11750 1 IN THE IOWA DISTRICT COURT FOR POLK COUNTY ----------------------------------------------- 2 JOE COMES; RILEY PAINT, ) 3 INC., an Iowa Corporation;) SKEFFINGTON'S FORMAL ) 4 WEAR OF IOWA, INC., an ) NO. CL82311 Iowa Corporation; and ) 5 PATRICIA ANNE LARSEN; ) ) TRANSCRIPT OF 6 Plaintiffs, ) PROCEEDINGS ) VOLUME XLIII 7 vs. ) ) 8 MICROSOFT CORPORATION, ) a Washington Corporation, ) 9 ) Defendant. ) 10 ----------------------------------------------- 11 The above-entitled matter came on for 12 trial before the Honorable Scott D. Rosenberg 13 and a jury commencing at 8:30 a.m., February 2, 14 2007, in Room 302 of the Polk County 15 Courthouse, Des Moines, Iowa. 16 17 18 19 20 HUNEY-VAUGHN COURT REPORTERS, LTD. 21 Suite 307, 604 Locust Street 22 Des Moines, Iowa 50309 23 (515)288-4910 24 25 11751 1 A P P E A R A N C E S 2 Plaintiffs by: BRADLEY M. BEAMAN 3 Attorney at Law Roxanne Conlin & Associates, PC 4 Suite 600 319 Seventh Street 5 Des Moines, IA 50309 (515) 283-1111 6 ROBERT J. GRALEWSKI, JR. 7 Attorney at Law Gergosian & Gralewski 8 550 West C Street Suite 1600 9 San Diego, CA 92101 (619) 230-0104 10 KENT WILLIAMS 11 Attorney at Law Williams Law Firm 12 1632 Homestead Trail Long Lake, MN 55356 13 (612) 940-4452 14 15 16 17 18 19 20 21 22 23 24 25 11752 1 Defendant by: DAVID B. TULCHIN 2 SHARON L. NELLES Attorneys at Law 3 Sullivan & Cromwell, LLP 125 Broad Street 4 New York, NY 10004-2498 (212) 558-3749 5 STEPHEN A. TUGGY 6 Attorney at Law Heller Ehrman, LLP 7 333 South Hope Street Suite 3900 8 Los Angeles, CA 90071-3043 (213) 689-0200 9 DAVID E. JONES 10 Attorney at Law Heller Ehrman, LLP 11 One East Main Street Suite 201 12 Madison, WI 53703-5118 (608) 663-7460 13 BRENT B. GREEN 14 Attorney at Law Duncan, Green, Brown & 15 Langeness, PC Suite 380 16 400 Locust Street Des Moines, IA 50309 17 (515) 288-6440 18 19 20 21 22 23 24 25 11753 1 STEVEN J. AESCHBACHER Attorney at Law 2 Microsoft Corporation One Microsoft Way 3 Redmond, WA 98052 (425) 882-8080 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 11754 1 (The following record was made in the 2 presence of the jury at 8:33 a.m.) 3 THE COURT: Everyone else may be 4 seated. 5 Before we begin, Plaintiffs' Exhibit 6 1369 was previously admitted I was told by 7 Mr. Green, so -- 8 MR. TULCHIN: That's correct, Your 9 Honor. 10 THE COURT: So that's admitted. Okay. 11 MR. GRALEWSKI: Thank you, Your Honor. 12 Good morning. We are going to 13 continue with the videotape deposition of 14 Mr. Barrett. 15 THE COURT: Thank you. 16 (Whereupon, the following video was 17 played to the jury.) 18 Question: You've been handed Exhibit 19 9 to your deposition, which is another string 20 of e-mails produced to us by Microsoft. 21 I'm going to direct your attention to 22 the e-mail that begins at the upper left-hand 23 corner of the first page. 24 Do you recognize this as an e-mail 25 from a David Cole, C-o-l, to Brad Chase, 11755 1 yourself, and others at Microsoft dated July 2 17, 1991? 3 Answer: Yes, I do. 4 Question: And who is David C-o-l? 5 Answer: David Cole, C-o-l-e, is his 6 last name, was the -- I don't recall the exact 7 title, group program manager on Windows 3.1, 8 sort of my counterpart. 9 Question: And what was the 10 responsibility of a group program manager? 11 Answer: Well, program management is 12 -- I think can be classified in some ways as 13 project management, where they -- they 14 basically do all the things that are necessary 15 to get the project started and completed, other 16 than develop the software, write the code and 17 that sort of thing. 18 Question: I see. So you worked with 19 him on the development side, or the two of you 20 worked together to develop a 3.1, Windows 3.1? 21 Answer: Yes. Together we were 22 responsible for delivering Windows 3.1. 23 Question: All right. Mr. Cole says 24 at the beginning of his e-mail, I think we 25 should use Windows to get a Microsoft operating 11756 1 system back on to the NetWare clients which 2 will bundle or require DR-DOS. 3 What was NetWare? 4 Answer: NetWare is -- NetWare is a 5 product from Novell that basically was a 6 network operating system, for lack of a better 7 term, that allowed PCs to talk to one another 8 over a network. 9 NetWare client -- well, you didn't ask 10 that question, sorry. 11 Question: Tell us what a NetWare 12 client is. 13 Answer: NetWare client is the 14 software that runs on a PC, which is basically 15 the software that does the connecting of the PC 16 and the PC operating system to the network. 17 Question: At this time, did Microsoft 18 have products that were competitive with 19 NetWare and NetWare clients? 20 Answer: Yes. 21 Question: And what were those 22 Microsoft products? 23 Answer: It was called LAN manager and 24 OS/2 LAN manager. 25 Question: Mr. Cole goes on to say, we 11757 1 should alter our plain a bit and move all the 2 DOS 6 improvements directly into Windows. 3 When the user starts Windows, they get 4 the Microsoft operating system (including 5 networking) and all the other cool features 6 that go with that. 7 When they quit, they get NetWare and 8 DR-DOS and no Windows applications. The key is 9 getting a piece of Microsoft system software on 10 the client so that we can deliver our strategy 11 and vision. We can leverage Windows and 12 Windows applications to do this. 13 Can you explain to me in lay terms 14 what the strategy was that Mr. Cole was 15 proposing here? 16 Answer: Well, the vision was 17 basically Microsoft software doing all the 18 functions. And strategy is to use various 19 pieces of Microsoft software to do that. 20 Question: How did this work? How 21 could Windows and Windows applications be 22 leveraged to make everything run on Microsoft 23 software? 24 Answer: Well, basically -- it's 25 actually outlined in this e-mail fairly 11758 1 succinctly. If you use -- Microsoft DOS 2 requires Windows to have, required to be 3 running for Windows to run, then basically 4 you've cut DR-DOS out of the loop and made it 5 essentially a useless appendage. 6 An additional cost item, if you've 7 standardized on Windows applications, which was 8 happening at that time, that that would 9 basically make DR-DOS an ordered expense, an 10 unnecessary added expense. 11 Question: And is there some way that 12 moving all of the DOS 6 improvements directly 13 on to Windows accomplished that goal? 14 Answer: Yes. 15 Question: How is that? 16 Answer: Yes. That's exactly -- 17 that's exactly to the point by essentially 18 putting DOS 6 into Windows that caused -- that 19 would cause that short-circuit to happen, that 20 DR-DOS would no longer be germane. 21 Question: Now, when Mr. Cole talks 22 about putting DOS 6 on to Windows, is that a 23 form of merging, technically merging Windows 24 and DR-DOS? 25 Answer: Yeah, that's exactly -- 11759 1 exactly it. 2 Question: Mr. Cole ends his e-mail 3 message by saying, if we are going to take over 4 the desktop when Windows starts, it must be all 5 Microsoft written software since Novell won't 6 help us do that. 7 Do you understand what was meant by 8 that sentence? 9 Answer: Well, yeah. Again, it goes 10 back to what I said earlier. The whole point 11 is to have only Microsoft software running and 12 taking over the desktop means precisely that. 13 Question: Mr. Barrett, you've been 14 handed Exhibit 10 to your deposition which is 15 another conglomeration of e-mail messages that 16 was produced to us by Microsoft. 17 I apologize for the small print, but 18 this is how it was produced to us. 19 Are you able to read this? 20 Answer: With some effort. 21 Question: There are a number of 22 e-mail messages here, but I'm actually 23 interested in the e-mail message that begins at 24 the bottom of the second page of the exhibit 25 that's Bates stamped MS-PCA 1178453. 11760 1 Do you recognize this as an e-mail 2 message from you to Brad Silverberg dated July 3 16th, 1996, subject Novell? 4 Answer: Yes, I do recognize this is 5 my e-mail. 6 Question: Was this e-mail generated 7 in response to the e-mail that we were just 8 looking at from Mr. Cole, Exhibit 9? 9 Answer: Yes, I -- let me look at that 10 just briefly. 11 Yes, I believe it was. 12 Question: You started out your e-mail 13 by saying, I think we need to carefully measure 14 our response. Totally freezing them out will 15 force them to compete on a wider basis and 16 could cause the disaster scenario to occur. 17 Do you know what disaster scenario you 18 were referring to? 19 Answer: Yeah. Disaster scenario is 20 just basically forcing, causing our competitors 21 to get together and up the playing -- basically 22 make the stakes much higher by when they get 23 together, they could become more significant 24 force. 25 Question: You go on to say, not to 11761 1 mention the thin FTC ice that we would be on. 2 FTC refers to Federal Trade 3 Commission? 4 Answer: Yes, it does. 5 Question: And what -- what did you 6 mean by thin FTC ice that you would be on? 7 Answer: Well, I'm not a lawyer, so 8 it's been pointed out to me, had been pointed 9 out to me that I really didn't have a basis to 10 make a judgment on. 11 Although there was a lot of -- there 12 were a lot of FTC issues going on there. 13 Question: During this time period? 14 Answer: During that time period. So 15 I was sensitive to it. 16 Question: All right. And I know 17 you're not a lawyer, but I'm just simply 18 asking, what was it that you were referring to 19 here in your lay opinion? 20 Answer: It was -- it was my belief -- 21 my belief that we had won. I mean, we had 22 basically sold at that point, I don't know, 23 four to six million copies of Microsoft Windows 24 3.0 and the corporations were flocking to us in 25 droves and we had basically won. 11762 1 And basically freezing out our 2 competitors might have been a bad thing to do. 3 But, again, not being a lawyer, I really -- you 4 know, probably was not -- should not have even 5 made that comment. 6 Question: Then you go on to say, on 7 the other hand, we have an advantage with 8 Windows and should press it in the systems 9 area. 10 What did you mean by that? 11 Answer: Well, because Windows was 12 becoming very, very successful at that point. 13 It was -- basically that software would wind up 14 on every corporate desktop. 15 And so not unlike what David Cole was 16 saying, that's our entry. That's our, if you 17 will, our springboard to do things that would 18 allow us to dominate the market. 19 Question: All right. 20 Answer: In case -- sorry. In this 21 specific case, it was about going to war with 22 Novell. 23 Question: All right. You go on to 24 say, one approach would be to deny enhanced 25 mode to all but real mode DOS customers. 11763 1 What is enhanced mode and what is real 2 mode? 3 Answer: Enhanced mode refers to 4 protected mode. There are a number of terms 5 for it. But basically it's the mode that 6 Windows would run in that would give the most 7 benefit by creating a large amount of memory. 8 Windows applications were very memory 9 hungry, and real mode limited them to what's 10 called the 640 K box, which was a very, very 11 constrained box at the time. 12 Small amount of memory forced them to 13 do some complicated and slow things in order to 14 fit all these features into memory. And 15 because of that, any application that ran in, 16 that didn't run in enhanced mode was at a 17 significant advantage. 18 Question: And enhanced mode or 19 protected mode Windows came out with which 20 release of Windows? 21 Answer: Windows 3.0. It was the 22 biggest feature of Windows 3.0. The thing that 23 caused Windows 3.0 to be successful. 24 Question: Now, what were you 25 referring to here when you said, one approach 11764 1 would be to deny enhanced mode to all but real 2 mode DOS customers? 3 Answer: Well, looking at it, I think 4 I probably didn't write this very clearly. 5 But what I was trying to say was that 6 MS-DOS, anybody using MS-DOS as the operating 7 system that Windows was running on would be our 8 -- we would only allow MS-DOS -- sorry. I'm 9 trying to say this clearly and I'm not doing a 10 good job. 11 Question: Take your time. 12 Answer: By making essentially MS-DOS 13 a requirement underlying Windows running in 14 enhanced mode would basically force anybody who 15 wanted to use enhanced mode Windows, which 16 basically is everybody, they would have to use 17 MS-DOS. 18 Question: As opposed to any competing 19 DOS? 20 Answer: That is absolutely correct. 21 As opposed to any other DOS from some other 22 company. 23 Question: Do you know if that 24 proposal was ever implemented? 25 Answer: I don't believe it was. 11765 1 Question: You go on to say, this can 2 be accomplished in a number of ways, but the 3 single most effective one is to move kernel (or 4 major components of it) into a VxD and thus 5 significantly up the cost of entry to running 6 full Windows. 7 What was it that you were proposing 8 there? 9 Answer: Okay, this is a very complex 10 and highly technical. Put it in a nutshell, 11 the kernel is the innermost piece. It's the, 12 if you will, the heart and soul of Windows, a 13 small piece of code. 14 Putting it into a VxD is a fairly 15 technical term for putting it into a 16 specialized -- making it be a specialized piece 17 of the Windows system. 18 Question: So the kernel you were 19 referring to here was the DOS kernel, the 20 MS-DOS kernel? 21 Answer: The Windows kernel. 22 Question: The Windows kernel, all 23 right. 24 Answer: The Windows kernel. 25 By doing that, there's other -- some 11766 1 other advantages, some technical advantages to 2 doing that, but the whole idea was to put it 3 into what's called a virtual device driver, 4 VxD, and thus making it difficult for people to 5 get at the kernel who do things with it. 6 Question: And that's what would 7 prevent other DOS competitors from making use 8 of Windows? 9 Answer: Yeah, it would prevent DOS 10 competitors from running Windows. Although, it 11 would require them, as I say here, it would 12 require them to do a bunch of work. 13 Wouldn't necessarily prevent them, but 14 it would cause them to have to do a bunch of 15 work and thus delaying their product plans. 16 Question: And is that work, that 17 extra work what you're referring to, is that 18 what was meant by significantly up the cost of 19 entry to running full Windows? 20 Answer: Yes, that's exactly what I 21 meant by that statement. 22 Question: Do you know whether 23 Microsoft ever implemented this idea of moving 24 the kernel or major components of it into a 25 VxD? 11767 1 Answer: I believe it did not happen. 2 Question: And VxD is virtual device 3 driver? 4 Answer: Yes, that's correct. 5 Question: Mr. Barrett, you've been 6 handed Exhibit 11, which is another 7 conglomeration of Microsoft e-mails which were 8 produced to us by Microsoft. 9 I'd like to direct your attention to 10 an e-mail that begins at the very bottom of the 11 first page from you to a L-a-w-r-e-n. 12 Do you know who that is? 13 Answer: I don't recall. 14 Question: Dated September 19th, 1991. 15 Subject re: Bambi and DR-DOS 6.0, many files 16 turned into crosslinked mess. 17 And then embedded within that e-mail 18 do you see a message from LAWREN to an alias 19 called Winbug dated the same date, September 20 19, 1991, on the beginning of the second page? 21 Answer: I'm sorry, the question? 22 Question: Do you see embedded within 23 your e-mail is an e-mail message from LAWREN to 24 an alias called Winbug on that same date? 25 Answer: Yes, I do. 11768 1 Question: Okay. Let's start with 2 that embedded e-mail that you were replying to 3 in your e-mail. 4 It begins, Windows bug report and then 5 it goes on to say, title, colon Bambi and 6 DR-DOS 6.0, many files turned into cross-linked 7 mess. 8 Do you know what that refers to? 9 Answer: Yes. 10 Question: Okay. Can you tell us what 11 that was? 12 Answer: Okay. Bambi I think was code 13 name for MS-DOS 6.0. And DR-DOS 6.0 is 14 obviously the Digital Research DOS. 15 Many files turned crosslink means -- 16 simply means that there was some problem that 17 occurred, that damaged the file system and 18 basically destroyed data. 19 Question: All right. And then do you 20 recall what the situation was where was this 21 crosslink mess? What caused that crosslink 22 mess? 23 Answer: No, I don't recall the 24 specifics of it. 25 Question: And then the e-mail at the 11769 1 bottom, the beginning of the bottom of the 2 first page was your response to the -- to 3 LAWREN's e-mail about the crosslink mess? 4 Answer: Yeah. Yes, it is my 5 response. 6 Question: Exhibit 12 has just been 7 handed to you, Mr. Barrett. And, again, this 8 is an e-mail, a series of e-mails produced to 9 us by Microsoft. If you would just take a 10 moment to look them over. 11 Answer: Okay. 12 Question: Now, there are some 13 references to Bambi in these e-mails also. 14 Was Bambi a beta version of a disk 15 cache driver that was being developed at 16 Microsoft? 17 Answer: Thank you for reminding me. 18 It, in fact, was that. It was a 19 technology we were developing to significantly 20 speed up performance of both Windows and DOS. 21 Question: Did that ultimately get 22 released, Bambi? Was it released as 23 Microsoft's SmartDrive disk cache driver? 24 Answer: Yes, that is exactly correct. 25 It was the early version of SmartDrive. 11770 1 Question: Were you involved in the 2 development of this disk cache driver that was 3 code named Bambi? 4 Answer: I was actually intimately 5 involved in it, yes. 6 Question: All right. Let me direct 7 your attention to the e-mail that begins at the 8 top of the page. 9 Do you recognize that as an e-mail 10 message from you dated September 28th, 1991 to 11 a Chuckst, Mikedr and Scottq? 12 Answer: Yes, I do recognize them. 13 Question: Who are those gentlemen? 14 Answer: Chuck Straub, I think, Mike 15 Dryfoos, and Scott Quinn. 16 And Chuck and Scott were the Bambi 17 developers, and Mike Dryfoos was the tech lead, 18 technical lead. 19 Question: And then your e-mail 20 message or response to an e-mail from Chuck 21 Straub to yourself and others at Microsoft 22 dated September 29, 1991? 23 Answer: Yeah, it was. 24 Question: All right. I'd like to ask 25 you about -- let's start with Mr. Straub's 11771 1 e-mail that you were responding to. 2 It's entitled -- the subject is Bambi 3 on DR-DOS 6.0. 4 Mr. Straub says, I tracked down a 5 serious incompatibility with DR-DOS 6. They 6 don't use the, quote, normal, close quote, 7 device driver interface for 32 M partitions. 8 What was he referring to there, if you 9 know? 10 Answer: Yeah. First of all, greater 11 than 32 M partitions means greater than 32 12 megabyte partitions. 13 There was a hard limit problem inside 14 of DOS, the file structures that DOS defined on 15 limiting sizes over 32 megabytes of disk space. 16 It seems like it's a trivial amount of 17 space now, but back then disks were in that 18 range. So this was cutting edge stuff back 19 then. 20 And the incompatibility is there is 21 simply the data structures on the disk that 22 people would use to essentially be able to get 23 access to greater than 32 megabytes. 24 MS-DOS did it one way, DR-DOS did it 25 another way. And if you tried to put MS-DOS on 11772 1 a machine that had a disk formatted by DR-DOS, 2 it wouldn't work over 32 megabytes. 3 Question: Okay. Now, did that have 4 some impact on this Bambi disk cache driver? 5 Answer: Certainly, certainly, because 6 Bambi understood -- Bambi understood at a very 7 low level how the disks were organized, the 8 data structures on the disks. So Bambi needed 9 to know the specifics of that to be able to 10 operate correctly. 11 Question: And this Bambi diskette 12 cache driver, was that incorporated into a 13 Microsoft product? 14 Answer: Yes, it was incorporated in 15 several Microsoft products, including Windows 16 and DOS 6. 17 Question: The third paragraph from 18 the bottom of Mr. Straub's e-mail begins, I've 19 patched a version of Bambi to work with DR-DOS 20 6 and it seems to run Windows 3.1 without 21 difficulty. 22 This same problem may have caused 23 other problems with Windows 3.1 and the swap 24 file under DR-DOS 6. It is possible to make 25 Bambi work, assuming we can come up with a 11773 1 reasonably safe method for detecting DR-DOS 6. 2 What was Mr. Straub telling you here? 3 Answer: Basically, what he's saying 4 here is that if you can figure out you're 5 running on DR-DOS 6, you can basically modify 6 Bambi to work in that environment, or I should 7 say you could modify Bambi to when it detects 8 DR-DOS 6 to work correctly. 9 Question: And he had developed a 10 patched version of Bambi that would allow Bambi 11 to work on DR-DOS 6.0? 12 Answer: That's precisely what he's 13 saying in this e-mail. 14 Question: I was asking you on Exhibit 15 12 about an embedded message from Chuck Straub 16 to you and others at Microsoft. And then you 17 replied to that message where Mr. Straub said 18 it was possible to make Bambi work with a 19 patch. 20 Do you see that at the top of the 21 page? 22 Answer: Yes, I do. 23 Question: And you said there the 24 approach we will take is to detect DR 6 and 25 refuse to load. The error message should be 11774 1 something like invalid device driver interface, 2 close quote. What were you saying there? 3 Answer: I was basically saying that 4 when SmartDrive, a/k/a Bambi, detects DR-DOS 5 that it wouldn't work. 6 Question: Do you know who made the 7 decision to do that? 8 Answer: Yes, I made that decision. 9 Question: Okay. Handing you Exhibit 10 13 to your deposition, which is another series 11 of e-mails provided to us by Microsoft. 12 I want to direct your attention to the 13 message that begins second from the bottom of 14 this exhibit from David Cole to you and Brad 15 Silverberg dated September 30, 1991. Subject, 16 Bambi on DR-DOS 6.0. 17 Do you see that message? 18 Answer: Yes, I do. 19 Question: All right. Do you 20 recognize this as a message that Mr. Cole sent 21 to you? 22 Answer: Yes, I do. 23 Question: In the message, it's just 24 one line, it says, it should say unsupported 25 version of DOS. 11775 1 Was this Mr. Cole's response to your 2 message in Exhibit 12 that we were just 3 referring to? 4 Answer: Yes, I believe it was his 5 response to me, to my message. 6 Question: So he wanted the error 7 message to say instead of saying invalid device 8 driver interface, he wanted it to say 9 unsupported version of DOS; is that correct? 10 Answer: That's what it appears to say 11 here, yes. 12 Question: And Mr. Cole, what was his 13 -- I'm sorry, I think I asked you this before, 14 but can you clarify what -- his position at 15 Microsoft was what? 16 Answer: He was the group program 17 manager on Windows. 18 Question: And did you report to him? 19 Answer: No, we were peers. 20 Question: You worked together with 21 him? 22 Answer: Yes. 23 Question: Do you know whether 24 Mr. Silverberg approved of this decision to put 25 up a warning message and have Bambi not run on 11776 1 DR-DOS? 2 Answer: Yeah, he did. 3 Question: You've been handed Exhibit 4 14 to your deposition, Mr. Barrett. 5 This is another set of e-mails 6 provided to us by Microsoft. 7 I want to direct your attention to the 8 e-mail that begins at the bottom of the first 9 page of the exhibit Bates stamped 1179289. 10 Do you recognize that as an e-mail 11 that you sent to Brad Silverberg, copy to David 12 Cole on September 28th, 1991? 13 Answer: Yes, I do. 14 Question: All right, okay. You say 15 at the bottom of your e-mail message, my 16 proposal is to have Bambi refuse to run on this 17 alias -- actually on this alien operating 18 system. Then you ask for comments. 19 Do you see that? 20 Answer: Yes, I do. 21 Question: Do you recall receiving any 22 comments from the recipients of this e-mail to 23 your proposal? 24 Answer: Not specifically. 25 Question: Do you know whether there 11777 1 was a period of time when the device -- I'm 2 sorry -- when the disk cache driver that 3 started out as Bambi did refuse to run on 4 DR-DOS? 5 Answer: I believe so. I believe so. 6 Question: Mr. Barrett, you've been 7 handed Exhibit 15 to your deposition. 8 This is another series of e-mails 9 produced by Microsoft. I want to direct your 10 attention to the last e-mail on the second 11 page. This is the page that's Bates stamped 12 MS-PCA 1143038. 13 And it's an e-mail message from -- do 14 you recognize this as an e-mail message from 15 Scott Quinn to Bambi and Georgef dated Monday, 16 September 30th, 1991? 17 Answer: Yes. 18 Question: The subject is Bambi 19 version 3.5. 20 Do you see that? 21 Answer: Uh-huh. 22 Question: And in his e-mail message 23 Mr. Quinn says, Bambi version 3.5 has passed 24 developer testing. The primary change fixes a 25 major problem with accessing logical units on 11778 1 external hard disks. 2 Then he goes on to say, also DR-DOS is 3 detected (needs testing) and Bambi refuses to 4 load. 5 Is this consistent with your 6 recollection of the status of Bambi at the time 7 that it passed developer testing? 8 Answer: Yes, this is. Yes, this was 9 consistent. 10 Question: And the e-mail alias here 11 that's shown in the to line Bambi, was that a 12 group of Bambi developers? 13 Answer: Yes, that was the group of 14 people developing Bambi. 15 Question: What happened to Bambi 16 after it passed developer testing? I'm trying 17 to find out what the next stage of development 18 was. 19 Answer: That would go into the QA 20 group for more full testing. Developer testing 21 is simply the developers that built it ensure 22 that they believe that it was going enough 23 quality to try to test to the next level of 24 quality. 25 Question: You've been handed Exhibit 11779 1 16 to your deposition, which is another 2 conglomeration of e-mail provided by Microsoft. 3 I'd like to refer you to the e-mail on 4 the first page, Bates stamped MS-PCA 1179461. 5 Do you recognize this as an e-mail 6 from Chuck Straub to you and Scott Quinn dated 7 February 23, 1993? 8 Answer: Okay. Yep. 9 Question: Mr. Straub says, this is 10 the mail which announces -- which announced the 11 fixing of the, quote, bug, close quote, which 12 had previously prevented DR-DOS 6 from working. 13 As of this point, the DR-DOS 6 check could have 14 been removed. 15 And then embedded in his e-mail is a 16 copy of an e-mail that Mr. Quinn sent out to 17 the Bambi group on October 16, 1991. 18 Do you see that? 19 Answer: Yes, I do. 20 Question: Do you know whether the 21 DR-DOS 6 check was removed after the bug which 22 prevented DR-DOS 6 from working with Bambi had 23 it been fixed? 24 Answer: I'm not really sure. It does 25 not -- this does not ring a bell one way or the 11780 1 other. 2 Question: All right. Do you know 3 what the bug was that appears -- it refers to a 4 bug in quotes. Do you know what bug he was 5 referring to? 6 Answer: Well, I think it referred to 7 a bug -- well, it referred to the way that 8 Digital Research was handling the greater than 9 32 megabyte partitions. 10 Question: We covered that earlier? 11 Answer: Yes. 12 Question: You've been handed Exhibit 13 17 to your deposition, which is more e-mail 14 provided by Microsoft in this litigation. This 15 is Bates stamped MS-PCA 1148484. 16 Do you recognize this as an e-mail 17 from David Cole to you and an Karlst dated 18 September 30, 1991? 19 Answer: Yes. 20 Question: The subject line says 21 supported DOSes. 22 And Mr. Cole begins his e-mail by 23 saying, it's pretty clear we need to make sure 24 Windows 3.1 only runs on top of MS-DOS or an 25 OEM version of it. 11781 1 Why was it that Microsoft needed to 2 make sure Windows 3.1 only ran on top of MS-DOS 3 or its OEM version? 4 Answer: Well, this is -- this goes 5 back to some of the other documents we looked 6 at. 7 Pretty much Bill Gates wanted to make 8 sure that we put our competitive operating 9 systems at a disadvantage, and this is one way 10 that we would do it. 11 Question: Did you hear that, 12 Mr. Gates say that in words or substance? 13 Answer: Yes, I personally sat in 14 meetings where he said that. 15 Question: All right. Mr. Cole goes 16 on to say, I checked with legal and they are 17 working up some text we are supposed to display 18 if someone tries to set up or run Windows on an 19 alien operating system. 20 We are supposed to give the user the 21 option of continuing after the warning. 22 However, we should surely crash at some point 23 shortly later. 24 Were you aware of any plan to cause 25 Windows 3.1 to crash after a warning message 11782 1 came up if it was being used in conjunction 2 with an operating system other than MS-DOS? 3 Answer: Yeah, I was aware of this. 4 Question: Mr. Cole goes on to say, 5 now to the point of this mail. How shall we 6 proceed on the issue of making sure Windows 3.1 7 requires MS-DOS? We need to have some pretty 8 fancy internal checks to make sure we are on 9 the right one. 10 Maybe there are several very 11 sophisticated checks so the competitors get put 12 on a treadmill. 13 Do you know what that reference is to 14 putting the competitors on a treadmill, what 15 that refers to? 16 Answer: Certainly. The whole idea, 17 in general, in competition is to cause your 18 competitors to constantly be playing catchup. 19 And that's why the reference to treadmill, 20 running fast to stay in place. 21 Question: Was putting the competitors 22 on a treadmill, was that a phrase that you 23 heard occasionally at Microsoft? 24 Answer: More than occasionally. 25 Yeah, it was a very common phrase. I used it 11783 1 myself quite a bit. 2 Question: All right. Do you 3 understand how this suggestion by Mr. Cole in 4 this e-mail would accomplish putting 5 competitors on a treadmill? 6 Answer: Certainly. The whole idea is 7 -- well, Digital Research was very, very good 8 at modifying their operating system to overcome 9 any barriers we have put in front of them. 10 In fact, the bug -- the quote-unquote 11 bug from a previous document is an example 12 where they figured out what we were looking for 13 and then they fixed it so that they looked more 14 like MS-DOS than they did before. 15 Basically, what you're trying to do is 16 cause them to continually have to rerelease new 17 versions of their operating system in order to 18 stay compatible. And that's precisely what 19 this was talking about. 20 Question: You've been handed Exhibit 21 19 to your deposition. This is another set of 22 e-mails provided to us in discovery from 23 Microsoft. 24 I want to direct your attention to the 25 e-mail that begins at the bottom of the page 11784 1 from a Jenk to Cliffg and Scott Quinn dated -- 2 and with a copy to the DOS development group, 3 dated September 30, 1991. 4 Answer: Okay. 5 Question: And embedded in that is -- 6 or it's replying to a message from Cliffg where 7 he discusses, quote, the official way to detect 8 DR-DOS. 9 Do you see that? 10 Answer: Yes, I do. 11 Question: Okay. And then he 12 enumerates three steps. 13 Do you recognize those as steps that 14 would allow the program to detect the presence 15 of DR-DOS? 16 Answer: That certainly looks like it. 17 Question: Okay. Do you recall 18 receiving this e-mail? 19 Answer: No, I don't recall receiving 20 this e-mail. 21 Question: Cliffg refers to, quote, 22 the official way to detect DR-DOS. 23 Is that a reference to -- well, what 24 is the official way to detect DR-DOS? What 25 does that mean? 11785 1 Answer: Basically, I think it simply 2 means the Digital Research -- Digital Research 3 recommended way of detecting Digital Research 4 DOS. 5 Question: You've been handed Exhibit 6 20 to your deposition, which is another letter 7 that was provided to us in discovery by 8 Microsoft Bates stamped MS-PCA 1143151. 9 And it appears to be a letter from a 10 Bradley Kerth, K-e-r-t-h, senior technical 11 support engineer at Digital Research, to a 12 Roger Sour or director of Windows development 13 at Microsoft dated October 24, 1991. 14 Answer: Yes. 15 Question: Did this letter ever come 16 to your attention? 17 Answer: Yes, it did. 18 Question: All right. And do you know 19 who Roger Sour is, S-o-u-r? 20 Answer: I don't know anyone named 21 Roger Sour. 22 Question: Tell me the circumstances 23 of how this came to your attention and what 24 happened, if you can recall. 25 Answer: Well, this came to me because 11786 1 of the director of Windows development. 2 Question: That was you? 3 Answer: Yes, that was -- well, yeah, 4 I was the closest thing to it. 5 Question: And what was the issue that 6 was -- that came to light as a result of this 7 letter? 8 Answer: Well, I'm not sure how to 9 answer your question, to be honest. 10 This came to -- there was some problem 11 that somebody obviously called Digital Research 12 about and wanted to get more information. 13 Question: All right. Well, Mr. Kerth 14 says in his letter, it has come to my attention 15 that on September 30th, 1991, you, meaning 16 Mr. Sour, had contacted the Digital Research 17 technical support department for assistance 18 with DR-DOS 6.0. 19 You provided the serial number and 20 then he recites the serial number to our 21 technical support analyst, Andrew Dyson, 22 D-y-s-o-n, and proceeded to ask if there was a 23 way for a program to detect if it is running 24 under that operating system. 25 While this information is not 11787 1 generally handed out, we try to maintain a very 2 cooperative policy towards software 3 manufacturers. In following that policy, 4 Andrew described the technique to do so. 5 Do you know if anyone at Microsoft 6 made the phone call that's described here? 7 Answer: I have no personal knowledge 8 of anyone actually making that call. 9 Question: All right. Was any 10 investigation made to see whether that, at 11 Microsoft, made to determine whether this call 12 was made? 13 Answer: You use the word 14 investigation, probably overstates it, but, 15 yeah, I did some asking. 16 Question: You looked into it? 17 Answer: I looked into it, yes. 18 Question: And what did you find out? 19 Answer: I found nothing. 20 Question: Okay. Was there -- did it 21 ever come to light that this Cliffg made the 22 phone call that's described here? 23 Answer: I -- no. At least never came 24 to light to me. 25 Question: Okay. Mr. Kerth's letter 11788 1 goes on to say, when Andrew asked why you 2 needed the information, you indicated that you 3 were developing portions of the new cache 4 software for the future Windows 3.1, and you 5 had found a, quote, problem in the DR-DOS 6.0 6 memory control blocks (MCB), close quote. 7 Now, that reference to the cache 8 software, was that the Bambi disk cache driver? 9 Answer: Well, it certainly sounds 10 like an accurate description of it. 11 Question: Okay. What are the memory 12 control blocks? What does that refer to? 13 Answer: I honestly don't know 14 specifically what the intent there is. I'd 15 only be speculating. 16 Question: Earlier when we were 17 talking about the Bambi, the e-mails concerning 18 the Bambi development, you made reference to 19 the fact that there was a difference between 20 the way DR-DOS and MS-DOS handled something 21 that affected this disk cache driver. 22 Does that have -- what you were 23 referring to earlier, does that have something 24 to do with the memory control blocks? 25 Answer: That would be a very logical 11789 1 assumption. 2 THE WITNESS: Could we go -- could I 3 confer with counsel for just a second? 4 Question: Sure. Absolutely. 5 Mr. Barrett, during the break, you 6 said you wanted to clarify something. 7 Answer: Oh, yes. I wanted to clarify 8 Cliffg was Cliff Garrett. He was a temporary 9 employee, G-a-r-r-e-t-t, I believe. And he was 10 the -- he basically ran the -- he definitely 11 ran the SmartDrive beta test program. I 12 believe he also was part of the DOS 6 beta test 13 program. 14 Question: All right. And just to 15 connect it all up, the SmartDrive product was, 16 the code name for that during development was 17 Bambi? 18 Answer: That is correct. 19 Question: Thanks for that 20 clarification. I appreciate that. 21 Continuing with Exhibit 20. I want to 22 refer you to the middle of the second paragraph 23 where Mr. Kerth says, as I understand it, your 24 goal is to identify the presence of DR-DOS 6.0 25 so that your software will terminate itself 11790 1 after warning the end user that, quote, an 2 unsupported DOS, close quote, is being used. 3 Usually when a software manufacturer 4 feels that something in our operating system is 5 preventing their application from running well, 6 that company works with us to resolve the 7 actual perceived or potential conflicts. 8 Was that your experience in the 9 industry, that when one software manufacturer 10 found some conflict with another software, that 11 the two companies would generally try and work 12 together to resolve those conflicts? 13 Answer: I wouldn't say it's the rule, 14 but it's quite common for two companies to work 15 to eliminate conflicts or other problems 16 between the two pieces of software that they 17 have. 18 Question: All right. And generally 19 when that is done, what's the rationale for 20 that? Why would the two companies work 21 together to eliminate any conflicts between 22 their two software products? 23 Answer: Typically the goal, good 24 companies care about their customers' 25 experience. And so they want to make sure that 11791 1 the customers have a good experience regardless 2 of what software they're running. 3 Question: Mr. Barrett, you've been 4 handed Exhibit 21 to your deposition. And do 5 you recognize this as a letter that you sent to 6 Mr. Kerth at Digital Research on November 1, 7 1991? 8 Answer: Yes. 9 Question: All right. You say, I am 10 the development manager for Microsoft Windows 11 3.1. As such, I was given the letter sent by 12 you and addressed to, quote, Roger Sour, close 13 quote, or director of Windows development. 14 This was a very odd piece of mail to 15 receive, that there is no one at Microsoft by 16 the name of Roger Sour. 17 Further, whoever this Mr. Sour is, he 18 certainly does not speak for Microsoft. 19 Perhaps you may have been the victim of a 20 prank. 21 Did you ever directly ask Cliff 22 Garrett whether he had sent this -- or whether 23 he had made this call under the name of Roger 24 Sour? 25 Answer: No. 11792 1 Question: You never asked him that? 2 Answer: Sorry. No, I did not. 3 Question: The detection code that 4 Mr. Garrett obtained from DRI, Digital 5 Research, is that used to detect DR-DOS? Was 6 it used in Microsoft's products? 7 Answer: I honestly don't recall. At 8 least at this moment. 9 Question: You've been handed Exhibit 10 22 to your deposition. 11 Do you recognize this as a letter that 12 Mr. Kerth, Bradley Kerth from Digital Research 13 sent to you on November 14th, 1991? 14 Answer: Yes, I do. 15 Question: At the end of his letter 16 Mr. Kerth says, if at any point Microsoft 17 detects a compatibility problem or potential 18 problem between Microsoft Windows 3.1 and 19 DR-DOS 6.0, please call myself or Brian -- I'm 20 sorry, Byron Tomingas immediately to discuss 21 the situation. 22 Did you or anyone else that you know 23 of at Microsoft call Mr. Kerth or Mr. Tomingas 24 at Digital Research to try and resolve the 25 conflict between the Bambi product and DR-DOS? 11793 1 Answer: No. 2 Question: Before you look at Exhibit 3 23, I just want to go back to 22 for a moment. 4 In Exhibit 22 Mr. Kerth says, the 5 first paragraph, I take it from your letter 6 that you do not know who Roger Sour is. That 7 there is no problem between Windows 3.1 and the 8 DR-DOS 6.0 memory control blocks, and then 9 Microsoft has no intention of identifying the 10 presence of DR-DOS 6.0 so that Windows 3.1 will 11 terminate itself or otherwise inconvenience the 12 end user because the end user is running DR-DOS 13 instead of MS-DOS. 14 He then asks you to let him know 15 immediately if his understanding is incorrect. 16 Do you see that? 17 Answer: Yes. 18 Question: Now, I want you to take a 19 look at Exhibit 23. 20 Was -- do you recognize Exhibit 23 as 21 a letter that you wrote to Mr. Kerth at Digital 22 Research on December 2nd, 1991, replying to his 23 letter of November 14th of Exhibit 22? 24 Answer: Yes, I do. 25 Question: And in it you say, thank 11794 1 you for your letter dated November 14, 1991. 2 Because Windows 3.1 is an unreleased product, I 3 cannot disclose details as to specific features 4 or functions in that product. 5 Was that a policy at that time at 6 Microsoft, that you did not release details as 7 to features or functions in unreleased 8 products? 9 Answer: No. 10 Question: Why were you telling 11 Mr. Kerth that you would not release the 12 details as to the specific features or 13 functions of the product? 14 Answer: Because -- well, because we 15 simply didn't want to go there. 16 Question: You didn't want to let him 17 know that you were going to continue to detect 18 DR-DOS? 19 Answer: That's correct. We did not 20 want to -- we didn't want to say anything that 21 could be perceived as tying our hands. 22 Question: Did you ever determine 23 where Digital Research was getting their 24 information about what Microsoft was doing with 25 the detecting Windows -- detecting the DR-DOS 11795 1 code? 2 Answer: I'm not 100 percent clear on 3 your question. Perhaps you'd like to rephrase 4 it. 5 Question: Yeah. Mr. Cole ends his 6 message by asking, you know, where the hell is 7 DRI getting their information. 8 Do you see that? 9 Answer: Yes, I do. 10 Question: All right. Did you or 11 anyone else that you know of at Microsoft make 12 an investigation to determine how Digital 13 Research was getting information about what 14 Microsoft was doing in this regard? 15 Answer: Yeah. I actually spent not a 16 lot of time, but I spent some time trying to 17 determine who this person might be, if there 18 was indeed a leak within Microsoft. 19 I went so far as to talk to the phone 20 people to see if there was a way to check phone 21 records, and also talked to a number of people 22 on the team and directly asked them if they 23 knew anything about it, of which they all said 24 no. 25 Question: Okay. So you never did 11796 1 find out how Digital Research was getting its 2 information? 3 Answer: No, did not. 4 Question: All right. As part of 5 looking into this, did you determine -- or did 6 you find out that Cliff Garrett had been in 7 touch with Digital Research to get the 8 detection code? 9 Answer: By the evidence document you 10 showed me earlier because I don't recall -- I 11 don't recall that specific situation, and yet I 12 would have thought that he would be the likely 13 candidate for this had I put two and two 14 together. 15 Question: Mr. Barrett, you've been 16 handed Exhibit 25 to your deposition, which is 17 a group of e-mails produced by Microsoft in 18 this case. 19 I'd like to direct your attention to 20 the second e-mail up from the bottom, the one 21 that appears to be from you. 22 Do you see that? 23 Answer: Yes, I do. 24 Question: Do you recognize that as an 25 e-mail message that you sent to Brad 11797 1 Silverberg, a Buckf and David Cole on November 2 6, 1991? 3 Answer: Yes. 4 Question: In that e-mail you say, 5 this is -- the subject line is re: FTC and 6 DR-DOS. 7 And was this a reply to the message 8 that you received from David Cole that we just 9 looked at in Exhibit 24? 10 Answer: Yes, it is. 11 Question: In the e-mail you say, they 12 are getting very accurate information from an 13 internal source. I don't want a witch hunt, 14 but this is out of control. 15 Buck, this is pretty important. I 16 talked to Brad and he is in agreement that we 17 should ferret out the leak. 18 Is this -- are you talking about the 19 leak to Digital Research? 20 Answer: Yes, that's precisely what 21 I'm referring to. 22 Question: At the end you say, I have 23 suspicions about one specific person, but don't 24 want to say in e-mail. 25 Who at that time did you suspect was 11798 1 leaking information to Digital Research? 2 Answer: Well, there was a -- and I 3 don't remember the guy's name, but there was a 4 difficult person, like another support person 5 on the beta program who I thought was -- I just 6 thought that they were disgruntled. I don't 7 believe that it was that person, though, and 8 I'm sorry, I don't recall their name. 9 Question: Okay. You've been handed 10 Exhibit 26, Mr. Barrett, which is yet another 11 set of e-mails produced to us in this case by 12 Microsoft. 13 I want to direct your attention to the 14 e-mail at the very top of the page. This is 15 Bates stamped MS-PCA 1179419. 16 Answer: Okay. The top piece of the 17 e-mail? 18 Question: The very top of the e-mail. 19 Answer: Okay, I see it. 20 Question: Do you recognize this as an 21 e-mail from Brad Silverberg to David Cole and 22 yourself on -- dated October 29, 1991? 23 Answer: Yep. Yes. 24 Question: The subject is DR and 25 Windows 3.1. 11799 1 And you'll see that embedded in 2 Mr. Silverberg's e-mail is an e-mail from Richf 3 to Brad Chase and Brad Silverberg. 4 Do you see that? 5 Answer: Yes. 6 Question: And who's Richf? Is that 7 Richard Fade? Do you know who Richardf is? 8 Answer: I believe Richardf is Richard 9 Fade. 10 Question: All right. In his e-mail 11 Mr. Fade says, I loaded DR 6 on my own system 12 about a week ago and have been using it since. 13 And then he goes on to talk about some 14 details of what he experienced while running 15 it. 16 Answer: Uh-huh. 17 Question: But then he concludes his 18 e-mail by saying, in short, I haven't seen any 19 basic kernel incompatibilities. 20 Are you able to tell what he's talking 21 about there? Incompatibilities between what 22 and what? 23 Answer: Incompatibilities between -- 24 the incompatibilities are between Windows 3.1 25 and DR-DOS. 11800 1 Question: Okay. And he wasn't able 2 to find any kernel incompatibilities? 3 Answer: Yes, although Richard Fade 4 wouldn't know a kernel if it hit him. He's 5 basically saying Windows 3.1. 6 Question: Windows 3.1 what? 7 Answer: With DR-DOS. 8 Question: Okay. And then Brad 9 Silverberg says in the e-mail that Mr. Fade's 10 e-mail is embedded in, he says, looks like 11 DR-DOS works with Windows 3.1. 12 Was that a conclusion that your team 13 reached at the end of October, that DR-DOS 14 actually did work with Windows 3.1 at that 15 point in time? 16 Answer: Yeah, I believe so. I think 17 they had made some changes or something. 18 I do want to point out that there are 19 a set of incompatibilities that still existed, 20 particularly using their EMM 386, which is 21 their enhanced memory manager, which was a way 22 of providing a set of features to DOS 23 applications so that they had more memory. 24 Question: All right. And that's what 25 he's talking about, what Mr. Fade is talking 11801 1 about in his first paragraph of the embedded 2 e-mail? 3 Answer: Yes, that's correct. 4 Question: Are you able to tell or did 5 you know how that particular incompatibility, 6 how much of an impact that had on being able to 7 run DR-DOS and Windows together? 8 Answer: It wasn't a big impact. 9 (Whereupon, the playing of the video 10 adjourned.) 11 THE COURT: We'll take our recess 12 right now. 13 Remember the admonition previously 14 given. 15 We'll be in recess for about ten 16 minutes. Leave your notebooks here. 17 Thank you. 18 (A recess was taken from 9:47 a.m. 19 to 10:02 a.m.) 20 THE COURT: Everyone else may be 21 seated. 22 You may continue. 23 (Whereupon, the following video was 24 played to the jury.) 25 Question: Exhibit 27 is additional 11802 1 e-mail provided to us from Microsoft, Bates 2 stamped MS-PCA 1143105. 3 And I'd like to refer you -- there are 4 two e-mails here. I'd like to refer you to the 5 bottom one. 6 Do you recognize this as an e-mail 7 from Brad Silverberg to Brad Chase, Steve 8 Ballmer and Tom Lennon? 9 Answer: I'm sorry. What was the 10 question again? 11 Question: Let me ask you a question 12 about the bottom e-mail. 13 Answer: Okay. 14 Question: You see it appears to be an 15 e-mail from Brad Silverberg to Brad Chase and 16 Steve Ballmer and Tom Lennon? 17 Do you see that? 18 Answer: Yes, I do. 19 Question: All right. And it's dated 20 October 8, 1991. Subject, thought you'd 21 appreciate. 22 And Mr. Silverberg says, the following 23 excerpt from Janineh's Windows 3.1 beta report. 24 Do you know who Janineh was? 25 Answer: Yes, I know who Janineh was, 11803 1 although I don't recall her last name. 2 Question: All right. What was her 3 responsibility or position at Microsoft? 4 Answer: She was a program manager in 5 charge of the 3.1 beta program. 6 Question: All right. And I think 7 you've mentioned before that you also worked on 8 the 3.1 beta program; is that correct? 9 Answer: Yes. 10 Question: Did you work -- did you 11 have any input to these beta reports that's 12 referred to here, the Windows 3.1 beta report? 13 Answer: Yeah, I'm sure I had some 14 input. In general, it was more reporting about 15 what was going on in the beta program, status 16 and that sort of thing. 17 Question: Now, Janineh, there's a 18 quote from her 3.1 beta report that's embedded 19 in the e-mail. 20 Do you see that at the bottom? 21 Answer: Yes, I do. 22 Question: And it says weekly -- under 23 the heading weekly activity in the report, it 24 says, Windows 3.1 doesn't run with DR-DOS. 25 And Janineh reports, we sent one 11804 1 person MS-DOS 5.0 to use and Randym is working 2 with another large account. This may stop them 3 from going to DR-DOS 6.0. 4 Do you know who Randym was? 5 Answer: No, I don't recall at this 6 moment. 7 Question: Do you know whether 8 Microsoft's sales and marketing people were 9 using the alleged incompatibilities between 10 DR-DOS and Windows to try and stop customers 11 from using DR-DOS? 12 Answer: I don't have any direct 13 personal knowledge of that. 14 Question: Did you learn of it 15 indirectly? 16 Answer: It was common knowledge that 17 the differences were being used by the sales 18 groups. 19 Question: Did people talk about that 20 fact at Microsoft? 21 Answer: It was fairly common, fairly 22 common discussion point. 23 Question: You've been handed Exhibit 24 28 to your deposition. 25 Do you recognize this as an e-mail 11805 1 message from David Cole to Brad Silverberg with 2 a copy to you dated November 4, 1991? 3 Answer: Yes. 4 Question: In this Mr. Cole talks 5 about where things stand on the wording of the 6 warning, what he refers to as the warning. 7 Do you know whether either of these 8 versions of the warning were actually 9 implemented in Windows 3.1? 10 Answer: I don't recall if these 11 specific ones were put in, but I know that 12 something very similar to the bottom one was 13 put into the beta release of 3.1. 14 Question: You've been handed Exhibit 15 29 to your deposition. 16 Do you recognize this as an e-mail 17 message from David Cole to Royha and Timbr with 18 a copy to you dated November 8th, 1991? 19 Answer: Yes, I do. 20 Question: Do you recall who Royha 21 was? 22 Answer: No, I don't. 23 Question: Or Timbr? 24 Answer: No, I don't. 25 Question: You mentioned earlier that 11806 1 -- well, this e-mail begins, in order to serve 2 our Windows customers properly, we are going to 3 warn them if they set up Windows or run MSD on 4 systems not using MS-DOS. 5 MSD refers to? 6 Answer: Microsoft diagnostics, I 7 believe. 8 Question: Okay. Mr. Cole goes on to 9 say, Aaronr -- is that Aaron Reynolds? 10 Answer: That is correct. 11 Question: Aaron Reynolds is 12 developing a piece of code which checks for the 13 absence of MS-DOS. Phil Barrett is working 14 with him. 15 Now, what was the code that Aaron 16 Reynolds was working on at this point in time? 17 Answer: That was a piece of code to 18 detect whether or not we're running on DR-DOS. 19 Question: And what work were you 20 doing with Mr. Reynolds in connection with 21 that? 22 Answer: Well -- 23 Question: Mr. Barrett, you've been 24 handed Exhibit 32 to your deposition. 25 Do you recognize this as an e-mail 11807 1 message from Aaron Reynolds to Karlst, yourself 2 and Sandyps? 3 Answer: Yes, I do. 4 Question: And it's dated February 24, 5 1993. 6 In the message, Mr. Reynolds begins by 7 saying, yesterday it was decided to do this 8 detection and display the message if 9 appropriate every time we run Windows. Aaron 10 has some code that you add to win.com. 11 Of course, the subject of this is 12 MS-DOS detection-hot job for you. 13 In December of 1991, was it, in fact, 14 decided that the Windows detection warning 15 message would be -- would be displayed every 16 time Windows is run? 17 Answer: Yes, it was. 18 I do want to point out that the date 19 on this is actually Friday, December 6th, 1991, 20 not Wednesday, February 24th, 1993. 21 Question: Thanks for making that 22 clarification. 23 You've been handed Exhibit 33, which 24 is another series of e-mail messages produced 25 by Microsoft in this case. 11808 1 I'd like to begin by asking you about 2 the message embedded in the first message. 3 You'll see that appears on page 2, and 4 it appears to be from Chuckst. Is that Chuck 5 Straub? 6 Answer: Yes, that's Chuck Straub. 7 Question: To you dated December 16, 8 1991, is that correct? 9 Answer: Yes. 10 Question: Mr. Straub in his e-mail 11 message to you says, sorry I wasn't able to get 12 these things done this evening. I had a prior 13 commitment, and he lists three things, one of 14 which was itemized Number 2, MS-DOS detection 15 code change. 16 Aaronr already came by my office with 17 a new object module which we checked in. And 18 then he goes on to talk about some details 19 about the MS-DOS detection code change. 20 And then what I want to ask you about 21 is whether this detection code was important 22 for the beta release of Windows 3.1? 23 Answer: Yeah, it was absolutely -- 24 yeah, it was absolutely considered important at 25 that time. 11809 1 Question: And was that the substance 2 of your message above where you said, business 3 as usual ain't good enough. We've got to 4 absolutely jump on this stuff. It can't wait 5 even a couple of hours, let alone a half a day. 6 I want you to make this your number one 7 priority. This beta release is critical to the 8 success of Windows 3.1. 9 Was that what you were trying to 10 convey to him? 11 Answer: Yes, that's exactly what I 12 wrote. 13 Question: Why the rush here? Was 14 there some urgency to getting this beta release 15 out? 16 Answer: Well, there were two things 17 going on. One was, yes, it was -- we had as a 18 management team decided that we were going to 19 do this. 20 Secondly, Chuck, who wasn't obvious -- 21 not terribly obvious in this mail, was a 22 performance problem and he was not doing his 23 job and he didn't come in. And so there were 24 two purposes of this mail. 25 One is to make it clear that we were 11810 1 going to put the MS-DOS detection code in the 2 beta and we were going to get it done. 3 And, secondly, was putting him on 4 notice that he was not doing his job, and that 5 he was on thin ice. 6 And this is called building a paper 7 trail when you have a performance problem. 8 Question: You've been handed Exhibit 9 35 to your deposition. 10 Mr. Barrett, do you recognize this as 11 an e-mail message from David Cole to Aaron 12 Reynolds, yourself, and others at Microsoft 13 dated February 11, 1992? 14 Answer: Yes, it is. 15 Question: Mr. Cole begins his e-mail 16 message by saying, after a long round of 17 debates, the executives have decided that no 18 message should be displayed when we fail to 19 detect MS-DOS. 20 The PR impact of showing a message 21 would be great and would consume much of our 22 time just dealing with the reaction from the 23 press and others. 24 First of all, were you involved in the 25 decision to not make use of the warning message 11811 1 in the final released product of Windows 3.1? 2 Answer: Yes, I was. 3 Question: And can you tell me how 4 that decision came about? 5 Answer: Well, I wasn't subject to all 6 of the discussions that Steve Ballmer and Brad 7 Silverberg had, but, basically, Steve Ballmer, 8 and I personally heard him say it, he just 9 thought it was a stupid thing to do. And that 10 went on for probably two weeks back and forth. 11 We finally reached consensus to not do it. 12 Question: All right. Now, at the 13 time that this decision was made, had the 14 Windows 3.1 beta release gone out to the beta 15 test sites? 16 Answer: Yes, it had. 17 Question: And the warning message did 18 appear in those beta copies; is that correct? 19 Answer: Yes, it did. 20 Question: Mr. Cole ends his message 21 by saying, I want to thank everyone for their 22 effort on this in putting up with the changes. 23 It was a close call on what to do. 24 Then he says, we will have a readme 25 item that explains we test it only on MS-DOS. 11812 1 What was readme? 2 Answer: Readme is a file, usually the 3 name of the file is read me.text and read 4 me.doc, which almost all software products, 5 when they ship, they include a readme file that 6 has very late breaking information that might 7 not have made it into the manual or something 8 like that. 9 The readme item specifically refers 10 to, you know, one item. They're typically 11 numbered or lettered, and it refers to one 12 specifically numbered or lettered item that 13 would be in there. 14 Question: All right. And what is the 15 purpose of a readme file? How does the end 16 user make use of that? 17 Answer: Typically, it's to 18 communicate very late breaking information or 19 incompatibilities or problems that have been 20 found and gives the users ways to work around 21 those problems. Basically, it's a 22 communication vehicle for very, very recent 23 information that you hadn't been able to 24 formalize. 25 Question: Exhibit 36 has just been 11813 1 handed to you, and this is, again, a series of 2 e-mail messages that was produced to us in this 3 litigation by Microsoft. 4 Let me ask you about the top e-mail 5 message. 6 Do you recognize this as an e-mail 7 message sent by Richab to Brad Silverberg, 8 David Cole, and yourself dated Friday, November 9 30, 1990? 10 Answer: Yes, I do. 11 Question: And was this a response to 12 the e-mail messages that appear below it? 13 Answer: Yes, I believe it is. 14 Question: In the e-mail message at 15 the bottom of the page from it appears to be 16 Amitc. Do you know who that is? 17 Answer: Yes, that's Amit Chaterjee. 18 He's a software engineer on Windows. 19 Question: To a Danq. Who's Danq, do 20 you know? 21 Answer: Doesn't -- doesn't ring a 22 bell. 23 Question: Dated November 27, 1990. 24 Amit Chaterjee says, I have been told 25 that I will not need to answer support 11814 1 questions on the VxD anymore. 2 That is virtual device driver? 3 Answer: Yes, that's correct. 4 Question: Some people still have my 5 number and they call me. I told them that we 6 now have a different group handling support 7 issues and that Tim is the person to talk to. 8 However, I think it has been decided 9 that Digital Research will not be supported. 10 Rich Abel should have the list of vendors who 11 we do not want to support. 12 Let me ask you, why did Microsoft 13 decide not to support Digital Research? 14 Answer: Because they were a 15 competitor. 16 Question: He makes reference to a 17 list of vendors who we do not want to support. 18 Are you familiar with that list? 19 Answer: I'm familiar that it existed. 20 Question: Was that sometimes referred 21 to as the beta blacklist? 22 Answer: I never heard it called that 23 when I was at Microsoft. 24 Question: What was this list called? 25 Answer: The list of companies we 11815 1 didn't want to support. I don't believe it 2 actually had a specific name. 3 Question: All right. Do you know who 4 was in charge of keeping that list, maintaining 5 it? 6 Answer: I honestly don't know. I'm 7 not sure that anyone really was. I think it 8 existed more in the form of, you know, word of 9 mouth and sort of common understanding. 10 Question: Do you know how long the 11 list was? 12 Answer: At this point I can only 13 recall of one company that would have been on a 14 list, if it existed in physical form, and 15 that's Digital Research. 16 Question: All right. It goes on say, 17 quite some time back DRI was sent a very early 18 version of VxD, which, again, you said is 19 virtual device driver. 20 I don't know what to tell them. I 21 guess we must somehow politely let them know 22 that we don't want to support them. I don't 23 feel very comfortable in this situation and 24 would not want to deal with Digital myself. 25 Do you know why Microsoft gave Digital 11816 1 Research a copy of the digital device driver if 2 Microsoft was not going to support Digital 3 Research? 4 Answer: I believe that it was a 5 mistake that it got sent to them in the first 6 place. 7 Question: I see. 8 Answer: I know that it was a mistake. 9 Question: All right. You've been 10 handed Exhibit 38 to your deposition, 11 Mr. Barrett. 12 I want to refer you only to the e-mail 13 at the bottom of the first page. 14 I'm sorry, I take that back. 15 There's some embedded e-mails, but 16 it's an e-mail about a third of the way up from 17 the bottom from Brad Silverberg to a number of 18 people, including yourself, on July 26, 1991. 19 Answer: Yes. 20 Question: Subject re: Digital 21 Research. 22 Do you recognize this as an e-mail 23 that Mr. Silverberg sent to you and others on 24 July 26, 1991? 25 Answer: Yes, I do. 11817 1 Question: Mr. Silverberg asked the 2 question how does DRI have Windows 3.1 in the 3 first place? They are on the beta blacklist. 4 I was just -- I wanted to show this to 5 you to ask you whether this refreshes your 6 recollection as to whether there was something 7 called the beta blacklist at Microsoft? 8 Answer: I guess this is pretty black 9 and white. I guess I never recalled it as the 10 beta blacklist, but, obviously, Brad 11 Silverberg's words speak for themselves. 12 Question: Was there someone who had 13 authority to determine who would be denied 14 support, who would appear on this, 15 Mr. Silverberg calls it the beta blacklist? 16 Answer: Sorry. Yes, Silverberg. 17 Question: Mr. Silverberg made the 18 determination as to who would be denied 19 support? 20 Answer: Yeah. 21 Question: Just to satisfy counsel, 22 were you personally present at a meeting where 23 Mr. Silverberg talked about denying Digital 24 Research the 3.1 beta? 25 Answer: Yes, I was. 11818 1 Question: And what did he say? 2 Answer: He said that there -- well, I 3 won't say precisely the way he said it, but he 4 basically said that DR would never receive a 5 copy of the 3.1 beta. They're a competitor. 6 Question: Do you recall precisely how 7 he put that? 8 Answer: Yes, I do. 9 Question: And would you tell us, even 10 though if -- even if it is -- 11 Answer: He said there's no fucking 12 way in hell that Digital Research will ever 13 receive a copy of the 3.1 beta. They're a 14 goddamn competitor. 15 Question: Thank you. 16 You never saw a physical list of 17 people who were on this beta blacklist? 18 Answer: No, I never personally saw 19 one. 20 Question: You've been handed Exhibit 21 39 to your deposition, Mr. Barrett. 22 This is another series of e-mail 23 messages provided by Microsoft in this case. 24 You'll see in the middle of the page 25 is an e-mail message that appears to be from 11819 1 Brad Silverberg to a Jancl and a Pattys, 2 subject Digital Research. 3 Do you see that? 4 Answer: Uh-huh. Yes. 5 Question: And it starts by saying, we 6 should not be providing Digital Research any 7 assistance getting their operating system to 8 work with our software. Our software supports 9 MS-DOS, not DR-DOS. It's completely up to them 10 to figure out how to resolve any problems that 11 may occur. 12 Did that statement there reflect 13 statements, other statements that you heard 14 from Mr. Silverberg with respect to supporting 15 Digital Research? 16 Answer: I think this is a very polite 17 way of saying we don't support our competitors. 18 It's their problem. 19 Question: And then above that is -- 20 at the very top of the page is an e-mail 21 message. 22 Do you recognize that as an e-mail 23 message sent by Mr. Silverberg to you and 24 others at Microsoft on Monday, July 29th, 1991? 25 Answer: Yes. 11820 1 Question: There Mr. Silverberg says, 2 you bet. DRI is on the DOS/Win blacklist. 3 Nothing gets shipped from this BU without the 4 blacklist being checked. 5 First of all, what was the reference 6 to a BU? 7 Answer: That's business unit. The 8 Windows and DOS business unit. 9 Question: All right. Now, he makes a 10 reference to the blacklist being checked. Was 11 there -- what was the process, if you know, for 12 checking the beta blacklist? 13 Answer: I don't think there was a 14 process. 15 Question: Okay. And then embedded in 16 that message is a message about a third of the 17 way down the page. 18 Do you recognize that as an e-mail 19 message from Steve Ballmer to Brad Silverberg, 20 yourself, and others at Microsoft dated July 21 28, 1991? 22 Answer: Yes, I do. 23 Question: He starts that out by 24 reference to BARD. Do you know what that is? 25 Answer: Yeah, that's a misspelling of 11821 1 Brad. 2 Question: Oh, okay. 3 Brad, please make sure we are not 4 supporting DRI anywhere in the company with 5 this stuff, thanks. 6 Other than receiving this e-mail from 7 Mr. Ballmer, did you ever hear Mr. Ballmer make 8 statements of that nature? 9 Answer: I don't recall specifically 10 hearing him say that. 11 Question: Was this company policy at 12 Microsoft that you are not going to support DRI 13 anywhere in the company? 14 Answer: I'm really not sure how to 15 answer your question. You know, policy makes 16 it sound formal. 17 Question: There was nothing written 18 other than this e-mail that you're aware of? 19 Answer: No, I'm not -- sorry. I'm 20 not aware. 21 Question: Mr. Barrett, you've been 22 handed Exhibit 42 to your deposition. 23 This is, again, another series of 24 e-mails produced to us by Microsoft. 25 I'd like to direct your attention to 11822 1 the e-mail that begins at the bottom of the 2 first page. This is Bates marked X 575838. 3 Do you recognize this as an e-mail 4 message sent to you and others at Microsoft by 5 Ben Slivka on November 26, 1990? 6 Answer: Yes, I do. 7 Question: The subject line says, to 8 share or not to share, that is the question. 9 And the next series of exhibits is 10 going to -- Mr. Barrett, is going to concern 11 Microsoft's actions with the independent 12 software vendors. 13 Here Mr. Slivka begins his e-mail by 14 asking or stating the issue should Microsoft 15 applications (and other ISVs) have access to 16 Windows source code. 17 And you had stated earlier that ISVs 18 refers to independent software vendors? 19 Answer: That's correct. 20 Question: Mr. Slivka says -- you do 21 recognize this is an e-mail sent by Mr. Slivka 22 to you? 23 Answer: Yes, I do. 24 Question: All right. Mr. Slivka goes 25 on to say under the heading complications, and 11823 1 he says, number one, if Microsoft applications 2 have access to the Windows sources, then all 3 independent software vendors should have 4 access, else we are subject restraint of trade 5 complaints (to say nothing of the morality of 6 the situation). 7 Now, what was Mr. Slivka's position at 8 Microsoft? 9 Answer: Ben Slivka was a -- probably 10 still is, a software engineer, fairly senior 11 one. Project lead, technical lead type at that 12 point. I don't recall his specific role at 13 that time, though. 14 Question: At the time of this e-mail 15 message in 1990, were some independent software 16 vendors complaining that Microsoft applications 17 had some unfair advantage because they had 18 access to Windows source code? 19 Answer: I don't specifically recall 20 if it was happening at this time frame, but it 21 was happening. 22 Question: Well, let me ask you this. 23 Was it an advantage for applications developers 24 to have access to Windows source code? 25 Answer: Absolutely, yes. 11824 1 Question: And why is that? 2 Answer: Well, as it pointed out in 3 this memo, which I completely agree with, 4 Windows is poorly documented and there are 5 undocumented APIs, application program 6 interfaces, and there's behavior that is poorly 7 understood even by the Windows developers. 8 Question: And how does access to the 9 source code provide an applications developer 10 with some advantage? 11 Answer: Application developers, by 12 looking at the source code, while a fairly 13 arduous and difficult process, one that a small 14 ISV probably couldn't do or couldn't afford to 15 do, it would allow them to dedicate a person to 16 become a Windows internal expert and to be able 17 to advise their development teams on the best 18 way to structure their applications to perform 19 the best on Windows, in the Windows 20 environment. 21 Question: Does access to Windows 22 source code provide applications developers 23 with a time advantage? Does it make it easier 24 or faster to develop an application if you have 25 that access? 11825 1 Answer: Certainly time advantage and 2 efficiency advantage can be gotten by 3 understanding the internals of Windows. 4 Question: Does access to Windows 5 source code provide application developers with 6 any sort of cost advantage? 7 Answer: I'd have to say yes, although 8 it's not as simple a formula. 9 Question: If you'll look at 10 Mr. Slivka's e-mail message, as I said, it 11 starts with the heading issue and then it goes 12 on with complication. We read complication 13 number one. 14 He goes on in complication number two 15 to say, if outside independent software vendors 16 have access to our sources, then we make it 17 much easier for another company to come along 18 and clone Windows. 19 Let me ask you this. Were there 20 companies outside of Microsoft that were given 21 access to Windows source? 22 Answer: Yes, there were. 23 Question: And did Microsoft protect 24 themselves from the potential of cloning 25 Windows with a nondisclosure agreement? 11826 1 Answer: I can't answer one way or the 2 other. I simply don't know. 3 Question: All right. Now, when you 4 worked at Microsoft, you had access to Windows 5 source code, correct? 6 Answer: Yes. 7 Question: And you currently work for 8 a competitor of Microsoft; is that correct? 9 Answer: That's correct. 10 Question: Did you sign a 11 nondisclosure agreement at Microsoft that would 12 prevent you from disclosing Windows source code 13 to a competitor that you might later work for? 14 Answer: Yes. 15 Question: And I assume you have not 16 violated that agreement? 17 Answer: Absolutely not. 18 Question: Mr. Slivka goes on in his 19 e-mail message about a quarter of the way down 20 the page to say, the state machine that is 21 user.exe is barely documented in SDK. 22 Let's see if we can define some terms 23 here. SKD is the software development kit? 24 Answer: Software development kit, 25 yes. 11827 1 Question: And what is user.exe? 2 Answer: User.exe is one of the key 3 modules of Windows, parts of Windows. User 4 basically implemented all of the user 5 interface, Windows, menu, sliders, all that 6 stuff. That's what user.exe implemented. 7 Question: And the term state machine, 8 what does that refer to? 9 Answer: It's a technical term from 10 mathematics and electrical engineering that 11 refers to -- it's called finite state machine 12 is actually the correct terminology for it. 13 And it basically is -- there's a 14 concept of a state, which might be, say, 20 or 15 30 or 100 or a thousand different states and 16 inputs that change the state. 17 To put it in simple terms, think of it 18 as a machine that when you turn a dial, it 19 changes the internal gears. Very, very similar 20 sort of thing happens, although it's all 21 electronic. 22 And -- boy, I'm sorry, it's a very -- 23 it's a very specific technical term to 24 electrical engineering and computer science. 25 Question: I appreciate your effort to 11828 1 explain it to a lay person like myself. 2 Was it important to an applications 3 developer to have well-documented user.exe? 4 Answer: Clearly it was very 5 important. 6 As it pointed out in Ben's extremely 7 clear memo, in order to create an application 8 that has a good user interaction, that feels 9 normal, that feels like it, feels natural to 10 the user, you have to understand how the 11 windowing, the user interface system of Windows 12 works. 13 And you have to understand it in a way 14 that when you make a change here, that the 15 various other parts of it line up properly. 16 And so that when you pull down a menu, 17 that it looks right and that it reacts as if it 18 was effortless rather than, you know, laborious 19 or something like that. 20 So it's extremely important to 21 understand the internal workings of a windowing 22 system in order to write a window -- a 23 windowing application. 24 Question: Did you agree with 25 Mr. Slivka's statement here, that the user.exe 11829 1 was barely documented in the software 2 development kit? 3 Answer: Absolutely agree. Completely 4 right. 5 Question: And without getting too 6 technical, if you can, can you describe in 7 general terms what was not documented that 8 should have been? Are there general categories 9 of types of information? 10 Answer: Yeah. For example, in order 11 to get a window to -- the way Windows work is, 12 you would write some data into a window and 13 then you'd tell Windows to redisplay it. 14 So say if you were going to change the 15 contents of a menu, you'd have to write into 16 the menu and redisplay it. 17 In certain circumstances, those 18 redisplay messages or commands would get lost. 19 And so it became something that people learned 20 about and sort of folklore that you send -- in 21 certain circumstances, you send two redisplay 22 messages in order to get a window to actually 23 display. 24 And that's a very, you know, that's a 25 common case. There were others similar to that 11830 1 where it's just people figured out how to make 2 a Windows application work well by, in fact, 3 what does he say, experimentation. And that's 4 -- that's why it's clearly -- that's an aspect 5 of it being poorly documented. 6 Question: Mr. Slivka goes on to say, 7 any independent software vendor that wants to 8 write a great Windows application ends up 9 looking at the source code (like our 10 applications group) unassembling the DLLs or 11 writing experiment code to devine the actual 12 behavior of the system. 13 First of all, what are DLLs? 14 Answer: Dynamic link library is what 15 DLL stands for. And, effectively, it's the 16 building blocks of Windows. All the various 17 Windows pieces, device drivers, everything are 18 DLLs, dynamic link libraries. 19 Question: Now, in this sentence 20 Mr. Slivka identifies three ways that an ISV 21 can write, quote, great Windows applications. 22 They can either look at the source 23 code like the Microsoft applications group. Do 24 you see that? 25 That's the first one he identifies. 11831 1 Did the -- or to your knowledge, did 2 the Microsoft applications group have access to 3 Windows source code? 4 Answer: I believe they did. 5 Question: And can you tell me what 6 the basis for that belief is? 7 Answer: There were a number of -- 8 there were a lot of friendships between people 9 in the systems group and the applications 10 group. There was a lot of informal sharing 11 going on. 12 And application developers would show 13 up in the area where the Windows developers 14 worked and ask questions of developers, and 15 they'd sit there and look at code and answer 16 questions. 17 I believe that the apps group had 18 access to -- they did indeed have access to our 19 source code control system and they could look 20 and find out what was going on in the source 21 code. So they were only to look at the source 22 code. 23 Question: And what is the source code 24 control system? 25 Answer: It's a computer program that 11832 1 basically allows whoever the manager of the 2 source code to control access by giving people 3 passwords and also maintaining the integrity of 4 the source code. 5 People could check out the source code 6 and make changes and check it back in. If 7 somebody had a piece of source code checked 8 out, nobody else could check it out because, 9 you know, didn't want to have two people 10 writing in the same document, if you will, and 11 then having to merge the changes back. The 12 document in this case being the piece of code. 13 Question: Did anyone from the 14 Microsoft Excel spreadsheet application group 15 have access to the expertise of the Windows 16 development group? 17 Answer: Absolutely, yes. 18 Question: Can you tell me who that 19 was and what the circumstances were? 20 Answer: I believe there were more 21 than one, but in particular Chris Peters, the 22 leader of that time, and subsequently vice 23 president of Microsoft was -- frequented the 24 Windows development area and was actually quite 25 good friends with a number of people there and 11833 1 spent a lot of time talking with the Windows 2 developers and looking at the Windows code and 3 really understanding it. 4 Question: Did Mr. Peters work 5 evenings in the Windows development building? 6 Answer: He was a frequent visitor, 7 yes, in the evenings. 8 Question: Looking back at 9 Mr. Slivka's e-mail message in Exhibit 42, 10 after referring to looking at Windows source 11 code like Microsoft's application group, he 12 identifies another way that this can be done. 13 He refers to unassembling the DLLs. 14 Can you describe how that works? 15 Answer: Unassembling means to take -- 16 or disassembling actually. The more common 17 term is disassembling. Means the same thing. 18 Unassembling, in this case is 19 referring to taking a piece of -- taking a 20 program and using a special program called a 21 disassembler and running it on that program and 22 turning the machine, basically the numbers, the 23 operation codes and other data into what they 24 call pneumonics or a human readable form that's 25 describing the sequence of instructions that 11834 1 the binary code refers to. 2 And that would allow someone to look 3 at how a piece of software works. It's not 4 particularly efficient. And it's actually very 5 error fraught, but it does allow somebody to 6 understand what's going on if they allow enough 7 time and thought to it. 8 Question: And I want to understand 9 what you said. It is error fraught or it's not 10 error fraught? 11 Answer: It's highly -- it's highly 12 error fraught. It's complex and difficult to 13 do. And only a few people -- there's a small 14 number of people in the world that are good at 15 it. 16 Question: So susceptible to error, is 17 that what you're saying? 18 Answer: Yes. 19 I'm sorry. 20 Question: Then Mr. Slivka also talks 21 about an ISV looking at source -- I'm sorry, an 22 ISV writing great Windows apps by writing 23 experiment code. 24 Do you see that? 25 Answer: Yes. 11835 1 Question: To devine the actual 2 behavior of the system. What is that? 3 Answer: What that refers to is 4 writing a piece of code that does a specific 5 thing. I don't know. Calls the menu code or 6 whatever, Windows management code and tries 7 different things. Tries different inputs. 8 Allows the developer that's doing 9 this, the experimenter, if you will, to try 10 different things to see how the system behaves 11 based on different input. 12 And when different input -- by giving 13 different input to the system and seeing how it 14 behaves, you can start to form a model of how 15 the system behaves given different input. 16 That allows the developers to do a 17 better job of working with the system as it 18 actually works, not the way it's documented. 19 And this is essential when the 20 documentation does not match the actual 21 behavior of the system, which was indeed the 22 case with Windows 3.0 in many ways. 23 Question: Can you think of any 24 specifics as to how Windows 3.0 documentation 25 did not match the behavior of the system? 11836 1 Answer: Well, certainly the repaint 2 -- or the redraw thing I pointed out before, 3 where in some cases you had to send two redraw 4 messages to a window to cause it to actually 5 display its updated contents. 6 Question: That was not documented? 7 Answer: No. That was documented 8 nowhere. 9 Question: Can you think of any others 10 offhand? 11 Answer: Not off the top of my head. 12 Question: Is this system of writing 13 experiment code to d