816 1 UNITED STATES DISTRICT COURT SOUTHERN DISTRICT OF NEW YORK 2 ------------------------------x 3 UNIVERSAL CITY STUDIOS, INC., ]et al, 4 Plaintiffs, 5 v. 00 Civ. 277 (LAK) 6 SHAWN C. REIMERDES, et al, 7 Defendants. 8 ------------------------------x 9 July 21, 2000 10 9:00 a.m. 11 Before: 12 HON. LEWIS A. KAPLAN, 13 District Judge 14 APPEARANCES 15 PROSKAUER, ROSE, L.L.P. Attorneys for Plaintiffs 16 BY: LEON P. GOLD CHARLES S. SIMS 17 SCOTT P. COOPER 18 FRANKFURT, GARBUS, KLEIN & SELZ Attorneys for Defendants 19 BY: MARTIN GARBUS ERNEST HERNSTADT 20 DAVID ATLAS 21 22 23 24 25 817 1 (Trial resumed) 2 THE COURT: A couple of preliminary things on my list 3 before we get started. First of all, have you been able to 4 reach agreement or otherwise come to a view on the specific 5 findings of the Microsoft case that I asked whether you would 6 stipulate to? 7 MR. SIMS: Your Honor we haven't yet. We will over 8 the weekend. 9 THE COURT: Just so I am clear, if I don't hear an 10 objection on the specific findings that I have identified, I 11 will take judicial notice of them and the deadline for any 12 objections is 11 o'clock on Monday. As I say, they are in the 13 nature of definitions and I -- 14 MR. GARBER: I can't see why it would be a problem? 15 MR. SIMS: I am advised we have gotten to it, we have 16 no objection. 17 THE COURT: You might consider also because I will be 18 looking further at that opinion for other things that might be 19 useful by way of background here, to see whether there is a 20 description in it of Linux that you can agree on as to what it 21 is, what its relationship to Microsoft windows is. You may be 22 able to do that, you may be able to save time. OK. 23 The second item on my list is that I am going to 24 alert you to an issue I thought I had because I am going to 25 ask for a brief on this if by the end of the trial I think it 818 1 is relevant and you might as well start thinking about it. If 2 I am understanding at least certain of the arguments here, the 3 defendants are arguing that there are encryption utilities 4 other than DeCSS that would decrypt CSS protected DVDs, that 5 there is no proof, direct or otherwise -- as to direct I think 6 it is stipulated -- that any particular decrypted movie that 7 may have been offered over the Internet or on a file-sharing 8 utility or whatever other relevant modes of distribution may 9 be, that was decrypted with DeCSS as opposed to some other 10 decryption utility and what the defendants I think would ask 11 me to conclude from that is that the plaintiffs therefore 12 can't have had any injury actually or threatened by reason of 13 these defendants' actions. 14 Just by way of a parenthetical digression, I am well 15 aware that the issue of the extent to which there is 16 sufficient proof that there are in fact decrypted movies 17 available over the Internet has not been decided and we are 18 going to have further discussion about that. My question 19 presupposes, just on a hypothetical basis, if there is 20 adequate proof to find that there are such movies out there. 21 The thought that occurred to me last night and on 22 which I may want briefs is whether the burden of proving that 23 whatever decryptions are out there came from the defendant or 24 more broadly DeCSS is on the plaintiff as distinguished from 25 the burden being on the defendant to prove that they did not. 819 1 The case that prompted me to think about that, I know we all 2 read this in law school a long time ago, Summers v. Tyce, 3 decided by the California Supreme Court in 1948 which seems at 4 least superficially to be analogous to this problem. 5 The case involved a group of hunters out hunting with 6 shotguns, two hunters fired simultaneously. One shotgun 7 pellet hit the plaintiff. Because they were shotgun pellets 8 as opposed to bullets out of a rifled weapon, it was 9 impossible to tell which shotgun the pellet came from. The 10 California Supreme Court said that the burden of proof was on 11 the defendants to show that their shotgun did not discharge 12 the bullet that hit the plaintiff and in the absence of such 13 proof, everybody who fired a shotgun at that time was jointly 14 and severally liable. 15 I am not inviting argument on that this morning, but 16 I wanted you to know that I am thinking about that and to give 17 you an opportunity to address it at the appropriate time. 18 OK. That said -- and I should add there is a lot of 19 development in the law of products liability, I am sure you 20 all the know the case Summers v. Tyce as its starting point 21 and some of you may know that. Let's continue where we are. 22 Mr. Corley, come on up. You are still under owes. 23 MR. GARBER: We had that open question before you 24 concerning certain aspects of Mr. Corley's testimony. I 25 presume we are not getting into it until such time as a 820 1 further discussion of the court. 2 THE COURT: Absolutely. Absolutely. If it is 3 something you are going to want to go into Mr. Gold, tell me 4 you need to talk to me privately with Mr. Garbus. 5 MR. GOLD: Thank you, your Honor. 6 CROSS-EXAMINATION 7 BY MR. GOLD: 8 Q. Good morning, Mr. Goldstein. 9 A. Good morning. 10 Q. We left off yesterday discussing a statement that you had 11 made and I am going to give it to you again. Did you not at 12 one point in time, after October of '99, state that DeCSS was 13 a free DVD decoder that allows people to copy DVDs? 14 A. That was a statement that was made on our web site which I 15 did not write but I take responsibility for. It was at some 16 point after it was published on our web site I believe in 17 November that we realized that it was a lot more complicated 18 than just that and we -- future writings on the web site 19 reflected this. 20 Q. What do you mean when you say it was a lot more 21 complicated than that? 22 A. It wasn't simply a bit of code that allowed someone to 23 copy DVDs. At that point in time, I wasn't even that familiar 24 with the technology. I wasn't aware of how the encryption 25 worked, I wasn't aware of the extent of existing DVD piracy 821 1 and wasn't aware that simply copying a DVD was a trivial thing 2 that had been done for years, so I misread the actual facts of 3 the case. When I realized it was more complicated than what 4 we had reported, the case actually became very much more 5 interesting to me. 6 Q. Did you discover after you first made the statement we 7 have just referred to that it was incorrect? 8 A. Did I discover -- first of all, I didn't make the 9 statement. I discovered after looking at various public 10 forums, discussions about the subject, that this was not a 11 case about copying. It had nothing to do with copying. It 12 was about something called access control which was, prior to 13 November, I had been unfamiliar with that. 14 THE COURT: The question, Mr. Corley, wasn't about 15 what this case was about. The question was about the 16 statement that appeared on your web site. 17 THE WITNESS: I am sorry. Not the case but what was 18 going on at the time, the people whose web sites were being 19 shut down, the story itself, that apparently was not about 20 copying DVDs as we had initially stated. 21 Q. Let me ask you, Mr. Goldstein, is it not true that DeCSS 22 is a free DVD decoder that allows people to copy DVDs? 23 A. I believe it allows the data to be copied to a hard drive 24 at some point during its operation as do many other utilities 25 but I believe that it can do that if you write it in a 822 1 particular way. 2 Q. Then the prior statement you made was correct, I guess? 3 A. Not entirely because it did not exist for the purpose of 4 copying DVDs. One would not run that program to copy DVDs. 5 Q. But you said that it was a free DVD decoder that allows 6 people to copy DVDs and if I understand your testimony this 7 morning, you are saying the same thing? 8 A. It allows you to do that but that's not the purpose of it. 9 Again, I am looking at this from a journalistic perspective. 10 Q. I am not asking you the purpose, I am asking you what it 11 does. 12 MR. GARBER: I will object to Mr. Gold cutting off 13 the witness' answer. 14 THE COURT: The witness is being unresponsive. 15 A. I am explaining it as best I can. Basically the program 16 exists to allow someone to bypass the CSS encryption on the 17 DVD. Now, whether that's the purpose of that is to copy a 18 DVD, that's something that we initially didn't understand 19 properly and I realize, I realized in November that it was -- 20 that was not the case at all, that it was about defeating 21 access control, being able to have a player that worked under 22 the Linux operating system among other things. 23 Q. You began posting DeCSS on your web site in November of 24 1999? 25 A. That's correct. 823 1 Q. Is it your testimony that you did that as a journalist to 2 write a story? 3 A. That's correct. 4 Q. Could you have written the identical story without the 5 posting, using the letters DeCSS as many times as you wanted 6 in the story? 7 A. Not writing a story that would have been respected as a 8 journalistic piece, no, because in a journalistic world, you 9 have to pretty much put up or shut up. You have to show your 10 evidence and in this particular case, we would be writing an 11 evidence without showing what we were talking about and 12 particularly in the magazine that I work for, people want to 13 see specifically what it is that we are referring to, what bit 14 of technology that doesn't work, what new advancement, what 15 evidence do we have and simply saying that somebody else said 16 something just won't cut it. 17 So in this particular case, we pointed to the 18 evidence itself which was already firmly established out there 19 in the Internet world. We just put it up on our site so we 20 could write our perspective on it and show the world what it 21 was all about. 22 Q. Getting back to the question, in November of 1999, was it 23 possible for you to write a story about DeCSS on your web 24 site, using the letters DeCSS next to each other as many times 25 as you wanted, without posting DeCSS? 824 1 A. I will take another shot at it. I -- basically, the story 2 would not hold any value to our readers if we simply printed 3 allegations without showing evidence. 4 Q. I am not asking you, sir, what value your story would 5 have. 6 THE COURT: Mr. Garbus. 7 MR. GARBER: I object. The questions are 8 argumentative. And bad as to form. 9 THE COURT: Certainly the one in the process of 10 winding up is. 11 Do you have another question, Mr. Gold? Ask it, 12 please. 13 Q. In order to post DeCSS, didn't you have to go to your 14 computer and get it? 15 A. Go to my computer and get it? 16 Q. Doesn't anyone have to -- yes, go to your computer, the 17 same one you used to write the story? 18 A. We had to get DeCSS from a computer, not our computer 19 because we didn't have it. We took it from one of the other 20 sites that had it at the time and posted it based on that. 21 Q. Once you did that, when you wrote DeCSS in your story, in 22 effect now you had posted it, isn't that right? 23 A. That's correct. 24 Q. Couldn't you have written the same story using the same 25 exact words and using DeCSS without going out and getting 825 1 DeCSS and posting it? 2 A. No. It would not have been the same story. It is 3 analogous to printing a story about a picture and not printing 4 the picture. People want to see what you are talking about. 5 Q. Does your web site now have a slogan "Stop the MPAA"? 6 A. It is probably on there somewhere, yes. 7 Q. Doesn't it stand out in bold letters? 8 A. I'm not exactly sure where it would stand out, but 9 wherever it is, I'm sure it stands out graphically. 10 Q. And what is the meaning of that phrase, what do you mean 11 to convey by stopping the MPAA? 12 A. Well, it is basically a phrase that I didn't coin myself. 13 It has come from a number of people that see the actions of 14 this lawsuit and other events over the past few months as 15 ominous and something that should be stood up to and that's 16 what people are doing simply by vocalizing that. 17 Q. Have you tried to stop the MPAA by writing stories about 18 the MPAA and describing what they do as inappropriate? 19 A. I don't know if they carry that much weight where I could 20 stop the MPAA by writing a story, but I have written stories 21 to educate people as to the facts of the case. 22 Q. Have you ever tried to stop the MPAA by making 23 contributions to organizations that were espousing the repeal 24 of the antisurf protection laws? 25 A. We have tried to raise funds, I am not in a position to 826 1 make contributions unfortunately, but we are trying to raise 2 funds to help pay for our defense. 3 Q. Have you promoted through your web site the taking or 4 downloading of DeCSS? 5 A. We have promoted the linking to various sites that still 6 have it posted because that has not been ruled illegal in any 7 way. 8 Q. Is it your belief as a reporter that to write a story 9 about trafficking in illegal drugs in Manhattan, you would 10 have to spend time actually trafficking in illegal drugs? 11 A. No, I don't think it is similar at all. One is a computer 12 file, one is an actual substance. 13 Q. Is it true that the number of sites posting CSS has 14 increased since January 2000 when you started linking? 15 A. It is hard to say exactly how many there are at any one 16 particular point in time. There were a lot already up when we 17 started the story. We came in kind of late. I believe CSS 18 was cracked, I am hearing dates of October, and we didn't post 19 the story until nearly a month later. 20 At that point in time, there were already hundreds of 21 sites that had it and that was a good part of the story that 22 we posted. I don't know how many there are now, and I don't 23 know how many came up in January. I do know there was an 24 increase after the preliminary injunction against us and I'm 25 not sure where that stands today or how many sites actually 827 1 have it or how many sites that haven't been reported that have 2 it. 3 Q. Mr. Goldstein, are you a leader in the hacker community? 4 A. Some people may say that. I don't like to refer to myself 5 as a leader but I am certainly somebody who certainly follows 6 that particular aspect of society and I try to speak 7 intelligently on it. 8 Q. You publish the Hacker Quarterly? 9 A. Yes. 10 Q. Your web site 2600 is known to virtually every hacker 11 throughout the United States? 12 A. And quite a number of nonhackers as well, yes. 13 Q. You have appeared on radio and television? 14 THE COURT: We are not going to repeat what Mr. 15 Garbus did yesterday, are we? We went through the TV programs 16 and magazine articles. Let's go, Mr. Gold. He has a right 17 not to like this law and he has a right to say so. 18 MR. GOLD: Absolutely. 19 THE COURT: The question is whether he violated it 20 and that's a different question. 21 MR. GOLD: My point, yes. 22 Q. There was a hacker convention last weekend, is that true? 23 A. That's correct, the third HOPE convention, call H2K. 24 Q. And there was a Mr. Johansen appeared at that? 25 A. Joe Johansen appeared on panel. 828 1 Q. And he spoke? 2 A. He and his father both spoke. 3 Q. Was a mock trial held of this case on -- 4 A. Yes, actually we had scheduled that before the trial had 5 been scheduled so the timing was kind of strange but yeah, 6 that was organized by somebody else. 7 Q. Is it fair to say that the hacker community in the United 8 States is following this trial carefully? 9 A. I would say the hacker community, open source community, 10 the Linux community, yes. 11 THE COURT: What do you mean by the open source 12 community? 13 THE WITNESS: The open source community is basically 14 people who write software, release the source code, share 15 information, it overlaps into the Linux world. I am actually 16 not a part of that community. I didn't know much about that 17 community before this case. 18 THE COURT: OK, thank you. 19 Q. Is it true that the 2600.com site provides a form for 20 people to fill out so they can add their site to the 2600.com 21 mirror list? 22 A. Yes, we have two forms. We have a form where people know 23 about a site that carries DeCSS or know of a site that carries 24 DeCSS, they can submit that, or if they know of a site that we 25 are listing that no longer has it, they can fill that out as 829 1 well. But I should also point out that that has not been 2 worked upon for a couple of months simply because we are too 3 busy with other projects. 4 Q. But the form has been there for many months? 5 A. The form has been there from the beginning, yes. 6 Q. And it still is? 7 A. I believe so. 8 Q. You don't know? 9 A. I haven't looked at our web site in a while but I don't 10 believe any changes were made, no. 11 Q. How long a while? 12 A. I don't look at every page. 13 Q. How long a while haven't you looked at your web site? 14 A. I have looked at my web site. That particular page I have 15 not looked at probably in a couple of weeks but I believe it 16 is still up there because we don't make changes. 17 Q. Do you understand that it is wrong to break through a 18 protective device that protects a digital copyrighted work in 19 order to take that work? 20 MR. GARBER: I object to the question. 21 THE COURT: Sustained. 22 Let me ask you something about the mirror list and 23 some of the sites to which you have linked. Are there sites 24 to which you are linked on which by clicking on the link on 25 your site, you reach a page at the linked site that has a 830 1 variety of content on it which may or may not include DeCSS on 2 the first page that comes up? 3 THE WITNESS: Yes, that's correct. I would say 4 that's probably the case more times than not. 5 THE COURT: Do some of those sites then give the 6 person who reaches it through clicking on the hyperlink on 7 your site then give the user or visitor, whatever you want to 8 call it, the option to go further in the -- I'll call it the 9 transferee site if the visitor wishes to download DeCSS? 10 THE WITNESS: The person is given the choice as to 11 whether or not they want to proceed further? Yes. 12 THE COURT: Are there sites to which you are linked 13 or have been linked where the effect of clicking on the 14 hyperlink on your site is to begin a download to the user of 15 DeCSS without the user or the viewer taking any further 16 action? 17 THE WITNESS: You would always have to take some bit 18 of further action even for something as simple as verifying 19 where you want to put the file on your hard drive. There are 20 no instances that I know of where you can only click once on 21 one of our hyperlinks and immediately start a download. 22 THE COURT: From your understanding, is it possible 23 to construct a link of the type I have described, whether it 24 is on your site or not? 25 THE WITNESS: I imagine anything is possible but I'm 831 1 not familiar with that particular application. 2 THE COURT: Are there links listed on your mirror 3 list in which the user, when the user clicks on the link is 4 transferred to a point on a transferee site that has no 5 content that comes up on the screen other than DeCSS or a 6 dialog box that requires the user to do something like verify 7 that the user wishes to download DeCSS? That is to say, no 8 other content? 9 THE WITNESS: Yes, that's certainly possible. 10 THE COURT: OK, thank you. Go ahead, Mr. Gold. 11 Q. Mr. Goldstein, have you ever publicly advocated that it is 12 wrong to infringe on anyone's copyright? 13 A. Yes, I support the copyright laws as I understand them. 14 Q. Have you, sir, publicly advocated that helping someone to 15 commit a copyright infringement is wrong? 16 A. I believe that helping somebody to break the law is wrong, 17 yes. 18 Q. What do you mean by wrong? 19 A. Illegal, immoral, something I wouldn't do. 20 Q. Now, at the end of 1999, did you know that CSS was a 21 protected device protecting digital copyrighted movies? 22 A. I knew it was an encryption standard that had been applied 23 to DVDs. I did not see it as the same thing as preventing 24 illegal copying, no. 25 Q. Why not? What's the difference? 832 1 A. Again, I'm not a lawyer, so my understanding of the 2 nuances might be a bit vague, but -- 3 Q. Sir, I would like you to not give me any legal opinions. 4 THE COURT: Mr. Gold, he was trying to answer so 5 let's hear the answer. 6 MR. GOLD: Thank you, your Honor. 7 A. My understanding is the protection, the copyright law is 8 meant to protect the owner of the copyright from having 9 illegal works distributed whether by copying or any other kind 10 of infringement where the work or the copyright holder is not 11 compensated for the work. And my analysis of CSS and what CSS 12 accomplished was not the same thing as that. In other words, 13 copying of DVDs was not affected by whether or not one 14 decrypted CSS or one did not decrypt CSS. 15 Q. Did you know at the end of 1999 that the movies made by 16 the major Hollywood studios were copyrighted? 17 A. Did I know they were copyrighted? Yes, of course. 18 Q. Is the quarterly published, Hacker -- forgive me -- 19 A. Hacker Quarterly. 20 Q. Is that copyrighted? 21 A. Yes, we copyright both the magazine and material on the 22 web page. 23 Q. Why? 24 A. We don't wish for people to be able to simply copy 25 everything on our site and claim ownership as theirs. 833 1 Q. Did you make the copyright application or authorize that 2 it be made? 3 A. Yes. 4 Q. Now, by providing links through the 2600.com site, are you 5 making it possible for any member of the general public to 6 download DeCSS if they have a computer? 7 MR. GARBER: I object to the form of the question. 8 THE COURT: What's the objection? 9 MR. GARBER: Let me hear the question again. 10 (Record read) 11 MR. GARBER: I will withdraw the objection. 12 A. If they have a computer and a connection and they choose 13 to go to our site, yes, they can download it that way. 14 Q. Is it your understanding that once someone has downloaded 15 DeCSS through the 2600.com site, neither you nor 2600 has any 16 control over what they do with that, what they did with DeCSS? 17 A. That's correct, we have no control what people do after 18 they leave our site. 19 Q. Did you make any efforts to confine your providing DeCSS 20 to people who were going to use it for any specific purpose? 21 A. I don't think such control is possible on the net. So the 22 answer would be no. 23 Q. Is it true that anyone who downloads DeCSS can use it for 24 any purpose that they want to put it to? 25 A. Any purpose they want -- 834 1 Q. Any purpose at all? 2 A. I'm not -- I am having trouble following that. DeCSS can 3 only do a limited number of things. It can't do anything 4 somebody wants. 5 Q. Is it not true that all DeCSS can do is decrypt CSS? 6 A. DeCSS exists for the purpose of bypassing CSS, correct. 7 Q. Is it true that your posting of DeCSS and your linking to 8 other sites that post DeCSS is being done for the sole reason 9 that you are a journalist? 10 A. The reason the story appeared was because it was a 11 journalistic piece. The reason we continued to write about it 12 and talk about it is because we believe it continues to be 13 that kind of a story. 14 Q. Writing it and being a journalist was your only purpose of 15 posting or linking to other sites that -- 16 A. That's the purpose for our site, that's the purpose for 17 our magazine. It is a journalistic endeavor, yes. 18 Q. Did you ever write that the mirroring of DeCSS was a 19 demonstration of electronic civil disobedience? 20 A. I can't say that I wrote those specific words but I 21 believe those words may have appeared on our web site. 22 Q. And you're responsible for them? 23 A. I take responsibility for what appears on our web site, 24 yes. 25 Q. What is your definition of civil disobedience? 835 1 MR. GARBER: I object. 2 THE COURT: Sustained. 3 Q. When those words appeared on your web site, what did you 4 understand them to mean? 5 A. My definition of civil disobedience, electronic civil 6 disobedience, I suppose simply people taking a stand in a way 7 they perceive as morally just. I really think it should be 8 left up to the individual person to make that definition on 9 their own. That's how I would picture it on the web site. 10 Again, I don't know the exact context of how it appeared on 11 our web site. I would have to look at that. 12 Q. Is it not true, to your knowledge, that civil disobedience 13 means violating some law for the purpose of making a 14 statement? 15 A. It many cases, it does involve sitting in front of a door, 16 for instance, a minor violation to prove a point, yes. 17 Q. It is limited only to minor violations? 18 A. It can be bigger than that, I suppose. 19 Q. Now, do you write a news story on your web site or does a 20 news story appear on your web site relating to DeCSS every 21 single day? 22 A. I write a story occasionally. There are other people who 23 write stories occasionally on the web site. They do not 24 appear every day, no. 25 Q. How often in the course of a week in the last three months 836 1 have stories about DeCSS appeared on your web site? 2 A. I would say probably an average of about one a week. And 3 that's simply because we don't have very much in the way of 4 staff. 5 Q. But I gather that you link to sites that post DeCSS every 6 single day, every single minute, every hour of the year? 7 A. It is not an ongoing thing. We are not consciously doing 8 it -- basically every story we write on our web page stays up 9 until the end of time or until we are ordered to take it down. 10 Q. Is it possible for you to remove it? 11 A. Yes. 12 Q. You haven't done that? 13 A. No, I have not. 14 Q. And you don't take it down on each day when you are not 15 writing a story about DeCSS, is that true? 16 A. No, we believe in keeping our stories up even when a story 17 has a factual error in it. Our philosophy is it would be 18 wrong to rewrite history and pretend we didn't say something 19 that was wrong. 20 Q. Have you made any money by posting DeCSS or linking it to 21 other sites that post DeCSS? 22 A. I would strongly suspect not. 23 Q. Why would you only suspect? Do you know? 24 A. I'm certain I haven't made any money from posting DeCSS. 25 I have probably lost quite a bit of money because I have not 837 1 been able to devote my time to the things I am supposed to be 2 doing. 3 Q. There has been a great increase in people coming to your 4 web site since you first started to post DeCSS, is that not 5 true? 6 A. I have no way of knowing that since we haven't kept a 7 counter since last summer. 8 Q. Wasn't that because so many people were coming to your web 9 site? 10 A. That was last summer before all this started. I have no 11 way of knowing how many people are hitting our site. We have 12 no ads on our site so we have no economic need for more people 13 to come there. I do note a lot of people are talking about 14 it. I can suspect that more people are coming to our site but 15 I have no real way of proving that. 16 Q. I gather al of your income comes from the Hacker 17 Quarterly? 18 A. That's correct. 19 THE COURT: Mr. Gold, isn't that whole line 20 irrelevant, in line of Greeging and a whole line of cases all 21 the way back to Greeging? 22 Q. Have you talked with anyone at all with about 23 advertisements in the Hacker Quarterly in the last several 24 months, the possibility of advertising? 25 A. No, our policy is never to accept advertising. 838 1 Q. And you have had a discussion with no one about that? 2 A. No. 3 Q. I gather, Mr. Goldstein, that you have never been involved 4 personally in reverse engineering? 5 A. No, I'm not an engineer. 6 Q. And I gather you have never been involved in cryptographic 7 research? 8 A. No, I am not a cryptologist. 9 MR. GOLD: Your Honor, I believe it would be 10 appropriate to have a sidebar at this point before I get into 11 another subject. 12 THE COURT: Come to the sidebar. 13 14 (Continued on next page) 15 (Pages 839-842 filed under seal) 16 17 18 19 20 21 22 23 24 25 843 1 (In open court) 2 MR. GARBER: I would ask that this part of the record 3 be deemed confidential. 4 THE COURT: The transcript of the sidebar will be 5 separately bound and filed under seal. It will be available 6 to counsel but not otherwise. 7 Implicit in that of course is there will be no 8 discussion by counsel or revelation to anybody about what the 9 sidebar was all about. 10 THE COURT: Anything else, Mr. Gold? 11 MR. GOLD: Your Honor, I would like to offer pages 58 12 and 59 of Mr. Goldstein's deposition transcript as an exhibit 13 in the record at this point. 14 THE COURT: You have your objection, Mr. Garbus. I 15 take it subject to the objection. If I don't refer to it in 16 my decision, you should assume that I sustained the objection. 17 BY MR. GOLD: 18 Q. Mr. Corley, I have noticed that a picture of a telephone 19 and a telephone booth appears to an appear on the back cover 20 of every issue of the Hacker Quarterly? 21 A. That's correct. We print foreign pay phone pictures. 22 Q. For what purpose? 23 A. People seem to be interested in what pay phones in other 24 countries look like so they send us their photos when they go 25 on vacation or if they live in that particular country and we 844 1 compile them together and print four pictures on the back of 2 every issue. 3 Q. And the 2600, that title of your quarterly, is that -- 4 does that relate to a situation occurring with frequency in 5 the '80s and '90s relating to phone service? 6 A. No, actually as I said yesterday, that dates back to the 7 1960s. It is something known as in-band signaling where a 8 particular frequency sent over a long distance telephone line 9 cause certain conditions to occur. And committing a 2600 10 hertz tone would drop a subscriber into what was known as 11 operator mode where they could route phone calls, route 12 themselves to internal operator and explore the entire system. 13 That has not worked for quite a while and it is a symbolic 14 type of name. 15 Q. Going back to the description you just gave us, was it 16 possible to use what you described to make phone calls without 17 paying for them? 18 A. That's one use, yes, it could have been used in that way 19 as well. 20 Q. Yesterday we were reviewing certain articles or your 21 counsel was reviewing with you articles that appeared in the 22 Hacker Quarterly. I am not going to go through them line by 23 line at this point. However, do you recall whether that 24 warning about -- that goes something like if I remember it, we 25 don't advocate the breaking of the law so we are not 845 1 advocating that you perform the acts described above, did that 2 appear in every column in the last four or five years? 3 A. It is something that has recurred. We don't print it on 4 every page in every issue. So for instance, if somebody 5 writes a letter to the editor and I think you could point to 6 virtually every expression of this, somebody expresses an 7 intent to commit a crime, we will without fail advise them not 8 to do that or try to convince them that's a really bad idea. 9 Q. Such descriptive articles about acts which were in fact 10 illegal did not contain that warning? 11 A. The articles are written from the author's perspective. 12 If the author wrote the article from a different perspective 13 as ours, there would be no such correction in there. 14 Q. By the way, how do we know that those authors weren't you? 15 A. Well, I can tell you they weren't me but other than that, 16 I don't know how I can prove it. 17 Q. So -- 18 A. Without violating -- 19 Q. So Bull Finch and Crypton and whoever people sign, are you 20 sure those people aren't you? 21 A. I'm very sure, very sure. For one thing, I don't have the 22 technical expertise to write many of those articles. I write 23 the editorial, I write the responses to the letters and 24 occasionally news updates. 25 Q. Do you recall one such article where there was a 846 1 description of how one could climb up a tree on his street 2 where there would be a black box that related to telephone 3 service in the community -- 4 A. I believe you made reference to that yesterday. It is not 5 really a black box. 6 Q. What color is it? 7 A. It is different colors, but beige, white, silver. 8 Q. Those are the boxes? 9 A. Basically that particular story entailed people who could 10 climb poles or go into basements or closets in buildings and 11 make phone calls off subscriber lines. 12 Q. Do you remember in one paragraph, there was a description 13 of how you could climb the tree or pole and get to the black 14 box and whatever color it was and rip it apart so there would 15 be no telephone service to the community? 16 A. I don't recall specifically that paragraph but that could 17 have been something that the author wrote. 18 Q. And you did not write that article? 19 A. No, I did not. 20 Q. Does that article, if you remember, contain a warning 21 about don't break the law and don't do the things that are set 22 forth above? 23 A. No, as I said, we warned people through letters and our 24 editorial policy. We don't insert our comments into other 25 people's articles. 847 1 Q. I am asking if you remember that that particular article 2 did not contain such a warning? 3 A. I don't remember the particular article but our policy 4 states we don't insert our editorial policy into articles. 5 THE COURT: Mr. Gold, I think you made your point. 6 MR. GOLD: Thank you, Mr. Goldstein. Thank you I 7 have no further questions. 8 Oh, yes, if I may, there is one area that I haven't 9 gotten into. I thank my colleague. 10 Q. Mr. Goldstein, is it true that unless you are enjoined, 11 you intend to keep providing links on 2600.com to other sites 12 where DeCSS is available, do you not? 13 A. It is my understanding that that's legal so until we are 14 told otherwise, yes. 15 Q. The answer is yes? 16 A. Yes. 17 MR. GOLD: Thank you. 18 THE COURT: Thank you, Mr. Gold. 19 MR. GARBER: I have one question. 20 REDIRECT EXAMINATION 21 BY MR. GARBER: 22 Q. The judge asked you about linking. If www.garb.com comes 23 to you and asks you to link, do you know what material I will 24 have on my site when you give me permission to link to you? 25 A. Well, first of all, I believe you mean permission for us 848 1 to link to you. 2 Q. Yes? 3 A. What we had been doing while we were keeping the list 4 updated was basically going to the site, seeing if the file 5 names were there, were present, and if so, and this is because 6 we were getting so many submissions, we would just say yes, 7 OK, we will post this one. 8 What we later found out, there were some dummy files 9 out there. Some sites had the file names but they weren't 10 really the files. They were gibberish or even empty files. 11 It is simply impossible for us to go out there and verify that 12 each file is exactly what it says it is. That's one of the 13 reasons we stopped doing it, because it was kind of pointless 14 after a while. 15 Q. Turning to the judge's question, do you know if you are 16 going to link to www.garb.com who allegedly has DeCSS there, 17 whether it is on the first page, second page, fifth page or 18 what other information is contained on that site? 19 A. I am sorry, could you rephrase that a little? 20 Q. Let's, let's assume on your site 2600.com, you had 21 ww.garb.com as a linking site. Do you know -- and that 22 linking site allegedly has something about DeCSS, do you know 23 whether -- it has comments on DeCSS, do you know whether it 24 has the actual mirror and do you know where that appears on 25 the first page, the fifth page of the site? 849 1 MR. GOLD: Objection, your Honor. 2 THE COURT: Overruled. 3 A. We would know when we initially verified the existence of 4 the site if it was on the fifth page or first page or wherever 5 but we wouldn't retain that knowledge and that could change, 6 too. A person could take down DeCSS the day after we link to 7 them and until somebody told us that, we wouldn't know to take 8 it off the list. As I mentioned over the past few months, we 9 haven't maintained it at all. 10 Q. Do you know how many clicks it takes once you get to that 11 site to get to the actual mirror of DeCSS? 12 A. It differs with every site. Some may have a diatribe of 13 have a various political ideology that they want you to read 14 before you get to the particular mirror of the file where you 15 click on it. Some might have nothing but the file name when 16 you get to their site that you click on. So it varies with 17 every site. We have no way of controlling that. 18 THE COURT: Ever heard of the term "deep linking"? 19 THE WITNESS: I have heard the term. I am a little 20 confused by it because I know of one type of linking. I might 21 have an idea of what that is. I think it involves going into 22 a site beyond the main page, I think Ticket Master was doing 23 something like that or somebody was doing something like that 24 to Ticket Master. 25 THE COURT: Now, your web site has a home page, 850 1 right? 2 THE WITNESS: Correct. 3 THE COURT: That's the page you get if you enter the 4 universal resource locator www.2600.com into your browser, 5 true? 6 THE WITNESS: That is right. 7 THE COURT: You have other pages that are, so to 8 speak, behind the home page, right? 9 THE WITNESS: That's correct. 10 THE COURT: Every one of those other pages has its 11 own universal resource locator or URL, correct? 12 THE WITNESS: That's correct, yes. 13 THE COURT: So someone who wished to link to your 14 site and not to your home page but rather to some specific 15 page beyond your home page could enter in his own web page a 16 hyperlink not to your home page, but rather to the page behind 17 your home page in which the person was interested in linking, 18 right? 19 THE WITNESS: That's true, yes. 20 THE COURT: So if, for example, let's just make an 21 assumption, the New York Times, which is www.NYTimes.com, has 22 a web site, the home page has the URL that I just indicated, 23 right? 24 THE WITNESS: That's correct. 25 THE COURT: And let's just suppose that as somebody 851 1 suggested somewhere in the bowels of the New York Times web 2 site is the DeCSS code, all right? Are you with me? 3 THE WITNESS: Yes. 4 THE COURT: That code would appear on a page or out 5 there in hyperspace with its own unique URL, true? 6 THE WITNESS: Correct. 7 THE COURT: Now, if the New York Times wanted you to 8 carry a link to its mirror, you might link to that mirror in 9 more than one way, isn't that true? 10 THE WITNESS: There are several ways you can do it, 11 yes. 12 THE COURT: You could link by inserting the hyperlink 13 to www.NYTimes.com and then leave the user to navigate through 14 the New York Times web site, true? 15 THE WITNESS: Correct. 16 THE COURT: Or you could link by inserting the 17 appropriate URL to the specific page on the New York Times web 18 site that had the DeCSS code, am I correct? 19 THE WITNESS: That's correct. 20 THE COURT: And it was your practice to verify the 21 existence of proposed mirrors before you linked to them, when 22 you put them on your mirror list, right? 23 THE WITNESS: That's correct. 24 THE COURT: So the means of knowing whether the link 25 that you posted on your mirror list took someone who clicked 852 1 on that hyperlink to a page containing a whole bunch of 2 content, some of it relating to DeCSS and others relating to 3 God only knows what, as compared to taking you directly to the 4 source code for DeCSS was in your hands, true? 5 THE WITNESS: If we had spent more time on it, we 6 could have refined it so it took you directly to the actual 7 files, yes. 8 THE COURT: Or not? 9 THE WITNESS: Yes. What we did was we took what they 10 gave us and looked at that and said does this go to DeCSS, 11 either on the first page, second page or whatever and if it 12 did, we just put it in the way they gave it to us. 13 THE COURT: OK. Anybody want any further 14 examination? 15 MR. GARBER: One thing. 16 If the court were to issue a declaratory judgment 17 saying that the posting of DeCSS were illegal or 18 inappropriate, would you stop posting the linking? 19 THE WITNESS: Yes, of course. 20 THE COURT: Mr. Gold? 21 RECROSS-EXAMINATION 22 BY MR. GOLD: 23 Q. One question, do you know what a declaratory judgment is? 24 A. I am sorry. 25 Q. Do you know what a declaratory judgment is? 853 1 A. It is similar to an injunction, preliminary injunction. 2 MR. GOLD: Thank you. 3 THE COURT: OK, anything else? Mr. Corley, you are 4 excused. Thank you. 5 (Witness excused) 6 MR. GARBER: Your Honor, can I approach the bench for 7 a moment. 8 THE COURT: OK, you and Mr. Cooper or Mr. Sims, 9 whoever wants to come. 10 (At the sidebar) 11 MR. GARBER: I think the Summer v. Tyce issue that 12 you raised is an issue that we have been conscious of since 13 the beginning and just to use the analogy, if one of the 14 bullets is defective -- 15 THE COURT: I don't want to hear argument about it 16 now. 17 MR. GARBER: I don't want to also, but I just want to 18 go into the question and relate it and I don't want to get 19 contumacious, the whole question of discovery and the question 20 of robustness is directly related to that issue and it has 21 always been our argument, which we have not had a chance to 22 document through discovery, there is the Merden report, the 23 Macrovision report which indicated that DeCSS was not a bullet 24 in the same way that the other utilities were bullets and it 25 has been, that's been our position from the beginning. 854 1 THE COURT: I am glad to hear that, Mr. Garbus, 2 because this is the very first time you have said it. 3 MR. GARBER: That robustness was an issue and the 4 efficacy of DeCSS -- 5 THE COURT: You just told me the Summer v. Tyce issue 6 has been in your mind from the beginning and the first person 7 to raise it in this case is me on the 5th day of trial. You 8 are trying to make a record for another purpose. If you want 9 to make a motion at an appropriate point for whatever relief 10 you think is appropriate, you make that motion, but we are 11 going to hear evidence right now. 12 (In open court) 13 THE COURT: Mr. Hernstadt, your next witness? 14 MR. HERNSTADT: Very briefly, before you go to the 15 next witness, yesterday we sent a letter in asking the Court's 16 permission to use a computer to do a demonstration. We 17 thought it was for next week but we have managed to bring the 18 witnesses in for today, so we have another letter asking 19 permission to do it this afternoon. 20 THE COURT: Tell me what it is about. Tell me what 21 the demonstration is supposed to be about, Mr. Hernstadt. 22 MR. HERNSTADT: This afternoon we would like to do a 23 demonstration of the LiVid Linux DVD player. The software is 24 still in an alpha -- in other words, even before it is 25 released for general testing, but it is functioning software 855 1 so we would like to show you how it works. 2 THE COURT: All right, I have signed the letter. I 3 didn't see the letter you sent yesterday. 4 MR. HERNSTADT: The other possible demonstration is 5 similar to the one Dr. Shamos did. 6 THE COURT: Mr. Cooper? 7 MR. COOPER: Your Honor, we have had no notice of 8 either of these demonstrations to my knowledge. It is 9 difficult for me to understand the relevancy of the first 10 described demonstration to any of the issues in this lawsuit 11 and we would like an opportunity to get more detail from 12 opposing counsel regarding what the demonstrations are. 13 THE COURT: How long is it going to take, Mr. 14 Hernstadt? 15 MR. HERNSTADT: The demonstration? 16 THE COURT: Yes. 17 MR. HERNSTADT: A few minutes. 18 THE COURT: Relax, Mr. Cooper. 19 Next witness? 20 MR. HERNSTADT: The defendants call Larry Peterson. 21 22 LARRY PETERSON, 23 called as a witness by the defendant, 24 having been duly sworn, testified as follows: 25 DIRECT EXAMINATION 856 1 BY MR. HERNSTADT: 2 Q. Good morning, Professor Peterson. Can you tell us about 3 your educational background please? 4 A. I have a Ph.D. in computer science from Purdue University. 5 Q. What year did you graduate? 6 A. 1985. 7 Q. What was the subject of your dissertation? 8 A. It had to do with several issues relating to computer 9 networks and focused primarily on e-mail and use of e-mail. 10 Q. After you received your degree, where were you employed? 11 A. I then took an assistant professorship position at 12 University of Arizona. I was there for 13 years, last two of 13 which I was department head. 14 Q. Of which department? 15 A. The computer science department. 16 Q. For the last couple of years, where have you been? 17 A. Princeton University. 18 Q. What courses do you teach at Princeton? 19 A. At Princeton, I have been teaching the computer networks 20 class and introductory programming class. 21 Q. What is involved in the computer networks class? 22 A. Primarily I walk students through the motivation behind 23 the Internet, the problems that cropped up over a number of 24 years as the Internet evolved, and how researchers had solved 25 them to evolve the Internet to where it is today. 857 1 Q. How long have you been studying the Internet? 2 A. Since I was a graduate student in 1980. 3 Q. How long has the Internet been around? 4 A. For a few more years than that. In 1980 is when it took 5 on the form that we are familiar with today roughly. Before 6 that, it was a packet switch network called the Arpanet. 7 Because of the advent of ethernet and wanting to bring more 8 networks into the fold, the Internet technology was deployed 9 in the very early '80s. 10 Q. And what are your research subjects? 11 A. My is research crosses between networks and operating 12 systems. Both of them are broadly the study of computer 13 systems. We build systems, we measure them, and evaluate how 14 they work, so it primarily has been focusing on how we can 15 build operating systems so that networks can be more effective 16 for users. We take a very, what we call, end-to-end 17 perspective so we are concerned not just with one particular 18 thing and how fast you can transmit over it or even the fact 19 that the links are catenated together to form the network or 20 that there is a computer attached to either end and 21 application programming or processes right on those computers, 22 so we have been focused on how to get data from an application 23 on this side of the network to an application on the other 24 side of the network. 25 Q. Have you ever researched a high speed connections from a 858 1 network to a computer? 2 A. Back during the early '90s, we were a part of what was 3 called the gigabit test, it was a national initiative to push 4 the edges of gigabit network technology and at that time, what 5 we in particular were doing was trying to move bits between a 6 processor on one computer to a processor on another computer 7 at the same speeds that the link connecting those two 8 computers operated which at the time was 622 megabits a 9 second, technology called OC12, and we were in fact able to 10 move the data from one machine to another at that speed. 11 There are a number of obstacles we had to overcome to do that. 12 Q. Is OC12, could you explain what OC12 is? 13 A. It's stands for Optical Carrier 12. There is a base unit 14 of transmission -- let me go back. There is basically the 15 technology having to do with fiber optics and there is a base 16 rate at which people can transmit which is roughly 55 megabits 17 per second and you keep multiplying that by 2 or 3 or 6 or 12, 18 so 6 times that base rate gives you 622. something megabits 19 per second. It is the way the bandwidth is measured on 20 optical links. 21 Q. Is OC12 used in the Internet today? 22 A. Yes, it is. 23 Q. Where do you find it? 24 A. You would find it in the backbone of the Internet. So let 25 me just give you a broad picture of the Internet and I can 859 1 tell you where different technology applies. 2 You can characterize the very core of the Internet 3 which is a set of long hold networks running across the 4 country from Los Angeles to New York and so on and cities in 5 between as the backbone of the Internet. A number of carriers 6 have provided backbones. In the original days of the 7 Internet, they were typically provided by government agencies 8 so there was the NSF net in the early '90s and eventually 9 commercial companies replaced that. So today you go to Sprint 10 or AT&T and they have a backbone. The links that make up the 11 backbone running between Chicago and Houston or whatnot might 12 be at 622 megabits per second. 13 Out near the edges of that backbone then, you connect 14 in various ISPs, Internet Service Providers, so your local 15 cable company might in one direction provide you a service at 16 home, but they have to turn around and connect into the 17 Internet at the high end, they would possibly use a 622 18 megabit link from their site into SprintNet, for example. 19 MR. HERNSTADT: Your Honor, may I approach? 20 THE COURT: Yes. 21 Q. I have handed you Exhibit BDC. By way of explanation, 22 could you tell us what this is? 23 A. This is just the layout of Sprint's Internet backbone 24 network. I found it on the web a few days ago. It shows 25 links connecting various cities. I know Sprint has a 860 1 significant presence in Kansas City, for example, so it will 2 show you the link running from Kansas City to Fort Worth and 3 collectively this would make up the Sprint backbone. 4 Q. Can you explain what the different links are? 5 A. This is color coded and it is hard to tell for sure which 6 are which, but these are the various links technologies being 7 used in the Sprint backbone, DS3, an older technology, 45 8 megabits per second. OC3 is 155; OC12, 622; 0C48, that's 6.4 9 gigabits per second. 10 MR. COOPER: From a foundational standpoint, can we 11 get some information from the witness about where this 12 document comes from? 13 THE COURT: He said he found it on the Internet a 14 couple of days ago. 15 MR. COOPER: We have not seen it before and I am 16 looking for some indication about how he went about locating 17 it and whether this is some area of his expertise or whether 18 he did some research and pulled it down. 19 THE COURT: Is that about the size of it? 20 THE WITNESS: Yeah, I went to Sprint.com as a good 21 guess of where to find Sprint's backbone, some links off 22 there, I couldn't give you the exact URL, and I found this 23 page. 24 THE COURT: What do you want me to do with it if 25 anything? 861 1 MR. COOPER: I am trying to explore the foundation of 2 documents of this type so we have an even playing field as far 3 as cross-examination is concerned, your Honor. 4 MR. HERNSTADT: Your Honor, we would move this into 5 evidence insofar as it assists your Honor in understanding 6 what this -- 7 THE COURT: Now you are not worried that things that 8 appear on the Internet that are hearsay. But there is no 9 objection so I will receive it. 10 MR. HERNSTADT: Thank you. 11 (Defendant's Exhibit BDC received in evidence) 12 Q. Professor Peterson, have you received any honors or awards 13 in connection with your work? 14 A. Several of my papers have been Outstanding Papers at 15 different conferences, two in particular were Student Paper 16 awards where I was co-author with one of my students at 17 SigComm which is premiere conference for the networking 18 community. 19 Q. Are you an editor now of one of the -- 20 A. I am editor and chief of ECM which is the computer 21 sciences professional organizations tracking on computer 22 systems. That's the premiere journal on computer systems. 23 Q. Have you authored any publications? 24 A. Several. I think I counted this morning over 50. 25 Technical papers. 862 1 THE COURT: Do you have a CV? 2 MR. HERNSTADT: I was about to ask if I could 3 approach. 4 THE COURT: Yes. 5 Q. Professor Peterson, is this your curriculum vitae? 6 A. Yes, it is. 7 Q. Does this accurately reflect your publications? 8 A. Yeah, I believe it does. 9 MR. HERNSTADT: Your Honor, we offer this into 10 evidence. 11 THE COURT: Received. 12 (Defendant's Exhibit BBD received in evidence) 13 Q. Professor Peterson, what is Internet 2? 14 A. Internet 2 is a consortium of 150-odd universities. In 15 the old days, the Internet was a play thing of the researchers 16 at universities, but of course it has become a very commercial 17 entity today and that has impacted the way research is 18 conducted. So primarily research universities have gotten 19 together to build an another piece, another corner of the 20 Internet that they can use for research purposes and done in 21 collaboration with some of the carriers, equipment companies 22 and the like, Cisco and Nortel, I am not precisely sure who 23 all the players are, so that they can have high speed 24 connectivity so their astrophysicists can send huge files to 25 each other in the national labs. 863 1 Q. Will it replace the Internet? 2 A. No, it is supplemental purely for the use of education. 3 In fact, to join the consortium, you have to agree you will 4 only use the connectivity and bandwidth for research and 5 teaching purposes. 6 Q. Are you being paid for your appearance today? 7 A. No, I am not. 8 Q. Or for any of the work you have done in connection with 9 this matter? 10 A. No. 11 Q. Can you describe how the Internet works by, for example, 12 if you were to send a file from your office to a colleague's 13 office, what are the various steps it would take? 14 A. As I have already briefly described that there is a 15 backbone and off the edge -- there are multiple backbones. 16 Off the edge of those backbones will be individual service 17 providers. They may be cable companies, local phone 18 companies. There are a variety of people who have gotten in 19 the business of providing Internet service. They will then 20 connect at some number of links into the various backbones 21 They might connect for one or might connect to more. 22 Collectively, you can think of that as the majority of the 23 Internet in terms of its capacity. 24 What happens at the edges then is that individual 25 users or consumers will buy some link into one of those 864 1 service providers. So maybe it is just they have a 56-kilobit 2 modem and they dial into the service provider periodically or 3 maybe they have released a line which is sometimes referred to 4 as the last mile, last mile link into the Internet. 5 So now we have the backbone surrounded by service 6 providers surrounded by the last mile links. That constitutes 7 then all the bandwidth that makes up the Internet. One point 8 I should make, if you go into the Internet, it is not just 9 links. There are nodes, computers, in essence, that connect 10 those links together sometimes called packet switchers, 11 sometimes called routers. So at every one of the connection 12 points, there is a router of some sort. Of course at the 13 edges, individual PCs and laptops and whatnot that the end 14 user has is connected to that last mile link. 15 Q. OK, could you tell us what file transfer speed is? 16 A. File transfer speed would be the number of bytes in the 17 file divided by how long it took to transfer the file. That 18 would give you a bit rate or byte rate. 19 Q. Are connections to the Internet or pieces of the Internet 20 rated by how fast you are capable of -- 21 A. Any one of these links, you can talk about the bandwidth 22 it could potentially deliver. Whether or not you get the same 23 transfer speed as you have individual links depends on a 24 number of factors. It certainly depends on the fastest link 25 between the sender and receiver. So I couldn't go -- the 865 1 transfer time couldn't be any faster than the slowest link 2 between the source and the destination. There are other 3 factors that come into play though. 4 Q. In practical terms, does one ever get the maximum 5 theoretical speed of any piece of the Internet? 6 MR. COOPER: Your Honor, it is ambiguous and lacks 7 foundation. 8 THE COURT: Overruled. 9 A. We get it in the line. We can get conditions set up just 10 right that we can transfer at the theoretical maximum rate 11 that the link would provide. On any given day at any given 12 time, you might see something approaching that speed, might 13 see half of it, a tenth of it, it is hard to predict. It 14 depends on a number of factors, not the least of which is how 15 many other people are attempting to use the network at the 16 same time and so -- I could expand upon that, but there is 17 clearly going to be a point at which, as I have described it, 18 we have got millions and millions of end users connecting to 19 their ISPs, connecting to backbone. There is some point at 20 which there is a link that goes from Kansas City to Fort Worth 21 that is trying to carry all of that traffic and that link 22 could potentially be a bottleneck, a pinch point in the 23 network. 24 Q. Is that, what you have just described, is that what is 25 known as congestion? 866 1 A. Let me be more precise, congestion is a state where you 2 have multiple incoming flows of packets attempting to go out 3 on a shared link and the input rate exceeds the output 4 capability. So what happens is the packets get skewed as much 5 as they can be, but there is a limited amount of memory in any 6 one of those routers. As soon as that queue is full, packets 7 are dropped and they simply are not delivered. The state at 8 which you drop packets is the state of congestion. 9 Q. What is the impact of the congestion on the rates of 10 speed? 11 A. Because congestion means packets are being dropped and not 12 being delivered, if you go to the software on the in-points, 13 they are charged with the responsibility of making sure every 14 packet gets through exactly as it was presented, so as the 15 packets are coming out is exactly the same as the packets that 16 go in. If you ever detect a packet didn't get through, then 17 the sending software is responsible for retransmitting that 18 packet. But the bottom line, you will not get the same rate 19 that the theoretical capacity of the links might have 20 suggested you would. 21 THE COURT: When you say they are dropped, does that 22 mean they are never transmitted or they are delayed? 23 THE WITNESS: They are dropped. They will be delayed 24 as long as they are queued in any one of the routers, but they 25 are literally, if I have no room to hold them, I reject the 867 1 packet as it is coming in. 2 THE COURT: The question whether a packet is dropped 3 or simply delayed depends on the capacity of the router? 4 THE WITNESS: That is right. 5 THE COURT: Go ahead. 6 Q. What are some of the other pinch points that you 7 mentioned? 8 A. Well, the software that I have been talking about is TCP, 9 transmission control protocol. It is the key protocol of the 10 Internet technology. TCP is a very delicate protocol in that 11 it has to be tuned just right to get the transfer rates that 12 the underlying bandwidth might suggest. So, for example, TCP 13 will allow some number of packets to be in transit before 14 getting an acknowledgement which is a packet that says I got 15 packet 42, before getting an acknowledgement back. 16 You would not want to be in a situation where I sent 17 you one packet and waited for you to respond I got it and then 18 send a second packet and then respond because then I have only 19 one packet in flight and cross-country legacies being what 20 they are, you have very low bandwidth, you get very far from 21 the capability of underlying links. 22 So what TCP does, it will send multiple packets and 23 multiple packets in flight, ideally as many as the Internet 24 capacity can hold. As I said, it is a simple matter of 25 tuning. If your TCP has not been tuned to keep enough packets 868 1 in flight, you will get far less bandwidth than you might have 2 expected. 3 Q. What are some of the solutions if any for congestion? 4 A. Well, there is no solution for congestion aside from 5 putting in more capacity. All we can do is program nodes so 6 they detect when congestion is happening so the sources stop 7 sending so fast because if they continue to send fast, the 8 network will eventually collapse, which is get no useful work 9 done, I am spending all my time sending packets that will 10 eventually be dropped so nothing gets through. So the 11 solution is what TCP does today which is slow down whenever it 12 detects that congestion is happening on the network. 13 Q. Have you ever heard of technology called Napster? 14 A. Yes, I have. 15 Q. Is Napster or use of Napster clogging the Internet? 16 MR. COOPER: Foundation. 17 THE COURT: Sustained. 18 Q. Are you aware of any impact on the Internet of Napster? 19 A. I am not aware of any studies that say Napster is 20 accounting for "X" percentage of the packets being exchanged. 21 It has not come up in any circles I have been in where people 22 point to Napster as detectable in the traffic. 23 Q. Are you aware of any other file-sharing technology similar 24 to Napster? 25 A. Gnutella is the only one I can think of. 869 1 Q. Are you aware of any impact on the Internet of Gnutella? 2 A. No. 3 Q. Is there, at Princeton University, is there what is 4 reported in the press as a Napster problem? Do you understand 5 what I mean by that? 6 A. Princeton has not taken any action to limit the use of 7 Napster is all that I know. 8 Q. What is the network topology? 9 A. Well, roughly speaking it is like this, and I'm not the 10 Princeton system administrator so I am sure there are details 11 that I don't know. Each dorm at Princeton currently has a 10- 12 megabit shared network for all the students within that dorm 13 that would then be connected by 100-megabit ethernet into the 14 central facilities of the university. So it is one of a -- 15 like a tree, like the central facility, the core, 100 megabits 16 going down the dorms and various departments as well and in 17 the dorms, you would have 10 megabits shared. Within 18 individual apartments, you have richer connectivity because 19 there are other things they are trying to do. 20 Q. Are you aware of use of Napster in uploading, downloading 21 MP3 files at Princeton? 22 A. Not that I am aware of, no. 23 Q. Could the Internet sustain file transfers of a size of 650 24 megabytes in any kind of significant volume? 25 MR. COOPER: Object to the form, and foundation. 870 1 THE COURT: Yes, sustained. 2 Q. Are you aware of the size of a Napster file? 3 A. Napster files are in the neighborhood of 3 or 4 megabytes. 4 Q. Are you aware of a general volume of Napster files that 5 are being transferred over the Internet? 6 A. I'm not aware of the volume of transfers, no. 7 Q. Could the Internet as it is now configured handle a 8 significant number of transfers of 650 megabit files? 9 MR. COOPER: Object to the form and foundation. 10 THE COURT: Sustained as to both. 11 Q. Based on your experience and your expertise with 12 networking the Internet, is it your opinion that the Internet 13 as it is presently configured could handle thousands of 650 14 megabyte files transferred? 15 MR. COOPER: Same objection. 16 THE COURT: Sustained. 17 Q. I would like you to assume the following scenario, that 18 tens of thousands of 650-megabit files are being transferred 19 each day and could you tell us what your opinion -- 20 A. A minute ago you asked me if I knew what the volume of 21 Napster was and I don't know in any detail what the volume of 22 Napster is. But I can say, I can make a comparison with 650 23 megabyte video files and say for whatever fault that volume 24 is, I can talk about how that would strain the Internet. 25 Q. Please do so. 871 1 A. It doesn't matter what that volume is. If you take 2 Napster to be a problem and 10,000, 100,000 or a million and 3 you suddenly just translated 3 megabytes into 650 megabytes, 4 then I could talk about that. 5 MR. COOPER: I would still like a foundation if I 6 could, your Honor. 7 THE COURT: I think you need it. 8 MR. HERNSTADT: I am asking him to assume -- he 9 stated he knows the size of a Napster file. I am asking him 10 to assume there are tens of thousands -- 11 THE COURT: Look, this is a little bit along these 12 lines, you have the Tappan Zee bridge going across the Hudson 13 River up there, and we don't know how many cars are going 14 across it, and so far we don't know how many lanes it is and 15 you are saying if instead of whatever numbers of cars are 16 going across it, we had a lot more very big trucks, would 17 there be a problem, and I think you can see that there are 18 some problems with that question. 19 MR. GARBER: Your Honor, can we take a five-minute 20 break? 21 THE COURT: Yes. 22 (Recess) 23 Q. Professor Peterson, have you ever heard of DivX? 24 A. I heard of DivX, the intake for compression tool, yes. 25 Q. Do you know what -- do you understand a part of this 872 1 lawsuit is about the threat that plaintiff has alleged will be 2 posed by movies being on DVDs being decrypted by DeCSS 3 compressed to a 150-megabyte file and then sent on the 4 Internet? 5 A. Yes, I understand that's the basic idea. 6 Q. Does that as a new technology and new threat -- in other 7 words, a year ago, DeCSS has been around since approximately 8 October of 1999. 9 THE COURT: Is there a question? 10 MR. HERNSTADT: Yes. 11 Q. Let me ask you to assume that as of July 2000, there is 12 additional traffic on the Internet consisting of these 13 650-megabyte DivX files of DVD movies? 14 A. If people started to transmit 650-megabyte DivX -- 15 MR. COOPER: Your Honor, is there a question? 16 THE COURT: There is not yet. 17 Patience, Mr. Peterson. 18 Q. My question to you, Professor Peterson, based on your 19 experience, do you have an opinion as to what the effect on 20 the Internet would be posed by a new flow of 650-megabyte 21 files being added to it? 22 MR. COOPER: Objection to the form. It is an 23 incomplete hypothetical. 24 THE COURT: It sure is. I am going to hear it sooner 25 or later, so I may as well hear it now and then I will ask the 873 1 next question and then we will be past this I think. 2 Go ahead, answer. 3 A. I believe the question being asked is if the transfer of 4 650-megabyte files in the Internet -- I'm not sure if that's 5 what he asked -- I am sorry. I am having a little trouble 6 knowing exactly what it is I should answer. 7 THE COURT: I think I know where Mr. Hernstadt is 8 going and I will help him get there. 9 The net has a finite capacity today, right? 10 THE WITNESS: Correct. 11 THE COURT: It has whatever level of congestion or 12 lack of congestion that it has, right? 13 THE WITNESS: Right. 14 THE COURT: If there is a big new load placed on it, 15 depending on how big the load is and what the current capacity 16 is, it may or may not put a strain on the capacity of the 17 system, true? 18 THE WITNESS: True. 19 THE COURT: When networks such as the Internet in the 20 past have encountered limitations due to the volume of traffic 21 pressing their capacity, the tendency has been to expand the 22 capacity, right? 23 THE WITNESS: Right. 24 THE COURT: Kind of like the interstate highway 25 system? 874 1 THE WITNESS: That's an accurate analogy. 2 THE COURT: All right, let's go, Mr. Hernstadt. 3 MR. HERNSTADT: Thank you. Actually, I thought you 4 were doing very well, your Honor. 5 Q. Professor Peterson, in such an instance, how much of an 6 expansion would there have to be? 7 A. Well -- 8 MR. COOPER: Your Honor. 9 THE COURT: Obviously the answer is it depends. How 10 many messages, what the capacity is, whether there is 11 overcapacity or undercapacity, unless you have got some 12 specifics here, you are nowhere, and even if you have the 13 specifics, you have to address the expansion question. 14 Q. If we are talking about the addition of 650-megabyte files 15 being traded or being sent via the Internet by let's say -- 16 and let us assume for the purposes of this at an additional 17 rate of 10,000, 10,000 650-megabyte files per day, will that 18 pose a strain on the Internet? 19 A. I can't say whether 10,000 transfers exactly would pose a 20 strain. What I can say, if you look at the Internet today as 21 to what is a common-sized file that's transferred and make the 22 assumption that the Internet is engineered for what is 23 happening today, that you see files in the neighborhood of 24 like the -- like the music example, MP3, 3, 4, 5, 6, 7 25 megabytes, that's a typical transfer file size at the high 875 1 end. There are certainly smaller transfers, but that is 2 considered to be a sizable transfer in today's Internet, 650 3 megabytes is two orders of magnitude or a hundred times larger 4 than that. 5 (Continued on next page) 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 876 1 BY MR. HERNSTADT: 2 Q. And based upon your experience, do you have an opinion as 3 to how long it would take for the backbone of the Internet to 4 be increased by two orders of magnitude? 5 MR. COOPER: No foundation for that question, your 6 Honor. 7 THE COURT: Sustained. 8 Q. Professor Peterson, do you have any experience with the 9 growth of the backbone of the Internet? 10 A. Yes, I do. 11 Q. Do you know how fast the Internet, the backbone of the 12 Internet, has developed, how much speed on the Internet has 13 increased over the last ten year? 14 A. Yes, I could walk through that if you'd like. 15 Q. Please. 16 MR. COOPER: Your Honor, I would like to have some 17 foundation for the information the professor is about to give. 18 THE COURT: I'm satisfied. Overruled. 19 A. Sir, there are different ways of measuring exactly what we 20 are talking about here, but let me give you a couple of 21 example ways of looking at this. 22 If you start in 1990, approximately ten years ago, 23 the Internet backbone at that time run by the National Science 24 Foundation consisted of 1.5 megabits per second links. In 25 1993 that was upgraded to 45 megabit per second links. 877 1 THE COURT: What was the second one? 2 THE WITNESS: 45 megabits. If you go to 1995, we are 3 now seeing some new technology introduced, we start to see the 4 commercial carriers become bigger and bigger players. A 5 cutting age network that was a follow onto the Nation Science 6 Foundation's network was running at 155 megabits per second. 7 That was the cutting edge technology in 1995 for the long haul 8 lengths that make up the Internet. 9 If you keep following that history, and I don't have 10 an exact date here, but it's in the '98 neighborhood, I don't 11 know exactly when this particular turnover took place, we 12 started to go see the OC12 that we were talking about a little 13 earlier in the backbone, 622 megabits per second. 14 Today you can find OC48 links, they are at 2.4 15 gigabits. So if you look at the peak technology available 16 over a ten year period, you will see that it's increased two 17 orders of magnitude over that ten year period, if I've got my 18 numbers right. 19 If you focus just on the last five years, which I 20 think is probably a little bit more appropriate because that's 21 the technology that's carrying us today, OC3, OC12, OC48, 22 we've gone from bandwidths measured in the hundreds of 23 megabits per second to gigabits per second. That's one order 24 of magnitude improvement in the backbone's capacity in a five 25 year period. It could be a little bit more precise, but for 878 1 now I'll stay with orders of magnitude. 2 We are looking at if you take the assumption that you 3 need to improve the capacity of Internet by two orders of 4 magnitude, so that 650 megabyte file transfers, because as 5 common place as today, 3 megabyte or 4 or 5 megabyte 6 transfers, that's going to take two orders of magnitude 7 improvement in capacity. As I just said, it was a five year 8 period to see the last order of magnitude improvement. 9 Q. Could you define what is an order of magnitude? 10 A. Factor of ten. Two orders of magnitude is a factor of 11 100. 12 Q. What is the connection between the increased capacity of 13 the backbone and the increased connection speed for the 14 average consumer, connection to the home? 15 MR. COOPER: Can we have a foundation, your Honor? 16 THE COURT: I don't even understand the question. 17 Try again. 18 Q. Professor Peterson, is there a connection between how fast 19 a connection is available to the home and how much capacity is 20 available on the backbone? 21 A. Yes, because more consumers in the home start sending in 22 and receiving data, that data is traveling over the shared 23 capacity of the backbone, the backbone capacity has to keep 24 pace as you add more users or you add larger files that those 25 users are transferring. 879 1 Q. Professor Peterson, in your opinion, what would be the 2 effect on the effective transfer schemes of files if the 3 average file size being transferred increased but the Internet 4 backbone did not? 5 MR. COOPER: I object to the form of the question, 6 because it's an incomplete hypothetical, and I think it still 7 lacks foundation. I think it's the same question. 8 THE COURT: Look, I am going to take it, but let me 9 say this: Obviously I have a lot of respect for Professor 10 Peterson's expertise, but this just has so many limitations 11 that it's close to being of no value, because there are so 12 many factors here that you are not controlling for and so many 13 things that we don't know about, that it's just not of much 14 help. But go ahead. 15 Q. Professor Peterson, what are the factors that determine 16 the transfer rate that a home user would obtain when sending a 17 file from his or her home to somebody else? 18 A. The factors that would come into play would be the 19 capacity of the links between the sender and receiver, and we 20 have already been talking about the backbone, but it might 21 most likely be limited by those two edge links which would be 22 the uplink speed of the sender and the download speed of the 23 receiver, because these links are typically not symmetric into 24 the home. And while I may be the only one using my end link, 25 the links that are in the middle of the network are being 880 1 shared with thousands and millions of other users, and so 2 congestion is going to happen on those links, and of course it 3 depends on the level of congestion exactly what transfer fee 4 I'm going to get. 5 Q. Is the Internet capacity today sufficient so that 6 congestion is only an occasional problem in terms of slowing 7 transfer rate down? 8 A. Congestion happens all the time on the Internet today. 9 Q. If everything else remained the same but the size of the 10 average file being sent increased, is it predictable what the 11 impact would be on the transfer speeds? 12 MR. COOPER: Same objections, your Honor. 13 THE COURT: Overruled. 14 A. There would be increased congestion. I would expect there 15 to be increased congestion. 16 Q. And what is the connection between increased congestion 17 and transfer speeds? 18 A. Congestion is probably the predominant factor after the 19 actual link speed that affect the transfer rate. 20 Q. And to get a sense of what two orders of magnitude means, 21 could you compare that to an increase, for example, in 22 microprocessor speed of two orders of magnitude? 23 A. Just as an example a lot of people today would consider 24 having a 500 megahertz PC as being pretty good. Two orders of 25 magnitude back in time, that was 5 megahertz, which has been 881 1 quite a while ago in a lot of people's memory. 2 MR. HERNSTADT: Thank you very much, Professor 3 Peterson. 4 THE COURT: Thank you. Mr. Cooper. 5 CROSS-EXAMINATION 6 BY MR. COOPER: 7 Q. Good morning, professor. 8 A. Good morning. 9 Q. I believe you testified regarding some information 10 regarding the Internet backbone and made reference to OC12. 11 Those are routers, is that correct? 12 A. OC12 is a link speed. 13 Q. Okay. And the speed is 622.08 megabytes, is that correct? 14 A. I forget if the .08 is correct, but it's 622. Something 15 megabits per second. 16 Q. And the speed at which the switchers and transfers are 17 I'll to operate now have increased significantly from the 18 OC12, have they not? 19 A. Today there are OC48 links in the backbone. 20 Q. Are there not OC96? 21 A. There are probably some OC96. I'm not aware specifically 22 but that's possible. 23 Q. Are there not OC92 that have an effective rate of 9.93 24 megabytes current any use? 25 A. I'm not aware of their use in the Internet backbone. 882 1 Q. Are you aware of their use in connection with work being 2 done at the high speed connectivity consortium? 3 A. I'm not specifically aware of that consortium. 4 Q. Have you not heard of the consortium, a consortium that 5 involves the Carnegie Mellon University, Cisco Systems Inc., 6 the Corporation for National Research Initiatives among 7 others? 8 A. There are lots of consortiums like that. I'm not familiar 9 with that particular one. 10 Q. Your writing partner in your book is Mr. David, is that 11 correct? 12 A. Um-hum. 13 Q. He is a fellow with Cisco Systems, is he not? 14 A. That's correct. 15 Q. The only experiment you conducted in connection with your 16 testimony today, as I understand it, was to check the 17 effective download speed of your home DSL, is that correct? 18 A. That's correct. 19 Q. And the effective download speed that you experienced was 20 2 meg, correct? 21 A. 2 megabits per second, yes. 22 MR. COOPER: I have no further questions, your Honor. 23 THE COURT: Mr. Hernstadt. 24 REDIRECT EXAMINATION 25 BY MR. HERNSTADT: 883 1 Q. One question. Professor Peterson, what is your Internet 2 connection from your home? 3 A. It's a DSL. 4 Q. Where does it go? 5 A. It goes directly into the department, so I have a leased 6 line directly into the department, so the department is in 7 essence my ISP. 8 Q. What is the connection of your department? 9 A. The department into the Internet? 10 Q. Yes. 11 A. We are connected by 100 megabit into the campus. The 12 campus is then connected to the Internet at I believe 50 13 megabits per second. 14 MR. HERNSTADT: Thank you very much. 15 THE COURT: Professor Peterson. Thank you very much. 16 MR. ATLAS: We are going to be calling Professor 17 Peter Ramadage. I wonder if we could have a short break so he 18 can set up a laptop computer. 19 THE COURT: Sure. We will take 15 minutes. 20 (Recess) 21 THE COURT: Mr. Atlas. 22 MR. ATLAS: Good morning, your Honor. Defense calls 23 professor Peter Ramadage. 24 PETER RAMADAGE, 25 called as a witness by the Defendants, 884 1 having been duly sworn, testified as follows: 2 DEPUTY COURT CLERK: State your name, spelling your 3 last name. 4 THE WITNESS: Peter Ramadage, R-A-M-A-D-G-E. 5 THE COURT: Proceed, Mr. Atlas, please. 6 DIRECT EXAMINATION 7 BY MR. ATLAS: 8 Q. Professor Ramadage, where are you presently employed? 9 A. Princeton University. 10 Q. What do you do at Princeton? 11 A. A professor in the department of electrical engineering at 12 Princeton University, so I am engaged in teaching both 13 graduate and undergraduate students, and also research. 14 Q. Do you teach in any specialized area? 15 A. Yes, I do teaching and research in the area of digital 16 video libraries, digital signal processing for video, search 17 techniques for digital video, video compression and 18 transcoding of digital video. 19 Q. Could you briefly go through each one of those areas in 20 which you teach and just describe them for the Court, please. 21 A. My primary focus at the moment is in digital video 22 libraries. With the ever increasing amount of digital video 23 that's available, it becomes important to have an indexing, 24 search and browsing mechanism for retrieving video content 25 once it has been archived. We are developing tools which will 885 1 go into producing systems to achieve that goal. This is a 2 relatively new area, so we are only at the very beginning of 3 the research. We are developing very elementary building 4 blocks for different types of searches, different types of 5 browsing mechanisms to aid in the formation of digital 6 libraries. 7 Q. Could you describe the area of signal processing for us, 8 please. 9 A. Okay. Signal processing generally is the area in which 10 you input a given signal, let's suppose it's a video signal, 11 for example, and then you pass those bits that you are waiting 12 through an algorithm, whose purpose is to come up with either 13 a different version of what has been input or to answer some 14 question that the user has input based on the video content 15 that you are searching or inputting into the algorithm. 16 Q. I think you also mentioned compression technology. 17 A. Yes. 18 Q. What do you do in compression technology? 19 A. Let's restrict our attention to video files. Because 20 video is a very, very intensive signal to convert to digital 21 format, it requires very many bits to convert adequately into 22 the right format. It results in a very large file. Now to 23 store those files in an efficient way, or to transmit them to 24 another person in an efficient way one normally uses a 25 compression technology. That compression technology comes in 886 1 two forms. It's a form called loss less compression, and in 2 loss less compression the file is transformed into a format 3 which is smaller and requires less space but which can be 4 decompressed and you get back the original content completely 5 without any change. So it's called loss less, because no 6 information was lost. 7 Another type of compression technology is lossy 8 compression, and lossy compression you are willing to trade 9 off accuracy of the decompressed files, so the decompressed 10 file would be an approximation to the original, but you will 11 benefit by being able to get much, much greater compression. 12 That's typically used for consumer video, video on the 13 Internet, for example, video phones, wireless multimedia, they 14 all use lossy compression. 15 Q. Are you familiar with any other types of compression 16 technology? 17 A. Those are the main two. There is also a general purpose 18 compression technology used on computer files. Typically it's 19 under the name of zip files or LZW double compression 20 techniques. That's a loss less compression technology. 21 Q. You also mentioned transcoding. If you could briefly 22 describe that for us. 23 A. Okay. There are various applications when after a video 24 has been produced and it has been stored in a compressed 25 format that one needs to actually compress it further or to 887 1 transform it inn some other way to a different format. 2 To give you an example, if I have a video stored and 3 I wanted to transmit it to someone over a wireless 4 communication channel, there simply may not be enough 5 bandwidth on the wireless communication channel to transmit 6 the video in a reasonable amount of time in its current 7 format, so you would take that current format and you would 8 transcode it or basically you would take an existing 9 compressed file, process it, and to achieve greater 10 compression or change other attributes of the video such as 11 the frame rate or the frame size to make it better suited to 12 transmit over the wireless communication channel. That would 13 be one application of transcoding. 14 Q. How long have you taught at Princeton? 15 A. 16 years. 16 Q. What degrees do you hold and from what schools? 17 A. I have two undergraduate degrees, one in physics, one in 18 electrical engineering from the University of New Castle in 19 Australia. I have a masters degree in electrical engineer 20 from the University of New Castle. I have a Ph.D. from the 21 University of Toronto in Canada. 22 Q. Have you received any professional honors? 23 A. Yes. I have been awarded a complication medal from my 24 undergraduate institution. I have been awarded a best paper 25 award by the IEEE. I have been awarded several teaching 888 1 awards both from inside Princeton University and agencies 2 outside Princeton University; NSF research initiation grants; 3 IBM young investigator awards, I think. 4 Q. What is the IEEE? 5 A. The Institute of Electrical and Electronic Engineers. 6 Q. Have you received any teaching awards separate and award 7 from what you described? 8 A. Those are the main ones. 9 Q. Have you been published? 10 A. Yes, I have published over 80 referee journals and 11 conference articles. 12 MR. ATLAS: I am going to show the witness what we 13 have marked as Defendants' Exhibit BDE. It's a copy. I will 14 let the witness describe what it is. 15 THE COURT: Any objection? 16 MR. MERVIS: No objection, your Honor. 17 THE COURT: Your name is? 18 MR. MERVIS: Michael Mervis. 19 THE COURT: BDE is received. 20 (Defendants' Exhibit BDE received in evidence) 21 Q. Can you tell us what BDE is? 22 A. This is my curriculum vitae. I would say it's the latest 23 one that I just produced about two or three weeks ago. 24 Q. Does this reflect your published articles and conference 25 papers? 889 1 A. Yes. Yes, it's got the most recent articles we will be 2 presenting this September at the IEEE conference on image 3 conferencing. Those have been added. 4 Q. Have you ever testified before as an expert? 5 A. Not in court, no. I have been employed as an expert 6 witness on other court cases, but I haven't actually 7 testified. 8 Q. Have you been deposed before as an expert? 9 A. Yes, I have. 10 Q. In which case? 11 A. There is a case I. Omega v. Cyquest concerning the patent 12 infringement on removable media disk drives. 13 Q. Were you being compensated for your time in that case? 14 A. Yes, I was. I was being paid an hourly rate of -- it 15 started at $200 an hour and it went up to $250 an hour by the 16 end of the case. 17 Q. Are you being compensated for your time in this case? 18 A. No, I am not. 19 Q. Do I take that to understand that you are offering your 20 service as an expert for no compensation? 21 A. That's correct. I believed it is important to present my 22 point of view and perhaps the point of view of other 23 researchers in this area, that it's important for academics 24 and researchers in industry to have fair access use of digital 25 content in my own area of research. That happens to be 890 1 digital video. And I think it's very important for 2 researchers in digital video today to have access to the 3 digital video that's in the marketplace for fair use, use in 4 research. And down the road when this technology is mature I 5 think it would be important for other researchers in 6 disciplines not necessarily related to technology also to have 7 access to this digital information. 8 Q. Can you describe in a little bit more detail the specific 9 areas of research you are currently engaged in now at 10 Princeton? 11 A. Okay. Sir, we have I think several projects running 12 currently. One project has to do with searching video by 13 example, so we have stored in a data base a large amount of 14 video. This is a hypothetical example. What we would like to 15 do is someone comes in and says I would like to see what 16 you've got in your data base, clips that look like this 17 example, and they bring with them an example of what they 18 would like to see, what they would like to retrieve from the 19 data base, and we have been working on algorithms for quickly 20 searching through the data base to try and match the example 21 of what is in the data base and pull those thing out of the 22 data base. 23 Now because the video in the data base is stored in 24 compressed form to save space, it is important that these 25 algorithms actually operate on the compressed video. So we 891 1 have been developing algorithms that do that. That's one 2 project. 3 Another project has to do with multicamera video. We 4 have video from two cameras taking a video of a scene 5 simultaneously, and we are creating a synthetic video as seen 6 from a virtual camera on the baseline between the two real 7 cameras. That might have application in replays of sporting 8 events, it might have application in surveillance. The Navy 9 is very interested in this because they are putting cameras on 10 the decks of aircraft carriers, so we will also be talking to 11 the Navy about these possible applications. 12 Another project is using the information which is 13 embedded in the compressed video to quickly search through 14 that based on camera motion, try to estimate how the camera 15 was moving by using the information embedded in the compressed 16 video. And once you extracted what the camera was doing, that 17 can often help you say what is happening in the video. 18 To give you a specific example, in a basketball game 19 the camera often follows the ball very closely, so by 20 determining the camera motion you can quite easily pick out 21 things like fast breaks, jump shots, lay-ups and things like 22 this. 23 Q. Any other areas of research you are currently engaged in? 24 A. We also are developing some novel compression algorithms. 25 Based on some other work we think we might have some ideas 892 1 that will lead to novel compression algorithms. These are not 2 general purpose algorithms. These are rather specialized 3 compression algorithms that might be applicable in computer 4 graphics, compressing computer graphics or compressing video 5 game video. 6 Q. What is the general purpose of your research in this area? 7 A. We are working towards providing tools for the browsing, 8 manipulation and searching and indexing of digital video. 9 That's our general objective. This is a very new discipline, 10 so we are not working on everything right now. We are working 11 taking the very first steps and working on the very elementary 12 building blocks that would go into building such a system. 13 Q. When you say this is a new discipline, you are referring 14 specifically to your research or the general research in this 15 area? 16 A. The general research in this area is relatively new, 17 really has come to the forefront in the past five years. 18 Q. Do you have any expectation of how this research will 19 develop in the future? 20 A. The expectation is that after maybe in ten years time, 21 after we have all the elementary pieces worked out, people 22 will start to put these together into commercial systems or 23 even public domain systems, and then people from the 24 humanities, people from the social sciences, as well as people 25 from technology and the sciences will be able to use these 893 1 tools as part of their research. It will become a research 2 tool for people outside of the immediate technological area 3 where they were developed. 4 Q. Now in your research, and specifically in the development 5 of the algorithms you testified about, do you use digital 6 content? 7 A. Almost exclusively everything is based on digital content. 8 Q. How do you use the digital content? 9 A. One of the advantages of having digital video is that you 10 can search through it in a very quick way. Digital video 11 enables many things which are not possible with analog video. 12 You can search through the video, you can jump into it in a 13 random access type of way rather than sequentially have to 14 start from the beginning. You can store it very conveniently, 15 and you can create data bases and libraries of it very 16 conveniently. The idea is you can create a library of video 17 content, both video, sound as well as text. 18 Q. What are the sources of digital video content that you use 19 in your research? 20 A. We have a variety of sources, and I also should tell you 21 that at different stages of the research we use different 22 types of video, so we have our own camera, it's a digital 23 camera and it stores its video on a high quality digital tape. 24 From there we can transcode it into different formats, 25 whatever format we find most convenient. We find MPEG1 the 894 1 most convenient to work with, so we often transcode into MPEG 2 1 format. 3 Q. Explain what MPEG1 is. 4 A. MPEG 1 -- there are various standard compression 5 technologies. One of the very first is called MPEG1. MPEG 6 stands for Motion Picture Expert Group. It was a consortium 7 of people interested, industries interested in digital video, 8 as well as representatives of the standards organizations, and 9 they produced over a period of three to four years an ISO 10 standard called MPEG1. 11 We also download pieces of video that other 12 researchers have, and they make small clips of video available 13 on their web pages, either displaying the result of their 14 algorithms, or often in conjunction with that they will put up 15 the original video, unprocessed video, to allow other people 16 to try to replicated their results on videos. Those are very 17 short videos, less than ten seconds. 18 Occasionally my students are able to download some 19 digital video from the web that a movie studio has put up as a 20 publicity piece for a movie or something like this. So we 21 also have something like that. 22 In addition, on two occasions we have been able to 23 negotiate through an industrial partner permission to use an 24 extensive piece of video, copyrighted video from the content 25 producer. That is subject to much more restricted use. 895 1 Q. Those are the four primary sources of digital content you 2 have available to you now? 3 A. There is another source, and that is we can take analog 4 tapes and convert them into digital form. Several years ago 5 that was how we relied for getting digital video, but it 6 wasn't totally satisfactory. It's subject to the noise of the 7 analog recording process, then the noise of the analog 8 playback process, and then the peculiarity of the particular 9 digitizer that you use and compression hardware that you use 10 to actually do the digitization and compression. So we prefer 11 not to have to rely on that too much. 12 As I said, algorithms rely on processing the 13 compressed domain video, and we don't want to get tied to a 14 particular compressor, a particular piece of hardware that 15 does the compression, because then maybe our algorithms will 16 only work with that particular piece of hardware, so it's 17 important for us not to get too dependent on that mechanism. 18 Q. In terms of the variety of digital content that you have 19 just gone through, the five categories, do you find the 20 variety available in those categories best suits the type of 21 work you're doing? 22 A. There are problems with each of those video sources that I 23 have mentioned. Let me elaborate on that a little bit. 24 First we have our own camera, but it's quite a steep 25 learning curve to learn how to use the digital camera, so our 896 1 graduate students spend some time learning it if they need to 2 produce video, but the resultant video is not particularly 3 high quality. They are not professional cameramen, and they 4 are faced with various hurdles that they have to come up with, 5 and they are quite innovative in trying to overcome those 6 hurdles, but the resulting video is not commercial quality 7 video. For example, we can't hold the camera very steady. 8 It's a hand-held camera. We can't take multicamera video. We 9 only have one camera. And we can't control the lighting very 10 well. 11 The video that we download from the web from other 12 researchers is typically very, very short and sometimes has a 13 lower frame rate than we would like to work with, and 14 sometimes a very small frame size, small number of pixels in 15 it. 16 Video that we digitize from analog tape I think I 17 said already there is a couple of problems with that. I won't 18 go over those again. 19 By far the best quality digital video that we can 20 obtain is from DVDs. 21 Q. Have you heard of DeCSS? 22 A. Yes, I have. 23 Q. Have you used DeCSS? 24 A. Yes, I have. I have used it. I went to the -- 25 Q. Wait. What do you understand DeCSS to be? 897 1 A. It's a program which will read the contents, the table of 2 contents of a DVD disk, and then you can ask it to descramble 3 the scrambled VOB files, the video object files on that disk, 4 and store them to your hard drive. 5 Q. You testified a moment ago that you used DeCSS. How did 6 you obtain DeCSS? 7 A. I went to a search engine, Google.com and I just typed in 8 DeCSS and did a search. It came up with about over 7,000 9 hits, so I did a bit of a search through those hits until I 10 found a site that had the software, and I downloaded the 11 source and executable program. 12 Q. Do you recall the site you downloaded it from? 13 A. No, because it involved a search through all of the 14 various hits I don't actually remember the site I eventually 15 found it on. Sometimes when you go to these sites you don't 16 actually end up on the front page of the site, you end up on 17 some lower page that has the feel you want, but it's not 18 immediately clear what site this is ultimately connected to. 19 Q. Do you know whether it's 2600.com? 20 A. I don't believe it was 2600.com, no. 21 Q. What form was DeCSS in when you downloaded it? 22 A. It was a zip file which is a loss less compression, that's 23 a general type of loss less compression used to compress 24 general computer files. When I decompressed that, I believe 25 it gave me the source, the executable and maybe another 898 1 document, I can't remember, a "read me" file or something like 2 that. 3 Q. Do you have an opinion on whether DeCSS would be useful to 4 you in your area of research? 5 MR. MERVIS: Objection, your Honor. There is no 6 foundation. 7 THE COURT: Overruled. 8 A. My initial experiments have indicated that's by far the 9 best source of high quality digital video available to us 10 today. 11 Q. What is the best? 12 A. The DVD. In terms of the breadth of video that's 13 available to us, the wide variety of video available to us 14 from various sources, and the high quality of the content, 15 DVDs are the best source for us. 16 Q. What I'm asking you is do you have an opinion on whether 17 the DeCSS utility would be useful to you in your area of 18 research? 19 A. Oh, absolutely, because the video content on DVDs is 20 scrambled, and so to get access to it we have to unscramble it 21 and that's exactly what DeCSS does, it unscrambles that video 22 content. 23 Q. If you were to have access to a wide variety of high 24 quality digital content like on DVDs, why would that be 25 helpful to your research? 899 1 A. Okay. Initially when we first start developing an 2 algorithm, we usually use very short pieces of video, because 3 videos are a very time consuming object to work with and it 4 takes a lot of space. But after we've got the prototype 5 working, it's very important to test the algorithm on a wide 6 variety of different video. It's important for two reasons. 7 First, you want to make sure your algorithm isn't somehow 8 dependent on the type of video you took or the type of encoder 9 you used. That's step number one. 10 Step two is you would like to search for video for 11 which it doesn't work, and to do that you need to get out 12 there and search through a whole range of different types of 13 video. You are specifically looking for video where your 14 algorithm fails to help you improve your algorithm or maybe 15 start up a different research direction. 16 Q. In order to use the video, the high quality video that's 17 available on DVDs, do you have an understanding of whether you 18 need to decrypt them first? 19 A. Yes. There are several types of files in the DVDs. There 20 are files with the extension IFO, which I believe is an 21 abbreviation for information, and those contain like a table 22 of content type information about what is on the DVD. There 23 are files with the extension BUP, which I believe stands for 24 back up. Those I think are back-ups for the IFO files. Then 25 there are files with the extension VOB, which I believe is an 900 1 abbreviation for video object. And those actually contain the 2 video and those are scrambled, and those are the things that 3 need to be unscrambled before you can actually use the video. 4 Now I should also add that that video is already in a 5 compressed form. It's in a compressed form called MPEG2, 6 which was the second extension of the ISO standard from MPEG1 7 to MPEG2, and that dealt with a higher quality video. The 8 initial base standard for MPEG2 was intended to produce 9 compressed video of comparable quality to broadcast video. 10 But it also includes higher level video standards as well. 11 Q. In terms of the digital video content that's commercially 12 available on DVDs, is that preferable to the digital content 13 that you described before, the digital content that's 14 currently available here? 15 A. Yes, it's very high quality, clean, no noise. It is 16 already digitized, which is excellent, which means we avoid 17 having to digitize it ourselves and then pick out any 18 peculiarities our own digitizer exhibits. So, it's a very 19 sound and preferable source of digital content. Also because 20 of the wide variety of DVDs available, it satisfies our need 21 for a large source of different varieties of video. 22 And in addition, I want to add one other thing. When 23 we use our own camera we try to avoid introducing any bias. 24 We don't want to sort of take video which is too favorable to 25 what we are trying to do, because the scientific method 901 1 demands that if you want to thoroughly test your algorithm you 2 have to use the video which has been taken independently by 3 somebody else. That's a basic fundamental premise of the 4 scientific method. 5 Q. Is there a relationship between the amount or variety of 6 digital video that you would use in connection with your 7 research and the confidence you would have in the results of 8 your research? 9 A. Absolutely. One of the criticisms -- I have been to 10 conferences. I have also given talks in industry, and one of 11 the criticisms that industry often has mentioned is that you 12 really need to test your algorithms on a wider variety of 13 video. I often hear that from people in the industry. 14 Q. Are there any other applications of your research that 15 would benefit from having access to the high quality digital 16 content of DVDs? 17 A. I think I have covered the main points. 18 Q. In terms of compression technology, do you use compression 19 technology in your research? 20 A. Yes, as I have said, our algorithms are designed to work 21 with compressed files, and because of that we need to have a 22 working knowledge of the compression technologies that are 23 employed. We also do our own transcoding of video from 24 different compression standards to a second compression 25 standard in order to test our algorithms on a variety of 902 1 compression standards. 2 Q. I believe earlier you testified that there was lossy 3 compression and loss less compression? 4 A. Yes. 5 Q. Do you know what type of compression is used on 6 consumer-oriented video like the DVDs you buy in a store? 7 A. Yes, both the MPEG1 standard and the MPEG2 standard are 8 lossy compression technologies. That means when the video is 9 compressed, information is thrown away. If it's a good 10 encoder, a good compressing algorithm, it will first try to 11 throw away information which is least perceptually 12 significant, but the more compression you ask for, the more 13 bits it has to discard or the more information it has to 14 discard, and eventually it is discarding important 15 information, and that shows up as artifacts in the resultant 16 video when you uncompress it. 17 Q. As a general matter, why would you use lossy compression 18 if you can use loss less compression? 19 A. Okay. Loss less compression that can achieve compression 20 factors of 2 is quite typical. If you are very lucky you 21 might get higher than 2. So, a 2 gigabyte file would be 22 reduced to 1 gigabyte with a compression factor of 2. By 23 applying loss compression, you can -- I'm quoting here from 24 the MPEG standard -- you can achieve compression factors of 25 around 20 to 30 without any visible perceptual difference 903 1 between the uncompressed video and the compressed video. So, 2 that's a factor of ten better at least than the loss less 3 compression technology. And in practice you often find 4 compression factors as high as 40 used. 5 Q. Now, could you just describe for us a little bit in 6 greater detail as you use more and more compression on content 7 what happens to that content. 8 A. Okay. Almost all of these compression technologies are 9 block based. What I mean by that is that each frame of the 10 video is divided into small blocks, nonoverlapping blocks in 11 the simplest case. These blocks are about 16 by 16, 16 pixels 12 horizontally by 16 pixels vertically. 13 The compression is based on those elementary blocks, 14 so when you start to throw away too many pieces of 15 information, those blocks don't get represented correctly in 16 the uncompressed video, so you start to actually visibly see 17 these blocks in the uncompressed video, and you start to see 18 miscoloring of the blocks. 19 One of the things that underlies most of the 20 compression technologies is that high frequency information is 21 discarded first. Now, high frequency information encodes 22 things like edges, where there are sharp transitions, sudden 23 changes in the image, so when you have discarded enough of 24 this high frequency information or too much of this high 25 frequency information, you start to see artifacts at the edge 904 1 boundaries, a phenomenon called ringing, where there seems to 2 be a time varying fluctuation around the edge, and you can 3 visibly see this in highly compressed video. 4 Q. Is that what you referred to before as an artifact? 5 A. That's an artifact. The fact that you can see the blocks 6 is an artifact. The fact that the blocks are miscolored is an 7 artifact, and this ringing is an artifact. There are other 8 artifacts as well, but those are the main ones. 9 Q. Do you have an understanding of an average size of a DVD 10 film? 11 A. Well, I have looked at the DVD for the movie Contact, and 12 the total amount of information on that DVD disk was about 13 seven and a half gigabytes. Most DVDs contain the original 14 movie as well as some additional add-ons such as director's 15 comments, or things that were cut from the movie might be 16 added on there, or special effects might be added on. 17 In the particular case of the DVD Contact, the actual 18 video file and the audio file took up about 6 gigabytes, and 19 the remaining gigabyte and a half were the extras and the 20 add-ons. 21 Q. Do you have an understanding of how much an average CDR 22 can hold in terms of megabytes or gigabytes? 23 A. The standard is 650 megabytes. 24 Q. So, if I wanted to copy a DVD that I went out and 25 purchased and copied it onto a CD, I would have to compress 905 1 the content to go from 6 or 7 gigabytes down to 650 megabytes? 2 A. That's correct. 3 Q. How would that be done? 4 A. Well, you would use a transcoder. Your video that you 5 have from the DVD after you've descrambled it, you must 6 descramble it first or else it won't work at all. Let's say 7 you descrambled the VOB files, now they are in MPEG2 format. 8 That has a bit rate -- and I will explain what bit rate is in 9 a second -- it has a bit rate of about 6 to 10 megabits per 10 second. Now the bit rate is how many bits are coming out of 11 the player per second in order to display the video on the 12 screen on average. That's an average rate. Sometimes it's 13 higher, sometimes it's lower. But on average it's about 6.7 14 megabytes per second. You have to now transcode that down to 15 a much, much lower rate. 16 You have available on the CD a space of 650 17 megabytes. Let's take the movie Contact. That's a two and a 18 half hour movie, so I have to get two and a half hours at 6.7 19 megabits per second down into a 650 megabyte disk. If you do 20 the conversion of units to make all the units appropriate, and 21 then work out the math, the bit rate after you have done the 22 transcoding for both the audio and the video needs to be 23 around 590 kilobits per second. So you need to go from 6.7 24 megabits per second down to 490 kilobits per second. 25 That's an enormous compression. 906 1 Q. Would that type of compression result in artifacts being 2 present on the ultimate product you end up on the CDR? 3 A. Yes, most definitely. The artifacts would be most visible 4 when there are scene changes, camera motion or when there is 5 movement of objects in the scene. 6 Just as a point of comparison, MPEG1, the rate for 7 MPEG1 which is generally to be believed below broadcast 8 quality video is 1.5 megabits per second. 9 Q. Based on your experience, if someone were going to 10 compress a film file to send over the Internet, is it more 11 likely that such a person would use lossy or loss less 12 compression? 13 A. If they used loss less compression you would take a 6 14 gigabyte file down to about 3 gigabytes. That is way too big 15 to transmit over the Internet. So that forces you, you must 16 use lossy compression if you are going to get it down to any 17