============================================================================ The text file below was created from a document with richer character encoding than standard ASCII supports. If you are using a Unix system, it may be translated into standard ASCII using the command tr '\216\237\320\321\322\323\325' '\145\165\055\055\042\042\047' | sed -f script or you may use the following table to make the translations on another system: Octal code Rich Character Ascii equiv ------------------------------------------------------------- 216 accented e e 145 237 umlaut u u 165 250 Circle with R (trademark) trademark 251 Circle with C (copyright) copyright 320 en dash - 055 321 em dash - 055 322 left double quote " 042 323 right double quote " 042 325 right single quote (apostrophe) ' 047 ============================================================================== Intellectual Property Rights Abstract The issue of software patents and intellectual property rights is a controversial and timely one. It has a direct and important bearing on the computer graphics community (both technical and artistic) as well as the programming and end-user communities as a whole. The issues have been brought to the fore by a series of highly-publicized (and highly controversial) lawsuits in recent years involving major industry players including Xerox, Apple, Lotus, Microsoft, Hewlett-Packard, NEC, Intel, Fujutsu, and IBM over cases ranging from copyrights on microcode to the Òlook and feelÓ of graphical user interfaces. Equally important is the stream of patents now appearing covering specific computer algorithms. Despite the publicity (or perhaps because of it), there seems to be much confusion among programmers regarding the basic principles of trade secret law, copyright, and patent law, not to mention the ultimate implications of these developments. Even within the legal community there is considerable disagreement over specific interpretations of the law. This panel will examine the question of software copyrights, patents, and intellectual property rights in general in a discussion between parties on both sides of the fence, hopefully leading to a greater understanding of the issues involved and how they will affect the practitioner in the field of computer graphics. Chair: Michel Denber, Xerox Panelists: Pamela Samuelson, University of Pittsburgh John Barlow, Electronic Frontier Foundation Isaac Victor Kerlow, Pratt Institute Peter Deutsch, PARC Place Systems Michel Denber: I want to welcome everyone to our first panel session at SIGGRAPH Õ91. My name is Michel Denber. I work for Xerox and IÕm in image processing. IÕve been doing computer graphics for about eleven years now. IÕd like to welcome everyone to what I think is going to be a very interesting and very exciting panel. ItÕs certainly a timely topic and one that is guaranteed to be a very controversial topic too. I think that it is a topic whose time has definitely come. If you went to the opening session this morning, you heard both Jim Kajiya and Scott McNealy make references to issues associated with intellectual property rights. So, I think itÕs definitely time that SIGGRAPH address these issues. These are things that are going to affect, or have already affected, everyone in this room and everyone at SIGGRAPH either as producers or consumers of computer graphics. So there are some very interesting and difficult problems that need to be addressed. I think that weÕve got a great panel for you this morning. Let me first tell you briefly about the format of the panel and then IÕll introduce the speakers and weÕll get underway. What I would like to do is to have each speaker give a short presentation of about eight minutes or so. Each one will give their presentation; that should take us to about the halfway point. After that, I have a short one page survey which I would like to have distributed to everyone here. If you could take a little while to fill it out for me and leave the completed surveys at the back, IÕll tell you a little bit more about that when we get to the halfway point. IÕve reserved the entire second half of the panel for questions and comments from the audience. I really encourage you to come up to the microphones that are placed in the aisles throughout the room and ask questions of the panelists or just state your views. This is what this panel is for. Okay, letÕs get started. The first thing that I have to mention is that unfortunately we have a speaker change that was just at the last minute, too late to get in the final program. Unfortunately, Richard Stallman is unable to be with us today, however, he has provided us with an outstanding replacement. WeÕre very fortunate indeed to have Peter Deutsch with us as a Richard StallmanÕs replacement. Peter is Chief Scientist of PARC Place Systems out in California and former Research Fellow at Xerox PARC (Palo Alto Research Center) and also a long-time hacker (and thatÕs hacker in the positive sense of the word; the old-fashioned type of hacker, not what the media has made the word out to be lately.) Our other speakers are Isaac Kerlow, who is the Chairman of the Art Show here at SIGGRAPH Õ91. He is Chairman of the Computer Graphics department at the Pratt Institute in New York, and Associate Professor of computer graphics. He got involved in computer graphics in the early 1980Õs programming video games. He has an interesting talk for us, IÕm sure. In addition, we have John Barlow who is co-founder of the Electronic Frontier Foundation, technological observer, songwriter for the Grateful Dead and retired cattle rancher. If you were at SIGGRAPH Õ90 and you saw him on the panel on Virtual Reality, IÕm sure youÕll remember his talk, which was outstanding. Our first speaker will be Pamela Samuelson, who we are extraordinarily lucky to have at SIGGRAPH Õ91. Pam is a Professor of Law at the University of Pittsburgh and contributing editor for CACM. You may have read her column, ÒLegally SpeakingÓ which appears in CACM. She is very interested in issues of intellectual property, software patents, and copyright. Her talk is going to be split in two parts. First, she will give us a brief introduction to the topic; her main talk will be at the end to wrap things up for us. So with that out of the way, IÕll step down and let Pam start and weÕll go from there. Thank you for coming this morning. Pamela Samuelson: I want to start by talking about where intellectual property rights come from. The truth is that some countries donÕt have intellectual property rights systems. Some of those that have intellectual property rights systems have different theories about why there should be such things as intellectual property rights in the first place. Some countries think that by virtue of creating an intellectual work, you have kind of a natural right or a natural property right in the created thing and that the law simply recognizes that kind of natural right that you have in the thing by virtue of creating it. In the United States and in the tradition out of which the U.S. law has developed, we take a very different view of things. In the tradition out of which U.S. law arises, the rule is that while if you come up with an intellectual product and you keep it secret, then you own it and no one else can use it because nobody else knows about it. But, if you show it to other people and if you sell it on the open market, then unless you have acquired a set of rights that the government creates, then the thing is free to be copied by whoever wants to copy it. So the basic principle of our system is that the general rule is free copying and that only certain categories of intellectual products are eligible for protection and then only eligible if they meet the standards and follow the procedures that the law sets up. That means that the government decides to create intellectual property rights in order to give people incentives to make that intellectual work available to other people. So, to give incentives to inventors of new machines and other industrial technologies, the government has passed a law that gives seventeen years of exclusive rights to inventors of these new machines. But you have to have made an inventive advance over the state-of-the-prior-art and you have to follow a very precise set of procedures in order to get the patent from the government. Sometimes itÕs going to be worth that disclosure of your invention to the public, because for that seventeen years, you are the only person who can sell that product in the marketplace. You can make a lot of money and hopefully invest in new innovations. At the end of that seventeen year patent period, the invention is in the public domain and anyone who wants to practice it can do so. Similarly, to give incentives to authors to publish their writings, the government gives a set of exclusive rights to authors as to those writings. Nowadays it is life-of-the-author plus fifty years of protection of the expression that is embodied in the work but not to the ideas. Over time Congress has passed laws which have expanded the concept of authors to include, for example, graphic artists and those who have developed motion pictures and even now, software. But, the basic aim of the copyright system is to give incentives to authors to make those works available and you get from copyright law only as many rights as Congress has granted and only again for this specific period of time. So, if you havenÕt followed the procedures, you donÕt have any exclusive rights at all; your work is available to be copied by anyone. Now last night I went to the Flamingo Hilton and saw a Las Vegas show. In that show I saw an act by three brothers; IÕm going to call them the Zamboni brothers because I donÕt remember what their names are. Their bodies were all painted and there were lights on them and they performed a very slow set of sculptural kinds of movements. I looked at it and asked myself what kind of intellectual rights are attached to this? There were some very, very inventive moves in this act, but theyÕre not patentable because theyÕre not in the industrial or technological arts. Patents are only available for things in the technological arts. And itÕs not copyrightable because itÕs not fixed, because itÕs a live performance, although I thought that maybe if they wrote down choreographic directions, they could copyright that, and then this would be a performance of that work. It canÕt be a trade secret because you can see it and if somebody taped them, that would be a violation of what the law calls their right of publicity. But the right of publicity would only protect their performance of the act, not the act itself. So, basically if somebody wants to do that act, they can do it. Their act, however, is a situation where you donÕt really need intellectual property rights because the skill level it takes to be able to do this really protects against cheap competitive imitations. That illustrates the way that the law tends to look at issues and says that unless it fits in a category and meets certain standards, itÕs not going to be protected. Sometimes something new comes along and Congress will pass a special law to protect it. With semiconductor chip designs, thatÕs in fact what Congress did. It passed a law that was very much copyright-like to protect chip designs because the whole design of a chip was not inventive enough to be patentable and yet it was very expensive to develop and very cheap to copy. So the semiconductor law was passed to provide some incentives and protection to designers of semiconductors. Where does software fit into all of this framework? Originally when software was first being developed it was made available to the public either free, because it came with some machines, or it was licensed on a kind of a trade secret licensing basis. In the middle 1960Õs, both the copyright office and the patent office tried to figure out what they should do about software. The copyright office decided that it would accept registration of computer programs as literary works because you write them sort of like you write books, as long as you deposited a full text of the source code that would be available for public inspection. But even then they had some doubts as to whether or not computer programs could be copyrighted because of a case that had been decided by the U.S. Supreme Court which had held that piano rolls were not copies of music because they were mechanical. They were parts of machines. So, the Office wasnÕt sure that it could be copyrighted but they said, well weÕll try this and see what the courts think about it. In 1966, the patent office, in line with the recommendations of a presidential commission, issued a policy statement that said that computer programs were not patentable subject matter. Although the patent statute says that processes are patentable, what the patent office thought that that word meant was that processes that involved the transformation of matter, as most industrial processes do, are patentable process, but mental processes, like those imbedded in software, werenÕt patentable processes. So, for a long time the copyright office was registering a few copyrights but not that many because people didnÕt want to deposit the full text of their source code and there were a few patent applications going on. But mainly during the 60Õs and the 70Õs, software was available on trade secret licensing basis and not protected by other intellectual property rights. A number of companies thought that software inventions really should be patentable and a couple of times went and took cases to the U.S. Supreme Court about this. The first of these cases, in 1972, the case of Gottschalk vs. Benson, the Supreme Court ruled that an algorithm for converting binary coded decimals to pure binary form was not patentable subject matter. It was a mathematical idea and it didnÕt involve the transformation of matter and therefore was unpatentable. In the 1970Õs, a commission was studying some new technology issues in copyright and decided to recommend to Congress to pass a law which would make computer programs protectable by copyright, because it was becoming increasingly apparent that trade secret licensing was unsuitable for mass marketed software and there was then a vision that software really could be a mass marketed product. In response to this commissionÕs recommendations, Congress passed in 1980 a law which basically granted copyright protection for computer programs. But the report that was written about this didnÕt give very much guidance about how the law should be interpreted as to software. A lot of what we see in the case law that you read about in the newspapers is trying to figure out what the law really means as applied to software. In the early 1980Õs, the patent office in response to a Supreme Court decision in 1981 (or allegedly in response to a 1981 Supreme Court decision), decided to take a very broad view of this particular decision as though it had overruled Gottschalk vs. Benson and removed all of the obstacles to the patenting of software. Since about 1983 or so, the patent office has been issuing patents on a wide range of software related inventions. ThatÕs a quick overview about the history of how we got to where we are now. I want to leave you with just this one view. The traditional intellectual property system in the United States has assumed that something is either a writing, in which case it is in the copyright realm, or that it is a machine, in which case it is in the patent realm. ItÕs assumed that something has to be either a writing or a machine and canÕt be both at the same time. I think that a lot of the problem weÕre having with software intellectual property law right now is that software is both at the same time and we just havenÕt figured out what is the right way to deal with that. Michel Denber: Thanks a lot, Pam; that was extremely interesting. IÕm sure that a lot of people here donÕt really know the history of these issues and it is something new for many people. Our next speaker is Peter Deutsch. He is also a member, as Richard Stallman is, of the League for Programming Freedom, although he tells me that heÕs not here in an official capacity to represent the league. So, if you will welcome Peter for his talk. Peter Deutsch: Thank you. I have been involved with issues of software patent and copyright from several different angles. I used to be a researcher at Xerox PARC. Xerox takes these things very seriously. I am currently at ParcPlace Systems. I am also an independent software developer and my software product is an implementation of a programming language. There is an issue as to whether programming languages can be copyrighted. So, I have gotten involved with this from several angles. So, let me say that I think what we are talking about here has some very, very deep philosophical questions behind it. I see the debate as being about what is the appropriate way to think about the social process of creation. In particular, what I believe we are seeing in the software industry is that, over the last five years, presumably under the impact of the spread of mass market software, as Dr. Samuelson suggested, we are seeing a shift in social paradigm from a cooperative one to a competitive one. Now, I know that competition is the American way and so on, but if you look around at the progress of the technology, the progress of the society, I believe that you get greater benefits when the prevailing ethic is one of cooperation. I am not going to attempt to defend this with examples and case histories. But, it is a thought that I would like for you to bear in mind as I go through my opinions on some of this stuff. So let me just say up front what my position is, what I think is appropriate to do about software patents and copyrights. I do not think that there should be patent protection for software and I think that copyright ought to be employed in a very traditional sense on the literal text of code and the literal appearance of visuals. I do not think that copyright protection should be extended to such things as look and feel or interfaces or even languages. So, let me say a word or two about patents. I think Dr. Samuelson made an interesting point to the effect that in this country we donÕt have inherent property rights. We have these limited monopoly rights that are created legislatively. So, if you look in the Constitution, there is a section about patents, the purpose of patents being to promote, I donÕt remember exactly what the phrase is but, basically to promote progress in the useful arts. So, it seems to me that in looking as to whether we think patents ought to be applied to software, the question that ought to be foremost in our minds is, do software patents in fact have the effect of promoting the useful arts. (There is also a long history of government created monopolies, even temporary monopolies, which I wonÕt go into, although I do have some backup material on that if people want to talk to me later.) So the first point that IÕd like to make is that software patents as they are presently being implemented are not working. Many bad patents are being awarded. I will give you some examples. First, there are patents being awarded for things that clearly have prior art. There was a patent awarded to Iwahashi in the early 1970Õs for a technique for speeding up multiplication which was in routine use when I was in high school in the early 1960Õs. Second, overlapping patents are being awarded for essentially the same technology. Patents have been awarded to both UNISYS and IBM for essentially the same compression method which is commonly known as the LZW compression method. Third, software patents are being awarded for things that are obvious. This includes the Cadtrak XOR cursor patent; a patent on using a standard byte order in files to overcome differences in byte ordering between machines; and I would put the Refac natural order recalc patent in that category as well, since in my opinion it grows directly out of published work on data flow machines. Patents are being awarded on algorithms and broad classes of algorithms despite the claim by the patent office that they are not doing this. A blatant example is the RSA encryption algorithm which a fair amount has been written about. Very injurious is the fact that what amounts to retroactive patents are being awarded. That is, that patents are being awarded many, many years after something has become widely practiced. The two examples that come to mind here are the AT&T patent on backing store, on a particular implementation of backing store, and again the LZW patent. A lot of the problem of this is that the current staff of the patent office is just not competent to be aware of prior art, to analyze the patent submissions, to understand the cross domain applicability of much of this technology. So, my first major point is that the current software patenting system is not working. The second point, and I am going to pass over this rather briefly because it is really a matter of history, is that I donÕt believe that software patents are needed. A claim is often made that software patents, or patents in general serve to promote the progress of the useful arts by making sure that the technology gets out in the open for licensing as opposed to being held as a trade secret. Software patents on the other hand were not awarded, were not recognized as legitimate, roughly speaking, before a decision in 1986, which I guess Dr. Samuelson didnÕt really have time to go into. I donÕt think that anyone would claim that progress in the software art has been markedly greater since 1986 than before 1986, and in fact I would be prepared to argue the proposition that it has been less rapid since 1986. Finally, I would say that software patents are actually actively harmful to the practice of the technology and to the furtherance of the good that comes from it. They raise the entry costs, in that the small developer now has to be aware of hundreds, if not thousands, of patents potentially affecting every aspect of the software design process. The claim has been made that such patents are not prosecuted, but that is irrelevant, in my opinion. The fact that someone is not being prosecuted for breaking the law does not mean that they are not in jeopardy. That is really all I have time to say about patents. Let me talk briefly about copyrights. Copyrights have been traditionally applied to the expression of ideas rather than to ideas themselves or concepts. I believe that the appropriate application of copyrights is to the text of code and to the particular visual details of interfaces. As Dr. Samuelson said, there is this ÒwritingÓ versus ÒmachineÓ issue for software which is unique among things that have existed in the past. My opinion is that, for example, computer languages which some people claim are copyrighted, should not be treated in the same way. Computer languages, even the description, a reference manual of a language, is something that is meant to be used like a machine, as opposed to read like a work. I would claim that languages in fact embody underlying concepts and that the expression that should be protected by copyright is not the language description but its implementation. An argument has been made that, you know, you can always implement another language to accomplish the same function. In my opinion this is a gratuitous incompatibility which hurts the buyer and has no benefits for the originators. I would extent this principle even to user interfaces and even to some rather detailed aspects of user interfaces. What should we do instead? I think that companies ought to compete the way that they have competed all along, namely, on originality and on quality of implementation; and trade secret protection is sufficient for the latter. So, finally, let me close with two short comments. One is that as I said at the beginning, I think this is a debate about the social process of creation, and I think that it should be pretty clear where I stand. The other comment is that in a debate on-line Rob Pike from AT&T asked the question, ÒWho owns an idea? DoesnÕt the inventor of an idea own it?Ó Well, no. The inventor of an idea, I think has, in some principled sense, some right to benefit from it, but, I do not think that they should have a monopoly right at the expense of other independent creators of the idea, which is in fact a very, very common situation in software today. Thank you! Michel Denber: Thank you, Peter. Our next speaker is going to be Isaac Kerlow and in fact he hasnÕt told me what heÕs going to talk about, so itÕs going to be a surprise to me as well as for all of you. Please welcome Isaac. Isaac Kerlow: Hi! I have three questions for you and three answers to start this very short presentation. Have you all thought about this? In terms of interface, what is the main difference between these two magazines? We are not talking about the contents obviously. The centerfold. That is a way of using, a way of interfacing with a magazine that some magazines use and some magazines do not use. Does Playboy copyright it? Does Penthouse copyright it? I donÕt think so. Now what is the point that I am trying to make here? My point is that often we forget how previous technologies have dealt with all of these issues of look and feel. The centerfold is a look and feel issue. This is how the magazine is interacted with. So, my first piece of advice is to look back and see how other technologies have done this in the past, have dealt with these issues. Next question. This actually happened to me two years ago. I gave a client all of the files that I created with a drawing program only to find later, a couple of weeks later, that they had changed things around. They had moved the elements of the drawing, they changed the typeface and they didnÕt even tell me. They didnÕt feel like they had to ask me because they had the diskette. Well, I was upset. I told them, you know, you cannot do this. If you are going to do this, you have to talk to me about it. Of course, I have learned a lesson. The lesson is that you have to be informed about these issues. Because, they are both complex and tricky. The laws you have been hearing about, intellectual property as both copyright protection and patent protection, are very unclear. Question number three. In the world of design, you know people are famous because people get famous or rich or popular, successful or not because of the ways in which they do things. The ways in which they use type, the ways in which they combine colors, the ways in which they interpret an idea. Can they patent their way of doing things on the basis that it is a process that transforms ideas? No, they canÕt. What if a programmer or the designer himself or herself codes a designerÕs way of doing things? Can they copyright the program? Can they copyright the algorithm? That is an interesting question. I would say no. My piece of advice here is that things are changing and they are changing right now. It is really up to us to decide what is going to happen with them. Okay, now you have heard about this before. Well this is a very important distinction to understand. It is an obsolete distinction, as the previous speakers have pointed out. Copyright law has traditionally covered intellectual property and patent law has covered manufacturing processes that result in tangible goods. I donÕt know if this is the legal definition or not but this is the way I understand it. Now, traditionally, this is before computers and computer graphics arrived, the results of intellectual work were entirely independent of the tools that were used to create them. In other words, after you typed your novel on a piece of paper you could smash the typewriter and you still have a novel. You could take your manuscript to the publisher and print a book. But today, with this fabulous technology, things are very different. Intellectual works in many cases are very dependent on the tools that were used for creating them. For example, if you have created a novel in a hypermedia format and your computer system breaks down, there is no novel. Nothing to show, nothing to consume, nothing to express. So, this is an example of how things are very different from just 15 or 20 years ago. We are talking about an entirely different animal. It is very obvious that the traditional distinction between copyright and patent does not apply to many of the things that we do today. Okay, I am going to go very, very quickly through the next seven or eight screens just to give you a sense of how cumbersome the copyright law can be. I am not a lawyer but I get a sense of how complicated and all of these subheadings that it has. First of all, copyright means a lot of different things. So when you say copyright or patents, should they be patented or not. I donÕt think that the issue is a yes or a no. Yes, we will patent it or no, we will not. Maybe we can patent some of it, and maybe we cannot patent some other parts of our work. This is very interesting for me because the coverage given by copyright law, which is the one that I understand better then the patent law, varies a lot in my opinion. Just to give you an idea of how much it varies, I will talk about three intellectual property fields: type designÑthe design of type faces, the creation of musical composition, and the creation of drawings. Type designs, as you can see there on the screen, have traditionally not been granted copyright protection. Well, actually creators of type designs have not been granted copyright to their works, because the government feels that would not be in the public interest. They can, however, get protection on the name of the type face so that when they market it, they have something to sell and something to be identified by. The original, in the case of a type face today, and for 10-15 years already, has been a digital matrix or a photographic matrix which is very easy to duplicate. So here we are talking about something that cannot be copyrighted, only the name. The data is given away. The way the market dealt with it is that the distributors of this type of intellectual property pay royalties to the creator. The final consumer, that is the client or the designer who use it, they donÕt really pay the designer or the typist. I am sure that many of you are familiar with this process. Now music is quite different. The ownership of a musical performance, like a musical recording in the form of a CD that you would go and buy in a store is usually split between dozens of different people. It has always been like this and it hasnÕt been a problem. Well, I mean, problems can exist but we are used to that. The original piece can be the musical score which is in a way like a program. It is just instructions to be interpreted by someone, or the particular expression of that idea using the terms that we were mentioning before would be the recording. The specific recording of a song with a specific set of musicians, a specific set of instruments, singers, recording studios, etc. So now we start to get a little tricky here. Of course the way that it is marketed is by consumers paying all of the people who own the copyright through a very complex network that collects all of these royalties and sends them to the composer and the recording studio and the composers union and all of these other things. I have three screens left and then I am done. Finally, drawing as intellectual property. This is I would say the most traditional example of copyright protection. Drawings are original things that you draw on paper and if you duplicate them, you can tell that they have been duplicated. So, unlike type designs or music, where you can give the originals away, the algorithms away, and the code away, here you canÕt. There is no code to give away, there is only a physical thing that you can tear off or frame. And of course, the artist has the rights to the piece and he or she can sell them, all of them, or some of them. Finally, the drawing exists, I already said it before, as a physical original as opposed to some codes written somewhere, and artists can sell some or all of the rights to the piece. That is all I have to say, thank you very much. Michel Denber: Thank you. Our next speaker will be John Barlow. John Barlow: Well, IÕve got to say that IÕm really sorry that Richard Stallman didnÕt come today. I can usually count on him to redefine the left. I was figuring that no matter how fanatical I got, he would make me sound like the voice of sweet reason. For Richard, I probably wonÕt sound like the voice of sweet reason. IÕve got to preface this by saying that I make my living with intellectual property. So before any of you think that IÕm trying to throw out your rights, my own would be in danger, I suppose, if there were any real effect to what IÕm proposing. I used to make my living in tangibilities but I found that there was a lot more money in bullshit than bulls, and IÕve moved to something that the market seems to favor. I think that the term Òintellectual propertyÓ is an oxymoron. If you donÕt understand that, I would direct your attention to a great cartoon that I just saw in the Bulletin of Atomic Scientists. ItÕs got a guy with his hands up in the air and a gunman on him. The gunman is saying, ÒGive me all your ideas.Ó I think the reason that it has appeared to work as a section of law as long as it has, is because it was so firmly vested in the physical world. Copyright worked as long as it was physically difficult to make a book. Patent worked well as long as it involved, as Pamela said, the transformation of materials. When you were making a widget and not something that was not totally insubstantial, it worked pretty well. But right now IÕd say that itÕs not only breaking down, itÕs already broken down. IÕve got a little poll that IÕd like to take at this point. How many people in this audience can honestly say that they have no unlicensed software on their hard disks? A few hands, good for you folks! Hang in there! Well, I would suggest that when you have such a divergence between the law and social practice, things are going to go to hell in a handbasket. We can send a whole horde of lawyers down into the Valley of Death trying to fight what society has decided is an acceptable practice, but itÕs not going to further the march of technology. When we started coming up with creations that had no investiture in physicality, we entered into a whole new world. Now I have this hippie- mystic sense that ideas canÕt really belong to anybody because, I donÕt know where yours come from, but mine seem to come from some other place. IÕm not going to get into my personal theology but it seems like they come to everybody when they come through me. As long as I was transforming those ideas into something physical, then my protection lay in execution. These days, execution is a much more difficult thing to define but IÕll get back to that in a second. We donÕt really know what intellectual property is in terms of intangibility. We think of property as a tangible thing and weÕre trying to protect intellectual property as though it were tangible, but intellectual property is very different. I mean if I steal your horse, you canÕt ride; if I steal your ideas, you can go on using them. That is a fundamental difference. We also seem to assume, and the institutional response is that this stuff is somehow like real estate or pig iron, in that if you have more of it around, it will automatically decrease in value. I think that you can look around and see plenty of evidence that thatÕs not necessarily the case. I would say that, for example, Lotus 1-2-3 continues to exercise an iron standard in the spreadsheet market, in part because itÕs also the most pirated software in the world. A little closer to home, back in the beginning of the Grateful Dead, we decided that we didnÕt have a problem with people taping our concerts. We didnÕt think that it was going to cut into the sales of our records, which were minimal anyway. Over the course of time, those tapes have proliferated. There are millions and millions of Grateful Dead tapes out there, and before the tapes became such an item of commerce among Dead-heads, we werenÕt selling very many records. Now we sell a lot of records, so I wouldnÕt say that theyÕve cut into our proceeds from our intellectual property, quite the contrary. You can also take a look at the software industry, which as you know is a multi-billion dollar a year deal, in spite of what we just saw when those few brave souls put their hands in the air. Now somebody is paying for this stuff. Somebody is paying for video, in spite of the fact that itÕs very easy to copy that movie you rent. I would say that the reason for both things is that thereÕs something vested in tangibility in software and in video. Most people who pirate software try it out for a while and if they really get to like it, they buy it for several reasons, some of them perhaps moral but largely because they want to have the manuals and they want to have the support. IÕm going to suggest that support is a manifestation of tangibility. What we have to do is to start looking at where we take the ideas and turn them into either goods or services. ThatÕs where we start to get paid for the work that we do with our minds. This gets us back to the origin of copyright law which protected not ideas but their execution. The execution is very important. I take an entirely different view towards somebody whoÕs out there selling exact replicas of Built to Last on CD than I do of somebody who is trading around copies of all of the songs on it. ItÕs a very different deal; there has to be protection against that. That protection is based on the physical resemblance of that bootleg CD to the actual item. There are many examples of this. I just found myself giving a speech on the same lectern as Edward Teller and we found ourselves in amazing agreement on this (a disorienting experience). Edward Teller said something very important, which is that everybody, and IÕm one of them, knows how to make an atomic bomb. The Progressive published the recipe years ago. What protects us from having everybody out there making atomic bombs is that it is very difficult to execute. So the real right of protection in that, as in many other cases lies in, ÒCan you deliver?Ó Where I come from, people are not highly regarded for the quality of their ideas. They are very highly regarded for the quality of what they can do in putting those ideas to use. ThatÕs a basic and fundamental principle that I think we have to start following. Now a lot of people would suggest that, if we operate on that principle, thatÕs going to give additional protection to large companies that have a superior capacity to execute, but I think that you could look around you at the superior capacity to execute of the likes of IBM, or increasingly Apple, and give that some serious thought. I think if youÕre small and light and able to move quickly, youÕre a lot more likely to be able to protect your ideas through swift delivery and execution than you can through any great industrial capacity in that area. Well IÕm running out of time here, but I think that weÕve got some serious problems with figuring out how weÕre going to get paid for the work that we do with our minds. IÕm not here to suggest that we should simply throw the whole thing out and start from scratch. But for those of you who are involved in issues like multimedia, I donÕt think that that whole endeavor is going to get very far unless we greatly expand our sense of what constitutes fair use. We start to make some assumptions that are different from the ones that weÕre legally veering toward that would give you the right to use commercially, images that have been produced by other people. Something is going to break here and I have a feeling that itÕs going to be the law and not social practice. The real danger from my standpoint is in the fact that software and other kinds of intellectual property are indistinguishable from speech in many respects. If we start regulating intellectual property with a kind of increased capacity and zest for control, weÕre going to be regulating speech. I donÕt think that any of us wants to live in a country where our rights of expression are controlled under patent and copyright law. Thanks! Question and Answer Session Michel Denber: Okay, thank you. IÕm going to throw the floor open to questions from the audience at this point. If we have time towards the end, then Pam has some more material that she can present. At this time I also have some surveys that IÕd like distributed among the audience. If you could take a minute just to fill these out IÕd appreciate that a lot. This is essentially the same survey that was handed out at CHI Õ89. I think itÕs going to be really interesting to see how attitudes have changed, if any, since that time, given all of the recent developments in this field. Hopefully these will be published in either the Panel Proceedings or in the SIGGRAPH Quarterly, so keep an eye out for those results. Your participation today will help insure the success of that. So first question from over here please. Q. My name is Rick Spear, IÕm with Animation Research. My question is for Professor Samuelson. IÕm curious about her opinion on the patenting of computer algorithms. Is this good for the computing community and for society in general? Is it working or can it be made to work? Pamela Samuelson: Boy, thatÕs a complicated question. Let me just say that, under the existing state of law, as the Supreme Court has interpreted it, algorithms are not patentable. The patent office is issuing algorithms patents notwithstanding the fact that they are clearly not justified under the Supreme Court decision. So in my view, if algorithms are patented, it should be either after the Supreme Court has overruled its prior three decisions in which it has said that computer program algorithms are not patentable, or the legislature, namely Congress, should pass another piece of legislation about it. In the process of passing such legislation, there would be an opportunity for people in the community to air their views about whether or not this would be in the public interest. From the standpoint of the existing legal system, I donÕt see that those patents are valid. So the issue of whether there should be patents on algorithms, it seems to me is a very complex public policy issue and I can make arguments pro and con. It does seem to me that thereÕs something technological in character to program algorithms; on the other hand it does seem to me that patents may not be the ideal system for algorithms. ThereÕs another possibility, to say that we could pass a different kind of law for algorithms if they needed some degree of protection, or we may decide collectively that algorithms are the sorts of things that people ought to be able to share. It seems to me that the public policy issue is one that you are probably in a better position to make a judgment about than I am; IÕm a lawyer. Michel Denber: Okay next question from the center isle please. Q. My name is Ben Delaney, IÕm the publisher of CyberAge Journal and I too make my living on intellectual property. I agree strongly with JohnÕs position that getting it out there and doing something with it is more important then spending a lot of energy on protecting it. But I would like to raise an issue (and itÕs more of a point IÕd like to make than a question), which is that thereÕs a new form of governmental control which is happening, notably in California where IÕm from, which is that the government has decided to attach a sales tax to publications. This I feel infringes on the right of expression, especially of people who are producing free publications, who now have to pay tax on their printing which can greatly curtail their ability to express themselves. It also adds a burden to those who wish to learn about things that are happening, because they are required to pay a tax on publications that are sold for a price. So I just wanted to make the point that this is another, I think insidious, way that the government is trying to control expression and I would like to urge everyone to do their best to oppose this. IÕd appreciate any comments from the panel. Michel Denber: Does anyone want to make any comment about that? Pamela Samuelson: I will make one comment that one of the things the First Amendment to the U.S. Constitution gives the public is the right to be free from government interference with expression in speech. There have been times when special taxes or other kinds of special regulations directed to the publishing industries have been struck down as, in fact, violations of the First Amendment. But itÕs a complicated thing about whether that should apply completely across the board to any kind of tax situation. So things have been struck down before but itÕs not easy to do. Michel Denber: Okay next question from the aisle on your right. Q. Hi, William Burken from the University of Washington. This question is for whoever wants to handle it on the panel. WeÕre working with algorithms that convert symbolic mathematics to a physical representation. In other words we can take logic, arithmetic, and algebra and convert it into piles of blocks. The formal semantics and transformation of these mathematics is rearranging the stacks of physical blocks. What are my protective rights with the manipulations of these blocks? John Barlow: You know, William, IÕve thought about that a lot, mostly with reference to art. The question is, is a virtual object which is sensed as though it were a physical object, a physical object? I donÕt know the answer to that because the difference between a VR object and a real object is that a real object has a way of being, as a fixed thing. There is a final cut when you make it. The difference between this and VR is that it is always malleable. If something is malleable by anybody at any time, I think it probably doesnÕt afford you much protection. William Burken: Actually, John, the question IÕm asking is about the physical embodiment of a mathematical technique. This is not a VR question. John Barlow: The physical embodiment, you mean what you wrote down on paper? William Burken: No, IÕm saying that we can set up control structures of programs as configurations of physical blocks stacked on a table. When I manipulate those blocks, IÕm doing program control. Do I have any rights as to the physical manipulation of a mathematical system? Peter Deutsch: IÕd like to take a stab at that with a much simpler exampleÑthe abacus. There is a set of algorithms for doing arithmetic on an abacus entirely by manipulation of physical objects. IÕm not a lawyer, I donÕt know what kinds of distinctions would stand up, but when I was preparing this talk late last night, the following distinction occurred to me, namely that a test of whether something should be patentable is whether the essence of the process is a manipulation of information, or whether the essence of the process is in the manipulation of physical objects. In other words, if the process would still be recognizably the same or equivalent if it were expressed in some other medium, perhaps an intangible one, then to me that places it squarely in the camp of algorithms. If thatÕs not the case, it places it squarely in the camp of physical processes. According to that line of thinking, your rights as to the manipulation of blocks are exactly the same as your rights as to the expression of the algorithm in any other way, tangible or intangible. William Burken: Just one more harmony on this. It seems that I was trying also to draw an analogy to what an actual computer chip does. Peter Deutsch: I donÕt think that the algorithms expressed in a computer chip should be protectable, but I do think that the physical design of the chip should be protectable in a narrow sense under copyright law. Pamela Samuelson: It seems to me that the kind of question you raised is the one that causes all the lawyers to go crazy, and the simple response is well, of course, the abacus. We can understand the abacus because the abacus is a machine, thatÕs easy. ThatÕs patentable, no problem. But I think that, in some sense, weÕre talking about something that we donÕt have any experience with. If you want to go to the patent office today, they will issue you one. Whether it is worth anything or not is a good question. Michel Denber: Okay if we could go on please. Q: Kelly Booth, University of British Columbia. Question for Peter Deutsch. You made it clear that you feel that software should not be copyrightable. I think it was implicit in your presentation that traditional novels, etc. are still copyrightable, at least I think thatÕs what you were saying. What happens when the line begins to fuzz between the two? IÕll take as an example a novel written in HyperCard, where much of the expression of the ideas is in the novel interweaving of the way the links are made. In fact, if the links are active, which requires execution of an algorithm dependent on how you previously explored this hypernovel, I can see that becoming very close to software and almost getting to the point where, if you claim to be an author, itÕs copyrightable. But, an identical work if you claim to be a software programmer would not be copyrightable. Peter? Peter Deutsch: Hi, Kelly. What I hoped I was saying was that I believe that the literal expression of software, that is the code, should be copyrightable. With respect to the example that you raised, I donÕt believe that either the idea of that kind of hypermedia work, nor the specific algorithms used to choose paths on the basis of past explorations should be copyrightable. I believe that the specific set of words, images, and links, taken as a whole, should be treated as an expression, just as a complete written work is an expression. In other words.. Pamela Samuelson: That is how traditional copyright law would view the issue. Peter Deutsch: So I would be legally, under your definition, within my rights as a reader of this document, to create a new algorithm which had a provably identical effect for every reader, I would not be violating your copyright if I wrote a different code but used the same algorithm. Kelly Booth: No, on the contrary, that is if the work that resulted from your activities was the same work, in the sense of producing the same effects when someone read it, then you would be violating copyright. Peter Deutsch: I think that would be embodied in the algorithm. John Barlow: You know that what youÕre raising here though is a very important point. ThereÕs a fellow at UCLA that talks about secondary orality as being characteristic of the culture that weÕre moving into. We used to have, for a long time in the Western tradition, the ability to fix pieces of expression. But now that theyÕre all fluid, anybody can take those pieces of expression and re-express them. ThereÕs a band called KLF thatÕs doing quite well by taking other peoplesÕ work and remixing it in all kinds of various ways. ItÕs a real serious question as to whether or not what theyÕre doing is a violation of copyright. Kelly Booth: If I might just close with the observation that if you think that two identical works, meaning that no matter what path you take you always get the same result, you quickly get into undecidable issues which would really be fun for the lawyers. Pamela Samuelson: If you will give me your card after this event, I have a paper that I just wrote on that subject and IÕd be glad to send it to you. Michel Denber: LetÕs go on please. On this side. Q: Hi, IÕm Tim Kay at Cal Tech. I think all of the speakers were terrific so far. My question is addressed to the chairman. It seems like weÕre preaching to the already converted and outside the patent and trademark office, are there no proponents of look and feel copyrights and patents, software patents? Michel Denber: Well, IÕm glad you asked that question because as it turns out that is in fact a problem. In fact, thatÕs three problems. Unfortunately, the problem that IÕve been wrestling with for just about a year now is that a lot of the people that are in a position to argue in favor of that are not the kind of people who would ordinarily attend SIGGRAPH. Many of the strongest proponents of this are in the legal profession and would not ordinarily come out here. ThereÕs a problem with getting SIGGRAPH to pay expenses, because SIGGRAPH generally has a policy of not paying expenses for people to participate on panels. Problem number two is that as soon as I talk to people about this and tell them a little bit about what SIGGRAPH is, they sort of suddenly turn very wary of being mugged in a dark alley or something on the way out. So thereÕs sort of a natural disinclination in that they realize the type of audience this is and they feel that they donÕt have any hope of converting people and they just donÕt want to be standing up here and having rotten fruit thrown at them. The third problem is that, what is the third problem, John Barlow: They are mostly busy in court is the third problem. Michel Denber: Yeah. Pamela Samuelson: And the fourth problem is that people who work for some of the litigants in cases are forbidden by their lawyers from attending meetings like this to argue in favor. They want to wait until after they have the solid decision from the courts and then theyÕll come in and tell you that this is the law now and youÕre going to have to follow it. Michel Denber: Yeah, thanks Pam, thatÕs exactly what I was going to say. That did happen to me; I got a number of turn-downs for exactly that reason. Either they didnÕt want to personally because of liability, or their companies told them not to show up here because their lawyers said, ÒSorry, you better not talk about this kind of stuff.Ó So I did the best I could; I realize that the panel may be biased. Maybe thatÕs some kind of Freudian reflection of my own opinions, which IÕve tried to keep out of this. But itÕs a problem putting together something like this and getting a completely balanced treatment of the topic. Q. IÕm Joel Merrit from Intelligent Resources. I have a question for Professor Samuelson and whoever else might want to contribute. Basically what I was wondering is, if I encounter a patent which I feel is clearly demonstrated in prior art, or something very close to it is demonstrated in prior art, what recourse is available to get the patent overturned or limited, or something to that effect? Pamela Samuelson: ThereÕs something called a re-examination process, if you have a prior art reference that is absolutely four-square that shows that this patent was issued on something that was prior art, you can send that prior art reference to the patent office, and for a fee which isnÕt huge (but isnÕt nothing either), you can ask the patent office to re-examine the patent and take into account this prior art reference. Something that companies tend to do instead is to hold the prior art reference and go ahead and use the thing anyway, and see if the other people sue them, and if they do then disclose the prior art reference in the course of negotiations. It may end up being that you could license it on a low royalty basis. Right now thereÕs a proposal being considered to have a re-examination of a wide range of software patents because thereÕs been so much publicity about bad software patents, and in order to restore the publicÕs confidence in the patent system as applied to software, having the patent office initiate some re-examination on their own, or allowing people to mail things in. This might be a way of doing it. ThereÕs also an Office of Technology Assessment report that is due to the U.S. Congress, I think sometime this fall, that may contain some recommendations about how this Òbad patentsÓ problem can be handled. Michel Denber: Okay, on this side please. Q. Stewart Feldman, Lightscape Graphics. What is your practical advice to small developers who may or may not be using algorithms that are patented? You donÕt knowÑthere are so many of them out there. Is the advice to forget about it? Is the advice to hire lawyers to look into all of the things that youÕre doing? How would you handle it? Michel Denber: Are you asking for legal advice here? Q. I am asking for practical advice. Peter Deutsch: Okay, write to Congress, cross your fingers, and hope that nobody notices you. John Barlow: Just do it! Michel Denber: I should add that that advice is probably more true than people know, Òcross your fingers and do it anyway.Ó That happens a lot in industry, so I think that might not be a bad line to pursue. Of course IÕm not a lawyer but we have our own lawyer here. Do you have anything to add to that Pam? Pamela Samuelson: Well, part of the problem right now with software patents is that the classification system that the patent office has is so screwed up, as to software, that if you hired a lawyer to do a prior art search, itÕs not clear at all that the patent office would be able to find a prior art reference. Think about it: for thirty years, they didnÕt accept any software patent applications at all. So sometimes people would go ahead and claim programmed computers, claim a machine that did a particular function as though it was just a piece of hardware and not for software implementations. That would be classified under whatever kind of hardware that thing was. If itÕs really a software invention, itÕs not going to be found in the part of the system where software now is classified, and they donÕt have any patent examiners with any computer science background. In fact, until I think about a month or two ago, it wasnÕt acceptable to work for the patent office and to have a degree in computer science, mathematics, or any other software related subject because only hardware was considered patentable. They donÕt have anybody with expertise, they donÕt have any prior art database and they donÕt have a classification system. So I would say that youÕre better off, unless you know of a patent for sure, crossing your fingers and doing your work and hoping that the world treats you well. John Barlow: Just get big enough quick enough to afford some good lawyers when they come for you. Michel Denber: Okay, letÕs go back to the center isle with the back microphone. Q. My name is Joe Sykose from Purdue University and also an independent consultant. First of all, maybe Pamela can shed some light on this, but itÕs not exactly clear what is patentable and especially if you look at this BCD to binary conversion, I donÕt see much difference between that and say, the alpha channel. Both of them are equally expressible in hardware and software. IÕm going to follow this with a kind of a personal view and then another quick question and turn over the mike. I share many ideas with Peter Deutsch, especially with patenting software concepts that have existed for years and some for many decades. For example, I believe that AT&T has patented the concept of include files. Peter Deutsch: Not exactly. There was a good deal of discussion on that on USENET and it turned out that they patented something more limited. It probably doesnÕt cover include files per se. Q. Okay. Do you see any software patents extending to new concepts of software applications? For example VISICALC was a spreadsheet application, and while we did this with paper and pencils and produced the same results, the new thing was that there was a computer program available to do this and then later followed a whole variety of companies which produced spreadsheet programs. I think that if this were patentable it would sure stifle freedom of speech, because people wouldnÕt be able to express comparable approaches and different ideas to solving the same problem. Thank you. Peter Deutsch: I know that your question was directed primarily to Dr. Samuelson but IÕd like to make one real quick comment on your last point, which is that the fact a patent was issued to RSA Incorporated on a what appears to be a class of algorithms for public key crypto systems is in fact very likely to have the kind of stifling effect that you just named. RSA has been very, very aggressive in terms of defending what they believe are their rights against not only people implementing their particular algorithms, but people implementing algorithms that are mathematically related. So thereÕs at least one pretty clear example of the kind of danger that youÕre talking about thatÕs in process right now. Pamela Samuelson: I think that itÕs worth noting that Dan Bricklin went to a lawyer to see whether he could patent some parts of the VISICALC program and was told by lawyers at the time that, no, software wasnÕt patentable and so he didnÕt pursue it further. The general view of the lawyers today is that if Bricklin came to them today, they would say something different and that the spreadsheet market would obviously have been affected by that. The issue of what is patentable depends on what the patent office has as a policy. At the moment I could give you this eight page set of guidelines they have about what is patentable about software. You could read them and you wouldnÕt understand them because they make distinctions which donÕt make a lot of sense at the moment. But they are issuing some very, very broad patents on software implementations and it does seem to me that itÕs partly the breadth of some of the software patents that are of concern, not just specific algorithms. Michel Denber: Okay, can we go on please. On this side. Q. IÕm David Levine from Microsoft Corporation. I have a question for John Barlow. You mentioned at the end about some particular problems in this area that will be coming up with multimedia and I was wondering if you could elaborate on that. John Barlow: Well the difficulty with multimedia and copyright is that (I think Isaac Kerlow expressed it pretty well), if youÕre going to take an image and youÕre going to somehow track the ownership of that image and see to the payment of royalty through all the various different manifestations and changes and re-expressions of the image, thatÕs going to have a stifling effect on the development of real broad based multimedia fairly obviously, because I donÕt think that multimedia is going to work very well unless itÕs got a great deal of random access to a lot of different kinds of images. If you had to do the equivalent of a title search every time you do that random access, youÕre not going to have multimedia and itÕs as simple as that. Isaac Kerlow: But I donÕt think that when you do a search, you have to pay royalties by the search. You pay royalties for the right to include that image in your database regardless of how many times you search for it. I disagree, I donÕt think that paying royalties for an image that is used in a multimedia project will stifle the development of the field. I think if you donÕt want to pay twenty dollars to use someoneÕs photograph of a mountain, you can take a plane, get your camera, go to the mountain and three thousand dollars later you have a photograph of a mountain. I think that paying royalties for something that someone invested a lot of time, money, and talent creating is a fair thing to do. John Barlow: Dependent on your source though. I mean that IÕd love to agree with you but the problem is that you may not get that from anything close to the original source of the image. It may be floating around cyberspace for a long time before you get it. Isaac Kerlow: I think that I agree with you about the point you mentioned earlier that there could be stifling effects of patent law and copyright on multimedia but, I think that would be more like awarding someone a patent for an algorithm that allows a specific type of computer to do certain things with a video disc player. If you award a patent on something as general as that, then you would be giving one company the monopoly on that technology and I think thatÕs where we both agree. I donÕt think that should happen. Pamela Samuelson: What we may really need to do is to invent some new institutions and so that the, I mean ASCAP and BMI are the examples that lawyers typically trot out as something which basically solves the transaction costs associated with redistributing money. But I think that the transaction cost problem is a real serious one. John Barlow: Speaking as a member of ASCAP, I hope that model is not followed. There are some serious problems with it. Michel Denber: Okay, center isle front please. Q. Hi, IÕm Joel Welling from Pittsburgh Supercomputing Center and I have a question for Pam Samuelson. LetÕs suppose that I have just written a hundred thousand line program or something like that, what would the cost in dollars be for me to hire the necessary people to find out what patents I was infringing and what do you think, if given the rate that software patents increase, that cost would be like in a couple of years? Pamela Samuelson: Oh, it will rise for sure. The rough estimate that you hear for doing patent searches is in the neighborhood of maybe two thousand dollars. Q. But IÕm likely to infringe on a awful lot of patents with my hundred thousand line program. Pamela Samuelson: It could get big, I donÕt know. I think that the point youÕre making is, think about it for a second, how much do lawyers generally charge per hour? Very few of them charge less than $100 an hour. Figure out yourself how long would they be likely to have to search. At present the patent office doesnÕt have that much in their files, so now it doesnÕt take terribly long. Although, as I mentioned before, the classification problem makes it difficult to be really sure that youÕve gotten everything thatÕs out there. ItÕs definitely going to cost a lot of money. Michel Denber: Okay, far aisle please. Q. IÕm Randall Kreps from Evans and Sutherland. I think that itÕs very appropriate for this discussion to be taking place in Las Vegas, Nevada, as many of the issues here are based on the almighty buck. IÕd like Mr. Barlow to respond to the following statement: The software developer has the same relationship to his customers as the televangelist has to his congregation. Both the Hacker and the Preacher have a moral obligation to fleece the suckers for everything they can get away with. John Barlow: Gee, I thought that was America in general! Well you know, thereÕs a correspondence in the sense that both of those individuals are providing a service for which people are willing to pay, doing so with skills that are unique to them and abilities to express those skills that are unique to them. I think that generally speaking, the market determines their value and pays them accordingly. ThereÕs plenty of software that I donÕt think that IÕd want to pay as much money for as they charge and there are plenty of televangelists that IÕm not about to give any money to. Michel Denber: Okay, time is running short, if we could get in as many more questions as we could. Please keep your questions short and hopefully your responses as well. Center aisle in the back, please. Q. Yes, my name is Ken Terkowski from Apple Computer and IÕd like to bring up an issue to the panel in general about the realities of litigation in dealing with technological patents. Generally a company will go after a small company and try to beat them out of the patent game and try to drive them out of business or otherwise get a lot of money from them and then they successively go after other small companies. Is there a possibility for these small companies to somehow get together and increase mass so that it would be a little bit more difficult for the larger company to do that? Are there any comments? Peter Deutsch: ItÕs funny that you should mention that, given that you work for Apple, not that, no, not for the reason that you think that IÕm going to mention. A friend of mine with whom I have violent disagreements on this exact subject, Paul Heckel, managed to do exactly the same thing in reverse and Apple was the first company that he managed to (and you should pardon the expression), extort blackmail money from, by bringing to them a patent that might possibly have been decided as infringing certain aspects of HyperCard. So, I think the problem works in both directions, in all directions. Increasing the mass of small companies is not going to solve it. John Barlow: You bring up an interesting point though, which is that most people justify increasingly severe restrictions on intellectual property because of the protection that this will give the little guy against the big guy. In actual practice, what I see happening is those restrictions are used by the big guy to quash competition from little guys. So this is to me a strong argument for much more liberal laws on intellectual property. Michel Denber: Okay, can we go on? This side please. Q. My name is Carley Cardon, IÕm with Tarna Systems. For Peter Deutsch, two questions. One is what is the status of protection for computer language design, for computer languages themselves like C, FORTRAN, and so on, and what is the fundamental difference between a computer language like C, and a Lotus screen, and Windows? Those are programming languages as well, they just have a visual component. Peter Deutsch: Okay, Dr. Samuelson I hope, will correct me if IÕm wrong about this. There have been attempts to assert that computer languages are copyrightable. My understanding is that all that has actually been established in law is that the book or the writing describing the language may be copyrighted but that the design of the language itself may not be. There was a great deal of furor when Adobe attempted to aggressively assert copyright ownership over the PostScript language. This happens to be a particular issue for me because my consulting companyÕs only product is a PostScript renderer. The other languages that you mentioned, as far as I know, no attempt has been made to assert ownership rights over them. So at the present moment, I believe that in practice, languages are not copyrightable. I would like Dr. Samuelson to comment after I finish addressing your second point. With respect to your second point, the command interface to a system like Lotus or a command system or a system like Windows is indeed a language. I donÕt think that it should be copyrightable. On the grounds that I mentioned before, that for someone to bring out a spreadsheet with essentially the same or similar functions to the Lotus spreadsheet but with a different system of menus, or a different particular set of commands is what I call a gratuitous incompatibility which hurts the consumer but doesnÕt benefit anyone. And so I donÕt think that languages per se should be copyrightable. Michel Denber: Center please, front. Q. Jackie Byrd, student at the University of Illinois in Chicago. This is directed towards the entire panel. What intellectual property rights do computer art students have for original software design, interfaces, applications, and processes which theyÕve produced while students in the university and possibly after theyÕve left the university? Isaac Kerlow: ThatÕs a good question, that IÕve heard before many times. All I can say is that it really depends on what University youÕre talking about, because different universities and different colleges have very different rules. Some universities will tell you, from the very first moment you apply, that anything that you do at the university belongs to them. Of course if you are accepted and enroll and study there, you have agreed to that condition. Whether thatÕs right or not is not even discussed. ThatÕs just the way it is. Other schools donÕt have such a rule. The tacit understanding or the written understanding is that students can do whatever they want with their work. For example, at Pratt Institute where I teach, we have this understanding that the copyright belongs to the student. Since weÕre not structured as a research lab thatÕs interested in marketing technology after it has been developed, we let the students have the copyright, but we do require that they leave a copy of their work at the university for teaching purposes. In other words, I donÕt want people to walk away with everything that they did without leaving something that I can show other students, because that would defeat the whole teaching environment. But many universities will tell you from scratch that they own the work that you do; thatÕs debatable. Michel Denber: Okay, we have time only for one or two more questions, I think. Far aisle please. Q. IÕm David Oliver with the University of Massachusetts and this is directed to Dr. Samuelson. You started by making the distinction between the so-called natural law thatÕs practiced in other countries and the government granted laws that are practiced here. Can you tell us what the practical distinction, that is the distinction that exists today, what would be the ramification of that in other countries? Pamela Samuelson: Well, the clearest example is with the concept of what is known as moral rights. If you have a view that any creation by an individual is a kind of expression of their personality and that they have a natural right to that thing as an expression of their personality, then after that thing has been sold, say to some member of the public, does the person with this natural right in it have a continuing existing interest in what that person who bought it does with it? In natural right countries the answer is yes, so if for example you buy a painting or a sculpture, letÕs say in France, and then you decide to tie red ribbons around neck of the sculpture or paint a mustache on the painting, you are violating the moral rights of the authors because thatÕs like defacing them personally, whereas in the United States, all weÕre interested in is money. As a consequence, if you buy a copy of a sculpture or if you buy a copy of a painting (well, painting we actually just did something special with), but in general copyright law says that if you want to put red ribbons on it or if you want to add new things to make a sculpture on top of it, you are free to do that. You can even destroy it. The original artist has no interest, no underlying interest. They still own the financial interest of the copyright but they donÕt have any kind of abiding interest in that physical manifestation. Michel Denber: Okay, last question, center back please. Q. My name is Francis Hammit, IÕm with CyberAge Journal. IÕve been covering various forms of technology for about twelve years and one of the things that IÕve observed is that when it comes to copyright, patent, and other intellectual property, thereÕs a lot of hypocrisy. You read the license on a software package and it promises everything short of death and damnation if you violate it, but itÕs completely unenforceable. On the other hand, you talk to a lot of people who produce these packages, they feel perfectly free to copy the work of others beyond fair use. I suggest that you canÕt have it both ways, that the industry has to decide what constitutes fair use, whether or not they want open and free distribution, and stop trying to beat each other over the head with lawsuits. Any comments? John Barlow: Right! Michel Denber: Okay, and on that note, I think that everybody is eager for lunch. I want to thank our panelists and thank you all for coming this morning. Copyright © 1991 by the Association for Computing Machinery, Inc. Copying without fee is permitted provided that the copies are not made or distributed for direct commercial advantage and credit to the source is given. Abstracting with credit is permitted. For other copying of articles that carry a code at the bottom of the first page, copying is permitted provided that the per-copy fee indicated in the code is paid through the Copyright Clearance Center, 27 Congress Street, Salem, MA 01970. For permission to republish, write to Director of Publications, Association for Computing Machinery. To copy otherwise, or republish, requires a fee and/or specific permission. .