Here is IBM's Redacted Memorandum in Support of Motion for Partial Summary Judgment On Breach of Contract Claims, filed by IBM on Friday. It's a hundred-page PDF. As you will see, they are going for the jugular now. Astoundingly, they say that all parties involved in the contract between AT&T and IBM have now provided testimony in discovery that IBM has the right to do whatever it wishes with its own code, contrary to SCO's claims, or as the memorandum puts it, they all provided "unequivocal testimony that the agreements were not intended and should not be understood to preclude IBM's use and disclosure of homegrown code and contemporaneous documents reflect this interpretation of the licenses".
SCO thought it was going to find evidence in discovery to bolster its case, but it has worked out exactly the opposite. I haven't finished reading the memorandum myself, because I wanted to share it with you immediately, but it looks like this is the heart of what's left of SCO's case. Everything you hoped IBM would say to the judge, they are saying, including pointing out that Novell has waived any breach, even if there had been one, which there wasn't. If IBM wins this motion, I think I might be in my red dress soon.
You don't want to miss reading page 76. It's where IBM tells the judge that as recently as August 4, 2004, SCO was *still* offering the Linux 2.4 kernel for download, the very code it is suing IBM over. My understanding of the significance of this revelation is that SCO has now released knowingly all the code at issue under the GPL. In the beginning of the case, they said they didn't know the allegedly infringing code was in there when they released Linux under the GPL. They can't say that since they filed the lawsuit in March of 2003. Now, in August of 2004, they are *still* distributing the same code under the GPL. Under the terms of the GPL, there is no taking that code back that I know of. I think, therefore, that SCO's case just went poof, on this one issue alone.
The Memorandum's preliminary statement goes like this:
Although SCO for months perpetuated the illusion that it had evidence that IBM took confidential source code from UNIX System V and "dumped" it into Linux, it has become clear that SCO has no such evidence. Instead, SCO's claims that IBM breached its agreements with AT&T depend entirely on the allegation that IBM improperly contributed certain of IBM's original source code, contained in its own AIX and Dynix operating systems (each of which contains tens of millions of lines of source code), to Linux. According to SCO, because AIX and Dynix allegedly contain some small component of source code from UNIX System V (SCO claims there are approximately 74,000 lines of UNIX System V code in AIX and approximately 78,000 lines in Dynix, which amounts to less than one percent of the total lines of code in AIX and Dynix), IBM is prohibited by its licensing agreements from disclosing any of the other millions of lines of code in AIX or Dynix, even if that code was created by or for IBM and contains no UNIX System V code. And note that on page 10, the memorandum is supported by no less than *16* declarations. Oh, my. We will be busy bees this week.
SCO is wrong as a matter of law, and IBM is entitled to partial summary judgment on SCO's contract claims, for at least two independent reasons.
First, the AT&T agreements upon which SCO's claims are based do not preclude IBM from using and disclosing source code that is written by IBM and does not include UNIX System V code (referred to herein as "homegrown" code):
1. The plain and unambiguous language of the agreements imposes no restrictions on the use or disclosure of source code that does not contain UNIX System V code (See Section I.A.)
2. The individuals who executed the licenses and were involved in their negotiation, on behalf of both AT&T and IBM, have offered unequivocal testimony that the agreements were not intended and should not be understood to preclude IBM's use and disclosure of homegrown code and contemporaneous documents reflect this interpretation of the licenses. (See Section I.B.)
3. Interpreting the licenses to prohibit the disclosure of homegrown code would be patently unreasonable. (See Section I.C.)
Second, even if the AT&T agreements could be read to preclude the disclosure of homegrown code -- and they cannot be -- any breach based upon such a reading has been waived by Novell, Inc. ('Novell') on behalf of SCO, and by SCO itself:
1. Novell, which at one time owned all rights in the AT&T agreements at issue, retains the right to waive alleged breaches of the agreements, and Novell has exercised that right to effect a waiver of the alleged breaches in this case. (See Section II.A.)
2. SCO itself sold or otherwise made available to its customers and the public the code it claims IBM should not have revealed. By its own conduct, therefore, SCO has waived any right to claim that IBM acted improperly by contributing its code to Linux. ( See Section II.B.)
For these reasons, partial summary judgment should be entered on behalf of IBM on SCO's claims for breach of contract (SCO's First, Second, Third and Fourth Causes of Action).
Also, note paragraphs 60 and 61 on pages 23 and 24:
60. It is plain from SCO's discovery responses that SCO has no evidence that any of the source code IBM contributed to Linux is either literally copied from source code in UNIX System V or is derived from source code in UNIX Sytem V. (See Exs. 6 & 27.) Indeed, SCO has purported to identify the lines of UNIX System V code that are present in AIX and Dynix, and none of those lines are among the lines of code SCO claims IBM improperly contributed to Linux. (See Ex. 6 at Exc. E & F.)
61. In addition, Dr. Randall Davis, Professor of Computer Science and Engineering at the Massachusetts Institute of Technology, has analyzed the specific lines of source code from AIX and Dynix that SCO claims IBM contributed to Linux. (Davis Decl. ¶ 23.) As Dr. Davis has concluded, that code does not contain any portion of source code from UNIX System V and is not substantially similar to any source code in UNIX System V. (See id. ¶ 48.) Accordingly, Dr. Davis opines that the specific code IBM allegedly contributed from AIX and Dynix is neither a modification nor a derivative work of UNIX System V. (See id. ¶ 49.)
So, an MIT deep diver surfaces at last. A real one. With a name. And on IBM's side.
And you thought IBM lawyers had no sense of humor.
Also of interest is Otis Wilson's testimony. Note page 25, paragraph 68, and page 28, paragraph 82:
68. At the time the agreements were signed, Mr. Wilson was the head of AT&T's department responsible for licensing AT&T's UNIX software, including UNIX System V, worldwide. . . .
So Sequent is clean too, according to Mr. Wilson. He ought to know. He was head of AT&T's licensing department -- not IBM's -- at the time.
82. For example, Mr. Wilson states in his declaration:
"These provisions set forth our licensees' rights as they relate to the UNIX System V source code and related materials -- the 'SOFTWARE PRODUCT' or 'SOFTWARE PRODUCTS' -- that AT&T provided to them. At least as I understood these sections and discussed them with our licensees, they do not, and were not intended to, restrict our licensees' right to use, export, disclose or transfer their own products and source code, as long as they did not use, export, disclose or transfer AT&T's UNIX System V source code along with it. I never understood AT&T's software agreements to place any restrictions on our customers' use of their own original work."
Mr. Wilson raises two more points on page 30: first, that no licensee would have been willing to sign themselves into indentured servitude by agreeing that all code they wrote or paid others to write or licensed from others from that day forward would belong to AT&T, and two, that had they been willing to do so, it would have raised antitrust issues for AT&T at the time. This, obviously, directly contradicts the main thrust of SCO's claim that all modifications, derivatives, methods, blah blah belong to them. Remember, Mr. Wilson is saying, that AT&T had antitrust concerns back then, and it is implausible that such a one-sided contract could have passed muster:
85. The witnesses do not dispute that the 'resulting materials' referenced in Section 2.01 do not include homegrown code, even when that code is itself contained in a modification or derivative work based on UNIX System V. (See Wilson Decl. ¶ ¶ 14-15; Ex. 34 at 55:23-57:14; Frasure Decl. ¶ ¶ 13-16; Ex. 35 at 42:17-48:25; McDonough Decl. ¶ ¶ 11-20; Rodgers Decl. ¶ ¶ 7-9; Ex. 36 at 25:15-30:20, 100:1-102:5; DeFazio Decl. ¶ ¶ 16-17; Kistenberg Decl. ¶ ¶ 11-12; Vuksanovich Decl. ¶ ¶ 14-15; Cronan Decl. ¶ 11; Mobley Decl. ¶ ¶ 8-9; Swanson Decl. ¶ 10.) Then AT&T's Stephen Vuksanovich in his declaration, paragraph 91 on page 31, takes it even further:
86. Mr. Wilson of AT&T states:
As my staff and I communicated to our licensees, this provision [Section 2.01] was only intended to ensure that if a licensee were to create a modification or derivative work based on UNIX System V, any material portion of the original UNIX System V source code provided by AT&T or USL that was included in the modification or derivative work would remain subject to the confidentiality and other restrictions of the software agreement. As we understood Section 2.01, any source code developed by or for a licensee and included in a modification or a derivative work would not constitute 'resulting materials' to be treated as part of the original software product, except for any material proprietary UNIX System V source code provided by AT&T or USL and included therein."
87. Indeed, Mr. Wilson states that he "do[es] not believe that our licensees would have been willing to enter into the software agreement if they understood Section 2.01 to grant AT&T or USL the right to own or control source code developed by the licensee or provided to the licensee by a third party". (Wilson Decl. ¶ 16; see also Ex. 34 at 57:15-58:20.) Mr. Wilson, in fact, was of the view at the time "that we could not claim any rights to non-UNIX System V code source . . . without raising serious antitrust issues. (Id. ¶ 18; see also Ex. 34 at 57:15-58:20, 59:15-61:3.)
91. ...Our standard software agreements also gave licensees the right to modify UNIX System V source code and to prepare derivative works based upon the code. As I believe we intended the agreements, and as I told our licensees, our licensees owned their modifications and derivative works they prepared based on UNIX System V, and were therefore permitted to do as they wished with those modifications and derivative works, so long as they treated those portions of the modifications or derivative works consisting of any UNIX System V source code the same way they treated the UNIX System V source code that we provided to them. I recall that during our negotiations IBM specifically wanted to make sure that IBM, and not AT&T, would own and control code that was developed by or for IBM, even if that code was mixed with AT&T's UNIX System V code in a product. I assured IBM that we had the same understanding. So the issue came up in negotiations and IBM was told modifications belonged to IBM, and so long as they kept System V code out, they were free to do whatever they wished with their own code, even modifications to System V code they wrote themselves. IBM can hardly be accused of violating any contract if that is what they were told. And with all these witnesses, who negotiated the contract on both AT&T and IBM's sides, saying the same thing, who in the world can SCO produce to counter this testimony?
IBM sums it up on page 62:
The plain and unambiguous language of IBM's and Sequent's agreements with AT&T therefore establishes that the use and disclosure restrictions of Sections 2.01, 2.05, 4.01, 7.06(a) and 7.10 of the IBM and Sequent Software Agreements do not apply to original code created by IBM and Sequent. To the contrary, AT&T expressly disclaimed any purported right to control the use and disclosure of IBM's or Sequent's homegrown code.
And on page 63, IBM states that SCO has provided no evidence that the code IBM is alleged to have improperly contributed to Linux from AIX and Dynix -- such as RCU, JFS, EVMS and AIO -- contains any UNIX System V code:
Therefore, such code is not subject to any restrictions under the IBM and Sequent Software Agreements, and SCO's contract claims fail as a matter of law.