He said one of the SCO attorneys, Robert Silver, seemed to be asleep at one point. At any rate, his eyes were closed and his head kept bobbing. There was, at Brent Hatch's request, a "vertical break" for a few minutes so the room could stand up and stretch. In short, it was a marathon.

Frank's impression is that Kimball had made up his mind by the time it was over, and if he had to guess, he'd guess that he is going to rule against SCO on its motion and for IBM on its motion. SCO did bring up its Expedited Motion to Enforce the Scheduling Order, and Judge Kimball asked IBM to answer SCO's motion by next Wednesday, and SCO to respond by Friday, so Judge Kimball could get matters resolved. So I don't expect a ruling on today's matters until the following week. Frank noticed that Judge Well's assistant was in the audience, taking notes, by the way.

SCO's Motion

First, they argued SCO's motion, their Motion to Dismiss or Stay. Hatch spoke for SCO first, saying that they wanted to talk about scheduling issues. Kimball asked if the matter was full briefed yet, and Hatch said no. And Kimball said, 'You're not likely to get that." Then he said they'd talk about it at the end.

Then Mark James spoke for SCO. He argued that IBM's counterclaim 10 was extremely broad and introduced significant new issues. This, he said, is a contract case. A copyright counterclaim isn't a mirror of that. IBM says their counterclaim is compulsory, but James said SCO claims it is permissive. Leaving the 10th counterclaim in the case would confuse the jury.

Kimball then asked, "If Counterclaim 10 is gone, the jury won't be confused any more?" The room burst out laughing.

James then argued that there would be a need for more discovery and there were already discovery problems and issues. "IBM has stonewalled us," he said.

Then Evan Chesler spoke for IBM.³ SCO pleads one thing and then argues another, he said. He said he couldn't hear either party's pleadings being addressed in what SCO is arguing now. He went back to the complaint to show why the counterclaim is compulsory. Their 2nd Amended Complaint said that IBM infringed, induced and contributed to the infringement of their copyrights. So, IBM's counterclaim is a mirror image of that.

Chesler then mentioned the Red Hat and AutoZone cases. SCO threatened Red Hat, too, and they brought a declaratory judgment action that SCO also sought to dismiss but failed. In AutoZone, they said the issues are similar to IBM's. And that's about copyrights. So is Red Hat. They claimed the issues there are similar to IBM's too. SCO's strategy, Chesler said, was "Anywhere But Here," meaning litigate the matter anywhere else but where they are currently.

James for SCO then said they didn't place at issue *all* IBM's contributions. IBM did. They are injecting that into the case.

Then Robert Silver spoke for SCO, talking about Red Hat and AutoZone, but Frank couldn't understand what he was saying specifically, but the point was that the cases were actually different than the IBM case.⁵ Someone send a telegram to DE please.

That was the end of the arguments on SCO's motion to dismiss. Then they started the arguments on IBM's motions.

IBM's Motions

David Marriott spoke for IBM. Frank says he is so easy to listen to. He's very organized, he tells you how many points he will address and then he labels them, "First,..." so naturally that appeals to the logically trained geek mind. Marriott said that there were three reasons the judge should grant IBM's motion for a declaratory judgment of non-infringement:

1. SCO has failed to present evidence of infringement;

2. The alleged evidence they have presented (Gupta et al declarations) are no impediment to granting IBM's motion;

3. SCO's proposition is that they need more time to do discovery, but all that they need they've had from the beginning.

On point 1, he read off a list of public statements SCO has made that they own copyrights and that there has been substantial infringement by IBM. Frank said it was like reading Groklaw. All the quotes we have about the MIT deep divers, the pattern recognition experts, the mountains of code. Frank thinks Marriott jokingly at one point mentioned icebergs of code. But IBM still has no response from SCO on IBM's Interrogatories 12 and 13, despite Judge Wells' order. IBM, Marriott said, believes that is because SCO has no evidence.

There then ensued a discussion about a confidential email, by a Mike Davidson or Davison, Frank wasn't sure which, saying that SCO did a comparison of Linux and Unix and found nothing. The email is from 2002. There is a Mike Davidson of SCO, so I assume that is who is meant.

On point 2, Marriott said that the Gupta declaration was inadmissible and should be stricken. But even if it was admissible, Gupta failed to do an abstraction, filtration, comparison test. He referred to the sealed Brian Kernighan declaration, and apparently he went through the Gupta declaration material and concluded that there were less than 300 lines and they were all filterable and not protectable. Marriott's point was that these 300 lines, even if they were not filterable, are not substantial quantifiably or qualitatively.

On point three, he said there was no need for more discovery. They have what they need. For purposes of this motion, IBM has copied and has induced other to copy Linux. 25,000 man years are not needed. That's a "straw man", Marriott said. Showing access is not relevant, but for the purposes of this motion, he said, youi may assume access.

At this point, as Hatch was given the cue that it was his turn to speak, he asked for the "vertical break." After the break, he talked about discovery again, and how badly they need it. Kimball asked him, "You've got Linux and Unix. What more is it that you need?" Hatch said Fred Frei would answer that separately. Then they brought forth a memo from October 4, 1999, a consultant's report by Robert Swarts. Frank is guessing at the spelling, but the point of the memo was to undo the confidential email about finding nothing. IBM said they'd only been given this memo several hours ago. Excuse me. 1999?

Then Frei spoke. ⁶ This was on the 56(f) motion to strike. He talked about why they need more. There is nonliteral copying, structures, sequences and organization, and algorithms and data structures. He showed examples, using a big screen, where he could scan in a page and then show it on the screen. ¹ The examples he showed were these:

sem.h
elf.h
msg.h
perror.c

They look similar, he said, but it's hard to compare them.⁴ Sem.h, for example, in Linux, it's one variable per line, whereas in Unix, it's all on one line, with commas, as in:

In Linux, it is:

int a;
int b;
int c;

In Unix, sem.h is:

int a,b,c;

Then he showed elf.h and how they are identical. Duh. For those new to this subject, I suggest you might wish to read this article on ELF.

With regard to IBM's motion to strike, he said Sontag's declaration shows the enormity of the task of comparison; Gupta's show what they could likely find with more discovery; and Harrop's shows the procedural history of the discovery issues. Because more discovery would likely find more items like Gupta mentions, he said the judge shouldn't rule on IBM's motion to strike now.

He said both Sontag and Gupta do qualify as experts and he seemed to be trying to qualify them on the spot. He said SCO wants a road map. He used the comparison: If you want to drive from Aurora, NY to Los Angeles, you need a map, and without discovery from IBM, it's like they won't give SCO the road map. Judge Kimball asked, "Where's Aurora, NY?" I like his sense of humor. And he does pay attention to detail, does he not?

Marriott then got to speak, and he said about the road map: They claim we are not giving them the road map, but giving them what they are asking for in discovery is like giving them a road map to China. You won't make it to LA with that road map. You can't get a Linux-Unix comparison by looking at AIX and Dynix code.

He addressed the memo, saying SCO is trying to explain away their email saying they found nothing. And, by the way, it says that someone went to BestBuy, bought Red Hat 5.2 and compared it with numerous versions of Unix, so clearly it can be done and it didn't take them 25,000 man years. He also addressed the experts issue, and argued that Gupta and Sontag do not qualify as experts.²

Kimball then asked IBM's Marriott: SCO says IBM is only now introducing the issue of copyright ownership in its Reply Memorandum. Marriott then showed their Motion and Supporting Memorandum did mention that SCO must show that they own the copyrights and that there is substantial copyright infringement, so that is addressing the issue of ownership right there.

SCO's Expedited Motion

Hatch and James both spoke. They complained that they thought they'd have their discovery issues solved yesterday, on the 14th, but IBM got it postponed to October 19. That's terribly unfair to SCO and they said SCO really needs help with the discovery matter. They wanted Kimball to have the benefit of the discovery issues before Judge Wells before today's hearing. So Kimball asked IBM how much time it was given to answer the expedition motion to enforce. ("Expedition" is a Freudian slip on my part. I meant to write "expedited". I am leaving it as is, though, because, like many Freudian slips, it reflects what I am really thinking. As in fishing.) IBM answered and said they intend to respond on time. Then Kimball said, he'd like IBM to answer by one week from today and asked SCO to reply two days later. That is fast. If a hearing was needed, he said, they'd do it by phone. He wants the issues resolved.

I'd like to stress that these are one man's eyewitness impressions, but they are not legal impressions or predictive in any way. Others were there and they will add to the story.⁷ One has already emailed me that you really can't believe how good IBM's lawyers are unless you see them in person and in action. He says they speak off the cuff just as well as when they are handling matters that were prepared in advance. cxd says about Marriott:

"I will tell all of you. He is the best courtroom litigator I have ever seen. He is very organized. He is to the point and the Judge likes that. He did the community proud today. He is my new vision of a super hero."

All in all, it sounds like it went very, very well.

We have also gotten some more exhibits. A lot of them. One is a second declaration from Randall Davis. He did a Unix-Linux comparison, and he tells how long it took. Not 25,000 man years, I'm guessing.

² A new document has just shown up on Pacer that may help us to understand the point being made here, IBM's Reply Memorandum in Support of Motion to Strike Materials Submitted by SCO in Opposition to IBM's Cross-Motion for Partial Summary Judgment, which says: "In response to IBM's motion to strike certain incompetent and inadmissible materials that SCO had submitted in support of its opposition to IBM's cross-motion for partial summary judgment on its Tenth Counterclaim, SCO concedes that the Declaration of Chris Sontag, Sandeep Gupta, and John Harrop were not submitted for the purpose of attempting to show that a genuine issue of material fact exists that would preclude summary judgment. Rather SCO now insists, these declarations were submitted only 'for a very narrow purpose: to provide the Court with Rule 56(f) facts' (SCO Opp'n at 1), and were not intended by SCO to oppose IBM's motion 'on the merits'. (Id. at 30.) In light of SCO's admission, then, the Court should not consider any of the statements in the Sontag, Gupta, or Harrop Declarations (or any portions of SCO's opposition to IBM's summary judgment motion that cite such testimony) in deciding whether a genuine issue of material fact exists with respect to IBM's motion."

³Another eyewitness, cranesable, provides some more detail about Mr. Chesler's segment:

"Chesler for IBM --

           If you look at the arguments SCO has put forward, the interesting thing is that we didn’t hear their pleadings motion. [Hands JK a copy of the 10th CC pleading.] Go to pg. 40, par. 171 and 173. The scope is narrow.  IBM doesn’t infringe in *its* activities. If you want to know what a claim says, it ought to be read.

   Par. 179, pg. 52 of SCO’s claim.  Quotes 179, and points out that the verbs IBM chose were the exact verbs SCO used in their complaint.  These three verbs that SCO tells you are incredibly broad came right from their complaint.  We didn’t make this stuff up.  They said it. Par. 62 of 2nd Amended Complaint.  “IBM-Related Agreements.”  Note those words are capitalized, referring to specific agreements. SCO says IBM has “infringed, induced infringement and contributed to infringement.” [these are the 3 magic verbs] Par. 113.  “IBM has violated contract by using and assisting others to use SYSV for external purposes… by actively supporting transfer of Unix technology into Linux. Par. 118 on pg. 35.  “Transferring portions of Unix SYSV by… giving it to Linus Torvalds for open distribution to the public.”  [I remember several times he referred to the code being in the “public domain.”  So, even one of the Nazgul seems to not quite understand the distinction between public domain and GPL. Par. 122.  “Breach of confidentiality by contributing to Linux.” Choice quote.  “We didn’t pick this fight.” All we want is a declaration that we didn’t do what they say we did.  They’ve been publicly accusing us of these terrible things for over a year.

Look at what SCO has told other courts. [Hands judge a book full of SCO’s public declarations.] In SCO v. IBM, SCO argued, “This precise issue will be litigated in a case against Autozone.”  Moved to dismiss or stay, pending Autozone.  Has since dropped that argument. RedHat:  “Infringement issues are currently before Judge Dale Kimball.” Now they say litigating that issue would expand this case. In May, RedHat moved to lift stay.  SCO argues, “It would be a waste… while a substantially similar issue is before Judge Kimball.” They cannot reconcile their representations today with their representations in RedHat. SCO’s strategy is clearly one of “anywhere but here.”  Wherever we are, the issue should be litigated elsewhere. “We have a right to litigate it here.” We plainly satisfy the 4 criteria SCO brought up. We are not seeking a declaration on what somebody else did.  They have defined the scope of our activities.

Regardless of all that has been said, even if it were just about AIX and Dynix, we’d still be whether we put their code into Linux.  The basis of their purported termination [he likes to say this] of our license was our contributions to Linux. Even if we ignore all the other arguments, and you decide this motion is permissive, you still have discretion to grant it, and you should exercise your discretion to do so, because:

•      Interrogatories 12 & 13 (What is yours that we took, and what did we do with it) have not been answered.  They keep arguing that they shouldn’t have to answer those.
  
• Their failure to provide those answers is the basis for our motion for PSJ.
  
•     They have all the source code they need to do a comparison.
  
•     If they don’t have all of the code they need for a comparison, how can they have a basis to sue Autozone?

⁴Frank has now reached his computer and provides this further information:

Computer comparison does not always detect similarities, even if the similarity is obvious enough that even a lay person may note the remarkable similarity. For an example, he showed the following comparison of sem.h: (Note: Though this example looks almost identical to what was shown in court, this code actually comes from the OpenBSD sem.h.)

UNIX	Linux
struct seminfo { int semmni, semmns, semmnu, semmsl, semopm, semume, semusz, semvmx, semaem; };	struct seminfo { int semmap; int semmni; int semmns; int semmnu; int semmsl; int semopm; int semume; int semusz; int semvmx; int semaem; };

He was trying to show that an automated comparison would not catch the similarity between

--
int semmni,
semmns;
--
and
--
int semmni;
int semmns;
--

These two sections of code are functionally the same (both declare two variables as 'int' or integers), but automated comparison would not catch that they are the same. According to SCO, this is an example of 'non-literal' copyright infringement, and shows how someone could potentially copy source code and obfuscate it so that it is undetectable.

*******

I also heard from Dr Stupid, who points out:

The "struct seminfo" was released under a permissive license with old SCO's blessing as part of the lxrun package as late as 1999, as was a msg.h file. The COPYING file from that release refers the reader to the source files for the copyright information, but neither sem.h nor msg.h in that package carries a copyright notice.

⁵ Another eyewitness, Totosplatz reports: "Mr. Silver made these points: (1) this case is not about copyright, and (2) the case is only about contracts."

⁶Totosplatz has a bit more on Mr. Frei's argument:

Mr. Frei had some quite amazing claims to make, and he made those claims twice, basically that there is a direct path from UNIX through AIX and Dynix to Linux. He said IBM contributed AIX and Dynix to Linux and he stated that there is a direct path from UNIX through AIX and Dynix to Linux. But his basic point is that a comparison of UNIX to Linux is not enough, more is needed.

Judge Kimball asked, "More time, more from IBM, or both?" Mr Frei: "Both." Judge Kimball: "One might assume that a comparison of UNIX to Linux might have been done before filing a lawsuit." Mr Frei responded, "Yes, but that was Red Hat 5.2 in 1999 when Linux only had 2.9 million lines of code, Linux now has an additional 2 million lines of code."

⁷Yet another eyewitness report from Chris Brown:

I'm sure it's a given in these things, but I've got to make the disclaimers that I'm not a lawyer, I'm not a court reporter, & I have no fiduciary interest in IBM or SCO. I'm a humble Linux geek who lives in . . . Utah. Therefore, my coverage of the September 15th, 2004 hearing are just observations of what is best described as a layman, and subject therefor to omissions, errors, bias, and general sloppiness. Just the same I took notes and attempted to gather as much information as I could. Unfortunately, when IBM's counsel was speaking (esp. Mr. Marriott and Mr. Chesler) I had difficulty taking notes as they were spellbinding. Outstanding orrators, very compelling, focused, prepared, and in every way professionals.

Mr. Hatch spoke first for SCO to tell Judge Kimball what SCO wished to address. Mr. Hatch brought up the Expedited Motion on Enforcement of the Scheduling Order. Judge Kimball expressed incredulity that Mr. Hatch wished to discuss it, as it was neither in front of the Judge nor yet responded to by IBM's counsel. The Judge said that he's "not likely to get that." There was general laughter from this comment. The Judge was aware of what SCO was seeking today and expressed it as "no motions on summary judgement until discovery is completed."

Mr. James, speaking for SCO, addressed SCO's motion to dismiss or stay IBM's 10th counterclaim. He asserted that IBM's 10th CC is permissive as opposed to IBM's assertion it is compulsury. SCO's argument was that IBM's 10th counterclaim was overly broad, attempting to inject issues far outside SCO's complaint, including 3rd-party Linux contributors making contributions as much as 10 years before IBM made their first Linux contribution. He said that there were literally hundreds of such contributors whose contributions would need to be researched. Mr. James repeatedly brought up the significant expense and time that would be incurred by such. He asserted that, in essence, to rule on IBM's counterclaim all such contributions would have to be judged non-infringing as well.

SCO used a digital projector and screen to display various PowerPoint-like slides to accompany their presentation. On one slide, they attempted to show how the 10th counterclaim raises legal issues that are not "largely the same", nor concern the same issues as SCO's complaint. They stated that for the CC to be compulsory it must match in "Time, Space, Origin, and Motivation" to SCO's complaint. SCO insisted it did not meet these tests, that the contributions to Linux by 3rd parties extend back as much as ten years before IBM first made a contribution to Linux, that in space the counterclaim neccessarily encompasses hundreds of contributors rather than just IBM. As to origin I believe he referred to code from the hundreds of Linux contributors rather than just IBM. And as to motivation he said the motivation of those 3rd-party contributors were necessarily different than IBM's motivation. SCO asserted that the "logical relationship" between CC10 and SCO's complaint was absent and that having "some overlap" was insufficient to make CC10 compulsory. He cited "Millelacs Band of Chippewa Indians v. Minnesota". He "harps" on discovery and claimed that IBM is stone-walling and blocking discovery. Mr. James used the terms "complex" or "complexities" about 13 times before I started losing count. Oh, and "jury confusion" thrown in a few times too.

Mr. Chesler spoke for IBM in opposition to SCO's motion to dismiss or stay IBM's 10th counterclaim. While Mr. James' presentation was OK, Mr. Chesler was eloquent and very focused. He asserted that IBM's 10th counterclaim is about *IBM's* activity. He said he'd been taught in school that if you wanted to know what a plaintiff's claim was, you went to the source. He directed attention to page 40 of CC10, paragraphs 171 and 173. Paragraph 173 is saying IBM is entitled to declaratory judgment on *IBM's* activities. He referred to paragraph 179 on page 52 about IBM not infringing copyright. He said CC10 comes directly from SCO's claim.

He spoke about "SCO's chairman running around the country making claims". Mr. Chesler referred to SCO's Second Amended Complaint. . . Mr. Chesler states that in these references SCO is specifically asserting that IBM infringed SCO's alleged copyrights by copying SYS V source code into Linux. He says IBM's 10th counterclaim is a direct answer to those assertions, and is therefore compusory.

Mr. Chesler went on to cite a brutal list of SCO's contradictory claims and statements made in the Autozone and Red Hat cases. That in those claims SCO indicates to each of those courts that Linux copyright issues will be addressed in a different court, and that those assertions pointedly indicate that they will be heard here before Judge Kimball. He goes on to point out that in *this* court, SCO is discaiming copyright issues. He said SCO's strategy seems to be an "Anywhere But Here Strategy." Mr. Chesler spoke further on this but since he was so compelling I had trouble keeping notes. He discussed SCO's failure to respond to IBM's Interogatories #12 and #13 where SCO is to state what protected property of SCO's IBM is infringing, and in what way. This despite *two* court orders to do so.

Mr. James rebutted Mr. Chesler. He seemed to claim both that thousands of 3rd-party contributions are at issue and also that *only* a few 3rd-party IBM subcontractors are involved. I'm not sure what his point was on this. The Judge had asked Mr. James to reply to IBM's claims about SCO's conflicting statements in Autozone and RedHat. Mr. James said that SCO counsel Mr. Silver is a better person to speak to these issues. Mr. Silver took the podium, but there commenced a most interesting commentary where Mr. Silver seemed to dispute Mr. James' characterization of his abilities to speak to the Red Hat issues. Judge Kimball seemed upset about this and the next minute or so could have come from a Laurel and Hardy skit. Eventually Mr. Silver simply stated that SCO has not made misleading or false statements in any of the other courts. Mr. Silver & Mr. James quite simply did not address or resolve the contadictory statements, to my satisfaction.

My word! This was a long hearing: over three hours. My note-taking skills were sorely pressed, especially in light of my being captivated by IBM's counsels' skills. The courtroom was packed, I counted about 60 people, though that includes about 12 lawyers, 2 court clerks, 2 interns, the Judge, reporter, and whoever-it-is-that-sits-next-to-the-reporter. There were several press reporters there, including Robert Mims from the Salt Lake Tribune whose reports we've read before. ... I also recognized a RedHat employee I've had conversations with before. He stated that there was another RedHat employee there but that *he* didn't know who it was. Frank pointed out several people he knew that had been at earlier hearings. SCO appears to have brought a half-dozen or so employees, most of whom neither Frank nor I knew. Ryan Tibbits was there, seated in the "cheap seats." Most of the rest, on my impression, were Linux supporters. One person actually planned a vacation from Oregon around being here and was bummed about the Sept. 14th hearing being rescheduled. I was too, but obviously I need only drove 20 minutes to attend.

Next we heard arguments on IBM's motion for PSJ on CC10 and motion to strike Sandeep Gupta's et al statements. Mr. Marriott spoke for IBM. He started off by saying he had three points to make. The first was that SCO had failed to supply evidence. He quoted *numerous* Darl McBride quotes made to the press over the last 15 months. All of which we've repeated often here. Quotes of SCO having "deep dived" into the code, three teams working on comparing code including one from the "MIT Math Department", that they've compared the code "every which way but Tuesday", and that there are thousands of lines of line-for-line copying from UNIX to Linux. There were numerous additional quotes. Mr. Marriott indicated how these all pointed to SCO claiming they *already* did the code comparisons necessary between UNIX and Linux, and that they *already* have evidence, but that SCO has failed to answer IBM's Interrogatories #12 and #13 in which they *should* have stated what SCO property IBM is reputed to have misused and in what way. And that despite TWO court orders by Judge Wells, they have not answered them to this day.

Mr. Marriott stated that if SCO wished to defeat IBM's counterclaim 10, THIS would have been the method to do so. He stated that SCO's failure it do so is the principle proof that SCO has no evidence and that IBM's CC should be granted.

Mr. Marriott then handed out an email he referred to as "the Davidson email". The email had been marked Confidential by SCO and he said he was therefore unable to read aloud any of the content. I will discuss this email further later in my report. At this time he simply directed the Judge's attention to several items in the email that he stated supported his position. Secondly, Mr. Marriott discussed the "alleged evidence" in the Sandeep Gupta Declaration. He said that declaration was not filed by SCO in response to IBM's interogatories numbered 12 and 13, but was instead attached to a motion by SCO. . . . Mr. Marriott shows a sheet of paper of Gupta's table of code that Gupta purported to be infringing and he also handed out a *BOOK* of the original source code from which the lines of code on the table were taken. The book, he indicated, had the indicated lines of code highlighted in yellow. He points out, thus, that Gupta's code examples are cherry-picked and juxtaposed in an attempt to indicated infringement. Further, he pointed to the code in Gupta's table (where SCO code was on a column on one side with the Linux "equivalent" in a column on the other side) and indicated to the Judge that it is even obvious to a *layman* that the code, the cherry-picked code, is not similar.

Third he discusses the need for SCO to compare *UNIX* to Linux. (This was a theme of IBM's thoughout the afternoon, that to prove infringement, SCO must compare SCO's "own" UNIX code to Linux and show copying.) He discussed Chris Sontag's claim that up to 25,000 man-years may be needed to compare the versions of AIX they have received to Linux. He discussed the time issue of comparison, and especially that SCO has *already* demonstrated its ability to compare code quicker through its earlier claims of the deep dives & etc. He stated that interim versions of AIX, or especially AIX itself, were not required.

Mr. Hatch, before replying, asked for a 10 minute vertical break. The court so granted. When returning from break, Judge Kimball specifically asked Mr. Hatch to address IBM's assertion that SCO already has access to all the code they need. The Judge specifically said that SCO "owns" the UNIX code and that anyone could get the Linux code. Mr. Hatch replied that "that's disingenous". Judge Kimball snapped back quickly with "How's that disingenous?". SCO counsel Mr. Silver at this point falls asleep. Indeed through most of the rest of the hearing he is either nodding off, or completely asleep. . . .

Mr. Hatch goes on about not getting discovery from IBM including discussions we've heard already about AIX/Dynix code (all versions) and the 7,200 names IBM gave them. He complained that IBM's list, undoubtedly, included secretaries and others who are not relavent. (Though I specifically seem to recall SCO had asked for all those with *access*. It would seem improper had IBM *not* included them if they indeed had access.)

Mr. Hatch discussed the Gupta declaration, pronouncing Gupta three different ways, varying soft and hard vowel sounds. He quotes from Proctor & Gamble v. Amway to support PSJ must be denied as premature. Judge asks again about how & why more discovery is needed when SCO could do the comparison with the SCO UNIX and Linux source they have. Mr. Hatch does not answer Judge but goes back to the P&G presentation stating that SCO needs expert discovery.

He said "substantial similarity test" is fact-intensive and is a "classic" issue for the jury. He shows Robert Swartz' memorandum. Slide is titled "1999 Report (Oct 4, 1999) Line for Line Identical To UNIX" shows how report's author claims to have found line-for-line identical code. Not highlighted is author's comment about "fragments of code in various programs from various areas of the OS, not whole modules." Mr. Hatch turns time over to Mr. Frei.

Mr. Frei says he is "going to show why comparing SCO UNIX to LInux is not all that is needed." He uses a slide and compares software to a book. He discusses how a plot in a book is analagous to the function of a software product. He says that you cannot sit down a 6-year-old and have him compare code and discover similarities in the "plot." Nor can an automated tool do so, but it instead needs a very time-consuming process by experts.

He shows a slide from the Gupta Declaration showing comparisons of sem.h, and msg.h, and some other header file. He uses these to explain how in some cases copying can be obvious, identical, code and in others how semantics of commas or semicolons make the code not identical, but still obvious copying. Further he demonstrates how some may not even look similar. He explained how they have used automated comparators that have not turned up anything, but with an expert knowing what areas to compare, they were able to find substantially similar infringment. And that without such knowledge, they must make the laborious comparison of everything to everything that could take up to the 25,000 man-years. He says "we seek a roadmap to make the comparison" from IBM. He said they need to know who worked on AIX, Dynix, Linux and that they are going to be seeking admissions of copying and obfuscation.

He goes on to discuss CMVC and how they need not just releases of AIX, but all versions of AIX and Dynix. He shows perror.c from sontag declaration. Shows comparison between version 1 and version 17. He explains how much change is evident from one version to another, and that they need all versions of AIX so they can show see those changes. He states: "AIX started in UNIX and ended in Linux." On motion to strike he states that the Sontag and Gupta declarations are offered in support of 56(f) motion.

Mr. Marriott replies stating that he will only take 10 minutes (places wrist watch on podium and glances at clock. Judge notes clock as well.) (Mr. Hatch and Mr. Frei had spoken for a considerable time). Mr. Marriott talks fast, and to-the-point in refuting SCO's assertions. He uses frequent hand gestures, especially when making his repeated point about SCO needing to compare UNIX (left hand in air) and Linux (right hand in air). He speaks about SCO's roadmap analogy about driving from Aurora NY to Los Angeles (UNIX and Linux) and that having a map to China (or of some other place, speaking of AIX) will be of no help to them. Says SCO asks for a roadmap (from AIX) to Linux, but that doesn't help them compare Unix to Linux. I missed a lot of what Mr. Marriott had said as it was so fast and so compelling, I stopped taking notes. Sorry.

Judge Kimball says he will take it under advisement. He then discusses, briefly, the timetable on SCO's expedited request to enforce the scheduling order. SCO asserts they brought the expedited request solely because the Sept 14th hearing was moved to October, and they explain how that's going to give very little time for SCO to meet the schedule. Judge Kimball asks IBM to reply to SCO's expedited request by a week from today and for SCO to respond to IBM's reply by a week from Friday. Judge Kimball indicates preliminarily that he is not inclined to interfere with Judge Well's schedule. Judge Kimball recesses court.