See related files:
http://www.eff.org/IP/Video (EFF Archive)
http://cryptome.org/cryptout.htm#DVD-DeCSS
(Cryptome Archive)
http://www.2600.com/dvd/docs (2600 Archive)
http://eon.law.harvard.edu/openlaw/dvd/ (Harvard DVD OpenLaw Project)
1
1 UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF NEW YORK
2 00 Civ. 0277 (LAK)
------------------------------------
3 UNIVERSAL CITY STUDIOS, INC., :
PARAMOUNT PICTURES CORPORATION,
4 METRO-GOLDWYN-MAYER STUDIOS INC., :
TRISTAR PICTURES, INC., COLUMBIA
5 PICTURES INDUSTRIES, INC., TIME :
WARNER ENTERTAINMENT CO., L.P.,
6 DISNEY ENTERPRISES, INC., and :
TWENTIETH CENTURY FOX FILM
7 CORPORATION, :
8 Plaintiffs, :
DEPOSITION OF:
9 vs. LARRY PETERSON
:
10 SHAWN C. REIMERDES, ERIC CORLEY,
a/k/a "EMMANUEL GOLDSTEIN" and :
11 ROMAN KAZAN and 2600 ENTERPRISES,
INC., :
12
Defendants. :
13 ------------------------------------
14
TRANSCRIPT of the stenographic notes of
15 the videotape proceedings in the above-entitled
matter, as taken by and before JOMANNA CASTANO, a
16 Certified Shorthand Reporter and Notary Public,
held at the offices of PROSKAUER ROSE, LLP
17 1585 Broadway, New York, New York, on Monday,
July 10, 2000, commencing at 10:30 in the
18 morning.
19
20
21
22
23 NEW YORK REPORTING COMPANY (USA), LTD.
245 Park Avenue
24 39th Floor
New York, New York 10167
25 (212) 792-5623 Fax: (212) 792-5624
2
1
A P P E A R A N C E S:
2
PROSKAUER ROSE, LLP
3 BY: WILLIAM M. HART, ESQ.
1585 Broadway
4 New York, New York 10036-8299
(212) 969-3095
5 Attorneys for Plaintiffs
6 FRANKFURT, GARBUS, KLEIN & SELZ, P.C.
BY: MARTIN GARBUS, ESQ.
7 488 Madison Avenue
New York, New York 10022
8 (212) 826-5582
Attorneys for Defendant Eric Corley
9
ALSO PRESENT: Eileen Dougherty, Videographer
10
11
12
13
14
15
16
17
18
19
20
21
22
23 NEW YORK REPORTING COMPANY (USA), LTD.
245 Park Avenue
24 39th Floor
New York, New York 10167
25 (212) 792-5623 Fax: (212) 792-5624
3
1 I N D E X
2 E X A M I N A T I O N S
3 WITNESS BY PAGE
4 LARRY PETERSON MR. HART 4
5
6
7 E X H I B I T S
8 NUMBER DESCRIPTION PAGE
9 1 A Document 4
2 Declaration 4
10
3 Table of contents 20
11
12 (Exhibits were retained by counsel)
13
14
15
16
17
18
19
20
21
22
23
24
25
4
1
2 (Whereupon, Exhibits Nos. 1 and 2 are
3 marked for identification.)
4 THE VIDEOGRAPHER: Eileen Dougherty,
5 member of the National Legal Video Association for
6 New York Reporting. We're on the record at 10:29
7 a.m. on July 10th, 2000. We're here for the case
8 Universal versus Reimerdes. The witness today is
9 Larry Peterson. We're at the location of 1835
10 Broadway, New York, New York -- 1585 Broadway, New
11 York, New York.
12 Will the attorneys please state their
13 appearances for the record.
14 MR. HART: I'm Bill Hart from Proskauer
15 Rose for the plaintiffs.
16 MR. GARBUS: Martin Garbus for the
17 defendants.
18 THE VIDEOGRAPHER: And will the court
19 reporter please administer the oath.
20
21
22
23
24 DIRECT EXAMINATION BY MR. HART:
25 Q. Good morning, Professor. How would you
5
1 LARRY PETERSON
2 prefer that I address you?
3 A. Professor is fine.
4 Q. Professor? Okay. Thank you.
5 Would you give us your home address for
6 the record, please.
7 CONFIDENTIAL
8
9 Q. Thank you. Are you employed, sir?
10 A. Yes.
11 Q. By whom?
12 A. Princeton University.
13 Q. How long have you been employed by
14 Princeton?
15 A. Two years.
16 Q. And what is your job at Princeton?
17 A. I'm a full professor. I teach classes
18 and advise students, conduct research.
19 Q. Do you also consult?
20 A. Some, yes.
21 Q. Have you ever testified in any court
22 proceedings before?
23 A. No.
24 Q. Have you ever been deposed in
25 connection with any court proceedings before?
6
1 LARRY PETERSON
2 A. No.
3 Q. Do you have a current copy of your
4 curriculum vitae?
5 A. I sent one by e-mail at one point.
6 MR. GARBUS: Yes. Did we not forward
7 it to you? I understood we had as part of the
8 affidavit.
9 Q. Okay. Let me mark Exhibits 1 and 2 and
10 ask you if you can tell me whether your curriculum
11 vitae is among either of the exhibits I've just
12 handed you.
13 A. No, I don't see it.
14 MR. HART: Could we ask you to contact
15 your office and zip that over to us?
16 MR. GARBUS: Sure.
17 Q. I don't want to take a long time
18 getting your track record, but where were you
19 before Princeton?
20 A. I was at the University of Arizona for
21 13 years on the faculty.
22 Q. On the faculty?
23 A. Yes.
24 Q. Are there any areas of specialization
25 that you have as a professor?
7
1 LARRY PETERSON
2 A. Broadly, computer systems, computer
3 software. And more narrowly, computer networks
4 and operating systems.
5 Q. And how long have you been focused, as
6 you say, "more narrowly" on computer systems and
7 networks?
8 A. Since my graduate student days. I was
9 at the -- a student at Purdue University. I
10 started working on networking-related things back
11 in 1980. I've been basically on that area since.
12 Q. Just so we start out on the same page,
13 can you tell me what you mean when you say
14 "computer network" or "network-related
15 activities."
16 A. Primarily, I have focused on the
17 Internet since the very early days of the
18 Internet. So, it's not the telephone system, it's
19 not the cable system in particular, but it's the
20 Internet as it's evolved.
21 Q. Okay. And is there any special meaning
22 to the word "network" in the context of the
23 Internet?
24 A. I'm not sure I understand the question.
25 The Internet is a network.
8
1 LARRY PETERSON
2 Q. Is there more than one kind of network
3 in computer science?
4 A. Yes, you could say that there are
5 different specialized networks. The telephone
6 network is a network. It's evolved over the
7 years. It's much different than it used to be.
8 Q. Why is that?
9 A. As they have attempted to accommodate
10 all the data that's being transmitted and the
11 Internet, they have adapted to that.
12 Q. How have they adapted to that?
13 A. Trying to bring the Internet technology
14 into the way they deliver -- trying to integrate
15 voice and data, essentially.
16 Q. How is that done technically, without
17 getting too technical?
18 A. You would like a single set of physical
19 resources linked, transmission cables and switches
20 and routers, to be able to deliver both the data
21 that computers want to exchange and the voice that
22 typical users at home want to use.
23 Q. Do improvements in telephone network
24 systems include such things as digitized signals,
25 compression, and high-speed switching?
9
1 LARRY PETERSON
2 A. Those are all elements of it, yes.
3 Q. Are there any other elements?
4 A. What was the original question again?
5 Q. We were talking about the improvements,
6 if you will, to accommodate more data and talking
7 about some of the technical things that facilitate
8 that. And my last question was whether
9 digitization of content, whether compression
10 technology and whether high-speed switching
11 improved the data transmission, and you said, yes,
12 and you said those were some of the factors.
13 And my question now -- I apologize for
14 being so laborious, but my question now is what
15 are the other factors or considerations?
16 A. So, there is both the technology
17 required to deliver content, which includes the
18 rate at which I can transmit data and it includes
19 the switches. Those are called various things
20 depending on technical circumstances, hubs,
21 switches, routers, there's nuances there.
22 And then there's the content, whatever
23 it is you want to deliver, and some technologies
24 have made -- so you're trying to match the
25 content, in essence, with the capabilities of the
10
1 LARRY PETERSON
2 delivery system. Some technologies, like
3 compression, help bring the content closer to what
4 the technology is able to deliver.
5 So, you asked for other things.
6 Q. Yes.
7 A. The other thing is the control of all
8 of that technology. And the hard problem there is
9 that the transmission facilities and the switches
10 are all shared resources, many different people's
11 data is being transmitted over them. And the
12 management of who gets to transmit their data
13 when, who gets to achieve this bandwidth and who
14 gets to achieve that bandwidth, is probably the
15 single greatest conditional ingredient.
16 Q. How is that management function
17 performed? Is that a technological function, is
18 that something that involves human interaction, or
19 both?
20 A. At different levels it's both technical
21 and human. At the very highest level, people
22 decide we need a new trunk between San Francisco
23 and Boston. So, the very high-level decisions
24 like that are made. But on a packet-by-packet
25 basis, there is software, in essence, that
11
1 LARRY PETERSON
2 controls whose packet gets to go next, and that,
3 in the end, influences whose content is delivered
4 and what rate.
5 Q. Just so the record is clear, when you
6 say "packet," you're referring to --
7 A. "Packet" is a fragment or a chunk of
8 whatever data you're trying to transmit.
9 Q. Is it characteristic of modern
10 telecommunications and content delivery systems
11 that data is broken up -- a particular set of data
12 to be transmitted is broken up into packets and
13 those packets are transmitted separately and at
14 different times or at different rates?
15 A. That is the chief philosophy behind the
16 Internet, and what, in some sense, distinguishes
17 it from the earlier telephone networks. It is a
18 packet switch network, so that's the key idea.
19 Q. Do you have any specialized knowledge,
20 background, or expertise in the area of video
21 codex?
22 A. Not in particular.
23 Q. Do you have any specialized knowledge
24 or expertise in the area of transmission of
25 audiovisual content over the Internet?
12
1 LARRY PETERSON
2 A. I know a little bit about MPEG-2,
3 MPEG-1 compression.
4 Q. Okay. What do you know?
5 A. And then my chief focus is on the
6 interplay between compression and the transmission
7 capabilities, so the interplay between compressing
8 a piece of video, for example, and then
9 transmitting it.
10 To your question, I know, at some level
11 of detail, how MPEG-2 works in terms of a video is
12 a sequence of frames, still images. There is a --
13 different ways that you compress each one. Some
14 frames are compressed relative to earlier frames
15 because the content didn't change much, you can
16 just give the differences.
17 Q. Right.
18 A. Other frames are called reference
19 frames, and they're compressed in a different way.
20 Q. Have you ever used any video
21 compression devices or software?
22 A. My students have implemented MPEG-2
23 decompression in software.
24 Q. Decompression.
25 A. Decompression. We have done a little
13
1 LARRY PETERSON
2 bit of work in the compression, but not very much.
3 Q. In answering that you referred to your
4 students.
5 A. My students and I, right.
6 Q. My question is have you personally,
7 Professor, had experience using video codex?
8 A. Not codex.
9 Q. I'm sorry, video compression.
10 A. Compression?
11 Q. Devices or software.
12 A. Define "personally." I've watched MPEG
13 videos. I have looked at pieces of the software.
14 I have talked with my students and we've drawn
15 diagrams on the board explaining how things
16 worked.
17 Q. Okay. Have you ever taken a data file
18 containing audiovisual content and caused it to be
19 compressed using some kind of video compression
20 software?
21 A. No, not that I can recall.
22 Q. Do you know what DivX is?
23 A. I know of two DivX's.
24 Q. The second one, sir.
25 A. My understanding is that DivX is a
14
1 LARRY PETERSON
2 program that will do MPEG-4 video compression.
3 Q. Have you ever used DivX or MPEG-4
4 compression tools?
5 A. No.
6 Q. Have you ever seen the picture that
7 results from a video file that has been compressed
8 using DivX or MPEG-4?
9 A. By "picture" you mean?
10 Q. Display, screen display. Have you ever
11 seen the resulting --
12 A. I have not seen a video. What I have
13 seen is a still image, a snapshot of a video on a
14 DivX site. I was curious and I went and looked at
15 what was there. They have some trailers that they
16 had compressed and they were still shots.
17 Q. Do you remember what site that was?
18 A. Divx.ctw.cc, I believe.
19 Q. And does that site provide DivX
20 software utilities for download?
21 A. As I recall, it does.
22 Q. Does it also provide instructions to
23 users on how to use such software?
24 A. I don't -- I didn't look that far.
25 Q. Can you tell me, to the best of your
15
1 LARRY PETERSON
2 recollection, why you looked at that site?
3 A. Because I had seen an earlier
4 deposition or declaration that referred to DivX
5 and wanted to find out a little bit about it.
6 Q. When did you first get involved in this
7 case?
8 A. It was very near the end of June. I
9 forget the exact day.
10 Q. And how did that come to occur?
11 A. Someone from the firm was visiting
12 Andrew Appel, who was going to be deposed, Ed
13 Hernstadt, and he was at Princeton. The subject
14 of networking must have come up because Andrew
15 brought him by my office. I saw him for five
16 minutes.
17 Q. Okay. And you talked with
18 Mr. Hernstadt during that five minutes?
19 A. Yes.
20 Q. And he talked to you?
21 A. Yes.
22 Q. Can you give us the gist of that
23 conversation?
24 A. He said something to the effect that he
25 was involved in a case, and the issue of network
16
1 LARRY PETERSON
2 capabilities had come up, did I know anything
3 about networking or what did I know about
4 networking.
5 Q. And what did you say?
6 A. I said I do research in that area, I
7 teach in that area, I wrote a book in that area.
8 I know a little bit about it.
9 Q. Now you recently revised your book. Is
10 that right?
11 A. That's right.
12 Q. And the revised edition was published
13 in October of '99. Is that correct?
14 A. That's correct.
15 Q. And how extensive was that revised
16 edition compared -- how extensive were the
17 revisions in the revised edition, if you will,
18 compared to the first edition?
19 A. We estimated, my coauthor and I
20 estimated that we changed about a third of it.
21 Q. Okay. And can you tell me, in general,
22 what the subject matters of the changes were?
23 A. If I had a copy in front of me, I could
24 thumb through and give you a better answer here.
25 Q. You don't have a copy of your book with
17
1 LARRY PETERSON
2 you; do you?
3 A. No, I don't.
4 MR. HART: Mr. Garbus, I had requested
5 one. I don't know if defendants --
6 MR. GARBUS: And you can buy it. The
7 only copy that I have in my office is a copy with
8 my notes on it. If you want to continue the
9 deposition to some other day, which I don't think
10 you want to do, we will certainly get it for you.
11 It's also orderable on Amazon. It's orderable on
12 several networks. I don't know what else to tell
13 you.
14 THE WITNESS: I can give you a -- so,
15 as I sit here thinking a little bit more, I can
16 say a little bit more, if it would help.
17 Q. Sure.
18 A. Mostly we did two things to the second
19 edition. One, we took out a specific piece of
20 software that caused people not to adopt the book
21 because they were afraid they had to use our
22 software to use the book. We had used a system
23 that my research group had developed over the
24 years to implement network protocols as an
25 illustrative technique, so we had used little
18
1 LARRY PETERSON
2 snippets of software throughout the book, and
3 people were under the impression that they had to
4 adopt my software to use the book, that the book
5 would be useless to them otherwise. So, we
6 changed how we did the software in the book.
7 And the second thing we did --
8 Q. Wait. Let's stop right there. What
9 software did you then refer to in the book?
10 A. Well, this was mostly a matter of
11 removing a chapter that described the software
12 system. We left all the code fragments as they
13 were except making very small changes to make it a
14 little bit more generic.
15 Q. Okay. I'm sorry to interrupt you. Go
16 ahead.
17 A. The second major change is that we
18 focused a little bit more on applications that we
19 had originally, some applications of the
20 protocols.
21 Q. If you could explain that?
22 A. Well, as I recall now, we included a
23 little bit of information about a protocol that
24 is, in fact, used to deliver video and audio over
25 the Internet.
19
1 LARRY PETERSON
2 Q. And what is that protocol?
3 A. That protocol is called RTP.
4 Q. What does RTP stand for?
5 A. RTP stands for real-time protocol.
6 Q. Is that used for streaming video
7 content over the Net?
8 A. Yes.
9 Q. Is it also used to transmit content for
10 downloading by the recipient?
11 A. If by "downloading" you mean pulling a
12 file across the network which contains a video so
13 that I can play it as opposed to streaming where
14 the video is arriving as I'm watching it, it's not
15 used for downloading.
16 Q. Got it. Now, does streaming save
17 network resources in terms of consuming bandwidth
18 time and space?
19 A. No.
20 MR. GARBUS: Object to the question.
21 MR. HART: Let me mark as Exhibit 3
22 what I'll show you in a moment, after the reporter
23 marks it.
24 (Whereupon, Exhibit No. 3 is marked for
25 identification.)
20
1 LARRY PETERSON
2 Q. I ask you if you can identify what
3 we've just marked as Exhibit 3, Professor.
4 A. This is the table of contents of my
5 book pulled off the publisher's web site. I'm
6 just trying to verify that it's the second
7 edition. And it is.
8 Q. Based on a quick review of Exhibit 3,
9 are there any other --
10 MR. GARBUS: May I have a copy?
11 MR. HART: You do. You were talking on
12 the cell phone when I gave it to you.
13 Q. Based on your review of Exhibit 3, are
14 there any other areas of subject matters of your
15 book that were changed or updated between the
16 first and second editions?
17 A. Well, there were certainly small
18 updates throughout. So, I'll just focus on the
19 bigger stuff.
20 Section 2.8 on wireless was new.
21 Actually, if you look closely, the changes are
22 there.
23 Q. How so?
24 A. The boxes. The publisher included the
25 changes so that adopters could see what had
21
1 LARRY PETERSON
2 changed.
3 Q. When you say "adopters," what do you
4 mean?
5 A. The book is primarily targeted as a
6 textbook, so universities will adopt it.
7 Q. Is this a basic text? Is it an
8 advanced text?
9 A. This is used either in upper division,
10 senior level, undergrad classes or graduate
11 classes. So it is for more advanced students.
12 Q. It is for more advanced students?
13 A. It is for advanced students, but it's
14 typically the first class they've taken in
15 networking.
16 Q. Now, a little bit earlier I had asked
17 you about different kinds of networks, and I
18 believe you had talked about telecommunications as
19 an example, you talked about the Internet as one
20 big network, I believe. I would like to get a
21 little bit more specific in terms of the different
22 kinds of networks that currently exist.
23 For example, tell me what you know
24 about the network or networks at Princeton, just
25 to use an illustration.
22
1 LARRY PETERSON
2 MR. GARBUS: Let me just object to
3 these questions. I will not stop the witness from
4 answering. Basically, what Mr. Hart is doing is
5 using this as a lecture so that he can understand
6 the situation involved in the case, that he is not
7 asking questions about the witness's testimony,
8 although I'm sure he will claim that it's
9 relevant. The witness is here for a specific
10 purpose, which is to deal with the issues raised
11 in his affidavit and not to bring Mr. Hart up to
12 date either by furnishing his books or a general
13 lecture on how the Internet works, presumable
14 Mr. Hart knows that or can get it from his own
15 witnesses.
16 MR. HART: Thank you, Mr. Garbus. I'm
17 sorry for the interruption.
18 Q. The question was: Tell me about
19 Princeton's network or networks as an illustrative
20 example.
21 A. Princeton -- a typical campus will have
22 a campus backbone which connects the buildings.
23 Each building is then typically wired with a local
24 area network or a collection of local area
25 networks. Princeton, in the computer science
23
1 LARRY PETERSON
2 department, we primarily have a 100-megabit
3 ethernet within the department into every office.
4 We have a smattering of 1-gigabit lengths; those
5 are still typically used for research and not part
6 of our production network. We are then connected
7 to the campus backbone, as all the other
8 departments are. I believe Princeton's campus is
9 still in the 100 megabit neighborhood.
10 Q. When you say "still," you're suggesting
11 that that's been superseded?
12 A. That's a little -- "still" as in that's
13 potentially upgradeable in the not too distant
14 future. I don't know any specifics, but.
15 Q. Potentially upgradeable to what?
16 A. 155 megabits, possibly a 622 megabits,
17 those are some of the next steps that are
18 sometimes taken. I actually have no specific
19 knowledge of what the backbone at Princeton is in
20 detail.
21 Q. Or its plans to grade it in the near
22 future?
23 A. Or its plans to upgrade.
24 So, generically, 100-megabit,
25 10-megabit ethernet are the two chief technologies
24
1 LARRY PETERSON
2 within a campus.
3 Q. And is that true with respect to dorm
4 rooms and other facilities within the campus where
5 students have access to the networks?
6 A. It is typically the case that the dorms
7 are a little behind what would generally be
8 available. I do happen to know that Princeton's
9 dorms are connected by 10-megabit ethernet. At
10 the current time, each dorm share one 10-megabit
11 amongst all the students in that dorm. So we're
12 talking in the neighborhood of 200 students
13 sharing a 10-megabit ethernet.
14 Q. And that's for an entire dorm building
15 or per floor?
16 A. I believe that's dorm, but I couldn't
17 say for sure.
18 Q. Apart from the declaration which we've
19 marked as Exhibit 2, have you prepared any
20 materials in connection with your involvement in
21 this case? By "materials," I mean whether it's in
22 the form of an English writing or computer
23 software or demonstrations or anything like that?
24 A. I worked through some numbers of how
25 long it would take to transmit a file if given
25
1 LARRY PETERSON
2 different technologies. That's all I've done.
3 Q. And when you say you "worked through
4 some numbers," how did you do that? On a
5 computer? On a notepad with a pen?
6 A. Basically, for my own collecting my
7 thoughts, wrote or just typed it into a file the
8 numbers, used a calculator to make sure I had the
9 numbers right, and stated what my assumptions
10 were.
11 Q. And do you have that document with you
12 here today?
13 A. No, I do not.
14 Q. Can you access it from a computer?
15 A. Yes, I should be able to.
16 Q. Can we take a break and let me ask you
17 to get on-line here and hook up to your computer
18 in your office or wherever this data is resident
19 and get me a copy of your notes?
20 A. That's fine.
21 MR. GARBUS: Well, I would get a copy
22 of the notes, if you can, and I will see it first
23 before he sees it, and then I will see whether or
24 not there are any objections with respect to it.
25 You and I will discuss it.
26
1 LARRY PETERSON
2 THE VIDEOGRAPHER: Off the record at
3 10:53.
4 (Whereupon, a brief recess was taken.)
5 THE VIDEOGRAPHER: Back on the record
6 at 11:00.
7 MR. HART: We took a quick break to see
8 if Professor Peterson could locate his notes.
9 There were some technical difficulties which my
10 information specialist is going to try and help
11 on, and we're going to try and proceed now. We
12 may get interrupted.
13 MR. GARBUS: And what happens then?
14 Professor Peterson goes back to the machine and
15 tries to get the document?
16 MR. HART: Well, Marty, I haven't
17 figured out the mechanics of it.
18 MR. GARBUS: In other words, we see the
19 document before you see the document.
20 MR. HART: Marty, I'm not going to see
21 it before you see it. Okay? Let's proceed,
22 Professor. Sorry for the interruption.
23 Q. Can you tell me, to the best of your
24 recollection, as you sit here right now, what
25 kinds of assumptions you made in the calculations
27
1 LARRY PETERSON
2 you were describing.
3 A. I had assumed that we were going to do
4 a large file transfer independent of its content.
5 Q. And how large a file?
6 A. I did the math for both 6 gigabytes and
7 650 megabytes.
8 Q. Okay.
9 A. I assumed for first calculations --
10 Q. By the way, I'm sorry to interrupt you,
11 what's the relative size difference between those
12 two files that you mentioned?
13 A. It's a factor of 10 in order of
14 magnitude.
15 Q. Okay. Go ahead.
16 A. And how long that would take on a
17 1-megabit-per-second link using 1 megabit as an
18 example of what one might have into the home.
19 My recollection is the numbers worked
20 out that the 6-gigabyte file took 11 hours and the
21 650 megabytes took a tenth of that, it was an hour
22 and a half or something, an hour 20 minutes.
23 Q. So, does that mean that if one had a
24 650-megabyte file using the type of Internet
25 bandwidth that one might find in a home that one
28
1 LARRY PETERSON
2 could effectively transfer that entire file in how
3 long?
4 MR. GARBUS: I'll object if there are
5 any other variables.
6 A. So, what I said was how fast could you
7 download at one megabit per second.
8 Q. Right.
9 A. And I had came up with the number 1
10 megabit because that is an example of what people
11 are claiming you can do into the home.
12 Q. Okay. And how long did that take
13 again?
14 A. An hour 20 minutes, an hour and a half,
15 I forget exactly.
16 Q. Wouldn't you say that that's a fair
17 reflection of file transfer time for a file of
18 that size over a network connection of that size?
19 MR. GARBUS: I'll object to the
20 question.
21 A. Depends.
22 Q. All right. What are the factors?
23 A. The factors are do I actually get the
24 whole megabit. Let's suppose that I have 1
25 megabit into my home, that's what's been
29
1 LARRY PETERSON
2 advertised to me. I'm only going to achieve that
3 if I can, in fact, get the 1 megabit from where
4 the data started into my home. So, I know that
5 the last piece runs at 1 megabit, the Internet is
6 in the middle, and there's going to be the -- what
7 would be called the upload or uplink at the other
8 end, the connection at the other side as into the
9 Internet. So, if that were a megabit and I could
10 get a whole megabit through the Internet and I
11 could get a megabit across that last link into my
12 home, then I could do it in that time.
13 Q. Did you actually run that as an
14 experiment?
15 A. I measured -- no, I did not do that as
16 an experiment.
17 Q. Now, were there any other scenarios
18 that you postulated in --
19 A. So, another scenario is that the uplink
20 wasn't a megabit. And I thought about that
21 scenario a little bit because much of the
22 technology out there, cable modem in particular,
23 ADSL is a second example, are asymmetric, and I
24 upload at different speed than I can download.
25 Q. Higher or lower?
30
1 LARRY PETERSON
2 A. Much lower, quite a bit lower.
3 (Whereupon, the requested portion of
4 the testimony is read back by the reporter.)
5 (Whereupon, a discussion takes place
6 off the record.)
7 Q. What was the outcome of your
8 contemplating that scenario?
9 A. I reran the numbers again for 128
10 kilobits as an example of what an uplink might be.
11 Q. Using both a 6-gig file and a 650-meg
12 file?
13 A. That's right, and that's an order of
14 magnitude less bandwidth and so you would multiply
15 those times by roughly 10 to get the total
16 transfer time.
17 Q. Any other scenarios that you had
18 contemplated?
19 A. If you had 10 megabits, so going in the
20 other direction, instead of 1 megabit if you had
21 10 megabits, then you would basically divide the
22 time by 10.
23 Q. Okay. Which means that our
24 hypothetical 650-meg size file over a 10-meg
25 Internet connection would yield us what in terms
31
1 LARRY PETERSON
2 of transfer time?
3 MR. GARBUS: I object to the question.
4 A. So, let me be clear that I am, at this
5 point, just doing the math. "If" I can get 10
6 megabits, "if" I can get 100 megabits, whatever
7 the number is.
8 Q. And I'm asking you what your math
9 yielded based on the assumption that you had a
10 10-megabit Internet connection?
11 A. If you had a 10-megabit link to the
12 data, it would transfer -- the 6 gigabytes would
13 transfer in an hour, roughly.
14 Q. Right. And the 650-meg file?
15 A. In a tenth of an hour, six minutes.
16 Q. Six minutes.
17 A. Well, it would be 10 minutes, in there
18 somewhere, because it was a little more than an
19 hour before.
20 Q. Okay. Did you study, contemplate, or
21 do the math for any other scenarios?
22 A. Not that I recall.
23 Q. Have we just exhausted, to the best of
24 your recollection, the content of the notes that
25 you referred to a moment ago?
32
1 LARRY PETERSON
2 A. No.
3 Q. What else was in there?
4 A. So, the chief other point made that I
5 jotted down in my notes is that looking just at
6 the uplink speed or the downlink speed was a
7 little simplistic because the Internet is going to
8 be in the middle. And it's quite likely that the
9 speed that you see would be limited by what the
10 Internet could deliver.
11 Q. Can you give me what your fair
12 approximation is of how that would factor into
13 these numbers?
14 A. It's very difficult to say. And the
15 reason is the Internet is a shared resource and
16 millions of people are using it constantly, and
17 the rate that you see at any given time depends on
18 exactly what set of links the data had to traverse
19 and what other users were using that. So, it's
20 difficult to say.
21 Q. When you say "what set of links," are
22 links the same as hops?
23 A. Sure.
24 Q. Can you give me an average for the
25 average Internet connection, number of hops?
33
1 LARRY PETERSON
2 A. I'm not sure if I can give you an
3 average, but you tend to see 10 to 16 hops between
4 source and destination. That's not untypical.
5 Q. And in terms of traffic, which I
6 believe you said was another variable when you're
7 looking at the Internet in the middle, are there,
8 in your experience, ways to optimize the speed or
9 time which it takes to send something in light of
10 traffic problems?
11 MR. GARBUS: Object to the form of the
12 question. It's not understandable.
13 MR. HART: Do you -- I can rephrase it
14 if you don't understand it.
15 THE WITNESS: Let's try it again.
16 Q. Given what you said about traffic on
17 the Internet as another variable, are there ways
18 to optimize delivery time in light of such
19 traffic, and if so, what ways are there?
20 MR. GARBUS: You say other ways to
21 optimize? Excuse me, go ahead.
22 A. There's two possible solutions to the
23 traffic. This problem is often called congestion,
24 it's exactly the kind of thing you experience on
25 the road.
34
1 LARRY PETERSON
2 Q. Right.
3 A. And either you have a -- shall I stop
4 here?
5 (Whereupon, a discussion takes place
6 off the record.)
7 (Whereupon, the answer is read back by
8 the reporter.)
9 THE WITNESS: I do an analogy to the
10 congestion on the roadways.
11 MR. HART: Okay. We'll just continue
12 from there.
13 THE WITNESS: Could we go ahead and do
14 this. I've already --
15 MR. HART: You've switched gears.
16 THE WITNESS: I've already switched
17 gears.
18 MR. HART: Okay. Sure. Then we'll
19 stop. And I apologize for the distraction.
20 THE VIDEOGRAPHER: Off the record,
21 11:10.
22 THE WITNESS: It's on my account I need
23 to tell it to my account so I can put it
24 TECHNICAL PERSON: It's on a local PC?
25 THE WITNESS: It's on the Net. I need
35
1 LARRY PETERSON
2 some way of remotely logging into it. The file is
3 not in a place that's accessible from the web.
4 MR. HART: I don't -- I want to go off
5 the record at this point.
6 MR. GARBUS: I insist that we stay on
7 the record.
8 MR. HART: I don't know what the point
9 of this is. I think this should be off the
10 record. It's my deposition.
11 MR. GARBUS: I would like this on the
12 record. He had no idea.
13 MR. HART: I resent your slighting me,
14 and I ask you to apologize. I think it's totally
15 unnecessary.
16 MR. GARBUS: I apologize for saying you
17 know nothing.
18 MR. HART: I think it's rude and
19 unprofessional and a lawyer of your stature should
20 help me because you didn't bring what you should
21 have to the deposition.
22 THE VIDEOGRAPHER: Back on the record,
23 11:13.
24 MR. HART: We were interrupted to see
25 if we could retrieve this document from your
36
1 LARRY PETERSON
2 computer at Princeton, and apparently it's
3 complicated, and I don't want to belabor this
4 anymore, so I just want to proceed with the
5 deposition. So, we're going to kind of shift
6 gears back to the questioning.
7 (Whereupon, the last question is read
8 back by the reporter.)
9 A. So you can do one of two things. One
10 is you can add more capacity. It's like adding
11 additional lanes, I suppose, and that's typically
12 how these problems are solved, more users come on
13 to the Internet, they get higher speed links, you
14 have to add more capacity at the backbone.
15 The second possibility is that you give
16 some traffic preferential treatment, that one
17 packet is somehow more important, so you create a
18 carpool lane, to continue the analogy. That sort
19 of technology is in I would call the research
20 stages right now, people are trying to understand
21 how to do that. The problem is complex because
22 it's a huge distributed network.
23 Q. "It" being the Internet?
24 A. "It" being the Internet. And
25 coordinating that sort of agreement that this is
37
1 LARRY PETERSON
2 an important packet across all those hops is a
3 very difficult problem.
4 Q. Is there a third solution, to borrow
5 from your road traffic analogy, and that is don't
6 drive at rush hour?
7 A. Certainly, don't drive at rush hour.
8 Q. In your experience, is that often how
9 users get around network traffic problems, they
10 choose to send data over the Internet at off
11 hours?
12 A. I can't say in my experience because I
13 don't typically send data at off hours.
14 Q. If there is network congestion because
15 of the intervention of the Internet as part of the
16 series of links or hops that you described, and
17 someone attempts to send packets of data and
18 congestion is slowing that down, does that data
19 get queued or sort of stored someplace until it
20 can get fully transmitted through the process?
21 A. The way the Internet works, the sender,
22 the software running at the source of the data, is
23 paced by the receiver's successful receipt of
24 data, and basically the data then is buffered back
25 at the sender. It's just not transmitted until
38
1 LARRY PETERSON
2 the sender has detected there's enough capacity.
3 Q. And are there other technological ways
4 that are employed to buffer the data in between
5 the sender and receiver to take account of
6 congestion?
7 A. Not data now. We're now dealing with
8 packets.
9 Q. Okay.
10 A. Individual packets are queued at the
11 routers -- at each hop along the Internet, but
12 that's very limited how much is queued there.
13 It's just temporary storage until the link becomes
14 accessible. So, we're dealing in milliseconds of
15 time.
16 Q. I see. Now, taking all of that into
17 account in terms of what you know and what you're
18 prepared to testify to under oath in connection
19 with this case, can you give me any fair estimates
20 of -- and they could be in a range from low to
21 high -- of the transfer time involved when one is
22 transferring a file the size of a 650-meg file
23 over the Internet?
24 A. If a single person wants to transmit a
25 650-megabyte file, so we're not talking about now
39
1 LARRY PETERSON
2 the whole world is also transferring 650-megabyte
3 files, it could take anywhere from two to ten
4 times what you thought based on your end, your
5 last link that you should have seen. So, if you
6 had a megabit, you might get a tenth of --
7 somewhere between a half and a tenth of that in
8 practice.
9 Q. Just so the record is clear here, could
10 you state that rather than as a fraction of the
11 lead time, then just express it in terms of hours
12 or minutes?
13 A. So, back to the 650 megabytes -- so if
14 I have a 1-megabit link into my home and I'm
15 transferring 650 megabits, that would, at a
16 megabit, take, let's say the number, and again I'm
17 approximating here, an hour and 20 minutes, then
18 that take anywhere from two hours and 40 minutes
19 to ten hours.
20 Q. Gotcha.
21 A. And that's actually presuming that in
22 that time the connection didn't die for some
23 reason, which is also a possibility. It's a long
24 time, much longer than connections typically are
25 asked to stay alive transferring data.
40
1 LARRY PETERSON
2 Q. Is it common practice to your
3 experience to break such files into even smaller
4 file sizes to make them smaller and to make them
5 transfer faster. So, take your 650-megabyte file,
6 cutting it in half, for example, and sending it in
7 two pieces, is that another way to optimize --
8 A. That is done. That doesn't optimize
9 anything except helping you survive the
10 possibility that the link failed during any one
11 transfer.
12 Q. Now, you said a few minutes ago that
13 that assumes one transfer occurring at a given
14 point in time, but that it doesn't account for
15 everyone in the whole world transferring the same
16 size file at the same time.
17 A. That's right.
18 Q. I would just like to spend a few
19 minutes on that subject, if you would.
20 Are there congestion problems on the
21 Internet today?
22 A. Yes.
23 Q. And is that attributed to the fact that
24 there are a lot of people sending and receiving
25 stuff all over the world?
41
1 LARRY PETERSON
2 A. Absolutely.
3 Q. Do we have any idea how many millions
4 of people are on the Internet at a given time?
5 A. There are all kinds of estimates, 30 or
6 40 million maybe.
7 Q. At any one point?
8 A. Oh, I'm sorry. At any given point int
9 time? I've never heard a number like that.
10 MR. HART: Okay, Marty, I'm going ask
11 you to stop playing with your cell phone.
12 MR. GARBUS: Well, will you just give
13 me a moment so that I can call my office?
14 MR. HART: Well, you know -- sure I'll
15 give you a moment to call your office.
16 MR. GARBUS: Just let the record
17 indicate that it relates to scheduling the
18 Marshall King deposition because --
19 MR. HART: One of the reasons why we're
20 taking so long is because you keep doing
21 distracting things, including making telephone
22 calls while we're on the record in the course of
23 testimony.
24 MR. GARBUS: I have no objection to you
25 continuing the deposition right now.
42
1 LARRY PETERSON
2 MR. HART: It's distracting and it's
3 inappropriate and I've complained about it before.
4 (Whereupon, a discussion takes place
5 off the record.)
6 MR. HART: Are we ready to resume, Mr.
7 Garbus?
8 MR. GARBUS: Yes.
9 MR. HART: I would prefer in the
10 future, I don't want to get into arguments with
11 you, Marty, but if you need to make a telephone
12 call, just indicate for the record that you are
13 going to do that and step out and we'll stop the
14 deposition, because I find it very distracting
15 for you to be pressing things that make beeping
16 noises while I'm trying to ask the witness a
17 question or think about his answer. Thank you.
18 Q. So, we don't have an estimate of, to
19 your knowledge --
20 A. I can't say that I've ever heard an
21 estimate like that.
22 Q. Is there anything else that you
23 postulated, considered, or analyzed mathematically
24 or otherwise in connection with your testimony in
25 this case?
43
1 LARRY PETERSON
2 A. You're talking specifically about my
3 notes?
4 Q. Well, let's finish the notes first.
5 Have we exhausted what you think was in your
6 notes?
7 A. I'm pretty sure we've exhausted that.
8 Q. Now, leaving aside the notes, are there
9 any other scenarios that you contemplated or
10 assumptions that you made where you did math or
11 calculations or projections that relate to
12 testimony that you may give in connection with
13 this case?
14 A. I gave some thought to the possibility
15 of 650 megabyte or 6-gigabyte transfers being
16 commonplace and what kind of stress that would put
17 on the Internet.
18 One of the things I did in relation to
19 that was look at what was available on the web
20 today. In particular I went to the @Home web
21 site. @Home is a company that sells cable modem
22 service, one of the first to do that. @Home is
23 now teamed with Excite, which is a web portal, I
24 guess you could say.
25 And at the @Home page, they were
44
1 LARRY PETERSON
2 bragging about "Get @Home and you can download all
3 of this wonderful content. See Excite." So, I
4 went to the Excite page and looked around and I
5 found some high-end data.
6 Q. When you say "high-end data," what do
7 you mean?
8 A. By "high-end data" I simply mean it was
9 video, audio, their high resolution, their good
10 stuff. In fact, they had several grades, as I
11 recall, that you could download this -- this high
12 resolution thing or if you don't have enough
13 capacity, you could download this low resolution
14 thing. So, looking at the high resolution, the
15 biggest file that I could find was 8 megabytes,
16 approximately.
17 So, I interpreted that as buy cable
18 modem and you, too, can download 8-megabyte files.
19 That is all the bigger they put up, and the reason
20 I believe that's the case is that if everyone
21 started downloading 8 megabytes or larger than 8
22 megabytes, they would have trouble in the
23 backbone.
24 Q. "They" being?
25 A. @Home or whoever the service provider
45
1 LARRY PETERSON
2 is.
3 Q. Gotcha.
4 A. So, I conclude from that, that 8
5 megabytes is what one might, as a high-end user,
6 get off the Internet today, and that's consistent
7 with what I've seen elsewhere as well. We're
8 basically talking two orders of magnitude to get
9 to 650 megabytes, a factor of 100.
10 So, one conclusion you can draw from
11 that is it's going to take something on the order
12 or two orders of magnitude improvement in the
13 Internet's infrastructure to make 650-megabyte
14 transfers as possible as 8 megabytes is today.
15 Q. Do you know how big an MP3 audio file
16 of a song or of a record album is in file size?
17 A. About a megabyte a minute, so 3 or 4
18 megabytes.
19 Q. That's for a single song?
20 A. I believe so.
21 Q. And an album would be, what, a function
22 10 times that if we assume 10 songs are on an
23 album?
24 A. Sure.
25 Q. Are you familiar with Napster as a
46
1 LARRY PETERSON
2 phenomenon?
3 A. I am.
4 Q. Has Napster contributed, to your
5 knowledge, to system congestion on the Internet?
6 A. I can't say whether it has or not.
7 Q. Are you aware of any universities or
8 colleges that have banned Napster because it has
9 overloaded their services?
10 A. I saw stories in the press, nothing
11 first hand.
12 Q. Hasn't happened at Princeton, for
13 example?
14 A. They have not banned Napster at
15 Princeton.
16 Q. Do you know if you have a Napster
17 problem at Princeton?
18 A. I don't know.
19 Q. Do you have any idea of the volume of
20 so-called file-sharing activity taking place on
21 the Internet today involving MP3 music files, like
22 how many users and how many files are exchanged?
23 A. I haven't seen any numbers like that.
24 Q. Is it immense, to your knowledge?
25 A. My impression is that it's not immense,
47
1 LARRY PETERSON
2 but I couldn't say for sure.
3 Q. What do you base that impression on,
4 sir?
5 A. How I haven't heard within the Internet
6 research community, the people thinking about the
7 new problems cropping up, Napster being a
8 particular problem. Or I should say Napster being
9 singled out as a particular problem.
10 Q. Leaving aside Napster, same question
11 generally with respect to MP3 audio so-called file
12 sharing as a problem, have you --
13 A. Not that I've heard of.
14 Reflecting on that, make sure that I'm
15 being specific enough here, that's in reference to
16 the Internet as a whole or a particular campus?
17 Q. I understood that your testimony
18 related to the Internet as a whole. Is that how
19 you meant your testimony?
20 A. Yes.
21 Q. Is there any kind of number or figure
22 that you can give me as an approximation which
23 would indicate the saturation point for the
24 Internet in terms of the number of files or the
25 amount of files, the amount of users sharing
48
1 LARRY PETERSON
2 certain amount of files that would cause the
3 Internet to reach saturation?
4 A. I'm not sure what you mean by
5 "saturation." One way to look at it is that the
6 Internet is today as big as it needs to be, has
7 the capacity that it needs to have. And because
8 it's expensive to put these facilities in, it's no
9 bigger than it needs to be. Which is to say it's
10 running very near saturation. Incrementally,
11 things get add -- capacity gets added and then it
12 gets consumed and it gets added, it gets consumed.
13 So, we're always running near full capacity with
14 the current facilities.
15 Q. But things still manage to get through.
16 Right?
17 A. Eventually.
18 Q. Let's turn to campus life for a couple
19 of minutes and talk now about the kinds of
20 networks that exist on a college campus like
21 Princeton's, and I'm going to have several
22 questions about that.
23 First, is it fair to treat such a
24 system as sort of a closed system in the sense
25 that you don't need to make the Internet part of
49
1 LARRY PETERSON
2 the chain to transfer files from one computer on a
3 university network to another computer on a
4 university network?
5 A. You do not -- those files need not go
6 out into the Internet to do that. They will pass
7 through a point within the campus that's connected
8 to both the Internet and the campus, so there is a
9 piece of equipment there, a router, that's the
10 bridge between the two.
11 Q. Now, have you done any reflection,
12 studies, mathematical extrapolations of transfer
13 times based on a university system that does not
14 involve passing data through the Internet?
15 A. We did the 10 megabit number a little
16 earlier, which is reflective of what one might
17 find on the campus.
18 Q. Have you done any studies, mathematical
19 calculations, or contemplated what the load factor
20 would be, if that's a word, on the campus network
21 if people were transferring 650-meg files on that
22 network?
23 A. Well, not specifically. If you have a
24 shared 10-megabit-per-second ethernet, then the
25 numbers we talked about earlier for 10 megabits
50
1 LARRY PETERSON
2 will apply if only one person is doing it and you
3 will get in the ballpark of half of that if two
4 people are trying to transfer at the same time and
5 so on.
6 Q. So, just, again, to be clear, I'm sorry
7 to be painstaking about this, I need to make the
8 record, this is assuming people are transferring
9 the 650-meg file at the same time or
10 simultaneously. Right?
11 A. Right.
12 Q. And if, as you said, 10 people are
13 doing it, you would get an effective transfer time
14 that looked more like the 1-megabit connection in
15 that scenario.
16 A. The first approximation, yes.
17 Q. Now, aside from shared networks, is it
18 also common to have switch networks at university
19 campuses?
20 A. Yes.
21 Q. And does your answer, again, going back
22 to this hypothetical of one person over a 10-meg
23 line or 10 people over a 10-meg line, differ in
24 any way based on whether it's a shared network or
25 a switch network?
51
1 LARRY PETERSON
2 A. Typically, the switch network -- I
3 have -- I'm the only one that can use the link
4 between my computer and the switch. The switch is
5 now shared in the same way that the cable was
6 before. So, I would have 10 megabits for my use
7 and all of the other people plugged into that
8 switch would have 10 megabits, in this example,
9 for their use. It's possible there's still
10 contention on the switch.
11 Q. I'm sorry, it's possible there's
12 still --
13 A. Contention is still possible, so let me
14 just --
15 Q. "Contention" is the word you used?
16 A. "Contention" is the word I used. This
17 is what happens when multiple people share the
18 same link or switch.
19 Q. Okay.
20 A. So, if it's the case that -- it depends
21 on the extent to which these parties are operating
22 independently. If it were to happen that one user
23 was providing data, and he has a
24 10-megabit-per-second link to the switch, and the
25 other nine users each had their own
52
1 LARRY PETERSON
2 10-megabit-per-second connection to the switch,
3 but they all wanted to download this one file,
4 then the link, that one 10-megabit link, even
5 though I'm the only one using it, connected to it,
6 would still be shared between all nine people
7 trying to download from it.
8 Q. But if there were a one-to-one
9 relationship between two 10-megabit switch users
10 on the network within the university, how many
11 different simultaneous one-to-one transfers could
12 be occurring given the university backbone?
13 A. So, in this simple scenario, we'll stay
14 with that for now, A could talk to B and C could
15 talk to D and those would be independent of each
16 other.
17 Q. Correct.
18 A. That will typically depend on how many
19 users I could connect to that one switch, so a
20 common switch is maybe 16, 32, 64 ports,
21 independent connections. There still, depending
22 on how the system is architected, the possibility
23 of contention within the switch. It may, in fact,
24 not have the capacity to simultaneously switch 64
25 connections because the designs engineered it in
53
1 LARRY PETERSON
2 such a way that they didn't expect that to happen.
3 Q. And we're talking about people with
4 10-meg connections right now in this scenario?
5 A. Right.
6 Q. And just remind me again, I'm not
7 trying to be difficult here, I know you answered
8 it, but I just forgot the answer, what is the file
9 transfer time of a 650-meg file over a 10-megabit
10 connection?
11 A. 650 megabytes at 10 megabits was
12 somewhere between 6 and 10 minutes.
13 Q. Okay, fine. So, is it fair to say that
14 it's at least theoretically possible on a switch
15 network with, I believe what you said was, up to
16 64 ports for our example, that you could have 64
17 different simultaneous transmissions of that
18 hypothetical 650-meg file occurring, assuming
19 there were no contention problems at the switch?
20 A. It is possible.
21 Q. Okay.
22 A. I'm sorry, 32 because --
23 Q. Sending and receiving.
24 A. Yeah, right.
25 Q. Okay. So, in that scenario, we would
54
1 LARRY PETERSON
2 be contemplating 32 effective file transfers of a
3 650-meg file within roughly 6 to 10 minutes?
4 MR. GARBUS: In a theoretical concept.
5 A. It's possible.
6 Q. And if that were being done, would that
7 overload the university network?
8 A. It would completely saturate that
9 portion of the network.
10 Q. And when you say "that portion," we're
11 talking about in relation to those computers
12 hooked up to that network talking to each other at
13 the same time?
14 A. That particular switch. There are more
15 than 64 users within a university. So, the
16 university has switches connected not just to end
17 users but to other switches and to other switches
18 still. So, depending on how isolated this
19 particular load was, it might totally load one
20 switch. It may not impact the rest of the campus,
21 possibly.
22 Q. And that problem could be meliorated in
23 a number of ways. Right? For example, we don't
24 necessarily have to have 32 senders and 32 users
25 sending a 650-meg file simultaneously on the same
55
1 LARRY PETERSON
2 switch network. Right? It could be fewer users
3 and fewer senders?
4 A. Certainly could be fewer, right.
5 MR. GARBUS: Just note my objection
6 that we're continuing to talk about theoretical.
7 And I think it would be helpful if you phrase that
8 each time in your question so that we don't lose
9 sight of it in the transcript.
10 Q. And it would also meliorate the
11 situation if we took the 650-meg file and cut it
12 in half so that we're dealing with two packages
13 rather than one larger one?
14 A. Not really, because I still -- assuming
15 that I want to get the whole file, I still have to
16 transfer both of them.
17 Q. What if you added additional servers to
18 that mix, would that change your analysis in any
19 way?
20 A. That doesn't really change it because
21 I'm talking about sources. Whether you call them
22 servers or they're students in their dorm rooms,
23 they're still the sources of the data.
24 Q. Now, have you done any real world tests
25 taking into account some of these file transfer
56
1 LARRY PETERSON
2 times that we've been talking about in a
3 university context?
4 A. In the context of this case since I've
5 been thinking about it?
6 Q. Well, I'll take that as the first
7 question, sure.
8 A. The only experiment I ran since hearing
9 about this case was to see how much I could get at
10 home off my DSL line. So, I transferred 30-some
11 megabytes just to try it, and I got 1 megabit in
12 one direction and 2 megabits in the other
13 direction. That's the only experiment I've run.
14 That was between my home and the department at the
15 other end. So, I didn't go out on the Internet.
16 That was just into my office.
17 Q. And you're at 1 meg at home?
18 A. I downloaded at 2 meg, I uploaded at 1
19 meg.
20 Q. Do you recall what the file contained?
21 Was it audiovisual material?
22 A. It didn't matter. I can't even recall
23 what it was now.
24 Q. Did you look at the result?
25 A. The result is guaranteed to be the same
57
1 LARRY PETERSON
2 that I transmitted.
3 Q. Are there any other areas that you have
4 considered for possible testimony in connection
5 with this case, and if so, what are they?
6 A. Are we talking areas now broader than
7 networking?
8 Q. Either within the networking topic or
9 outside of the networking area.
10 A. I'm not sure what I am going to be
11 asked to testify on. I mostly concentrated on
12 networking.
13 Q. And were you asked to consider any
14 other subjects outside of networking by Mr. Garbus
15 or his firm?
16 A. I think the only question I can recall
17 outside of that domain was if I knew anything
18 about Linux.
19 Q. Do you?
20 A. I use it on my desktop. We use it in
21 my research lab.
22 Q. Have you ever used DeCSS?
23 A. No.
24 Q. Do you know what it is?
25 A. Yes.
58
1 LARRY PETERSON
2 Q. How do know what it is.
3 A. Just through my own investigations
4 after having talked to Ed.
5 Q. And what did your investigations
6 consist of?
7 A. I think I got a pointer to crytome.org,
8 which I believe is posting some of the depositions
9 or declarations. Went through those a little bit
10 but not in any detail.
11 Q. When you say you got a "pointer," what
12 do you mean?
13 A. It came up in conversation that there
14 was material relating to this case on crytome.org.
15 Q. Someone told you to go to crytome.org?
16 A. No. It just came up in conversation
17 that -- I picked up in the conversation that this
18 was a place that had -- I didn't want to look
19 foolish not knowing anything about this case.
20 Q. I understand. I didn't mean to sound
21 accusatory. I was just trying to differentiate
22 between someone told you versus you found it
23 through a search engine. When you said you got a
24 pointer, I didn't know it that was a term of art.
25 That's all.
59
1 LARRY PETERSON
2 Are you getting paid for your testimony
3 in this case?
4 A. No.
5 Q. And is it your intention to appear at
6 the trial in this case?
7 A. If I can schedule it, yes.
8 Q. Is there anything that you're aware of
9 presently that would impede your ability to appear
10 at the trial of this case?
11 A. I promised to take my kids to a Yankees
12 game on the 20th.
13 Q. And other than that?
14 A. Other than that, that's the only thing
15 I can think of. I've got tickets.
16 Q. Now, back within the networking area,
17 are there any other subjects that you have
18 contemplated for possible testimony in this case
19 regarding networking, if I can use that term as
20 generally as possible, that we have not already
21 covered here today?
22 THE WITNESS: Can I talk to Marty for a
23 second?
24 MR. HART: If you need to, sure.
25 THE VIDEOGRAPHER: Off the record,
60
1 LARRY PETERSON
2 11:33.
3 (Whereupon, a brief recess was taken.)
4 THE VIDEOGRAPHER: Back on the record,
5 11:43 a.m.
6 MR. HART: Let the record reflect that
7 we're waiting for Mr. Garbus to get off the phone
8 even though he was summoned.
9 (Whereupon, the last question is read
10 back by the reporter.)
11 A. So, the answer is the big question that
12 I've been thinking about in the last few days is
13 how practical is it to download 650-megabit files
14 or 6-gigabyte files over Internet.
15 Q. Over the Internet as opposed to through
16 university campus networks?
17 A. Over the Internet.
18 Q. Okay, because if you look at your
19 declaration, and I think we marked that as Exhibit
20 2, take a look at paragraph No. 3 you say that one
21 of the areas that you've been asked to testify
22 about in this case is "the theoretical and
23 practical transfer speeds available on college and
24 university local area networks." Do you see that?
25 A. Yes.
61
1 LARRY PETERSON
2 Q. And my question is: When you referred
3 to --
4 MR. GARBUS: Excuse me, don't I see
5 paragraph 3 "and to the public generally"?
6 Haven't I read that correctly?
7 MR. HART: Okay. I'm sorry, I would
8 really like to ask the questions and not have you
9 testify.
10 MR. GARBUS: Well, I don't want you to
11 misstate.
12 THE WITNESS: No, I'm fine.
13 MR. HART: Okay. I haven't asked the
14 question yet, Marty. If you have an objection --
15 MR. GARBUS: You've read selectively
16 from the affidavit and you've implicated --
17 MR. HART: Are you done?
18 MR. GARBUS: -- and implied to the
19 witness --
20 MR. HART: Are you done testifying?
21 MR. GARBUS: No, I'm not.
22 MR. HART: Are you done coaching?
23 MR. GARBUS: And you have implied to
24 the witness --
25 MR. HART: I haven't implied anything.
62
1 LARRY PETERSON
2 I haven't asked the question yet.
3 MR. GARBUS: -- what he said with
4 respect to the affidavit. What I'm asking for is
5 an accurate reading of what the man has signed his
6 name to.
7 MR. HART: Are you done now?
8 MR. GARBUS: Go ahead.
9 MR. HART: Thanks. Sorry for that
10 again.
11 Q. Paragraph 3. We talked about
12 university networks in the sense that they weren't
13 connected through the Internet or the Internet was
14 not part of the intermediate chain in the transfer
15 process. Right?
16 A. Right.
17 Q. And then separately we talked about
18 using the Internet.
19 A. That's right.
20 Q. Right.
21 A. And you asked me if I had any more to
22 say about networks defined generally.
23 Q. Right. And you said you had been
24 contemplating the issue of transferring a 650-meg
25 file over the Internet in practical terms.
63
1 LARRY PETERSON
2 A. In practical terms.
3 Q. Right. And I guess the question that I
4 was trying to get to there was, have you also
5 contemplated the practical transfer times involved
6 with university local area networks not using the
7 Internet as part of the chain?
8 A. Some, but not beyond what we've already
9 covered, that there are shared networks and there
10 are switches within universities. Nothing to add
11 to that.
12 Q. Okay. Fair enough.
13 Now, let's turn to the question about
14 what you've been contemplating in terms of the
15 practical transfer speeds available on the
16 Internet for a 650-meg file?
17 A. We touched on this earlier. I observed
18 that what was considered a big transfer on the
19 Internet now was this file in the neighborhood of
20 8 megabytes, and that is a pretty sizable transfer
21 on the Internet today.
22 Q. And that's based on your earlier
23 testimony about looking at @Home and making
24 assumptions about why their file sizes are 8 meg?
25 A. Not making assumptions about, but
64
1 LARRY PETERSON
2 finding that was the biggest file I could find for
3 them to demonstrate their cable service.
4 Q. Okay.
5 A. And as I said, that number is
6 consistent with what I would expect people to be
7 able to transfer over the Internet today.
8 So, the point here is that, and I think
9 I made this earlier, that it's two orders of
10 magnitude to get to 650 megabytes, a factor of
11 100. That's restating what we've already talked
12 about.
13 Q. Fair enough.
14 A. A number -- set of data consistent with
15 that is that today most people still have 28
16 kilobit modems, maybe they've got 56 kilobits.
17 It's a relatively few number of people that
18 actually have this megabit-per-second connection
19 that we're talking about some of us having. So,
20 the majority of the people that are on the
21 Internet still have two orders of magnitude to
22 catch up with the high end.
23 So, I asked myself the question: Even
24 assuming that 1 megabit is enough to do a
25 650-megabyte download, again, that's an hour's
65
1 LARRY PETERSON
2 plus transfer, which is a sizable transfer, the
3 question, then, is how -- I've got two pieces of
4 data that suggest to me we need to have two orders
5 of magnitude improvement in the Internet's
6 backbone core capacity to accommodate commonplace
7 transfers of 650-megabytes files. How long is it
8 going to be before we get to that point? What
9 kind of changes is going to be necessary?
10 Q. On the Internet generally.
11 A. Yes, generally. The only way I can
12 answer that question is to go back in time and see
13 how long did it take to do the two previous orders
14 of magnitude improvement.
15 Q. And how long did that take?
16 A. That was a ten-year process.
17 Q. In your experience, have improvements
18 increased in speed over time on the Internet?
19 Let me rephrase that question because I
20 don't confuse speed with transfer time.
21 MR. GARBUS: Object to the form of the
22 question.
23 MR. HART: I just said that I would
24 rephrase it. Thank you.
25 Q. You were looking backwards to look at
66
1 LARRY PETERSON
2 how long certain levels of improvement took on the
3 Internet, and I guess my question in simple terms
4 is the rate of change that's appearing in
5 technology --
6 A. Is it getting faster, faster?
7 Q. Yes. That's my question.
8 A. It's hard to say. If you go back and
9 try to divide that time out, you find that in
10 1990, the backbone was made up mostly of --
11 actually, let's go back a little farther than
12 that. No, let's not do that.
13 If you go back to 1990, the backbone
14 consisted mostly of 1-megabit-per-second links,
15 order of magnitude 1 megabit per second. If you
16 go to 1995, you measured backbone capacity in the
17 10-megabit-per-second range, and now we're
18 measuring in the 100-megabit-per-second range.
19 Q. For backbone?
20 A. For the backbone. So, we're seeing, at
21 least across -- with the intermediate point of
22 five years, a fairly steady increase in capacity.
23 Q. Do you have any understanding of what
24 projections are for backbone bandwidth increases,
25 say, over the next year to five years on the
67
1 LARRY PETERSON
2 Internet?
3 A. So, looking out, the gigabit is the
4 next one order of magnitude change and then we go
5 to 10 megabits, just to project out. The
6 10-megabit-per-second work is in, I would call it,
7 the research advance development stage. It's
8 still years out before that's going to be
9 commonplace.
10 Q. You mean commonplace in the home?
11 A. No, not in the home, in the backbone.
12 MR. GARBUS: Object to the form of the
13 question.
14 MR. HART: Okay.
15 THE WITNESS: Could you restate the
16 question?
17 MR. HART: Read it back. I'm sorry.
18 (Whereupon, the last question is read
19 back by the reporter.)
20 MR. GARBUS: Mr. Hart, if you want to
21 testify, swear yourself in. Otherwise, allow the
22 witness to answer the question.
23 MR. HART: Thank you, Mr. Garbus.
24 Go ahead.
25 A. No, not in the home, in the backbone.
68
1 LARRY PETERSON
2 Q. In the backbone on the Internet as
3 people use it in their homes today?
4 A. The backbone that people in their homes
5 connect to.
6 Q. Right. As distinguished, for example,
7 from the local area networks that we were talking
8 about within universities and colleges?
9 A. That's correct. Either you use the
10 Internet or you're limited to those set of people
11 that are connected to that local area network.
12 Q. Gotcha.
13 MR. GARBUS: I would appreciate if
14 you'd stop saying "gotcha" after each question. I
15 find it as disruptive as you found my use of the
16 telephone.
17 MR. HART: Boy, you really are surly
18 this morning, Marty.
19 Q. Any other observations on your
20 practical views of transfer speeds on the Internet
21 and what the future holds in that regard?
22 A. Nothing in particular.
23 Q. I would like to go back to the
24 university scenario for a couple of minutes with
25 the university networks that don't depend on the
69
1 LARRY PETERSON
2 Internet as part of the chain. Okay?
3 A. Okay.
4 Q. If, for example, a whole bunch of
5 people were hypothetically transferring 650-meg
6 files over networks at Princeton, would that have
7 any effect, say, on the bandwidth or traffic on a
8 network, let's say, up at MIT or in Palo Alto at
9 another university where people might also be
10 engaged in the same thing?
11 A. No.
12 Q. Do you know how many university or
13 college students and other people that are around
14 universities and colleges and have access to
15 university and college networks there are in the
16 United States?
17 A. I don't know.
18 Q. What?
19 A. I have no idea how many students or
20 faculty there are.
21 Q. Now, the @Home/Excite portal that you
22 were talking about, that, to your knowledge, is a
23 commercial, legitimate portal. Right?
24 A. That's a commercial entity, yeah.
25 Q. Do you have any knowledge whatsoever
70
1 LARRY PETERSON
2 about the Internet practices of people that engage
3 in sharing unauthorized copies of content?
4 A. Do I have any practical knowledge?
5 MR. HART: Let me read it back, because
6 I may have misphrased the question. I apologize
7 if I did.
8 (Whereupon, the last question is read
9 back by the reporter.)
10 MR. GARBUS: Object to the form.
11 A. I'm not sure what it means to be an
12 "Internet practice."
13 Q. Let me try it this way: Are you aware
14 that there are people that use the Internet that
15 are fairly sophisticated computer users?
16 MR. GARBUS: Object to the form of the
17 question.
18 Go ahead, if you can answer it.
19 A. Am I aware of their being sophisticated
20 computer users?
21 Q. Yes.
22 A. Yes.
23 Q. Are you aware whether there are people
24 that are sophisticated computer users that can
25 take advantage of the latest technologies in terms
71
1 LARRY PETERSON
2 of bandwidth and computational power of computers?
3 MR. GARBUS: I object to the form of
4 the question. It's so silly.
5 The witness can answer.
6 A. If they can afford to.
7 Q. Do you have any knowledge whatsoever
8 about pirated content files being transferred on
9 the Internet?
10 A. None first hand.
11 Q. What do you know anecdotally?
12 A. Pirated files. I guess I don't know of
13 any pirated files. Apparently, something is
14 happening in that the MPA has taken action, but
15 that's the extent of it.
16 Q. Are you aware of the legal
17 controversies surrounding the so-called file
18 sharing of MPEG-3 files involving audio
19 recordings?
20 MR. GARBUS: I object to it. It has no
21 relationship to this case.
22 (Whereupon, the last question is read
23 back by the reporter.)
24 A. I'm aware of it.
25 Q. Are you aware whether there is any kind
72
1 LARRY PETERSON
2 of traffic, say, in illegitimate copies of
3 computer software application programs via the
4 Internet?
5 A. Not that I'm aware of.
6 Q. Do you have any knowledge that there
7 exists, essentially, an underground market on the
8 Internet through which people trade copyrighted
9 content by so-called file sharing or file
10 transfer?
11 MR. GARBUS: I object to the form of
12 the question. It's not an appropriate way to ask
13 the question. If you want to testify as a witness
14 and you want to swear yourself in, I'll allow the
15 witness to answer the question.
16 A. I'm not aware.
17 Q. Now, given what you said about
18 increases in Internet backbone bandwidth, do
19 advances in compression technology offer a partial
20 solution, in your estimation?
21 MR. GARBUS: Object to the form of the
22 question.
23 A. Partial solution to what?
24 Q. To traffic -- in other words --
25 MR. GARBUS: I object to the form of
73
1 LARRY PETERSON
2 the question.
3 MR. HART: Fine, Marty.
4 A. To the extent compression can reduce
5 the size of the data, that would mean there's less
6 data to transmit, it can be transmitted in less
7 time.
8 Q. Excellent. That was precisely what I
9 was asking. I thank you for that. That was
10 eloquent.
11 MR. HART: I want to take five minutes
12 and just see if I have -- what I have left, I want
13 to consolidate it.
14 MR. GARBUS: I have a 12:30 lunch that
15 I have to leave for at 12:15. I advised
16 Mr. Cooper of that beforehand.
17 MR. HART: Right. And given the fact
18 that I was taking this deposition, you didn't see
19 fit to advise me of it, but that's okay, Marty.
20 MR. GARBUS: My memory is I did advise
21 you at the beginning that there was a 12:30 lunch,
22 and I think it's on the record.
23 THE VIDEOGRAPHER: Off the record,
24 12:07.
25 (Whereupon, a brief recess was taken.)
74
1 LARRY PETERSON
2 THE VIDEOGRAPHER: Back on the record,
3 12:12.
4 Q. Do you have any knowledge of what the
5 projections are for improvements in video
6 compression technology as we sit here today?
7 A. No.
8 Q. Do you own a DVD player?
9 A. Yes, I do.
10 Q. Is it a computer-based DVD player?
11 A. No.
12 Q. Or a set top box?
13 A. It sits next to my television.
14 Q. Have you ever tried to copy a movie
15 from your DVD player?
16 A. No.
17 Q. Would you know how to do that?
18 A. With the help of my 17-year-old son I
19 could probably do it.
20 Q. Why do you say that?
21 MR. GARBUS: Objection. It's obvious.
22 It's a silly question. Go ahead.
23 Q. Do you adopt Mr. Garbus's statement
24 that it's obvious?
25 A. Yes.
75
1 LARRY PETERSON
2 Q. Okay. Why is that so?
3 A. Because it's only 17-year-old kids that
4 understand how to run my stereo system.
5 Q. Have you written or published anything
6 that's inconsistent with any of the testimony that
7 you've given us here today?
8 A. Not that I know of.
9 Q. Assuming that the electronic transfer
10 of movies in digital form were feasible, do you
11 regard that as something that would be harmful to
12 the copyright owners of movies?
13 MR. GARBUS: I object to the form of
14 the question. That's not for this witness to
15 decide. He can answer it, if he wants. He's not
16 a lawyer and he's not a judge.
17 A. I wouldn't consider it not necessarily
18 harmful.
19 Q. Why is that?
20 A. I'm sure there is a wealth of business
21 opportunities in delivering movies over the
22 Internet.
23 Q. You mean with authority of the
24 copyright owners?
25 MR. GARBUS: I object to the form of
76
1 LARRY PETERSON
2 the question. If you want to testify, Mr. Hart,
3 why don't you swear yourself in.
4 MR. HART: Thank you, Judge Garbus.
5 THE WITNESS: I'm sorry, the question
6 again was?
7 Q. You're last answer was, and we can have
8 it read back, that there were a wealth of business
9 opportunities. And my question was in light of
10 that last answer, is that based on the assumption
11 that that would involve the authority of the
12 copyright owner?
13 A. I'm not sure what it means to involve
14 the authority of the copyright owner.
15 Q. What did you mean by a "wealth of
16 business opportunities," then?
17 A. My only point was that the Internet is
18 another way of delivering content and surely
19 someone will figure out how to take advantage of
20 that.
21 MR. HART: I don't have any further
22 questions. Thank you for your candor and your
23 time, Professor.
24 THE WITNESS: You're welcome.
25 THE VIDEOGRAPHER: Off the record,
77
1 LARRY PETERSON
2 12:15.
3 (Whereupon, the deposition
4 of LARRY PETERSON is concluded at 12:15p.m.)
5
6
7
8
9
10 o O o
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
78
1
2 J U R A T
3
4 I DO HEREBY CERTIFY that I have read
5 the foregoing transcript of my deposition
6 testimony and I certify that it is true and
7 correct to the best of my knowledge, subject
8 to the changes, if any, shown on the attached
9 page(s).
10
11
12 ___________________________
LARRY PETERSON
13
14
15
16 SWORN TO AND SUBSCRIBED
17 BEFORE ME ON THIS ________
18 DAY OF _______________, 2000.
19
20 _____________________________
NOTARY PUBLIC
21
22
23
24
25
79
1
2 C E R T I F I C A T E
3
4 I, JOMANNA CASTANO, a Certified Shorthand
5 Reporter and Notary Public, do hereby certify
6 that prior to the commencement of the
7 examination LARRY PETERSON was sworn by me
8 to testify to the truth, the whole truth
9 and nothing but the truth.
10 I DO FURTHER CERTIFY that the foregoing is a
11 true and accurate transcript of the testimony as
12 taken stenographically by and before me at the
13 time, place and on the date hereinbefore set
14 forth.
15 I DO FURTHER CERTIFY that I am neither of
16 counsel nor attorney for any party in this action
17 and that I am not interested in the event nor
18 outcome of this litigation.
19
20
21
_______________________________
22 JOMANNA CASTANO
Certification Number 2216930
23 My Commission Expires August 19, 2003
24
25