% fortune -ae paul murphy

SCO vs. IBM: as push approaches shove

Just recently groklaw published both IBM's motion for summary judgement on SCO's contractual claims and SCO's rebuttal argument.

The judge could rule on those today or later or this week, but the outcome is less interesting than the documents themselves because here, for the first time, both SCO and IBM state their cases clearly and relatively simply.

As the motions make clear everything ultimately comes down to one issue: did IBM breach contracts now held by SCO?

For a judge to grant a motion for summary judgement, however, the judge has to agree that the facts are undisputed, and since they're not my belief is that IBM's lawyers had to know that the only motion that counts, the one on contractual issues, would not succeed.

The key question, therefore, is why they decided to waste the court's time with it - and my guess, because as regular readers known my belief is that SCO has a strong case, is first that they're trying to use IBM's financial strength to bankrupt SCO, and, secondly, that they're trying to establish a fundamental misrepresentation of the key issues as fact.

Here's part of the critical text, from the SCO side:

First, IBM argues briefly otherwise, but the plain language of the standard AT&T UNIX license agreement required the licensee to hold in confidence all parts of the modifications and derivative works the licensee developed based on the licensed UNIX software product. IBM does not dispute that it developed such a derivative work, AIX, and acquired another such derivative work, Dynix/ptx, when it acquired Sequent Computer Systems, Inc. in the late 1990s. IBM also does not dispute that it made hundreds of contributions or disclosures of technology to Linux development from AIX and Dynix/ptx. There is no reading of the agreement to support IBM's theory that the source code, methods and concepts it developed fall outside the confidentiality restrictions when they were created as part of or added to such derivative works.

Second, although IBM entered into a side letter with AT&T that amended aspects of its standard UNIX license agreement, none of those amendments permitted IBM openly to disclose the source code of a product, such as AIX, into which IBM had indisputably copied UNIX source code. In addition, as stated by the clear language of the AT&T's UNIX license agreements, the amendments of the IBM Side Letter did not affect the terms of the standard UNIX license agreement into which Sequent had entered. The vast majority of contributions to Linux that SCO challenges here were made from Sequent's Dynix/ptx derivative work.

Third, in addition to the fact that any contractual ambiguity in the agreements at issue precludes summary judgment under New York law, IBM's reliance on only selected extrinsic evidence is improper. IBM disregards the sworn testimony of numerous witnesses responsible for negotiating, overseeing, interpreting and enforcing the UNIX license agreements for decades and contemporaneous documents contradicting the position that IBM and selected declarants now advance. That evidence underscores the narrow sample of evidence on which IBM relies in its Motion and easily permits the inference that the testimony IBM does cite is unreliable. In addition, the parties' respective businesses and economic incentives at the time of contracting -evidence supported by SCO expert witness Dr. Gary Pisano and that IBM concedes to be relevant in the event of ambiguity in the Agreements - bears out the reasonable scope of the Agreements as SCO has explained them.

Fourth, in addition to the substantial number of relevant witnesses whose testimony IBM's Motion fails to account for or reconcile, the selected witnesses on whom IBM does rely have given other testimony belying IBM's arguments.

...(some examples)

Such evidence reveals a pattern whereby witnesses have signed declarations, drafted by IBM, that turn out not to reflect the witnesses' true understanding of or intent in the agreements and transactions at issue, or that plainly contradict testimony the witnesses have previously given or contradict contemporaneous documents that IBM simply chooses to ignore. The clarifications of declarations as they have been drafted or used by IBM underscores the need for the trier of fact to reach conclusions and draw inferences from those disputed facts.

Fifth, IBM's arguments for "estoppel" and "waiver" suffer from the same defects of unambiguous contractual language, contradictory extrinsic evidence, and improper inference. IBM further argues that Novell had the right to and did "waive" SCO's claims against IBM by virtue of mid-1990s agreements between Novell and Santa Cruz, but substantial, relevant evidence of those transactions - including the language of the agreements and amendments thereto and the testimony of all of the principal negotiators on both sides of the transactions - contradicts IBM's "facts" and arguments. Both Mr. Chatlos and Mr. Bouffard, for example, explain why IBM is wrong.

You'd think IBM would pretty much say the opposite, but they don't. Instead, they seem to make two distinct kinds of argument: one based on legal technicalities, and the other on an appeal to natural justice.

To judge the former you'd have to be a lawyer - although you can get a feel for the tenor of IBM's arguments by looking at extreme examples like this one:

The statute of limitations for breach of contract is six years under New York law, which governs SCO's contract claims. IBM publically disclosed the allegedly misused material relating to RCU in connection with a patent application and the resulting patent, which issued in 1995, more than six years prior to the date that SCO filed its lawsuit, Thus SCO's claims relating to RCU are barred by the statute of limitations.

In other words, they argue that because they got away with it before, they should get away with it forever.

In contrast the natural justice arguments are easier to assess - here's the shortest statement I could find in IBM's filing of something they repeat endlessly:

Neither IBM not Sequent would have invested in AIX and Dynix as they did if they had believed that AT&T or its successors, instead of IBM and Sequent, owned and had the right to control IBM's or Sequent's original works, whether or not they are part of a modification or derivitive work of Unix System V.

Notice that the main argument advanced here: that IBM would never have broken the terms of the contract if they'd believed that the terms applied to them; is just a wrapper for another justification entirely: that IBM has a right, if not indeed a duty to the law, to oppose SCO because SCO asserts that the contracts gave AT&T or its successors both control over, and ownership of, derivitive works.

If I understand things correctly the entire ownership and control thing has to be a deliberate misrepresentation of SCO's position. The AT&T contracts licensed IBM to use AT&T intellectual property code for its own purposes, including direct use of AT&T source and the derivation of other works, provided that IBM paid agreed fees and did not release any of the original or derived materials to the public. If so, there is no claim of ownership nor assertion of control over anything beyond the embedded intelletual property.

Given that the contractual remedy was to suspend IBM's license to use the IP and that this has the effect of barring IBM from selling or otherwise benefiting from works incorporating that IP, the extension to product ownership and control is a natural misinterpretation - but that doesn't make it less wrong and using it as the basis of argument in a legal document amounts to misrepresentation.

So if this usage is an emotionally manipulative appeal for support, are we to believe that IBM's lawyers are a bunch of sloppy thinkers committed to a freshman mistake or that this is a deliberate strategy aimed at building support among potential jurors, the groklaw hordes, and anyone at IBM who can read a contract? - because the plain language of the thing is against them, and their one real chance at law, proving that absolutely everything they disclosed to anyone, had already been disclosed by someone else, is awfully thin.

Far too thin, I'd think, for IBM's board to bet something like ten billion shareholder dollars on - and that means the (inevitable?) dismissal of IBM's motion for summary judgement on contractual issues should pave the way for someone at the top to intervene: demanding that this mess be settled as quickly, quietly, and cheaply as possible.

As I've said before, there is a natural basis for a settlement under which SCO would drop the case and IBM would pay SCO to release all of its remaining rights in Unix code to the public -allowing IBM to leave the field in triumph after having achieved something of value to the Linux community.


Paul Murphy wrote and published The Unix Guide to Defenestration. Murphy is a 25-year veteran of the I.T. consulting industry, specializing in Unix and Unix-related management issues.