end-user uses metered resources, 
and credited any time the end-user purchases more meter units 
for cash. The end user may be a real person (based on whatever 
the local system uses for user identifier), budget pseudo-people 
(virtual users set up simply as means to implement budgeting 
classifications), or an installation as a whole if finer grain 
accounting is not required. 

Periodically payments are made to the vendors whose resource's 
accounts had positive balances. 

In summary, resources deal in meter units, while vendors (and 
users) deal in monetary currency (e.g. U.S Dollars, Deutsche 
Marks, etc.). 

The Structure of a Resource 

A software component uses other software components to 
implement its functionality. Correspondingly, a resource uses 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 8 of 35 

other resources, which become its sub-resources. In Figure 1, 
resource A uses resources X, Y, and Z, which are its sub- 

Each resource is identified by a resource identifier (or a rid). 
Resource identifiers are universally unique, so that accounting 
integrity is preserved. 

Thus, conceptually, a resource has the following components: 
• a resource identifier, 


• a set of sub-resources, 

• an account with a balance, and 

• an associated software component. 

Resources are implemented as persistent objects, subject to the 
same persistence and visibility properties as any other object in 
the Kala system. 

Licensed vs. Metered Use 

The use relationship between a resource and any of its sub- 
resources can be based either on a license or on a metered basis. 
The subsections below explore in detail each of these two 

Licensed ("pay-per-user") use 

A license is a deal between a grantee resource M and a grantor 
resource N whereby N grants to M's component access to N's 
component for a certain period of time, in return for a pre- 
negotiated license fee (see Figure 2). 


In this context, the license fee is the monetary exchange between 
the vendors that own the two resources M and N, or between an 
end-user represented by resource M and the vendor that owns 
resource N. 

A license is implemented as a persistent object in the system. The 
mere existence of a license object implies that the license fee has 


DP Workshop - Simmel/Godard: Metering and Licensing 

Page 9 of 35 

been paid. The connection between the existence of a license 
object and the actual payment for the license can be enforced by 
the software under certain circumstances, explored in more detail 
in the Section on "The Architecture". 

The license duration is the time for which a license exists. 
License durations can be expressed either as elapsed time 
(measured in days) or as a fixed date, denoting the license's 
expiration time. A perpetual license is a license whose duration is 
infinite. A temporary license is a license whose duration is finite, 
and usually shorter than the usefulness of the granting 

Thus, a license consists of: 

• Grantor. This is the rid (resource id) of the resource granting 
access to its component. 

• Grantee. This is the rid of the resource whose component 
gains access to another component. This may be 
wildcarded, i.e. "any using resource' or specified by 

• Duration. This is the time validity of the license. 

A licensed use engenders a fee regardless of whether or not the 
licensed resource (the grantor) is used or not. Through a license, 
the grantee gains access to the grantor's associated component 
without additional charge, but the grantee need not actually 
access it. 

Metered ("pay-per-use") Use 

A metered charge is an exchange between a grantee resource M 
and a grantor resource N taking place at execution time, whereby 
N grants to M's component access to N's component in return for 
a metered fee. 


As in the licensed case, there is a grantor resource and a grantee 
resource. However, the relation between them is established at 
runtime, based on two pieces of information: 

• The potential grantor's provisional charge. Each resource 
defines charges (in meter units) that it will ask for if its 
component is requested. A provisional charge is an object 
that specifies both a potential grantor and a potential 


EP Workshop - Simmel/Godard: Metering and Licensing 

Page 10 of 35 

grantee, thus allowing different charges to be applied to 
different requestors. These party specifications can be 
wildcarded. The model does not specify how the charge is 
defined, what are the terms, etc. This is a matter of policy, 
left to the users of the resource manager to negotiate and 

• The potential grantee's acceptable provisional charge. 

Each resource defines a method by which it decides whether 
it will accept a charge or not While the model allows this 
definition to be associated with a resource, it does not 
specify how this acceptance should be determined. This is a 
matter of policy, left to the users of the resource manager to 
define for their resources. 

The metered use involves a runtime provisional charge 
acceptance step. Once the potential grantee accepts the potential 
grantor's provisional charge, the access is granted. Thereafter, 
metering engenders a fee only if the granted resource is actually 
used. Upon access (use), the resource manager debits metered 
units from the grantee's account and credits them to the grantor's 

The transfer is done by the resource manager after completion of 
all use of the granted component, according an actual charge 
definition presented by the grantor resource and accepted by the 
grantee. This actual charge is checked for error and inconsistency 
against the provisional charge and acceptance is recorded during 
the resource acquisition phase prior to the grant of access. 

The Resource Graph 

For both licensed and metered use relationships between 
resources and their sub-resources, the totality of resources that 
make up an application form a directed graph, called the resource 
graph. The resource graph is not a tree because some 
components may be independently used by several different 
components that enter into the making of an application. 

For the purposes of the model detailed in this paper, we assume 
the resource graph to be static. That is, its structure is determined 
at a time prior to the execution of the application. For example, this 
could be at static linking time, or at application definition time.[4] In 
other words, for simplicity we assume that the knowledge of all 
components that could be potentially used by an application is 
present before the application is actually launched. 

The model is easily extended to allow for a dynamic resource 
graph. This provides for the general case in which the set of 

http ://\\wwxni . org/docs/m 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 11 of 35 

components potentially used by an application is not known at 
application launch time. 

The resource graph has a root node, corresponding to the 
resource associated with the application. In the example in Figure 
4, the root node is A. The set of all nodes in the graph is obtained 
by the transitive closure of the use relationship. 

The example in Figure 4 shows the resource U corresponding to 
the end-user who runs the application. It also shows B and C 
having self-pointing arrows. These arrows indicate that both B and 
C add value beyond the value acquired from their own sub- 
resources (K in C's case, none in B's case), and therefore their 
own use is not free. 


For an application to run successfully, the top level code (the 
application's main program) must gain access to all its 
components, and so on recursively. This translates into the 
resource A acquiring its sub-resources, its sub-resources 
acquiring their own sub-resources, and so forth until all resources 
in the resource graph have been acquired. 

Acquiring a Resource: The Basic Algorithm 

Acquiring a resource is a recursive process, starting from the root 
of a resource graph and working its way down until all resources 
have acquired all their sub-resources. There are three kinds of 
information that are used in the process: 

• the resource graph, 

• the existing license objects, and 

• the provisional charge and charge acceptance definitions 
associated with each resource. 

When an attempt is made to acquire a resource X, the following 
algorithm is followed: 

Step 1: For each of X ? s sub-resources Y: 
l.a Attempt to acquire Y. 

l.b If successful, then go to next sub-resource, 
if any. 

l.c If Y requests a charge, and there is a 

license from Y to X or any parent of X, then 
acquire Y with license. 

l.d If Y requests a charge, and there is no 
license for Y, then accumulate charge. 


DP Workshop - Simmel/Godard: Metering and Licensing 

Page 12 of 35 

Step 2: Add local added value to accumulated charge. 

Step 3: If X's parent has a license for X, then 
3. a Accept provisional charges from all X's 

sub-resources, and remember to pay all charges 
to X's sub-resources from X's account. 
3.b Return success. 

Step 4: If X's parent has no license for X, 

then propose to charge accumulated charge 
to parent. 

Step 5: If X's parent accepts provisional charge, 

5. a Accept provisional charges from all X's 

sub-resources, and remember to pay all charges 
to X's sub-resources from the actual charge 
received from X's parent. 

5.b Return success. 

Step 6: If X's parent refuses the provisional 
charge, then return failure. 

As the algorithm shows, negotiations take place between a 
resource and each of its sub-resources. The negotiations take 
place entirely outside the resource manager, which is totally 
unaware of the nature, methods, and means of these negotiations. 
A negotiation may be a complex dialog between the two 
resources, or may be empty (no negotiation at all). 

However determined, the result of the negotiation is 
communicated to the resource manager by both parties. The 
potential grantor communicates to the resource manager a 
provisional charge (in Step 4), while the potential grantee 
communicates a provisional acceptable charge (in Steps 3a and 

The resource is able to acquire the sub-resource if it either has a 
license for it or is ready to accept metered charges, based on the 
provisional description of the charges presented by the sub- 
resource. If neither of these happens, the resource is unable to 
acquire the sub-resource, and the entire algorithm fails, all the way 
up to the root of the resource graph. In other words, an application 
can execute if and only if it is able to acquire all resources in its 
resource graph.[5] 

When proposing or accepting a provisional charge, the resources 
inform the resource manager of the fact by supplying a pair of two 
numbers (a range). The lower number represents the provisional 
minimum charge: if the provisional charge is accepted then the 
grantor resource's account will be credited at least that many 
meter units, whether or not the grantee accepts the final actual 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 13 of 35 

charge (see Section on "Charging and Disbursing"). The 
provisional minimum charge may be zero, but not negative. 

The upper number represents the provisional maximum charge. 
It means that the grantor resource's account will be credited no 
more than that many meter units (see Section on "Charging and 
Disbursing"). If the final actual charge exceeds the provisional 
maximum charge, the resource manager considers this an error (a 
sign of potential run-away charges), credits the grantor's account 
with only the provisional minimum charge, and refuses to acquire 
the grantor resource for the grantee resource until the bug is fixed. 
The provisional maximum charge may be infinity, but not smaller 
than the provisional minimum charge. 

Charging and Disbursing 

After having successfully acquired all the resources it needs, an 
application can now execute. As it runs, some resources are 
actually used. Some resources may never be used. 

As they are used, those components that have been acquired on a 
metered basis (as opposed to a licensed basis) tally their running 
charges using their internal algorithms and internal data structures 
to hold the running tallies. At the same time, those components 
which acquired other components on a metered basis may also 
keep a tally of the charges they are expecting to eventually receive 
from those components based on the actual pattern of usage. This 
tally too is performed entirely by internal algorithms, possibly 
based upon information about the grantor resources which was 
obtained during the negotiation phase. 

At the end of the application execution, the grantor resource 
presents the accumulated actual charge tally and the grantee 
resource presents the accumulated expected charge tally (as an 
actual acceptable charge) to the resource manager, which 
compares them to each other and against the provisional charges 
specified by the grantor and provisional limits accepted by the 
grantee resource. 

If the actual charges conform to the agreement represented by the 
agreed provisional charges, the resource manager transfers the 
amount of metered units actually charged from the account of the 
grantee resource to the account of the grantor resource. The same 
process occurs for all components which were provided on a 
metered basis. These transfers are known as actual charges. 
They may include charges to the end-user's account. 

The tally by the grantee of expected actual charges is intended to 
provide a check on the veracity of the grantor beyond the rather 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 14 of 35 

broad limits of the provisional agreement. The grantee presents 
the expected actual charge to the resource manager in the form of 
a range, and so need not be exact in its calculation of the 
expected actual charges. Indeed if the grantor is trusted, the 
grantee may omit the expected tally altogether and present the 
resource manager with an expected actual charge range of zero to 
infinity, effectively taking the grantor's word for the actual charges. 


Periodically, the resource accounts must be converted into cash, 
so that the corresponding vendors can be paid cash for the use of 
their components. This can be done at specific times (such as 
whenever end-users refill their own resource accounts), or 
regularly (for example, on a quarterly basis). 

The resource accounts conversion is an activity of summarizing 
the balances of all resources in a resource manager installation, 
communicating the summary to the agency that does the 
conversion (the equivalent of the bank), the agency paying the 
corresponding vendors the equivalent amounts of cash, and finally 
resetting the paid accounts to zero, ready for the next cycle. 

The resource accounts conversion can be carried out manually or 
mechanically. A manual process involves running a batch program 
at the installation site, which will create an accounting dump file; 
sending the accounting dump file to the metering agency (the 
resource manager vendor); and finally reinitializing vendor account 
balances at the site. 

The same activity can be carried out entirely mechanically over a 
modem or other transmission line. The activity can be manually 
started or could even be started automatically (by the resource 
manager itself), thus making it largely transparent to the end 


To provide a simple example, this section traces the execution of 
the basic resource acquisition algorithm presented in Section on 
"Acquiring a Resource: The Basic Algorithm" . 

The Resource Graph 

The example involves four resources: A (the root resource, likely 
standing for the application), B, C, and K. The resource graph is 
shown in Figure 5. 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 15 of 35 


Here, resource A uses resources B and C, and resource C uses 
resource K. Resource B does not use anything else, but adds its 
own value B\ 

We also assume that A has a license for C, and that there are no 
other licenses. Let's assume that K submits a provisional charge 
with a minimum of uK meter units and a maximum of infinity (with 
the actual charge computed on usage), and that B' proposes a 
provisional charge with a minimum of uB units and a maximum of 
uB units as well (fixed fee). 

Tracing Through a Resource Acquisition 

The execution of the algorithm in Section on "Acquiring a 
Resource: The Basic Algorithm" to acquire resource A entails the 
following sequence of events: 

1 . The end-user attempts to run the application, i.e. the end 
user's resource attempts to acquire the application's 
resource A. 

2. A attempts to acquire B (cf. A's step 1 ). 

3. B has no sub-resources, but has a local charge of a 
minimum uB meter units (cf. B's step 2). 

4. B proposes the provisional charge (uB, uB) back to A (cf. B's 
step 4). This provisional charge is accumulated by A (cf. A's 
step 1d). 

5. A attempts to acquire C (cf. A's step 1a). 

6. C attempts to acquire K (cf. C's step 1a). 

7. K has no sub-resources, but has a provisional local charge 
of (uK, [[infinity]]) which it proposes to C (cf. K's step 4). 

8. C has accumulated provisional charges of {(uK, [[infinity]])} 
(cf. C's steps 1 and 2), and proposes them up to A. 


9. A has a license for C, and presents it in response to C's 
provisional charge (cf. A's step 1c). 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 16 of 35 

10. C accepts K's provisional charge, but must commit to pay 
out of its own account (cf. C's step 3a). 

1 1 . Since C accepted K's provisional charge, K returns success 
(cf. K's step 5b). 

12. C has successfully acquired all its sub-resources, and it 
returns success (cf. C's step 3b). 

13. A still has an accumulated provisional charge (from B) of 
{(uB, uB)}, and it asks the end-user resource whether it 
accepts this provisional charge (cf. A's step 4). 

14. The end-user resource accepts, either by silent pre-specified 
provisional acceptance of the charges or interactively by 
responding "Yes" to the "Accept charges?" dialog box (cf. 
A's step 5). 

1 5. A accepts B's provisional charge of (uB, uB) meter units (cf. 
A's step 5a). 

16. B receives the acceptance to its provisional charge and 
returns success (cf. B's step 5b). 

17. A has acquired all its sub-resources, and it returns success 
(cf. A's step 5b). 

18. All components of the resource graph have returned 
success, and so their respective components have been 
acquired. A proceeds with execution of its associated 

19. During execution, K internally accumulates the actual 
charges for the actual use made of its component (as 
opposed to the provisional charges used during resource 
acquisition). B need not do this as it is using flat fee 


20. At completion of execution (signaled by A) B and K present 
their final bills based on actual usage to the resource 
manager. B presents a uB* actual charge equal to the uB 
provisional. K presents a uK* actual charge, greater than or 
equal to the provisional charge uK. 

21 . The user resource, which knows that it is expecting a uB 
charge from B, presents an acceptance of a {(uB, uB)} actual 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 17 of 35 

charge to the resource manager. C, which has been keeping 
rough track of the use it has been making of K, presents an 
acceptance of a {(uK, uK)} actual charge to the Resource 
Manager, where uK is somewhat larger than the uK* 
computed by K but low enough to catch any error or 
cheating by K. 

22. After verification, the Resource Manager debits uB* units 
from the user (resource U) and credits them to B, and uK* 
units from C and credits them to K. The transaction is 

Resulting Charges 

Since the charge contracts between A and B, and between C and 
K involve metered usage, the balances of the U (end-user), B, C, 
and K accounts are modified: 

• U's balance was debited uB* meter units, to pay for the use 
of B (there was no license for B f nor was there one for any of 
its ancestors up the chain to U). 

• B's balance was credited uB* units. 

• C's balance was debited uK* meter units, to pay for the use 
of K (C had to pay this itself, because A had a license for C, 
and so C could not ask A for payment). 

• K's account was credited uK* units. 

This section provides an overview of the architecture supporting 
the main activities of the resource model. 

Three Activities 

The resource management model involves three distinct activities 
and related sets of objects: 

• The Definition Activity. This activity creates and modifies 
resource and license objects. This activity is described in 
detail in the Section on "The Definition Activity". 

• The Resource Acquisition and Charging Activity. This 
activity takes place every time an application executes for an 
end-user. This activity is described in detail in the Section on 
"The Resource Model". 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 18 of 35 

• The Accounting Activity. This activity takes place 
periodically, both at each installation and at the account 
agency site. This activity is described in the Section on "The 
Accounting Activity". 

The Definition Activity 

Definition is the activity by which resources are defined, licenses 
are installed, and meters (resource accounts) are filled with meter 
currency. While the times and frequency of the resource 
definitions differ from those of installing licenses and filling up 
meters, the notions are related and share several common 

• They all involve interactions between a customer, using one 
or more applications based on Kala's resource manager, and 
a vendor, able to provide the customer with both the 
application and related licenses and meter currency. 

Note that by "vendor" we don't necessarily mean a vendor of 
software - ultimately, it is not software that's being sold here. By 
vendor, we mean any agency that has acquired the legal right to 
sell and administer licenses and meter currency for various 
components and applications. This may be a software distributor, 
a software manufacturer, an Independent Software Vendor (ISV), 
an independent licensing and metering authority, or the 
manufacturers of the resource management software itself. 

• They all reflect cash transactions between the customer and 
the vendor. The cash exchanges don't necessarily have to 
mimic the definition actions. However, each such definition 
action is related to either a past, a concurrent, or a future 
cash exchange between the vendor and the customer. 

• As a consequence of the previous item, they all must be 
totally safe, thus offering trusted protection against fraud. 
The Kala Resource Manager insures this safety through a 
cookie exchange protocol described below. 

The Cookie Exchange Protocol 

The "Cookie Exchange Protocol" is a mechanism designed to 
insure the safe and unforgeable definition of resources, installation 
of licenses and filling of meters at the customer sites. The 
mechanism is based on the exchange of "magic numbers" 
- between vendors and customers. These magic numbers are also 
known as cookies. 

For each definition action, two cookies are involved: 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 19 of 35 

a. The Customer Cookie. This cookie is generated by the 
customer on the customer computer, and transmitted (either 
manually or mechanically) to the vendor. The customer 
generates the customer cookie using the Cookie utility 
program or a version of it embedded in the application itself. 
The customer cookie encodes information that makes it 
unique, reflecting this particular customer site (specifically, 
this particular Kala-based resource manager instance). The 
customer cookie identifies the customer uniquely throughout 
the subsequent activity. 

b. The Vendor Cookie. This cookie is generated by the 
vendor in response to a customer cookie and to information 
the vendor has about the associated cash transaction (such 
as whether the customer's check cleared or the credit card 
transaction went through). The vendor cookie encodes the 
originating customer cookie, the vendor's identity, and the 
action to be performed on the basis of customer payment. It 
is generated by the vendor, using a unique per-vendor copy 
of the NewCookie program. It is passed back to the 
customer, who uses it to perform the definition action. 

The cookie exchange protocol is safe against fraud as long as: 

• The vendor's NewCookie program is safeguarded by the 
vendor. Since the vendor has material interests in preventing 
fraud, we assume that it will take good precautionary 
measures to insure the program's safety (for example, 
installing it only on a physically protected, standalone (un- 
networked) machine. 

• The customer's site is able to generate customer cookies 
that uniquely identify the site. Satisfaction of this requirement 
is guaranteed by the mechanism Kala uses to guarantee 
universal uniqueness of identifiers [6]. 

• Valid cookies from NewCookie only work on sites which 
generated the original customer cookie. 

• Invalid cookies don't work at all. 


Since the customer cookie is not useful to anyone other than the 
vendor who generates a vendor cookie with it, and the vendor 
cookie is useful only to the originator of the customer cookie the 
vendor cookie was generated from, cookies can be transmitted 
safely over unsafe communication media, such as regular 
electronic mail or voice telephone. 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 20 of 35 

The application developer may embed the customer cookie 
generation in the application itself, so it can benefit from the 
already existing Graphical User Interface (GUI), communication 
facilities, etc. This is done by calling the Kala Cookie API function, 
which returns a datum of type cookie. The datum can then be 
either displayed to the end-user (for manual communication to the 
vendor), or embedded into a message silently sent to the vendor 
via some communication link (e.g., modem connection, electronic 
mail, etc.). 

Defining Resources 

The sub-activity of defining resources is part of the process of 
"installing" the software on the customer's computer from the 
delivery medium (e.g., diskettes, tapes, network, CDs, etc.). A 
resource is defined in the Resource Manager for each installed 

To define a resource, the installation program calls 
DefineResource, a function that is part of Kala's API. For example, 
the following defines the K in the example resource graph in the 
Section on "The Resource Graph". This resource has a resource 
identifier K, no sub-resources, and represents some given 
component. The definition is controlled by a vendor-supplied 

rid K = 

cookie magicNumberK = . . . ; 
DefineResource (K, nilRow, 

magicNumber, component, state) ; 

The following code fragment defines the C resource in the same 
example in the Section on "The Resource Graph": 

rid C = 

cookie magicNumberC = . . . ; 
DefineResource (C, Only (K, 1), 

magicNumberC, component C, 

stateC) ; 

Defining Charges and Acceptable Charges 

Once a resource is defined, one can define the provisional charge 
the resource would present to a potential grantee (another 
resource attempting to acquire this resource) using the 
DefineProvisionalCharge function. For example, the following 
defines the charge presented by the K resource defined above 
(see the section on "Defining Resources"): 

span uK = . . . ; 
DefineProvisionalCharge (K, Godard.html 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 21 of 35 

Range (uK, maxlnt) , 
magicNumberK) ; 

Note that DefineProvisionalCharge (like DefineResource) is 
controlled by a vendor cookie. The same is true for the next 
function, DefineProvisionalAcceptableCharge. 

This defines an acceptable charge that a grantee (using) resource 
can accept when charged back by any of its sub-resources: 

De fine Provisional Accept ableCharge ( 
C, K, Range (uK, maxlnt), 
magicNumberC) ; 

Installing Licenses and Refilling Meters 

Finally, an application's code can install licenses or refill meters 
using two additional Kala API functions: InstallLicense and 
Refill Account. They both require a vendor cookie to operate. 

For example, the following installs 2 one-year licenses of resource 
C to resource A: 

rid A = . . . , C = 

InstallLicense (A, C, 2, 365, 
magicNumberC) ; 

To refill U's account with 100,000 meter units, you call: 

rid U - 

cookie magicNumberU = . . . ; 

/* from meter vendor */ 
RefillAccount (U, 100000, 

magicNumberU) ; 

The Accounting Activity 

The accounting activity has three parts; the first and the last take 
place at the customer site, and the middle takes place at the 
accounting vendor's site: 

a. Balance Information Collection. This activity takes place 
periodically. It consists of (i) summarizing the account 
balances (their magnitude, not the actual units or money) for 
all resources defined at the customer site, and (ii) 
communicating this data to the accounting vendor. The 
communication can be either manual (via paper, removable 
magnetic media, etc.) or mechanized (via modem and 
telephone lines, high-speed network, etc.). 

b. Accounting. This activity takes place at the accounting 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 22 of 35 

vendor site. The balance for each resource is merged with 
the corresponding accounts for the same resource from 
other customers, and those with net positive balances trigger 
cash payments to the corresponding vendors. 

c. Balance Reinitialization. Upon successful delivery of 
balance summary information (per step a. above), the 
balance of each resource account at the customer site is 
brought to zero. 

The Resource Acquisition and Charging Activity 

This activity was described in detail in the Section on "Acquiring a 
Resource: The Basic Algorithm". Here, we need only mention two 
Kala API functions called in the process. The first is the function 
that starts the resource acquisition activity: AcquireResource. This 
function is called by the application as part of its initialization. For 
example, if the application's resource id is r, the following call 
acquires the application component on behalf of the end-user: 

p = AcquireResource (r, 

ResourceOf Client (myCid) , 1) ; 

The call above makes use of the second relevant Kala API 
function: ResourceOfClient. Given a client identifier (for example, 
myCid, the well known wildcard identifier of the calling client 
process itself), ResourceOfClient returns the resource identifier 
associated with that client. 

If resource acquisition (and all sub-acquisition) was successful, 
AcquireResource returns a pointer to the component associated 
with the acquired resource. The Resource Manage automatically 
loads the component from the persistent store. If the resource 
acquisition activity fails, the pointer is nil. Because the 
components are kept on protected and inaccessible persistent 
store by Kala, the application has no way to get at any of the 
components other than through a successful AcquireResource. 


The resource management model and its implementation as part 
of the Kala technology is useful in practice only as long as it is 
able to support useful and desired business models. A good test is 
to explore its support for a few simple ones in wide use. 

While this section only explores a few simple models, many more 
can be implemented atop Kala's resource management primitives, 
opening the doors to creative business deals that are both 
mechanically and legally enforceable, and inexpensive to 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 23 of 35 


Perpetual Floating Licensing 

Contemporary software implemented on network computers 
commonly employs perpetual floating license schemes. These 
schemes allow up to a specific number of users to use the 
application concurrently. The number can be increased by 
purchasing more licenses. Once purchased, a license never 
expires. If a site has N licenses and the (N+1 )[th] user attempts to 
access the application, she gets a message informing her that the 
system is temporarily out of licenses, and that she needs to wait 
until one of the current users logs off this application. 

Such a perpetual floating licensing scheme can be implemented 
trivially using Kala's resource management mechanism. Perpetual 
licenses are implemented as Kala licenses with infinite durations 
(represented as maxlnt values). For each application (or 
component) subject to such a licensing scheme, a resource will be 
created as part of that application's (or component's) installation. 


For example, let's assume an application that uses Kala to store 
its persistent data. The application uses no other component that 
is subject to resource management. The resource graph is shown 
in Figure 9. A is the resource associated with the application, and 
K is the resource associated with Kala itself, viewed as a software 

We assume that Kala's own installation (the coldKala program) 
installs the K resource. The application's own installation program 
contains the following code fragment: 

rid A = . ... ; 

mid midOf AnEntryPoint = . . . ; 
Def ineResource (A, 

Only (Account (K, 0) ) , 


midOf AnEntryPoint, nilMid) ; 

In the fragment above, midOfAnEntryPoint is the identifier of 
application As entry point segment, to be loaded into memory 
from the persistent store and branched to for execution upon a 
successful acquisition of As resource. 

The application has a user interface (e.g., part of its GUI) that 
provides the application administrator the means to: 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 24 of 3 5 

• generate customer cookies to request additional floating 

• add new (paid for) licenses, using vendor cookies, 
communicated either by telephone (voice), or mechanically 
via some form of connection between the application site 
and the vendor's site (e.g., modem connection over regular 
phone lines), and 

• inquire about the status of licenses and meters, such as how 
many licenses are currently installed, etc.). 

The application can run either with or without a license for Kala. In 
the former case, no further exchange takes place between A and 
K. In the latter case, the meter units consumed by application 
execution will be debited from A's account and credited to K's 
account when application execution ceases. 

A typical user site scenario is: 

a. The application administrator (likely the same person who 
administers networks, etc.) clicks on the application's menu 
item reading "Add another license". 

b. The application presents the administrator with a dialog 
box. The application administrator fills out the quantity 
desired (defaulting to 1) and the credit card number and 
expiration date (no default, for privacy reasons). Then he 
clicks on the OK button. 

c. The application silently sends electronic mail to the 
vendor, containing the above information and a locally 
generated client cookie. 

d. The vendor silently responds with a vendor cookie. 

e. The application receives the electronic mail from the 
vendor and issues a call to 

Def ineLicense (A, installation, 1, 
forever, vendorCookieA) ; 

to install the license to A. 

f. If the agreement between A's vendor and Kala's vendor 
specified that each sale of a license for A will also include a 
sale of a license for Kala (one of the many possible 
arrangements), then a license to Kala is also installed: 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 25 of 35 

Def ineLicense (K, A, 1, forever, 
vendorCookieK) ; 

Note that a perfectly valid alternative is to assume that 
licenses for Kala are obtained through a completely separate 
channel. For example, Kala may already be installed on that 
network computer in support of other applications, and 
blanket licenses (that is, licenses to Kala for anyone who 
needs them) may already exist. 

g. The application sends e-mail to the application 
administrator, notifying him that the license(s) have been 
successfully installed. 

There are many variations and extensions of the perpetual floating 
license scheme, both with respect to what is being licensed and to 
how the licenses are administered. The example above suggests 
a more mechanized, transparent and easy-to-use approach, 
involving a minimum amount of effort on the customer's side and a 
minimum amount of labor on the vendor's side. Indeed, a win-win 

Flat Annual Royalty Arrangements 

Another commonly practiced licensing scheme is that of a flat 
annual royalty. In this scheme, if an application A uses a 
component B, As vendor obtains the permission to dispense an 
unlimited number of licenses to B in return for a flat annual fee. 
Other legal limitations may occur, and some may be enforceable 
through software. However, we will ignore them here for simplicity. 
We also assume that As vendor sells its application under a 
perpetual license agreement. 

When As vendor pays B's vendor the agreed-upon annual fee, B's 
vendor gives As vendor a NewCookie module that generates 
cookies for 1-year licenses from B to A. As vendor delivers the B 
upgrade module to its own customers as part of an annual 
upgrade of the A application. The B upgrade module silently 
installs up-to-date licenses to B, so As customers can continue to 
use the embedded B without charge (they may actually not be 
even aware of the existence of B!). 

This scheme assumes an explicit module upgrade. While these 
upgrades of B-to-A licenses are not relevant to A's customers, 
they can be hidden inside other kinds of upgrades of A, such as 
annual A software releases, etc. 

If the customer does not install the upgrade, A will continue to run 
even though the local (sub)license from B to A has expired, so 
long as the B component was written to fall back to pay-per-use. 


EP Workshop - Simmel/Godard: Metering and Licensing 

Page 26 of 35 

Absent a license, B will run off a meter so long as the end-user's 
resource accepts the provisional and actual charges. This 
inconvenience can be avoided by providing the customer with 
strong enough incentives to upgrade. 

If A's vendor (who has sold perpetual licenses to its customers) 
fails to buy and deliver to its customer the necessary B module . 
upgrade (so that the customer can use B as part of A without 
charge), then the customer has the same legal complaint against 
A as for any other failure to deliver contracted upgrades. This 
presents an incentive to A's vendor to provide the necessary 

As in the example in Section on "Perpetual Floating Licensing", 
the process can be made quite smooth by employing silent 
transfer mechanisms such as electronic mail or direct connect via 
modems over phone lines. Since most sites (organizational and 
residential) are now equipped with such devices, and since the 
amount of data to be transferred is fairly modest, these options are 
now more practical than ever. 

Metered Use of Kala as a Repository 

Another simple application occurs where passive objects are 
provided under a metered arrangement. For example, a vendor of 
clip art may place an entire clip art collection on a Kala-managed 
CD-ROM (see [7], [8] for more information on Kala's persistent 
store functionality). 

The vendor distributes the collection for no or low cost (perhaps 
enough to cover manufacturing costs in part). The arrangement is 
that each clip is paid for on a usage basis: a small amount is 
charged every time a drawing program loads a clip and inserts it 
into a drawing. 

The drawing program is implemented so that it cannot be used to 
make any further copies of the clip. Kala's Resource Manager 
doesn't really enforce this if the clip is to be truly passive, i.e. 
usable by a lot of applications. The application must have an 
internal (memory) representation for the art clip, and Kala has no 
means to prevent the application from writing out the internal 
representation and reusing it. Here, the actual protection comes 
from the fact that commercial drawing package developers will 
have no incentive to allow such loopholes -- it is in their best 
interest to protect the economic interest of clip artists, so that they 
can continue to supply them with high quality clip art. 

An individual programmer may find ways to break this protection; 
however, if the cost of clips is low enough, it would be too much 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 27 of 35 

trouble for what it was worth. Pricing strategy plays an important 
role here. The vendor of a clip art library should price the product 
such that if every clip was referenced once, the total meter credit 
would be more than what he would sell an unlimited license for 

In the implementation, Kala holds clips as persistent data. Access 
to these clips is securely controlled using Kala's data visibility 
functionality. Drawings using these clips can also be held as Kala 

Each clip has a resource object associated with it. Each "handle" 
to a clip is set up to acquire the clip's associated resource. The 
model supports embedded clips as well. 

The end-user (or the site, if accounting is on a site basis) 
periodically fills up his account with some quantity of meter 
currency. The end-user resource accepts charges as long as its 
account balance is positive.[6] 

When a clip needs to be loaded into a drawing, if the proposed 
charge is accepted by the end-user resource, the actual charge is 
credited to the clip's resource account. Periodically (for example, 
on a quarterly basis or when the customer purchases more meter 
currency), the accumulated accounting information is sent to the 

Clip art may be supplied by many artists. The clip art vendor can, 
using the detailed accounting data received from the customer, 
distribute the proportional revenues to the clip artists. Thus, the 
small but highly skilled contributor can get actual revenue from his 
or her work, without placing an excessive burden on the revenue 
collection system -- a simple accounting computation. The 
underlying mechanism is very similar to the one ASCAP uses to 
convey a portion of each coin in a jukebox to the author of the 
song played. 

Free (Unlicensed) Use 

A degenerated use of the model occurs when a component is 
offered for free use to anyone who needs it. In this case, the code 
fragment that installs such a "nil license" for a resource X is: 

Def ineLicense (X, anyone, maxlnt, 
forever, aCookie) ; 

Here, a virtually unlimited number of users can use component X 



EP Workshop - Simmel/Godard: Metering and Licensing 

Page 28 of 35 


This section summarizes Kala's resource management interface. 
The interface (API) is shown as C function declarations, although 
other language interfaces will be supported. 

Resources, Relationships and Accounts 

DefineResource defines a resource and its "use" relationship to a 
suite of other resources, expressed as a row of resource. It 
associates a resource with a component. 


DefineResource (rid resource, 

rowRid subResource, 
cookie magicNumber, 
mid component, 
mid state) ; 

InstallLicense defines one or more pay-per-user (license) 
relationships between a grantor resource and a grantee resource, 
based on a cash payment. The license is for a given number of 
seats, and has a given time validity, measured in days. 


InstallLicense (rid grantor, 
rid grantee, 
span seats, 
duration validity, 
cookie magicNumber) ; 

RefillAccount credits a resource's account with a given number of 
meter units, based on a cash payment. 


RefillAccount (rid resource, 
span units, 
cookie magicNumber) ; 

Provisional and Actual Charges 

DefineProvisionalCharge states that a resource can allow 
access to its associated component (see DefineResource in the 
Section on "Resources, Relationships and Accounts") in return for 
an expected charge. A provisional charge is defined to be a range, 
indicating a minimum and a maximum (see the Section on 
"Resources and Sub-resources"). 



(rid grantor, 
rid grantee, 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 29 of 35 

urange charge, 
cookie magicNumber) ; 

DefineAcceptableProvisionalCharge states that a grantee 
resource is willing to accept an indicated provisional charge. 


Def ineAcceptableProvisionalCharge 
(rid grantor, 
rid grantee, 
urange charge, 
cookie magicNumber) ; 

Charge allows a grantor resource to present a grantee resource 
with a given actual charge, measured in meter units. 


Charge (rid grantor, 
rid grantee, 
span units) ; 

AcceptCharge allows a grantee resource to accept an actual 
charge from a grantor resource. The accepted actual charge is 
given as a min/max urange (range of unsigned integers), so that 
the match does not have to be exact. 


AcceptCharge (rid grantor, 
rid grantee, 
urange charge) ; 

Acquiring Resources 

AcquireResource acquires a resource for a potential grantee 
resource (the caller of this function). If successful, it secures 
access to the resource's component for the grantee's component. 
It can acquire the resource a quantity number of times. If 
successful, it returns a pointer to the grantor component, now in 


AcquireResource (rid resource, 
rid grantee, 
span quantity) ; 

ResourceOfClient returns a client's associated resource, 
expressed by its identifier (rid). The client is identified by its client 
unique identifier (cid). 


ResourceOfClient (cid client); 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 30 of 35 


Cookie generates a new local installation cookie, to be passed to 
the vendor, along with some request and cash payment. 


Cookie (void) ; 


CreateAccountingSummary computes the summary of accounts 
in an internal format and uses Kala facilities to create a new Kala 
persistent datum to hold the summary. The newly created datum 
is pointed to by a Kala handle located at <kin, basket>, using 
usual Kala addressing [7]. The resulting datum can thereafter be 
copied to a file, sent to the vendor site via electronic mail, or 
moved between Kala installations using regular Kala facilities. 


CreateAccountingSummary (kid kin, 

bid basket) ; 


In the Section on "Revenue Collection" we outlined a few 
requirements for a practical revenue collection mechanism 
independent of the software distribution mechanism. To conclude 
this brief overview of Kala's resource management functionality, 
we are reviewing how Kala meets these requirements. 

a. Be safe. Kala meets this requirement by a combination of 
several factors: (i) the resource model requires components 
to be accessible only via their resource objects; (ii) Kala 
provides a secure storage of data, whereby applications 
have full control over who can see what and when, via the 
data visibility primitives (see [7] for a complete discussion of 
Kala's main data visibility mechanism, the Kala Basket); and 
(iii) the dual cookie device permits rights to be 
communicated between vendors and customers without the 
need for expensive or cumbersome communication safety 

b. Be recursive. Kala's resource management model meets 
this requirement by recognizing that, with respect to 
licensing or metering, arbitrary components are no different 
from applications which are no different from passive data. 
The model explicitly calls for resource graphs. 

c. Be flexible. Kala's model does not impose any policies. It 
imposes no practical restrictions on how the business deal is 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 31 of 35 

structured, how the licenses or the meters are administered, 
what mechanism is used to communicate between vendors 
and customers, etc. As long as a policy does not violate the 
other requirements discussed here (for example, the safety 
requirement), it is implementable using Kala's resource 
management primitives. This feature sharply distinguishes 
Kala from all other license managers (commercial or 
research). The immediate effect is simplicity and interface 

d. Be efficient. Kala's model was designed to minimize the 
number of communications between the 
application/component code and the resource manager. The 
result is an implementation that brings very little overhead, 
unnoticeable in practice. 

e. Be invisible. Kala's toolkit approach makes it possible to 
fully integrate all license and meter management with the 
application (and under some schemes even hide them 

f. Provide Dual Techniques. Kala provides support for both 
the "pay-per-user" and "pay-per-use" approaches, and does 
so by linking them together, so that "pay-per-use" becomes 
the default in the absence of available licenses. 

A restricted version of this model has been part of the Kala 
persistent data server product since its 2.2a version. This model is 
part of Kala's 3.x version. Extensions to it are also expected. 


1. Kala is a Trademark of Penobscot Development Corporation. 
Portions of the technology described herein are covered by 
pending US and international patents. 

2. Many have pointed out that there is more software and 
information being used without payment than there is legitimate 

3. We are, however, encouraging everyone to explore these topics 
and devise ways to overcome these serious barriers. 

4. Application definition with respect to the component repository, 
i.e., at application installation. 

5. This restriction can be relaxed in the case of a dynamic 
resource graph. 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 32 of 35 

6. It is possible to allow account balances to become negative. 
This corresponds to extending some amount of credit to a 
customer, much as a charge account bank would do. 


[1] Cox, Brad, What if There Is A Silver Bullet, Journal of Object 
Oriented Programming, June 1992. 

[2] Hemnes, Thomas M.S., Esq., Software Revenue Generation in 
Network Environments, Ropes and Gray, Massachusetts 
Computer Software Council Annual Legal Update Program, 
November 1992. 

[3] Miller, Mark, On Software Pay-Per-Use, private conversations, 
June 1 992 and January 1 993. 

[4] Mori, Ryoichi and Maraji Kawahara, Superdistribution: An 
Overview and The Current Status, Technical Reports of the 
Institute of Electronics, Information, and Communication 
Engineers, Volume 89, Number 44. 

[5] Penobscot Development Corporation, The Kala Archives, 
available via anonymous ftp from at ~ftp/pub/kala. 
Send mail to for more information. 

[6] Penobscot Development Corporation, The Kala Forum, by free 
on-line subscription only. Send requests to kala- 
req u est@ world . std . com . 

[7] Simmel, Sergiu S. and Ivan Godard, The Kala Basket - A 
Semantic Primitive Unifying Object Transactions, Access Control, 
Versions, and Configurations, Proceedings of OOPSLA'91, 
October 1991, pp. 240-261. 

[8] Simmel, Sergiu S. and Ivan Godard, Objects of Substance, 
BYTE, December 1992, Volume 14, Number 17, pp. 167-170. 

[9] Simmel, Sergiu, Software Licensing and Metering with Kala 
Infrastructure for a New Economics of Software, Hotline on 
Object-Oriented Technology, Volume 3, Number 4, pp. 3-7. 


account - constituent of a resource, holding a balance of meter 

actual charge - the amount of meter units a grantor resource 
asks a grantee resource to pay for its actual use of the grantor's 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 33 of 35 

services during an execution 

actual acceptable charge - he amount of meter units a grantee 
resource expects or is willing to be charged by a grantor resource 
for its actual use of the grantor's services during an execution 

component - small-grain constituent of a software system 

customer cookie -- magic number generated by the customer on 
the customer computer, and transmitted (either manually or 
mechanically) to the vendor 

definition -- activity by which resources are defined, licenses are 
installed, and meters (resource accounts) are filled with meter 

distribution mechanism by which a component is made 
available to another component or final consumer (end-user) 

end-user real person (based on whatever the local system uses 
for user identifier), budget pseudo-people, or an installation as a 
whole if finer grain accounting is not required 

license - deal between a grantee resource M and a grantor 
resource N whereby N grants to M's component access to NTs 
component for a certain period of time, in return for a pre- 
negotiated license fee 

license fee -- monetary exchange between the vendors that own 
the two resources M and N, or between an end-user represented 
by resource M and the vendor that owns resource N 

license duration -- time for which a license exists 

licensing - see pay-per-user 

meter units -- currency in which all transactions between 
resources take place 

metering see pay-per-use 

pay-per-use -- arrangement whereby the consumer pays the 
producer for as much of the producer's software as the 
consumer's software actually uses 

pay-per-user -- arrangement whereby the consumer's software is 
permitted to use the producer's software for a fixed period of time 
in return for a fixed fee, independent of whether or how much the 
consumer's software actually uses the producer's software 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 34 of 35 

provisional charge - a formulation of the amount of meter units a 
grantor resource may charge if the resource is used 

provisional acceptable charge - a formulation of the amount of 
meter units a resource M would accept to be charged for the use 
of another resource N 

resource - abstraction representing access to a software 

resource accounts conversion - activity of summarizing the 
balances of all resources in a Resource Manager installation, 
communicating the summary to the agency that does the 
conversion (the equivalent of the bank), paying the corresponding 
vendors the equivalent amounts of cash, and finally resetting the 
paid accounts to zero, ready for the next cycle 

revenue collection - mechanism by which payment for a 
component is made to reach the producing vendor, regardless of 
the context in which the component is actually used 

resource graph - the graph of resources related by the user 
relationship, with the application's resource as the entry node 

rid - resource identifier 

subassembly medium-grain constituent of a software system 

vendor - manufacturer of a component, an agent for the 
manufacturer, or anyone who has acquired the legal right to sell 
access to a component 

vendor cookie -- magic number generated by the vendor in 
response to a customer cookie and to information the vendor has 
about the associated cash transaction (such as whether the 
customer's check cleared or the credit card transaction went 


Sergiu S. Simmel, President and Co-Founder of Penobscot 
Development Corporation, has been involved in the KALA project 
since 1987. He holds a Master's Degree in Computer and 
Information Sciences from University of Minnesota. He has been 
working in the computer industry as a software engineer and 
technical manager since 1981. His areas of expertise include 
CASE systems, hypermedia document management, and object 
oriented databases. 


IP Workshop - Simmel/Godard: Metering and Licensing 

Page 35 of 35 

Ivan Godard, Chief Technologist, is Kala's main designer and 
implementor, being the main force of the KALA project throughout. 
Mr. Godard has been involved in the computing industry since 
1968 as a scientist, engineer, consultant, and entrepreneur. He 
contributed to the Aldo168 Revised Report and the design of the 
Ada Language (the "green" version). His areas of expertise 
include language design, translation technologies, and object 
oriented databases. Mr. Godard has taught computer science at 
graduate and post-graduate level at several universities including 
Carnegie-Mellon University and University of Maine. 

The authors can be reached at: 

Penobscot Development Corporation 

One Kendall Square, Building 200, Suite 2200 

Cambridge, MA 02139-1564 

voice: (617) 267-KALA 

fax: (617) 859-9597 

Internet : inf or@Kala . com 

Copyright (c) 1993, Penobscot Development Corporation 

© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


IP Workshop - Kahn: Deposit, Registration and Recordation 

Pagel of 11 

Sponsored by: 


About CUi 

Task Force 





Site Map 

Search our she 


[Coalition for Networked Information 

Deposit Registration and 
Recordation in an 
Electronic Copyright 
Mangement System 

by Robert E. Kahn 


This document proposes the development of a testbed for 
deposit, registration and recordation of copyright material in 
a computer network environment. The testbed will involve 
the Library of Congress and provide for electronic deposit of 
information in any of several standard formats, automated 
submission of claims to copyright, notification of registration 
and support for on-line clearance of rights in an interactive 
network. "Digital signatures" and "privacy enhanced mail" will 
be used for registration and transfer of exclusive rights and 
other copyright related documents. Electronic mail will be 
used for licensing of non-exclusive rights with or without 
recordation. Verification and authentication of deposits can 
be carried out within the testbed using the original digital 
signatures. A system of distributed redundant "Repositories" 
is assumed to hold user deposits of electronic information. 
The testbed provides an experimental platform for concept 
development and evaluation, a working prototype for system 
implementation and a basis for subsequent deployment, if 


Deposit, registration and recordation of copyright material 
and its associated claims to rights have generally been 
handled manually. Over the past two decades, the 
economics of information technology has enabled an 
electronic foundation for such material and claims. The key 


IP Workshop - Kahn: Deposit, Registration and Recordation 

Page 2 of 11 

elements of this foundation are the personal computers, 
workstations, computer networks and peripheral devices 
such as scanners, printers and digital storage systems which 
have now become sufficiently powerful and cost effective to 
be put into widespread use. It is now essential that the 
underlying systems used to manage copyright be conformed 
to be compatible with the promise of this new computer 
networking environment. This paper addresses several 
essential steps that should now be taken to facilitate that 

In the current manual system, claims to copyright are 
registered with the Copyright Office, Library of Congress. 
Deposits are accepted and stored in physical form including 
tapes and diskettes as well as paper and other substances. 
Notification of registration is also made in physical form. In 
addition, documents transferring copyright ownership and 
other documents pertaining to copyright may be submitted to 
the Copyright Office for recordation. While an on-line record 
of recent registrations and recordations may be accessed at 
the Copyright Office, there is only limited external 
dissemination of this information in electronic form for access 
at remote sites. 

This approach requires considerable physical storage at the 
Library of Congress for deposited materials which can only 
increase over time. Materials stored in physical form will 
slowly degrade unless deposited in digital media in which 
case the contents may be reproduced subsequently without 
loss of information but at some cost for duplication. Even if it 
is available digitally, much, if not most, of this material will 
not generally be accessible on-line from any source. Rights 
to use the information in a computer network environment 
cannot usually be acquired easily or quickly, even if the 
identity of the rightsholder is accurately known. Fortunately, 
these limitations can also be overcome with the use of 
information technology and only minor modification to the 
current manual system. 


This document proposes building a testbed to develop and 
evaluate key elements of an electronic copyright 
management system. These elements include: 

a. Automated copyright registration and recordation 

b. Automated transactional framework for on-line 
clearance of rights 


IP Workshop - Kahn: Deposit, Registration and Recordation 

Page 3 of 11 

c. Privacy enhanced mail and digital signatures to 
facilitate on-line transactions 

d. Methodology for deposit, registration, recordation 
and clearance 

Current registration and recordation activities of the Library 
of Congress would be maintained and enhanced in the 
proposed testbed. It provides for repositories and recordation 
systems both within and without the Library of Congress, 
which would serve as agents for authors and other copyright 
owners which seek to register works with the library. In 
addition, the testbed provides for automated rights 
clearance, outside of but linked to the library, which would 
accelerate permissions and royalty transfers between users 
and rightsholders. 

Electronic Copyright Management Testbed 

A testbed is proposed to develop and evaluate these 
concepts and to obtain experience in the implementation and 
operation of an experimental system (see Figure 1). The 
proposed testbed consists of a Registration and Recording 
System (RRS), a Digital Library System (DLS) and a Rights 
Management System (RMS). The RRS will be operated by 
the Library of Congress and will permit automated 
registration of claims to copyright and recordation of transfer 
of ownership and other copyright related documents. The 
RRS would also provide evidence of "chain of title." The DLS 
will be a distributed system involving authors, publishers, 
database providers, users, and numerous organizations both 
public and private. It will be a repository of network 
accessible digital information and contain a powerful network 
based method of deposit, search and retrieval. The RMS will 
be an interactive distributed system that grants certain rights 
on-line and permits the selective use of copyright material on 
the network. 


Information may be stored in the DLS, located within the DLS 
and retrieved from the DLS using any of several mechanisms 
such as file transfer, electronic mail or agents such as 
Knowbot programs. Material may be imported into the DLS 
from other independent systems, from paper and other 
sources or exported from the DLS to other independent 
systems, to paper or to other materials such as CD-ROM, 
DAT, and microcassettes. The electronic copyright 
management system described in this document would be 


IP Workshop - Kahn: Deposit, Registration and Recordation 

Page 4 of 1 1 

directly linked to the DLS. 

The testbed would contain a digital storage system 
connected to an applications gateway (which is, in turn, 
connected to multiple communication systems including the 
Internet) to which documents would be submitted. The 
storage system would constitute an experimental repository 
for information. The applications gateway would be designed 
to support multiple access methods including direct login. 
The RRS and RMS would be servers connected to the 
Internet. Initially, they would be on a common machine, but 
they could later be easily separated. After development, the 
RRS would be relocated to the Library of Congress or its 
designated agent prior to being placed in operation. After 
initial implementation, the repository and the RMS would be 
replicable at other sites. 

Electronic Bibliographic Records 

An electronic bibliographic record (EBR) is created by the 
user for each digital document submission and supplied with 
the document for registration. The EBR is also suitable for 
use in cataloging and retrieval. The EBR may be supplied to 
other systems without the actual document but with a pointer 
to it. The EBR must contain a unique name for the document 
per author. If a name is provided that has already been used 
by the same author, it will be rejected with an explanation. 
An acknowledgment of deposit will be returned to the user 
along with a unique numerical identifier and a retrieval 
pointer to the document, and, in the event of a claim to 
copyright, a certificate of registration from the RRS. 

Claims Registration 

When the EBR indicates a claim to copyright, the RRS will 
be supplied a copy of the EBR by the repository along with a 
digital signature (to be described shortly) that can be used to 
verify the accuracy of a deposit at a later time. The actual 
work would remain in the repository. The digital signature 
consists of a few hundred bytes of data and is approximately 
the size of the EBR. It should allow the authenticity of the 
retrieved document to be formally established at any time for 
legal and other purposes. 


The RRS need not be collocated with a repository. It is 
expected that an operational RRS would be operated by the 
Library of Congress. The repositories would be operated by 


IP Workshop - Kahn: Deposit, Registration and Recordation 

Page 5 of 11 

the Library of Congress as well as other organizations or 
individuals. Deposits in certain qualified repositories will 
constitute deposit for public record purposes. The Library of 
Congress will maintain its own repository of selected 

Although a set of distributed repositories is envisioned for a 
widely deployed system, the proposed testbed will only have 
a single repository for experimentation. The repositories 
would be established in such a way as to insure the survival 
of the deposited information with perhaps different degrees 
of confidence (much like the treasury, banks and brokerage 
houses, for example). Certain information would probably not 
be deposited for purposes of registration and might be stored 
at the users local site or in a commercial repository. Highly 
valued information could be stored in rated repositories (5- 
star down to 1-star) with x varying degrees of backup and 
corresponding costs. The most critical information, as 
determined by Copyright Office regulations, might be stored 
at the Library of Congress or the National Archives as a 
safeguard. The structure of such a system of repositories 
should be developed as part of the project. 

The advantages of a distributed repository system are: 

1 . Large amounts of physical storage is not required to be 
made available at the Library of Congress. 

2. Access to the original documentation is guaranteed by 
the DLS to the confidence level selected by the user's 
choice of repository (again like the banks). 

3. Repositories serve as interfaces to the users, thus 
offloading and insulating any central servers and 
systems such as the RRS from potentially large user 
loadings and specialized customer service requests, 

4. Access to the RRS in transaction mode is available 
only to authorized repositories and RMSs that are 
qualified to use the RRS in that mode. An individual 
author, a collective licensing organization, a 
government or corporate entity or others may run an 
RMS. Authors and other copyright owners, as well as 
users may also connect directly to the RRS through a 
separate interactive user interface. 

The Computer Network Environment 

There are three specific actions of concern in a network 


IP Workshop - Kahn: Deposit, Registration and Recordation 

Page 6 of 11 

environment. One is the movement of information already 
contained in a computer network environment thereby 
greatly facilitating the creation of multiple copies in multiple 
machines in fractions of a second. The second is the 
importation of external information, such as print material or 
isolated CD-ROM based material, which must first be 
scanned or read into the system before it can be used. The 
third is export of internal network based information to paper 
using digital printers or facsimile machines or copied to 
separable media such as tape or DAT for external transport 
to others. Some of these actions, such as local use on paper 
in very small quantities, may or may not be covered by fair 
use provisions. However, non fair use actions would require 
approval of rightsholders. 

In addition to the above three actions, there is a fourth action 
that is facilitated by the computer network environment. 
Information in digital form has the property of being easily 
manipulated on a computer to produce derivative works. 
Such derivative works can also be easily moved about in a 
computer network environment and be subject to further 
manipulation by other parties. The technology makes it 
possible for parallel and concurrent manipulation of such 
information to result in an exponential proliferation of such 
derivative works. 

Rights Management System 

The four actions described above form a basis for a rights 
management system. In general, there will be many such 
systems operated by rightsholders or their agents for 
required permissions on either an exclusive or non-exclusive 
basis for a given type of action. To locate an RMS, a user 
requires the existence of a domain server that knows about 
the network names and addresses of all RMS servers. 
Transactions involving rights may be handled by direct 
exchange on-line between the user system and the 
corresponding RMS. Typically, this exchange would occur 
rapidly on-line, and we refer to this as the interactive 
clearance of rights. Privacy enhanced electronic mail would 
be available for exclusive licenses and other transfers of 
rights. Non-exclusive licenses might be handled by regular 
electronic mail. 

Transfer of copyright ownership would usually involve 
recordation in the RRS and could conceivably be handled 
automatically by the RMS on behalf of the rightsholder and 
the user to facilitate matters. The confirmation from the RRS 
would also be passed back to the rightsholder and user 


IP Workshop - Kahn: Deposit, Registration and Recordation 

Page 7 of 1 1 

directly or via the RMS using privacy enhanced mail. Various 
enabling mechanisms in the normal screen-based computer 
interface could be developed and invoked by a user to 
achieve rapid clearance. If included in the user interface, this 
capability would have the effect of creating an instant 
electronic marketplace for such information. 

Recordation is defined to mean the official keeping of 
records of transfers of copyright ownership and other 
documents pertaining to copyright by the Copyright Office, 
Library of Congress. For legal purposes, proof of official 
registration of claims and recordations will be provided by the 
Copyright Office (via the RRS). Other registrations (at 
repositories) and non-exclusive licenses (via RMSs) will be 
certified by privacy enhanced mail. It will be up to the parties 
to such registrations and recordations to maintain electronic 
records of their transactions. These could also be stored 
within the DLS. 

Identification Systems 

The electronic copyright management system actually 
requires several types of domain servers. First, documents 
can be easily retrieved via the DLS if the citation is 
accurately known or through one or more search and 
browsing processes otherwise. However, the mapping of a 
bibliographic pointer (to the designated repository) into its 
network name and address may require a separate server. 
Second, the above mentioned domain server for RMSs is 
needed. Third, the date and time that transactions have been 
requested and taken may need to be formally validated. An 
electronic notary and time server would provide such a 
capability as part of the privacy enhanced mail system. 

Retrieval, Appearance and Submission of Documents 

Retrieval of documents from the DLS is generally a two-step 
process. The initial step is to identify the document and to 
retrieve its EBR. This record will also identify the rightsholder 
and any terms and conditions on the use of the document or 
a pointer to a designated contact for rights and permissions. 
Rules would have to be formulated and posted to inform 
clearly what obligations a user incurs when accessing the 
system. For example, it may be specified that a submitted 
request with a valid EBR will then be taken to mean 
acceptance of the terms and conditions, including any 
implementation and usage restrictions or payment 
requirements. The rightsholder may also wish to place 
restrictions on the appearance of documents for certain 


IP Workshop - Kahn: Deposit, Registration and Recordation 

Page 8 of 11 


As part of the process of document submission, a valid EBR 
will have been produced which can be used in the author's 
system. Each author or other owner of copyright (or such 
owner's successor in title or duly authorized agent) will 
maintain his or her own collection of EBRs. Searches and 
requests will typically be made to the user's home system 
unless the rights have been transferred or delegated 
elsewhere (e.g. to a publisher, agent, or database provider). 
In applying for registration of claims to copyright at the 
Copyright Office, a user could be required to certify that he or 
she has the rights to the material and sign the submission 


This section briefly describes several key technologies to 
handle privacy and authentication in the digital network 
environment. Four such technologies are described below, 
namely: 1) Public Key Cryptography, 2) Digital Signatures, 3) 
Privacy Enhanced Mail, and 4) Notarization. 

Public Key Cryptography 

In conventional cryptography, a mathematical function and a 
"secret key" are shared by parties who wish to communicate 
confidentially. Each message to be sent is "encrypted" using 
the function and key and the recipient(s) "decrypt" it using 
the same function and key. This may be thought of as 
sharing a locked box in which several individuals have the 
key and any of them can lock or unlock the box at will. 

In the late 1970's, two Stanford University researchers, 
Martin Hellmann and Whitfield Diffie speculated that it might 
be possible to devise paired cryptographic functions which 
had the interesting property that one function would encrypt 
and the other would decrypt. In fact, the concept was slightly 
more sophisticated in that any message encrypted with 
either one of the functions could only be decrypted by the 
other. In other words, having access to the function which did 
the encrypting does not help when it is time to decrypt. Using 
the box analogy, the public key cryptography system would 
be like having a box with a two- key lock. If one of the keys is 
used to lock the box, the other must be used to unlock it. A 
person holding a key used for locking could not use it for 

One of the biggest problems with conventional cryptography 


IP Workshop - Kahn: Deposit, Registration and Recordation 

Page 9 of 11 

is that the keys must be kept secret and must be distributed 
by secure means. The notions of Hellman and Diffie opened 
up a new way of thinking about key management. One key 
could be made public (e.g. the one to be used for encryption) 
and the other kept private. Anyone knowing the public part of 
a pair of keys could use it to prepare a message which would 
remain confidential until the person knowing the private key 
used it to decrypt the message. The public keys could be 
listed in public directories without any special protection 
since knowing them did not help anyone decrypt messages 
encrypted using the public key. This feature makes it far 
simpler to manage key distribution since the public part need 
not be protected. 

Three researchers at MIT, Rivest, Shamir and Adelman 
developed a pair of functions meeting the requirements 
specified by Diffie and Hellman. These functions are now 
known as the RSA algorithms (from the last names of the 

Digital Signatures 

Since either key of a public key cryptography pair can be 
used to perform the initial encryption, an interesting effect 
can be achieved by using the secret key of the pair to 
encrypt messages to be sent. Anyone with access to the 
public key can decrypt the message and on doing so 
successfully, knows that the message must have been sent 
by the person holding the corresponding secret key. The use 
of the secret key acts like a "signature" since the decryption 
only works with the matching public key. 

Buyers could send digitally signed messages to sellers and 
the sellers could verify the identity of the sender by looking 
up the public key of the sender in a public directory and 
using it to verify the source of the message by successfully 
decrypting it. 

Privacy- Enhanced Mail (PEM) 

Public key cryptography can be combined with electronic 
mail to provide a flexible way to send confidential messages 
or digitally signed messages or both. In actual practice, a 
combination of public key, conventional secret key and 
another special function called cryptographic hashing is used 
to implement the features of privacy- enhanced mail. The 
public key algorithms require a substantial amount of 
computing power compared to conventional secret key 
algorithms. The older secret key algorithms, such as the 


IP Workshop - Kahn: Deposit, Registration and Recordation 

Page 10 of 11 

Data Encryption Standard (DES) developed by the National 
Institutes of Standards and Technology (NIST), are much 
more efficient. Consequently, confidential messages are 
typically encrypted using a conventional secret key which, 
itself, is sent, encrypted in the public key of the recipient. 
Thus, only the recipient can decrypt the conventional secret 
key and, eventually, decrypt the message. 

To send digitally- signed messages, each message is run 
through a "hashing" algorithm which produces a compressed 
residue which is then encrypted in the private key of the 
sender. The message itself is left in plain text form. The 
recipient can apply the same hashing algorithm and compare 
the compressed residue against the one that was sent (after 
decrypting it with the sender's public key). 

One of the basic problems with this application of public key 
cryptography is knowing whether the public key found in the 
directory for a given correspondent is really that 
correspondent's key or a bogus one inserted by a malicious 
person. The way this is dealt with in the Privacy- Enhanced 
Mail system is to create certificates containing the name of 
the owner of the public key and the public key itself, all of 
which are digitally signed by a. well- known issuing authority. 
The public key of the issuing authority is widely publicized so 
it is possible to determine whether a given certificate is valid. 
The actual system is more complex because it has a 
hierarchy of certificate issuers, but the principles remain the 


Using digital signatures, it is possible to establish an on- line 
notarization service which accepts messages, time- stamps 
them and digitally signs them, then returns them in that form. 
If the person desiring notarization digitally signs the message 
at the time it is sent to the notarizing service, then it will be 
possible, later, to establish that the person requesting the 
notarizing had the document/message in question at the time 
it was notarized. One can imagine that the originator of a 
message might have it notarized for the record and the 
recipient might independently do so. By this means, for 
instance, evidence of a contract's existence in the hands of 
each party at particular times might be established. 


The verification process uses stored digital signatures to 
ascertain whether a given copy is identical to the version 


IP Workshop - Kahn: Deposit, Registration and Recordation 

Page 11 of 11 

which was originally deposited. If any portion of the copy 
differs from the original, the verification process will fail. 
Authentication or formal certification of deposits may be 
provided to a requesting party in traditional ways or via 
electronic mail. Privacy enhanced mail would be used to 
certify the authenticity of a deposit, as well as to certify 
registration and recordation records, for legal purposes. 

The deployment of an electronic deposit, registration and 
recordation capability for use in a computer network 
environment would greatly facilitate and accelerate the move 
to a network base for information creation and dissemination. 
The system would be compatible with the current manual 
system and would support the ability of the Library of 
Congress to provide automated registration and recordation 
services. It would provide a foundation for straightforward 
and easy expansion and evolution and provide a direct 
linkage for the Library of Congress to the DLS. It would 
provide a prime working example for all other kinds of 
activities where claims registration and rights management 
come into play. Verification and authentication of copies of 
deposits may be performed electronically using digital 
signatures. Formal certification of deposits, as well as 
registration and recordation records, using privacy enhanced 
mail may be provided for legal purposes. A testbed which 
demonstrates the relevant concepts and ideas can be 
implemented within a two to three year period with initial 
limited use within a year. 

Robert E. Kahn, Ph.D. 

Corporation for National Research Initiatives 
Suite 100 

1895 Preston White Drive 
Reston, VA 22091-5434 


© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


IP Workshop - Tygar/Yee: Dyad 

Page 1 of 42 

Sponsorec by: 



[ Coalition for Networked Information 

About CNI 

Task Force 





Site Map 

Search our site 

Dyad: A System for Using 
Physically Secure 

by J. D. Tygar and Bennet Yee 


Physically secure coprocessors, as used in the Dyad project 
at Carnegie Mellon University, provide easily implementable 
solutions to perplexing security problems. This paper 
presents the solutions to five problems: (1) protecting the 
integrity of publicly accessible workstations; (2) tamper-proof 
accounting/audit trails; (3) copy protection; (4) electronic 
currency without centralized servers; and (5) electronic 


The Dyad project at Carnegie Mellon University is using 
physically secure coprocessors to achieve new protocols and 
systems addressing a number of perplexing security 
problems. These coprocessors can be produced as boards 
or integrated circuit chips and can be directly inserted in 
standard workstations or PC-style computers. This paper 
presents a set of security problems and easily implementable 
solutions that exploit the power of physically secure 

Standard textbook treatments of computer security assert 
that physical security is a necessary precondition to 
achieving overall system security. While meeting this 
condition may seem reasonable for yesterday's computer 
centers with their large mainframes, it is no longer so easy 
today. Most modern computer facilities consist of 
workstations within offices or of personal computers 
arranged in public access clusters, all of which are 


IP Workshop - Tygar/Yee: Dyad 

Page 2 of 42 

networked to file servers. In situations such as these where 
computation is distributed, physical security is very difficult—if 
not impossible--to realize. Neither computer clusters, nor 
offices, nor networks are secure against intruders. An even 
more difficult problem is posed by a user who may wish to 
subvert his own machine; for example, a user who wishes to 
steal a copy protected program can do so by gaining read 
access to the system memory while the system runs the 
nominally "execute only" program. By making the processing 
power of workstations widely and easily available, we've also 
made the system hardware accessible to casual interlopers. 
How do we remedy this? 

Researchers have recognized the vulnerability of network 
wires and have brought the tools from cryptography to bear 
on the problem of non-secure communication networks, and 
this has led to a variety of key exchange and authentication 
protocols [15, 16, 20, 35, 37, 45, 46, 53, 54] for use with end- 
to-end encryption to provide privacy on network 
communications. Others have noted the vulnerability of 
workstations and their disk storage to physical attacks in the 
office workstation environment, and this has led to a variety 
of secret sharing algorithms for protecting data from isolated 
attacks [24, 42, 48]. Tools from the field of consensus 
protocols can also be applied [24]. These techniques, while 
powerful, still depend on some measure of physical security. 

Cryptography allows us to slightly relax our assumptions 
about physical security; with cryptography we no longer need 
to assume that our network is physically shielded. However, 
we still need to make strong assumptions about the physical 
protection of hosts. We cannot entirely eliminate the need for 
physical security. 

All security algorithms and protocols depend on physical 
security. Cryptographic systems depend on the secrecy of 
keys, and authorization and access control mechanisms 
crucially depend on the integrity of the access control 
database. The use of physical security to provide privacy and 
integrity is the foundation upon which security mechanisms 
are built. With the proliferation of workstations to the office 
and to open computation clusters, the physical security 
assumption is no longer valid. The recent advent of powerful 
mobile computers only exacerbates this problem, since the 
machines may easily be physically removed. 

The gap between the reality of physically unprotected 
systems and this assumption of physical security must be 
closed. With traditional mainframe systems, the security 


IP Workshop - Tygar/Yee: Dyad 

Page 3 of 42 

firewall was between the users 1 terminals and the computer 
itself—the mainframe was the physically secure component in 
the system.[N1] With loosely administered, physically 
accessible workstations, the security partition can no longer 
encompass all the machines. Indeed, with most 
commercially available workstations, the best that can be 
found is a simple lock in the front panel which can be easily 
picked or bypassed-there really is no physically secure 
component in these systems. 

This paper discusses the use of physically secure 
processors to achieve new, powerful solutions to system 
security problems. (Physically secure coprocessors were first 
introduced in [6].) A secure coprocessor embodies a 
physically secure hardware module; it achieves this security 
by advanced packaging technology [62]. We focus on 
systems and protocols that can exploit the physical shielding 
to achieve novel solutions to challenging problems. There 
are many applications that need to use secure coprocessors; 
we discuss five of them here. 

1 . Consider the problem of protecting the integrity of 
publicly accessible workstations. For normal 
workstations or PCs, it is very easy to steal or modify 
data and programs on the hard disks. Operating 
system software could be modified to log keystrokes to 
extract encryption keys that you may have used to 
protect data. There is neither privacy nor integrity when 
the attacker has physical access to the machine, even 
if we prevent the attacker from adding Trojan horses to 
hardware (e.g., a modified keyboard which records 
keystrokes or a network interface board which sends 
the contents of the system memory to the attacker). 

2. The problem of providing tamper-proof audit trails and 
accounting logs is similar to that of workstation 
integrity, except that instead of protecting largely static 
data (operating system kernels and system programs), 
the goal is to make the generated logs unforgeable. For 
normal workstations or PCs, nothing prevents attackers 
from modifying system logs to erase evidence of 
intrusion. Similarly, secure system accounting is 
impossible because nothing protects the integrity of the 
accounting logs. 

3. The problem of providing copy protection for 
proprietary programs is also insoluble on traditional 
hardware. Distributing software in encrypted form does 
not help, since the user's machine must have the 


IP Workshop - Tygar/Yee: Dyad 

Page 4 of 42 

software decrypted in its memory to run it. Because we 
cannot guarantee the integrity of the machine's 
operating system, we have no assurances as to the 
privacy of this in-memory copy of the software. 

4. Another difficult problem is that of providing electronic 
currency without centralized control. Any electronic 
representation of currency is subject to duplication- 
data stored in computers can always be copied, 
regardless of how our software may choose to interpret 
them. When electronic currency no longer remains on 
trusted, centralized server machines, there is no way to 
guarantee against tampering. 

Given that we cannot trust the system software on our 
publicly accessible computers, any electronic currency 
on our machines might be arbitrarily created, 
destroyed, or sent over a network to the attacker. 
Alternatively, an untrustworthy user can record the 
state of his computer prior to "spending" his electronic 
currency, after which he simply resets the state of his 
computer to the saved state. Without a way to securely 
manage currency, attackers may "print" money at will. 
Furthermore, the attacker may take advantage of a 
partitioned network in order to use the same electronic 
currency in transactions with machines in different 
partitions. Since no communication is possible between 
these machines, users (or computers acting as service- 
providers) have no way to check for duplicity. 

5. A closely related problem arises when we want to 
provide "electronic contracts" that obligate secure 
coprocessors to perform certain actions or enforce 
certain restrictions. The notion of an electronic contract 
provides a mechanism for controlling the configuration 
of security constraints listed in the above items. Note 
that it does not suffice to merely cryptographically sign 
contracts; we must ensure that contracts are enforced 
on all machines that are parties in the contract. 

All of these problems are vulnerable to physical attacks 
which result in a loss of privacy and integrity. Software 
protection systems crucially rely on the physical security of 
the underlying hardware and are completely useless when 
the physical security assumption is violated. 

We can, however, close the assumption/reality gap in 
computer security. By adding physically secure coprocessors 
to computer systems, real, practical security systems can be 


IP Workshop - Tygar/Yee: Dyad 

PageS of 42 

built. Not only are secure coprocessors necessary and 
sufficient for security systems to be built; placing the security 
partition around a coprocessor is the natural model for 
providing security for workstations. Moreover, they are cost 
effective and can be made largely transparent to the end 

The rest of this paper outlines the theory of secure 
coprocessors. First we discuss a model for physically secure 
coprocessors and describe a number of platforms that use 
secure coprocessor technology. Then we consider several 
important security problems that are solved by using secure 
coprocessors. We next present a hierarchy of traditional and 
new approaches to physical security, and demonstrate 
naturally induced security partitions within these systems. 
We give an approach which allows secure coprocessors to 
be integrated into existing operating systems; we continue 
with a machine-user authentication section, which tackles the 
problem of verifying the presence of a secure coprocessor to 
users. Finally, we discuss previous work. For the sake of 
readers who are not computer scientists, we include a 
glossary of technical terms at the end of the paper. 


What do we mean by the term secure coprocessor? A 
secure coprocessor is a hardware module containing (1) a 
CPU, (2) ROM, and (3) NVM (non-volatile memory). This 
hardware module is physically shielded from penetration, 
and the I/O interface to this module is the only means by 
which access to the internal state of the module can be 
achieved. (Examples of packaging technology are discussed 
later in this section.) Such a hardware module can store 
cryptographic keys without risk of release. More generally, 
the CPU can perform arbitrary computations (under control 
of the operating system) and thus the hardware module, 
when added to a computer, becomes a true coprocessor. 
Often, the secure coprocessor will contain special-purpose 
hardware in addition to the CPU and memory; for example, 
high speed encryption/decryption hardware may be used. 

Secure coprocessors must be packaged so that physical 
attempts to gain access to the internal state of the 
coprocessor will result in resetting the state of the secure 
coprocessor (i.e., erasure of the NVM contents and CPU 
registers). An intruder might be able to break into a secure 
coprocessor and see how it is constructed; the intruder 
cannot, however, learn or change the internal state of the 
secure coprocessor except through normal I/O channels or 


IP Workshop - Tygar/Yee: Dyad 

Page 6 of 42 

by forcibly resetting the entire secure coprocessor. The 
guarantees about the privacy and integrity of non-volatile 
memory provide the foundations needed to build security 

Physical Assumptions for Security 

All security systems rely on a nucleus of assumptions. For 
example, it is often assumed that it is infeasible to 
successfully cryptoanalyze the encryption system used for 
security. Our basic assumption is that the coprocessor 
provides private and tamper-proof memory and processing. 
Just as attackers can exhaustively search cryptographic key 
spaces, it may be possible to falsify the physical security 
hypothesis by expending enormous resources (possibly 
feasible for very large corporations or government agencies), 
but we will assume the physical security of the system as an 
axiom. This is a physical work-factor argument, similar in 
spirit to intractability assumptions of cryptography. Our 
secure coprocessor model does not depend on the particular 
technology used to satisfy the work-factor assumption. Just 
as cryptographic schemes may be scaled to increase the 
resources required to penetrate a cryptographic system, 
current security packaging techniques may be scaled or 
different packaging techniques may be employed to increase 
the work-factor necessary to successfully bypass the 
physical security measures. 

In the section on applications, we will see examples of how 
we can build secure subsystems running partially on a 
secure coprocessor by leveraging off the physical security of 
the coprocessor. 

Limitations of Model 

Even though confining all computation within secure 
coprocessors would ideally suit our security needs, in reality 
we cannot-and should not-convert all of our processors into 
secure coprocessors. There are two main reasons: the first is 
the inherent limitations of the physical security techniques in 
packaging circuits, and the second is the need to keep the 
system maintainable. Fortunately, as we shall see in the 
section below, the entire computer need not be physically 
shielded. It suffices to physically protect only a portion of the 

Current packaging technology limits us to approximately one 
printed circuit board in size to allow for heat dissipation. 
Future developments may eventually relax this and allow us 


IP Workshop - Tygar/Yee: Dyad 

Page 7 of 42 

to make more of the solid-state components of a 
multiprocessor workstation physically secure, perhaps an 
entire card cage; the security problems of external mass 
storage and networks, however, will in all likelihood remain a 

While it may be possible to securely package an entire 
multiprocessor in a physically secure manner, it is likely to be 
impractical and is unnecessary besides. If we can obtain 
similar functionalities by placing the security concerns within 
a single coprocessor, we can avoid the cost of making all the 
processors (in multiprocessors) secure. 

Making a system easy to maintain requires a modular 
design. Once a hardware module is encapsulated in a 
physically secure package, disassembling the module to fix 
or replace some component will probably be impossible. 
Moreover, packaging considerations, as well as the extra 
hardware development time required, imply that the 
technology used within a secure coprocessor may lag slightly 
behind the technology used within the host system-perhaps 
by one generation. The right balance between physically 
shielded and unshielded components will depend on the 
class of applications for which the system is intended. For 
many applications, only a small portion of the system must 
be protected. 

Potential Platforms 

Several real instances of physically secure processing exist. 
This subsection describes some of these platforms, giving 
the types of attacks which these systems are prepared 
against, and the limitations placed on the system due to the 
approaches taken to protect against physical intrusion. 

The mABYSS [62] and Citadel [64] systems provide physical 
security by-employing board-level protection. The systems 
include an off-the-shelf microprocessor and some non- 
volatile (battery backed) memory, as well as special sensing 
circuitry which detects intrusion into a protective casing 
around the circuit board. The security circuitry erases the 
non-volatile memory before attackers can penetrate far 
enough to disable the sensors or to read the memory 
contents from the memory chips. The Citadel system 
expands on mABYSS, incorporating substantially greater 
processing power; the physical security mechanisms remain 

Physical security mechanisms must protect against many 


IP Workshop - Tygar/Yee: Dyad 

Page 8 of 42 

types of physical attacks. In the mABYSS and Citadel 
systems, it is assumed that in order for intruders to penetrate 
the system, they must be able to probe through a hole of one 
millimeter in diameter (probe pin voltages, destroy sensing 
circuitry, etc). To prevent direct intrusion, these systems 
incorporate sensors consisting of fine (40 gauge) nichrome 
wire, very low power sensing circuits, and a long life-time 
battery. The wires are loosely but densely wrapped in many 
layers about the circuit board and the entire assembly is then 
dipped in a potting material. The loose and dense wrapping 
makes the exact position of the wires in the epoxy 
unpredictable. The sensing electronics can detect open 
circuits or short circuits in the wires and erase the non- 
volatile memory if intrusion is attempted. The designs show 
that physical intrusion by mechanical means (e.g., drilling) 
cannot penetrate the epoxy without breaking one of these 

Another physical attack is the use of solvents to dissolve the 
potting material to expose the sensor wires. To block this 
kind of attack, the potting material is designed to be 
chemically "stronger" than the sensor wires. This implies that 
solvents will destroy at least one of the wires-and thus 
create an open-circuit condition-before the intruder can 
bypass the potting material and access the circuit board. 

Yet another physical attack uses low temperatures. 
Semiconductor memories retain state at very low 
temperatures even without power, so an attacker could 
freeze the secure coprocessor to disable the battery and 
then extract the memory contents at leisure. The designers 
have blocked this attack by the addition of temperature 
sensors which trigger erasure of secrets before the low 
temperature reaches the dangerous level. (The system must 
have enough thermal mass to prevent quick freezing--by 
being dipped into liquid nitrogen or helium, for example-so 
this places some limitations on the minimum size of the 

The next step in sophistication is the high-powered laser 
attack. Here, the idea is that a high powered (ultraviolet) 
laser may be able to cut through the protective potting 
material and selectively cut a run on the circuit board or 
destroy the battery before the sensing circuitry has time to 
react. To protect against such an attack, alumina or silica is 
added to the epoxy potting material which causes it to 
absorb ultraviolet light. The generated heat will cause 
mechanical stress, which will cause one or more of the 
sensing wires to break. 


IP Workshop - Tygar/Y ee: Dyad 

Page 9 of 42 

Instead of the board-level approach, physical security can be 
provided for smaller, chip-level packages. Clipper and 
Capstone, the NSA's proposed DES replacements [3, 56, 57] 
are special-purpose encryption chips. The integrated circuit 
chips are designed in such a way that key information (and 
perhaps other important encryption parameters-the 
encryption algorithm is supposed to be secret as well) are 
destroyed when attempts are made to open the integrated 
circuit chips' packaging. The types of attacks which this 
system can withstand are unknown. 

Another approach to physically secure processing appears in 
smart-cards [30], A smart-card is essentially a credit-card- 
size microcomputer which can be carried in a wallet. While 
the processor is limited by size constraints and thus is not as 
powerful as that found in board-level systems, no special 
sensing circuitry is necessary since physical security is 
maintained by the virtue of its portability. Users may carry 
their smart-cards with them at all times and can provide the 
necessary physical security. Authentication techniques for 
smart-cards have been widely studied [1, 30]. 

These platforms and their implementation parameters 
together provide the technology envelope within which 
secure coprocessor hardware will likely reside and this 
envelope will provide constraints on what class of algorithms 
is reasonable. As more computation power moves into 
mobile computers and smart-cards and better physical 
protection mechanisms are devised, this envelope will grow 
larger with time. 


Because secure coprocessors can process secrets as well 
as store them, they can do much more than just keep secrets 
confidential. We can use the ability to compute privately to 
provide many security related features, including (1) host 
integrity verification; (2) tamper-proof audit trails; (3) copy 
protection; (4) electronic currency; (5) and electronic 
contracts .[N 2] None of these are realistically possible on 
physically exposed machines. 

Host Integrity Check 

Trojan horse software dates back to the 1960s, if not earlier. 
Bogus login programs are the most common, though games 
and fake utilities are also widely used to set up back doors 
as well. Computer viruses exacerbate the problem of host 
integrity— the system may easily be inadvertently corrupted 


EP Workshop - Tygar/Yee: Dyad 

Page 10 of 42 

during normal use. 

The host integrity problem can be partially ameliorated by 
guaranteeing that all programs have been inspected and 
approved by a trusted authority, but this is at best an 
incomplete solution. With computers getting smaller and 
workstations often physically accessible in public computer 
clusters, attackers can easily bypass any logical safeguards 
to modify the disks. How can you tell if even the operating 
system kernel is correct? The integrity of the computer needs 
to be verified. The integrity of the kernel image and system 
utilities stored on disk must be verified to be unaltered since 
the last system release.[N3] 

There are two main cases to examine. The first is that of 
stand-alone workstations that are not connected to any 
networks, and the second is that of networked workstations 
with access to distributed services such as AFS [52] or 
Athena [4]. While publicly accessible stand-alone 
workstations have fewer avenues of attack, there are also 
fewer options for countering attacks. We will examine both 
cases concurrently in the following discussion. 

Using a secure coprocessor to perform the necessary 
integrity checks solves the host integrity problem. Because of 
the privacy and integrity guarantees on secure coprocessor 
memory and processing, we can use a secure coprocessor 
to check the integrity of the host's state at boot-up and have 
confidence in the results. At boot time, the secure 
coprocessor is the first to gain control of the system and can 
decide whether to allow the host CPU to continue by first 
checking the disk-resident bootstrap program, operating 
system kernel, and all system utilities for evidence of 

The cryptographic checksums of system images must be 
stored in the secure coprocessor's NVM and protected both 
against modification and (depending on the cryptographic 
checksum algorithm chosen) against exposure. Of course, 
tables of cryptographic checksums can be paged out to host 
memory or disk after first checksumming and encrypting 
them within the secure coprocessor; this can be handled as 
an extension to normal virtual memory paging. We have 
more to say on this subject in the section on system 
architecture. Since the integrity of the cryptographic 
checksums is guaranteed by the secure coprocessor, we can 
detect any modifications to the system objects and protect 
ourselves against attacks on the external storage. 



IP Workshop - Tygar/Yee: Dyad 

Page 11 of 42 

One alternative model is to eliminate external storage for 
networked workstations-to use trusted file servers and 
access a remote, distributed file system for all external 
storage. Any paging needed to implement virtual memory 
would go across the network to a trusted server with disk 

What are the difficulties with this trusted file server model? 
First, note that non-publicly readable files and virtual memory 
pages must be encrypted before being transferred over the 
network and so some hardware support is probably required 
anyway for performance reasons. A more serious problem is 
that the workstations must be able to authenticate the 
identity of the trusted file servers (the host-to-host 
authentication problem). Since workstations cannot keep 
secrets, we cannot use shared secrets to encrypt and 
authenticate data between the workstation and the file 
servers. The best that we can do is to have the file servers 
use public key cryptography to cryptographically sign the 
kernel image when we boot over the network, but we must 
be able to store the public keys of the trusted file servers 
somewhere. With exposed workstations, there's no safe 
place to store them. Attackers can always modify the public 
keys (and network addresses) of the file servers so that the 
workstation would contact a false server. Obtaining public 
keys from some external key server only pushes the problem 
one level deeper-the workstation would need to authenticate 
the identity of the key server, and attackers need only to 
modify the stored public key of the key server. 

If we page virtual memory over the network (which we 
assume is not secure), the problem only becomes worse. 
Nothing guarantees the privacy or integrity of the virtual 
memory as it is transferred over the network. If the data is 
transferred in plaintext, an attacker can simply record 
network packets to break privacy and modify/substitute 
network traffic to destroy integrity. Without the ability to keep 
secrets, encryption is useless for protecting their memory- 
attackers can obtain the encryption keys by physical means 
and destroy privacy and integrity as before. 

A second alternative model, which is a partial solution to the 
host integrity problem, is to use a secure-boot floppy 
containing system integrity verification code to bring 
machines up. Let's look at the assumptions involved here. 
First, note that we are assuming that the host hardware has 
not been compromised. If the host hardware has been 
compromised, the "secure" boot floppy can easily be ignored 
or even modified when used, whereas secure coprocessors 
cannot. The model of using a secure removable medium for 


DP Workshop - Tygar/Yee: Dyad 

Page 12 of 42 

booting assumes that untrusted users get a (new) copy of a 
master boot floppy from the trusted operators each time a 
machine is rebooted from an unknown state. Users must not 
have access to the master boot floppy since it must not be 

What problems are there? Boot floppies cannot keep 
secrets-encryption does not help, since the workstation 
must be able to decrypt them and workstations cannot keep 
secrets (encryption keys) either. The only way to assure 
integrity without completely reloading the system software is 
to check it by checking some kind of cryptographic checksum 
on the system images. 

There are a variety of cryptographic checksum functions 
available, and all obviously require that the integrity of the 
checksums for the "correct" data be maintained: when we 
check the system images on the disk of a suspect 
workstation, we must recompute new checksums and 
compare them with the original ones. This is essentially the 
same procedure used by secure coprocessors, except that 
instead of providing integrity within a piece of secure 
hardware we use trusted operators. The problem then 
becomes that of making sure that operators and users follow 
the proper security procedures. Requiring that users obtain a 
fresh copy of the integrity check software and data each time 
they need to reboot a machine is cumbersome. Furthermore, 
requiring a centralized database of all the software that 
requires integrity checks for all versions of that software on 
the various machines will be a management nightmare. Any 
centralized database is necessarily a central point of attack. 
Destroying this database will deny service to anybody who 
wishes to securely bootstrap their machine. 

Both secure coprocessors and secure boot floppies can be 
fooled by a sufficiently faithful emulation of the system which 
simulates a normal disk during integrity checks, but secure 
coprocessors allow us to employ more powerful integrity 
check techniques to provide better security. Furthermore, 
careless use (i.e., reuse) of boot floppies becomes another 
channel of attack-boot floppies can easily be made into viral 

Along with integrity, secure coprocessors offer privacy; this 
property allows the use of a wider class of cryptographic 
checksum functions. There are many cryptographic 
checksum functions that might be used, including Rivest's 
MD5 [44], Merkle's Snefru [31], Jueneman's Message 
Authentication Code (MAC) code [26], IBM's Manipulation 


IP Workshop - Tygar/Y ee: Dyad 

Page 13 of 42 

Detection Code (MDC) [25], chained DES [60], and Karp and 
Rabin's family of fingerprint functions [28]. All of these 
require integrity; the last three require privacy of keys. The 
strength of these rely on the difficulty of finding collisions- 
two different inputs with the same checksum. The 
intractability arguments for the first four of these are based 
on conjectured numbers of bit operations required to find 
collisions, and so are weak with respect to theoretical 
foundations. MDC, chained DES, and the fingerprint 
functions also keep the identity of the particular checksum 
function used secret-with MDC and DES it corresponds to 
keeping encryption keys (which select particular encryption 
functions) secret, and with fingerprint functions it 
corresponds to keeping an irreducible polynomial (which 
defines the fingerprint function) secret. DES is less well 
understood than the Karp-Rabin functions. 

The secrecy requirement of MDC, chained DES, and the 
Karp-Rabin functions is a stronger assumption which can be 
provided by a secure coprocessor and it allows us to use 
cryptographic functions with better theoretical underpinnings, 
thus improving the bounds on the security provided. Secrecy, 
however, cannot be provided by a boot floppy. The Karp- 
Rabip fingerprint functions are superior to chained DES in 
that they are much faster and much easier to implement 
(thus the implementation is less likely to contain bugs), and 
there are no proven strong lower bounds on the difficulty of 
breaking DES. 

Secure coprocessors also greatly simplify the problem of 
system upgrades. This is especially important when there 
are large numbers of machines on a network: systems can 
be securely upgraded remotely through the network. 
Furthermore, it's easy to keep the system images encrypted 
while they are being transferred over the network and while 
they are resident on secondary storage. This provides us 
with the ability to keep proprietary code protected against 
most attacks. As noted below in the section on copy 
protection, we can run (portions of) the proprietary software 
only within the secure coprocessor, allowing vendors to have 
execute-only semantics-proprietary software need never 
appear in plaintext outside of a secure coprocessor. 

The later section on operational requirements discusses the 
details of host integrity check as it relates to secure 
coprocessor architectural requirements, and the section on 
key management discusses how system upgrades would be 
handled by a secure coprocessor. Also relevant is the 
problem of how the user can know if a secure coprocessor is 
running properly in a system; our section on machine-user 


IP Workshop - Tygar/Yee: Dyad 

Page 14 of 42 

authentication discusses this. 
Audit Trails 

In order to properly perform system accounting and to 
provide data for tracing and detecting intruders on the host 
system, audit trails must be kept in a secure manner. First, 
note that the integrity of the auditing and accounting logs 
cannot be completely guaranteed (since the entire physically 
accessible machine, including the secure coprocessor, may 
be destroyed). The logs, however, can be made tamper 
evident. This is quite important for detecting intrusions- 
forging system logs to eliminate evidence of penetration is 
one of the first things that a system cracker will attempt to 
do. The privacy and integrity of the system accounting logs 
and audit trails can be guaranteed (unless the secure 
coprocessor is removed or destroyed) simply by holding 
them inside the secure coprocessor. It is awkward to have to 
keep everything inside the secure coprocessor since 
accounting and audit logs can grow very large and resources 
within the secure coprocessor are likely to be tight. 
Fortunately, it is also unnecessary. 

To provide secure logging, we use the secure coprocessor to 
seal the data against tampering with one of the cryptographic 
checksum functions discussed above; we then write the 
logging information out to the file system. The sealing 
operation must be performed within the secure coprocessor, 
since all keys used in this operation must be kept secret. By 
later verifying these cryptographic checksums we make 
tampering of log data evident, since the probability that an 
attacker can forge logging data to match the old data's 
checksums is astronomically low. This technique reduces the 
secure coprocessor storage requirement from large logs to 
the memory sufficient to store the cryptographic keys and 
checksums, typically several words per page of logged 
memory. If the space requirement for the keys and 
checksums is still too large, they can be similarly written out 
to secondary storage after being encrypted and 
checksummed by master keys. 

Additional cryptographic techniques can be used for the 
cryptographic sealing, depending on the system 
requirements. Cryptographic checksums can provide the 
basic tamper detection and are sufficient if we are only 
concerned about the integrity of the logs. If the accounting 
and auditing logs may contain sensitive information, privacy 
can be provided by using encryption. If redundancy is 
required, techniques such as secure quorum consensus [24] 


IP Workshop - Tygar/Yee: Dyad 

Page 15 of 42 

and secret sharing [48] may be used to distribute the data 
over the network to several machines without greatly 
expanding the space requirements. 

Copy Protection 

A common way of charging for software is licensing the 
software on a per-CPU, per-site, or per-use basis. Software 
licenses usually prohibit making copies for use on unlicensed 
machines. Without a secure coprocessor, this injunction 
against copying is unenforceable. If the user can execute the 
code on any physically accessible workstation, the user can 
also read that code. Even if we assume that attackers cannot 
read the workstation memory while it is running, we are 
implicitly assuming that the workstation was booted correctly- 
-verifying this property, as discussed above, requires the use 
of a secure coprocessor. 

When secure coprocessors are added to a system, however, 
we can quite easily protect executables from being copied 
and illegally utilized by attackers. The proprietary code to be 
protected-or at least some critical portion of it-can be 
distributed and stored in encrypted form, so copying it 
without obtaining the code decryption key is futile.[N4] Public 
key cryptography may be used to encrypt the entire software 
package or a key for use with a private key system such as 
DES. When a user pays for the use of a program, a digitally 
signed certificate of the public key used by his secure 
coprocessor is sent to the software vendor. This certificate is 
signed by a key management center verifying that a given 
public key corresponds to a secure coprocessor, and is 
prima facie evidence that the public key is valid. The 
corresponding private key is stored only within the NVM of 
the secure coprocessor; thus, only the secure coprocessor 
will have full access to the proprietary software. 

What if the code size is larger than the memory capacity of 
the secure coprocessor? We have two alternatives: we can 
use crypto-paging or we can split the code into protected and 
unprotected segments. 

We discuss crypto-paging in greater detail in the section on 
system architecture below, but the basic idea is to 
dynamically load the relevant sections of memory from the 
disk as needed. Since good encryption chips are fast, we 
can decrypt on the fly with little performance penalty. 
Similarly, when we run out of memory space on the 
coprocessor, we encrypt the data as we flush it out onto 
secondary storage. 



IP Workshop - Tygar/Yee: Dyad 

Page 16 of 42 

Splitting the code is an alternative to this approach. We can 
divide the code into a security-critical section and an 
unprotected section. The security-critical section is encrypted 
and runs only on the secure coprocessor. The unprotected 
section runs in parallel on the main host processor. An 
adversary can copy the unprotected section, but if the 
division is done well, he or she will not be able to run the 
code without the secure portion. 

A more primitive version of the copy protection application for 
secure coprocessors originally appeared in [63]. 

Electronic Currency 

With the ability to keep licensed proprietary software 
encrypted and allow execute access only, a natural 
application would be to allow for charging on a pay-per-use 
basis. In addition to controlling access to the software 
according to the terms of software licenses, some 
mechanism must be available to perform cost accounting, 
whether it is just keeping track of the number of times a 
program has run or keeping track of dollars in the users' 
account. More generally, this accounting software provides 
an electronic currency abstraction. Correctly implementing 
electronic currency requires that account data be protected 
against tampering— if we cannot guarantee integrity, 
attackers will be able to create electronic money at will. 
Privacy, while perhaps less important here, is a property that 
users expect for their bank balance and wallet contents; 
similarly, electronic money account balances should also be 

There are several models that can be adopted for handling 
electronic funds. The first is the cash analogy. Electronic 
funds can have similar properties to cash: (1) exchanges of 
cash can be effectively anonymous; (2) cash cannot be 
created or destroyed; and (3) cash exchanges require no 
central authority. (Note that these properties are actually 
stronger than that provided by currency-serial numbers can 
be recorded to trace transactions, and national treasuries 
regularly print and destroy money.) 

The second model is the credit cards/checks analogy. 
Electronic funds are not transferred directly; rather, promises 
of payment-perhaps cryptographically signed to prove 
authenticity-are transferred instead. A straightforward 
implementation of the credit card model fails to exhibit any of 
the three properties above. By applying cryptographic 
techniques, anonymity can be achieved [10], but the latter 


IP Workshop - Tygar/Yee: Dyad 

Page 17 of 42 

two requirements remain insurmountable. Checks must be 
signed and validated at central authorities (banks), and 
checks/credit payments en route "create" temporary money. 
Furthermore, the potential for reuse of cryptographic signed 
checks requires that the payee must be able to validate the 
check with the central authority prior to committing to a 

The third model is analogous to a rendezvous at the bank. 
This model uses a centralized authority to authenticate all 
transactions and so is even worse for large distributed 
applications. The bank is the sole arbiter of the account 
balance and can easily implement the access controls 
needed to ensure privacy and integrity of the data. This is 
essentially the model used in Electronic Funds Transfer 
(EFT) services provided by many banks-there are no access 
restrictions on deposits into accounts, so only the depositor 
for the source account needs to be authenticated. 

Let us examine these models one by one. What sort of 
properties must electronic cash have? We must be able to 
easily transfer money from one account to another. 
Electronic money must not be created or destroyed by any 
but a very few trusted users who regulate the electronic 
version of the Treasury. 

With electronic currency, integrity of the accounts data is 
crucial. We can establish a secure communication channel 
between two secure coprocessors by using a key exchange 
cryptographic protocol and thus maintain privacy when 
transferring funds. To ensure that electronic money is 
conserved (neither created nor destroyed), the transfer of 
funds should be failure atomic, i.e., the transaction must 
terminate in such a way as to either fail completely or fully 
succeed-transfer transactions cannot terminate with the 
source balance decremented without having incremented the 
destination balance or vice versa. By running a transaction 
protocol such as two-phase commit [8, 12, 65] on top of the 
secure channel, the secure coprocessors can transfer 
electronic funds from one account to another in a safe 
manner, providing privacy as well as ensuring that money is 
conserved throughout. With most transaction protocols, 
some "stable storage" for transaction logging is needed to 
enable the system to be restored to the state prior to the 
transaction when a transaction aborts. On large transaction 
systems this typically has meant mirrored disks with 
uninterruptible power supplies. With the simple transfer 
transactions here, the per-transaction log typically is not that 
large, and the log can be truncated once transactions 
commit. Because each secure coprocessor needs to handle 


IP Workshop - Tygar/Yee: Dyad 

Page 18 of 42 

only a handful of users, large amounts of stable storage 
should not be needed-because we have non-volatile 
memory in secure coprocessors, we only need to reserve 
some of this memory for logging. The log, the accounts data, 
and the controlling code are all protected from modification 
by the secure coprocessor, so account data are safe from all 
but bugs and catastrophic failures. Of course, the system 
should be designed so that users should have little or no 
incentive to destroy secure coprocessors that they can 
access-which should be natural when their own balances 
are stored on secure coprocessors, much like cash in 

Note that this type of decentralized electronic currency is not 
appropriate for smart cards unless they can be made 
physically secure from attacks by their owners. Smart cards 
are only quasi-physically-secure in that their privacy 
guarantees stem solely from their portability. Secrets may be 
stored within smart cards because their users can provide 
the physical security necessary. Malicious users, however, 
can easily violate smart card integrity and insert false data. 

If there is insufficient memory within the secure coprocessor 
to hold the account data for all its users, the code and the 
accounts database may be cryptographically paged to host 
memory or disk by first obtaining a cryptographic checksum. 
For the accounts data, encryption may also be employed 
since privacy is typically desired as well. The same 
considerations as those for checksums of system images 
apply here as well. 

This electronic currency transfer is analogous to the transfer 
of rights (not to be confused with the copying of rights) in a 
capability-based protection system. Using the electronic 
money-e.g., expended when running a pay-per-use 
program-is analogous to the revocation of a capability. 

What about the other models for handling electronic funds? 
With the credit card/check analogy, the authenticity of the 
promise of payment must be established. When the 
computer cannot keep secrets for users, there can be no 
authentication because nothing uniquely identifies users. 
Even when we assume that users can enter their passwords 
into a workstation without having the secrecy of their 
password compromised, we are still faced with the problem 
of providing privacy and integrity guarantees for network 
communication. We have similar problems as in host-to-host 
authentication in that cryptographic keys need to be 
exchanged somehow. If communication is in plaintext, 



IP Workshop - Tygar/Yee: Dyad 

Page 19 of 42 

attackers may simply record a transferral of a promise of 
payment and replay it to temporarily create cash. While 
security systems such as Kerberos [53], if properly 
implemented, can help to authenticate entities and create 
session keys, they revert to the use of a centralized server 
and have similar problems to the bank rendezvous model. 

With the bank rendezvous model, a "bank" server supervises 
the transfer of funds. While it is easy to enforce the access 
controls on account data, this suffers from problems with 
non-scalability, loss of anonymity, and easy denial of service 
from excessive centralization. 

Because every transaction must contact the bank server, 
access to the bank service will be a performance bottleneck. 
The system does not scale well to a large user base-when 
the bank system must move from running on a single 
computer to several machines, distributed transaction 
systems techniques must be brought to bear in any case, so 
this model has no real advantage over the use of secure 
coprocessors in ease of implementation. Furthermore, if the 
bank host becomes inaccessible, either maliciously or as a 
result of normal hardware failures, no agent can make use of 
any bank transfers. This model does not exhibit graceful 
degradation with system failures. 

The model of electronic currency managed on a secure 
coprocessor not only can provide the properties of (1 ) 
anonymity, (2) conservation, and (3) decentralization, but it 
also degrades gracefully when secure coprocessors fail. 
Note that secure coprocessor data may be mirrored on disk 
and backed up after being properly encrypted, and so even 
the immediately affected users of a failed secure 
coprocessor should be able to recover their balance. The 
security administrators who initialized the secure 
coprocessor software will presumably have access to the 
decryption keys for this purpose. Careful procedural security 
must be required here, both for the protection of the 
decryption key and for auditing for double spending, since 
dishonest users might attempt to back up their secure 
coprocessor data, spend electronic money, and then 
intentionally destroy their coprocessor in the hopes of using 
their electronic currency twice. The amount of redundancy 
and the frequency of backups depends on the reliability 
guarantees desired; in reliable systems secure coprocessors 
may continually run self-checks when idle and warn of 
impending failures. 

Contract Model 


IP Workshop - Tygar/Yee: Dyad 

Page 20 of 42 

Our electronic contract model is built on the following two 
secure coprocessor-provided primitive objects: (1) 
unforgeable tokens and (2) computer-enforced contracts. 

Tokens are protected objects that are conserved by the 
secure coprocessors; they are freely transferable, but they 
can be created and destroyed only by the agent that issued 
them. Tokens are useful as electronic currency and to 
represent the execute-only right to a piece of software (much 
as in capability systems). In the case of rights such as 
execute-only rights, the token provides access to 
cryptographic keys that may be used (only) within the secure 
coprocessors to run code. 

Contracts are another class of protected objects. They are 
created when two parties agree on a contract draft. 
Contracts contain binding clauses specifying actions that 
each of the parties must perform-or, in reality, actions that 
the secure coprocessors will enforce-and "method" clauses 
that may be invoked by certain parties (not necessarily 
restricted to just the parties who agreed on the contract). 
Time-based clauses and other event-based clauses may 
also exist. Contractual obligations may force the transfer of 
tokens between parties. 

Contract drafts are typically instantiated from a contract 
template. We may think of a contract template as a 
standardized contract with blanks which are filled in by the 
two parties involved, though certainly "custom" contracts are 
possible. Contract negotiation consists of an offerer sending 
a contract template along with the bindings (values with 
which to fill in blanks) to the offeree. The offeree either 
accepts or rejects the contract. If it is accepted, a contract 
instance is created whereby the contract bindings are 
permanent, and any immediate clauses are executed. If the 
draft is rejected, the offeree may take the contract template 
and re-instantiate a new draft with different bindings to create 
a counter-offer, whereupon the roles of offerer and offeree 
are reversed. 

From the time that a contract is accepted until it terminates, 
the contract is an active object running in one or more secure 
coprocessors. Methods may be invoked by users or triggered 
by external events (messages from the host, timer 
expiration). The method clauses of a contract are access 
controlled: they may be optionally invoked by only one party 
involved in the contract--or even by a third party who is under 
no contractual obligations. Contractual clauses may require 
one of the parties to accept further contracts of contractual 


IP Workshop - Tygar/Yee: Dyad 

Page 21 of 42 

obligations. Contractual clauses may require one of the 
parties to accept further contracts of certain types. One 
example of this is a requirement for some action to be 
performed prior to a certain time. Another is a contract 
between a distributor and a software house, where the 
software house requires the distributor to accept sales 
contracts from users for upgrading a piece of software. 


Network hosts, regardless of whether they use cryptography, 
have a de facto security partitioning that arises because 
different system components have different vulnerabilities to 
various attacks. Some of these vulnerabilities diminish when 
cryptography is used; similarly, the use of a secure 
coprocessor can be thought of as adding another layer with 
fewer vulnerabilities to the partitioning. By bootstrapping our 
system using a secure coprocessor and thus ensuring that 
the correct operating system is running, we can provide 
privacy and integrity guarantees on memory that were not 
possible before. In particular, public workstations can use 
secure coprocessors and cryptography to guarantee the 
privacy of disk storage and provide integrity checks. Let us 
see what kind of privacy/integrity guarantees are already 
available in the system and what new ones we can provide. 

Table 1 shows the vulnerabilities of various types of memory 
when no cryptographic techniques are used. That memory 
within a secure coprocessor is protected against physical 
access is one of our axioms, and correctly using that to 
provide privacy and integrity at the logical level is a matter of 
using the appropriate software protection mechanisms. With 
the proper protection mechanisms within a secure 
coprocessor, data stored within a secure coprocessor can be 
neither read nor tampered with. Since we assume that we 
have a working secure coprocessor, we will also assume that 
the operating system was booted correctly and thus host 
RAM is protected against unauthorized logical access. [N5] It 
is not, however, well protected against physical access-it is 
a simple matter to connect logic analyzers to the memory 
bus to listen passively to memory traffic. Furthermore, 
replacing the memory subsystem with multi-ported memory 
in order to allow remote unauthorized memory accesses is 
also a conceivable attack. While the effort required to do this 
in a way that is invisible to users may make it impractical, 
this line of attack can certainly not be entirely ruled out. 
Secondary storage may be more easily attacked than RAM 


IP Workshop - Tygar/Y ee: Dyad 

Page 22 of 42 

since the data can be modified off-line; to do this, however, 
an attacker must gain physical access to the disk. Network 
communication is completely vulnerable to on-line 
eavesdropping and off-line analysis, as well as on-line 
message tampering. Since networks are inherently used for 
remote communication, it is clear that these may be remote 

What protection guarantees can we provide when we use 
encryption? By using encryption when appropriate, we can 
guarantee privacy. Integrity of the data, however, is not 
guaranteed. The same vulnerabilities which allowed data 
modifications still exist as before; tampering, however, can 
be detected by using cryptographic checksums as long as 
the checksum values are stored in tamper-proof memory. 
Note also that the privacy that can be provided is relative to 
the data usage. If data in host RAM is to be processed by the 
host CPU, encrypting it within the secure coprocessor is 
useless-the data must remain vulnerable to on-line physical 
attacks on the host since it must appear in plaintext form to 
the host CPU. If the host RAM data is simply serving as 
backing store for secure coprocessor data pages, however, 
encryption is appropriate. Similarly, encrypting the secondary 
store via the host CPU protects that data against off-line 
privacy loss but not on-line attacks, whereas encrypting that 
data within the secure coprocessor protects that data against 
on-line privacy attacks as well, as long as that data need not 
ever appear in plaintext form in the host memory. 


For example, if we wish to send and read secure electronic 
mail, the encryption and decryption can be performed in the 
host processor since the data must reside within both hosts 
for the sender to compose it and for the receiver to read it. 
The exchange of the encryption key used for the message, 
however, requires secure coprocessor computation: the 
encryption for the key exchange needs to use secrets that 
must remain within the secure coprocessor, regardless of 
whether the key exchange uses a shared secret key or a 
public key scheme. [N6] 


This section discusses one possible architecture for a secure 
coprocessor software system. We will start off with a 
discussion of the constraints placed upon a secure 
coprocessor by the operational requirements of a security 
system-during system initialization and during normal, 


IP Workshop - Tygar/Yee: Dyad 

Page 23 of 42 

steady-state operation. We will next refine these constraints, 
examining various security functions and what their 
assumptions imply about trade-offs in a secure coprocessor. 
Following this, we will discuss the structure of the software in 
a secure coprocessor, ranging from a secure coprocessor 
kernel and its interactions with the host system to user-level 

Operational Requirements 

We will start by examining how a secure coprocessor must 
interact with the host hardware and software during the 
bootstrap process and then proceed with the kinds of system 
services that a secure coprocessor should provide to the 
host operating system and user software. The first issue to 
consider is how to fit a secure coprocessor into a system. 
This will guide us in the specification of the secure 
coprocessor software. 

To be sure that a system is bootstrapped securely, secure 
hardware must be involved in the bootstrap process. 
Depending on the host hardware-whether a secure 
coprocessor could halt the boot process if it detects an 
anomaly-we may need to assume that the bootstrap ROM is 
secure. To ensure this, the system's address space either 
could be configured such that the boot vector and the boot 
code are provided by a secure coprocessor directly or we 
may simply assume that the boot ROM itself is a piece of 
secure hardware. Regardless, a secure coprocessor verifies 
the system software (operating system kernel, system 
related user-level software) by checking the software's 
signature against known values. We need to convince 
ourselves that the version of the software present in external, 
non-secure, non-volatile store (disk) is the same as that 
installed by a trusted party. Note that this interaction has the 
same problems faced by two hosts communicating via a non- 
secure network: if an attacker can completely emulate the 
interaction that the secure coprocessor would have had with 
a normal host system, it is impossible for the secure 
coprocessor to detect this. With network communication, we 
can assume that both hosts can keep secrets and build 
protocols based upon those secrets. With secure 
coprocessor/host interaction, we can make very few 
assumptions about the host--the best that we can do is to 
assume that the cost of completely emulating the host at 
boot time is prohibitive. 

At boot time, the primary duty of a secure coprocessor is to 
make sure that the system boots up securely; after booting, a 


LP Workshop - Tygar/Yee: Dyad 

Page 24 of 42 

secure coprocessor's role is to aid the host operating system 
by providing security functions not otherwise available. A 
secure coprocessor does not enforce the system's security 
policy-that is the job of the host operating system; since we 
know from the secure boot procedure that the correct 
operating system is running, we may rely on the host to 
enforce policy. When the system is up and running, a secure 
coprocessor provides the following security services to the 
host operating system: the host may use the secure 
coprocessor to verify the integrity of any data in the same 
manner that the secure coprocessor checks the integrity of 
system software; it may use the secure coprocessor to 
encrypt data to boost the natural security of storage media 
(see section above on security partitions); and it may use the 
secure coprocessor to establish secure, encrypted 
connections with remote hosts (key exchange, 
authentication, private key encryption, etc).[N7] 

Secure Coprocessor Architecture 

The bootstrapping procedure described above made 
assumptions about the capability of a secure coprocessor. 
Let us refine what requirements we have on the secure 
coprocessor software and hardware. 

When a secure coprocessor verifies that the system software 
is the correct version, we are assuming that a secure 
coprocessor has secure, tamper-proof memory which 
remembers a description of the correct version of the system 
software. If we assume that proposed functions such as MD5 
[44], multi-round Snefru [31], or IBM's MDC [25] are one-way 
hash functions, then the only requirement is that the memory 
is protected from writing by unauthorized individuals. 
Otherwise, we must use cryptographic checksums such as 
Karp and Rabin's technique of fingerprinting, which uses a 
family of hash functions with good error-detection 
capabilities. This technique requires that the memory be 
protected against read access as well, since both the hash 
value and the index selecting the particular hash function 
must be secret. In a similar manner, cryptographic 
operations such as authentication, key exchange, and secret 
key encryption all require that secrets be kept. Thus a secure 
coprocessor must have memory that is inaccessible by 
everybody except the secure coprocessor-enough private 
NVM to store the secrets, plus possibly volatile private 
memory for intermediate calculations in running the 

There are a number of architectural tradeoffs for a secure 


IP Workshop - Tygar/Yee: Dyad 

Page 25 of 42 

coprocessor, the crucial dimensions being processor speed 
and memory size. They together determine the class of 
cryptographic algorithms that are practical. 

The speed of the secure coprocessor may be traded off for 
memory in the implementation of the cryptographic 
algorithms. We observed in [54] that Karp-Rabin 
fingerprinting may be sped up by about 25% with a 256-fold 
table-size increase. Intermediate size tables may be used to 
yield intermediate speedups at a slightly higher increase in 
code size. Similar tradeoffs can be found for software 
implementations of the DES. 

The amount of real memory required may be traded off for 
speed by employing cryptographic techniques: we need only 
enough private memory for an encryption key and a data 
cache, plus enough memory to perform the encryption if no 
encryption hardware is present. Depending on the 
throughput requirements, hardware assist for encryption may 
be included-where software is used to implement 
encryption, private memory must be provided for 
intermediate calculations. A secure coprocessor can 
securely page its private memory to either the host's physical 
memory (and perhaps eventually to an external disk) by first 
encrypting it to ensure privacy. Cryptographic checksums 
can provide error detection, and any error correcting 
encoding should be done after the encryption. This 
cryptographic paging is analogous to paging of physical 
pages to virtual memory on disk, except for different cost 
coefficients, and well-known analysis techniques can be 
used to tune such a system. The variance in costs will likely 
lead to new tradeoffs: cryptographic checksums are easier to 
calculate than encryption (and therefore faster modulo 
hardware support), so providing integrity alone is less 
expensive than providing privacy as well. On the other hand, 
if the computation can reside entirely on a secure 
coprocessor, both privacy and integrity can be provided for 

Secure Coprocessor Software 

With partitioned applications that must have parts loaded into 
a secure coprocessor to run and perhaps paging of secure 
coprocessor tasks, a small, simple security kernel is needed 
for the secure coprocessor. What makes this kernel different 
from other security kernels is the partitioned system 

Like normal workstation (host) kernels, the secure 


IP Workshop - Tygar/Yee: Dyad 

Page 26 of 42 

coprocessor kernel must provide separate address space if 
vendor and user code is to be loaded into the secure 
coprocessor-even if we implicitly trust vendor and user 
code, providing separate address spaces helps to isolate the 
effects of programming errors. Unlike the host's kernel, many 
services are not required: terminal, network, disk, and other 
device drivers need not be part of the secure coprocessor. 
Indeed, since both the network and disk drives are 
susceptible to tampering, requiring their drivers to reside in 
the secure coprocessor's kernel is overkill-network and file- 
system services from secure coprocessor tasks can simply 
be forwarded to the host kernel for processing. Normal 
operating system services such as printer service, electronic 
mail, etc. are entirely inappropriate in a secure coprocessor- 
these system daemons can be eliminated entirely. 

The only services that are crucial to the operation of the 
secure coprocessor are (1) secure coprocessor resource 
management; (2) communications; (3) key management; and 
(4) encryption services. Within resource management we 
include task allocation and scheduling, virtual memory 
allocation and paging, and allocation of communication ports. 
Under communications we include both communication 
among s.ecure coprocessor tasks and communication to host 
tasks; it is by communicating with host system tasks that 
proxy services are obtained. Under key management we 
include the management of secrets for authentication 
protocols, cryptographic keys for protecting data as well as 
execute-only software, and system fingerprints for verifying 
the integrity of system software. With the limited number of 
services needed, we can easily envision using a microkernel 
such as Mach 3.0 [22]: we need to add a communications 
server and include a key management service to manage 
non-volatile key memory. The kernel must be small for us to 
trust it; we have more confidence that it can be debugged 
and verified. 

Key Management 

A core portion of the secure coprocessor software is code to 
manage keys. Authentication, key management, fingerprints, 
and encryption crucially protect the integrity of the secure 
coprocessor software and the secrecy of private data, 
including the secure coprocessor kernel itself. A permanent 
part of a bootstrap loader, in ROM or in NVM, controls the 
bootstrap process of the secure coprocessor itself. Like 
bootstrapping the host processor, this loader verifies the 
secure coprocessor kernel before transferring control to it. 


IP Workshop - Tygar/Yee: Dyad 

Page 27 of 42 

The system fingerprints needed for checking system integrity 
must reside entirely in NVM or be protected by encryption 
while being stored on an external storage device-the key for 
which must reside solely in the secure NVM. If the latter 
approach is chosen, new keys must be selected[N8] to 
prevent replay attacks where old, potentially buggy secure 
coprocessor software is reintroduced into the system. 
Depending on the cryptographic assumptions made in the 
algorithm, the storage of the fingerprint information may 
require just integrity or both integrity and secrecy. For the 
cases of MD4, MDC, and Snefru, integrity of the integrity 
check information is sufficient; for the case of the Karp-Rabin 
fingerprint, both integrity and secrecy are required. 

Other protected data held within the secure coprocessor's 
NVM include administrative authentication information 
needed to update the secure coprocessor software. We 
assume that a security administrator is authorized to upgrade 
secure coprocessor software, and that only the administrator 
may authenticate his identity properly to the secure 
coprocessor. The authentication data for this operation can 
be updated along with the rest of the secure coprocessor 
system software; in either case, the upgrade must appear 
transactional, that is, it must have, the properties of 
permanence, where results of completed transactions are 
never lost; serializability, where there is a sequential, non- 
overlapping view of the transactions; and failure atomicity, 
where transactions either complete or fail such that any 
partial results are undone. The non-volatility of the memory 
gives us permanence automatically, if we assume that only 
catastrophic failures (or intentional sabotage) can destroy the 
NVM; serializability, while important for multi-threaded 
applications, can be easily enforced if we permit only a single 
upgrade operation to be in progress at a time (this is an 
infrequent operation and does not require parallelism); and 
the failure atomicity guarantee can be provided easily as 
long as the non-volatile memory subsystem provides an 
atomic store operation. Update transactions need not be 
distributed nor nested; this simplifies the implementation 


With secure coprocessors, we can perform all the necessary 
security functions to verify the integrity of the host system. 
The secure coprocessor may believe that the host system is 
clean, but how is the user to be convinced of this? After all, 
the secure coprocessor within the computer may have been 
replaced with a Trojan horse unit. 


IP Workshop - Tygar/Yee: Dyad 

Page 28 of 42 


One solution to this is the use of smart-cards. Users can use 
advanced smart-cards to run an authentication procedure to 
verify the secure coprocessor's identity. Since secure 
coprocessors' identity-proofs can be based on a zero- 
knowledge protocol, no secret information needs to be stored 
in smart-cards unless smart-cards are to also aid users in 
authenticating themselves to systems, in which case the only 
secrets would be those belonging to the users. By the virtue 
of their portability, users can carry smart-cards at all times 
and thus provide the physical security needed. 

Remote Services 

Another way to verify that a secure coprocessor is present is 
to ask a third-party entity-such as a physically sealed third- 
party computer--to check for the user. Often, this service can 
also be provided by normal network-servers machine such 
as file-servers. The remote services must be difficult to 
emulate by attackers. Users may rely on noticing the 
absence of these services to detect that something is amiss 
with the secure coprocessor. This necessarily implies that 
these remote services must be available before the users . 
authenticate to the system. 

Unlike authentication protocols reliant on accessing central 
authentication servers, this authentication happens once, at 
boot time. The identity being proven is that of the secure 
coprocessor-users may be confident that the workstation 
contains an authentic secure coprocessor if access to any 
normal remote service can be obtained. This is because in 
order to successfully authenticate to obtain the service, 
attackers must either break the authentication protocol, 
break the physical security in the secure coprocessor, or 
bypass the physical security around the remote server. As 
long as the remote service is sufficiently complex, attackers 
will not be able to emulate it. 


Partitioning security is not new. The method of embodying 
physical security in a secure coprocessor, however, is new, 
and it has been made possible only recently due to advances 
in packaging technology [62]. Certainly, the need for physical 
security is widely described in standard textbooks. For 
example, one book states that "physical security controls 
(locked rooms, guards, and the like) are an integral part of 
the security solution for a central computing facility."[18] 


IP Workshop - Tygar/Yee: Dyad 

Page 29 of 42 

We can trace several analogs to this approach of partitioning 
security in previous work. The logical partitioning of security 
in the literature [58, 61] of dividing the system into a "Trusted 
Computing Base" (TCB) and applications in some sense 
heralds this idea-the security partition was firmly drawn 
between the user and the machine; it not only included the 
logical security of the operating system part of the TCB, but 
also the physical security of the TCB hardware installation 
(machine rooms, etc). 

Systems such as Kerberos [53] move that security partition 
for distributed systems toward including just one trusted 
server behind locked doors. This approach, however, still 
has serious security problems: client machines are often 
physically exposed and users are provided with no real 
assurances of their logical integrity, and the centralized 
server approach offers attackers a central point of attack-the 
system catastrophically fails when the central server is 
compromised [5]. Certainly, it does not offer much in terms of 
providing fault tolerance with distributed computing. 

More recently, the partitioning in Strongbox [54] more clearly 
points the way toward minimizing the number of assumptions 
about trusted components in a secure system and clearly 
defining the security partition boundaries and security 
assumptions. In that system, the base security system was 
divided into trusted servers which, assuming protected 
address spaces, allowed security to be bootstrapped to 
application servers and clients. Unfortunately, while the 
system has better degradation properties, it could deliver 
system integrity assurances only by assuming trusted- 
operator-assisted bootstrapping. Table 3 shows the various 
types of systems and their basic assumptions as well as 
typical cryptographic assumptions. 


The secure coprocessor approach minimizes the basic 
assumptions and can address all of the problems with the 
approaches cited above. By implementing cryptographic 
protocols within a secure coprocessor, we can be assured 
that they will execute correctly and that the secrets required 
by the various protocols are indeed kept secret. By using the 
secure coprocessor to verify the integrity of the rest of the 
system, we can give users greater assurance that the 
system has not been compromised and that the system has 
securely bootstrapped. 

In addition to the work mentioned above, there are many 


IP Workshop - Tygar/Y ee: Dyad 

Page 30 of 42 

other relevant works on security related issues: [3, 56, 57, 63] 
discuss issues in the design and implementation of physically 
secure system components. Research on cryptosystems and 
cryptographic protocols which are important tools for secure 
network communication can be found in [2, 5, 7, 1 1 , 15, 16, 
17, 19, 20, 21, 24, 29, 34, 35, 37, 42, 45, 47, 48, 49, 53]. 
More general information on some of the number theoretic 
tools behind many of these protocols may be found in [33, 36, 
40, 51]. The tools for checking data integrity are described in 
[27, 28, 38,41]. 

Research on protection systems and general distributed 
system security may be found in [39, 43, 46]. 

[9] provides a logic for analyzing authentication protocols, 
and [23] extends the formalism. 

General security/cryptography information can be found in 
[14], the new proposed federal criteria for computer security 
[61], and in older standards such as the "Orange book" [58] 
and the "Red book" [59]. General information on 
cryptography can be found in [13, 32]. 


At the suggestion of the editors of this anthology, we include 
a small glossary of technical terms which may be unfamiliar 
to readers who are not computer scientists. Readers who are 
interested in operating systems may wish to read [50]; those 
who are interested in cryptography may wish to read [13, 32], 

authentication The process by which identity (or other 
credentials) of a user or computer are verified. Authentication 
protocols are series of stylized exchanges which prove 
identity. The simplest example is that of the password, which 
is a secret shared between the two parties involved; 
password-based schemes are cryptographically weak 
because any eavesdropper who overhears the password may 
subsequently impersonate one of the parties. 

More sophisticated authentication protocols use techniques 
from cryptography to eliminate the problem with 
eavesdroppers. In these systems, the password is used as a 
key to parameterize the cryptographic function. Some of 
these protocols depend on the strength of the particular 
cryptographic function involved and may leak a little bit of 
information about the keys used each time the protocol is run. 
A more powerful class of authentication protocols known as 
zero knowledge authentication probably do not leak any 

http ://www . . ip - workshop/Ty gar . Yee . html 


IP Workshop - Tygar/Yee: Dyad 

Page 31 of 42 

information. Zero knowledge protocols are an important 
special case of zero knowledge proofs, which have a number 
of important applications in computer science. 

authentication protocol See authentication. 

bootstrap The process of initializing a computer. 

checksum A small output value computed using some 
known checksum function from input data. The checksum 
function is chosen so that changes in the input will likely 
result in a different output checksum. Checksums are 
intended to guard against the corruption of the original data, 
typically due to noisy communication channels or bad 
storage media. 

checksum, cryptographic A checksum computed using a 
function where it is infeasible (other than by exhaustively 
searching through all possible input) to find another input 
which would have the same output value. Cryptographic 
checksums may be used to guard against malicious as well 
as unintentional modification of the data, since the correct 
checksum may be delivered by a (more expensive) tamper- 
proof means, and the recipient of the data may recompute 
the checksum to verify that the data has not been corrupted. 

Cryptographic checksums are generally computed by 
applying a conjectured[N9] one-way hash function to the 
input. One-way hash functions have the property that they 
are computationally infeasible (except by exhaustive search) 
to invert-that is, given only the output value, one cannot 
easily find an input to the function that would give that output. 

Another approach to cryptographic checksums is based on 
parameterizing the checksum function by a secret key. The 
Karp-Rabin fingerprint system uses secret keys to checksum 
data, forcing attackers to guess the secret key correctly; the 
secret key and resultant checksum must be delivered via a 
secure communication channel which guarantees against 
both tampering and eavesdropping. 

ciphertext See cryptography. 

client-server model A model of distributed computing where 
client software on one computer makes requests to server 
software on the same or a different computer. Examples of 
the client-server model include distributed file systems, 
electronic mail, and file transfer. 


EP Workshop - Tygar/Yee: Dyad 

Page 32 of 42 

client See client-server model. 

cryptography The enciphering and deciphering of 
messages in secret codes. The enciphered messages are 
known as ciphertext] the original (or decoded) messages are 
known as plaintext. Cryptographic functions are used to 
transform between plaintext and ciphertext, and keys are 
used to parameterize the cryptographic function used (or 
alternatively, select the cryptographic function from a set of 
functions). The security of cryptographic systems depends 
on the secrecy of the keys. Generally cryptographic systems 
may be classified into two different kinds: private key (or 
symmetric) systems; and public key (or asymmetric) 

In private key systems, a single value or key is used to 
parametrically encode and decode messages. Thus, both the 
sender and the receiver of enciphered messages must know 
the same key. In contrast, public key systems are 
parameterized by pairs of keys, one for encryption and the 
other for decryption. Knowing one of the two keys in a pair 
does not help in determining the other. Thus, the encryption 
key may be widely published while the decryption key is kept 
secret; anyone may send encrypted messages that can only 
be read by the owner of the secret key. 

Alternatively, the decryption key may be widely published 
while the encryption key is kept secret; this is used in digital 
or cryptographic signature schemes where the owner of the 
encryption key uses the secret key to encrypt, or sign a 
digital document. The result is a digital signature, which may 
be decrypted by anyone to verify that it corresponds to the 
original digital document. Since the secret key is known by 
the owner of that key, the cryptographic signature serves as 
evidence that the data has been processed by that person. 

Cryptographic systems are attacked in two main ways. The 
first is that of exhaustive search, where the attacker tries all 
possible keys in an attempt to decipher messages or derive 
the key used. The second class is that of short cut attacks 
where properties of the cryptographic system are used to 
speed up the search. 

Cryptographic attacks may be further classified by the 
amount of information available to the attacker. A ciphertext- 
only attack is one where the only information available to the 
attacker is the ciphertext. A known-plaintext attack is one 
where the attacker has available corresponding pairs of 
plaintext and ciphertext. A chosen-plaintext attack is one 


IP Workshop - Tygar/Yee: Dyad 

Page 33 of 42 

where the attacker may chose plaintext messages and 
obtain the corresponding ciphertext in an attempt to decrypt 
other messages or derive the key. A chosen-ciphertext 
attack is one where the attacker may chose some ciphertext 
messages and obtain their corresponding plaintext in an 
attempt to derive the key used. 

cryptographic checksum See checksum, cryptographic. 

cryptographic signature See cryptography. 

cryptography, private key See cryptography. 

cryptography, public key See cryptography. 

daemon a program which runs unattended, providing system 
services to users and application programs; examples 
includes electronic mail transport, printer spooling, and 
remote login service. 

Data Encryption Standard A U.S. federal data encryption 
standard adopted in 1976. The NSA recommended that the 
Federal Reserve Board use DES for electronic funds transfer 
applications. [60] 

digital signature See cryptography. 

executable binary A file containing an executable program; 
typically includes standardized headers. 

fingerprint See checksum, cryptographic. 

kernel The program which is the core of an operating 
system, providing the lowest level services upon which all 
other programs are built. There are two major schools of 
designing kernels. The traditional method of monolithic 
kernels places all basic system services within the kernel. In 
a more modular kernel design approach, microkernels 
provide many of the system services in a set of separate 
system servers rather than the kernel itself. 

known-plaintext attack See cryptography. 

microkernel See kernel. 

monolithic kernel See kernel. 

National Security Agency (NSA) The U.S. agency 
responsible for military cryptography and signals intelligence; 


IP Workshop - Tygar/Yee: Dyad 

Page 34 of 42 

recently more active in civilian cryptography. 

nonvolatile memory (NVM) Memory that retains its contents 
even after power is removed. 

one-way hash function See cryptography. 

operating system The collection of programs which provide 
the interface between the hardware and the user and 
application programs. 

paging See virtual memory. 

plaintext See cryptography. 

private key cryptography See cryptography. 

public key cryptography See cryptography. 

server See client-server model. 

virtual memory A technique of providing the appearance of 
having more primary memory than actually exists by moving 
primary memory (RAM) contents to/from secondary storage 
(disk) as needed. The transfer of sections of this virtual 
memory between physical memory and secondary storage is 
called paging. 

zero knowledge protocol See authentication. 

1 . Even greater security would be achieved if the terminals 
were also secure; otherwise the users would have the right 
to wonder whether their every keystroke is being spied upon. 

2. In recent work [55], we have also demonstrated the 
feasibility of cryptographically protecting postage franking 
marks printed possibly by PC-based electronic postage 
meters. Because such meters are located at customer sites, 
secure coprocessors were crucial for the protection of 
cryptographic keys. 

3. Sufficiently sophisticated hardware emulation can fool 
both users and any integrity checks. If an attacker replaced a 
disk controller with one which would provide the expected 
data during system integrity verification but would return 
Trojan horse data (system programs) for execution, there 
would be no completely reliable way to detect this. Similarly, Yee.html 


EP Workshop - Tygar/Yee: Dyad 

Page 35 of 42 

it would be very difficult to detect if the CPU were substituted 
with one which fails to correctly run specific pieces of code in 
the operating system protection system. One limited defense 
against hardware modifications is to have the secure 
coprocessor do behavior and timing checks at random 
intervals. There is no absolute defense against this form of 
attack, however, and the best that we can do is to make such 
emulation difficult and force the hardware hackers to build 
more perfect Trojan horse hardware. 

4. Allowing the encrypted form of the code to be copied 
means that we can back up the workstation against disk 
failures. Even giving attackers access to the backup tapes 
will not release any of the proprietary code. Note that our 
encryption function should be resistant to known-plaintext 
attacks, since executable binaries typically have 
standardized formats. 

5. We can assume that the operating system provides 
protected address spaces. Paging is assumed to be 
performed on either a local disk which is immune to all but 
physical attacks or a remote disk via encrypted network 
communication (see section below on coprocessor 
architecture). If we wish to protect against physical attacks 
for the former case, we may need to encrypt the data 
anyway or ensure that we can erase the paging data from 
the disk prior to shutting down. 

6. The public key encryption requires no secrets and may be 
performed in the host; signing the message, however, 
requires the use of secret values and thus must be 
performed within the secure coprocessor. 

7. Presumably the remote hosts will also contain a secure 
coprocessor, though everything will work fine as long as the 
remote hosts follow the appropriate protcols. The final design 
must take into consideration the possibility of remote hosts 
without secure coprocessors. 

8. One way is to use a cryptographically secure random 
number generator, the state of which resides entirely in 

9. Theorists have not been able to prove any particular 
function to be one-way. However, there are several functions 
that seem to work in practice. 



IP Workshop - Tygar/Yee: Dyad 

Page 36 of 42 

[I] M. Abadi, M. Burrows, C. Kaufman, and B. Lampson. 
Authentication and delegation with smart-cards. Technical 
Report 67, DEC Systems Research Center, October 1990. 

[2] W. Alexi, B. Chor, O. Goldreich, and C. P. Schnorr. RSA 
and Rabin functions: Certain parts are as hard as the whole. 
SIAM Journal on Computing, 17(2): 194-209, April 1988. 

[3] R. G. Andersen. The destiny of DES. Datamation, 33(5), 
March 1987. 

[4] E. Balkovich, S. R. Lerman, and R. P. Parmelee. 
Computing in higher education: The Athena experience. 
Communications of the ACM, 28(1 1):1 21 4-1 224, November 

[5] S. M. Bellovin and M. Merritt. Limitations of the Kerberos 
authentication system. Submitted to Computer 
Communication Review, 1990. 

[6] Robert M. Best. Preventing software piracy with crypto- 
microprocessors. In Proceedings of IEEE Spring COMPCON 
80, page 466, February 1980. 

[7] Manuel Blum and Silvio Micali. How to generate 
cryptographically strong sequences of pseudo-random bits. 
SIAM Journal on Computing, 1 3(4):850--864, November 

[8] Andrea J. Borr. Transaction monitoring in Encompass 
(TM): Reliable distributed transaction processing. In 
Proceedings of the Very Large Database Conference, pages 
155-165, September 1981. 

[9] Michael Burrows, Martin Abadi, and Roger Needham. A 
logic of authentication. In Proceedings of the Twelfth ACM 
Symposium on Operation Systems Principles, 1989. 

[10] David Chaum. Security without identification: 
Transaction systems to make big brother obsolete. 
Communications of the ACM, 28(1 0):1 030-1 044, October 

[II] Ben-Zion Chor. Two Issues in Public Key Cryptography: 
RSA Bit Security and a New Knapsack Type System. ACM 
Distinguished Dissertations. MIT Press, Cambridge, 

[12] C. J. Date. An Introduction to Database Systems 


IP Workshop - Tygar/Yee: Dyad 

Page 37 of 42 

Volume 2. The System Programming Series. Addison- 
Wesley, Reading, MA, 1983. 

[13] Donald Watts Davies and W. L. Price. Security for 
Computer Networks: An Introduction to Data Security in 
Teleprocessing and Electronic Funds Transfer, 2nd Edition. 
Wiley, 1989. 

[14] Dorothy Denning. Cryptography and Data Security. 
Addison-Wesley, Reading, MA, 1982. 

[15] W. Diffie and M. E. Hellrnan. New directions in 
cryptography. IEEE Transactions on Information Theory, IT- 
26(6):644--654, November 1976. 

[16] Uriel Feige, Amos Fiat, and Adi Shamir. Zero knowledge 
proofs of identity. In Proceedings of the 19th ACM Symp. on 
Theory of Computing, pages 210-217, May 1987. 

[17] U. Fiege and A. Shamir. Witness indistinguishable and 
witness hiding protocols. In Proceedings of the 22nd ACM 
Symp. on Theory of Computing, pages 416-426, May 1990. 

[18] Morrie Gasser. Building a Secure Computer System. 
Van Nostrand Reinhold Co, New York, 1988. 

[19] S. Goldwasser and M. Sipser. Arthur Merlin games 
versus zero interactive proof systems. In Proceedings of the 
17th ACM Symp. on Theory of Computing, pages 59-68, 
May 1985. 

[20] Shafi Goldwasser and Silvio Micali. Probabilistic 
encryption and how to play mental poker keeping secret all 
partial information. In Proceedings of the Fourteenth Annual 
ACM Symposium on Theory of Computing, 1982. 

[21] Shafi Goldwasser, Silvio Micali, and Charles Rackoff. 
The knowledge complexity of interactive proof systems. In 
Proceedings of the Seventeenth Annual ACM Symposium on 
Theory of Computing, May 1985. 

[22] David Golub, Randall Dean, Alessandro Forin, and 
Richard Rashid. Unix as an Application Program. In 
Proceedings of the Summer 1990 USENIX Conference, 
pages 87-95, June 1990. 

[23] Nevin Heintze and J. D. Tygar. A critique of Burrows 1 , 
Abadi's, and Needham's a logic of authentication. To Appear. 


IP Workshop - Tygar/Yee: Dyad 

Page 38 of 42 

[24] Maurice P. Heriihy and J. D. Tygar. How to make 
replicated data secure. In Advances in Cryptology, CRYPTO- 
87. Springer- Verlag, August 1987. To appear in Journal of 

[25] IBM Corporation. Common Cryptographic Architecture: 
Cryptographic Application Programming Interface Reference, 
sc40-1 675-1 edition. 

[26] R. R. Jueneman, S. M. Matyas, and C. H. Meyer. 
Message authentication codes. IEEE Communications 
Magazine, 23(9):29--40, September 1985. 

[27] Richard M. Karp. 1985 Turing award lecture: 
Combinatorics, complexity, and randomness. 
Communications of the ACM, 29(2):98»109, February 1986. 

[28] Richard M. Karp and Michael O. Rabin. Efficient 
randomized pattern-matching algorithms. Technical Report 
TR-3 1-81, Aiken Laboratory, Harvard University, December 

[29] Michael Luby and Charles Rackoff. Pseudo-random 
permutation generators and cryptographic composition. In 
Proceedings of the 18th ACM Symp. on Theory of 
Computing, pages 356-363, May 1986. 

[30] J. McCrindle. Smart Cards. Springer Verlag, 1990. 

[31] R. Merkle. A software one-way function. Technical 
report, Xerox PARC, March 1990. 

[32] C. Meyer and S. Matyas. Cryptography. Wiley, 1982. 

[33] G. L Miller. Riemann's hypothesis and a test for 
primality. Journal of Computing and Systems Science, 
13:300-317, 1976. 

[34] R. M. Needham. Using cryptography for authentication. 
In Sape Mullender, editor, Distributed Systems. ACM Press 
and Addison-Wesley Publishing Company, New York, 1989. 

[35] Roger M. Needham and Michael D. Schroeder. Using 
encryption for authentication in large networks of computers. 
Communications of the ACM, 21(12):993-999, December 
1978. Also Xerox Research Report, CSL-78-4, Xerox 
Research Center, Palo Alto, CA. 

[36] I. Niven and H. S. Zuckerman. An Introduction to the^ 


IP Workshop - Tygar/Y ee: Dyad 

Page 39 of 42 

Theory of Numbers. Wiley, 1960. 

[37] Michael Rabin. Digitized signatures and public-key 
functions as intractable as factorization. Technical Report 
MIT/LCS/TR-21 2, Laboratory for Computer Science, 
Massachusetts Institute of Technology, January 1979. 

[38] Michael Rabin. Fingerprinting by random polynomials. 
Technical Report TR-81-15, Center for Research in 
Computing Technology, Aiken Laboratory, Harvard 
University, May 1981. 

[39] Michael Rabin and J. D. Tygar. An integrated toolkit for 
operating system security (revised version). Technical 
Report TR-05-87R, Center for Research in Computing 
Technology, Aiken Laboratory, Harvard University, August 

[40] Michael O. Rabin. Probabilistic algorithm for testing 
primality. Journal of Number Theory, 12:128-138, 1980. 

[41] Michael O. Rabin. Probabilistic algorithms in finite fields. 
SI AM Journal on Computing, 9:273-280, 1980. 

[42] Michael O. Rabin. Efficient dispersal of information for 
security and fault tolerance. Technical Report TR-02-87, 
Aiken Laboratory, Harvard University, April 1987. 

[43] B. Randell and J. Dobson. Reliability and security issues 
in distributed computing systems. In Proceedings of the Fifth 
IEEE Symposium on Reliability in Distributed Software and 
Database Systems, pages 113-118, January 1985. 

[44] R. Rivest and S. Dusse. The MD5 message-digest 
algorithm. Manuscript, July 1991. 

[45] R. Rivest, A. Shamir, and L. Adleman. A method for 
obtaining digital signatures and public-key cryptosystems. 
Communications of the ACM, 21(2):120-126, February 

[46] M. Satyanarayanan. Integrating security in a large 
distributed environment. ACM Transactions on Computer 
Systems, 7(3):247-280, August 1989. 

[47] A. W. Schrift and A. Shamir. The discrete log is very 
discreet. In Proceedings of the 22nd ACM Symp. on Theory 
of Computing, pages 405-415, May 1990. 


IP Workshop - Tygar/Y ee: Dyad 

Page 40 of 42 

[48] A. Shamir. How to share a secret. Communications of ' 
the ACM, 22(11):612--614, November 1979. 

[49] Adi Shamir and Eli Biham. Differential cryptanalysis of 
DES-like cryptosystems. In Advances in Cryptology, 
CRYPTO-90. Springer- Verlag, August 1990. 

[50] Abraham Silberschatz, James L. Peterson, and Peter B. 
Galvin. Operating System Concepts, 3rd Edition. Addison- 
Wesley, 1991. 

[51] R. Solovay and V. Strassen. A fast Monte-Carlo test for 
primality. SIAM Journal on Computing, 6:84-85, March 

[52] Alfred Z. Spector and Michael L. Kazar. Wide area file 
service and the AFS experimental system. Unix Review, 7 
(3), March 1989. 

[53] J. G. Steiner, C. Neuman, and J. I. Schiller. Kerberos: 
An authentication service for open network systems. In 
USENIX Conference Proceedings, pages 191-200, Winter 

[54] J. D. Tygar and Bennet S. Yee. Strongbox: A system for 
self securing programs. In Richard F. Rashid, editor, CMU 
Computer Science: 25th Anniversary Commemorative. 
Addison-Wesley, 1991. 

[55] J. D. Tygar and Bennet S. Yee. Cryptography: It's not 
just for electronic mail anymore. Technical Report CMU-CS- 
93-107, Carnegie Mellon University, March 1993. 

[56] U. S. National Institute of Standards and Technology. 
Capstone chip technology press release, April 1993. 

[57] U. S. National Institute of Standards and Technology. 
Clipper chip technology press release, April 1993. 

[58] U.S. Department of Defense, Computer Security Center. 
Trusted computer system evaluation criteria, December 

[59] U.S. Department of Defense, Computer Security Center. 
Trusted network interpretation, July 1987. 

[60] U.S. National Bureau of Standards. Federal information 
processing standards publication 46: Data encryption 
standard, January 1977. 


IP Workshop - Tygar/Yee: Dyad 

Page 41 of 42 

[61] U.S. National Institute of Standards and Technology and 
National Security Agency. Federal criteria for information 
technology security, December 1992. Draft. 

[62] Steve H. Weingart. Physical security for the mABYSS 
system. In Proceedings of the IEEE Computer Society 
Conference on Security and Privacy, pages 52-58, 1987. 

[63] Steve R. White and Liam Comerford. ABYSS: A trusted 
architecture for software protection. In Proceedings of the 
IEEE Computer Society Conference on Security and Privacy, 
pages 38-51, 1987. 

[64] Steve R. White, Steve H. Weingart, William C. Arnold, 
and Elaine R. Palmer. Introduction to the Citadel 
Architecture: Security in Physically Exposed Environments. 
Technical Report RC16672, Distributed Security Systems 
Group, IBM Thomas J. Watson Research Center, March 
1991. Version 1.3. 

[65] Jeannette Wing, Maurice Herlihy, Stewart Clamen, 
David Detlefs, Karen Kietzke, Richard Lerner, and Su-Yuen 
Ling. The Avalon language: A tutorial introduction. In Jeffery 
L. Eppinger, Lily B. Mummert, and Alfred Z. Spector, editors, 
Camelot and Avalon: A Distributed Transaction Facility. 
Morgan Kaufmann, 1991. 


Dr. J. Douglas Tygar is Associate Professor of Computer 
Science, Carnegie Mellon University, where he is active in 
computer security and cryptography research, and directs 
the Dyad and StrongBox projects. 


Pittsburgh PA 15213-3891 
Telephone: (412) 268-6340 

Bennet Yee is a doctoral candidate in the School of 
Computer Science at Carnegie Mellon University. He 
received a B.S. in Mathematics and a B.S. in Computer 
Engineering from Oregon State University in 1986. His 
primary research interest is in computer security, and his 
current research centers on Dyad, a project with Doug Tygar 
that explores the use of physically secure coprocessors to 
solve otherwise intractable security problems. 


Pittsburgh PA 15213-3891 


IP Workshop - Tygar/Yee: Dyad 

Page 42 of 42 

Telephone: (412) 268-7571 

Partial support [for the Dyad project] was provided by the 
Avionics Laboratory, Wright Research and Development 
Center, Aeronautical Systems Division (AFSC), U.S. Air 
Force, Wright-Patterson AFB, OH 45433-6543 under 
Contract F33615-90-C-1465, ARPA Order No. 7597. 
Additional support was provided in part under a Presidential 
Young nvestigator Award, Contract No. CCR-8858087 and 
matching funds from Motorola Inc. and TRW. Additional 
partial support was provided by a contract from the U.S. 
Postal Service. We gratefully acknowledge the generous 
support of IBM through equipment grants and loans. 

The views and conclusions contained in this document are 
those of the authors and should not be interpreted as 
representing the official policies, either expressed or implied, 
of the US Government. 


© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


IP Workshop - Graham: Intellectual Preservation 

Page 1 of 19 

Sponsorec by: 


Coalition for Networked Information 

About CNI 

Task Force 





Site Map 

Search our site 

Intellectual Preservation and 
Electronic Intellectual Property 

by Peter S. Graham 


Preserving intellectual property means protecting it from easy chang 
in electronic form. Change can be accidental, well-intended or 
fraudulent; protection must be for terms longer than human lifetimes 
Three possible solutions for authenticating electronic texts are 
described: encryption (least useful), hashing, and (with the most 
potential) digital time-stamping, which can fix document existence at 
point in time using public techniques. 


This conference is concerned with means of protecting intellectual 
property in the networked environment. This paper will focus on the 
authenticity of electronic information content, that is, on intellectual 
preservation. [1J 

The concern with authentication arises from the concerns of 
librarianship, which has the imperatives of identifying information on 
behalf of users and of providing it to them, intact, when they need it. 
The professional paradigm librarians speak of is that they acquire 
information, organize it, make it available and preserve it. The 
paradigm is appropriate for electronic information just as for print ov< 
the last several centuries. 

For printed texts preservation of the work has meant preservation of 
the artifact that contains the work. Indeed, for most people there has 
been no distinction between the book and the text, though the more 
sophisticated analytical bibliographers and librarians have discussec 
that distinction for some decades. But now, in the electronic 
environment, the work (which may be a text or may be graphic, 
numeric or multimedia information) can migrate from medium to 
medium and has no necessary residence on any one of them. The 


IP Workshop - Graham: Intellectual Preservation 

Page 2 of 19 

preservation of the work independent of its medium takes on 
importance in its own right. 

Librarians have as their professional responsibility the serving up of 
the information placed in their custody as true to its original 
intellectual content as they can. This conference's concern is with 
protection of intellectual property, a related concern. Such protectior 
must extend not only to intellectual rights over the property, but to th 
property itself: how can we preserve information content from 
unauthorized, intentional or accidental change? The exercise of 
property rights includes purchase and sale. Both the buyer and selle 
have an interest in the property being what it is said to be, that is, in 
authenticating the property or text. Authentication is an interest of 
librarians as well. 

Barry Neavill, a professor at the library school at the University of 
Alabama, wrote presciently almost ten years ago that no one had 
"addressed the issue of the long-term survival of information. . . . Th- 
survival of information in an electronic environment becomes an 
intellectual and technological problem in its own right."[21 If we want 
assure permanence of the intellectual record that is published 
electronically, he said, then it will be necessary consciously to desig 
and build the required mechanisms within electronic systems. We ar 
still in need of those mechanisms. 

To address this need, this paper is in two parts. First, it will briefly 
describe some of the issues associated with preservation of the 
objects containing electronic information: medium preservation. 
Second, it will discuss the challenge of intellectual preservation, or tl 
protection and authentication of information which exists in electroni' 
form. Several potential methods of electronic preservation will be 
described, and one will be recommended for further attention. 


In the electronic environment it is unlikely that a focus of critical stud 
will be upon the electronic medium itself. To begin with, there is 
nothing in an electronic text that necessarily indicates how it was 
created; and the ease with which electronic texts can be transferred 
from disk to disk, or networked from computer to computer, means 
that there is no necessary indication of the source medium or even i 
the information has been copied at all. We are not likely to see sale 
catalog references in the future, therefore, which remark on the fine 
quality of the floppy disk's exterior label, or which remark on the 
electronic text's provenance ("Moby Dick on the original Seagate 
drive; never reformatted, very fine"). [3] 

The preservation of the information will still require the preservation 
whatever medium it is contained on at any given time. This is mostly 


IP Workshop - Graham: Intellectual Preservation 

Page 3 of 19 

what has been meant up to now when electronic preservation has 
been discussed. But there is another kind of preservation required fc 
information media: not only the preservation of the physical medium 
on which the information resides, but the preservation of the storage 
technology that makes use of that mediurn.[4J 

The physical preservation of media do not need extensive address 
here, for at any given time the physical characteristics of the mediun 
in use are well understood and the problems inherent in preserving i 
are simply financial and managerial: Who should pay for the 
necessary equipment and for the properly designed and acclimatize 
space, how often should backups be made, and who keeps track of 
backups and sees that they happen? These issues cause expenses 
for the electronic collection, but they raise only routine technological 

The storage obsolescence problem is quite another matter. A brief 
sequence of storage media many of us have seen in our lifetimes 
would include: 

• punched cards*, in at least three formats (80-column, 90-col, 

• 7-track half-inch tape* (at densities of 200, 556 and 800 bits pe 

• 9-track half-inch tape* for mainframes, with various recording 
modes and densities up to 3200 bpi and beyond; 

• 9-track half-inch tape cassettes* for mainframes ("square 
tapes", as they are known in contradistinction to the earlier 
"round tapes"); 

• RAMAC disk storage; 

• magnetic drum storage; 

• data cell drives*; 

• removable disk packs*; 

• Winchester (sealed removable) disk packs*; 

• mass storage devices (honeycombs of high-density tape 

• sealed disk drives; 

• floppy disks* of 3 sizes so far; and at least 3 storage densities 
so far; 

• cartridge tapes* of very high density (e.g. Exabyte) for use in 
workstation backups and data storage; 

• removable disk storage media on PCs; 

• laser-encoded disks* (CD-ROMs and laser disks); 

• magneto-optical disks*, both WORM (write-once-read-many) 
and rewritable. 

• = considered by some to have long-term storage potentia 

Some of the storage options appearing now and in the near future 


IP Workshop - Graham: Intellectual Preservation 

Page 4 of 19 

include new floppy disk sizes and storage densities, and "flash 
cards" (PCMCIA), or memory cards for use with very small 
computers. One sees discussion of storage crystals, encoded by 
laser beams and having the advantage of great capacity without 
moving parts, and probably even as stable as good paper. 

Technologies are superseding each other at a rapid rate. We know 
that authors and agencies are now storing long-term information on 
floppy disks of all sizes, but we don't know for how long we are goint 
to be able to read them. No competent authorities yet express 
confidence in the long-term storage capabilities or technological life 
any present electronic storage medium. CD-ROMs are an example. 
Their economical use in librarianship derives from their mass marke' 
use for entertainment; that mass market may be threatened by DVI 
(digital video interactive) technology, by DAT technology, or by othei 
now being actively promoted by entertainment vendors. If forms 
alternative to CDs win out in entertainment, the production of 
equipment for CDs and therefore CD-ROMs will be quickly curtailed. 

There are perhaps three possible long-term solutions for preserving 
storage media in the face of obsolescence (as opposed to physical 
decay), and they vary in practicality: preserve the storage technolog 1 
migrate the information to newer technologies, or migrate the 
information to paper or other long-term eye-readable hard copy. 

The prospect for the first option, preserving older technologies, is no 
bright: equipment ages and breaks, documentation disappears, 
vendor support vanishes, and the storage medium as well as the 
equipment deteriorates. 

The second option is migration. Most character-based data could be 
preserved by migrating it from one storage medium to another as th< 
become decrepit or obsolete. To do this requires a computer which 
can read in the old mode and write in the new; with present network 
capabilities, this is usually not difficult to arrange. 

Whether "refreshing" data is practical for large quantities of 
information over long periods of time is another matter. The present 
view of the Commission on Preservation and Access is expressed ir 
a report entitled Preservation of New Technology by Michael Lesk 
(see fn. 4). His view is that "refreshing" is the necessary and essenti 
means of preserving information as media obsolesce; I do not belies 
it -will be possible for more than a fraction of recorded information. Tl 
investment necessary to migrate files of data will involve skilled labo 
complex record-keeping, physical piece management, checking for 
successful outcomes, space and equipment. A comparable library 
data migration cost and complexity at approximately this order of 
magnitude would be the orderly photocopying of books in the 
collection every five years. This is not practical. In any case, this 


IP Workshop - Graham: Intellectual Preservation 

Page5 of 19 

migration solution will only work easily for ASCII text data. Migrating 
graphic, image, moving or sound data, or even formatted text, will 
only work as long as the software application can also be migrated t 
the next computing platform. 

The third option - practical but unexciting - is to migrate informatior 
from high-technology electronic form to stable hard copy, either papi 
or microform. In the near term, for certain classes of high-value 
archival material, this is likely to be the permanent medium of choice 
It offers known long life, eye readability and freedom from 
technological obsolescence. It also, of course, discards the flexibility 
in use and transport of information in electronic form. But until we 
have long-term stable electronic storage media, it offers the medium 
preservation mode most likely to be used. 


The Problem 

The more challenging problem is intellectual preservation - 
preserving not just the medium on which information is stored, but tt* 
information itself. Electronic information must be dealt with separate 
from its medium, much more so than with books, as it is so easily 
transferable. The great asset of digital information is also its great 
liability: the ease with which an identical copy can be quickly and 
flawlessly made is paralleled by the ease with which a flawed copy 
may be undetectably made. Barry Neavill wrote in 1984 of the 
"malleability" of electronic information, that is, its ability to be easily 
transformed and manipulated.[6] For an author or information provid 
concerned with the integrity of their documents, there are new 
problems in electronic forms that were not present in print. 

The issue may be framed by asking several questions which confror 
the user of an electronic document (which may be a text or may be 
graphic, numeric or multimedia information, for the problems are 
similar). How can I be sure that what I am reading is what I want? 
How do I know that the document I have found is the same one that 
you read and made reference to in your bibliography? How can I be 
sure that the document I am using has not been changed since you 
produced it, or since the last time I read it? How can I be sure that th 
information you sell me is that which I wanted to buy? To put it most 
generally: How can a reader be sure that the document being used i 
the one intended? 

We properly take for granted the fixity of text in the print world: the 
printed journal article I examine because of the footnote you gave is 
beyond question the same text that you read, and it is the same one 
that the author proofread and approved. Therefore we have 
confidence that our discussion is based upon a common foundation. 


IP Workshop - Graham: Intellectual Preservation 

Page 6 of 19 

The present state of electronic texts is such that we no longer can 
have that confidence. 

Taxonomy of Changes 

Let us examine three possibilities of change or damage which 
electronic texts can undergo that confront us with the need for 
intellectual preservation: 

1. accidental change; 

2. intended change that is well-meant; 

3. intended change that is not well-meant; that is, fraud. 
Accidental change 

A document can sometimes be damaged accidentally, perhaps by 
data loss during transfer or through inadvertent mistakes in 
manipulation. For example, data may be corrupted in being sent ove 
a network or between disks and memory on a computer; this happer 
seldom, but it is possible. 

More likely is the loss of sections of a document, or a whole version 
a document, due to accidents in updating. For example, if a docume 
exists in multiple versions, or drafts, the final version might be lost 
leaving only the previous version; many of us have had this 
experience. It is easy for the reader or author not to notice that text 
had been lost in this way. 

Just as common in word-processing is the experience of incorrectly 
updating the original version that was supposed to be retained in 
pristine form. In such a case only an earlier draft (if it still exists) and 
the incorrectly updated version remain. Again, a reader or author rru 
not be aware of the corruption. Note that in both cases backup 
mechanisms and the need for them are not the issue, but rather how 
we know what we have or don't have. 

Intended change ~ well-meaning 

There are at least three possibilities for well-meaning change. The 
change might result in a specific new version; the change might be e 
structural update that is normal and expected; or the change might t 
the normal outcome of working with an interactive document. 

New versions and drafts are familiar to those of us who create 
authorial texts, for example, or to those working with legislative bills, 
or with revisions of working papers. It is desirable to keep track 
bibliographically of the distinction between one version and another. 


IP Workshop - Graham: Intellectual Preservation 

Page 7 of 19 

In the past we have been accustomed to drafts being numbered and 
edition statements being explicit We are accustomed to visual cues 
to tell us when a version is different; in addition to explicit numbering 
we observe the page format, the typos, the producer's name, the 
binding, the paper itself. These cues are no longer dependable for 
distinguishing electronic versions, for they can vary for identical 
informational texts when produced in hard copies. It is for this reasoi 
that the Text Encoding Initiative Guidelines Project has called for 
indications of version change in electronic texts even if a single 
character has been changed. [7] 

It is important to know the difference between versions so that our 
discussion is properly founded. Harvey Wheeler, a professor at the 
University of Southern California, is enthusiastic about what he calls 
"dynamic document," continually reflecting the development of an 
author's thinking.[8] But scholars and readers need to know what the 
changes are and when they are made. Authors have an interest in 
their intellectual property. There is a sense in which the scholarly 
community has an interest in this property as well, at least to the 
extent of being able properly to identify it. 

Structural updates, changes that are inherent in the document, also 
cause changes in information content. A dynamic data base by its 
nature is frequently updated: Books in Print, for example, or a 
university directory ("White Pages"). Boilerplate such as a funding 
proposal might also be updated often by various authors. In each of 
these cases it is appropriate and expected for the information to 
change constant!y.[9] Yet it is also appropriate for the information to 
be shared and analyzed at a given point in time. In print form, for 
example, BIP gives us a historical record of printing in the United 
States; the directory tells us who was a member of the university in i 
given year. In electronic form there is no historical record unless a 
snapshot is taken at a given point in time. How do we identify that 
snapshot and authenticate it at a later time?[10] 

Another form of well-meaning change occurs in interactive 
documents. Consider the note-taking capabilities of the Voyager 
Extended Books, and the interactive HyperCard novels. [11] We can 
expect someone to want snapshots of these documents, inadequate 
though they may be. We need an authoritative way to distinguish on 
snapshot from another. 

Intended change fraud 

The third kind of change that can occur is intentional change for 
fraudulent reasons. The change might be of one's own work, to cove 
one's tracks or change evidence for a variety of reasons, or it might 
be to damage the work of another. In an electronic future the 
opportunities for a Stalinist revision of history will be multiplied. An 


IP Workshop - Graham: Intellectual Preservation 

Page 8 of 19 

unscrupulous researcher could change experimental data without a 
trace. A financial dealer might wish to cover tracks to hide improper 
business, or a political figure might wish to hide or modify 
inconvenient earlier views. 

Imagine that the only evidence of the Iran-Contra scandal was in 
electronic mail, or that the only record of Bill Clinton's draft 
correspondence was in e-mail. Consider the political benefit that 
might derive if each of the parties could modify their own past 
correspondence without detection. Then consider the case if each o 
them could modify the other's correspondence without detection. W€ 
need a defense against both cases. 


The solution is to fix a text or document in some way so that a user 
can be sure of the original text when it is needed. This solution is 
called authentication. There are three important electronic technique 
proposed for authentication: encryption, hashing and digital time- 
stamping. While encryption offers a form of data security, only 
hashing and digital time-stamping are useful for long-term scholarly 
communication and for providing protection against change of an 
intellectual creation. 


The two best-known forms of encryption are DES and RSA. DES is 
the Data Encryption Standard, first established about 1975 and 
adopted by many business and government agencies. RSA is an 
encryption process developed by three mathematicians from MIT 
(Rivest, Shamir and Adleman) at about the same time, and rnarkete< 
privately. It is regarded by many as superior to the Data Encryption 

Encryption depends upon mathematical transformation of a 
document. The transformation uses an algorithm requiring a particul 
number as the basis of the computation. This number, or key, is alsc 
required to decode the resulting encrypted text; the key is typically 
many digits long, perhaps 100 or more. Modern encryption depends 
upon the process being so complex that decoding by chance or 
merely human effort is impossible. It also depends upon the great 
difficulty of decoding by brute force. Computational trial-and-error 
methods would take unreasonably long periods of time, perhaps 
hundreds or thousands of years even using modern supercomputers 

Therefore the key is crucial to DES encryption. It is also the problem 
for passing the key to authorized persons turns out to be the Achilles 
heel of the process. How is the key sent to someone -- on paper in 
the mail? By messenger? These introduce the usual vulnerabilities 


IP Workshop - Graham: Intellectual Preservation 

Page 9 of 19 

dramatized in thriller literature. Do you send the key electronically? 
Sending it as plain text doesn't seem like a good idea, and sending i 
in encrypted form -- well, you see the problem. This is a recognized 
flaw in the widely-used DES encryption method. 

The RSA encryption technique is called public key encryption. The 
computational algorithm depends upon a specific pair of numbers, a 
public key and a private key; data encoded by one number cannot b 
decoded using the same number but can only be decoded by the 
other number, and vice versa (see Fig. 1). A correspondent B keeps 
one of the pair of numbers secret as a private key and makes the 
other number available as a public key. The public key can be used 
by anyone, for example her friend A, for coding messages which he 
sends to B; only B can decode them, because only she has the othe 
number of the pair. She sends an encrypted message back to A usir 
not her private key, but As public key, and only he can decode it, 
mutatis mutandum. 

Alternatively, B can code a simple message using her private key; 
anyone can decode it using her public key. This functions as a digita 
signature, allowing her messages to be authenticated, since only sh< 
is able to create such messages. The usefulness is evident in 
financial transfers, for example, or in authenticating e-mail or 
electronic purchase orders. 

Encryption is valuable for security. But neither the DES nor the RSA 
form is useful as an authentication system. Encryption could perhap: 
be used to authenticate a text if one considered it as an envelope wi 
contents presumed to be intact, but this would only work if the text 
had not been changed and re-encrypted. Encryption also has seven 
drawbacks as a long-term authentication means. No matter which 
method is used, encryption requires keys specific to the reader and 


IP Workshop - Graham: Intellectual Preservation 

Page 10 of 19 

writer. If the keys are generally available, as they would need to be f 
wide document access, then authentication is not possible, for the 
document could easily be modified and re-encrypted using the same 
keys. In addition, one of our concerns in librarianship is authenticatic 
over periods of time longer than a normal human lifetime. Secret ke} 
may be lost over such periods of time, making encrypted documents 


Another technique is called hashing; it is a shorthand means by whic 
the uniqueness of a document may be established. Hashing depend 
upon the assignment of arbitrary values to each portion of the 
document, and thence upon the resulting computation of specific bu 
contentless values called "hash totals" or "hashes." They are 
"contentless" because the specific computed hash totals have no 
value other than themselves. In particular, it is impossible or 
infeasible to compute backward from the hash to the original 
document. The hash may be a number of a hundred digits or so, bul 
is much shorter than the document it was computed from. Thus a 
hash has several virtues: it is much smaller than the original 
document; it preserves the privacy of the original document; and it 
uniquely describes the original document. 

Fig. 2 allows a simplified description of how a hash is created. If eac 
letter is assigned a value from 1 to 26, then a word will have a 
numeric total if its letters are summed. In the first example, EAT has 
the value of 26. The problem is, the word TEA (composed of the 
same letters) has the same value in this scheme. The scheme can fc 
made more complicated, as shown in the second pair of examples, i 
the letter-values are also multiplied by a place value. In this scheme 
the two words composed of the same letters end up with different 
totals. For the sake of illustration, the numbers at the right are showi 
as summed to the value 52 at the bottom; in fact the total is 152, but 
the leftmost digit can be discarded without materially affecting the fa 
that a specific hash total has been found: contentless, private, and (i 
this simple example) reasonably distinctive of the particular words in 
the "document." 


EP Workshop - Graham: Intellectual Preservation 

Page 11 of 19 

j 2s hashing T 
KA$HIN<ir arbitrary values, e<wvt*ntl££i tout* 
















A = 1 
8 w2 -' 

T a 20 






•;5' ■ 










This is a very simplistic description of a process that can be made 
excessively complicated for human computation. Using cryptograph! 
techniques, it is easy for current computing technology to compute 
quite complex hashes for any kind of document; paradoxically, these 
hashes are beyond the reach of computers to phony up or break in 
the perceived future. Hashing as a means of authentication is a topic 
of interest to the business and governmental communities and there 
have been several recent mathematical papers on it, including 
descriptions of recent patents. 

How might authors use hashing as an authentication technique? 
Above all it must be easy to use. It is typical for a document to be 
mundane at the time of its creation; it is only later that a document 
becomes important. Therefore an authentication mechanism must b< 
so cheap and easy that documents can be authenticated as a matte 
of routine. First, there must be an agreement on a hashing algorithnr 
that is generally trusted. Second, the algorithm must be widely 
distributable in a useful form, perhaps as a menu or hot-key 
command on a microcomputer or even embedded as a routine 
operating system option. To be useful, the selected algorithm must t 
commercially licensed and so cheap that there is no barrier to 
hashing documents at will. 

In such a scheme, each time a document or a draft is created or 
saved the hash is created and saved with it and is separately 
retrievable. If the document is electronically published, it is publishec 
with its hash; and if the document is cited, the hash is part of the 
citation. If a reader using the document then wishes to know if she 
has the unaltered form, she computes the hash easily on her own 


IP Workshop - Graham: Intellectual Preservation 

Page 12 of 19 

computer using the standard algorithm and compares it with the 
published hash. If they are the same, she has confidence she has th 
correct, untampered version of the document before her. 


Digital time-stamping takes the process a step further. Time-stampir 
is a means of authenticating not only a document but its existence a 
a specific time. It is analogous to the rubber-stamping of incoming 
mail with the date and time it was received. An electronic technique 
has been developed by two researchers at Bellcore in New Jersey, 
Stuart Haber and Scott Stornetta.[13] Their efforts initially were 
prompted by charges of intellectual fraud made against a biologist, 
and they became interested in the problem of demonstrating whethe 
or not electronic evidence had been tampered with. In addition, they 
are aware that their technique is useful as a means for determining 
priority of thought, for example in the patenting process, so that 
electronic claims for intellectual priority could be unambiguously 

Their technique depends on a mathematical procedure involving the 
entire specific contents of the document, which means they have 
provided a tool for determining change as well as for fixing the date < 
the document. A great advantage of their procedure is that it is 
entirely public, except (if desired) for the contents of the document 
itself. Thus it is very useful for the library community, which wishes t 
keep documents available rather than hide them, and which needs t 
do so over periods of time beyond those it can immediately control. I 
is also likely to be useful for segments of the publishing community 
which will want to provide a means for buyers to authenticate what 
they have purchased. 

The time-stamping process envisioned by Haber and Stornetta 
depends upon hashing as the first step. Assume, in Fig. 3, that Auth 
A creates Document A and wishes to establish it as of a certain time 
First he creates a hash for Document A using a standard, publicly- 
available program. He then sends this hash over the network to a 
time-stamping server. Note that he has thus preserved the privacy o 
his document for as long as he wishes, as it is only the hash that is 
sent to the server. The time-stamping server uses standard, publicly 
available software to combine this hash with two other numbers: a 
hash from the just-previous document that it has authenticated, and 
hash derived from the current time and date. The resulting number ii 
called a certificate, and the server returns this certificate to Author A 
The author now preserves this certificate, a number, and transmits ii 
with Document A and uses it when referring to Document A (e.g. in ; 
bibliography) in order to distinguish it from other versions of the 


IP Workshop - Graham: Intellectual Preservation 

Page 13 of 19 

fig* 3; TIMf-STAMFtHfl } 

A&U&* N creates 
{tocument A (private) 

Author, tf create* hash 

Author N 

Author N>1 creates 

rthfcr N*l create* hash 
Author H*r sends hash — 


(previous hash, 

create* new ha$h 
with DAT! & TIM£ 

H(h4sh) } 

creates new hash 

. * 


Crtaxt hash. 

W times) 

The time-stamping server has one other important function: It 
combines the certificate hash with others for that week into a numbe 
which, once a week, is now published in the personals column of Th 
New York Times ("Commercial and Public Notices"), as in Fig. 4. Th 
public nature of this number (what Stornetta calls an example of a 
"widely-witnessed event") assures that it cannot be tampered with. 

The privacy of the document been preserved for as long as Author / 
wishes; there is also no other secrecy in this process. All steps are 
taken in public using available programs and procedures. Note too 
that no other document will result in the same certificate, for 
Document A's certificate is dependent not only upon the algorithms 
and the document's hash total, but also upon the hash of the 
particular and unpredictable document that was immediately previoi 
Once Document A has been authenticated, it becomes itself the 
previous document for the authentication of Document B. 


IP Workshop - Graham: Intellectual Preservation 

Page 14 of 19 


"PUBLIC" AND •'«H|1> r 

Now let us consider Reader C, who wishes to determine the 
authenticity of the electronic document before her. Perhaps it is an 
electronic press release from a senatorial campaign, or an index 
purchased over the network from an electronic publisher, or perhaps 
it is the year 2093 and the document is an electronic text of Author fi 
Reader C has available the certificate for Document A. If she can 
validate that number from the document she can be sure she has th» 
authenticated contents. Using the standard software, she recreates 
the hash for the document and sends the hash over the network, wit 
the certificate, to the time-stamping server. The server reports back 
on the validity of the certificate for that document. 

But let us suppose that it is the year 2093 and the server is nowhere 
to be found. Reader C then searches out the microfilm of The New 
York Times for the putative date of the document in question and 
determines the published hash number; using that number and the 
standard software she tests the authenticity of her document just as 
the server would. 

What I have described are simplified forms of methods for identifying 
a unique document, and for authenticating a document as created a 
a specific point in time with a specific content. Whether the specific 
tools of hashing or time-stamping are those we will use in future is 
open to question. It is however the first time that authors, publishers 
librarians and end-users have been offered electronic authentication 
tools that provide generality, flexibility, ease of use, openness, low 
cost, and functionality over long periods of time on the human scale. 
Using such tools (or similar ones yet to be developed), an author cai 
have confidence that the document being read is the one he or she 
published, and that it has not been altered without the reader being 
aware of it. Such tools are essential for every player in the chain of 
scholarly communication. 



P Workshop - Graham: Intellectual Preservation 

Page 15 of 19 

It may be asked why librarians make such authentication issues thei 
concern. Why do they do this - why do they bother? The short 
answer is that it is what librarians do. As noted earlier, the basic 
professional paradigm for librarians is to acquire information, organi; 
it, preserve it and make it available. 

It is the preservation imperative that is particularly important for this 
audience of authors and publishers as well as for librarians. Authors 
and publishers have an interest in seeing that their works are 
preserved and provided in uncorrupted form, but neither have taken 
on the responsibility for doing so; librarians have. Authors have a 
specific interest in the uncorrupted longevity of their works, and both 
authors and research libraries have long periods of time as their 
concern. Librarians have taken on the particular responsibility to see 
that authors 1 works (and the graphic culture in general) are preserve 
and organized for use, not only by our generation but by succeeding 
generations of scholars and students. On behalf of future readers, 
librarians have the general responsibility for preserving against moth 
rust and change. If librarians do not preserve works for the long hau 
no one else will; once again, it is what librarians do. 

Speaking pessimistically for a moment, it is possible that the job 
cannot be done. We may all - librarians, authors and publishers - b 
swimming against the tide. Our society is obsessed with the present 
and is generally uncaring of the past and of its records. 
Technologically refined tools are now available which allow and 
encourage the quick and easy modification of text, of pictures, and c 
sounds. It is becoming routine to produce ad hoc versions of 
performances, and to produce technical reports in tailored versions < 
demand. Post-modernist critical theory detaches authorial intention 
from works, and demeans the importance of historical context. The 
technology that allows us to interact with information itself inhibits us 
from preserving our interaction. 

However, there is cause for optimism. In our house there are many 
mansions; there will continue to be people who want history, who ca 
what authors say, and who wish the human record to last. They will 
support the efforts of librarians to achieve these goals. We are 
fortunate that electronic preservation is of some interest to other 
communities for the mundane commercial reasons. The financial, 
publishing and other business communities have a stake in the 
authenticity of their electronic communications. The business and 
computing communities wish to protect against the undesired loss o 
data in the short term. The governmental and business communities 
profess an interest in the security of systems. 

The protection of intellectual property in the internetworked 
multimedia environment is the concern of this conference. The 
preservation of the actual information content is a prerequisite to the 


IP Workshop - Graham: Intellectual Preservation 

Page 16 of 19 

protection of property rights. Recognizing the need for authenticating 
and preserving our intellectual productivity is a common ground for 
authors, publishers and librarians. 


1 . Parts of this paper are drawn from the author's presentation at th€ 
1992 annual preconference of the Rare Books and Manuscripts 
Section of the Association of College and Research Libraries, and 
published in Robert S. Martin, ed., Scholarly Communication in an 
Electronic Environment: Issues for Research Libraries (Chicago: 
American Library Association, 1993), as "Preserving the Intellectual 
Record and the Electronic Environment" (pp. 71-101). 

2. Gordon B. Neavill, "Electronic Publishing, Libraries, and the 
Survival of Information," Library Resources & Technical Services 
28:76-89 (Jan. 1984), p. 78. 

3. However, see the recent work by Stuart Moulthrop, Victory Garde 
(Cambridge, Mass.: Eastgate Systems, 1991 [800 MB disk (signed 
and numbered 226/250 by author) for Macintosh + 16 p. brochure 
with introduction by Michael Joyce and explanatory matter, in plastic 
casing labeled "first edition"]). 

4. There is a third kind, the obsolescence of software designed to 
read a specific medium. For example, Kathleen Kluegel has pointed 
out how CD-ROM software updates have left unreadable older disks 
of the same published data base. She fears CD-ROM ending up 
"being the 8-track tape of the information industry" in "CD-ROM 
Longevity," message on PACS-L (Iistserv@uhupvm1.bitnet, April 29 

The best discussion of medium preservation, and the distinctions 
between the various kinds of obsolescence, is in Michael Lesk, 
Preservation of New Technology: A Report of the Technology 
Assessment Advisory Committee to the Commission on Preservation 
and Access (Washington, DC: CPA, 1992). 

5. See especially Lesk, but also Janice Mohlhenrich, ed., 
Preservation of Electronic Formats: Electronic Formats for 
Preservation (Fort Atkinson, Wis.: Highsmith, 1993), the proceeding; 
of the 1992 WISPPR preservation conference. 

6. Neavill, 1984, p. 77. 

7. TEI P1, Guidelines, Version 1.1: Chapter 4, Bibliographic Control, 
Encoding Declarations and Version Control (Draft Version 1.1, 
October 1990); sec. 4.1.6, Revision History, p. 55: "...[l]f the file 
changes at all, even if only by the correction of a single typographic 


IP Workshop - Graham: Intellectual Preservation 

Page 17 of 19 

error, the change should be mentioned.... The principle here is that 
any researcher using the file, including the person who made the 
changes, should be able to find a record of the history of the file's 

8. Harvey Wheeler, keynote speech at the October, 1988 LITA 
conference (Boston, Mass.). The issue arises in a different context ii 
the ESTC note below. 

9. A peculiar case is the transportation time-table; theoretically it coi 
be dynamically updated in electronic form, yet it is the timetable's 
hard-copy publication that signals to the users that a change has 

10. An electronic catalog is a similar case. Librarians never pretende 
that card catalogs were static, but the electronic catalogs (particularl 
when on the network) are so accessible as to raise citation problems 
Robin Alston, in Searching the Eighteenth Century (London: British 
Library, 1983), claimed superiority for the Eighteenth Century Short 
Title Catalog (ESTC) on the grounds that "machine-readable 
data.. .can be always provisional." Hugh Amory, a Harvard rare book 
cataloger, responded in a review by noting: "The permanence of prir 
has its own advantages, moreover: who will wish to cite a catalogue 
that can change without notice?" Papers of the Bibliographical Socie 
of America (PBSA) Vol. 79 (1985), p. 130. 

1 1 . See the discussion of hypertext books in Robert Coover, "The Ei 
of Books," The New York Times Book Review (June 21,1 992), p. 1 , 
23-25. Examples of such works include Moulthrop (n. 2 above), 
Michael Joyce, Afternoon: A Story (Cambridge, Mass.: Eastgate 
Systems, 1987), and Carolyn Guyer and Martha Petry, "Izme Pass," 
Writing on the Edge Vol. 2, no. 2 (Spring, 1991), attached Macintosh 

12. DES is described in FIPS Publication 46-1: Data Encryption 
Standard, National Bureau of Standards, January 1988. RSA Data 
Security, from whom information is available about their product, is i 
10 Twin Dolphin Drive, Redwood City, California 94065; the original 
description of RSAs method is in R. L. Rivest, A. Shamir, and L. 
Adleman, "A Method for Obtaining Digital Signatures and Public-key 
Cryptosystems," Communications of the ACM, Vol. 21 , No. 2 (Feb. 
1978), p. 120-126. 

A few readily available popular articles on the two schemes include 
John Markoff, "A Public Battle Over Secret Codes," The New York 
Times (May 7, 1992), p. D1; Michael Alexander, "Encryption Pact in 
Works," Computerworld, Vol. 25, No. 15 (April 15, 1991); G. Pascal 
Zachary, "U.S. Agency Stands in Way of Computer-security Tool," 
The Wall Street Journal (Monday, July 9, 1990); D. James Bidzos ar 


IP Workshop - Graham: Intellectual Preservation 

Page 18 of 19 

Burt S. Kaliski, Jr., "An Overview of Cryptography," LAN Times 
(February 1990). More technical and with many references is W. 
Diffie, "The First Ten Years of Public-key Cryptography," Proceedinc 
of the IEEE, Vol. 76, No. 5 (May 1988), p. 560-577. 

13. Stuart Haber and W. Scott Stornetta, "How to Time-stamp a 
Digital Document," Journal of Cryptology (1991) 3:99-111; also, und 
the same title, as DIMACS Technical Report 90-80 ([Morristown,] 
New Jersey: December, 1990). DIMACS is the Center for Discrete 
Mathematics and Theoretical Computer Science, "a cooperative 
project of Rutgers University, Princeton University, AT&T Bell 
Laboratories and Bellcore." The authors are Bellcore employees. 

D. Bayer, S. Haber. and W. S. Stornetta, "Improving the Efficiency 
and Reliability of Digital Time-stamping," Sequences II: Methods in 
Communication, Security, and Computer Science, ed. R. M. Capoce 
et al (New York: Springer-Verlag, 1993), p. 329-334. 

A brief popular account of digital time-stamping is in John Markoff, 
"Experimenting with an Unbreachable Electronic Cipher," The New 
York Times (Jan. 12, 1992), p. F9. A better and more recent summa 
is by Barry Cipra, "Electronic Time-Stamping: The Notary Public Go* 
Digital," Science Vol. 261 (July 9, 1993), p. 162-163. 


Peter S. Graham, Associate University Librarian for Technical and 
Networked Information Services at Rutgers University, co-leads the 
Working Group on Legislation, Codes, Policies and Practices of the 
Coalition for Networked Information, and serves on the Council of th 
American Library Association. Holding an M.L.S., he has been a 
senior administrator of university libraries and computing centers. 

Peter S. Graham 

Associate University Librarian for Technical 

and Networked Information Services 
Rutgers University Libraries 
169 College Ave. 
New Brunswick, N.J. 08903 
(908) 932-5908 
fax (908) 932-5888 

e-mail : psgraham@gandalf . rutgers . edu 

© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


IP Workshop - Graham: Intellectual Preservation Page 19 of 19 10/10/2002 

IP Workshop - Griswold: Method for Protecting Copyright on Networks 

Pagel of 12 

Sponsorec by: ^T^T§*^ Tl 

ARL I Coalition for Networked Information 

About CN1 

Task Force 





Site Map 

Search our she 

A Method for Protecting 
Copyright on Networks 

by Gary N. Griswold 


This solution to copyright protection uses software envelopes which 
authenticate each access by communicating with an authorization 
server on a wide area network. It decrypts the information for display 
print, or copying when the authorization is approved. This method is 
specifically suited to controlling information which has been delivere 
to customer machines over a wide area network. 


Many celebrate the freer environment of electronic networks: the ea: 
of data modification, copying, and multiple use usher in a relaxed 
attitude toward copyright. They believe that copyright holders must 
accept the less controlled environment of electronic networks. 
However, they are ignoring the property rights granted to authors an 
publishers in article 1, section 8, item 8 of the U.S. Constitution. The 
decision to place intellectual property on electronic networks is the 
prerogative of rights holders. 

Because publishers do not share this relaxed vision of copyright, the 
current providers of electronic services are delivering information wh 
does not require extraordinary protection, for example: open 
discussions, such as USENET; perishable information, such as new 
services; and government information, such as patent databases. 
However, any new system which wishes to leverage its content from 
the trillions of dollars in intellectual property already existing in the 
world, must address the property owner's concerns of property 
protection, or risk losing their cooperation. 

Mr. Timothy King, Vice President of Corporate Development at John 
Wiley and Sons, has identified the following key concerns. 


IP Workshop - Griswold: Method for Protecting Copyright on Networks 

Page 2 of 12 

• Will the integrity of information be preserved? 

• Will attribution for all information be ensured? 

• Will the quality of the content and form of information be 
maintained? Will creators and copyright holders be able to 
control the use of their work and to receive compensation for tl 

The legal problem must be solved. The High Performance Computir 
and Communications Act of 1991 (HPCC) specifically requires that t 
National Research and Education Network (NREN) include a means 
protect copyright: 

(c) NETWORK CHARACTERISTICS. — The Network shall — ... 

(5) be designed and operated so as to ensure the 
continued application of laws that provide network and 
information resources security measures, including thos 
that protect copyright and other intellectual property 
rights, and those that control access to data bases and 
protect national security; 

(6) have accounting mechanisms which allow users or 
groups of users to be charged for their usage of 
copyrighted materials available over the Network and, 
where appropriate and technically feasible, for their 
usage of the Network; [2] 

To date, the problem remains unresolved. In his December 8th, 199; 
presentation to Congress on the NREN, Dr. Allen Bromley, Director 
the Office of Science and Technology Policy, had the following to sa 
about the current status of copyright protection. 

The technical mechanism appropriate to protect copyright 
of material distributed over the network is as yet unclea 
. . . Because consensus has not been reached in this comple 
area, implementation of technical measures on the Network 
has not yet been scheduled. [3] 

There are an abundance of applications which require a solution to t 
problem if they are to be performed legally and without negative 
implications for publishers. Libraries, which are currently using FAX 
inter-library loan, are looking forward to delivering the journals over 1 
NREN. Likewise, the Colorado Alliance of Research Libraries (CARl 
Uncover Project and Engineering Information's Article Express are 
looking forward to NREN delivery. The CUPID project (Consortium c 
University Publishing and Information Distribution) is planning a 
distributed network architecture that will permit university presses to 
establish servers containing their copyrighted products in electronic 
form. These university press servers will be used for distributed 
publishing on the Internet. Libraries have made extensive progress i 


IP Workshop - Griswold: Method for Protecting Copyright on Networks 

Page 3 of 12 

putting bibliographic information on-line, and look forward to 
implementing digital libraries in which they deliver copyrighted 
information. Also, information retrieval systems, such as Wide-Area 
Information Service (WAIS), deliver the query result to the machine < 
the customer. At present, such systems are not being used for the 
delivery of information that requires protection. One can also concer 
of additional applications which could appear once adequate copyric 
protection were available. For example, news could be delivered by 
broadcast over the NREN, but only received by subscribers. Means 
filter the information to subscriber requirements would also be part o 
such a system. Journal subscriptions could be delivered electronical 
That is, each month a copy of the latest journals could be file 
transferred to the machines of each subscriber. Also, an electronic 
retail service could be provided so that customers could search by 
author, title, and subject indexes and request electronic delivery of 
titles they wished to purchase. 


Many solutions to this problem have been suggested. The following 
a discussion of some of the more important. 

Many have suggested a simple system: A library charges for each 
transmitted article and pays the publisher or the Copyright Clearano 
Center a royalty for each copy. This method is being used effectively 
by CARL (Colorado Alliance of Research Libraries) for FAXed journ< 
articles.[4] However, as we move to the electronic distribution of 
information, the ease with which information can be repeatedly 
distributed, for no fee after the first distribution, threatens the pruden 
of using this approach on computer networks. 

Digital signature use of public key encryption has been suggested a; 
means to protect copyright. A hashing algorithm is used to create a 
unique number from the content of a document. This number is 
encrypted with the private key of the originator. The receiver of such 
document can obtain the public key of the assumed source of the 
document from a central key facility.[51 However, while this importan 
technology verifies the source and content of the document, it does 
nothing to prevent the creation or use of copies. 

Public key encryption has also been suggested as a way to encrypt 
information. By using the public key of the receiver, only the receivei 
can decrypt it with their private key. However, while this is 
mathematically very secure, nothing prevents people from distribute 
encrypted information along with their private keys. The elegant 
security of public key encryption prevents anyone from identifying th 
source of the offending private key and copyright infringement. 

John H. Ryder and Susanna Smith describe a simple solution for the 


IP Workshop - Griswold: Method for Protecting Copyright on Networks 

Page 4 of 12 

electronic dissemination of software. Before the customer receives t 
copyrighted software product in working form, he or she is presentee 
with a number of screens of text which display a license agreement. 
The customer must follow certain steps on the keyboard to signify th 
they agree to the terms of the license agreement.JBJ However, while 
this method makes certain the customer understands their licensing 
rights, it does nothing to insure that the customer lives up to those 

Martin E. Hellman describes a means to limit access and bill usage < 
software, video games, video disks, and videotapes. This is 
accomplished via an encrypted authorization code, which contains 
information related to an identification of the computer, a product, a 
number of uses requested, and a random or non-repeating number. 
When entered into the customer's base unit, the authorization code 
permits use of the specified software product for the specified time.[ 

Victor H. Shear describes a system and method to meter the usage 
distributed databases, such as CD-ROM. This method describes a 
hardware module which must be part of the computer used to acces 
the distributed database. This module retains records of the intellect 
property viewed. Once the module becomes full, it must be removed 
and delivered to someone who will charge for the usage and set the 
module back to zero.[8] 

Hellman's and Shear's methods both require hardware modules, wh 
must be constructed into the customer's computer, in order to contrc 
access. These methods will not be practical until a very large numbe 
of computers contain these modules. Hardware manufacturers will b 
hesitant to include these modules in the design of their computers ui 
there is sufficient demand for these specific systems. 


A solution to the copyright protection problem is described in the 
following section. Patent applications have been filed on the pivotal 
aspects of the innovation ^, 10,11] 

Description of the Innovation 

Our approach is as follows: copyrighted information is transmitted in 
encrypted form, and is transmitted in a software "envelope". The 
copyrighted information and the software envelope together compris 
an executable program which can decrypt the copyrighted informatic 
and present it to the user. The capabilities of the envelope intentions 
limit the user's access to the copyrighted information to those 
capabilities which are appropriate under copyright law for the specifi 
kind of copyrighted information contained. For database information, 
the software envelope should enable the user to search indexes anc 



EP Workshop - Griswold: Method for Protecting Copyright on Networks 

Page 5 of 12 

display text. For CAD information, the software envelope should pen 
the display of the information and permit the user to manipulate 
attributes of the display. For video information, the software envelop 
should display the video. For audio information, the software envelo) 
should display the audio information. For text, the software envelope 
should display and turn pages. For hypertext information, the softwa 
envelope should allow the user to thread through the information. 
These are only some of the ways these software envelopes can con 
different kinds of copyrighted information. 

Finally, the software envelope uses a method to check for 
authorization to access and to track the usage of the software 
envelope and copyrighted information over the same 
telecommunication network used to transmit them to the user. The 
tracking method works as follows. Automatic messages are sent 
between the software envelope and a central authorizing site. Each 
time a customer starts to use a copyrighted work, a message is 
automatically sent from the work. Also, at a regular interval, addition 
messages are sent. Sent at regular intervals, they are a measure of 
use. When the messages arrive at the central authorizing server the 
are verified. A reply is sent back, which is an authorization to contini 
or a denial of authorization. If no valid message returns, a denial is 
assumed by the software envelope. Whenever a denial is received c 
assumed, the use of the software or copyrighted information producl 
discontinued. The diagram in Figure 1 illustrates this method of 
tracking copyrighted information. 

Benefits of the Innovation 

The system of authorization and usage measurement capabilities 
described above can be used to license information products in a 
variety of ways to suit a variety of information licensing policies. It ca 
be used to enforce site licenses by preventing off-site access and 
limiting the number of concurrent uses. It can be used to limit duratic 
of use, analogous to returning a book to a library, by disabling use o 
an information product after a period of time. It can be used to 
implement an electronic subscription by providing an unending 
duration of use of the product on one machine. It can also be used tt 
meter and charge for each use of the information. 


IP Workshop - Griswold: Method for Protecting Copyright on Networks 

Page 6 of 12 

The software envelope would provide the user with the ability to viev 
the information product, but it would not provide any way to edit or 
extract from it. This is needed, because otherwise the displayed 
information could be used as a source from which to create a new 
copy which is not subject to this copyright protection scheme. Secor 
it would insure the authenticity of the information products, by 
preventing the automatic creation of altered copies. Third, it would 
interfere with plagiarism, which has become an increasing problem 
because of the abundance of easily copyable electronic information. 
Fourth, it would prevent the automated generation of derivative work 

Other Licensing Requirements 

So far, we have only discussed controlling licenses for viewing 
information, but the same method can be used to control licensed 
printing. While the rightsholder may choose to give the customer a 
license to view and to print, they could require an additional expense 
for the action of printing. In this case, the authorization request woul' 
indicate that printing is requested and the reply would indicate whett 
the customer is licensed. The act of printing would be recorded for tf 
purpose of charging. In some computer operating system 
environments, insuring the security of the document will require the 
installation of a special print server, which is capable of decrypting 
while printing. 

This system permits unlimited copying on the network, and yet limits 
the use of those copies to licensed customers. However, a customet 
may need to take an electronic copy of a document onto a machine 
which is not connected to the Internet. For machines which contain 
internally readable serial numbers or firmware private keys, we can 
license and control the act of making copies. Each copy made will 
contain the internal identifiers of the machine on which it is to run. It 
still be encrypted, and requires a similar software envelope for 
presentation. Instead of checking for further authorization over the 
network, the software envelope checks that it is running on the 
machine to which it is licensed. 

Network Infrastructure 

This method assumes the existence of a network used in the deliver 
of electronic information. This network should also be capable of 
sending connectionless datagrams. Analog telephone is both too sic 
for sending large amounts of data, and would require an explicit 
telephone call with each use of an information product. Integrated 
Services Digital Network (ISDN) telephone, because of its minimum 
K bps speed, would be much more suitable for the transmission of 
information products. Also, the authorization datagrams which this 
method requires could be sent over the signaling channel without 
placing a call. Similarly, on the Internet, the authorization datagrams 


IP Workshop - Griswold: Method for Protecting Copyright on Networks 

Page 7 of 12 

can be most efficiently transmitted and processed as User Datagran 
Protocol (UDP) datagrams. Digital Cellular would also be a very 
suitable network. 


At this time, we have a demonstration version of our technology 
running on the Internet. The system consists of three main programs 
1) a license authorization program called "authorize"; 2) a program f 
creating protected files called "product"; 3) and a program for viewin 
the protected files called "read". The authorization server runs on on 
machine on the Internet in Albany NY, and will control access to any 
documents created using the "product" program. Copies of "product" 
and "read" are available upon request. 


While the above prototype has many capabilities, it has many 
limitations which make it less than a commercial product. While it do 
register the creation of new protected products, authorizes access, 
tracks usage, and permits customers to register upon receiving a 
denial, it does not include a customer billing module or a publisher 
payment module. While the software envelope provides the essentia 
features needed to display the decrypted information, it lacks the us< 
interface quality one would expect in a commercial product. Finally, ■ 
viewer program is written to run on Sparcstations. Versions are not ) 
available for other computers. Despite all of the above limitations, th 
Demonstration Prototype performs an important service by 
demonstrating how licenses can be managed over the Internet. 


We will be able to proceed with this step as soon as the necessary 
funding is available. This system should be limited in the number of 
products sold and the number of customers serviced in order to 
facilitate revision of the system as we learn from its use. However, tt 
system should provide the full scope of functionality required in a 
commercial version. That is, it should manage licenses for viewing, 
printing and node-locked copying, and it should maintain a full 
database about its customers and publishers, which should be used 
bill customers and pay publishers. The system should provide a high 
quality presentation program which is available on a wide variety of 
platforms. Such a viewer could be developed by InfoLogic, but it woi 
be more efficient to have the developers of an existing viewer integr; 
InfoLogic's license control mechanism into their viewer. Finally, the 
license server will be redundantly implemented to guarantee 100% 


IP Workshop - Griswold: Method for Protecting Copyright on Networks 

Page 8 of 12 


There are a variety of applications for which the described method o 
copyright license management would be very useful. These include: 
electronic retailing, inter-library loan, library circulation, and distribute 
information services. The following is a description of how each of 
these applications could function using the copyright protection 
mechanisms described in this report. 

Electronic Retailing 

Publishers and printers have automated their methods of production 
that typeset copies of books or journals exist in electronic forms, sue 
as Standard Graphics Markup Language (SGML) or Postscript. Fror 
these electronic copies, the pages are printed. These same electron 
forms are a useful source for electronic distribution. In addition, 
scanned copies of older books are a source of electronic distribution 

After printing their books and journals, the publisher could license th 
electronic sources to the electronic retailer. The only task the publisf 
needs to perform is signing the license agreement. There is no need 
for a second tier of distribution. The electronic retailer could offer to 
pay for each copy delivered to the customer. Considering the absent 
of printing, inventory, warehousing, and returns, the publisher could 
earn a considerably larger margin than they receive on paper copies 
Considering the absence of two-tier distribution in this model, the 
electronic retailer could sell the copies for less than the cost of pape 

Those currently connected to the Internet include most universities; 
most national laboratories; most private research laboratories doing 
government work, or collaborating with universities; and a growing 
number of smaller organizations, especially technical. As a result of 
this profile, it appears that PSP/STM (Professional Scholarly 
Publishing and Scientific, Technical and Medical) are the publishing 
segments where the demand will occur first. 

To begin using the system, the customer would request a copy of th 
electronic retailer's client program over the network. The client progr 
could be delivered free, or for a nominal charge. The first time the 
customer used this client program, they would be asked to enter 
identifying information. This program would enable them to browse 
through the title, author, and subject catalog of books and journals ir 
the electronic retail server. They could request any book, whereupor 
they would be required to enter charging information, such as a cred 
card number. The book or journal would be delivered to them 

For universities and organizations the system would permit the site 


IP Workshop - Griswold: Method for Protecting Copyright on Networks 

Page 9 of 12 

licensing of the information, while at the same time permitting the 
licensing to individuals or licensing by the duration of time used. 
People would be able to share electronic documents freely, and all 
accesses to a site licensed document within the site would be 
permitted. However, if someone off the licensed site were to receive 
copy, they would be denied access when they attempted to access i 

Inter-library Loan and Document Delivery 

Inter-library loan and document delivery services are very similar, 
except that one is a library service and the other commercial; one 
usually pays copyright royalties while the other usually does not. Usi 
this copyright management method they become even more similar. 

When a document is requested for delivery, it is located, scanned in* 
a computer, and immediately converted to an encrypted file. The 
protected file can be transferred to the requester's machine and a 
licensing entry permitting one concurrent use of the document can b 
made at the same time. Once received, the document can be freely 
accessed by the requester on the machine to which the document w 
sent. Should the requester pass the document along to others, they 
not be able to access the document until they have secured a licens 
to the document. At the same time that they receive a denial of acce 
from the license server, they will be given the opportunity to enter 
charging information on the screen which will permit them to access 
the information. 

On a periodic basis, the license management system will generate 
administrative reports which detail the following: 1) library charges fc 
documents delivered; 2) library receipts for documents provided; 3) 
copyright royalties for documents provided; 4) copyright royalties for 
additional licensees added to previously delivered documents. Thes- 
documents could be the basis for payments between libraries and Vr 
Copyright Clearance Center. 

Library Circulation 

A possible use of this technology is for each library to maintain a 
license server to manage the copies of books and periodicals which 
have been checked out from their library in electronic form. In additic 
to the technology previously described, the digital library card catalo 
must contain a record of the number of copies owned and number o 
copies borrowed for each item in the electronic card catalog. Such a 
system would work as follows. 

Each time someone wishes to check out an electronic copy of a boo 
or periodical, the current "number owned" by the library and the curr 
"number checked out" from the library would need to be looked up tc 
be certain that a copy is available. 


IP Workshop - Griswold: Method for Protecting Copyright on Networks 

Page 10 of 12 

When a book or article is checked out from the library, a licensing er 
for the user would be entered into the license database. A terminatic 
date, such as two weeks, would be entered in the license to represe 
the borrowing period. The card catalog's record of the number of 
copies checked out from the library would need to be updated to 
indicate that the copy has been removed from the library. 

When the two-week borrowing period of the book or periodical 
terminates, the copyrighted work would cease to be accessible by th 
library patron, even though the copy still exists on his or her computi 
On a nightly basis, the library's system could look in the licensing 
database for copies which have terminated on that day and decreas 
the "number of copies checked out" shown on the electronic card 
catalog. This action is analogous to returning the book or periodical ■ 
the library shelf. 

Advantages of Standardization 

If this technology were consistently implemented by libraries and 
electronic retail services, it would be possible for the holder of a cop; 
checked out from the library to purchase the same item from a retail 
service. The customer would use the software envelope of a retail 
service to try to access the library copy of the document. Upon gettir 
a denial of access, they would fill out the charging information 
requested on their screen by the electronic retailer. Once this step w 
completed, they would be purchasing a copy of the book or periodic; 

Distributed Information Services 

Currently, providers of on-line services fill their large computers with 
quantities of information and charge the customers for the use of the 
infrastructure needed to access that information. Using the methods 
this paper, much more efficient information services are possible. Fc 
example, one could provide a bibliographic information retrieval serv 
at no cost, since money would be made on the sale of information. 

Before using this system, the customer would need to provide certai 
charging information, such as corporate purchase orders, or credit c 
numbers. The customer would search the on-line bibliographic 
database for documents on particular topics. Once documents are 
selected by the user, the documents or abstracts of the documents 
could be delivered to the user by file transfer. Access to the informal 
could be measured in a variety of ways. By default, it may make sen 
to charge the customer for the time each document is accessed. Tin 
would be measured in intervals, such as every 15 minutes. In additic 
the customer could be charged for printing out a copy of the 
documents. Finally, the customer could be given the opportunity to 
purchase permanent electronic copies that they may store and view 
any time without further charge. The license servers can be apprise< 


IP Workshop - Griswold: Method for Protecting Copyright on Networks 

Page 11 of 12 

these events by automatic messages, sent between the software 
envelopes and the license server. 


One of the side effects of these methods of distribution is to lower th 
amount of infrastructure needed to deliver information, because mos 
of the information access occurs on the customer's own computer. 
Lowering the cost can in turn lower price and thus increase profit. Ar 
lowering of price of the currently expensive electronic information is 
to increase demand. We need to build into our selling systems a 
positive feedback loop which would lower costs of operation, to lowe 
prices, and increase demand. Increased demand would lower the pe 
unit production costs, which increases demand even more. At the 
same time, we must retain and even increase the use of peer review 
and editorial filtering to insure the availability of the highest quality 
information. This technology facilitates the lowering of operational 
costs, while providing a mechanism to compensate for the time and 
effort that went into production. 


1 . Tim King, "Critical Issues for Providers of Network Accessible 
Information", EDUCOM: Summer 1991, Page 82. 

2. High Performance Computing and Communications Act of 1991 
(HPCC), Section 15 USC 55112 (c). 

3. Dr. Allen Bromley, Director of the Office of Science and Technoloj 
Policy, "The National Research and Education Network Program: A 
Report to Congress", December 1992, Page 2. 

4. CARL Systems, Inc., Uncover and Uncover2--the Article Access t 
Delivery Solution, unpublished article, 1992. 

5. Public-Key Cryptography Standards, RSA Data Security, Inc., Jur 

6. John H. Ryder and Susanna R. Smith, "Self-verifying Receipt and 
Acceptance System for Electronically Delivered Data Objects", Unite 
States Patent 4,953,209; August 28, 1990. 

7. ? 

8. Victor H. Shear, "Database Usage Metering and Protection Systei 
and Method", United States Patent 4,977,594, December 1 1 , 1990. 

9. Gary N. Griswold, "License management system for information 
products located at user site periodically requesting usage 


IP Workshop - Griswold: Method for Protecting Copyright on Networks 

Page 12 of 12 

authorization via communication network", Application for Internatior 
PCT patent, 1992. 

10. Gary N. Griswold, "System and method for protecting and licens 
information products on an electronic network", Application for Unite 
States Patent, 1992. 

11. Gary N. Griswold, "System and method for protecting and licens 
software on an electronic network", Application for United States 
Patent, 1991. 


Gary Griswold is President of InfoLogic Software, Inc., a consulting 
firm which develops software in technical applications including: Ver 
Large Scale Integrated (VLSI), CAD, Image Recognition, Computer 
Aided Software Engineering (CASE), Manufacturing Automation, an< 
Management Information Systems. Recently, his primary technical 
interest has been copyright protection for networked information. He 
holds an M.S. (Union College, Schenectady, NY) and a B.S. 
(University of Washington, Seattle). 

Gary Griswold 

1223 Peoples Avenue 

Troy, NY 12180 

Tel: (518) 276-4840 

FAX: (518) 276-4841 



© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


IP Workshop - Macq/Quisquater: Digital Images 

Page 1 of 7 

Sponsorec by: 



Coalition for Networked Information 

About CNI 

Task Force 





Site Map 

Search our she 

Digital Images 
Multiresolution Encryption 

by Benoit Macq and Jean-Jacques Quisquater 


Digital image transmissions often require compression, 
secrecy and transparency. We have developed a 
multiresolution encryption algorithm, where the low- 
resolution information of the images (i.e. their icons) remains 


Nowadays conditional access systems for digital image 
transmission or storage are a necessity. Among their range 
of applications one can point out: 

• pay-TV, 

• medical images for transmission on LAN or for 

• confidential videoconferences and 

• secret facsimile transmissions. 

Digital images can be considered as a given number of bits 
and an encryption could be achieved by directly applying a 
conventional method, like the Data Encryption Standard 
(DES). The DES is a one-to-one mapping of blocks of 64 bits 
defined by a 56-bit secret key. This method would, however, 
have two major drawbacks: 

• First, the image is not a random amount of data: the 
pixels are connected by a correlation process which 
could offer a possible path for breaking the encryption. 


IP Workshop - Macq/Quisquater: Digital Images 

Page 2 of 7 

More precisely, the unknown key could be retrieved by 
a method giving the maximum correlation for the data 
at the output of the decoding. 

• The output of the DES is pseudo-random and no 
compression can be achieved after the encryption, 
since the apparent correlation has disappeared. 

Applying a method like the DES after a compression coding 
of the image seems attractive since the output of the coding 
is more or less random and already encoded at the required 
bit rate. However, this method is also not satisfactory, for 
three reasons: 

• A user could intend to protect his images independently 
from the nature of the transmission channel, i.e. 
independently from the compression algorithm in use in 
this channel. 

• Compression techniques are very sensitive to 
transmission errors and are specifically protected. 
Generally, a specific framing and synchronization is 
added to the compressed data. A DES encryption 
would decrease dramatically the efficiency of this 

• In many applications, the encryption has to be 
somewhat transparent: 

- A broadcaster of pay-TV does not always intend 
to prevent unauthorized receivers from receiving 
his program, but rather intends to promote a 
contract with non-paying watchers. 

- The access to the icons of a secret image bank 
could also remain unprotected. 

These observations have led us to propose a new image 
encryption technique. In our technique the encryption is 
achieved before the compression (see Figure 1). We 
propose a multiresolution scheme which produces a 
"compressible" image with a certain level of transparency. 


IP Workshop - Macq/Quisquater: Digital Images 

Page 3 of 7 


Our cryptosystem can be modeled as in Figure 1. In this 
figure, the encryption function is isolated from the other 
components of the transmission system. Our algorithm is 
based on the following specifications: 

• lossless: The encryption process has to be reversible, 
with perfect reconstruction of the image, DK(EK(I))=I. 

• multiresolution: The algorithm has to be somewhat 
transparent, encoding only the details above a given 
resolution. Furthermore, it allows conditional access for 
resolution: e.g., one could provide High- Definition TV 
with free access to the TV signal. More formally, the 
two first properties are related to two factors: 

o - the transparency, which is maximum when D(E 


o - the opacity, which is minimum when E(l)=l and 
maximum when E(l) is totally scrambled. So the 
variable opacity of the cryptosystem will allow the 
user of the system to decide on the degree of 
unrecognizability of the image. 

• compressible: The compression of the encrypted 
image has to remain efficient, i.e., the encrypted image 
must have similar statistical properties to a real picture, 
i.e., the compressions of / and E(l) for a given rate 
have to lead to similar coding distortions. 

• secure: The cryptosystem has to be resistant to any 
known attack. Attacks specific to high redundant 
messages like images are to be taken into account. 
Notice that there are some connections between the 
secure and the compressible conditions, since if the 
encrypted image is highly correlated it is highly 
compressible and also difficult to attack by maximizing 
the correlation. 


IP Workshop - Macq/Quisquater: Digital Images 

Page 4 of 7 

• low-complexity: The algorithm has to be based on 
low-cost operations. 


The core of the system is a one-to-one lossless 
multiresolution mapping of images based on a new operator 
that we define as the L-H mapping. The L-H mapping maps a 
pair of pixels (x(/-1), x(i)) into two numbers (xl, xh), xl being 
close to the half-sum of the pixels, xh being close to the pixel 
half-difference. The signals xh and xg can be interpreted as 
the approximation and the detail of the pixel pair. This new 
mapping is depicted in Figure 2 and can be easily 
implemented by using some logical gates. 

Figure 2: The L-H mapping 

The L-H mapping is applied first in the horizontal direction 
and then in the vertical direction, only on the horizontal 
approximation signal. The process is applied recursively on 
the approximation signal according to the decomposition 
pattern shown in Figure 3. A corresponding image is shown 
in Figure 4. We denote this decomposition as the Lossless 
Multiresolution Transform (LMT). A permutation of lines or 
columns after the LMT, followed by the corresponding 
inverse LMT, allows us to generate an encrypted image from 
which the original picture can be reconstructed. 


IP Workshop - Macq/Quisquater: Digital Images 

Page 5 of 7 



Figure 3: LMT and permutations 

Let us give some details on the process. We denote by xi[j] 
the value of the pixel at position (i,j) of the resulting image 
after a LMT. For the sake of simplicity, we assume that the 
number of pixels in a column or a row is a power of 2, that is, 
2[ I] for some /: these pixels are numbered from 0 to 2[l] - 1 . 
We denote by xi a column of pixels at position / and by x[j] a 
row of pixels at position j. A permutation of columns (resp. 
rows) of pixels is a reversible transformation from any subset 
of columns (resp. rows) into itself. We denote by PK a 
permutation indexed by K. This value K is related to the set 
of chosen permutations and is called the key when used in a 
cryptographic scheme. A set of consecutive columns (resp. 
rows) in the range [/1,...,/2], /1 [[sterling]] /2, is denoted by 
x/1,/2 (resp. x[H,i2]): the corresponding permutation of these 
columns {resp. rows) is denoted by xPK(/1/2) (resp. x[PK 
(i1 ,i2)]). Using L to denote the LMT, we have 

L[-l] (PK(L(I) ) ) = EK(I) 

DK{X) = L[-l] (PK[-1] (L(X) ) ) 

The opacity of the encryption can be modulated by the 
number of L-H decomposition. In Figure 3, we have a 3-level 

An encrypted image is shown in Figure 5. In order to 
increase the compressibility of the scheme, we could perform 
conditional permutations of the values; the detail values are 
permutated by data in the same context (we permute xh 
values having the same range for the corresponding xg value 
and neighborhood). 


IP Workshop - Macq/Quisquater: Digital Images 

Page 6 of 7 




The method proposed in this paper is preliminary. Further 
issues are related to the improvements (and how to measure 
them) of the algorithm properties (compressibility, security, 


The use of cryptographic scrambling for protecting 
handwritten signatures and signal television is very old; see 
the standard reference [3] for instance, and the two relevant 
old papers [4] and [5], 

A recent book about cryptology is [6]. 

[1] Proceedings of the First International Seminar on 
Conditional Access for Audiovisual Services, Rennes, 
France; June 1990. 

[2] Takeshi Kimura, Masafumi Saito and Seichi Namba, 
"Some studies on conditional access for DBS television 
service-Algorithms of permutation scrambling and an 
experimental decoder with smart card" in [1], pp. 107-122. 

[3] David Kahn, The codebreakers. Macmillan Publishing 
Co., New York, 1967, pp. 827-836. 

[4] Signature scrambler foils forgery, Management and 
Business Automation, Sept. 1960, p. 53. 

[5] Don Kirk, Engineering report on encoding television 
signals, Jerrold Electronics Corporation, Philadelphia, 1955. 

[6] Gus J. Simmons (Editor) Contemporary cryptology. The 
science of information integrity, IEEE Press, 1992. 


BenoTt Macq received the Mngenieur Civil Electricien' and the 
'Docteur en Sciences Appliquees* degrees from the 
Universite Catholique de Louvain (UCL), in 1984 and 1989, 
respectively. He has worked on telecommunication planning 


IP Workshop - Macq/Quisquater: Digital Images 

Page 7 of 7 

in the Tractionnel society in 1985, and on video coding in the 
Telecommunication Laboratory of the UCL from 1986 to 
1990. From 1990 to 1991 , he was with the Philips Research 
Laboratory Belgium. He is now permanent researcher of the 
Belgian NSF (Xhercheur Qualifie' du FNRS), at the 
Telecommunication Laboratory of the UCL. 

Laboratoire de Telecommunications 

2, Place du Levant 
B-1348 Louvain-la-Neuve 


Jean-Jacques Quisquater received his MS in applied 
mathematical engineering (1970) from the Universite 
Catholique de Louvain and his PhD in computer science 
(1987) from the University of Paris (Orsay). Formerly, he was 
project leader and senior scientist in information security and 
cryptology at Philips Research Laboratory Belgium. Since 
1992, he has been an associate professor at the UCL. He 
also teaches at the Ecole Normale Superieure (Paris) and at 
the University of Namur. 

Laboratoire de Microelectronique 

3, Place du Levant 
B-1348 Louvain-la-Neuve 

e-mail : 


© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


IP Workshop - Matsui/Tanaka: Video-Stenography 

Page 1 of 1 

Sponsor ec by: 


About CNI 

Task Force 





Site Map 


Coalition for Networked Information 

Video-Stenography: How to 
Secretly Embed a Signature 
in a Picture 

by Kineo Matsui and Kiyoshi Tanaka 

At this time, this document is unavailable electronically. 

Search our she 


©2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


IP Workshop - Jensen: Need-Based Intellectual Property Protection 

Page 1 of 11 

Sponsorec by: 


About CNI 

Task Force 





Site Map 

Search our she 



Coalition for Networked Information 

Need-Based Intellectual 
Property Protection and 
Networked University Press 

by Michael Jensen 


The needs of university presses for intellectual property 
protection are a good microcosm for understanding the 
needs of electronic publishers in general. Systems will need 
to be reasonably secure (rather than utterly secure), and be 
flexible enough to accommodate a wide range of content 
forms and transaction forms. Header-based security holds 


Pve heard speakers at various conferences say that 
publishers won't be necessary in the New Online World. I 
think that's wrong. Publishers will survive because people 
want authentication and validation, both as authors and as 
readers. In a networked environment, the greater the volume 
of information, the greater the need for distillation and 
dependability, which publishers will provide. 

University presses will survive because scholarship, 
academic prestige, and tenure committees will survive. An 
electronic publication by a university press will simply be 
more believable, trustworthy, and potentially important than 
an ftp-able file on WUarchive will be, or an electronic 
publication by Acme Publishing--not to mention more useful, 
attractive, and readable. Publication in high-quality form by a 
full-fledged publisher will be preferred by authors, and 
readers will prefer trustworthy documents as their mainstay 


IP Workshop - Jensen: Need-Based Intellectual Property Protection 

Page 2 of 11 

of information. New forms of publishing will inevitably unfold, 
but the institution of publishing will not die out. 

For the people gathered at this conference, considering 
methodologies for intellectual property protection, it's useful 
to understand the underpinnings of the sale of scholarly and 
academic information. Nonprofit publishers such as 
university presses are a particularly appropriate model, since 
profiteering is not one of our goals. The goal is rather to 
provide information of high value to the few people who'll 
value it highly, but who will not pay too high a price. 

Network publishing will not make information too cheap to 
meter. In fact, the printing costs of a book-the only variable 
that changes in the networked environment-are generally 
only 15% to 20% of the overall costs of publishing. 
Manuscript development, peer review, copyediting, 
production costs like design, typesetting (read code- 
enrichment) and proofreading must all be considered when 
assessing the costs of publishing, whether that's electronic 
or print publishing. There are also such non-luxuries as 
publicity, marketing, order-fulfillment, record-keeping, and 
accounting which must be paid for. The value added by 
publishers take humanpower and brainpower, which must be 
financially supported. Straight-from-the-author document 
transmission may be cheap, but publishing isn't. The security 
systems we're talking about today are essential for the 
continuation of peer-reviewed, well-edited, well-promoted, 
well-designed and well-produced documents; that's why I'm 
so pleased to be invited to be here today. 

Intellectual property concerns are at the heart of much 
informed hesitation to commit to electronic publishing. 
Protection of published information is essential, and without 
reasonably secure environments or systems, much of the 
best scholarship available will be very slow to go online. 

I use the phrase "reasonably secure" intentionally. Generally, 
like anything under lock and key, the more secure it is, the 
more hassle it is to get to. Publishers aren't interested in 
having those serial-port dongles attached to every electronic 
book. Nor are we willing to force users through arduous or 
costly verification procedures. 

Intellectual-property protection approaches must be flexible 
enough to vary according to the needs of the publisher 
(whether that's a university press, a scholarly society, an 
individual scholar, or a commercial publisher), and must be 
adaptable to the needs of the user, and to the technical 


IP Workshop - Jensen: Need-Based Intellectual Property Protection 

Page 3 of 11 

capacity of the user's system. 

It's clear that no single protection scheme will cover all 
security needs. Different kinds of documents will require 
different levels of protection, different forms and levels of 
access, as well as different subscription and pricing and 
distribution channels (which affect the protection demands). 
Therefore, before outlining specific strategies, I'd like to 
briefly overview some of the varied contents, and the varied 
protection demands called for by that content. 


Humanities texts, for example, are likely not to need the 
same degree of "timeliness" as the sciences, with which 
most of you, I think, are more likely to be familiar. Archival 
material is important: original sources. The scholar browses 
and mulls and finds references and makes notes. Makes 
marginalia for later thought. Highlights key passages. They 
(we) tend to want to have the entire document, in context, 
and easily available. The humanities scholar has a different 
"information-need model," if you will, than one in the 
sciences. In the Internet environment, humanities 
scholarship will require repeated and dependable access to 
the same documents, as well as easy interconnections to 
other similar documents during research. 

The information content of the sciences differs quite 
significantly from the standard humanities content. Current 
information is often much more important than archival 
information. Frequently, texts are read once, and only rarely 
re-referenced. The documents themselves are visually and 
operationally different: there tends to be much more 
reference material-tables, graphs, mathematical models, 
graphic representations. It lends itself more to multimedia 
work, and will need those sorts of tools-interactive graphs, 
interactive models, interactive algorithms. These last 
interactive content models may need a different protection 
system-and permission system-than the text within which it 

Journals have a different set of needs than individual texts; 
they're a more direct-to-customer form of publishing than 
book sales, which is why journal managers are often the 
most interested in Internet publishing. Timeliness is often 
tremendously important, for which the Internet is a boon. A 
single security check for a selected sequence of individual 
articles is required. 


IP Workshop - Jensen: Need-Based Intellectual Property Protection 

Page 4 of 1 1 

Monographs have been declared dead, but I doubt that. I 
think there's room for the monograph even in an e-mail 
soundbite world, because it allows for context to be built 
brick by brick like the walls of a house. Monographs may be 
more likely to be downloaded and printed out than reference 
works, journal articles, or scientific texts. Local site 
ownership is more likely than online access. 

Different disciplines and different forms have different 
information-access models, which in turn will demand 
different security models-most of which I can't predict. I can 
say that while university presses predominantly publish text- 
based information now, that will change to include sound and 
video as they become applicable. 


The content of the texts published will make demands upon 
any security structure, and must be integrated into the other 
great demand: working within the varied economic structures 
of publishing. These will change dramatically. Current 
theories imply that because delivery will be simpler, the 
business will be simpler. I think that's a misinterpretation of 
the complexity of the business of publishing. 

Our main objective-beyond the prime objective of economic 
survival-is to get it into the hands of interested people. 

Currently, to do that we have an intricate and interconnected 

Figure I, 

Current Sales Web 

Purchase Transaction 

mmm *Y*fc Tfaii Wtfon 

• * • * • Utwm fir Btyaity Tbtn$a<*ikMi 

Ihufettctfons M in eluded : Iraiuvlat ion 
rigltts, extracts, $ertnf$$i<m$ bmk ifuhs, 


IP Workshop - Jensen: Need-Based Intellectual Property Protection 

Page 5 of 11 

web of distributors, resellers, bookstores, and individuals we 
serve (see Fig. 1). Bookstores often buy our books from 
distributors and from wholesalers and from us directly. 
Individuals may call our 800 number to order, or may call up 
their bookstore, or a wholesaler, or a distributor. Libraries 
may order from us or from the library wholesaler or from 
both. Publishers sell units, which are then resold as units. 

It's easiest and cheapest for us to sell units in bulk, of 
course, because there's less humanpower involved. We like 
to sell to wholesalers, and bookstores, and libraries. 

But this business has been developed based on units-a 
commodity. Electronic publications are not units in the same 
way. When we shift to a network publishing framework, 
suddenly a welter of new connections, new possibilities, and 
new "networks" appear (see Fig. 2). 

Pmeniial Sales Web 

♦ . < Uccrisc or Royatty Itamsadiao ; 

We may sell a site-license to a library exclusively for the 
campus-wide network. We may license to a "virtual 
bookstore," which functions as a sort of "for-profit library." 
We may license to a new kind of entrepreneur, who builds a 
sort of tailored educational experience and rents it over the 
web, and for whom our book is one license and royalty 
among many he must calculate. We may license to a 
university the rights to sell/distribute/display a specific text for 
a course, but only for the duration of a course, for which the 
students all pay a small fee, of which the publisher and 
author receive some proportion. We may sell directly to the 
customer, providing client-server systems for online access 
directly, or "rent" access for referencing, or sell a text for 


IP Workshop - Jensen: Need-Based Intellectual Property Protection 

Page 6 of 11 

local ownership-even for printing out locally. We may use 
the Internet to connect up with books-on-demand printers 
using Docutech or Lionheart systems-high-speed PostScript 
printers/binders for generating reprint-like documents. 

Licensing becomes dramatically important, because the 
same electronic text can and will be used in a variety of 
forms, sold by a variety of vendors, and manipulated by a 
variety of users, each of which will have a different security 
model, usage model, and pricing model. 

In the networked world, we must design systems-or 
appropriate existing systems-that will allow us to rent, sell, 
and license texts, to allow these very different audiences with 
very different needs to view, search, annotate, copy in limited 
fashion, and/or virtually "own" these texts. We also must be 
ready to provide mixed models on demand. 

Scholars who "own" an annotated online text-say a server- 
based display-only collection of documents-will also want to 
make temporary connections to other publications-to check 
references, make glancing checks of related documents, etc. 
Currently, Scholar Smith owns one collection of books 
outright, books she purchased personally. She also has 
related books she's borrowed from the library. And she 
"rents" information via fair-use photocopying or interlibrary 
loan. In the near future, we must build electronic models that 
allow these interconnections, even foster them, thus 
providing scholars with what they want: to have validated, 
paid-for ownership, be able to "rent" certain brief connections 
to other titles or journal articles, and be able to borrow 
access from the library, which has purchased the title or 
journal from a publisher. 

Through all of this we must be able to make these sales (at 
differential costs), track these licenses and sales, confirm 
their use and their limits, collect payments, and pay royalties 
to our authors accordingly, as well as provide readers with 
some form of authenticity check. All without having the text 
easily copied by Scholar Smith to all her friends as a 

This is a tall order, and is why many models won't be put into 
practice right away. But it also needn't be done all at once, 
which is a relief. This web I describe is perhaps five years 
off, Td say-or longer (if ever), if security systems aren't 

Let me come back to "reasonable" security, and what 


IP Workshop - Jensen: Need-Based Intellectual Property Protection 

Page 7 of 11 

university presses need to make the previously described 
flexible desktop library possible. 


From what I've seen, I don't believe there's any way to 
effectively build absolute data security into any ftp-able or e- 
mailable file, without a prohibitively significant hassle factor. 
Hashing and public-key encryption could work for individual 
texts, but unless there's a universal yet specifically-designed 
front-end that handles the decryption on-the-fly-and which 
itself cannot be copied-then either a morass of document- 
specific codes would result, making a hard-disk-stored 
"bookshelf clumsy, or,we'd end up with an array of unique 
and mutually exclusive front-ends cluttering up one's virtual 

The viable models-in my opinion-are all variants of a client 
server, in which access is constrained and controlled by the 
server itself. This assumes a stable and direct network 
connection and appropriate display hardware and software, 
of course. The servers might belong to a library (to whom a 
site license is sold by a publisher), or a university, or a 
"virtual" bookstore, or the entrepreneur, or the on-demand 
printer, or the reference service, or the publisher itself. 

Reasonable security is all we require. Client-server systems 
can and will be cracked; consequently publishers (and other 
server owners) will need security structures that provide the 
authentication systems described by Dr. Graham, to be sure 
that the texts which are served are the authoritative version. 
This can be done, I suspect, relatively easily, via a separate 
archive which is copied back to the server periodically to 
assure that the "authoritative" version is always available. 

Occasional crackers who are simply borrowing or stealing 
access aren't so much the worry, any more than occasional 
shoplifters are a worry. I'm not even tremendously worried 
about commercial theft-to sell a text, its existence must be 
publicized; a thief doesn't publicize a theft. Black market 
bookstores simply aren't likely. I'm a bit concerned about 
international theft-out where copyright conventions aren't 
followed-but that's a matter more of trade policy and 
international law. 

Publishers are primarily, and justifiably, concerned about 
local abuse. If Scholar Smith purchases access to a title, 
either as an "owner" or a "renter"--then we want to be sure 
that she doesn't have easy means to copy or print files 


IP Workshop - Jensen: Need-Based Intellectual Property Protection 

Page 8 of 11 

without either notification to the publisher, payment of some 
secondary cost, or official permission. If Scholar Smith can 
copy and e-mail (or print and OCR) any title, article, or 
chapter, and give it to any other colleague who can then 
continue the copying, publishers will be reticent to make it 
available. What we want is reasonable security that 
precludes casual gross copying by well-meaning colleagues, 
and precludes "broadcasting" of a text by any individual. We 
don't want to be the Big Brother information police, but we do 
want means to protect our intellectual property rights. 

The Z39.50 communication protocols have been--if I 
understand them correctly-transformative, allowing a 
multiplicity of systems to be built that were internally 
compliant, and thus interconnectable. Gopher, WAIS, Panda, 
World Wide Web, and other publication access systems are 
internally compliant, and so can work apparently seamlessly 
together. I'm hoping this workshop begins the process of 
creating a similarly flexible set of security protocols. I want a 
scholar to be able to have access to a multiplicity of titles 
from a multiplicity of publishers from a multiplicity of sources, 
and be able, relatively seamlessly, to have a virtual desktop 
which allows easy connectivity to the titles he or she "owns" 
or "rents" or borrows. 


Header-based security-in natural conjunction with client- 
server security-looks the most promising for establishing the 
appropriately flexible security protocols. The following list of 
header information is a reasonable minimum for allowing a 
reasonable amount of protection within many client-server 
models, assuming that the headers themselves were 
reasonably secure. 

ISBN-the International Standard Book Number, a 
unique identifier for every published text. 

Copyright-holder information/Bibliographic 
information. It seems reasonable to have some variant 
of the standard "books in print" data included with a 
published document. 

Publisher's electronic address, to be used for a 
variety of purposes-communicating transactions, 
checking authenticity, perhaps verifying ownership via 
a message transaction sent to that address. 

Authentication-site. This is the address from which a 


IP Workshop - Jensen: Need-Based Intellectual Property Protection 

Page 9 of 11 

hash-number or other unique identifier-derived from 
the text itself--can be checked against the version 
onscreen. This may differ from the publisher's own 
address. A variant on the authentication-site might be 
an "access-site" tag, which would allow access only if 
the server's IP address matched the code. 

Printable/nonprintable/amount printable; 
Copyable/noncopyable/amount copyable. This 
would function as a "public-domain/non-public-domain" 
identifier as well, thus allowing those who didn't give a 
hoot about redistribution to provide a means of 
indicating that. This data might also allow some control 
over redistribution, while still allowing limited fair-use 

License information: n/a for individual sales, but 
otherwise would include a) number of concurrent 
viewers; b) access-site limits (as in "accept only 
readers with login addresses from the following nodes"; 
and c) identification of licensee (in case of illegitimate 

Hashed/NotHashed, encrypted/not encrypted. For 

some publishers and for some documents, encryption 
of some kind is likely, even if unwieldy. 

Time stamping, which for us would be "date of 

Duration of copyright on the work. 

Character set used by the document. 

Searchable/not searchable-if we have "knowbots" 
hunting around, we must have some scheme that 
allows searching without retrieving--so that my knowbot 
can tell me that there's a resource that's exactly what 
I've been looking for, if I want to buy it. 

Coding scheme (raw text, SGML-enriched, PostScript, 
Acrobat, TEX, etc.) 

Attached-file information-are illustrations, graphs, 
algorithms, figures, and tables original and subsumed 
under the overall copyright? If they are "permission" 
inclusions-elements copyrighted elsewhere for which 
permissions have been obtained-where do their 
permission-headers lie? How can those elements be 


IP Workshop - Jensen: Need-Based Intellectual Property Protection 

Page 10 of 11 

protected independently? 

One of my problems defining the list above is that security 
structures seem to be unavoidably intertwined with the 
access system using them. A security structure that is 
flexible enough to provide a wide range of architectures with 
tools for building systems is also probably flexible enough for 
there to be an underground of front-ends written that 
circumvent the restrictions-perhaps even those restrictions 
that are server-based, since the front-ends will be reading 
and responding to the headers. 

Some client-server systems could have a security system 
that validated access by comparing client codes, client codes 
plus account address, and/or server codes plus address plus 
password. But those security structures won't mean anything 
if the user can easily print out the entire file, or use the flash- 
OCR tools that are around the corner, or use some other tool 
for snaring the file as it displays on the screen. Some of that 
is unavoidable-what we want is that stealing be so awkward 
that it must be willful theft rather than a just a lapse into the 
ethical grey zone. 

It may be that "authoritative versions" are the final "security," 
and that having "authoritation centers" may be necessary. A 
Library of Congress-like bank of hash-scheme authoritative- 
version proofs for public-domain documents, and similar 
banks held by the publishers of copyrighted information, 
might be useful. 

I'm not able to say what system or combination of systems is 
best. Would that I could. But I'm hopeful that the sorts of 
solutions I'm hearing today, and hope to continue to hear, 
can be combined in a manner that allows publishers to feel 
secure enough on the Internet to make available the vast 
array of scholarship that we publish. 


What I hope I've done today is describe the publisher's 
perspective on the needs for security, and show the 
complexity of the interconnections between resellers, 
retailers, lenders, and individuals with which we deal every 
day. We want to provide scholars and students and the 
reading public with a variety of options which suit the needs 
of the text, the researcher's method, and the idiosyncratic 
needs of the reader. We want to be able to serve our 
customers, whoever and wherever they are. And we want to 
be able to feel reasonably secure that our publications aren't 


IP Workshop - Jensen: Need-Based Intellectual Property Protection 

Page 11 of 11 

being copied freely everywhere around the world. 

We want an environment where scholars, students, and 
interested readers can be sure that the information they're 
getting is dependably available, certain of worth, and 
unerringly trustworthy, and where millions of items are 
available relatively seamlessly. The best qualities of the 
present system-flexible and mixed distribution, flexible and 
mixed access, flexible and mixed ownership-need to be built 
into the security protocols that are devised. 

We can't do it alone-we don't have the programming 
expertise. But I'm hopeful that those protocols can be 
devised, and I'm hopeful that university presses can help 
structure and test those protocols in the real, virtual world of 
the Internet by being partners in the creation of the protocols. 


Michael Jensen is the Electronic Media Manager at the 
University of Nebraska Press, one of the ten largest 
university presses in the country, and the first to have a 
searchable publications catalog on the Internet. This paper is 
presented under the auspices of the Association of American 
University Presses. 

Michael Jensen 

University of Nebraska Press 
327 NH 

901 North 17th Street 
University of Nebraska 
Lincoln, NE 98588-0520 

© 2002 Coalition for Networked Information. Ail Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


IP Workshop - Massarsky: ASCAP, BMI and SESAC 

Page 1 of 12 

Sponsorec by: 


About CNI 

Task Force 






Site Map 

Search our she 


Coalition for Networked Information 

The Operating Dynamics 
Behind ASCAP, BMI and 
SESAC, The U.S. 
Performing Rights Societies 

by Barry M. Massarsky 


Existing copyright collective organizations such as ASCAP, 
BMI and SESAC have developed, on behalf of their music 
rights holders, intricate licensing and distribution 
mechanisms that may augur the intellectual property 
safeguards confronting the emerging interactive multimedia 


The following discussion highlights the essential operating 
dynamics utilized by the American Society of Composers, 
Authors and Publishers (ASCAP), Broadcast Music Inc. 
(BMI), and the Society of European Stage Authors and 
Composers (SESAC), to represent the public performance 
copyright interests on behalf of music copyright holders. This 
discussion will first concern itself with the role of the 
collective societies in licensing, identifying and distributing 
copyrighted musical works and then contrast the differences, 
when apparent, among the agent cooperatives. Second, 
parallel interests to networked information and multimedia 
will be provided, including the role of proxy evaluations as an 
alternative measurement device. 

By definition, the intellectual property concerns of multimedia 
properties are far more expansive than the traditional 
borders affecting music licensing interests. However, the 
business of copyright protection for music licensing rights 


IP Workshop - Massarsky: ASCAP, BMI and SESAC 

Page 2 of 12 

holders has roots dating back to 1914 when the original 
concept was devised to protect and administer public 
performance rights. The resulting entity, ASCAP, provides an 
appropriate window for the development of new copyright 
collective initiatives. ASCAP's competitors in the 
marketplace, BMI and SESAC, provide further insight into 
the learning process. 

The following treatment provides an overview from which to 
judge this industry's relevance to the emerging multimedia 


At this time, the licensing strategies for the three U.S. 
performing rights societies are similar. The bulk of the 
licensing effort concerns the application of the blanket 
license. The blanket license allows the music user 
unlimited access to the collectives' licensed repertory, 
for a contractual period of time, in exchange for a profit 
participation in the music user's economic growth. The 
following discussion will break down each component part of 
the aforementioned sentence. 

The blanket license allows the music user... 

The music user is defined as any organizational entity that 
wishes to use music, in a public performance form, for a 
commercial or non-commercial business purpose. This broad 
characterization includes radio stations, local commercial 
television stations, network television, public radio and 
television stations, cable television, background music 
services such as MUZAK, bars, grills, skating rinks, baseball 
stadiums, funeral parlors, etc. 

ASCAP defines its music user market through strategic 
litigation initiatives. A case in point would be clothing stores 
such as the Gap chain. ASCAP defined this relatively new 
market as commercial establishments deploying industrial 
radio speakers for use as a sales inducement. These stores 
use the music from already-licensed radio stations for a 
different motive; the music is used not as a source of private 
entertainment but rather to stimulate a sales environment for 
the product. ASCAPs legal forces prevailed on the "double- 
dip" licensing concept. BMI followed ASCAP's market lead. unlimited access to the collectives' licensed 
repertory... .ip-workshop/Massarsky.html 


DP Workshop - Massarsky: ASCAP, BMI and SESAC 

Page 3 of 12 

The concept of unlimited access to the licensed repertory is 
the heart of the blanket license strategy. ASCAP and BMI 
(and to a lesser extent, SESAC) maintain that the ease of 
access accommodates the music user's need to gain instant 
permission for copyright use and thus provides a true service 
to the licensee community. Blanket licensing, according to 
the societies, eliminates the structural impediment resulting 
from transactional licensing. Most importantly, it allows 
ASCAP, BMI and SESAC to minimize their administrative 
costs in providing a licensing structure for the music user 
community. As we shall see later, these virtues are now seen 
differently by the music user in a vastly changed, 
technologically-enhanced, and cost-containment conscious 
entertainment economy. 

...for a contractual period of time... 

The significance of this statement is twofold: (1) it ties up the 
licensee with the repertory for a period of time, allowing the 
collectives to enjoy a stable economic relationship; (2) it ties 
up the copyright holder to the individual collective 
representing its works. ASCAP refers to this phenomenon as 
"licenses in effect." When ASCAP negotiates a license 
agreement with a user group (traditionally, broadcasters and 
other music user types form negotiating committees that 
represent the industries 1 interests), it promises to that group 
that it represents the song catalogs owned by its writer and 
publisher members. ASCAP's membership rules allow a 
writer or publisher to resign at a fixed point each year, but 
the songs attributable to the catalog, as represented to the 
music user in a negotiated agreement, must stay with 
ASCAP through the duration of the agreement with the music 
user. exchange for a profit participation in the music 
user's economic growth. 

The blanket license calls for a negotiated fixed percentage of 
the music user's gross revenue (allowing for some 
deductions) as consideration for the unlimited access 
doctrine. Each industry group negotiates with the performing 
rights societies based on its valuation of the use and 
importance of music in its operation. 

The most intensive music user, the radio broadcasting 
industry, pays the highest rate (approximately 2.5% of gross 
revenues) for each of the 8,500 stations currently in 
operation. This fixed rate facilitates a simpler enforcement 
strategy by eliminating the need for customized agreements 


IP Workshop - Massarsky: ASCAP, BMI and SESAC 

Page 4 of 12 

with each station. The societies regularly audit the reported 
financial disclosures to determine the gross revenue base. 
The local television industry negotiates similarly but in recent 
years has been battling ASCAP for a viable alternative to the 
blanket license. All other music users are also licensed 
through a percentage-of-gross formula. The glaring 
exception had been the commercial television networks 
which had been paying on a flat sum basis. Pending the 
decision of a recent rate determination hearing between ABC 
and CBS against ASCAP, this flat sum licensing practice 
may end soon. 


ASCAP, BMI and SESAC negotiate the value of their 
licenses in such a similar approach that the ASCAP 
approach is representative of the efforts for all three 
performing rights societies. 

ASCAP's licensing relationship with significant music user 
groups is predicated on an historical base line which has 
evolved over the last five decades. Once ASCAP established 
market legitimacy through a series of strategic infringement 
lawsuits successfully litigated against the radio broadcasters, 
the subsequent licensing agreements with the radio industry, 
constructed during the 1 930*s, allowed for a subjective 
valuation of the significance of music in broadcast. The 
negotiation amounted to "horse trading" between the users 
and the creators of the intellectual property. 

As the license negotiation practice evolved, ASCAP did an 
economic analysis of financial data pertaining to the 
anticipated growth of the radio industry. ASCAP argued that 
music was an essential component of the profit-seeking 
broadcasters and thus, license fees should be linked to the 
industry's gross revenues. The broadcasters were more 
accustomed to a variable rate structure for securing rights 
with creative talent such as writers, actors, directors, etc. The 
notion of a creative element sharing in a revenue stream was 
anathema to their interests. To this day, broadcasters are 
irate about the idea that ASCAP is a silent partner in the 
ownership of a radio or television station. 

As these license agreements progressed overtime, ASCAP 
would monitor the use of its protected music on licensee 
stations and when positive trends were apparent, insist on an 
increase in the fixed percentage of gross revenues. Other 
macroeconomic conditions such as inflation required an 


IP Workshop - Massarsky: ASCAP, BMI and SESAC 

Page 5 of 12 

indexing of the rate into the 1970's. 

It is fair to judge that the relationship between the user and 
the creator became strained. Recently, the broadcasters 
have begun to exercise some of their contractual rights in 
seeking an effective alternative to the blanket license: 


ASCAP and BMI offer a per-program license for music users 
that require minimal access to their repertories. Typically, all- 
talk or all-news radio stations have been the prime 
beneficiary of such a licensing arrangement. Recently, local 
television stations have won the right to a per-program 
license for syndicated programming aired on non-network 

In practice, the per-program license has been an inefficient 
alternative to the blanket because ASCAP and BMI have 
insisted on passing high administrative costs along to the 
user. This tactic drives up the transactional costs, and when 
coupled with an onerous user reporting requirement, makes 
this option less attractive than the blanket. The world of per- 
program licensing has recently changed with the final rate 
determination decision handed down by Magistrate David 
Dollinger ( United States v. ASCAP In the Matter of the 
Application of Buffalo Broadcasting Co. et. al ) Civ. 13-95 
(WCC), governing the operating rules for determining a per- 
program license for local television stations. This ruling has 
opened up the per-program window by setting the initial rate 
at 140% of an applicable blanket license rate and then 
reducing the effective rate for those television programs 
which have no appreciable ASCAP music. The final outcome 
is likely to encourage more stations to choose a per-program 
alternative. The cable industry is expected to follow the local 
television broadcasters with a demand for a per-program 
license. These changes will widen the interpretation of 
ASCAP's Consent Decree with the Justice Department 
governing ASCAP's licensing offers with the music user 

Though in the infant stages of development, SESAC is 
planning to introduce an alternative license that leverages a 
new song detection technology which matches a digital 
imprint detected from actual airplay, with a digitally- 
recognized pattern resident in a database. This information 
will allow for a first-time application of a usage-based license 
formula. Use of this and other new technologies will allow the 
performing rights societies to capture more information at a 

http : //www . cni .org/docs/ima. ip- workshop/Mass arsky . htm 1 


EP Workshop - Massarsky: ASCAP, BMI and SESAC 

Page 6 of 12 

diminishing marginal cost. 

Historically, other licensing options have met with stiff 
resistance within the music community. For example, efforts 
to license the public performance of music directly with the 
creators, bypassing the collective agent, have been in large 
measure unsuccessful. This form of licensing is referred to 
as direct licensing. It is difficult for the music user to properly 
identify and locate the copyright owner. Often, the copyright 
owner does not want the administrative burden of direct 
licensing and refers the user to the performing rights 
societies. Direct licensing is more appropriate when good 
information is available which relates the copyright holders 
and users in the marketplace. The natural concern about 
infringement (intended or unintended) inhibits the growth of 
this license form. 


ASCAP, BMI and SESAC expend tremendous efforts to 
allocate royalties to their respective memberships. ASCAP's 
overhead is 18%; most of that cost is apportioned for survey 
and distribution expenses. If there are discernible differences 
among the collectives, their respective choice in allocation 
methodology is what sets them apart. A discussion of each 
methodology follows. 

ASCAP Survey Approach 

ASCAP's Consent Decree requires that an independent 
survey research firm govern the principles guiding ASCAP's 
survey and distribution efforts. These statistical principles 
can be summed up in one phrase: follow-the-dollar. An 
analysis of ASCAPs commercial local radio survey will serve 
as the paradigm for other ASCAP surveys in local television, 
cable, public television, etc. 

Stratified Sampling 

ASCAP's radio survey is stratified by geographic area, 
economic class, and type of community. These groupings 
allow ASCAP to properly represent the balance of all 
collections across the United States. The geographic 
consideration suggests that all regions should be sampled in 
relation to their pro-rata contribution of earnings. For 
example, ASCAP surveys 60,000 hours of radio airplay each 
year. If 10% of their radio collections come from New 
England radio stations, then 6,000 hours will be dedicated to 
stations in that region. 


IP Workshop - Massarsky: ASCAP, BMI and SESAC 

Page 7 of 12 

Disproportionate Sampling 

This principle provides that not all stations will be sampled 
and that each station's sampling allowance is not necessarily 
equal. ASCAP's methodology indicates that a radio station 
paying $10,000 is guaranteed to be sampled at least once 
during the year; stations paying in excess of $10,000 are 
sampled pro-rata to a $10,000 station; and stations paying 
less than $10,000 may or may not be sampled. ASCAP's 
follow-the-dollar strategy weights the larger stations more 
favorably in the more lucrative advertising markets. ASCAP 
does include small stations in the sample mix but to an 
increasingly smaller degree. 

Random Sampling 

ASCAP employs several random techniques in constructing 
its sample design. Again, as far as small stations are 
concerned, their eligibility for inclusion is predicated on a 
random draw. The actual dates selected, and times of day 
that taping takes place, are governed by statistical principles 
of random occurrences. As an example, ASCAP maintains 
that out of the 60,000 hours of annual radio taping, the 
probability of sampling a Monday is roughly one-seventh and 
the probability of drawing a morning tape (typically 7am-1pm) 
is one-fourth for the four, six-hour average dayparts in a 24- 
hour day. ASCAP prides itself on these techniques to assure 
that all performances have an equal opportunity of being 
sampled. The reality is less inviting: ASCAP's 60,000 hours 
represents only 0.1% of the universe of radio broadcast 
hours, suggesting that 99.9% of all performances go 
undetected. ASCAP makes the argument that a truly 
scientific sampling fairly represents the entire universe of 
possibility within allowable cost tolerances. 

BMI Survey Approach 

While ASCAP is dedicated to great precision generated from 
small samples, BMI looks to include a greater number of 
performances absent the rigorous precision. BMI's radio 
sample includes over 500,000 hours of logged radio 
performances. While the number of BMI's recognized 
performances is over eight times greater than ASCAP's, the 
system of relying on program logs creates some concerns. 

BMI requires that radio stations submit airplay logs on a 
regularly scheduled basis. BMI notes that it includes only a 
portion of the submitted logs for distribution purposes. The 
stations do not know if they are part of the selected group. .html 


IP Workshop - Massarsky: ASCAP, BMI and SESAC 

Page 8 of 12 

However, logs indicate the airplay schedule, not the actual 
performance. Computer-generated song rotational systems 
provide hour and minute listings of these airplay schedules. 
Often, these scheduling efforts are interrupted by last-minute 
insertions or deletions that are never revealed to BMI. Other 
stations fill out handwritten logs days after the actual 
performances have taken place. In these situations, the 
station employee is asked to recreate the playlist ex post. 
Many times, the employees have multiple tasks at the station 
and fail to promptly or accurately fill out the log requests. 

BMI has made greater strides in television. BMI was the first 
performing rights society o conduct a complete count or 
census of syndicated programming on both local and cable 
television. ASCAP is just beginning to catch up. 

SESAC Survey Approach 

Because of SESAC's limited repertory offering, a 
comprehensive sampling approach was deemed 
unnecessary. SESAC employs a passive allocation system 
that relies on published title rankings as the basis for 
payment. Such publications as Billboard, R & R,and The 
Gavin Report provide SESAC with a listing of chart songs 
ranked by sales volume. It is assumed that sales volume and 
airplay are positively correlated; in fact, that relationship is 
not as obvious when compared with the ASCAP and BMI 
systems. The SESAC system remains simple and cost- 
effective for the repertory it represents. 

SESAC is currently planning a major overhaul of its sampling 
and distribution strategy as it relates to specific music 
genres. At this time, the model description is deemed 


In general, the collectives have similar approaches to 
retaining, processing and weighting data. Again, a discussion 
of ASCAP's methodology also reflects those of the other two 
licensing organizations. 

ASCAP maintains a library of over 2 million song titles. To 
date, while this information is stored on massive tape and 
disk drives connected to its mainframe server, ASCAP also 
relies on hard copy index cards submitted by copyright 
holders, each of which identifies a copyright registration. A 
registration provides ASCAP with the copyrighted song title, 
writer(s), music publisher(s) and copyright date. 


IP Workshop - Massarsky: ASCAP, BMI and SESAC 

Page 9 of 12 

ASCAP dedicates a 30-person department to both manually 
and electronically update this information. The electronic 
version of this file is referred to as the title data base and is 
utilized extensively in the royalty allocation process. 

The final product of ASCAPs survey efforts is the 
identification of song titles picked up in ASCAP's various 
samplings of radio and television broadcasts. The sample 
data is linked to the title data base for matches on writer, 
publisher and society affiliation. 

Other databases play a critical role in directing the allocation 
system. The member databases provide essential 
information on name, address, social security number, 
authorized representative (in the case of publishers), and 
earnings history. The member databases also relate the 
song information stored in the title database. 

When a song is detected on a sampled radio station, the 
song traverses the other files for matches on second-level 
information. Once this is accomplished, the songs are 
grouped by writer name to form the first stage of the royalty 
payout. ASCAP also applies a weighting scheme on each 
performance to reflect the type of use (feature, theme, etc.), 
the origin (radio, network television, local television, etc.), 
and the sample time (pertaining to network prime time vs. 
non-prime time). 

The next stages of the distribution strategy route this 
information into the check creation phase for final payout 
instructions. Oftentimes, though song titles are known and 
corresponding writer and publisher information is provided, 
the physical delivery of royalty checks is hampered by a non- 
current address, a legal hold such as a tax lien or judgment, 
or an estate issue, such as identifying the rightful heirs to a 
deceased member's royalty earnings. Such research 
requirements are often overlooked when broadly describing 
the role of a collective organization. 

The concluding stage of the distribution process involves 
excruciating detail to record the final results and to convert 
the weighted performance recognition into available dollars 
for distribution. The incredible attention to detail and 
economic logic cannot be over stressed. 


The music performing rights societies have a unique 
advantage over the multimedia industry in that they can 


IP Workshop - Massarsky: ASCAP, BMI and SESAC 

Page 10 of 12 

readily measure copyrighted product without placing an 
onerous burden on the user. As an example, ASCAP can 
conveniently tape a radio station without the knowledge or 
cooperation of the station. BMI can request a station log 
requiring some licensee intervention but typically available in 
some form for another business purpose. SESAC's new 
contemplated system will provide total data security because 
the sampling target will be unaware of the data collection 

However, retrieving copyrighted uses in a multimedia 
environment poses many hazards. The scope of effort is 
demonstrably greater and will probably require some 
significant level of cooperation from the end user. 

In spite of these fundamental differences, a study of the 
music rights organizations may still be instructive as a 
primitive first step in organizing a cohesive copyright 
allocation strategy. For example, identifying copyrighted 
works in use is sometimes a burden for the music performing 
rights societies. As mentioned in an earlier section, all three 
organizations hold license agreements with non-broadcast 
entities such as bars, grills, hotels, dance halls, skating rinks, 
arenas, stadiums, conventions and expositions, fraternal 
organizations, etc. 

Unlike their broadcast counterparts, these general license 
establishments create a more difficult challenge in monitoring 
music usage. ASCAP has long argued that the operational 
costs for direct monitoring of these establishments would be 
prohibitive. Yet, a significant portion of ASCAP's gross 
revenues are attributable to general licensing venues and 
thus cannot be overlooked in the royalty allocation process. 

ASCAP employs a feature factor proxy to distribute royalties 
collected from general license establishments. ASCAP's goal 
is to predict the content of the music being performed in 
these establishments based not on direct measurement but 
rather on other available sources of information. The difficulty 
resides in the lack of congruence among the various general 
license types. The mix of music featured in a bar may vary 
depending on the nature of the bar's clientele. Therefore, 
ASCAP relies on its. sampling measurement of all radio and 
television performances to encourage a fair mix; accordingly, 
it allocates the general licensee revenue pro-rata to its 
existing allocation of radio and television license fees. 

ASCAP provides one further distinction in that the proxy only 
involves the sampling of feature performances, those uses 


IP Workshop - Massarsky: ASCAP, BMI and SESAC 

Page 11 of 12 

that are the principal focus of a radio listener or television 
viewer's attention. Most radio performances are classified as 
feature uses while a camera focus on a singer or singers is 
required for feature credit on television. 

ASCAP awards its highest credit valuation to features. All 
other types of uses such as theme, background, jingles, etc. 
are allocated a fractional value of a feature use. 


Multimedia organizations, as presently envisioned, will 
require a much broader and more intensive effort for 
copyright management than has arisen in the public 
performance arena. The sheer volume of transactions will 
dwarf the traditional information boundaries of ASCAP, BMI 
and SESAC combined. The complexity of multiple 
administrations for different copyright constituencies has little 
parallel in the world of music licensing. 

However, ASCAP, BMI and SESAC still prove to be the 
guiding working example of large-scale copyright 
management initiatives. Their development of license 
strategies is immensely useful in analyzing the pricing 
components of multimedia services. Their system 
organization provides useful insight into the inner workings of 
a massive copyright administration system geared to protect 
copyright holders. 

ASCAP, BMI and SESAC provide a tremendous historical 
basis from which to evaluate multimedia licensing. Their vast 
electronic warehouses of song titles, their aggressive 
approach to licensing access rather than transaction, and 
their collective ability to establish elaborate distribution 
mechanisms were all precedent-setting. Music copyright 
collectives are likely to represent the singularly best 
approach for guiding multimedia licensing and distribution 


Barry M. Massarsky, a consulting economist holding 
expertise in copyright-related industries, was formerly 
ASCAP's Senior Economist. He currently serves as 
economic counselor to SESAC, as consultant to the 
Recording Industry Association of America (RIAA), and as 
economic counsel in litigation-related music licensing 
matters. Mr. Massarsky's consulting practice is based in New 


EP Workshop - Massarsky: ASCAP, BMI and SESAC 

Page 12 of 12 


Barry Massarsky 

Barry M. Massarsky Consulting 

1120 Ave. of the Americas, Ste. 4100 

New York, NY 10036 

© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 



IP Workshop - Phillips: Meta-Information 

Page 1 of 1 1 

Sponsorec by: 

A & 



Coalition for Networked Information 

About CNI 

Task Force 





Site Map 

Search our she 

Meta-Information, The 
Network of the Future and 
Intellectual Property 

by Prof. Kenneth L Phillips 


Information is present when a more informed decision 
between two equally probable events can be made. 
Information loses half its value in an information half-life, 
which is shortening as the velocity and bandwidth of 
information flows increase. The tremendous economic 
incentives to collect and synthesize information about the 
use of information must be balanced against possible threats 
to individual privacy. 

The nature of information itself has changed fundamentally, 
as a result of advanced networking technologies, and in 
ways which will require the development of novel concepts 
and approaches to the protection of intellectual property. 
Technologies of telecommunications are never content- 
neutral, rendering the content/conduit distinction a legal 
fiction. As a result of these technological changes, new forms 
of information will develop, and along with them, increased 
incentives to sell these new forms, often complicating the 
development and enforcement of privacy and intellectual 
property concepts. 

Although even a cursory review of the trade press will reveal 
considerable debate over the ''future of the network," I feel 
secure in setting forth a few planning assumptions which I 
feel are not contentious: 

1 . The Network is moving away from the dedicated paths 


IP Workshop - Phillips: Meta-Information 

Page 2 of 11 

typical of circuit switched technologies, at all bandwidth 
levels, and in the direction of ^virtual" switched 
environments, routing information on a per-cell or per- 
packet basis. 

2. Switching will take place at the packet level, though it is 
more difficult to predict whether variable length, fixed 
length, synchronous, asynchronous or isochronous 
formats will be first choice. 

3. Packet processing at the baseband level will be at such 
a rate as to render transmission and switching of cells 
and packets of compressed video and other multimedia 
data practicable, on a real-time, low latency basis. 

4. Dynamic bandwidth allocation will become a 
fundamental feature of integrated networks of the 
future, as contrasted with the deterministic time division 
multiplexing methodologies typical of today's digital 
networks, used largely for highly predictable voice 
traffic. Application sectors experiencing the highest 
rates of future growth produce traffic characteristics 
which are intensely "bursty", where demand fluctuates 
drastically from millisecond to millisecond, and where 
peaking is not highly predictable (i.e., not Poisson-like). 

The interconnection of networks on a global level has 
resulted in an amplification of the spikes in traffic 
brought on by natural disasters, political change, and 
fundamental global financial trends in foreign 
exchange, rare metals, and international arbitrage. 

5. The variance in traffic arrival rates will grow further as 
demand rises for the simultaneous delivery of bit 
streams of information, ranging from the traditional 
56/64kb representative of basic voice telephony, to 
155Mb/s for high definition television, and as these 
technologies come on line. 

These basic changes in user requirements have dictated the 
development of cell relay switching methods using fixed 
length cells and Asynchronous Transfer Mode (ATM). Using 
fixed length cells running at heretofore unencountered 
speeds enables the network to basically utilize the benefits of 
the Law of Large Numbers to make the arrival distribution 
more predictable. Fixed length cell structure allows the 
pipelining of applications, further smoothing the mean arrival 
rate curves-ultimately increasing economy. 


IP Workshop - Phillips: Meta-Information 

Page 3 of 11 

Although these technologies will require solution sets to 
intellectual property issues having characteristics unlike 
anything we have developed in the past, the basic problems 
are surprisingly old. 

While the East Coast of the United States was experiencing 
the "storm of the century" a couple of months ago, I had the 
good fortune to have been working in Europe. One morning, 
while taking the train from Zurich to Basel, I purchased a 
copy of the Herald Tribune and noticed what struck me as a 
rather odd headline to deserve placement across the lower 
half of the front page. It chronicled the decommissioning of 
an M Elite French Army Squad", which has had as its principal 
duty the transmission of packets of information since nearly 
the time of Julius Caesar.[H This battalion saw its most 
heroic hour at the Battle of Verdun, in 1916, when its 
members carried messages back to base through poison gas 
appealing for assistance. Yet despite such bravery, this most 
recent announcement was but the fourth time in French 
history that this unique group has been threatened with 
dissolution. In the past, fear mounted that the messages 
would be intercepted and the identities of the senders, as 
well as the content, disclosed to those who would sell or 
otherwise pass such information to the enemy. 

Back in the late 1970's, while completing graduate school, I 
worked for the United States District Court for the Southern 
District of New York assisting the Court in criminal cases 
having complex backgrounds, often involving conflicting 
expert testimony. I remember a case in which the FBI, in 
attempting to locate a terrorist who had allegedly blown up 
microwave relay towers, visited the local public library and 
asked the staff to compile a list of patrons who had borrowed 
books on such subjects as making explosives. The polite 
ladies refused, arguing that such information about who 
sought information on a particular subject was private. The 
federal government sued, arguing that since the library was 
funded from public monies, its records were as public as the 
books it loaned. The government initially lost, but appealed 
and won. The matter was then joined by the ACLU and other 
groups, and again appealed, overturning the appellate court 
decision. The court finally held that absent a disclosure 
statement to the contrary, patrons of libraries have a 
reasonable expectation in the form of an implicit contract or 
guarantee that such information will not be sold or otherwise 
disclosed without their permission, except where a court of 
jurisdiction grants a warrant, which strangely, in the instant 
case, was not sought by the law enforcement organization. 

My purpose in telling you these things, which on the surface 


IP Workshop - Phillips: Meta-Information 

Page 4 of 1 1 

may strike you as unrelated to the subject of this meeting, is 
to alert you to a new form of information which, while not 
entirely new, becomes both more readily available and very 
much more valuable as a result of, and throughout the digital 
age: meta-in formation, or information about the use of 
information. Indeed, as will be seen shortly, this new form of 
information has the potential to alter pervasively the nature of 
some of our largest industries, such as telecommunications, 
retail, and finance, not to mention the enormous inducement 
it could provide to breach personal privacy in ways totally 
unheard of in the past. In addition, while both the legal and 
regulatory communities will have to revise their statutes and 
rules significantly in order to provide adequate protection and 
enforcement of intellectual property rights, history clearly 
teaches that we should not wait for changes to take place in 
these areas. Both federal regulatory and intellectual property 
law lag years behind the introduction of technologies altering 
the powers of those who use them, regardless of their 
intentions and motivations. 

Elsewhere, I have argued that the proliferation of meta- 
information, coupled with advanced telecommunications 
technologies, has profound implications for those whose 
notion of political sovereignty includes operating so-called 
M closed societies". Perhaps the most lucid discussion of this 
dynamic may be found in The Twilight of Sovereignty, [2] an 
exceptional volume authored by Walter Wriston, Citicorp's 
former Chairman. 

In order to understand the dynamics of meta-information, it is 
first necessary to recognize the basic unit of information, 
which I like to call the infon t \3\ a term first used by Keith 
Devlin. Though a more formal mathematical definition is 
possible, for our brief purposes suffice it to say that 
information is present if and only if the presence of 
information aids one in making a decision between two 
equally probable choices. Such a definition establishes a 
distinction between data and information. For example, the 
statement "We are at the Kennedy School" surely contains 
data, but not information, since it is reasonable to assume 
that everyone here knows where they are. An infon, 
therefore, is a basic unit of information and by definition must 
have some value, though at this juncture we have not agreed 
on how information should be valued. 

If we concede that information exists and that its basic unit 
may be called an infon, and that it has at least some minimal 
value, then in order to understand what must be done to 
protect that value, we must first look at the dynamics 
affecting value. These dynamics have changed significantly 


IP Workshop - Phillips: Meta-Information 

Page5 of 11 

at the hands of technology, and telecommunications in 

Perhaps the most impressive aspect of what has gone on in 
the technology of telecommunications in recent years is the 
increase in both the rate and the bandwidth at which 
information is transmitted, switched, processed and then 
sometimes retransmitted. It is generally assumed that the 
acceleration of information transfer rates to the speed of light 
minus some ever-decreasing variable is for the good. I shall 
hold true to my promise to the conference chair to leave the 
so-called M policy" issues for another time, but would like to 
remind you, through the use of a riddle, that these questions 
are more complex than they appear at first blush. The riddle I 
use in class is, M What do a greengrocer in the days prior to 
refrigeration and the modern information manager have in 
common?" The answer, of course, is that both are dealing 
with a terribly fragile commodity with a very short shelf life. 
Those who earn their keep from the sale of information in 
many ways have their lives made more difficult by the 
acceleration in velocity and bandwidth. For example, not 
many years ago, one could sell a quotation service offering 
the spot price of chromium, which is principally traded out of 
Zaire, on the London, New York or Zurich markets, based on 
transactions occurring 24 hours earlier. Today, such data 
has no value, because trading desks are linked to one 
another via broadband networks operating at SONET rates. 
Within a couple of seconds the latest spot price appears 
updated on electronic spreadsheets seen on hundreds of 
trading screens in over a dozen countries. Not only are 
traditional opportunities for spread-based arbitrage 
significantly reduced, but the base prices are subject to 
drastic fluctuations due to the simultaneous presentation of 
infons connected with either related metals, industries which 
are high consumers of chromium, or political events affecting 
Zaire. All of this sort of information is now available 
essentially at the speed of light. 

The value of an infon in this sort of environment becomes 
critically related to the amount of time that has elapsed since 
the receipt of the most recent infon dealing with the same 
matter. Accordingly, I would argue that it now makes sense 
to speak of information or infon half-lives: a measure of a 
quantum of time in which a given infon loses 50% of its 
value. Indeed, when it has lost 100% of its value it no longer 
constitutes information, since it can play no role in assisting 
one with the classical choice between two equally probable 

These notions are simple and I hope clear, and came to my 


IP Workshop - Phillips: Meta-Information 

Page 6 of 1 1 

mind as meaningful analogies to things I learned as a 
graduate student in physics. Information, it seems to me, 
suffers from the classical paradox of being considered to 
behave simultaneously as a wave-like phenomenon, and as 
discrete entities or particles/commodities of some kind. This 
is why most businessmen, with a few interesting exceptions, 
have such a hard time figuring out how to sell it. 

What stands to change this somewhat is the advent of such 
techniques of information transfer as Asynchronous Transfer 
Mode (ATM), where the advantages of fixed cell structures 
on network operation render it almost certain that high-level 
infons will require more than one cell or packet. Indeed, 
under the current wisdom, information is packaged into fixed- 
size cells of 53 octets. Cells are identified and switched 
throughout the network by means of a label in the header. 
ATM allows bit-rate allocation on demand, so the bit rates 
can be selected on a connection-by-connection basis. The 
actual channel mixture at the broadband interface point can 
change dynamically on very short notice. Theoretically, ATM 
supports channelization from low kb/sec. up to the entire 
payload capacity of the interface, minus some small 
overhead factor. 

The ATM header contains the label, which is comprised of a 
Virtual Path Identifier (VPI) and an error detection field. Error 
detection in ATM is limited to the header alone--a mixed 
blessing. Further content-based error correction takes place 
at the periphery of the network, within applications running 
on hosts and their interface nodes. The ATM cell format for 
user, as opposed to bearer, network interfaces is specified in 
CCITT Recommendation 1.361. The header, as usual, is 
transmitted first. However, inside the octet bits are sent in 
decreasing order, starting with bit 8. But octets are sent in 
increasing order, beginning with octet 1 . (The network node 
interface cell M NNP is identical to the layout in Figure 1 
except that the VPI occupies the entire first octet rather than 
just bits 1 through 4.) 

The ATM Cell Fields consist of the following: 

Generic Flow Control (GFC) Field. 

The 4-bit field allows encoding of 16 states for flow control. 
No standardization has yet occurred for coding values. The 
CCITT is presently considering several proposals. 

Routing Field (VPI/CV) 


IP Workshop - Phillips: Meta-Information 

Page 7 of 11 

24 bits are available for routing: 9 bits for the VPI and 16 for 
the VCI (Virtual Channel identification). Except for 2 reserved 
codes used for signaling, and VCI and for indicating general 
broadcast, the encoding methodology has yet to be set. This 
is very important, for reasons which will become clear 

Payload Type (PT) Field. 

Two bits are available or Payload Type identification, 
differentiating user information payloads from network 
information. In user information cells, the payload consists of 
user information and service adaptation information; in 
network information cells, the payload does not form part of 
the user's information transfer. 

Cell Loss Priority Field. (CLP). 

If the CLP field is set (CLP value is, 1 .), the cell is subject to 
discard, depending on network conditions. If the CLP is not 
set, and the value is 0, the cell has a higher priority rating. 

Header Error Control Field (HEC). 

This field consists of 8 bits and is used for error management 
of the header itself. 

Reserved Field. 

This field, consisting of 1 bit, is for further enhancement of 
existing cell header functions yet to be specified. 


Since large numbers of multiple cells are going to be 
required in literally all applications, and ATM and related 
technologies are not circuit switched, identification and 
addressability will have to be handled on a cell-by-cell basis. 
Indeed, such addressing information, regardless of whether it 
references dedicated virtual circuits or user identification 
numbers, constitutes in its own right infons, or what I have 
recently discovered is information for which some parties are 
willing to pay a great deal. 

For example, with the implementation of both the Line 
Interface Data Base (LIDB), justified to achieve 800-number 
portability for customers between long distance carriers, and 
the CCITT Signaling System VII (SS-VII), it is now possible 


IP Workshop - Phillips: Meta-Information 

Page 8 of 11 

for inter-LATA carriers to generate lists of customers by the 
800 number called. 

In a friendly deposition, the Direct Marketing Association 
(DMA) told the Committee of Corporate Telecommunications 
Users that its members would M be willing to pay $3 per name 
and address for a list of telephone subscribers sorted by 800 
number destination. For example an 800 number associated 
with a hotel charging at least x-amount for a room, or a 
contributions line to a charity or political party." Following 
discovery of this fact, a similar inquiry was made of AT&T: 
How many calls are processed per day, and could such a list 
be compiled. AT&T averred that in excess of 100,000 such 
calls were processed per day, that the exact number was not 
obtainable on short notice, and that indeed, given SS-VII 
capabilities, originating station information was captured and 
could be cross-referenced with customer account files and 
addresses lists printed out. 

Aware of the more recent fact that AT&T is now the second 
largest issuer of consumer credit cards in the United States, 
processing literally millions of transactions per month, I 
sought to determine the value of infons consisting of 
telephone traffic information and credit card purchasing data 
linked by Boolean operands. In other words, what would the 
value to the list brokers (or banks, law enforcement 
agencies, tax collectors, lobbyists, etc.) be of data 
assembled in the new format of lists of people who, for 
example, called a hotel reservations 800 number and also 
spent over $500/month on sports equipment? To my 
astonishment the DMA indicated that if the list had been 
generated within one month of their members receiving it, 
the brokers would pay between the earlier $3 and $7 per 
name. Given the traffic numbers provided by AT&T earlier, 
clearly there exists an opportunity of at least $300,000 to 
$700,000 per day, simply based on the AT&T traffic. 

All of this is just an example, and indeed one which AT&T 
rightly protests, since none of these practices is taking place 
at present. However, the writing is on the wall. Citicorp, with 
a much larger customer base, has used Thinking Machine's 
equipment to develop detailed customer purchasing profiles 
linking telephone numbers, to ZIP codes, to SMSA statistics 
and default rates. AT&T has issued letters of intent to 
purchase and lease similar equipment. Companies will 
eventually be forced to become far more open about such 
policies, just as nation states have had to as technology has 
forced the issue. In so doing, they will also become more 
profitable as a greater sphere of potential consumers of 
meta-information become customers. But so far, few have 


IP Workshop - Phillips: Meta-Information 

Page 9 of 11 

figured this out. Indeed, telephone companies and banks are 
especially covetous of this sort of information. (Just ask a 
telephone company for traffic statistics between various parts 
of a city or state, or a bank for the average number of 
Automated Teller Transactions on a time of 
day/neighborhood by neighborhood basis-all useful 
behavioral data.) 

This phenomenon, of infons describing the use of 
information, constitutes second-order information, what I first 
termed meta-information several years ago. When linked to 
the identity of the user or other classes of information, both 
the theft of intellectual property without the detection of the 
act, and the invasion of personal property become 
increasingly easy. Indeed, I believe that one might adopt the 
potentially draconian means of measuring the technological 
advancement of a given society by measuring how many 
sorts of interconnected data bases such as those containing 
meta-information are required in order to gain the identity of 
any given citizen. Alternatively, in the case of intellectual 
property protection, one would simply ask the same question 
pertaining to detecting the location of some file or piece of 
unique work, be it art, software, or your latest manuscript. 
This will all become most interesting as we move towards 
such future institutions as digital libraries, for-profit image- 
based archives, high-definition audio recording, and the like. 

The solution sets required of these problems are not at hand, 
but do bode of careful and thoughtful consideration of just 
what goes into such things as ATM Cell Fields. In non- 
dedicated route networks and in packetized environments- 
where the packet length is finite and small, resulting in a 
proliferation of transport cells--the identity of owners and 
users of intellectual property becomes far more accessible to 
the casual interloper as well as the professional thief. 

Incentives to obtain meta-information will increase at least 
geometrically as the number of interconnected sources goes 
up arithmetically. Indeed, the value of such information may 
be expected to approach a log function of the number of 
sources. Figure 2. (Courtesy of Privacy Journal) depicts 
basic meta-information flows between major categories of 
data collection in the United States. Clearly this is a booming 
business poised to take off, once the ''Network of the Future" 
becomes perceived as a meta-information engine. Profound 
business, policy, and regulatory issues attend all this 
development. A long-distance carrier may see a contribution 
to revenue from processing a transcontinental call of only 
9.7cents per minute, while the existence of the virtual path 
through the network generates $3 to $7 worth of meta- 


IP Workshop - Phillips: Meta-Information 

Page 10 of 11 

information per call. 

How much is the string of four letters representing the 
Adenine, Guanine, Cytosine, and Uracil (A,G,C,T) bases of 
the DNA found on a particular allele of your 18th 
chromosome worth to you, the police, your bank, or a 
genetics engineering company attempting to clone 
antibodies in order to replicate adaptive or otherwise positive 
immune responses in less healthy individuals? What is the 
meaning of Justice Brandeis' prescient equation of privacy 
with the right to be left alone, in light of these developments? 
I do not believe that there is cause for panic-but there is 
cause for pause and serious thought given these matters. 

Yet again, these are not new issues. In fact, earlier on, in 
mentioning the decommissioning of the French Army 
Division and past concerns over the identity of the senders of 
data, I told the truth, but nof the whole truth. Indeed, in the 
age of meta-information, lies of comission will become 
increasingly simple to spot while the detection of deception 
by omission, without violating privacy, will present some 
uniquely vexing problems. And on that note I close, but not 
before I tell you that all the members of the famous French 
Guard threatened with extinction are pigeons. 


1. International Herald Tribune., No. 34,229, March 18, 
1993., page 1. 

2. Wriston, Walter B. Twilight of Sovereignty, Scribner's & 
Sons, NY, 1992. 

3. Devlin, Keith. Logic and Information. Cambridge U. Press, 
1991, p.11,ff. 

4. Deposition of J. Rankel, DMA, 8/13/89, by CCTU; Reid & 

5. See end notes at conclusion of paper for other related 
papers by this author. 


Kenneth L. Phillips, Ph.D. has been Vice President for 
Telecommunications Policy at Citicorp for 15 years, where 
he is now Of Counsel. He is presently a Professor of 
Psychology at the Graduate Interactive Telecommunications 
Program at the Tisch School of New York University. 


IP Workshop - Phillips: Meta-Information 

Page 11 of 11 

© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


IP Workshop - CUPID: Protocols and Services 

Page 1 of 26 

|D7#DDDlDaUlDanaDaDaDaDDa(a(a(a(a( a2:alal,aixa(af ata»*a<?,a[]D 
a»atQa # a»a£Da # a»a*a*a«a* 

Sponsorec by 


About CNI 

Task Force 





[ Coalition for Netwoiiced Information 

Protocols and Services 
(Version 1): An Architectural 

Consortium for University Printing and Information 
Distribution (CUPID) 


Site Map 

Search our site 


The Consortium for University Printing and Information 
Distribution (CUPID) is sponsored by the Coalition for Networked 
Information (CNI), as an open consortium of Universities, 
supporting the development of distributed, high quality networked 
print services. 

This document proposes an architectural framework for the initial 
set of CUPID protocols and services, to support a range of 
applications. The framework is the basis for detailed functional 
and programming specifications. 


CUPID (Consortium for University Printing and Information 
Distribution) is an informal and open consortium of universities 
interested in the distributed printing over the Internet of finished, 
high-quality production documents. 

CUPID is concerned with the support and management at remote 
sites of most or all of the services performed by the production 
printshop or central reprographics organization of a college or 
university. Achieving this objective will depend upon the 
widespread availability of advanced-function, networked printers 
such as the Xerox Docutech or the Kodak Lionheart, although 
distributed applications may also make use of lesser-function 
networked printers. 


IP Workshop - CUPID: Protocols and Services 

Page 2 of 26 

CUPID has set itself a primary task of defining a suite of 
protocols and services that can be used as the core and 
foundation for a variety of applications (see Figure 1). The 
objective is not to develop software that can support an entire 
application. The objective is to extract from these applications 
that which is common (termed the "Common or Generic CUPID 
Infrastructure"), so as to avoid duplicate and costly development 
and to encourage the use of shared and open protocols. 
Applications developers will be encouraged to make use of these 
protocols and services. CUPID protocols define the interface 
between application-specific functions and generic CUPID 

These objectives support the Consortium's overall goal of 
encouraging the development and deployment of distributed 
publishing applications that nurture a shift from the traditional 
"centralized" publishing model of "print then distribute" to a 
decentralized model of "distribute electronically then view." In this 
context, "viewing" may occur either at the workstation or in 
printed form (CUPID concentrates on the latter), and can 
embrace the just-in-time concept of "print on demand." The 
Consortium endorses such a shift to provide functional and 
electronic alternatives to the centralized manufacturing model 
and its accompanying costs of distribution and inventory, and to 
reduce the delays between information creation and 
consumption, or between information requests and production. 

This document proposes a general architectural framework for 
the initial set of CUPID protocols and services, to be used as a 
basis for the further development of detailed functional 
specifications and programming specifications. Where there can 


IP Workshop - CUPID: Protocols and Services 

Page 3 of 26 

be no confusion in this document, we use the term "CUPID" to 
mean the Consortium itself or interchangeably the suite of CUPID 
protocols and services. 


The following are examples of CUPID applications: 

• A scholarly journal publisher who wishes to distribute a print 
journal electronically for local printing by site licensees. 

• A textbook publisher who wishes to adopt the same model 
allowing local printing by campus stores of all or parts of a 

• An author who wishes to distribute his/her monograph 
directly, bypassing traditional publishing channels. 

• A university press that wishes to use electronic channels for 
distribution of printed material. This could include, for 
example, the distribution of Harvard Business School case 

These and other examples all have common needs, including (a) 
the network delivery of print-ready electronic documents (b) the 
authorization of who is to print or distribute finished documents 
(c) the communication of information as to how the documents 
are to be printed and distributed, including the steps of proofing 
and estimating, and (d) the support of certain business functions 
such as payment for printing services and the specification and 
collection of royalties or other fees. Other functions that are 
required include support for security and for conversion of 
document formats. CUPID aims to provide the protocols and 
services necessary to support these common functions. 

Electronic versus Print, Push versus Pull 

Version 1 of CUPID focuses on the electronic distribution of 
documents that are ultimately intended to be printed, and printed 
in finished form. The Consortium believes that although an 
increasing number of documents will be distributed that are 
primarily intended for electronic viewing at the workstation with 
printing being an incidental side activity (such as printing a few 
pages at a local laser printer), the need will remain the need for 
production printing of many documents where the publisher 
wants to control the total appearance of the finished product. 
Nevertheless, many of the features of CUPID protocols and 
services may also apply to the delivery of electronic documents 
for viewing at workstations. 


IP Workshop - CUPID: Protocols and Services 

Page 4 of 26 

Version 1 of CUPID also focuses on the "push" model of 
operation, in which it is the publisher who initiates a request for 
production of a document. Subsequent versions of CUPID will 
also support the "pull" model, sometimes known as "print on 
demand." In the pull model, a request is initiated by someone 
other than a publisher, perhaps a printshop or a customer. The 
key distinction between push and pull is the relationship between 
the initiator of a print request and the documents being printed. In 
the push model, the initiator (the publisher) owns or controls the 
documents, and presumably has direct access to them. In the 
pull model, the initiator generally must acquire rights and/or 
access to the documents via some mechanism defined by the 
documents' owner(s). Again, much of the Version 1 CUPID 
services and protocols will apply equally to both push and pull 
models, and the architecture is designed to allow reuse of these 
common elements. See Section 6 for further discussion of how 
Version 1 can be extended to the pull model. 


CUPID defines three types of Parties who interact over the 
Internet with two types of CUPID Servers. The CUPID Parties are 
Publishers, who initiate requests for document production; 
Printshops, which produce and deliver the finished documents; 
and Agents who, on behalf of Publishers, perform or certify the 
performance of various actions. The requests for document 
production include, among other items, the contents of all 
documents to be printed and are termed CUPID Printjobs. 

The CUPID Servers are Printjob Origination Servers (or, for 
short, Origination Servers), which receive CUPID Printjobs from 
Publishers and maintain the state of those Printjobs; and 
Printshop Notification Servers (or Notification Servers), which 
hold information about one or more Printshops and receive 
notification of Printjobs submitted for printing at those Printshops. 

CUPID Parties communicate with CUPID Servers by means of 
special CUPID Clients. "Client" is used here as in the phrase 
"Client/Server Architecture." The ultimate recipients of CUPID 
documents, on the other hand, are termed "Customers" in the 
CUPID Architecture (see Section 2). 

CUPID Servers provide a set of generic services which are 
available to all CUPID applications. These services constitute the 
Generic CUPID Infrastructure. CUPID Clients, on the other hand, 
provide Application-Specific Functions, tailored both for the type 
of Party and for a particular application. Thus, one Publisher 
might use a Client specially written for the application of printing 
monthly journals at multiple locations, while another Publisher 


IP Workshop - CUPID: Protocols and Services 

Page 5 of 26 

might use a Client customized for the production of multiple 
versions of a single publication at a given site. Some Publishers 
might use both of these Clients, or perhaps a single Client written 
to handle a variety of applications. 

The relationship between CUPID Parties, their Clients, and 
CUPID Servers is shown in Figure 2. 

The remainder of this document describes the CUPID 
Architecture, including the most important CUPID services, the 
Parties to these services, and the CUPID Servers that provide 
the services. It also describes the structure and some of the 
content of the protocols that will be used to communicate 
between CUPID Clients and CUPID Servers (CUPID Exterior 
Protocols) and among the CUPID Servers themselves (CUPID 
Interior Protocols). This document is not, however, intended as a 
complete or detailed description of either the CUPID services or 
protocols. That task is left to the CUPID detailed-design 
document, which defines all protocols and services at the level 
necessary to allow independent developers to build CUPID 
Clients and CUPID Servers that interact with each other in a 
transparent fashion. 

Parties to CUPID Services 

The CUPID Architecture defines three generic Parties directly 
associated with CUPID services: Publishers, Printshops, and 
Agents . Different CUPID services are available to each Party. 

The names of these three Parties are quite generic in the CUPID 
context and are used in the broadest possible sense. A 
Publisher, for example, could be a researcher who wishes to 
cause a report she has authored to be printed at a number of 
different universities. 


IP Workshop - CUPID: Protocols and Services 

Page 6 of 26 

Customers, to whom printed documents are ultimately delivered, 
are considered to be indirect Parties to CUPID services. The 
names and addresses, for example, of Customers may be 
passed to CUPID by the Publisher's CUPID Client for subsequent 
use by Agents. This limited recognition of Customers applies only 
to CUPID Version 1. Subsequent versions may also extend direct 
services to Customers. 

In more detail, the CUPID Parties are: 

• Publishers, who use application-specific Clients to create 
CUPID Printjobs and place them on CUPID Origination 
Servers. A Publisher is the creator, originator, or owner of 
the document to be printed and subsequently delivered to 
Customers. In Version 1, CUPID presumes that the 
Publisher owns (or has been assigned) any rights required 
by the Printjob (but see the section below on future 

• Printshops, which print documents on (usually) high- 
performance production printers attached to printer servers, 
and perform other activities as specified in Printjobs, 
including delivery of finished printed documents. The 
printers and printer servers are not themselves part of 
CUPID. Instead, Printshops use one or more customized 
Printshop CUPID Clients to interact with both the generic 
CUPID Servers and with the local printers and printer 
servers (see Figure 2). The CUPID Architecture allows 
Printshop systems to be organized in a variety of ways. A 
single program, for example, might perform all the 
Printshop's CUPID Client functions and also act as the 
printer server. Alternatively, several programs running on 
several computers might act as specialized CUPID Clients, 
communicating with a printer server running on yet another 

Each CUPID Printshop is associated with a single 
Notification Server which contains a Printshop Specification 
Record for that Printshop. A Printshop Specification Record 
contains a unique CUPID Printshop ID for the Printshop 
and all relevant information about the Printshop's 

The main function of the Printshop Specification Record is 
to ensure that the Publisher is not requesting services of a 
Printshop that it cannot provide, or cannot provide at the 
desired level of quality. The Printshop Specification Record 
includes such information as which PostScript fonts (if any) 
are supported by the Printshop, the TRC (Tone 


IP Workshop - CUPID: Protocols and Services 

Page 7 of 26 

Reproduction Curve) characteristics of the Printshop's 
printers, and any special production capabilities of the 
Printshop. For example, a given Printshop might not offer a 
"heatset binding" option, in which case the Publisher may 
wish to select a "stapling" option instead. 

The Printshop Specification Record also contains any 
relevant standard pricing information the Printshop wishes 
to advertise, current lead times for common types of 
operations, and so forth. 

A CUPID Printjob received by an Origination Server 
specifies one or more Printshops to print a document by 
indicating the Notification Server that contains the Printshop 
Specification Record associated with each required 
Printshop. It does so by specifying the CUPID Printshop ID. 
This requires that a CUPID Address Map (which could, in 
future versions of CUPID, be an X.500 directory or some 
similar database) be maintained at one or more known 
Internet locations that map CUPID Printshop IDs into the 
DNS (Domain Name System) name of the Notification 
Server on which the Printshop's Specification Record is 
located. Printshop registration thus consists of two steps: 
placing a Printshop Specification Record on a CUPID 
Notification Server and updating the CUPID Address Map. 
Such registration and indirect addressing allows, for 
example, a Printshop to relocate to a different Notification 
Server without rendering obsolete the Publishers' existing 
Clients that create Printjobs referring to that Printshop. 

• Publishers Agents (or just "Agents"), which are third 
parties performing requested activities on behalf of a 
Publisher. Agents are individuals (or individuals acting for 
institutions) who operate according to specifications within 
a Printjob, either carrying out designated activities (such as 
delivering documents or collecting fees) or certifying that 
other activities have been carried out satisfactorily (such as 
by approving page proofs). A single Printjob may refer to 
multiple Agents, specifying which activities are to be 
performed by which Agents. A given Agent may perform on 
behalf of several Publishers, and a given Publisher may 
utilize the services of a variety of Agents. 

An Agent for a given activity, for example, could be a 
campus bookstore distributing documents on behalf of a 
commercial publisher, or a university press acting on behalf 
of another university press. An agent could also be an 
academic department, such as a business school that has 
entered directly into a contractual relationship with, say, the 


IP Workshop - CUPID: Protocols and Services 

Page 8 of 26 

Harvard Business School for local distribution of Harvard 
Case Studies. A publisher could be a commercial publisher, 
a university press, or even an individual faculty member 
publishing directly across the Internet with the assistance of 

Conceivably a Publisher's Agent for a given activity could 
be the Publisher itself. A Publisher's Agent could also be 
the Printshop itself. However, when a Publisher or a 
Printshop is acting as an Agent, they are acting in a 
conceptually separate role. It is also conceivable that the 
Agent and the Customer could be one and the same, but 
again are considered logically separate for purposes of 
defining CUPID. In future versions of CUPID, "Agent 
Specification Records" may be added to the Architecture, 
analogous to Printshop Specification Records, that 
"advertise" the capabilities of registered CUPID Agents. 

Because the CUPID Architecture provides for authentication of 
the Parties to a Printjob, all CUPID Parties must be registered 
within the scope of the authentication system chosen. 
Registration for purposes of authentication is conceptually 
distinct from the registration of CUPID Printshops already 
discussed. The current proposals for Privacy Enhanced Mail, as 
described in Internet Draft RFC's 1113-1115, provide a 
framework for CUPID's authentication-oriented registration 
requirements. Independent of any registration(s) required by the 
CUPID Architecture, it is anticipated that all CUPID Parties- 
Publishers, Printshops, and Agents-may need to have 
contractually or otherwise previously defined relationships 
outside of CUPID. 

CUPID Servers 

The CUPID Architecture defines two kinds of Servers: Origination 
Servers and Notification Servers. These terms refer both to the 
software (in UNIX terms, the daemons) that provides the 
specified services and to the computers upon which this software 
is running. A single computer could, of course, operate as both 
an Origination Server and a Notification Server. 

Communication with and among CUPID Servers utilizes a 
reliable byte-stream protocol such as TCP/IP as a transport 
mechanism. In a TCP/IP-based implementation, for example, 
Origination and Notification Servers would operate on separate 
designated Ports, which would be registered with the Internet 
Engineering Task Force. As Internet protocols evolve, CUPID will 
continue to operate on whatever new transport layer emerges. It 
is also likely that the CUPID Architecture will prove readily 


IP Workshop - CUPID: Protocols and Services 

Page 9 of 26 

implementable on proprietary networks. 

Almost all CUPID activity is centered around the Origination 
Server. CUPID Notification Servers exist solely as a means for 
CUPID Printshops to register their capabilities and to receive 
notification of incoming work. 

Version 1 of CUPID does not provide for a wide-area directory of 
CUPID Printshop capabilities other than what can indirectly be 
obtained through the CUPID Address Map (see section above on 
parties). Future versions of CUPID may utilize emerging network 
information services to "advertise" the identities of CUPID 
Printshops over the Internet. Such a service will allow Publishers 
to "shop around" for Printshops that provide the facilities required 
for a particular Printjob at acceptable terms. 

CUPID will evolve over time. CUPID protocols, however, will be 
defined so that Clients and Servers using different levels of the 
protocols will be able to interoperate to the greatest degree 

Communication among CUPID Servers and Clients assumes that 
the daemons responsible for Origination and Notification Servers 
are constantly running, but that a particular Client may or may not 
be operating at any point in time. Server-to-server 
communication, using CUPID Interior Protocols, is thus 
straightforward (but see below). For Client-Server 
communication, using CUPID Exterior Protocols, there are two 
cases: Client-initiated and Server-initiated. In the case of Client- 
initiated communication, the Client typically connects to the 
Server, requests information and/or issues commands, and 
eventually disconnects. Because the Client can assume the 
Server is always accessible, no special provisions are needed. 
On the other hand, when a Server, wishes to initiate 
communication with a Client (in order, for example, to inform a 
Publisher that part of a Printjob has completed), it is possible that 
the relevant Client is not currently running or not connected to 
CUPID. Such communication needs are managed by associating 
a CUPID Message Queue with each Printjob. The Message 
Queue resides on the Origination Server for that Printjob, and 
accumulates Messages related to the Printjob that are targeted 
for the Publisher, the Printshop, and any Agents referenced by 
the Printjob. A Client connecting to a Server may request the 
accumulated messages for the appropriate Publisher, Printshop, 
or Agent. Future Versions of CUPID may allow Publishers, 
Printshops, and Agents to be notified via electronic mail that one 
or more CUPID messages are waiting. 

Although it is assumed that Origination and Notification Servers 

http : //www. cni . org/docs/ima. ip - workshop/CUPID . html 


IP Workshop - CUPID: Protocols and Services 

Page 10 of 26 

are constantly running, network interruptions and other 
instabilities may temporarily disable communication with a given 
Server. Each Server and Client must therefore be prepared to 
find that any other Server is inaccessible at any moment. The 
amount of time allowed for recovery in such situations will be left 
to developers, along with issues of how such time limits may be 
configured by CUPID system administrators and users. 


To initiate CUPID activity, the Publisher's Client creates a CUPID 
Printjob and places it on a CUPID Origination Server. Each 
Printjob specifies a series of activities, or tasks to be performed 
at one or more CUPID Printshops, and also includes the contents 
of any documents referenced by those Tasks. After placing the 
Printjob on the Origination Server, the Publisher's Client will, in 
general, disconnect from the Server. 

For each CUPID Printshop referenced by the Printjob, the 
Origination Server informs the Printshop's CUPID Notification 
Server that a Printjob is ready. The Printshop receives this 
notification either immediately (if its Client happens to be online 
to the Notification Server at the time) or when it next connects. In 
either case, the Printshop then uses its Client(s) to interact with 
the CUPID Servers to execute the Tasks. The Printshop's Client 
retrieves the specified document(s) from the Origination Server 
and directs the document(s) to the appropriate printer server. The 
CUPID Architecture neither requires nor prohibits the caching of 
text, images, or other information at locations other than the 
Origination Server. This is an implementation consideration. The 
Architecture does require, however, that any such caching must 
be invisible to all Clients and must not violate any of CUPID's 
security provisions. 

Some Tasks are directly performed by the Printshop, and some 
by an Agent; others are performed by the Printshop and certified 
by an Agent. As each Task is performed and/or certified, the 
Printshop or Agent uses its Client to notify the Origination Server 
what has occurred. The Origination Server maintains a Message 
Queue for the Printjob, and these Messages are available to the 
Publisher's Client when it next connects to CUPID (or, if it 
remains constantly connected, in real time). 

To carry out the process summarized above, CUPID Servers 
provide the following services (among others): 

• Workflow Management Services.. These services begin 
with interactions between the Publisher's Client and the 
CUPID Origination Server (resulting in the creation of a 

http://www.cni . org/docs/ima. ip-workshop/CUPID.html 


IP Workshop - CUPID: Protocols and Services 

Page 11 of 26 

CUPID Printjob on that Server); continue by informing the 
Notification Server(s) that a Printjob is available; and 
conclude with the removal of the Printjob and all associated 
control information from the Origination Server at some 
defined interval of time following completion of all Printjob 

CUPID controls the flow of the Printjob in at least the 
following ways: The Origination Server maintains the status 
of the Printjob, including indications of which Tasks have 
been completed. This status can be queried by the 
Publisher, Printshop, and appropriate Agents, and forms 
the basis for CUPID to present a list of "next possible tasks" 
to Printshops and Agents. CUPID also ensures that no 
Task may be marked as complete until any prerequisite 
Tasks have been so marked. 

Part of CUPID's Workflow Management Services is a 
facility by which any of the Parties to a Printjob may send a 
free-text message to any other Party to that Printjob. An 
option on each such message is the requirement that all 
CUPID processing on the Printjob be suspended (at the 
next reasonable breakpoint) until an answer is received and 
the Printjob is "released" by the sender of the original 
message. Such messages may also be used by a Publisher 
to cancel a Printjob, although it should be noted that CUPID 
cannot guarantee the response time to such cancellation 

Yet another feature of CUPID's Workflow Management 
Services is maintaining (on the Origination Server) a log of 
all activity related to the Printjob, complete with timestamps. 
This log may be examined by the Publisher (and, to a 
limited extent, by other Parties) during the progress of the 
Printjob and may be archived by the Publisher as a 
permanent audit trail. The log may also be used for system 
recovery purposes (see System Services below). 

• Authentication and Access Control Services. CUPID 
Servers will have the ability to authenticate the identity of 
Publishers, Printshops, and Agents. CUPID Servers will 
also authenticate each other. CUPID limits all Parties and 
Servers to only those activities each is permitted to carry 

• Encryption Services. Within CUPID, all Client-Server and 
Server-Server communications will be end-to-end 
encrypted, using a suitable public- or private-key system. 
One particular encryption system will be selected and 


IP Workshop - CUPID: Protocols and Services 

Page 12 of 26 

described in the CUPID detailed-design document CUPID 
Origination and Notification Servers will offer the option of 
storing local information in encrypted form as well. The 
need for local encryption will depend on whether a 
particular CUPID Server is under the complete 
administrative control of the relevant Party or is, instead, a 
shared system. As a general design goal, CUPID provides 
the ability to ensure that all information related to the 
CUPID System-including the contents of all documents-is 
secure while under CUPID control. 

• Validation Services. CUPID ensures that all Server-Server 
and Client-Server communications conform to CUPID 
requirements. CUPID also ensures that Printjobs do not 
request services from Printshops which those Printshops 
cannot perform (based on the Printshop Specification 
Records stored on CUPID Notification Servers). 

• Document Assembly Services. Publishers are provided 
the ability to submit documents in parts (called 
Subdocuments) for assembly into one or more finished 
products. This facilitates, for example, the submission of a 
single Printjob to produce a variety of documents which 
differ among themselves only in their cover text, or the 
production of "personalized journals" based on customers 1 
registered areas of interest. 

• Image Conversion Services. CUPID provides for the 
conversion of various document image file formats and 
compression algorithms to standard CUPID file formats and 
compression algorithms (see section below on printjobs). 
This conversion is performed by the Origination Server at 
the time of Printjob submission. No further conversion or 
reconversion services are provided by CUPID. This implies 
that applications that make use of CUPID, or Printshops 
themselves, are responsible for any further conversion 
required to print CUPID Documents on a given printer. 

• System Services.. CUPID provides for server backup and 
recovery; audit trails; capacity (local site limits pertaining to 
a particular Server) control, including local duration and 
size storage limits placed on the temporary storage of 
documents and other CUPID files; version control of the 
CUPID software itself; standards control; and other system 
administration functions. 


A CUPID Printjob includes (among other things) the following 


IP Workshop - CUPID: Protocols and Services 

Page 13 of 26 


• An ordered sequence of zero or more Subdocument Files, 
each of which is a self-contained and printable 
Subdocument. The case of zero Subdocuments anticipates, 
for example, a Printjob whose only purpose is to obtain an 
estimate based on page count and other Printjob 
specifications that are independent of the contents of the 
document(s) that will eventually be printed. 

The acceptable formats for CUPID Subdocument Files are: 

o - TIFF, optionally compressed with either CCITT 
Group 3 or Group 4 (using the recently-adopted IETF 
image format standard); and 

o - PostScript Level 1 or Level 2, 

o - For CUPID Version 2, support for SGML-encoded 
documents may be added. 

It is not required that all of the Subdocument Files be of the 
same format, but each must be in print-ready (rather than 
make-ready form, and each must be self-contained and 
self-defining. Additional information about the nature of the 
File (such as its optimal Tone Reproduction Curve) may 
optionally be included. In the case of Files in PostScript 
format, CUPID takes note of the fonts used and verifies (by 
means of the Printshop Specification Record) that these 
fonts are, in fact, supported by the target Printshop. 

• One or more Printjob Orders(a\so called Orders). Each 
Order asks that a single CUPID Printshop carry out a set of 
Tasks, resulting in the printing of a single Document. 
(Orders, Documents, and Tasks are described in more 
detail below.) The different Orders in a given Printjob may 
specify different Printshops and different sets of Tasks. 
When a complete Printjob has been placed on an 
Origination Server, CUPID so informs the Notification 
Servers associated with all of the Printshops referenced by 
Orders in that Printjob. 

The above two Printjob components are created and placed on 
the Origination Server by the Publisher's Client. In addition, the 
CUPID Origination Server itself creates and maintains Printjob- 
related information, including: 

• Status Information, indicating the progress of the Printjob 
as a whole, as well as the progress of each Printjob Order; 


IP Workshop - CUPID: Protocols and Services 

Page 14 of 26 

• A Message Queue, containing messages transported by 
the CUPID System which are to be delivered to CUPID 
Clients operated by Publishers, Printshops, and Agents. 
These messages may be generated by internal CUPID 
System activity, or may result from interactions with 
application-specific Clients. 

Notification Servers are informed that a Printjob has been 
submitted only when the Printjob is complete on the Origination 
Server (in particular, all Subdocument Files must be present) and 
when all CUPID validity checks and conversions have been 
successfully performed (including confirming that there is a 
match between the requested operations and the capabilities at 
the target Printshop(s)). The Printjob remains on the Origination 
Server for some amount of time after all Printjob-related activity 
has been completed (that is, all Printjob Tasks have been 
completed), although the Publisher may explicitly purge a 
Printjob or have its retention period extended. 

As with all other CUPID Printjob components, the Status 
Information related to a CUPID Printjob resides on the 
Origination Server. Status changes are recorded by the 
Origination Server based on information and commands received 
from Notification Servers and from CUPID Parties (via their 

A CUPID Printjob also includes a Header, which contains a 
unique CUPID Printjob identification Number (CPJIN) which is 
generated and assigned by the Origination Server at the time the 
Printjob is submitted. It is presumed that all internal CUPID 
Printjob-related communication will use the CPJIN as key. The 
Printjob Header also contains the following items: 

• Publisher ID; 

• Date and time submitted; 

• Job Name, used for Publisher identification purposes, not 
necessarily the same as the Document title; 

• Job Limits (optional), used to extend or reduce the default 
Printjob retention period on the Origination Server; 

• Security Keys (if and as required); 

• General free-text comments, intended to be seen by all 
Parties working on this Printjob. 



IP Workshop - CUPID: Protocols and Services 

Page 15 of 26 

The basic unit of CUPID functionality is called a Printjob Order, 
several of which may appear in each CUPID Printjob. Through a 
Printjob Order, a Publisher may designate all of the operations, 
features, and options required to produce a final-form document 
at a specified Printshop, including (for example): 

• what document is to be printed (indicated as a selection of 

• which Printshop is to do the printing; 

• how the printing is to be done, including number of copies, 
binding, paper color, cover stock, etc.; 

• what, if any, pre-printing steps are required, such as 
estimation, proof-copy creation, color selection, etc.; 

• how and to whom the resulting output is to be distributed. 
This also includes, for example, identification of an Agent 
acting as the immediate recipient of the document (such as 
the campus store or the library) as well as distribution lists 
of ultimate Customers (for example, a list of journal 

• how payment is to be collected, including Job Accounting 
(payment to the Printshop for work performed) and 
Customer Accounting (collected by a designated Agent on 
behalf of the Publisher, and which may include royalty 
payments). This is discussed further in the section below on 
future extensions; 

• what step(s) may not proceed until some previous step(s) 
have been explicitly evaluated and certified by some 
authorized Agent and, for each such case, the identity of 
the authorized Agent (e.g., the final print run must wait for 
approval of a proof copy). 

A Printjob Order contains a Header (see below) and the following 
two items (among others): 

• a single Document, composed of a designated sequence of 
Subdocument Files; 

• a set of one or more Tasks, called a Task List, where each 
Task specifies: 

o - a CUPID Operation; 

o -an Object; 


IP Workshop - CUPID: Protocols and Services 

Page 16 of 26 

o -Operation Specifications; 
o -an Agent; 

o -a Prerequisite Task List. 

Of the above items, all but the CUPID Operation are optional. 
That is, a CUPID Task must specify an Operation, and may in 
addition specify any or all of the other four elements. 

The Printjob Order Document is represented as a Publisher- 
specified sequence of zero or more of the Printjob's 
Subdocument Files. It is legitimate for a particular Subdocument 
File to appear in this list more than once. The most likely CUPID 
Printjob Order will simply request the production of some number 
of copies of the Document. To support requests involving less 
than the complete Document (such as for proofing purposes), 
arbitrary lists of Subdocument Files may also be used as Objects 
of Operations, as described below. 

The CUPID Architecture design allows all lists of Subdocument 
Files to be represented as sequences of integers. Each integer 
would be interpreted as the index into the sequence of 
Subdocument Files in the current Printjob, all of whose 
Subdocument Files reside on a single Origination Server. This 
design allows the CUPID Architecture to expand easily by 
generalizing the definition and use of Subdocument Files. For 
example, in the next Version of CUPID, Subdocument Files might 
be redefined to be (optionally) pointers to Subdocuments, rather 
than the actual contents of the Subdocuments. These pointers 
might refer to files outside of CUPID, and might also include keys 
or other access-control information. Such a generalization would 
facilitate CUPID f s inclusion of the "pull model". 

The Task List in a CUPID Printjob Order specifies the activities 
that the Publisher wishes the Printshop to carry out, any 
sequencing relationships among the activities that the Publisher 
wishes to impose, and all other details related to these activities. 
Each Task in the Task List identifies one such activity, called a 
CUPID Operation. Examples of CUPID Operations include 
"provide estimate", "print", "prepare proof, and "distribute 
output". The full set of CUPID Operations is given in the CUPID 
detailed-design document. 

In addition to all of the predefined, built-in CUPID Operations, the 
Architecture allows for application-specific Operations, whose 
meaning has been separately negotiated by the relevant Parties, 
but whose semantics are unknown to the CUPID System. These 
application-specific Operations will be generated and interpreted 


IP Workshop - CUPID: Protocols and Services 

Page 17 of 26 

by a compatible suite of Publisher, Printshop, and Agent Clients 
that are tailored to a particular application. The purpose of these 
application-specific Operations is to allow CUPID to transmit 
Tasks whose meaning is unknown to CUPID; responsibility for 
validation is left to the application-specific Clients. If, for example, 
the predefined CUPID Operations did not include "fan-fold," a 
suitably constructed pair of Publisher and Printshop Clients could 
provide for fan-folding as an application-defined Operation. 

Some CUPID Operations require an Object, which is either the 
Complete Document or else a list of Subdocument Files. Some 
Operations require (or allow) a set of Operation Specifications 
(Opspecs), such as deadlines, printing instructions, or a list of 
recipients for distributed output. Examples: 

• Operation: Print Proof 
Object: Subdocuments 2 and 5 
Opspecs: [optional; omitted] 

• Operation: Print 
Object: Document 

Opspecs: 20 copies; stitch left; light-blue legal-size paper; 
delivery required by November 30, 1992 

• Operation: Deliver 
Object: Document 

Opspecs: {list of Customer names and addresses} 

• Operation: Bill 

Opspec: 123456789 (Publisher's account number) 

The CUPID detailed-design document indicates which 
Operations require Objects and which require and allow 
Opspecs, and also describes the content of all Opspecs. 

Some Operations require (or allow) an Agent, which is generally 
a person or other entity designated either to carry out the 
Operation or to certify that the Operation has been satisfactorily 
carried out. Examples: 

• Operation: Approve proof 


IP Workshop - CUPED: Protocols and Services 

Page 18 of 26 

Agent: John Smith (local publisher's rep) 

• Operation: Charge customers 
Opspec: $0.20/copy 
Agent: Cornell Campus Store 

• Operation: Collect royalty payments 
Opspec: $0.01 /copy 

Agent: University of Michigan Library System 

For each operation, the CUPID detailed-design document 
indicates whether an agent is required or optional and the 
relationship of the agent to the operation. 

So as to allow the Publisher to indicate that certain Operations 
may not be performed until other Operations have been 
successfully completed, each Task in the Task List may 
optionally include a Prerequisite Task List {PTL). Impossible sets 
of PTLs and other PTL-related inconsistencies will be recognized 
by the Origination Server, causing rejection of the associated 
Printjob. CUPID will refuse to record a Task as "complete" until 
all of the Tasks in its PTL have been so recorded. 

While PTLs allow the Publisher to impose certain constraints on 
Task sequencing, the CUPID System itself "knows" that some 
sets of Operations can only reasonably take place in certain 
sequences. Thus, for example, if a Task List contains both 
"prepare proof and "print" Operations, CUPID will not permit 
"print" to be marked complete until "prepare proof has been so 

The Printjob Order Header contains a unique CUPID Printjob 
Order Number {CPJON) which, like the CPJIN, is generated and 
assigned by the Origination Server at the time the Printjob is 
submitted. The CPJON is simply the CPJIN suffixed by an 
integer indicating the index of the Order within the Printjob. As 
with the CPJIN, it is presumed that all internal CUPID Order- 
related communication will use the CPJON as key. The Printjob 
Order Header duplicates the Printjob-identifying information from 
the Printjob Header (Publisher ID, date and time submitted, job 
name), and also contains these additional items: 

• Printshop ID; 

• Order Name (used for Publisher identification purposes); 


IP Workshop - CUPID: Protocols and Services 

Page 19 of 26 

• Scheduling, priority, and/or deadline information; 

• Authorization codes, if any (i.e., authorization codes defined 
and known by the Publisher and the Printshop outside of 
CUPID, by virtue of separate contractual or other 
arrangements); and 

• General free-text comments (intended to be seen by all 
Parties working on this Order). 


CUPID Version 1 offers only rudimentary capabilities to support 
such business functions as granting permissions and payment of 
royalties. These and related functions are assumed to be 
performed "out-of-band." Version 1 does support the 
transmission of information related to these functions via the 
appropriate Task definitions, but does not provide any control 

Version 1 does lay the necessary groundwork, however, for 
extensions to support these business functions. As we have 
noted, extending Version 1 from the "push" model to the "pull 
model" mostly consists of replacing Subdocument Files located in 
Printjobs on Origination Servers by pointers to those 
Subdocument Files wherever they may be located outside of 
CUPID. However, these pointers could just as well be to 
"permission servers" that perform gatekeeping functions and in 
turn contain pointers to the Subdocument Files that they control. 
They can also point to corresponding "terms and condition 
servers" that contain business-related information on the 
payment and other conditions governing the printing of the 
associated Subdocuments. Finally, in conjunction with 
information contained in the Printjob Order, they can also point to 
designated "payment servers" that can cause the specified 
royalties or other payments to be charged to particular Customer 

These functions are all kept separate to allow for greater 
generality. For example, one clearinghouse may be able to clear 
a given set of Subdocuments in a manner defined by its 
permission server and terms-and-conditions server. The same 
set of documents could also be cleared by another clearing 
house through a different permission server and terms-and- 
conditions-server. The particular payment server defined will 
normally depend upon both the clearinghouse (which could be 
the Publisher) and on the particular customer being charged. 


IP Workshop - CUPID: Protocols and Services 

Page 20 of 26 

It is likely that a server containing Subdocuments can contain 
pointers to the permission servers that can "clear" those 

The precise definitions of and architectural relationships among 
these server concepts are beyond the scope of this Version 1 
overview. However, the foregoing sketch is consistent with 
Version 1 and the detailed extensions should not be overly 



The outline below summarizes the elements of a CUPID printjob. 
It does not specify the format, sequence, or encoding of those 
elements. Such issues are left to the CUPID detailed-design 

In the outline, brackets indicate an optional item, "(s)" indicates 
an item that may appear 1 or more times (0 or more times if in 
brackets). "*" indicates an item created and maintained by the 
CUPID system, rather than by a CUPID party. 



[Subdocument File(s)] 

Status* (includes Status of all Printjob elements) 
Message Queue* 
Printjob Order (s) 

[Complete Document] 
Task (s) 





[Prerequisite Task List] 


What is CUPID ? 

In 1990 the Coalition for Networked Information (CNI) was 
founded by the Association of Research Libraries (ARL), CAUSE 
and EDUCOM to foster the creation of and access to information 
resources in networked environments in order to enrich 
scholarship and enhance intellectual productivity. CNI now has 
over 150 members, including universities, libraries and 



IP Workshop - CUPED: Protocols and Services 

Page 21 of 26 

technology vendors. 

CUPID (Consortium for University Printing and information 
Distribution) is a working group of CNI, with members including 
Harvard, Cornell, Michigan, Princeton, the California State 
system, Virginia Tech, and Penn State. CUPID members, 
individually and in collaboration with other universities, libraries, 
and vendors, are prototyping applications, and developing the 
architectural framework for CUPID applications. 

The Cupid vision 

The goal of Cupid is to demonstrate the feasibility of distributed 
printing at remote sites of finished, high quality production 
documents. This utility can support a range of functions, including 
custom text production, personal publishing, networked print on 
demand services and rare book preservation. For instance, 
wouldn't it be nice if... 

Custom Text: 

• Professor X's section of English as a Second Language in 
Harvard's summer school met for the first time today and 
conducted a needs assessment in class. Now, at 1 1 am, 
Professor X sits down at her workstation to customize her 
course materials to the class profile. She accesses the ESL 
database over the University network and browses through 
sections of interest, scanning on-line sections of a grammar 
text, associated exercises, and readings from books, 
magazine and newspaper articles. Using a job ticket pulled 
down in a screen window, she modifies her earlier grammar 
selections, adds extra exercises on the use of the 
subjunctive, and chooses readings to complement class 
interests. After an hour, satisfied with the materials for the 
next four weeks, she sends the completed job ticket over 
the network to the printer at Harvard Copy, with instructions 
to print and bind 18 copies, with a table of contents, for 
delivery to the student pick up center by 4 pm. 

Distribute then Print: 

• Professor Y is teaching a class on business ethics at 
Alaska State University. From his desktop computer he 
dials up the Harvard Business School database of cases, 
searches the catalog of the 7500 titles on-line, and 
identifies three cases he would like to use. Opening the 
Cupid job ticket window he orders 50 copies of each case, 
to be printed and distributed from the ASU bookstore 
CUPID printer for the start of class in two days. 


IP Workshop - CUPED: Protocols and Services 

Page 22 of 26 

Rare Book Access: 

• A Ph.D. student in San Francisco accesses Cornell 
University Library on-line catalog. He locates a study 
published sixty years earlier which is a critical reference for 
the next chapter of his thesis, due for presentation at the 
MLA conference next month. The student has neither time 
nor funds for a trip to Cornell and the book is too fragile for 
inter-library loan. However, the librarian has a suggestion: 
the already microfilmed preservation version of the text can 
be converted to digital form, sent over the Internet to the 
library at Berkeley and printed and bound there in book 
facsimile form within 24 hours. 

On-Line Journal Articles: 

• The most recent issue of a leading science journal has a 
controversial article on cold fusion, which teaching fellow Z 
wants to distribute before tomorrow^ lecture in the 
Dynamics and Energy basic sciences course. The journal 
has disappeared from the library shelf. From her 
workstation, she accesses the on-line journal database at 
MIT, finds the article, and orders 150 copies printed at 
Harvard Copy for pick up before the 9 am lecture. 

Personal Publishing: 

• Professor A is editing a Festschrift for the retirement of the 
department chair. Articles by scholars and former students 
who now teach at universities worldwide have been 
circulated for editorial comments over the Internet. The 
completed volume will be published simultaneously at 
Harvard, Oxford University, the Sorbonne and St. 
Petersburg. Professor A assembles the print-ready copy at 
his desktop, and uses the CUPID application to send it to 
CUPID printers at each location for distributed publication. 

What are the benefits? 

The CUPID model promises efficiency and economy in new ways 
of working with information: 

• Lower cost: select and pay for units of text instead of 
multiple expensive textbooks; potentially cheaper than 
offset printing. 

• High quality: improvement over current class notes 
assembled from copies. 


IP Workshop - CUPID: Protocols and Services 

Page 23 of 26 

• Increased productivity: desk-top search, scan, select, 
assemble and send to print. 

• Flexibility: avoid long order lead times; customized texts 
and class materials for specialized needs; instant 
adjustment to unexpected class size changes. 

• Network access to stored digital information. 
Why is CUPID happening now? 

CUPID applications are enabled by several current technology 

• New digital copier/printers that can: 

o - accept and store electronic image input; 

o - accept and store scanned text or graphics; 

o - print at high resolution (up to 600 dpi), or close to 
off-set printing quality; 

o - print at high speed for volume production; 

o - collate, finish and bind for single process production. 

• Networked capability in digital copier technology. 

• Expansion of robust Internet as international connector. 

• Proliferation of LAN's which link desktops to Internet and 
world-wide networks. 

• High end workstations on the desktop. 

What else is needed to fulfil the potential of CUPID? 

CUPID initiatives also depend on the growth of related 
technology services: 

• On-line 

o - data bases; 

o - copyright clearance and royalty agreements; 
o - billing and accounting systems. 


IP Workshop - CUPID: Protocols and Services 

Page 24 of 26 

• Data base search and management tools: directories, 
catalogs, key word searching. 

• Network standards for information distribution. 
How will it work? 

The CUPID architecture outlines new Internet engineering 
standards to define the functional and programming 
specifications common to CUPID applications: 

• Internet-based utility that provides services to enable 
distributed printing. 

• Protocol to send document over network, with job 
instructions and status information. 

• Initial distributed services include: access control; 
authentication; encryption/decryption; images text 
conversion; routing; assembly; job status and resource 

• Future services may include: pointers to remote stored 
documents; end-user desktop assembly of custom 
documents; print-time merge of component materials; print- 
time final edit; etc. 

Where will CUPID take us? 

CUPID is a historic opportunity-The Second Gutenberg 
Revolution-with the potential to transform traditional modes of 

• Distribute-then-print defines new roles/relations for 
publishers, bookstores, copy shops, universities and 

• The global scholar: enhanced collaboration and 

• The author as publisher: individual printing of texts. 

• Learning enrichment: the customized text, just-in-time 

This paper was prepared for the Coalition for Networked 
Information by the CUPID Architecture Subcommittee: 

Scott Bradner 



IP Workshop - CUPID: Protocols and Services 

Page 25 of 26 

Robert Cowles 

( Cornell) 

Jim Ferrato 


Steve Hall 

( Ha rva rd) 

Tom Head 

(Virginia Tech) 

Ted Hanss 


Robert Knight 

( Princeton) 

Clifford Lynch 

(University of California) 

Chaiz*: M. Stuart Lynn 


Anne Margulies 

(Ha rva rd) 

Mark Resmer 

(California State University) 

Lawrence Sewell 

(Virginia Tech) 

Carol M. Taylor 


Jeff Wooden 


Steve Worona 


© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 

□uM€DD(A ( €□ □ Kyy9€d □ (A, □€□ □ Kyy j€D □ Kyy □ 8 □ □ □€ □ 0(A, €lOKyyn€D4(A, 
€»4Kyt>f€nH8!@|an!0TnTf"Q ©□□□□□□PDQD"D%oD§DADADDDDnGDHD„D... 
□ADADODyDt3D6DlDiDiD6DtD0qpD > r'DdD?ntDOD0nGD"DDndD>D%oDxn% 
□jD o nqpLD"njDe0; ; :1AU¥6B R u H " a / y QDDDaDDEDDDiCI|JD[][]KD 
[□uD a D □uD2D}D 1 D"5D*D|D£DnD;DDDTDDD\n-'DuDGn-D'DaD5D... 

□Td£De 7 * Y e\,\z\i" ,, M ,, <"0#)#y#A$Dr$|$e$t3%-%y%e&n&d& o &u , L , » , a(D(s("(a)2) 
k)»*D*Q*y*A+ +Y+ +T,D,:,?,CE I D-)-v-/E-a.7. > j/D/j/ o 0 

4U4£4i4u5l5-5 1 /46D606™6e757,7l8D8k8°9D909™9 1 9 1 /2: :R: :i;7;,;l<rjE>">6?D?n? 3 ? 
y@M@'@aA.A>A A0B!BiB|jB6C;C§ClDDD% 

l6mErriTnan,nBnYn£n"n0oEo oOpDp(pcphp°pTqDgkq»rDrXr"r0sFsts6t"tpt 1 uDuXu- 
uuvEvZv"v(lwnwiw > xDxRxixey2y€y'fzDzjzMz6{?{ {0|D|e|'}D}Q} 



IP Workshop - Nelson: Publishing and Royalty Model 

Page 1 of 4 

Sponsorec by: 

A & 


About CM 

Task Force 





Site Map 

Search our she 



Coalition for Networked Information 

A Publishing and Royalty 
Model for Networked 

by Theodor Holm Nelson 


For some years the Xanadu* project has been planning a 
royalty publishing service for networked documents. The 
proposed system has a number of aspects, some of which 
are subtle. 


"Publication" under most systems of law means making a 
document public. A publisher, then, is whoever, or whatever 
entity, commits the act of publishing. The role of publisher 
corresponds to the paper publisher: that person or entity 
which takes the initiative of publication, receives the profits 
and is sued for the contents. 

(The role of author may or may not be different from the 
publisher; but his or her presence is not formally 
acknowledged within the contractual system; it is assumed 
that the publisher has made appropriate arrangements with 
the author, and it is up to the publisher to pay the author in 
whatever way they have agreed.) 

The role of sen/ice provider is like that of printer and 
distributor in the paper world. The publisher contracts with 
the service provider for the material to be reproduced and 
distributed, just as the publisher now does in the paper 

The role of customer is like that of "customer" in the ordinary 
paper world. But in the paper world, granularity is large: 


IP Workshop - Nelson: Publishing and Royalty Model 

Page 2 of 4 

magazines, books and newspapers are the units of sale. In 
our model, the unit of sale can be as small as one byte. 

Note, of course, that these roles may overlap. In the network 
community we anticipate a customer will often be both 
author and publisher, as well. 


Publishing consists of network storage and delivery of 
documents, voluntarily and explicitly put on line by 
publishers, and delivery by such fragments as customers 
request. Nothing is sent but what the reader asks for. The 
customer pays on a per-byte basis for all published materials 
sent by the server. 


The publisher sets the price-per-byte of the document, or of 
its sections, if they differ. The customer sends for arbitrary 
portions (up to the whole document), paying the royalty for 
each byte transmitted. (Note that other royalty arrangements 
have been mooted during the life of the Xanadu project.) 

Contract between publisher and service provider 

The publisher contracts with the service provider for the 
storage of the document and its sale by arbitrarily small 
fragment. The service provider promises to send the 
royalties for each sent byte. The service provider also agrees 
to forward materials to other service providers as needed to 
provide the service throughout the network. 

In this contract, the publisher also represents that he/she/it is 
the rightful publisher, and further agrees to be responsible 
for any disagreeable consequences under law or tort 
(violations of national security, privacy, copyright, etc.). 
These same understandings ordinarily hold in paper 
publishing, but are not made explicit. 

Contract between customer and service provider 

The service provider agrees to send materials on demand to 
the customer. The customer agrees with the service provider 
to pay for materials sent-both royalties at rates specified by 
the publishers, and delivery fees to the service provider (as 
separately negotiated). The customer further agrees only to 
"fair use" of materials received-a copy for use, printed out if 
desired, and copies for backup~but no further distribution. 


IP Workshop - Nelson: Publishing and Royalty Model 

Page 3 of 4 

Contract between author and publisher 

This is assumed, but not formally required. The author 
agrees to the sale of materials by the small fragment, and to 
the various consequences thereof. 


In our software (still under development), we implement a 
special pointer which we call transclusion, a handy term for 
virtual inclusion by reference across a document boundary. 
A transclusion pointer from Document Y to a paragraph in 
Document Z means that the paragraph is logically and 
virtually a part of Document Y. 


A problem of universal concern is the issue of copyright 
violation (from the point of view of publishers) or the 
restriction of freedom (from the point of view of authors 
wishing to quote other documents). We believe our model 
nicely resolves the two motivational thrusts. 

The transclusion pointer means that any author is free to 
quote any document already published under this system, 
since the publisher of the other document has already given 
contractual permission for sale by fragment. The quoted 
materials are thus purchased automatically by the reader 
from the original publisher at the time of delivery. 


There is little question that publishing with royalty on 
electronic networks will become a principal feature of the 
world of information. Sale only of whole documents is a 
frustrating practice with limited usefulness. Sale by user- 
specified fragment makes transclusion widely practical, 
making both possible and fair to all parties many varieties of 
use which are currently frustrated within the system of 

However, without contractual recognition of the varied 
possible ramifications, many parties may get into difficult 


Nelson, Theodor Holm, Literary Machines 93.7. $25 prepaid 


IP Workshop - Nelson: Publishing and Royalty Model 

Page 4 of 4 

($40 foreign) from Mindful Press, 3020 Bridgeway #295, 
Sausalito, CA 94965. 

Xanadu Operating Company, "Xanadu Hypermedia Server 
Developer Documentation," July 1992. $150 prepaid ($200 
foreign) from Mindful Press, 3020 Bridgeway #295, 
Sausalito, CA 94965. 

Theodor Holm Nelson 
Project Xanadu 
3020 Bridgeway #295 
Sausalito, CA 94965 

* "Xanadu" is a service and trademark for services and 
software of Project Xanadu. 


©2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002. 


IP Workshop - Acronyms List 

Page 1 of 5 

Sponsorec by: 


About CNI 

Task Force 





Site Map 

Search our she 

[ Coalition for Networked Information 

IP Workshop - Acronyms 

ACATS: (FCC) Advisory Committee on Advanced TV 

ACLU: American Civil Liberties Union 

ADAPSO: earlier name for Information Technology 
Association of America 

ASCAP: American Society of Composers, Authors, 
and Publishers 

ASN: Abstract syntax notation 

ATM: Asynchronous transfer mode 

BDE: Base Development Environment (from Transarc 

BMI: Broadcast Music Inc. 

CAD: Computer-aided design 

CARL: Colorado Alliance of Research Libraries 

CASE: Computer-aided software engineering 

CAUSE: member newsletter of the Assn. for Managing 
& Using Info. Tech. in Higher Educ. 

cid: client unique identifier 

CLP: Cell loss priority 

CNI: Coalition for Networked Information 


IP Workshop - Acronyms List 

Page 2 of 5 

CNRI: Corporation for National Research Initiatives 

CPJIN: CUPID printjob ID number 

CPJON: CUPID printjob order number 

CPU: Central processing unit 

CUPID: Consortium for University Printing and 
Information Distribution 

DART: Digital audio recording technology 

DCE: Distributed computing environment 

DES: Data encryption standard 

DLS: Digital library system 

DMA: Direct Marketing Association 

DNS: Domain name system 

DVI: Digital video interactive 

EBR: Electronic bibliographic record 

EDI: Electronic data interchange 

EDUCOM: A non-profit consortium of higher-ed 
institutions which facilitates information research 

EFT: Electronic funds transfer 

FTP: File transfer protocol 

GFC: Generic flow control 

G4Fax: Group 4 facsimile 

HEC: Header error control 

HPC: Act High-Performance Computing Act 

HPCC Act: High-Performance Computing and 
Communications Act 

IBP: Internet billing protocol 


IP Workshop - Acronyms List 

Page 3 of 5 

IBS: Internet billing server 

IC: Integrated circuit 

IETF: Internet Engineering Task Force 

INI: Information Networking Institute 

IP: Intellectual property 

I/O: Input/output 

ISBN: International standard book number 

ISDN: Integrated Services Digital Network 

ISO: Organization for International Standardization 

LCS: Library collections services 

LIDB: Line interface data base 

LMT: Lossless multiresolution transform 

MH: Modified Hoffman 

MR: Modified read 

MMR: Modified modified read 

NIST: National Institute of Standards and Technology 

NREN: National Research and Education Network 

NSA: National Security Agency 

NSFnet: National Science Foundation Network 

NVM: Non-volatile memory 

OS: Operating system 

PC: Personal computer 

PEM: Privacy enhanced mail 

PIN: Personal identification number 


IP Workshop - Acronyms List 

Page 4 of 5 

PSP: Professional scholarly publishing 

PS/WP4 (FCC): Planning Subcommittee/Working 
Party 4 

PT: Payload type 

PTL: Prerequisite task list 

RAM: Random-access memory 

RFT: Request for technology 

RIAA: Recording Industry Association of America 

rid: resource identifier 

RL: Run length(s) 

RMS: Rights management system 

ROM: Read-only memory 

RPC: Remote procedure call 

RRS: Registration and recording system 

RSA: Rivest, Shamir & Adleman (developers of 
encryption standard) 

SESAC: Society of European Stage Authors and 

SGML: Standard graphics markup language 

SIDBA: Standard image database 

SMPTE: Society of Motion Picture and TV Engineers 

SMTP: Simple mail transfer protocol 

SPA: Software Publishers Association 

SS-VII: Signaling system VII 

STM: Scientific, technical & medical 

TCB: Trusted computing base 

http :/A\ 


IP Workshop - Acronyms List 

Page 5 of 5 

TCP: Transmission control protocol 

TCP/IP: Transmission control protocol/Internet 

TP4: Transport class 4 

TRC: Tone reproduction curve 

UDP: User datagram protocol 

VCI: Virtual channel identification 

VLSI: Very large-scale integrated 

VM: Virtual memory 

VPI: Virtual path identifier 

WAIS: Wide-area information service 


© 2002 Coalition for Networked Information. All Rights Reserved. 
Last updated Wednesday, July 3, 2002.