Internet Dreams
`
`Archetypes, Myths, and Metaphors
`
`Mark Stefik
`
`The MIT Press
`
`Cambridge, Massachusetts
`
`London, England
`
`Petitioner Apple Inc. - Exhibit 1044, p. 1
`
`

`
`Third printing, 2001
`© 1996 Massachusetts Institute of Technology
`
`All rights reserved. No part of this book may be reproduced in any form by any
`electronic or mechanical means (including photocopying, recording, or informa-
`tion storage and retrieval) without permission in writing from the publisher.
`
`This book was set in Sabon by Asco Trade Typesetting Ltd., Hong Kong and was
`printed and bound in the United States of America.
`
`Images of the metaphors are by Eric P. Stefik.
`
`Library of Congress Cataloging-in-Publication Data
`
`Internet dreams : archetypes, myths, and metaphors / Mark Stefik.
`p.
`cm.
`
`) and index.
`Includes bibliographical references (p.
`ISBN 0-262-19373-6 (HB), 0-262-69202-3 (PB)
`1. Information superhighway—United States. 2. Internet (Computer
`network)—~United States. I. Stefik, Mark.
`ZA3250.U6I5 8
`1996
`303.48’33-—-dc20
`
`96-28249
`CIP
`
`Petitioner Apple Inc. - Exhibit 1044, p. 2
`
`

`
`Letting Loose the Light: Igniting Commerce
`
`in Electronic Publication
`
`Mark Stefik
`
`Connections
`
`In “The Digital Library Project: The World of Knowbots” in Part 1,
`Robert Kahn and Vinton Cerf ask, “If a thousand books are combined
`
`on a single CD-ROM and the acquirer of the CD-ROM only intends
`to read one of them, what sort of royalty arrangement is appropriate
`to compensate the copyright owners? How would compensation be
`extended for cases in which electronic copies are provided to users?”
`Their questions show how, in 1988, issues ‘about copyright protection
`and payment for usinginformation arose in the context of early CD-
`ROM distribution.
`
`By 1994 copyright issues not only had not been settled,
`they were
`coming to a boil. Laura Fillmore’s effort to build a successful publishing
`business on the Internet reveals the limitations of what was practical in
`May of 1994. Although digital works were being sold on the Internet,
`provisions for commerce were primitive. Furthermore, the ease of copy-
`
`ing digital works had led many people to believe that digital information
`should be free. Fast access to the network had made trading programs or
`other data as easy as mixing songs on audio tape. In short, it had become
`
`much simpler for network users to infringe copyright than to uphold it.
`
`This is the context for the oft-quoted statement by ]ohn Perry Barlow
`
`of the Electronic Freedom Foundation, “Copyright is dead.” Advocates
`
`of free information argue that because you don’t lose the original when
`
`you make a copy of a digital work, there should be no charge for copying
`
`information. The conventional wisdom among publishers in late—1994,
`
`when this article was written, was that digital containers for software
`
`Petitioner Apple Inc. - Exhibit 1044, p. 3
`
`

`
`220
`
`Mark Stefik
`
`were inherently leaky vessels and that no viable solution would ever he
`found. The article suggests, however, a way to sustain commerce for
`those who want to sell information on the network.
`
`Throughout the time I’ve been groping around cyberspace, an immense, unsolved
`conundrum has remained at the root of nearly every legal, ethical, governmental,
`and social vexation to be found in the Virtual World. I refer to the problem of
`digitized property. The enigma is this: If our property can be infinitely reproduced
`and instantaneously distributed all over the planet without cost, without our
`knowledge, without its even leaving our possession, how can we protect it? How
`are we going to get paid for the work we do with our minds? And, if we can’t get
`paid, what will assure the continued creation and distribution of such work?
`
`John Perry Barlow, “The Economy of Ideas”
`
`No problemo.
`
`T-101 (Arnold Schwarzenegger) in Terminator 2
`
`It all depends on whether you really understand the idea of trusted systems. If you
`don’t understand them, then this whole approach to commerce and digital pub-
`lishing is utterly unthinkable. If you do understand them, then it all follows
`easily.
`’
`
`Ralph Merkle
`
`Across many cultures, knowledge and inner knowing are described as
`light. Letting loose the light refers to spreading knowledge in the world,
`typically in written form. Consistent with this metaphor, the period in
`the eighteenth century characterized by a burst of writings in philosophy
`and science is called the Enlightenment. In the present century the meta-
`phor of knowledge as light is both poetic and physically realized. Books,
`pictures, movies, musical performances, and other works can be con-
`veniently represented digitally. With fiber optics, digital works are actually
`transmitted by the shining and pulsing of light.
`The digital representation of works and their nearly instantaneous
`transmission has profound consequences for commercial publishing.
`Three of the fundamental economic factors affecting the publishing
`
`industry—printing costs, inventory costs, and transportation costs-—can
`be drastically reduced. Digital works can be copied at minuscule costs,-
`stored in almost no space, and transported instantly anywhere in the
`world.
`
`Petitioner Apple Inc. - Exhibit 1044, p. 4
`
`

`
`Letting Loose the Light
`
`221
`
`This portability opens up visions of a greater information age. For
`libraries, universal access to the world’s written knowledge is a centuries-
`old vision. Today many libraries have electronic catalogs accessible to
`anyone with a computer. Articles can be delivered to anyone with a fax
`machine. In the technophile’s idealized vision, books and magazines need
`never be printed on paper at all; any digital work could be made avail-
`able to anyone, anytime, anywhere in the world.
`
`However, the dream of universal digital access to high-quality works
`dangles just beyond reach. Such works are not usually available, because
`of publishers’ concerns that uncompensated copying will infringe and
`erode their ability to make a living. History suggests that this problem
`will not go away. Publishing thrives only when it is profitable, and prof-
`itability depends on limiting uncompensated copying.
`The conventional wisdom——based on the way computers are used
`for word processing, electronic mail, and computer networking—is that
`copying digital works is easy and, therefore, inevitable. There appears to
`be a clear, inherent conflict between representing works digitally and
`honoring the commercial and intellectual property interests of creators
`
`and publishers. Fortunately, computers need not be blind instruments of
`copyright infringement. Properly designed digital systems can be more
`powerful and flexible instruments of trade in publications than any other
`medium. The seeming conflict between digital publishing and commerce
`
`is merely a consequence of the way computer systems have been designed
`to date.
`
`The technological means for commerce in digital works are now at
`
`hand. New and unconventional when compared with today’s uses of
`computers, these means will enable us to buy, sell, and lend digital works
`much as we now buy, sell, and lend printed books and other pub-
`lications. They will change the way digital works are purchased and
`delivered and will give consumers access to all sorts of works at any time
`of the day—though not necessarily for free. Consumers will be able to
`
`sample works, borrow them, rent them for nominal fees, and make
`
`copies for friends. Creative people will be able to circulate their works to
`
`networks of friends while earning a reliable living from people who make
`copies of them. This technological system will affect everything from
`digital books to digital television, from digital music to digital video
`
`Petitioner Apple Inc. - Exhibit 1044, p. 5
`
`

`
`222
`
`Mark Stefi/2
`
`games. It will radically change our concepts of digital libraries, digital
`bookstores, digital music stores, digital newspapers, and digital television
`stations. Moreover, any competent technological company will be able to
`implement the required systems.
`Here is a road map to this new land. First, we discuss the history of
`copyright law and the reasons for the widespread, but incorrect, belief
`that works represented digitally will be copied without permission. We
`then describe the technological innovations that can enable and support
`commerce in digital publishing. Finally, we introduce the institutional
`and business challenges that lie ahead. What we require to overcome
`them is the wit, will, and means to create institutions that provide the
`necessary security, convenience, vision, and longevity.
`
`The Origin and Rationale of Copyright
`
`It is harder to be honest than to cheat when copying digital works on
`general-purpose computers. The license printed on the package of most
`purchased computer software authorizes a buyer to load the software
`into one computer and use it there. Getting another legal copy for a
`friend involves driving to the computer store and buying it. It is much
`easier, faster, and cheaper to simply load the same software into another
`computer. Such copying is so private and easy to do that most people do
`. it without thinking, and without guilt.
`Unauthorized copying on computers is not, of course, limited to pur-
`chased software. With a few keystrokes, it is often possible to copy a
`paragraph, an article, a book, or a life’s work without compensating its
`creators or publishers. Nor are unauthorized copying and use new phe-
`nomena. Anyone who ignores the FBI warning message on video tapes to
`make copies for friends infringes a copyright, as do people who copy
`compact discs onto cassettes. As a practical matter, it has not been fea-
`sible to enforce the copyright law in these cases. There are simply too
`many people with recording devices to make rigorous enforcement prac-
`tical or cost-effective.
`It is widely believed that there is no viable technical solution to this
`problem for digital information. John Perry Barlow, a prominent spokes-
`person in the computer industry, says that the idea of patents and copy-
`
`Petitioner Apple Inc. - Exhibit 1044, p. 6
`
`

`
`Letting Loose the Light
`
`223
`
`rights needs to be rethought in the digital age. Information, he argues,
`cannot be contained or owned. It wants to be free. Cyberspace is the new
`
`frontier, and its leaders and pioneers are today’s radical thinkers about
`freedom of information. Barlow suggests, in fact, that we abandon all
`notions of intellectual property and market regulation. This solution was
`tried at least once, and it didn’t work. Apparently, for high-quality works
`to spread in the world people need to be able to make a living from
`creating and distributing them.
`Barlow’s arguments are reminiscent of the intense debates about intel-
`lectual property that took place in France during the French Revolution.
`Like Barlow, revolutionaries argued that ideas cannot be owned and
`should not be regulated. During the revolution, many writers and under-
`ground publishers emerged as civic heroes of public enlightenment by
`arguing against tyranny and for freedom of the press. The revolution of
`the mind, they said, required the dismantling of the laws and institutions
`governing authorship, printing, publishing, and bookselling. Absolutely
`free communication was one of the most precious rights of man. All
`citizens should be able to speak, to write, and crucially, to print freely.
`According to this philosophical ideal, people had a will to know and
`_ should be allowed to read and learn from anything they liked. The wide
`availability of books and the right to publish were seen as keys to this
`spread of knowledge.
`In 1789, the revolutionary government wholly deregulated the press,
`believing that the works of the great writers of the Enlightenment would
`thus be made universally and cheaply available. The writers and pub-
`lishers certainly never expected what actually happened. Instead of works
`of enlightenment, the presses turned out mostly seditious pamphlets and
`pornography. Printers also competed with each other to bring out cheap
`editions of books others had spent money developing. So little money
`could be made producing the good books that quality declined; most
`editions were abridged and contained many errors. Publisher after pub-
`lisher went into bankruptcy and then out of business. The disastrous
`nature of an unregulated press, largely unanticipated in the heat of the
`revolution, became blatantly obvious as the publishing industry fell into
`shambles. The same leaders who had clamored for the freeing of the
`presses came belatedly to understand the folly of their action. In the
`
`Petitioner Apple Inc. - Exhibit 1044, p. 7
`
`

`
`224
`
`Mark Stefi/2
`
`chaos of the unregulated press, some prominent and popular writers even
`stopped publishing; because they could not control the printing of their
`works, they could not make a living by writing.
`In 1793 legislation to restore order to publishing was passed. It rec-
`ognized the rights of authors and grounded the publishing industry in
`the principles of the marketplace, establishing the author as creator,
`the book as property, and the reader as an elective consumer. This law
`reflected a fundamental shift in the Enlightenment perspective, which
`now saw that the widespread creation and publication of creative works
`was better served when the authors could own the products of their
`minds. This history of the treatment of intellectual property in France is
`discussed by Carla Hesse in her book Publishing and Cultural Politics in
`Revolutionary Paris, 1 789-1 81 0.
`
`Today, most people see the infringement of copyright on digital systems
`as unavoidable. In the remainder of this section, we describe the assump-
`tions about computer design behind this belief and argue that we need to
`go beyond conventional ways of thinking to solve the problem.
`Three main factors currently inhibit the development of digital pub-
`lishing:
`(1) the absence of high-contrast, low-power, cheap flat-panel
`displays; (2) lack of an inexpensive and reliable way of handling money
`digitally; and (3) the need for a widely accepted means of accounting for
`the use and copying of digital works. Improvements in technology will
`almost certainly solve the display problem in the next five to ten years.
`Most people see such displays as crucial to making electronic books
`and newspapers portable. They matter less, however, in applications for
`which desktop displays are satisfactory or where displays are not neces-
`sary—such as in transmitting musical works. The second factor—meth-
`ods of handling money digitally, in the form of checks, credit cards, or
`anonymous cash——has recently become the subject of much field exper-
`imentation. Our focus is on the third problem, techniques for commerce
`in what we call digital property rights or usage ‘rights, a generalization
`of the idea of copyright that delineates several kinds of rights besides
`
`_
`copying.
`Some publishers see illicit copying as too big a business risk and do not
`publish in digital form at all. Digital newspapers often leave out impor-
`
`Petitioner Apple Inc. - Exhibit 1044, p. 8
`
`

`
`Letting Loose the Light
`
`225
`
`tant and high-value content such as the pictures or graphics, and con-
`sumers of these lower-quality papers are unwilling to pay much for them.
`The perception of low quality leads to a chicken-and-egg problem in
`which the publishers make little money and consumers have few choices.
`Ironically, publishers of works that need periodic upgrading, such as
`computer software, have found that some leakage increases their cus-
`tomer base, even though it is often reported that there are more unau-
`thorized copies of a program in use than authorized ones. Software
`publishers have decided that the revenue losses of illegal copying are
`affordable, although they lead to unfair billing. Software publishers
`charge all users the same price, regardless of the use to which they put the
`program, arguably overbilling people who use the work infrequently.
`As computers and computer networks have proliferated, the need for a
`better approach to protecting digital works has become more widely
`appreciated. Moreover, as new kinds of works—such as music, video,
`and multimedia works that mix these forms——are now available digitally,
`people from different industries are searching for solutions. Given this
`wide acknowledgment of the need, why have solutions seemed so elu-
`sive? Apparently, we are stuck in a rut, assuming that things must be.
`_done the way they have always been done with electronic mail, word
`processing, and other current applications.
`Conventionally, we use general~purpose computers with general-
`purpose operating systems and general-purpose programs. The computer
`industry, grounded on the premise that computers can do anything that
`can be programmed in software, produces a wide range of programs—
`word processors, spreadsheets, databases, calendars, graphics programs,
`and computer games. Manufacturers accept no liability when someone
`uses a computer to copy a copyrighted file. After all, one company builds
`the computer, another writes the software that does the copying, and
`both hardware and software are intended for general purposes——that is,
`any purpose the user,wants to put them to. The manufacturer wants no
`responsibility for, someone who uses the- computer in a way that just
`happens to infringe a copyright, nor does the software publisher. The
`perpetrator is the consumer, who finds it easier to make an unauthorized
`copy than to be strictly honest.
`'
`
`Petitioner Apple Inc. - Exhibit 1044, p. 9
`
`

`
`226
`
`Mar}: Stefi/2
`
`Stuck within this framework the community of computer users pro-
`tests against any attempt to regulate the copying of digital property. If we
`continue to accept this framework, with all of its assumptions, no party
`will be motivated or empowered to break the cycle and no effective
`way to protect digital property will be developed. At present, without
`enforceable property rights, the writers of words, interactive games, and
`songs often are not compensated for their work. And without their
`‘works the world is a darker, poorer place. Honoring their creative work
`in the digital systems of tomorrow requires us to challenge the design
`assumptions of the systems we use today.
`
`A New Design for Digital Publishing
`
`The technical core of the approach we propose is based, on two ideas: (1)
`that digital works can be bought and sold among trusted systems, and (2)
`that works have attached usage rights that specify what can be done with
`them and what it costs to exercise those rights.
`
`Trusted Systems
`The term trusted system refers to computers that can be relied on to do
`certain things. For example, suppose that a creator or publisher forbids
`all copying of a particular digital work. A trusted system in this context
`would reliably and infallibly carry out that stipulation; no amount of
`shouting or coaxing would coerce it to copy the work. The trusted sys-
`tem might be very polite, but ultimately it would always refuse to make
`an unauthorized copy. Similarly, suppose that a trusted system could
`copy a work but only if it reliably records a set fee to be paid when it has
`done so. A trusted system would always record the fee whenever the
`work was copied. If the copying process is interrupted part way through,
`the trusted system would follow a standard policy; for example, it might
`delete the partial copy, record no fee, and note that a copying attempt
`was begun but not completed. Again, no amount of coaxing would
`change its behavior. It could always be counted on to follow the rules of
`the trust.
`
`A common but false analogy claiming to show why digital works
`cannot be protected in computers is that of genies and bottles. In this
`
`Petitioner Apple Inc. - Exhibit 1044, p. 10
`
`

`
`Letting Loose the Light
`
`227
`
`analogy, a valued digital work corresponds to a genie and the bottle is a
`place to store it When a digital work is sent to a computer, for example,
`it may be sent in coded form, so that even if the transmission is inter-
`cepted it is useless to a wiretapper. Once people have a legal copy of a
`digital work, however, they can make more copies of it. Since they have a
`key, they can just decode the work and make copies of it. Alternatively,
`they can copy the coded version and give away copies of the key. Once
`the content genie is out of the bottle, according to this scenario, you can’t
`put it back in and unauthorized digital copies are sure to circulate. This is
`the problem trusted systems can fix.
`Trusted systems speak a communications protocol with other trusted
`systems and will not transmit information to any system not recognized
`as another trusted system. This strategy ensures that copies of digital
`works are either inside trusted systems or they are encrypted. When they
`are inside trusted systems, usage is controlled. When they are outside
`trusted systems, usage is practically impossible without breaking the
`code. The important issue, however, is not just protection and contain-
`ment. The greater good is not served by simply limiting the flow of
`information. It is served by supporting and encouraging a lively trade
`- in information. Rather than just confining genies to specific bottles,
`we want to encourage them to travel between bottles under rules of
`commerce.
`
`A very concrete question about such a system is “Why couldn’t I just
`copy a file onto a diskette and give that away?” Unless there is permis-
`sion to do so, a trusted system would never copy a work to a diskette or
`anywhere else. Even if permission to copy a work is given, a trusted sys-
`tem would not make a copy on a diskette, because a diskette is not
`a trusted system. Nor are magnetic tapes, compact discs, or, even, the
`disk drives of trusted systems. Trusted systems contain computers, have
`internal protected storage, and communicate by protocol. From a user’s
`point of view, the trusted system is the storage device. Trusted systems
`only make copies of a digital work on themselves or on other trusted
`
`systems. Putting an unencrypted copy on a diskette is letting the genie out
`of the bottle onto an unprotected medium that can be accessed by a
`general-purpose computer that does not honor usage rights.
`
`Petitioner Apple Inc. - Exhibit 1044, p. 11
`
`

`
`228
`
`Mark Stefik
`
`There is an important issue about the perception of trusted systems.
`One way of looking at them is to say that trusted systems presume that
`the consumer is dishonest. This perception is unfortunate, and perhaps
`incorrect, but nonetheless real. Unless trusted systems offer consumers
`real advantages they will probably view them as nuisances that compli-
`cate our lives. A more favorable way to look at trusted systems is to
`compare them to vending machines. They make it possible to order dig-
`ital works any time of the day and get immediate delivery. Faster than a
`telephone—order pizza, a digital work can be delivered immediately over
`the same telephone line it was ordered over.
`In summary, the first key to commerce in digital works is to use trusted
`systems: We have spoken of these systems as computers, but they are not
`limited to devices like personal computers and need not seem like com~
`puters at all. They could be personal entertainment devices for playing
`music, video game devices, laptop reading devices, personal computers,
`devices for playing digital movies at home, credit—card-sized devices that
`fit in your pocket, or whatever. In the following discussion we refer to
`these trusted systems as repositories, an architectural plan that can have
`different embodiments. Repositories communicate digitally with other
`repositories and not with anything else. In contrast to such current passive
`media as compact discs, repositories have no externally defined limits on
`storage capacity; so successive generations of repositories could increase
`in capacity while remaining completely compatible with earlier systems.
`Digital works would be communicated between repositories using secure
`coded protocols. Repositories would read the rules that apply to a given
`digital work and follow them. This brings us to the next issue: How do
`repositories know what the rules are?
`
`Attached Usage Rights
`We start with an analogy. When we go to a store to buy a shirt, there are
`various tags attached to it. One kind of tag is a price tag. If we want to
`buy the shirt, we must pay the amount ongthe tag. Another tag gives
`cleaning instructions: for example, wash by hand in cold water or dry
`clean only. Still another tag might say something about the style of the
`shirt or the history of the shirt company.
`
`..~._é........:........._._._____
`
`Petitioner Apple Inc. - Exhibit 1044, p. 12
`
`

`
`Letting Loose the Light
`
`229
`
`This is roughly the idea of usage rights on digital works. Digital works
`would come with tags on them. The tags—put there by the creators,
`publishers, and distributors—would describe the usage rights for the
`digital work: what can be done with it and what it costs.
`
`There are some important differences from the shirt’s tags. The first is
`that the tags are digital and intended to be read and used by the reposi-
`tory itself, although consumers can also read the tags through the repo-
`sitory’s user interface. They are written in a machine-readable language
`and give the repository the rules for using the work; they are an elec-
`tronic contract enforced by the repository. Another difference is that the
`
`tags are not removable. Finally, there can be tags attached to different
`1 parts of a work. For a shirt, it is as if there were tags on the pockets, tags
`on the buttons, tags on the collar, tags on the sleeves, and so on. Each tag
`would grant rights to that part, and different rights could pertain to dif-
`ferent parts of a work. For example, a digital newspaper might have
`certain rights on local stories, others on photographs or wireline stories
`or advertisements, and so on.
`.
`
`the digital work is a piece of music. A statement
`Suppose that
`describing a right might say the following:
`
`This digital work can be played on a player of type Musica-13B. This right is
`valid from February 14, 1995 to February 14, 1996. The repository must have a
`security level of three. No other authorizations are needed. The fee for exercising
`this right is one cent per minute with a minimum of five cents in the first hour.
`Usage fees are paid to account 1997-200-5 67131.
`
`I
`
`Of course, such an internal statement would not be in English, although
`it should be in a well-defined computer language. Here is an example of a
`machine-readable statement in a usage rights language:
`
`Right Code:
`
`Play Player: Musica-13B
`
`Copy Count:
`
`1
`
`Tirne-Spec:
`
`From 95/02/14 Until: 96/02/14
`
`Access-Spec:
`
`Security-Level: 3
`
`Fee-Spec:
`
`Account:
`
`Fee: Metered $0.01 per 0:1:0
`Min: $0.05 per 0/1/O
`1997-200-5 67131
`
`Petitioner Apple Inc. - Exhibit 1044, p. 13
`
`

`
`23 0
`
`Mark Stefik
`
`Computer languages are more precise than natural languages and have
`formal grammars and semantics that define how to interpret each phrase
`in the language. Computer languages are not at all poetic, but they are
`much less ambiguous, if less expressive than natural languages. Because
`the sentences of a digital property language are parts of potential con-
`tracts between the creators of digital works and consumers, clarity and
`simplicity are exactly what we want. Interpreting a usage rights language
`is quite simple. In level of difficulty, it is more like reading bar codes from
`packages at the supermarket checkout than it is like reading and under-
`standing an English sentence in a story.
`A digital property language needs to define several different kinds of
`rights, mainly those concerned with how the work can be transported,
`how it can be rendered, and whether it can be used in derivative works.
`Other, special rights relate to making and restoring backup copies to
`protect against hardware failure. The easiest way to understand usage
`rights is to consider some examples.
`I
`
`Transferring Digital Works When we copy files for friends on a gen-
`eral-purpose computer, we increase the number of copies of a digital
`work, fail to compensate the work’s creator, and infringe the copyright.
`A repository, in contrast, never infringes copyright.
`Our first scenario illustrates how copy and transfer rights would work
`in a repository system. Suppose that Morgan buys a copy of a digital
`book, perhaps at the book kiosk at the supermarket. To do so, he exer-
`cises a right to copy the book and pays a fee; copying thebook records a
`transaction between the seller’s repository and (say) a card-sized reposi-
`tory that Morgan carries‘ with him. Alternatively, he could buy a copy of
`the digital book from home by telephone. In either case, the digital book
`is delivered electronically by a communications protocol between the
`vendor’s repository and Morgan’s repository. At the end of the trans-
`action, Morgan has spent some money, has a copy of the digital book in
`his repository, and can now read it on a reader. The book arrives with all
`its usage rights intact.
`Now suppose that, when Morgan finishes reading the book, his friend
`Andy asks to borrow it. They plug their repositories together, and Mor-
`gan exercises a transfer right to move the digital book to Andy’s reposi-
`
`Petitioner Apple Inc. - Exhibit 1044, p. 14
`
`

`
`Letting Loose the Light
`
`231
`
`tory. With paper books, once we have bought a book we can give it
`away or dispose of it in any way we please, and the same right could
`apply to Morgan’s digital book. At the end of the transfer transaction,
`the digital book resides on Andys repository and not on Morgan’s, and
`no money has been exchanged. Andy can now read the book, but Mor-
`gan cannot. The crucial point is that the transfer transaction preserves
`the number of copies of the digital book.
`We now consider a scenario involving a loan right. Again Morgan has
`a digital book that his friend Ryan wants to borrow for a week. They
`plug their repositories together and Morgan exercises a loan rig/at. Again,
`while the digital book is loaned out, Morgan cannot use it. Suppose,
`however, that Ryan goes off on vacation and, while he is playing volley-
`ball on a beach thousands of miles away, the week’s loan period runs
`out. He has completely forgotten the book. Because both repositories
`have clocks in them, Ryan’s repository deactivates its copy when the
`week is up. Meanwhile, Morgan’s repository also notices that the loan
`time is up and marks its temporarily deactivated copy as usable again.
`Without any action by either person, or even any communication be-
`tween their repositories, the digital book has been returned automati-
`cally. If Ryan still wants to access it later, he could pay a nominal fee to
`rent the work or to make his own copy. The point of both scenarios is
`that the repositories follow rules, which in this case mimic and improve
`on the rules of loaning for paper books. The ability to return loaned
`materials automatically would probably be widely used in digital
`libraries.
`
`Rendering Digital Works To read a digital book you have to be able to
`see it; to listen to digital music you have to be able to hear it; to enjoy a
`digital video game, you have to be able to see and hear it. We use the
`term render to mean the processing of a digital work so that it can be
`experienced. Like copying, transferring. and loaning, rendering is con-
`trolled by usage rights.
`We distinguish two forms of rendering: playing and printing. When we
`play a digital work we send it to another person through some kind of
`transducer so that he or she can experience it. The term play, usually
`employed in phrases like playing music or playing a movie, is also used to
`
`Petitioner Apple Inc. - Exhibit 1044, p. 15
`
`

`
`232
`
`Mark Stefik
`
`denote displaying part of a book, running a computer program, or run-
`ning an interactive video game. The term print in the digital context
`means to make a copy of the work on media outside usage rights control,
`either on paper or by writing a file to an external storage device.
`The concept of usage rights allows great flexibility in marketing digital
`works. Today, when you buy a compact disc at the music store, you pay
`for the copy and play it for free. The same is true for a book. You buy the
`book and read it as often as you want; generally, you aren’t supposed to
`make copies of it, but you can give it away. By contrast, keeping digital
`works in repositories would provide more flexibility.
`Suppose for example that Andrea’s mother is at the music store but
`does not know exactly what music her teenage daughter wants to buy.
`She transfers a selection of music to her own repository, choosing col-
`lections by half a dozen bands that she knows Andrea likes. At this point
`she does not need to pay anything for the right to make the copies. When
`she gets home, she transfers t