We were wondering what SCO's motion to enforce the scheduling order would be about. Well, here it is, SCO's Memorandum in Support of SCO's Expedited Motion to Enforce the Court's Amended Scheduling Order Dated June 10, 2004. What they want is for IBM's motions for summary judgment to be put on hold until discovery is scheduled to be complete, in February.
In a word, they are terrified. They would prefer that their ship sink into the sea in February instead of September. I don't know. The water is a lot colder in February.
So, they want more DELAY. What else doees SCO ever ask for? AIX? Well, yes. They ask for that too. In short, they have found a way to respond to IBM's motions more than once apiece. This motion answers and they get to reply to all the motions too.
I have only skimmed it, so I'll write some more about this after I have time to evaluate it, but in the meantime, we can read it together. Note that IBM's memorandum in support of its motion for partial summary judgment on SCO's breach of contract claims, as text in the article just before this one, has references to and quotations from the AT&T and IBM executives mentioned in this SCO document.
Update: This document was unsealed [PDF] about a year later, and the only redacted section, on page 22 of the document, or 24 of the PDF, reads like this:
Email from Bill Sandve to Kim Tran, January 22, 2002 at 2 (Exh. 4)
(“AIX was derived from System V.”); Agreement for Licensing of AIX Source Code and
Related Products between Argus Systems Group, Inc. and IBM ¶ 8.4 (Exh. 5) (“AIX is derived
form software under license from SCO.”); International Business Machines Corp. Royalty
Statement, June 30, 1987 (Exh. 6) (stating that AIX is “derived from System V”); IBM/Supplier
Technical Services Agreement at 1 (Exh. 7) (“‘AIX Operating System’ or ‘AIX’ shall mean the
UNIX operating system that operates on the Power, Power PC, Power 2, Power 3, and Power 4
architectures or derivatives or follow-on architectures irrespective of the names of such
architecture.” (emphasis added)); Rodgers Dep. 138 (Exh. 8) (“Dynix/ptx is almost certainly a
derivative work of Unix System V.”) Don't ask me why they'd redact that portion, unless perhaps they were afraid Groklaw would explain what derived from means with regard to AIX and how Dynix/ptx came to be.
There were exhibits attached as well, some of which were later made available as well. You can find them all, the ones publicly available, on Groklaw's IBM Timeline page, under docket number 292.
Brent O. Hatch (5715)
Mark F. James (5295)
HATCH, JAMES & DODGE
[address, phone, fax]
Robert Silver (admitted pro hac vice)
BOIES, SCHILLER & FLEXNER
[address, phone, fax]
Stephen N. Zack (admitted pro hac vice)
Mark J. Heise (admitted pro hac vice)
BOIES, SCHILLER & FLEXNER LLP
[address, phone, fax]
Attorneys for The SCO Group, Inc.
IN THE UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF UTAH
THE SCO GROUP,
MEMORANDUM IN SUPPORT OF
SCO’S EXPEDITED MOTION TO
ENFORCE THE COURT’S AMENDED
SCHEDULING ORDER DATED JUNE
Case No. 2:03CV0294DAK
Honorable Dale A. Kimball
Magistrate Judge Brooke C. Wells
Plaintiff The SCO Group, Inc. (“SCO”) respectfully submits this memorandum in support of its expedited motion to enforce the Court’s Amended Scheduling Order dated June 10, 2004. SCO hereby requests that the Court defer the scheduling and adjudication of any dispositive motions until after the end of the Court’s fact-discovery period, which the Amended Scheduling Order extended to February 11, 2005.
IBM has now filed three fact-intensive dispositive motions, six (or more) months ahead of the Court’s extended fact-discovery cut-off, which purport to cover almost all of the issues in the case. Adjudicating these motions on IBM’s schedule (the motions are now scheduled for hearing on December 9) could not eliminate the need for discovery. Discovery could not be suspended pending disposition of the motions without exposing SCO to the risk of having to resume and complete all discovery within a matter of weeks. Because deferring adjudication of these motions could not sacrifice any case-management efficiency, there is no benefit from adjudicating these motions before the close of fact discovery.
At the same time, there is a great cost associated with from adjudicating these motions now. By coordinating the accelerated timing of these dispositive motions while it continues to block critical discovery, IBM would achieve what this Court’s Amended Scheduling Order –
over IBM’s stringent objection – refused to allow IBM to do: force adjudication of this case without adequate opportunity for SCO to develop its proof. IBM’s coordinated motions would thus nullify the Amended Scheduling Order.
Given their content, moreover, IBM’s motions do not have any apparent additional purpose. Although IBM claims there is no issue of disputed fact on what it calls, for example, the dispositive contract issue in the case, IBM asks the Court to rely on the testimony of witnesses who previously gave sworn testimony that directly contradicts their current testimony on this key issue. See pp. 42-59, below. (In fact, what IBM labels SCO’s “New-Found Contract Theory” is the very theory that these same IBM witnesses described under oath more than ten years ago, when they were immersed in the day-to-day issues, and before IBM came to represent them or fund their representation.) This flatly contradictory testimony, which IBM has had for months, obviously precludes any possibility that IBM’s new interpretation – its effort to rewrite history – could be undisputed as a matter of law.
Similarly, while IBM rhetorically claims it is only seeking protection for its “homegrown” and “original” work, IBM asks in the fine print of its contract motion for the right to have copied and otherwise exploited with impunity a seminal, industry-changing innovation in ways that black-letter law would enjoin as misappropriation – before contractual protections are even considered. IBM also claims that SCO’s predecessor AT&T conferred that extraordinary gift even though the IBM and Sequent license agreements that gave IBM and Sequent the right to see and use AT&T’s (and now SCO’s) source code are replete with plain terms that had (and have) no purpose except to prohibit precisely the right to copy and exploit that IBM seeks.
IBM also presents witnesses claiming that AT&T was ready and willing to let IBM and Sequent remove those express terms if they had wanted to, and to ask nothing in return. But IBM also tells the Court that IBM and Sequent nevertheless simply chose not to do so – needlessly (and inexplicably) letting their claimed rights depend on linguistic contortions and extrinsic evidence (presumably including IBM’s ability, as needed, to have witnesses flatly contradict their prior testimony once IBM came to “represent” them).
IBM’s motions are riddled with such errors and overreaching – and thus confirm that through those motions IBM seeks simply to attack the Amended Scheduling Order and to accelerate adjudication of key issues in this case before SCO can obtain, and pursue, Court-ordered discovery that IBM continues to withhold.
The relief SCO proposes would protect this Court’s and the Magistrate Judge’s jurisdiction and orders. SCO shows below that the Court may and should impose such relief in its well-established discretion.
Background and Overview of Issues
In early June, when IBM opposed SCO’s motion to extend the scheduling order, IBM sought to force adjudication at a point where depositions would have had to proceed before IBM had produced much rudimentary discovery, where IBM had only recently produced even a single version of the AIX program at issue in the case (but had held back many critical versions), and where – as a result – dispositive motions would be scheduled to be filed and adjudicated (without such predicate discovery) this fall. In its Amended Scheduling Order, the Court overruled IBM’s position and extended the fact discovery cut-off by over six months, from August 4, 2004, until February 11, 2005.
Since that time, IBM has sought to nullify that Amended Scheduling Order by taking every step to achieve indirectly what the Order prevents IBM from accomplishing directly. SCO has received none of the predicate discovery it was seeking before June 10, including discovery that Magistrate Judge Wells ordered and IBM has now conceded it has not produced. IBM also refuses over Court order to produce even the identities of, and related information about, IBM and Sequent programmers, so as to allow the depositions of such programmers – sixibiblio months into a case about programming. SCO will not receive any such discovery until sometime in the fall and will not be able to use such discovery until later in the year, including much discovery that SCO has shown will contain material reflecting IBM and Sequent’s actual views and thus potential admissions about the contracts at issue – as well as information that IBM’s own declarants made centrally important when they provided prior testimony flatly contradicting what they and IBM now assert.
At the same time, IBM seeks the dismissal of almost every issue in the case, on the ground that SCO (allegedly) cannot produce the evidence it seeks through the discovery that IBM continues to block. IBM asks the Court to adjudicate its motions during the period the original scheduling order, now replaced, provided for such motions. IBM’s coordinated actions would force SCO to proceed with key depositions before being able to access even the basic Court-ordered discovery that IBM continues to withhold. And while IBM expressly asks this Court to rely on the absence of certain evidence, IBM asks the Magistrate Judge in separate filings to block SCO’s access to that very same evidence on the ground that it is “wholly irrelevant.”
IBM has an obvious interest in trying to force SCO to waste its deposition opportunities while IBM holds back other discovery and continues to litigate over Court-ordered discovery that was requested more than a year ago. IBM also has an obvious interest in trying to position the litigation as if the original scheduling order remained in place, while compromising SCO’s ability to use the Court’s extended discovery period to pursue the development of its proof. That is one reason why this Court, through its orders, controls the litigation.
In addition to timing its motions as if the original scheduling order were in place, IBM has taken care to make the return dates for the motions virtually coincide with the date of the hearing before the Magistrate Judge regarding SCO’s outstanding discovery requests – a hearing IBM did not want to occur – as well as with the hearing the next day before this Court on the first of IBM’s dispositive motions. To arrange for this tactical timing, IBM went so far as to serve its contracts motion on SCO’s counsel at his home at 9 p.m. on Friday, August 13 – an otherwise inexplicable act given the fact that IBM was filing its motion approximately six months ahead of the Court’s discovery cut-off. IBM served its dispositive motion on its Eighth Counterclaim on August 16, making it returnable one day after the hearing on IBM’s Tenth Counterclaim Motion.
IBM has since sought to leverage that conduct even further. IBM now asserts that it will agree to “reasonable” (IBM’s word) extensions of time to respond to their motions – but only if SCO agrees to forfeit its right to seek discovery to respond to IBM’s motions. See Letter from T. Shaughnessy to B. Hatch dated Aug. 27, 2004 (Exh. 1).
IBM has thereby tried to force SCO to forfeit the very discovery rights SCO has been seeking to vindicate for over a year – including rights that the Magistrate Judge has upheld in a March 3 Order with which IBM now concedes it has still not complied. Directly following this Court’s Amended Scheduling Order, SCO made clear to IBM that it needed predicate discovery before moving on to the stage of deposition discovery. SCO also made clear that a preliminary deposition plan would have to depend on when disputes were resolved over IBM’s refusal to provide such rudimentary discovery. In submissions filed this summer, SCO has further made clear precisely why it needs this predicate discovery including for the specific purpose of preparing for key contract depositions and the development of its contract case. IBM nevertheless continues to block all of that discovery (just as it did before the Amended Scheduling Order), and continues to contend that SCO should already have had to take its key depositions without access to any such discovery (just as it did before the Amended Scheduling Order).
In fact, IBM contends that SCO should move forward with a deposition plan even though IBM still will not even comply with a Court Order requiring it to simply provide the identities and other ordered basic information about programmers so that such predicate depositions could at least become possible. IBM admits that it possesses this Court-ordered information and has not produced it; IBM even admits that it pointed SCO to sources to obtain this information which did not, in fact, contain the information – a concession SCO had to litigate for months to secure. Yet IBM now has filed a dispositive motion based critically on testimony from witnesses – who when they gave prior contradictory testimony relied on just the kinds of facts about the programming process that those programming depositions would have allowed SCO to establish – but which IBM will not allow to occur, and will not provide the predicate discovery to allow SCO to take effectively.
IBM has likewise held back a repository of potential admissions with respect to issues relating to the scope of contract protections. IBM now tries to force SCO to forfeit its right to seek even the identities of these potential programming deponents, and to forfeit its right to seek a variety of other discovery, all sought (for over a year), simply at the price of securing what IBM admits would be a “reasonable” extension to IBM’s dispositive motions. IBM seeks to extort this forfeiture having first timed the filing of its motions to require responses at the most burdensome possible time for SCO and the most advantageous time for IBM.
In the discovery context here – in which, for example, the parties litigated whether discovery should go forward while IBM sought to hold back discovery, and IBM lost; in which IBM litigated to preserve the Court’s original scheduling order, and IBM lost; and in which IBM sought to schedule the hearings on SCO’s discovery motions in mid-September, and won – IBM now presumes to put SCO in a position where SCO either has to answer IBM’s motions without even a “reasonable” extension of time on the very dates of the discovery hearings IBM sought to avoid, or else waive its rights to seek discovery to answer IBM’s overtly fact-intensive motions. IBM’s efforts thereby to prevent SCO (and the Court) from examining relevant discovery, and to arrogate to itself the Court’s control over this litigation, are plainly improper.
IBM’s proposed course thus presents no benefit, but great cost. This conclusion is only underscored by IBM’s judgment arguments:
IBM’s Motion for Summary Judgment on Its Claim That No Contribution to Linux by Anyone Violates Any SCO Copyright. IBM has moved for summary judgment not only on all potential copyright issues raised in SCO’s claims, but also on a wide range of new copyright issues that IBM has injected into the case. Rather than undertaking any substantive examination of the issues, IBM’s motion seeks to preclude SCO from undertaking such an examination. IBM’s motion, among other things:
directly attacks the Court’s Amended Scheduling Order by seeking to impose a discovery cut-off 10 months earlier than the Court’s extended period;
directly attacks the Court’s Amended Scheduling Order by seeking to impose a discovery cut-off that would have eliminated SCO’s discovery rights less than one month after IBM produced even a single version of the IBM software at issue in the case;
leverages its attack on the Court’s Amended Scheduling Order by trying to nullify the Magistrate Judge’s March 3, 2004 Order allowing SCO to show that it needs further versions of IBM’s software (which SCO could not begin to do until it received at least a single version of that software, and which it could not do at all under the discovery cut-off that IBM sought);
leverages its attack on the Court’s Amended Scheduling Order by trying to nullify the Magistrate Judge’s March 3 Order requiring IBM to produce rudimentary discovery sought for over one year (which IBM’s cut-off would allow it to never produce – as it still has not);
leverages its attacks even further by trying to cut off SCO’s discovery rights less than two months after IBM injected a wide range of additional issues into the case, encompassing all contributions to Linux by parties other than IBM; and
purports to support these attacks on Court Orders with selective quotations that omit material from the Court Orders and from publicly available documents on which the Orders are based.
See SCO’s Memorandum in Opposition to IBM’s Motion for Summary Judgment on IBM’s Tenth Counterclaim.
2. IBM’s Motion for Summary Judgment on Its Claim That SCO Infringed IBM’s Copyright.
IBM also has moved for summary judgment on its Eighth Counterclaim, which seeks a declaration that SCO is liable for copying IBM code already contributed to Linux. That motion, among other things:
asserts that it is true as a matter of law that SCO is “licensing Linux,” see Part II.B.2, below;
does not inform the Court that SCO has never licensed anything but the right to use UNIX proprietary information – which SCO owns – whether found in Linux or elsewhere, see id.; and
does not inform the Court that SCO has licensed the right to use UNIX proprietary information in Linux in connection with the compromise of potential claims based on the literal or non-literal copying or other wrongful misappropriation of SCO’s UNIX proprietary information into Linux, see id.
IBM’s motion thus reduces to the indefensible claim that SCO may not license its own proprietary information, and may not settle or release valid claims in connection with that SCO-owned information, without incurring massive liability for somehow copying IBM code in Linux. IBM presents this theory as sufficiently serious and urgent that it warrants allowing IBM to accelerate adjudication while IBM also holds back discovery, returning the litigation to the pre-June 10 status.
3. IBM’s Motion for Summary Judgment on Its Claim That It Was Contractually Permitted to Copy and Distribute UNIX Material without Restriction. IBM also has moved for partial summary judgment on all of the contract claims in the case. IBM’s motion asks the Court to read the contracts as “giving” SCO far less protection for UNIX than SCO would have been able to claim without any contract, so that under IBM’s view SCO would have been better off with no contract at all. That motion, among other things:
claims that IBM’s and Sequent’s contracts with SCO’s predecessor AT&T allowed them to freely appropriate from UNIX so long as they did not literally copy UNIX code, see Part II.A.2, below;
claims that IBM only seeks protection for “homegrown” and “original” IBM work, but in fact asks the Court to interpret “homegrown” and “original” to include what even statutory and common law would call “copying” and “misappropriation,” see id.;
therefore claims that the licensor thus used its contracts to destroy the protections and claims it already had for its seminal innovation before entering into those agreements, see id.;
therefore claims that SCO’s predecessor AT&T gave IBM and Sequent the extraordinary gift of being able to exploit uncontrollably the seminal innovation embodied in UNIX – so long as IBM and Sequent made an easily evadable and thus empty commitment to AT&T, see id.;
asks the Court to disregard the plain language common to AT&T’s standard contracts, which prohibits the disclosure of “derivatives,” “modifications” and “methods or concepts” – terms specifically designed to support claims protecting proprietary information from misappropriation even when there is no literal copying of such information, see Part II.A.3, below;
asks the Court to disregard the plain language of the standard contracts, which required IBM and Sequent to treat each successive derivative or modification as if it were part of and subject to the very same protections as UNIX, thereby controlling the use and preventing the disclosure of the results of each covered step of the licensee’s UNIX-dependent program development, see id.;
asks the Court to disregard the plain language of the standard contracts, which do not contain any language limiting the licensor’s protections to the literal copying of code, see id.;
asks the Court to treat at least one key provision not merely as surplusage (which itself is improper), but as literally wrong – where simply inserting the word “not” would have accomplished what IBM claims was everyone’s intent, see id.;
thereby asks the Court to disregard these and other ways in which the license agreements did not reduce, but increased the licensor’s preexisting protections – as they had to given the increased risks the licensor was taking by granting the licensee permission to rely on the seminal UNIX innovation as the basis for its products, see id.;
asks the Court to rely on extrinsic evidence, even though black-letter law precludes summary judgment when the moving party must rely on such evidence, see Part II.A.4.a, below;
asks the Court to rely critically on the testimony of witnesses that in prior sworn testimony contradicted themselves on the key issues IBM places before the Court – testimony given long before IBM came to represent them, see Part II.A.4.b, below;
asks the Court to credit IBM’s current version of history as a matter of law while failing to disclose to the Court that IBM knows about its own witnesses’ directly inconsistent prior testimony, see id.;
asks the Court to credit its witnesses’ interpretation of the legal protections afforded to AT&T’s intellectual property when the AT&T attorney who was primarily responsible for supervising the drafting and enforcing of those protections – a witness who has provided IBM with a declaration that IBM has omitted from its summary judgment submission – has provided a sworn declaration that directly undermines, in numerous detailed ways, IBM’s reading of the contract, see id.;
mischaracterizes SCO’s contract theories by disregarding submissions SCO has made to the Magistrate Judge, setting forth theories IBM tells this Court SCO does not have, see Part II.A.1, below; and
supports its claim that SCO does not have such theories by pointing this Court to the fact that SCO has not yet produced certain types of evidence, while arguing to the Magistrate Judge that SCO should not be given access to that very same evidence because it is “wholly irrelevant,” see id.
It is on the basis of these extraordinary positions that IBM asks the Court to return the litigation to its pre-June 10 status and force SCO to waste its key depositions while IBM succeeds in holding back all of the discovery SCO has been seeking and which SCO is clearly entitled to receive.
IBM Continues to Refuse to Produce Discovery Relevant to Its Motions
By refusing to permit the Amended Scheduling Order’s discovery process to go forward, IBM has precluded SCO from obtaining discovery on numerous issues, including those bearing directly on IBM’s dispositive motions. In its July 9 Opposition Memorandum, for example, SCO has demonstrated that it has not had the opportunity to uncover a substantial amount of the evidence with which SCO would oppose IBM’s motion for summary judgment on the counterclaim in which IBM asks the Court to declare that no material ever contributed to Linux has violated any SCO copyright.
Similarly, IBM seeks to support its motion for summary judgment on SCO’s contract claims by alluding to SCO’s discovery responses and contending that SCO fails to offer any evidence that IBM’s contributions to Linux are “modifications or derivative works of any code in UNIX System V.” Yet the contract protects SCO from IBM disclosure and uncontrolled exploitation of any of its successive modifications or derivatives (all of which are defined as “resulting materials” and all of which are given the same protection from disclosure and uncontrolled exploitation as UNIX itself). IBM has nevertheless blocked the discovery SCO needs about this iterative development process to provide the specific responses whose absence IBM now seeks to exploit. In filings in both May and July, SCO has shown that only upon IBM’s production of specified discovery can SCO in any reasonable time frame – and with minimal efficiency – identify the evidence showing that IBM has made contributions to Linux through the very kinds of use and dependence on UNIX derivatives and modifications that the contracts protect from the very kind of disclosure and uncontrolled exploitation that IBM’s contributions to Linux involved. Yet when SCO tries to obtain the evidence to make that showing, IBM – having focused this Court on the absence of such evidence – tells the Magistrate Judge that the discovery should be blocked because such evidence is “wholly irrelevant.”
Over a year ago, moreover, SCO asked IBM to identify the programmers who have made contributions to AIX and Dynix and their specific contributions. In addition, the Magistrate Court ordered IBM to produce that information over five months ago, in its March Order. Yet IBM has failed to provide any of that information. IBM has also failed to produce early versions of AIX, which discovery is necessary to permit SCO to trace the progression of the kinds of dependence on UNIX System V material through AIX and Dynix that the contract envisioned and expressly prohibited licensees from exploiting, among its other protections. Without that information – which IBM does not dispute it has failed to provide – SCO could not depose the AIX and Dynix programmers in a remotely complete way. As one rudimentary example, SCO is plainly entitled to ask programmers questions about the code they have contributed to AIX and Dynix; without the early versions of AIX, SCO cannot sensibly depose any programmer about the contributions he made to AIX – because SCO will not even have the contributions in hand.
The longer IBM holds back the foregoing rudimentary, predicate information, the greater IBM will compromise SCO’s ability to use it effectively as a basis for further discovery and the development of its case. SCO has made this clear to IBM, and this is why even a deposition plan cannot be formed without disputes about such rudimentary discovery first being resolved. This is plainly true in the case, for example, of SCO’s ability effectively to take key contract depositions. With respect to IBM’s contracts motion, for example, only two of the witnesses whose testimony IBM presents have been cross-examined. At that time – when they were still immersed in the day-to-day issues and before their contact with IBM or its counsel – they testified to the reasons and justifications for the broad scope of the agreements, including the importance of the programming process by which the licensees would be developing products modified and derived from UNIX.
In preparing to take the depositions of the declarants on whom IBM relies, SCO is entitled first to take the depositions of the AIX and Dynix programmers, to show that the very risks inherent in the programming process on which IBM’s declarants focused before IBM came to represent them – and when they contradicted their present positions – were also concretely present in the programming processes at issue in this case. In other words, IBM improperly seeks to preclude SCO from cross-examining IBM’s summary judgment declarants with the very information that is the express basis of the contradictory sworn testimony of the only two declarants who were cross-examined regarding the scope of the license agreements. SCO’s capacity effectively to depose the foregoing contract witnesses depends on whether SCO can concretely show and remind these witnesses of the types of risks that concerned AT&T and underlie the license agreements. Moreover, increased risks to the licensor from the way the programming process works greatly inform why the increased contract protections were fashioned in the manner the plain contractual language reflects, in the manner IBM’s declarants previously explained, and in the manner IBM now denies. Because IBM relies on extrinsic evidence in supporting its position – not least on declarants who previously themselves relied on facts about the programming process when providing contradictory testimony – IBM’s refusal to allow discovery about that very same programming process while trying to force accelerated adjudication of contract interpretation issues is in an inexplicable and extraordinary position.
The multiple categories of documents and information that SCO has long sought not only are necessary predicates to the depositions, but would also bear directly on IBM’s motions. IBM continues to refuse to produce, for example, the AIX and Dynix revision information available on IBM’s Configuration Management Version Control (“CMVC”) System and Sequent’s Revision Control System (“RCS”). These are systems specifically designed to store all of the development history of AIX and Dynix, which would show step by step how the dependence of AIX and Dynix on UNIX was compounded over time in just the manner the contract protected from uncontrolled exploitation, by protecting against unrestricted use and disclosure of the results. That step-by-step process can only be seen through the program’s early history, and then only through examination of the development process itself – both of which IBM seeks to block. The same storage systems also contain programming notes, design documents and white papers. SCO has previously explained that those materials could readily contain admissions concerning the meaning of, and limitations imposed by, the license agreements; admissions regarding IBM’s liability for breaching those agreements; and admissions that the development of AIX and Dynix depended on UNIX System V. Under any theory of the agreements, the materials are relevant because they would permit SCO to test the validity of IBM’s claim that at some point AIX and Dynix ceased to be modifications or derivative works “based upon” UNIX System V and were no longer subject to the restrictions of the license agreements.
Even more basically, those documents would logically include programmers’ acknowledgements of the constraints they faced upon exposure to UNIX materials. SCO first requested all of this material on June 24, 2003. Because IBM did not produce even a single version of its AIX program until over a year into the case, SCO was not able to demonstrate its need for this additional discovery pursuant to the Magistrate Judge’s procedure until this summer. SCO will now not receive this discovery until sometime this fall, and will not be able to begin to analyze it and use until that time.
SCO respectfully submits that IBM should not be permitted to force the adjudication of its already premature and fact-intensive dispositive motions while also refusing to produce predicate discovery such as the foregoing. For all of these reasons, and as set forth in detail below, SCO asks the Court to defer the scheduling and adjudication of any dispositive motion until after the end of the Court’s fact-discovery period, which is now February 11, 2005.
I. UNDER ITS WELL-ESTABLISHED DISCRETION, THE COURT SHOULD DEFER BRIEFING ON IBM’S MOTIONS IN ORDER TO ENFORCE THE REVISED SCHEDULING ORDER AND TO PROMOTE THE FAIR AND EFFICIENT ADMINISTRATION OF THE CASE
It is black-letter law that a Court faced with a summary judgment motion may, even apart from its authority under Rule 56(f), “exercise its discretion under Rule 56” to “postpone consideration of the motion.” Charles A. Wright et al., Federal Practice and Procedure § 2728 (2004). The Court’s “discretion to postpone consideration of the motion for summary judgment” stems from a recognition that the “principles governing summary judgment procedure should be applied in a common sense manner to the realities of the litigation at hand.” Williams v. Howard Johnson’s, 323 F.2d 102, 106 (4th Cir. 1963); see also Young Enters., Inc. v. United States, 26 Cl. Ct. 858, 863 (Cl. Ct. 1992) (on summary judgment, the court “may also postpone a decision until it can be founded on a more complete factual record”); see, e.g., Adams v. R.R. Donnelley & Sons, Nos. 98 C 4025, 96 C 7717, 2001 WL 336830, at *7 (N.D. Ill. Apr. 6, 2001) (Exh. D) (deferring briefing and consideration of defendant’s summary judgment motion where plaintiffs reasonably had not taken discovery relevant to the motion). Indeed, the district court may even “deny summary judgment as a matter of discretion.” Toyoshima Corp. v. Gen. Footwear, 88 F.R.D. 559, 560 (S.D.N.Y. 1980).
Here, the Court should exercise its well-established discretion to defer consideration of IBM’s summary judgment motions until after fact discovery is completed in this case. For the reasons detailed further herein, it is clear that even under a plenary review of the relevant facts, consideration of IBM’s motions at this juncture would impede the fair and efficient administration of this litigation by failing to apply the summary judgment procedure “in a common sense manner to the realities of the litigation at hand.” Williams, 323 F.2d at 106. This case, however, presents two unique additional circumstances that strongly counsel in favor of the Court’s exercise of its discretion to defer consideration of IBM’s motions.
First, in June, this Court granted SCO’s motion to extend the discovery period in this case, over IBM’s strident objection, until February 10. IBM’s coordinated filing of three fact-intensive summary judgment motions now threatens to undo that Order, in several respects:
In filing the motions now, IBM proceeds just as if the Court had not entered the Amended Scheduling Order at all. IBM has produced no discovery since June 10. The result IBM sought to impose on SCO in unsuccessfully opposing the entry of the Amended Scheduling Order is the exact same result IBM seeks to impose through its recent motions – namely, to have the Court adjudicate core issues before IBM has permitted the relevant facts to see the light of day.
IBM filed its two most recent motions after having seen precisely the type of burdensome and time-consuming response to such motions SCO must make, given the fact-intensive nature of the motions and the limited discovery to date. SCO explained in great detail in opposition to IBM’s first motion for summary judgment, on IBM’s Tenth Counterclaim, that IBM’s failure to produce long-sought, predicate discovery made it impossible for SCO to respond to the motion in full on the merits. See SCO’s Mem. in Opposition to IBM’s Motion for Summary Judgment on Its Tenth Counterclaim (July 9, 2004). IBM’s motions thus threaten SCO’s ability to use the Amended Scheduling Order’s extended discovery period to take discovery (rather than divert resources to still more burdensome litigation over SCO’s indisputable right to do so).
IBM filed the motions during the pendency of SCO’s motions seeking to require IBM to produce rudimentary predicate discovery that pertains to the basic issues in the case, including (of course) IBM’s motions. In contrast even to its first summary judgment motion – which IBM in any event filed having failed to produce the rudimentary discovery at issue – IBM filed its most recent motions after SCO moved again to compel IBM to produce that discovery, and thus seeks to preempt SCO’s efforts to obtain that discovery. See SCO’s Mem. in Support of Its Renewed Motion to Compel (July 6, 2004) & Reply Mem. in Support of Renewed Motion to Compel (Aug. 26, 2004).
IBM timed the motions not only to take advantage of its refusal to produce, and to preempt SCO’s efforts to obtain, that discovery, but also to coincide with the hearing date on SCO’s pending discovery motions. Having argued for and received from SCO a mid-September hearing date on SCO’s discovery motions, IBM thereafter took care to impose mid-September return dates on the motions, going so far as to serve SCO’s counsel at his home on a Friday at 9 p.m. See Certificate of Service dated Aug. 13, 2004 (Exh. 2).
IBM has now coupled the foregoing conduct with its proposal that SCO would be entitled to “reasonable” extensions of time to respond to the motions only if SCO agrees not to invoke Rule 56(f). See Letter from T. Shaughnessy to B. Hatch dated Aug. 27, 2004 (Exh. 1). In that additional respect, IBM seeks to leverage its motions to nullify the right that any litigant has during fact discovery – namely, to demonstrate to the Court what further discovery it needs and expects to find to support its claims and to defend against its adversary’s claims – and thereby to further undermine the Court’s extended period for fact discovery.
Second, while an early dispositive motion may in certain cases facilitate the orderly preparation of the case by eliminating or focusing issues, in light of the timing of IBM’s motions, there are no potential benefits to be gained from allowing IBM’s fact-intensive motions to be accelerated no less than six months ahead of the close of the fact-discovery cut-off. The accelerated schedule IBM seeks to impose could not streamline or avoid discovery burdens or create any other administrative efficiencies in this case. This is plain wholly apart from the question of whether IBM should be able to force SCO to go forward with key depositions now or otherwise to litigate motions now without yet having secured the benefit of further discovery.
There is no reason to expect that on any even minimally reasonable schedule IBM’s motions could be decided before the end of the year (if then). Prior to that time, SCO’s need for discovery could not be held in suspension. To the contrary, SCO would need, and would be entitled, to secure discovery, for example, not only to rebut IBM’s extrinsic evidence with respect to the interpretation of the contracts, but also to establish IBM’s violations of the contracts. If IBM’s contract motion were at that time resolved in SCO’s favor, SCO could not begin such discovery only then – that course would leave SCO only weeks to accomplish what the Amended Scheduling Order contemplated would take eight months. Instead, to protect SCO’s discovery rights, SCO would need to proceed with all of this discovery during the pendency of the motions, as IBM would plan to do with its respective discovery. (In late August, for example, IBM noticed for September, October, and November over twenty depositions.)
In addition, the discovery the parties are entitled to continue to take through the close of fact discovery would itself bear on the merits of IBM’s motions. At the very least, SCO would be compelled to summarize and present that evidence to the Court at the time of the hearing on IBM’s motions (and, if appropriate, thereafter). That evidence would be relevant whether or not the parties have “fully briefed” the motions well before the hearing, and whether or not SCO has formally moved under Rule 56(f) in response to IBM’s motions.
By setting the schedule for dispositive motions – especially fact-intensive motions such as those at issue here – after the close of fact-discovery, the Amended Scheduling Order avoids such series of supplemental submissions. IBM’s approach not only does not avoid them, but essentially guarantees such inefficiencies: By forcing fact-intensive dispositive motions to be adjudicated no less than six months before the end of fact discovery, IBM presents the Court with a constantly moving target; IBM then aggravates the problem it created by holding back discovery – further increasing the likelihood that SCO will receive it late in the day and at a point when it will need to supplement its briefs on IBM’s needlessly accelerated dispositive motions.
Accordingly, in at least the two foregoing respects alone, a common sense regard for the realities of this litigation – as well as considerations of basic fairness and orderly procedure –militate in favor of deferring consideration of IBM’s summary judgment motions until after the February 11 fact-discovery cut-off. IBM’s foregoing conduct might not (at least not plainly) undercut the Amended Scheduling Order or impede the fair and efficient administration of this case if IBM were at all likely to prevail on its motions, if IBM had properly withheld the discovery at issue, or if the discovery SCO has long sought and continues to seek were collateral to the motions. But the opposite is true here.
SCO shows in Part II, below, that even cursory scrutiny of IBM’s dispositive motions reveals that they are overtly fact-intensive, premature, and meritless. SCO further shows in Part III, below, that the discovery that is currently at issue before the Magistrate Court not only is necessary predicate discovery to further, targeted discovery relevant to the motions, but also is itself directly relevant to the motions – and that IBM does not even dispute its failure to produce crucial parts of that discovery. Thus, by timing its motions as it has, IBM would impede the Court’s evaluation of the (supposed) merits of IBM’s motions (see Part II), and undercut the Court’s administration of the case (see Part III). SCO therefore submits, and further demonstrates in Parts II and III, below, that the Court should exercise its well-established discretion to defer its consideration of IBM’s summary judgment motions until after fact discovery in this case.
II. EVEN BRIEF SCRUTINY OF IBM’S SUMMARY JUDGMENT MOTIONS REVEALS THAT THEY ARE PREMATURE AND MERITLESS BECAUSE THEY REST ON DISPUTED FACT-INTENSIVE ARGUMENTS
Even a brief review of IBM’s summary judgment motions – which overtly rely on numerous fact-intensive arguments – reveals that those motions are both premature and meritless. Because SCO has already submitted full briefing demonstrating the serious factual and legal deficiencies in IBM’s summary judgment motion on its Tenth Counterclaim, this memorandum focuses on IBM’s two most recent motions, which concern SCO’s contract claims and IBM’s Eighth Counterclaim. SCO focuses particularly on IBM’s contract motion because its fact-intensive approach puts squarely at issue facts that are the subject of discovery IBM has improperly refused to produce for over a year.
A. IBM’s Fact-Intensive Summary Judgment Motion on SCO’s
Contract Claims Is Demonstrably Meritless and Premature
1. IBM’s Motion Depends Upon Its Mischaracterization of
SCO’s Contract Claims
As an initial matter, IBM’s contract motion rests on a significant mischaracterization of SCO’s contract claims. IBM inaccurately asserts that those claims “rest entirely on the proposition that ‘[t]he AIX work as a whole and the Dynix/ptx work as a whole are modifications of, or are derived from [UNIX] System V.’” IBM Contract Mem. ¶ 62. That is false. SCO does maintain that AIX and Dynix are subject to the restrictions of the licensing agreements because they are derivatives of UNIX, and consequently that IBM breached the license agreements by contributing to Linux any portion of those contractually-protected derivatives. As discussed below, this position is well-supported by both a plain reading of the license agreements and the prior sworn testimony of IBM’s own witnesses. See Parts II.A.3-4, below. Moreover, IBM and Sequent have repeatedly referred to each of AIX and Dynix, respectively, as derivatives of UNIX. See, e.g., Computerworld: IBM’s Mills sets sights on middleware, Linux, October 24, 2003 (Exh. 3) (“We took the Unix System V kernel and we made modifications”); Email from Bill Sandve to Kim Tran, January 22, 2002 at 2 (Exh. 4) (“AIX was derived from System V.”); Agreement for Licensing of AIX Source Code and Related Products between Argus Systems Group, Inc. and IBM ¶ 8.4 (Exh. 5) (“AIX is derived form software under license from SCO.”); International Business Machines Corp. Royalty Statement, June 30, 1987 (Exh. 6) (stating that AIX is “derived from System V”); IBM/Supplier Technical Services Agreement at 1 (Exh. 7) (“‘AIX Operating System’ or ‘AIX’ shall mean the UNIX operating system that operates on the Power, Power PC, Power 2, Power 3, and Power 4 architectures or derivatives or follow-on architectures irrespective of the names of such architecture.” (emphasis added)); Rodgers Dep. 138 (Exh. 8) (“Dynix/ptx is almost certainly a derivative work of Unix System V.”).
Nevertheless, contrary to IBM’s suggestion, SCO’s contract claims do not, by any stretch, “rest entirely” on the above-described theory of IBM’s contract liability. Rather, SCO advances a distinct, and additional, basis for IBM’s liability under the license agreements, which focuses specifically on the particular code contributions that IBM has made to Linux. SCO seeks to demonstrate in this case that IBM violated the license agreements by making particular contributions to Linux of code sections that are themselves UNIX modifications or derivatives. See, e.g., SCO’s Reply Mem. Re Discovery (July 12, 2004) (“Reply Mem. re Disc.”), at 11-14.
SCO has repeatedly described these different theories to IBM in numerous memoranda and discovery requests – including, recently, SCO’s July 12 Reply Mem. Re Discovery – but IBM continues to ignore SCO’s second theory of liability. IBM’s refusal to acknowledge this theory is remarkable, and convenient, because, as detailed below, IBM has consistently blocked the very discovery that IBM needs to pursue this second theory of contract liability. See Part III.B, below. As SCO has explained to IBM, SCO’s second theory requires a close tracing of the development of IBM’s code contributions from UNIX System V code through AIX and Dynix into Linux. As detailed in Part III, below, however, IBM has now been actively resisting for more than a year.
Indeed, in resisting the discovery that would permit SCO to show the UNIX origins of the AIX and Dynix code that IBM has dumped into Linux, IBM has told Magistrate Judge Wells that such discovery is irrelevant to the issues in this case. See IBM’s Opp. to SCO’s Renewed Mot. to Compel, at 9. Now, however, IBM seeks summary judgment from this Court on IBM’s contract claims, and relies on SCO’s failure to produce the very evidence that such discovery would provide in order to support its inaccurate assertion that SCO’s contract claims must “instead rest entirely” on its theory that AIX and Dynix are, in their entireties, modifications or derivatives of UNIX System V. IBM Contract Mem. ¶¶ 57-62. In this circular fashion, IBM’s mischaracterization of SCO’s contract claims thus seeks to exploit IBM’s failure to produce critical discovery and to expedite resolution of the key issues in this case before IBM is required to produce such discovery.
2. IBM’s Claim That the UNIX License Agreements Actually Reduced
AT&T’s Legal Rights By Giving IBM the Unfettered
Rights to Own and Control Such Copied Work Is Absurd.
In addition to mischaracterizing SCO’s contract theory, however, IBM attempts to advance (and prevail as a matter of law on) a contract theory of its own – a reading of the license agreements under which IBM and Sequent were free to do anything with any modifications or derivatives of the licensed UNIX product so long as those modifications or derivatives did not literally copy UNIX code. For the reasons discussed below, that contract interpretation defies economic logic, directly contradicts the plain language of the license agreements themselves, and finds support only in declarations that are either untested or have been totally undermined by flatly inconsistent prior testimony and contemporaneous writings of IBM’s witnesses.
In attempting to justify its proposed contract interpretation, IBM also mischaracterizes its own position. IBM insists that it seeks only to preserve ownership and control of IBM’s purely original work, casting SCO as contending that SCO should own and control what can “indisputably” be shown to be IBM’s so-called “homegrown” work. IBM Contract Mem. at 2. Divorced from its rhetoric, however, IBM’s argument reduces to an incredible claim that under the very license agreements whose purpose was to protect the UNIX intellectual property, IBM obtained the unfettered rights to own and to control work that IBM copied or derived from UNIX.
In other words, IBM contends that the agreements protected the licensor against only the literal copying of its code: IBM expressly defines “homegrown” code as “source code that is written by IBM and does not include UNIX System V code.” IBM Contract Mem. at 2. What IBM calls “homegrown” code turns out not be “homegrown” at all. Instead, that material includes – at a minimum – every non-literal copy IBM ever made of any form of UNIX material.
IBM’s argument is baseless. SCO shows in Parts II.A.2-3, below, that the agreements clearly and expressly (1) protect against the misappropriation and unauthorized disclosure of “modifications” and “derivative works” – without any showing of literal copying; (2) protect against misappropriation of UNIX “methods or concepts” – without any showing of literal copying; and (3) do not once mention any limitation on the licensor’s protection to protection against only literal copying.
Moreover, IBM’s claimed right to make and appropriate for itself copies of SCO’s code, so long as the copies are not literal copies, is one that IBM would not even have under the copyright laws. Basic copyright law has long recognized that copying can occur through the making of “derivatives” without any literal copying at all. Moreover, the common law not only protects “methods or concepts,” but also extends farther, beyond trade secrets, and far beyond literal copying. IBM’s claim thus reduces to the untenable proposition that under the terms of the license agreements, IBM can do what copyright and common law would forbid: make non-literal copies of another’s work and innovation, and thereby misappropriate that innovation for itself.
According to IBM’s argument, the parties to the license agreements struck an absurd (and, to use IBM’s terminology, “commercially unreasonable”) deal. IBM sought – and, by virtue of the license agreements, received – the benefit of using AT&T’s broadly-used (and customer-demanded) innovation UNIX to build IBM’s own self-described “flavor” and “modification” of AT&T’s product, so that IBM did not have to incur the expense and risk of developing its product from scratch. In order to confer this substantial benefit on IBM, AT&T had to allow IBM to immerse its engineers in AT&T’s proprietary and confidential material, which created heightened dangers of misappropriation, whether literal or non-literal, and which required AT&T to fashion suitably heightened contractual protections for its intellectual property. AT&T thus sought to protect itself against those dangers by obtaining, through contract, protections beyond that which were already afforded to it by copyright and common law. Indeed, the actual concerns underlying, and intent of, AT&T’s UNIX license agreements were aptly described by David Frasure – one of the declarants on whom IBM heavily relies in its motion – in his prior sworn testimony:
“if you’ve got a group of people over here developing a set of source code and they have never seen the Unix source code, they’ve never been too exposed to it and the[y] develop a product completely on their own, then that’s one thing. But if they’re developing a product with the benefit of Unix or perhaps they have used it for – for a number of years, ten years, and then they think they’re going to go off and develop something on their own that’s an operating system that may look like Unix, we had – we expressed our concern that – that we had an interest in that product.” Frasure Dep. (12/8/92) at 20 (Exh. 9) (emphasis added).
Yet IBM now contends that when AT&T gave IBM this benefit and exposed itself to these heightened risks, AT&T did not receive proportionally greater protection. To the contrary, IBM claims that AT&T received far less protection for its innovation than if IBM had not received the benefit – if IBM had been forced to innovate for itself. IBM claims that was the deal, even though it makes no economic sense; even though (as explained further in the next section) the key provisions of the AT&T contracts include clear and unambiguous language that had and has no recognized function except to protect against non-literal copying, and no language that restricts protection to literal copying; and even though (as explained in Part II.A.4, below) substantial extrinsic evidence developed thus far compels the opposite conclusion. SCO thus submits that apart from the plain contractual language and the extrinsic evidence that undermines IBM’s position, IBM’s remarkable contractual interpretation warrants at least full factual exposition of the circumstances under which the parties contracted and each party’s intent in contracting.
3. The Plain and Unambiguous Language of the License
Agreements Directly Contradicts IBM’s Interpretation.
In an attempt to support its economically irrational reading of the license agreements, IBM purports to rely on the plain language of AT&T’s standard license agreement and a side letter agreement that AT&T entered into with IBM (the “IBM Side Letter”) – but not with Sequent. There is no merit to IBM’s argument that it is entitled to summary judgment on the ground that these documents unambiguously resolve, in its favor, the issues on both contracts.
Section 2.01 of both license agreements expressly grants to each licensee “a personal, nontransferable and nonexclusive right to use in the United States each SOFTWARE PRODUCT identified in the one or more Supplements hereto solely for LICENSEE’S own internal business purposes and solely on or in conjunction with DESIGNATED CPUs for such SOFTWARE PRODUCT.” It is undisputed that the “SOFTWARE PRODUCT” includes UNIX System V. Moreover, Section 2.01 further provides: “Such right to use includes the right to modify such SOFTWARE PRODUCT and to prepare derivative works based on such SOFTWARE PRODUCT, provided the resulting materials are treated hereunder as part of the original SOFTWARE PRODUCT.” (Emphasis added). Thus, in direct contrast to IBM’s claim that the agreements protect against only a licensee’s literal copying of IBM source code, Section 2.01 unambiguously protects any modifications or derivatives of UNIX System V.
By requiring that “resulting materials” be treated “as part of” UNIX for purposes of the contract, Section 2.01 expressly and unambiguously precludes IBM’s claim that the agreements protect against only a licensee’s literal copying of UNIX source code. Section 2.01 protects the “materials” that “result” from (1) the licensee’s exercise of its “right to modify” the original “SOFTWARE PRODUCT” (which includes the UNIX System V source code) and (2) the licensee’s right “to prepare derivative works based on such SOFTWARE PRODUCT.” IBM’s proposed construction of the license agreements violates each of the well-settled principles of contract interpretation requiring (1) that no contractual provisions be rendered superfluous; and (2) that contractual provisions are not read to internally conflict with each other. See Merrill Lynch v. Adler, 651 N.Y.S.2d 38, 39 (App. Div. 1996) (“an agreement should be read to give effect to all of its provisions and to render them consistent with each other”) (emphasis added); Ruttenberg v. Davidge Data Sys. Corp., 626 N.Y.S.2d 174, 178 (App. Div. 1995) (citing the “rule of construction that a contract shall be read so as to give effect to each and every term”); Robshaw v. Health Mgmt., Inc., 470 N.Y.S.2d 226, 227 (App. Div. 1983) (employing “general rule that a contract should be interpreted to avoid inconsistencies and to give meaning to all of the terms” (citing 22 N.Y. Jur. 2d, Contracts §§ 221, 222)); Garza v. Marine Transport Lines, Inc., 861 F.2d 23, 27 (2d. Cir. 1988) (suggested interpretation of a contract that would “render at least one clause superfluous or meaningless” should be “avoided if possible”).
If IBM’s interpretation were correct, then the requirement that “resulting materials” be “treated as part of the original SOFTWARE PRODUCT” would be rendered superfluous, because UNIX source code is already clearly and indisputably contained within the original SOFTWARE PRODUCT. In other words, there would simply be no reason for the contract to treat source code contained in “resulting materials” “as part of the original SOFTWARE PRODUCT” if such source code is already automatically included in the original SOFTWARE PRODUCT.
Furthermore, IBM’s “source code only” argument is patently wrong for the additional reason that it is incompatible with the Section 2.01 language requiring derivatives or modifications to be treated “as part of” the original SOFTWARE PRODUCT. By definition, treating “resulting materials” – “modifications” and “derivatives” – “as part of the original SOFTWARE PRODUCT” must mean adding something to the “original SOFTWARE PRODUCT” that was not already there. In other words, since source code is already contained in the “original SOFTWARE PRODUCT,” then IBM’s reading of Section 2.01 (i.e., that it covers only UNIX source code) would provide nothing to add (i.e., nothing to treat “as part of”).
Even beyond the plain meaning of Section 2.01, Section 7.06(a) of the operative IBM and Sequent license agreements each include a specific provision requiring the licensee to hold the “SOFTWARE PRODUCTS subject to this Agreement in confidence for AT&T.” The Sequent agreement further specifies that the licensee “shall not make any disclosure of any or all of such SOFTWARE PRODUCTS (including methods or concepts utilized therein) to anyone, except to employees of LICENSEE to whom such disclosure is necessary to the use for which rights are granted hereunder.” Id. (emphasis added).
Thus, Section 7.06(a) of the license agreements not only expressly extended the requirement of confidentiality to all aspects of the “SOFTWARE PRODUCTS” (which, by virtue of Section 2.01, were treated identically to any modifications or derivatives based on such SOFTWARE PRODUCTS), but also specified (in the case of the Sequent agreement) that protection of the SOFTWARE PRODUCTS included protection of the “methods or concepts utilized therein.” Again, a plain reading of these provisions is directly at odds with the tortured construction advanced in IBM’s dispositive motion.
IBM tries to avoid the plain language of Sections 2.01 and 7.06 by claiming that a side letter that it entered into with IBM (but not Sequent) somehow nullified both of these cornerstone contractual provisions (both for itself and for Sequent and by extension all of the other thousands of UNIX licensees). Moreover, IBM contends that its side letter accomplished this drastic change merely by “clarifying” the agreements. IBM Contract Mem. at 54. IBM’s misuse of its side letter is indefensible, for several reasons.
First, in suggesting that the IBM Side Letter limited the protections of Section 2.01 to source code only, IBM ignores entirely that the plain language of the provision on which it relies concerned the ownership of derivatives and modifications, and not the contractual protections that were afforded the licensor concerning their use and disclosure: “Regarding Section 2.01, we agree that modifications and derivative works prepared by or for you are owned by you. However, ownership of any portion or portions of SOFTWARE PRODUCTS included in any such modification or derivative work remains with us.” (Emphasis added). Plainly, this language did not purport to compromise – and did not in fact compromise – AT&T’s right to control the use and disclosure, among other things, of such modifications or derivative works, as clearly set forth in the license agreements – particularly, Sections 2.01 and 7.06(a). As the licensor, AT&T acknowledged in the IBM Side Letter that it would not claim ownership of modifications or derivative works that had been prepared by a licensee, but it never agreed to give away its right to protect its intellectual property by ensuring that licensees would not use or disclose works that were copied (even non-literally) from its innovations.
In light of the fact that the IBM Side Letter left entirely in place the existing language in Section 2.01, IBM’s suggestion that the “Regarding 2.01” clause of the IBM Side Letter somehow “clarified” out of existence most of the contractual protections afforded by the UNIX license agreements is particularly incomprehensible. As discussed above, Section 2.01 plainly covers far more than mere literal source code, and, for the reasons previously discussed, IBM’s “source code only” reading of Section 2.01 renders the express contractual protection requiring “resulting materials” to be “treated as part of the original SOFTWARE PRODUCT” both superfluous and illogical.
Moreover, if the IBM Side Letter were intended to cut back the Section 2.01 protections to mere literal source code, as IBM contends, that would have been exceedingly easy to accomplish as a drafting matter. Instead of leaving the “resulting materials” clause of Section 2.01 entirely in place, the IBM Side Letter could have simply added the word “not” before “treated as part of the original SOFTWARE PRODUCT.” With that simple change, IBM could have effected the very change that it now reads into the side letter – economically, easily, and clearly. Without such a change, however, the plain language of the license agreements precludes the retroactive change that IBM is attempting to make.
IBM similarly relies on a misinterpretation of its side letter when it suggests that the amendment to the language of Section 7.06(a) somehow “clarified” all of the licensor’s contractual protections out of existence. As with IBM’s other side-letter argument, however, this claim contains the fatal flaw that the IBM Side Letter did not modify – or purport to modify – any of the restrictions contained in Section 2.01. In addition, the IBM Side Letter specifically reiterated that IBM was required to accord confidential treatment to all SOFTWARE PRODUCTS – and, therefore, pursuant to Section 2.01, all modifications and derivatives that were “based on” such SOFTWARE PRODUCTS. Accordingly, although the IBM Side Letter acknowledged that IBM could engage in “developing and marketing products or services employing ideas, concepts, know-how or techniques relating to data processing embodied in SOFTWARE PRODUCTS,” there is no merit to IBM’s suggestion that the IBM Side Letter somehow unambiguously removes, even for just IBM only, the license agreements’ use and disclosure restrictions on far more than just literally copied code.
Finally, and wholly apart from the absence of textual support for its sweeping (re)construction of the IBM Side Letter, IBM’s argument concerning the side letter would not entitle it to summary judgment for the independently-sufficient reason that the side letter did not cover Sequent. In fact, although the Sequent license agreement was entered into just a couple of months after the IBM agreement and side letter, none of the language on which IBM (incorrectly) relies from its side letter was included (by side letter or otherwise) in the Sequent agreement. Nor is there any support in the contractual language for IBM’s suggestion that Sequent and other licensees somehow received the benefit of such “clarifications” – and in the case of Sequent, even previously-executed “clarifications” – even though no such language had been included in their written contracts. IBM Contract Mem. at 53-54. To the contrary, each of the UNIX license agreements (including the Sequent agreement) contained clear and unambiguous integration language that specifically ruled out after-the-fact “me too” arguments such as those advanced here by IBM (on Sequent’s behalf):
“This Agreement and its Supplements set forth the entire agreement and understanding between the parties as to the subject matter hereof and merge all prior discussions between them, and neither of the parties shall be bound by any conditions, definitions, warranties, understandings or representations with respect to such subject matter other than as expressly provided herein or as duly set forth on or subsequent to the date of acceptance hereof in writing and signed by a proper and duly authorized representative of the party to be bound thereby. No provision appearing on any form originated by LICENSEE shall be applicable unless such provision is expressly accepted in writing by an authorized representative of AT&T.” Exh. 13, ¶ 4; Exh. 14, ¶ 4 (emphasis added).
Notably, IBM ignores entirely this integration clause in its strained attempt to transfer its misreading of the IBM Side Letter over to the Sequent license agreement.
4. IBM’s Reliance on Extrinsic Evidence Lacks Any Merit.
Faced with the plain language of the license agreements, IBM purports to support its contract summary judgment motion with extrinsic evidence in the form of (largely untested) witness declarations. That evidence, however, is both legally and factually incapable of supporting IBM’s motion.
In fact, the extrinsic evidence on which IBM relies is directly contradicted not only by the declaration submitted herewith from Martin Pfeffer, AT&T’s principal attorney responsible for the intellectual property protections provided by the UNIX software licenses, but also by the prior statements (including sworn testimony) of the only two IBM declarants who have been subjected to cross-examination (in a prior litigation). Even though none of this extrinsic evidence – which eviscerates IBM’s argument – is even mentioned in IBM’s brief, there can be no doubt that IBM is aware of it: IBM has obtained (but not used in its summary judgment motion) a declaration from Mr. Pfeffer, and IBM produced in discovery the transcripts of its own witnesses’ prior inconsistent statements.
Given the lack of textual support in the contracts for IBM’s interpretation, as well as the genuine issues of fact that already exist at this early stage of deposition discovery concerning the contract issues, IBM’s suggestion that those issues can be resolved in its favor as a matter of law – without any further discovery – is frivolous.
a. IBM’s reliance on extrinsic evidence precludes
summary judgment as a matter of law.
The weight of well-reasoned authority – including cases applying New York contract law – hold that “in a contract dispute, summary judgment may be granted only where the language of the contract is unambiguous.” Bouzo v. Citibank, 96 F.3d 51, 58 (2d Cir. 1996) (emphasis added). That is in part because the Court must construe ambiguous contractual terms in favor of the non-moving party; so if a contract is ambiguous, genuine issues of fact must exist. See id. Put another way:
“If the court must resort to extrinsic evidence to ascertain the correct and intended meaning of a term, material questions of fact necessarily exist. If the language is susceptible to different reasonable interpretations, and where there is relevant extrinsic evidence of the parties’ actual intent, then the contract’s meaning becomes an issue of fact precluding summary judgment.”
Alexander & Alexander Servs., Inc. v. These Certain Underwriters at Lloyd’s, London, 136 F.3d 82, 86 (2d Cir. 1998) (emphasis added). In addition to the purportedly “relevant extrinsic evidence” on which IBM so heavily relies, there is already available – even at this relatively early stage of SCO’s contract-related discovery substantial extrinsic evidence directly contradicting IBM’s interpretation of the agreement. See Part II.A.4, below.
The limited case law cited by IBM for its contention that “summary judgment is appropriate” here despite IBM’s heavy reliance on extrinsic evidence does not dictate a contrary result. See IBM Contract Mem. at 56, 66. For example, IBM selectively quotes Moncrief v. Williston Basin Interstate Pipeline Co., 174 F.3d 1150 (10th Cir. 1999), for the proposition that summary judgment may be appropriate “where [the evidence] so clearly weighs in one direction that there is no genuine issue of material fact left.” IBM Contract Mem. at 56 (citing Moncrief at 1173). But IBM ignores the unusual procedural posture of the Moncrief case: A bench trial had already been conducted on the issue of damages, during which a complete record was established on the issue of liability as well. And IBM omits to mention that it was for that specific reason – and that reason alone – that the appellate court declined to remand the contract interpretation issues to the district court, but instead ruled on those issues itself as a matter of summary judgment. See id. at 1174 (“Because of the unusual procedural posture of this issue – an issue decided on summary judgment, yet about which the parties introduced evidence and testimony at trial – we have before us all of the arguments and evidence relevant.” (emphasis added)). Indeed, had that not been the case, the Moncrief court itself recognized that in “most such cases” where the court finds contractual ambiguity, it “must remand . . . for findings regarding the interpretation of the ambiguous contract.” Id. at 1173 (emphasis added). IBM even conspicuously omits from its quotation of the Moncrief opinion, the first half of the very sentence it quotes, in which the court expressly acknowledged that it was breaking with the standard required practice (of remanding the case to the trier of fact to resolve the contract interpretation issues) because “all the extrinsic evidence is in, and where that evidence so clearly weighs in one direction that there is no genuine issue of material fact left for the district court to decide.” Id. at 1173 (emphasis added). Here, as IBM appears to recognize in selectively editing the Moncrief court’s opinion, there can be no argument that “all of the extrinsic evidence is in”; to the contrary, substantial fact discovery remains to be taken on the contract interpretation issues.
Moreover, IBM’s motion is legally unsustainable for a separate, and independently sufficient, reason: The motion relies substantially, and improperly, on the untested declarations of witnesses SCO has not yet deposed. The very New York courts on whose decisions IBM otherwise relies make clear that summary judgment cannot be granted in these circumstances. See G-I Holdings v. Baron & Budd, 213 F.R.D. 146, 149 (S.D.N.Y. 2003) (denying summary judgment on the ground that “untested affidavits are not acceptable”); Messina v. Mazzeo, 854 F. Supp. 116, 141-42 (E.D.N.Y. 1994) (holding that where non-moving party had not yet deposed important affiants, “it cannot be said with certainty that a deposition” would not “uncover material issues of disputed facts,” leading the court to rule that “summary judgment is premature” (emphasis in original)). A summary judgment motion cannot properly be founded on undeposed, untested affiants because, in considering such a motion, the Court must construe all evidence “in the light most favorable to the non-moving party.” Konikoff v. Prudential Ins. Co., 234 F.3d 92, 97 (2d. Cir. 2000). Moreover, the above rule applies with particular force here, where the few IBM declarants whose testimony has been tested through cross-examination gave sworn testimony that is directly contrary to the declarations on which IBM relies for its motion. See Part II.A.4, below. SCO has not yet deposed IBM’s untested declarants because, as explained at length above, IBM has withheld the critical predicate discovery that SCO needs to take effective depositions of the witnesses. See also Part III, below. While SCO pursues that discovery through litigation – including a litigation to enforce the Magistrate Court’s prior Order with which IBM has not complied – IBM cannot properly use premature summary judgment motions to force SCO to take depositions without necessary predicate discovery.
b. Even at this early stage of SCO’s contract-related discovery, summary
judgment is precluded by the numerous disputed issues of fact that
exist concerning the extrinsic evidence on which IBM relies
Only two of the declarants on whom IBM relies (Otis Wilson and David Frasure) have been subjected to any cross-examination. Each of those witnesses made prior statements and gave sworn testimony that directly contradict the key statements they have made in their declarations for IBM. Moreover, Messrs. Wilson and Frasure made their prior statements more than a dozen years closer to the events at issue, when the witnesses were immersed in those issues on a regular basis, and many years before IBM began representing them or paying for their representation in this case.
The prior inconsistent testimony and written statements of Mr. Wilson and Mr. Frasure not only directly undermine the credibility of the declarations that they have recently provided to IBM, but also provide clear support for SCO’s interpretation of the plain and unambiguous meaning of the license agreements. Moreover, SCO has submitted herewith the declaration of Martin Pfeffer, the AT&T attorney who had principal responsibility for overseeing the drafting and enforcement of AT&T’s intellectual property protections in its UNIX license agreements during the precise timeframe of the IBM and Sequent agreements.
Thus, to the extent that it is even relevant or admissible on the interpretation of the AT&T license agreements, the extrinsic evidence available to date both precludes summary judgment for IBM and highlights the need for further discovery so that SCO can test the declarations of IBM’s other witnesses.
The Prior Inconsistent Statements and Testimony of Otis Wilson and David Frasure
The Wilson and Frasure declarations on which IBM relies were signed within the past six months. Mr. Wilson was a former AT&T and USL employee; Mr. Frasure was his underling and is a recent felony insurance fraud convict. Frasure Dep. (12/8/92) at 116-19 (Exh. 9). Mr. Wilson is represented by IBM’s corporate counsel in this case, Wilson Dep. (12/10/92) at 7 (Exh. 15); and Mr. Frasure, whose legal bills are being paid by IBM, Frasure Dep. (12/8/92) at 114 (Exh. 9), is represented by an attorney closely associated with IBM.
IBM relies on those witness’ declarations to advance the following four claims critical to its contract motion: (1) IBM’s general claim the license agreements protected AT&T against only literal copying of actual UNIX System V source code; (2) IBM’s specific claim that the agreements did not protect anything more than source code in their coverage of modifications and derivatives of the original licensed software product; (3) IBM’s specific claim that the agreements did not protect methods and concepts embodied in the original product; and (4) IBM’s claim, based principally on its side letter, that the license agreements made no distinction between ownership rights, on the one hand, and contractual restrictions on licensees’ use, copying, distribution, and disclosure privileges, on the other. On each of these points, however, the witness’ prior inconsistent statements directly and unmistakably undermine IBM’s argument.
First, IBM relies on the recent declarations from Mr. Wilson and Mr. Frasure for its principal, and audacious, argument that the AT&T’s license agreements protected against only the literal copying of UNIX System V source code. See Wilson Decl. (April 2004) ¶ 12 (Exh. 16) (opining that UNIX license agreements “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”); Frasure Decl. (March 2004) (Exh. 18) ¶ 28 (claiming that licensees were free to use and disclose their products “so long as they did not disclose any UNIX System V code”).
But the witness’ prior statements, including their sworn testimony, are expressly and unequivocally to the contrary. These contradictions alone create issues of fact precluding summary judgment in IBM’s favor. In 1992, Mr. Wilson and Mr. Frasure testified in a lawsuit between UNIX Systems Laboratories (“USL”), assignee of AT&T’s rights to UNIX, and one of its licensees, Berkeley Software Design (“BSD”). In that case, Mr. Wilson testified that AT&T’s license agreements protected far more than just “UNIX System V code” – and that he had expressly communicated that view to AT&T’s licensee. Wilson Dep. (12/10/92) at 128 (Exh. 15). Explaining the contractual significance of the licensee’s “exposure” to AT&T’s original UNIX product, Mr. Wilson further testified:
“Q:If the University of California were to develop a software product that they contended was their property to do with whatever they felt was appropriate, regardless of AT&T, and the university, as you know, has had exposure to the UNIX operating system, has worked with it for many years, is it your understanding that that software that the University of California would develop would be subject to any of the restrictions in the AT&T licensing agreement?
A: Absolutely. It would be.
Q: And why is that?
A: As you characterized it, you said if they had developed some software with exposure to the licensed software would it be subject to the AT&T agreement. I’m saying absolutely that would be the case. That’s why they always used both. They were careful if they had any exposure to the licensed software, it was a given that you had to have a license from AT&T.” Id. at 127-128 (emphasis added).
Furthermore, in a letter that Mr. Wilson wrote to IBM on July 27, 1987, he clearly stated that AT&T intended for its license agreements to protect from disclosure even products that were based on derivatives of the licensed software products. At the beginning of the letter, Mr. Wilson explained: “We have sent a letter to all our licensees who have also acquired sublicensing rights to address an issue that is of great concern to AT&T.” Exh. 19 at 1. He then continued:
“We want to stress at the outset that licensees may not distribute, or permit access to, any licensed source code software product (either in modified or unmodified form) to any third party at any time, or distribute sublicensed products based on such software products or its derivatives, unless specifically permitted by the license agreement.” Id. (emphasis added).
Mr. Frasure’s prior testimony similarly contradicts his recent declaration for IBM. In the BSD case, Mr. Frasure testified specifically about AT&T’s protection of the entire universe of products that could be created as a result of UNIX exposure, regardless of whether the resulting product actually contained UNIX code:
“if you’ve got a group of people over here developing a set of source code and they have never seen the Unix source code, they’ve never been too exposed to it and the [sic] develop a product completely on their own, then that’s one thing. But if they’re developing a product with the benefit of Unix or perhaps they have used it for -- for a number of years, ten years, and then they think they’re going to go off and develop something on their own that’s an operating system that may look like Unix, we had – we expressed our concern that -- that we had an interest in that product.” Id. at 20 (emphasis added).
Mr. Frasure later reiterated that same point:
“Q: But assuming – when you said it’s possible, you were saying it’s possible that this agreement was saying they should distribute it to anyone so long as it did not include the licensed software of [sic] disclose methods or concepts used in the licensed software?
A: My interpretation of this is if – if you were developing software modifications without the benefit of the licensed software, then you could distribute those, but if you developed them with the benefit of the licensed software that you could not distribute them to anyone in this case if you mean anyone who is a non-licensee.” Frasure Dep. (12/8/92) at 42 (Exh. 9) (emphasis added and objection omitted).
When pressed about the scope of AT&T’s contractual protections, Mr. Frasure could cite only one “clear cut hypothetical case” that would leave a licensee free to develop a product that would not be subject to the use and disclosure restrictions of AT&T’s license agreement:
“I say if it was developed independent of the licensed software and without benefit of it, then to me that’s a -- that is the licensee’s product. In other words, if they were completely separate and divorced from this and the people who developed that product didn’t have the benefit of the knowledge of this product. I mean there are -- there are companies that developed software products all the time without the benefit of this, and introduce them.
* * *
That if a -- a company -- a licensee had a facility and people in it that never had -- had used the UNIX source code and they set off and they developed a product on their own, then I -- in my opinion, then AT and T has no interest in that product.” Id. at 166-167.
And, when asked about the implications of this view, Mr. Frasure reiterated that the AT&T agreements were designed to protect various possible products that could be developed from mere exposure to the licensed UNIX product (and precisely the types of developments and circumstances that IBM now attempts to defend in this action):
“Q: Forgive me. I don’t know that I understand your hypothetical completely. This is -- if an AT and T licensee has a licensed UNIX operating system and they’ve looked at it and they’ve worked with it, then they go ahead and develop a product that doesn’t have actual source code, object code, documentation, or methods and concepts of AT and T’s, is it my understanding -- do I understand you to be saying that that licensee’s product falls under the definition of licensed software?
A: I don’t -- I really don’t think that I can answer the -- the question clearly. I guess I have my -- it’s very difficult to try to answer some of these questions, because I think of the methods and the concepts issue. There’s a -- I’ve used the term at least one time, mental contamination if you will, of being -- being exposed to a product. You know, I think a clear cut hypothetical case would be that if you have a licensee and they have two physical locations and you’ve got this group out here that’s never seen the UNIX source code and they go off and they develop a product, then I would think clearly that that is that licensee’s product and AT and T has no interest in it. However, if the other location that had access to that source code and there were people working on that new product that had worked with the UNIX software to some extent -- I don’t know that that extent is because we’re talking in general terms here -- then I would think that there’s a chance that AT and T would have an interest in that product.” Id.
Lastly, and further illuminating the critical contractual concern about licensees’ exposure to the UNIX product, Mr. Frasure recalled a meeting that he had attended with Otis Wilson, Geoff Green (the sole attorney who has signed a declaration submitted with IBM’s motion), and representatives of one of AT&T’s licensees. Testifying about that meeting, which occurred in late 1984, id. at 96-97, Mr. Frasure repeatedly described the concept of “mental contamination” that he and Mr. Wilson had communicated to licensees concerning the expansive scope of the protections provided under the license agreements:
“Q: And if your code was not part of the product?
A: Well, if it was used as part of the development -- I really need to be careful here on words, I guess. If the source code, the Unix source code was -- was required, was used to generate the enhancement, was required to have the -- the rest of the enhancements work, then we had an interest in it. It’s been a long time. I’m not sure of the right -- the right key words to use, but we went through those discussions with them and what we felt the, you know, the agreement said.
We also discussed contractor provisions which allowed a licensee to contract with someone to develop software and then when that development was done everything had to come back to them and we expressed concern I guess with -- Otis [Wilson] and I used the term mental contamination, that if you had been exposed to the source code and its methods and concepts, even though you give something back to the -- the licensee, there was – there was concern there that someone could go off on their own and develop what they thought was their own product but really using the methods and concepts and techniques that were in the product that they had previously used.” Id. at 18-19 (emphasis added).
Second, IBM’s declarants now specifically contend that AT&T never intended to protect any UNIX modifications and derivatives made by licensees, so long as actual source code was not contained in the modification or derivative. Wilson Decl. (April 2004) ¶ 14 (Exh. 16) (“[A]ny 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.”); Frasure Decl. (March 2004) ¶ 28 (Exh. 18) (“In all cases, as I understand the agreements and believe they were intended, modifications and derivative works are not subject to restrictions contained in the license agreements on use, export, disclosure or transfer (except for any licensed UNIX System V source code actually included therein) because they are owned by the licensees.” (emphasis added)).
Again, however, in their prior sworn testimony, Mr. Wilson and Mr. Frasure affirmatively rejected such a restrictive interpretation of the AT&T license agreements and, in fact, ascribed to those agreements the broad protections clearly evinced by the contractual language. In the BSD case, Mr. Wilson testified: “My understanding is that anything created by the university with exposure to the licensed software, based on, contained, a part of, was a derivative work with regard to these documents and had to be treated as licensed software.” Wilson Dep. (12/10/92) at 51 (Exh. 15) (emphasis added). It is hard to imagine a broader statement regarding the protected material.
Mr. Wilson confirmed this view when asked about a letter in which AT&T clarified the meaning of “LICENSED SOFTWARE” in the University’s agreement:
“Q: And under this clarification, as you understand it, not every derivative work prepared by the licensee was required to be treated as licensed software, is that right?
As Mr. Wilson further explained: “That which was created by the university independent of any exposure to the licensed software belonged to the university and we would not even had these types of discussions. These discussions and these agreements only came up when there was exposure to the licensed software.” Id. at 49-50 (emphasis added). Thus, Mr. Wilson’s prior testimony was that, after the licensee had been “exposed” to the UNIX System V source code, “every derivative work prepared by the licensee” had to be treated like the licensed source code.
A: I did not say that.
Q: I’m asking you is that your understanding?
A: Not, it’s not.” Id. at 43 (emphasis added).
Similarly, in Mr. Frasure’s BSD testimony, he actually corrected the very mistake made by IBM in this case, in response to a question from BSD’s counsel:
“Q: Would that – could you tell me the components of what your understanding in May ’85 would have been included in the licensed software? If you recall what your understanding was in 1985.
A: Licensed software would have been . . .
Q: Would it have been the source code?
A: … would have been the source code or some portion of it, with modifications made to it by the licensee.
Q: Where do -- where do you get the understanding that a modification to the source code falls within the definition of licensed software?
A: It says licensee agrees that any modifications or derivative work prepared by it that contains any licensed software . . .
A: . . . shall be treated . . .
A: . . . as licensed software.” Frasure Dep. (12/8/92) at 146 (Exh. 9) (alterations in original) (emphasis added).
In fact, Mr. Frasure recounted how he and Mr. Wilson had been informed of a licensee’s plans to distribute a derivative product, and how they informed the licensee that “we’ll have to bring a suit against you to stop that.” Id. at 22. Asked to characterize the software, Mr. Frasure stated that it was an “add-on type product” to the original UNIX product. When confronted with the basis for such a lawsuit, Mr. Frasure denied the need for actual UNIX source code to have been present in the new product:
“Q: And what was it about what they were planning to distribute that caused you to consider that possibly a violation of the agreement if they pursued it?
A: Because Unix source code had been instrumental in its development.
Q: Was the Unix source code -- was Unix source code included in the product they were planning to distribute?
A: I don’t know, ‘cause we were never told.
Q: Was that something that you asked about at the meeting?
A: I’m sure we did. I don’t recall a specific instance of -- of that. I wish I could recall the specific software that we were talking about, but the -- we considered at the time that the ... the product was a --it was based on Unix, it was -- could be an enhancement or a modification to the operating system.” Id. at 23.
In further defining the concept of derivative and modification work, Mr. Frasure supported the very claim that IBM now belittles -- that even significant additions and enhancements to UNIX would be considered derivates and modifications subject to the licensing agreement protections. Mr. Frasure testified that he always “conveyed to the licensees or whoever we were talking to that a modification was an addition or change to or the deletion from.” Id. at 118. Mr. Frasure stressed that the modification label , once attached, remained on newly developed code, regardless of how much was added to the initial product: “a modification is a modification and that involves three things, and you can add one line of code or a thousand lines of code or ten thousand.” Id. at 121.
Mr. Frasure asserted that the licensing agreements’ protection of derivative work extended the above protections to modified products that could operate independently:
“Q: Okay. And then I would also like to ask you about derivative work, and I’m not asking you for a legal conclusion, I’m just asking for your common understanding of that in your dealings with the University and as an employee of AT and T. Did the definition of a derivative product or work vary in any way from what you’ve earlier testified is a modification, or is it somehow different?
A: Well, derivative product is a result of the modifications, and we -- I -- I guess I consider the derivative product to be something that you would -- it would be executable or usable in its entire form. In other words, it’s a -- it’s a product that will stand on its own, but it’s strictly as a result of the modifications that have been made to the -- to the software.” Id. at 119.
Third, IBM’s declarants now deny that AT&T sought to protect the methods and concepts embodied in UNIX System V, again in the face of plain contractual language to the contrary. Mr. Wilson’s declaration now states that, at least as of February 1985, AT&T did not seek to protect the methods and concepts embodied in UNIX System V:
“In fact, we were not aware of any particular ‘methods or concepts’ that needed to be protected. We made up this phrase as sort of a general catch-all. We were willing to delete the reference to methods and concepts in the IBM agreement because we were not aware of any UNIX System V methods or concepts that required protection. The fact that we had distributed the UNIX System V source code so broadly that the internal structure of the UNIX system V operating system was well known in the academic community and by computer programmers generally.” Wilson Decl. (Dec. 2003) ¶ 14 (Exh. 21).
See also Frasure Decl. ¶ 12 (Exh. 18) (claiming that the licensing agreement provisions were not intended “to restrict our licensees’ use export, disclosure or transfer of anything besides the licensed UNIX System V source code and related materials”).
But just prior to and well after the IBM Side Letter was signed in February 1985, Mr. Wilson (on behalf of AT&T) repeatedly and expressly sought to protect UNIX “methods and concepts” pursuant to AT&T’s license agreements. For example:
In his letter to Digital Equipment Corporation (“DEC”) dated Sept. 5, 1984, Mr. Wilson wrote:
“Regarding Section 2.01 of the referenced Software Agreement, you do not have all right, title and interest in derivative works based on a SOFTWARE PRODUCT if such derivative works include any of our code or embody any of the methods and concepts used in a SOFTWARE PRODUCT.” Exh. 22 at 2 (emphasis added).
In proposed (and later executed) letter agreements dated June 18, 1990, June 29, 1990, and August 3, 1990, Mr. Wilson stated to other licensees: “You will not provide access to any copy of the source code of the SOFTWARE PRODUCT (including methods and concepts contained therein), in whole or in part, to anyone other than your organization’s employees who have a need to know.” Exhs. 23, 24 & 25 (emphasis added).
Such statements plainly belie Mr. Wilson’s current claims that by the time of the IBM Side Letter in 1985, “we were not aware of any UNIX System V methods or concepts that required protection” and that AT&T “had distributed the UNIX System V source code so broadly” that no protection for methods and concepts could obtain.
Mr. Frasure, for his part, testified in the BSD case that AT&T sought to protect far more than merely UNIX source code through its “methods and concepts” provision. Mr. Frasure clearly stated, “the licensed software to me encompasses methods and concepts techniques,” so that “to the extent at some point [a licensee] gets involved with the source code, than I – and its methods and concepts, then I think there -- there’s a restriction on its -- on its use.” Frasure Dep. (12/8/92) at 45-46 (Exh. 9) (emphasis added). Acknowledging that AT&T’s software license agreement undoubtedly protected actual UNIX source code, Mr. Frasure stressed that there is “more to the agreement” and proceeded to point to “other portions of paragraphs that cover methods and concepts and – and stuff.” Id. at 37 (emphasis added). In fact, Mr. Frasure characterized these protections as a “key part of the agreement.” Id. at 147.
When asked point blank about the types of products covered by AT&T’s licensing agreement, Mr. Frasure responded as follows:
“Q: So licensed software – your understanding of licensed software in the 1985 time period in the System Five license agreement could include source code of AT and T’s; is that correct?
Q: Okay. And it could include object code of AT and T’s as well?
Q: Okay. And it could also include documentation of AT and T’s; is that correct?
Q: And it could also include methods and concepts of AT and T’s included in their System Five product; is that correct?
A: Yes.” Id. at 148.
Mr. Frasure also explicitly acknowledged the application of this methods and concepts protection to work created by a licensee’s programmers:
“if someone has been exposed to the source code and has used the source code and then goes off on their own and wanted to develop a product that subconsciously or through rote or whatever the words are developed a product that contained the methods and concepts, techniques that we’re using, then it’s -- I don’t know what the right word to use -- understanding, interpretation or what –-
that –- that AT and T would have a -- an interest in that product.” Id. at 147-148.
And Mr. Frasure expressed his doubts about a licensee’s ability to develop genuinely independent, but related products after such exposure to UNIX:
“Q: Under what circumstances would they have an interest -- would AT and T have an interest in the licensee’s product because they had had mental access to the AT and T Unix operating system?
A: Well, I think the fact that they -- they had access to it -- I’m really not trying to be -- to be stubborn with you but techniques, methods and concepts are things that -- that we use in every day life that we have been trained -- we have learned by experience to use. I think that for someone to say, well, we’re not using AT and T’s methods and concepts is ... is a -- it’s a statement that -- someone may think that
because they subconsciously aren’t aware that they were, but they could be, that’s why I have to answer the question by saying there’s a chance that they -- they do. I mean you do things for so long and develop styles and techniques, then for someone to say, well, it doesn’t contain AT and T’s methods and concepts, I -- I think may be hard to say if someone has worked with a product for -- for some number of -- period of time.” Id. at 169-70 (emphasis added).
Lastly, in describing the extent of AT&T’s protections of the “methods and concepts” derived from UNIX, Mr. Frasure delineated the scope of activities that could violate the provision prohibiting disclosure of such material:
“Q: Okay. But with respect to the methods and concepts, what is restricted in that paragraph is the disclosure of those methods and concepts.
Q: Is that right? Is that your understanding, in this paragraph?
A: Disclosure has many forms, so it -- if I look at the -- the word disclosure as having many forms, then -- then I would agree with that.
A: Not just -- not just one form, but many. You can reveal how things are done by the way you do other things, and that’s a disclosure.” Id. at 51 (emphasis added) (objection omitted).
In response to questioning about what products could be disclosed under the agreement, Mr. Frasure was hesitant to concede that any product remained unprotected, because even “if you develop your own system modifications and you use -- and it works with the software that was licensed, there is a -- there is a very good chance you are revealing methods and concepts just strictly through the interface technique that -- that’s used.” Id. at 41-42 (emphasis added). Mr. Frasure maintained this position even when asked about products that did not overtly disclose methods or concepts:
“Q: But assuming -- when you said it’s possible, you were saying it’s possible that this agreement was saying they should distribute it to anyone so long as it did not include the licensed software of [sic] disclose methods or concepts used in the licensed software?
A: My interpretation of this is if -- if you were developing software modifications without the benefit of the licensed software, then you could distribute those, but if you developed them with the benefit of the licensed software that you could not distribute them to anyone in this case if you mean anyone who is a non-licensee.” Id. (emphasis added) (objection omitted).
Fourth, IBM’s declarants now ignore any distinction between the ownership rights that AT&T expressly recognized through the IBM Side Letter and its subsequent license agreements, on the one hand, and the contractual restrictions that AT&T’s license agreements imposed, without exception, on the use, copying, distribution, and disclosure of products covered by those agreements. In so doing, they mistakenly claim that despite its plain language (which was expressly limited to ownership), the IBM Side Letter eviscerated all of the significant protections that AT&T’s license agreements afforded to the UNIX intellectual property. For example, Mr. Wilson now declares that by recognizing in the IBM Side Letter that AT&T would not claim any ownership interest in work developed by IBM, AT&T intended to give IBM
“the right to control modifications and derivative works prepared by or for IBM” such that “IBM (like all licensees under the agreements) fully owns any modifications of and derivative works based on UNIX System V prepared by or for IBM, and can freely use, copy, distribute or disclose such modifications and derivative works, provided that IBM does not copy, distribute or disclose any material portions of the original UNIX System V source code provided by AT&T or USL (except as otherwise permitted by the IBM Agreements).” Wilson Decl.
¶ 19-20 (Exh. 16) (emphasis added).
See also Frasure Decl. ¶ 18 (Exh. 18) (asserting that IBM Side Letter evinced AT&T’s intent not to “assert any right to control the use and disclosure of modifications and derivative works prepared by its licensees, except to the extent of the licensed UNIX System V source code included in such modifications and derivative works”).
Contrary to Mr. Wilson’s current declaration, however, at the time AT&T entered into its software license and side letter agreements with IBM, he repeatedly and expressly acknowledged the distinction he now ignores between licensees’ “ownership” of modifications and derivative works, on the one hand, and their “use, copy, distribute, or disclosure” of such works, on the other ? while confirming (again) that AT&T’s license agreements protected UNIX “methods and concepts.” In Mr. Wilson’s letter to DEC dated Sept. 5, 1984, for example, he explained:
“If such derivative work does include any of our code or embody any of our methods and concepts you may have a property right in such derivative works to the extent of any modifications that you have added, but the exercise of that property right is subject to the terms of the Software and Sublicensing Agreements, including to the restrictions on the use of the SOFTWARE PRODUCT (for example, it must be kept in confidence).” Exh. 22 at 2 (emphasis added).
Mr. Wilson reiterated this statement in a later letter to DEC on February 21, 1985 – twenty days after AT&T entered into the license agreement and side letter with IBM. See Exh. 26 at 3. These statements plainly belie Mr. Wilson’s contention that neither AT&T nor its license agreements drew any distinction between the licensee’s ownership of certain material and the restrictions on its use of even that material.
Likewise, when Mr. Frasure was asked in the BSD case about contractual language that was substantively similar to the “Regarding 2.01” provision in the IBM Side Letter, he made clear that the contractual statement of ownership did not in any way alter the contractual restrictions on use and disclosure:
“Q: So that if an AT and T licensee developed a software that was a modification or derivative work which did not contain source code, object code, documentation or methods and concepts of AT and T’s, AT and T did not assert an ownership interest in that product. Is that your understanding – was that your understanding in May of 1985?
A: It has no ownership in that portion of the product.
A: But it doesn’t mean that it’s not part of the licensed software, I don’t think. I think there’s two -- two things here that we’re talking about, licensed software and -- and ownership, which I think are two completely separate subjects.” Frasure Dep. (12/8/92) at 151-152 (Exh. 9).
Mr. Frasure reiterated that distinction numerous times in his BSD testimony:
“I still feel like we’re talking about -- about two separate things here. One is ownership and one is what is the licensed product.” Id. at 166-167.
And, in addressing the implications of that distinction, he further explained:
A: “I feel like there’s been perhaps some confusion or interchange of words incorrectly regarding one completely separate subject to me, which is ownership, and the other is licensed software, and the two are not interchangeable, so that was the reason that I attempted to make a distinction before.
* * *
“ownership, to me is a very clear and distinct subject from the licensed software product. If ... if you generate the code, add the code to it then I think, clearly the software agreement and the side letter said that you owned that. That didn’t say that it was not part of the – the software licensed product . . .
Q: So what are you suggesting? You owned it but because it was part of the agreement, what does that ... can you tell me the significance of that?
A: Well, I ? I’m just saying I want to make a distinction. I’m not sure where we are going with this. It’s just that I’m saying that the ownership belongs clearly to the originator of -- of that code, but it -- it could be construed to be part of the software product.” Id. at 157-59. (ellipses in original) (emphasis added).
Lastly, Mr. Frasure used this fundamental distinction to explain the textual change, later in 1985, to conform AT&T’s software agreement to the substantially similar ownership language in the IBM Side Letter: “all we were trying to do is to clarify that there was – they [the licensees] owned the – the modifications, any – any derivative or enhancements that they made to the product, but it was also going to be treated as part of the product.” Id. at 103; see also id. at 111 (specifying that the change to the language in AT&T’s standard license agreement was designed to “reduce the number of side letters to licensing agreements that would clarify the ownership issue”).
Therefore, at an absolute minimum, the prior testimony of Messrs. Wilson and Frasure not only contradicts IBM’s position that AT&T did not protect, and was not interested in protecting, anything other than literal UNIX source code, but also demonstrates that AT&T actually sought through its license agreements to protect any product that a licensee made after having been given access to UNIX. Based even on that limited reading of those witnesses’ prior testimony – given much closer in time to the execution of the agreements at issue – IBM could not be entitled to summary judgment based on the extrinsic evidence it has submitted (which, other than the testimony of Messrs. Wilson and Frasure, has not been tested at all). Furthermore, as explained in greater detail herein, these statements also demonstrate why the discovery that IBM has failed to produce regarding the evolution of AIX and Dynix is essential, to permit SCO to show that source code IBM has contributed to Linux was developed through “exposure” to and “with the benefit of” the licensed UNIX product, represented modifications or derivatives “based on” that product, and/or contained “methods and concepts” from the UNIX product. See Part III, below.
The Sworn Declaration of Martin Pfeffer,
AT&T’s Chief Attorney on UNIX Licensing Protections
In addition to the prior statements of IBM’s own declarants, testimonial evidence that SCO has thus far secured – in the face of the recurring discovery obstacles IBM has imposed – unmistakably supports SCO’s reading of the clear and unambiguous contractual language.
At the very time that AT&T entered into its licensing agreements with IBM and Sequent, AT&T General Attorney Martin Pfeffer, an intellectual property specialist, was primarily responsible for participating in, supervising, and approving AT&T legal department’s drafting of, and enforcing, the terms of AT&T’s UNIX license agreements. Pfeffer Decl. ¶¶ 3-4, 11 (Exh. 27). Mr. Pfeffer’s sworn declaration is consistent with SCO’s contract interpretation and the prior testimony of IBM’s declarants, and directly undermines IBM’s summary judgment motion. Mr. Pfeffer’s declaration confirms the following:
Section 2.01 of AT&T’s license agreements was intended to provide expansive coverage to any modifications or derivatives of the original licensed UNIX product. Mr. Pfeffer’s sworn declaration explains that Section 2.01 “set forth the parties’ intent and agreement that the ‘SOFTWARE PRODUCT’ licensed and protected under the terms of the license agreements included the full content of all of the ‘resulting materials’ created over time from the licensees’ exercise of their contractual ‘right to modify’ and ‘to prepare derivative works’ based on the original licensed material.” Id. ¶ 6. Thus, according to Mr. Pfeffer, “under Section 2.01, if a licensee created a modification or derivative work based on the original licensed product, then the agreement treated the ‘resulting’ work as if it had been part of the original SOFTWARE PRODUCT, and any further modifications or derivatives of that ‘resulting’ work would be treated in the same manner.” Id.
The ownership language contained in the IBM Side Letter was not intended to release any licensee from the contractual restrictions on its use, copying, distribution, or disclosure of materials covered by the license agreements. Mr. Pfeffer recalls that in 1985, AT&T added language to its license agreements “to address inquiries from certain of its licensee concerning the ownership of modifications and derivative works.” Id. ¶ 7. According to Mr. Pfeffer, the language that AT&T added to its agreements (which is substantially similar to the “Regarding 2.01” clause in the IBM Side Letter) “set forth the parties’ intent and agreement concerning only the ownership of modifications or derivative works, and that this language was not intended to change, and did not in fact change, AT&T’s right to control (for example, the use and disclosure) of such modifications and/or derivative works, as set forth in Section 2.01 and other provisions of the software license agreements.” Id. Moreover, Mr. Pfeffer does not recall during his tenure with AT&T and USL (which lasted from 1962 through 1993) “any such instances in which AT&T or USL changed or agreed to change the contractual limitations on use and disclosure that were set forth in the UNIX license agreement.” Id.
Section 7.06 of the AT&T license agreements was intended to provide broad confidentiality protections. According to Mr. Pfeffer’s sworn declaration, the general requirement of confidentiality (usually contained in the first sentence of Section 7.06(a)) of the license agreements, “set forth the parties’ intent and agreement that the ‘SOFTWARE PRODUCTS’ covered by the license agreements (and, pursuant to Section 2.01, any modifications or derivative works ‘based on such SOFTWARE PRODUCT’) would be subject to the restrictions set forth in AT&T’s confidentiality provision, regardless of the licensee’s ownership interest in any part of such SOFTWARE PRODUCTS.’” Id. ¶ 8. Moreover, Mr. Pfeffer explained, the further specific protection for “methods or concepts” found in the Sequent license agreement was “specifically designed to cast the widest possible net in order to ensure protection for AT&T’s proprietary information, beyond the common law protections that would otherwise have been available (absent the agreement) for AT&T’s trade secrets and ‘know-how.’” Id. ¶ 9.
AT&T’s UNIX license agreements were specifically designed to protect far more than just literal UNIX source code. According to Mr. Pfeffer, “it was AT&T’s intent to prevent through its UNIX license agreements the unauthorized use and disclosure of more than just literally copied UNIX source code.” Id. ¶ 10. Thus, “AT&T intended to protect, and through its standard license agreements expressly protected, its UNIX business by preventing anyone from using AT&T’s proprietary material in UNIX (including by literally or non-literally copying UNIX source code, disclosing methods or concepts from UNIX, and/or exploiting licensees’ access to the technology in UNIX) without paying UNIX license fees to AT&T.” Id. Mr. Pfeffer further explained: “AT&T effected this protection through its license agreements by, among other things, requiring licensees to treat all materials resulting from any exercise of their ‘right to modify’ and ‘to prepare derivative works’ as if such materials had been ‘part of the original SOFTWARE PRODUCT’; by specifically requiring licensees to hold AT&T’s ‘SOFTWARE PRODUCTS subject to this Agreement in confidence for AT&T’; and by strictly controlling the disclosure of ‘any or all of such SOFTWARE PRODUCTS (including methods or concepts utilized therein).’” Id.
The License agreements’ protection of more than literal UNIX source code was never subject to negotiation. Mr. Pfeffer does “not recall any instance during my tenure in which either AT&T or USL agreed (in any license agreement or supplement or modification thereof) to reduce its protection under a UNIX license to prevent the unauthorized use or disclosure of only source code.” Id. ¶ 11. According to Mr. Pfeffer, any such change would have represented “a significant and material change to the standard terms of AT&T’s license agreement” that would have required his, or to a lesser extent Burt Levine’s, approval. Id. Not only does Mr. Pfeffer confirm that no such change was ever made, but he does not even recall any such “proposed” modification to AT&T’s standard license agreement. Id.
The integration clause in the licensing agreements was intended to prevent the terms of any agreement between AT&T and one licensee from affecting the rights and obligations of any other licensee. Mr. Pfeffer’s sworn declaration reiterates that the standard integration language in AT&T’s licensing agreements “set forth the parties’ intent and agreement that only the terms of each licensee’s written agreements with AT&T would govern those parties’ contractual obligations. Id. ¶ 12. In other words, any agreements (including any side letter agreements) that AT&T entered into with one of its UNIX licensees would have no legal effect on the contractual obligations of AT&T vis-à-vis any of its other UNIX licensees.” Id.
Accordingly, the limited extrinsic evidence available to date defeats IBM’s motion and demonstrates why further extrinsic evidence discovery – including testing the other declarations on which IBM relies – is warranted (if extrinsic evidence is to be considered at all on the contract interpretation issues). The foregoing evidence also illustrates precisely why the license agreements clearly require that any modifications or derivatives of the licensed Software Product must be treated as part of that Software Product: AT&T was greatly concerned that as a result of the “contamination” that would follow from licensees’ access to AT&T’s intellectual property, those licensees would create modifications and derivative works that copied AT&T’s intellectual property (even if they did not include literal copying of UNIX source code). The foregoing deposition of Messrs. Wilson and Frasure also demonstrates why, given the evidence of AT&T’s intent in and own interpretation of its license agreements, SCO is entitled to discovery in order to demonstrate that the AIX and Dynix source code – including source code contributed to Linux – was written “with the benefit of” the UNIX source code.
B. IBM’s “Waiver” Arguments Are Patently Meritless
IBM also attempts to avoid any further scrutiny of its conduct by claiming entitlement to summary judgment on the ground that its breaches of the license agreements were purportedly “waived.” IBM Contract Mem. at 71-74. The question of waiver, however, is “generally ill-suited for summary adjudication.” In re Caldor, Inc., 217 B.R. 121, 133 (Bankr. S.D.N.Y. 1998); see also Boston Concessions Group v. Criterion Ctr. Corp., 606 N.Y.S.2d 696, 697 (App. Div. 1994) (“[U]nder New York law, the establishment of a waiver, requiring the relinquishment of a known right, is ordinarily a question of fact which precludes summary judgment.”). And, as explained below, IBM’s two “waiver” arguments provide no basis for departing from that rule.
1. IBM’s Claim that Novell Has Waived SCO’s Rights Relies on an
Indefensible Misreading of the UNIX Asset Purchase Agreement.
IBM first contends that Novell, on SCO’s behalf, has waived IBM’s breaches of the license agreements by the supposed authority of the Asset Purchase Agreement between Novell and SCO’s predecessor-in-interest The Santa Cruz Organization, Inc. (hereinafter “SCO”), which was dated as of September 19, 1995 (the “APA”) (Exh. 28). IBM Contract Mem. at 72. For this argument, IBM relies on an indefensible reading of the APA, which removes snippets of text from their context while ignoring the key contractual provisions as well as the unmistakable deal structure and parties’ intent.
In 1995, Novell sold to SCO its entire UNIX-related business. The APA expressly transferred to SCO “all of Seller’s right, title and interest in and to” the specified assets, which included, among other things, the following: (1) “All rights and ownership of UNIX and UnixWare,” including, among other things, the UNIX source code; (2) “All of Seller’s rights pertaining to UNIX and UnixWare under any software development contracts, licenses and any other contracts,” including, among others, the UNIX “Software and Sublicensing Agreements – This includes the source code and sublicensing agreements that Seller has with its OEM, End User and Educational customers”; and (3) “All contracts relating to the SVRX Licenses” for, among many others, all versions of UNIX System V that had been licensed to IBM and Sequent. See APA at Schedule 1.1(a), I, III, VI (Exh. 28).
The sole exception to the wholesale transfer of UNIX rights under the APA is Novell’s retention of the limited right to receive future royalties paid by UNIX binary code licensees under UNIX System V licenses. See id. at Schedule 1.1(b), VIII (reserving for Novell “All right, title and interest to the SVRX Royalties, less the 5% fee for administering the collection thereof pursuant to Section 4.16 hereof.”). This arrangement was designed as a pricing mechanism; because SCO could not pay Novell the full purchase price up front, SCO agreed to permit Novell to retain these future royalty streams to bridge the price gap. Id. § 1.2(a)-(b) (specifying as “Payments” for “the transfer of the Assets by Seller to Buyer” both SCO common stock and “the SVRX Royalties as defined and described in Section 4.16 hereof”).
Despite the clearly expressed purpose of the APA to transfer all rights relating to UNIX source code licenses to SCO, IBM quotes a portion of section 4.16(b), out of context, and contends that it gives Novell the power to “waive” all of SCO’s rights (including the intellectual property protections) under the UNIX System V licenses. The text and context of section 4.16 make clear, however, that the sole purpose of that section is to set out the rights and obligations of the parties with respect to the binary royalty income stream that Novell retained through the APA. Thus, Section 4.16(a) sets forth the procedures to be followed in the collection, administration, and payment of the SVRX binary royalties, granting Novell the right to audit SCO’s collection of those revenues in order to protect its going-forward interest in the SVRX Royalties. See id. § 4.16(a). And Section 4.16(b) further protects Novell’s binary royalty rights by prohibiting SCO from compromising those rights under the SVRX agreements without Novell’s consent and by requiring SCO to take certain actions under the SVRX agreements at Novell’s direction. IBM’s argument based on its selective quotation from Section 4.16(b) thus ignores the critical context of that provision which, together with the key terms of the APA, clearly demonstrate that 4.16(b) did not, and was not intended to, give Novell the unfettered right to destroy the value of the assets that SCO purchased by giving away SCO’s intellectual property protections under the SVRX source code software licenses.
SCO and Novell subsequently affirmed this understanding – in language that IBM’s motion overlooks entirely – through the October 1996 Amendment 2 to the Asset Purchase Agreement. Exh. 29. In addition to acknowledging that the APA had transferred to SCO “the copyrights and trademarks owned by Novell as of the date of the Agreement required for SCO to exercise its rights with respect to the acquisition of UNIX and UnixWare technologies,” id. ¶ A, Amendment 2 reiterated that Novell had no “right to increase any SVRX licensee’s rights to SVRX source code,” no “right to grant new SVRX source code licenses,” and no right to “prevent SCO from exercising its rights with respect to SVRX source code in accordance with the Agreement,” id. ¶ B.5.
IBM thus interprets the APA to create an unreasonable result that the parties never intended. See, e.g., In re Lipper Holdings, LLC, 766 N.Y.S.2d 561, 562 (N.Y. App. Div. 2003) (citations omitted) (a “contract should not be interpreted to produce a result that is absurd, commercially unreasonable or contrary to the reasonable expectations of the parties”). As noted above, under IBM’s interpretation, Novell could unilaterally permit any licensee to disclose material protected by the confidentiality provisions – thereby eviscerating the value of the assets SCO purchased. Moreover, if Novell could waive SCO’s intellectual property protections under the SVRX agreement, there would have been no reason for Novell to have taken out an SVRX source code license from SCO concurrent with the consummation of the APA, see APA § 1.6 (Exh. 28), and the express, stringent restrictions on Novell’s use of that code – e.g., “for internal purposes” only, id. – would have been worthless to SCO, the licensor. Indeed, if Novell could waive any of the restrictions of its own SVRX source code license, as IBM’s extreme contract reading suggests, then Novell could even continue its UNIX business in competition with SCO, rendering the APA itself a nullity. The Court could not reasonably, and certainly could not as a matter of law, ascribe any such reading to the APA or any such intent to the contracting parties.
2. IBM’s Argument for Summary Judgment on the Ground That
SCO Waived Its Own Rights Under the Contracts Is Meritless.
IBM also claims that SCO itself has “waived” IBM’s breaches relating to its improper contributions to Linux on the ground that SCO sold to customers, and allegedly made available on the Internet, code that IBM had improperly contributed to Linux. IBM Contract Mem. at 74. IBM’s argument is wrong, and it does not remotely raise an issue that the Court could resolve in IBM’s favor on summary judgment.
“A waiver is the intentional relinquishment of a known right with both knowledge of its existence and an intention to relinquish it.” City of New York v. State, 40 N.Y.2d 659, 669 (1976); see also Davison v. Kless, 280 N.Y. 252, 261 (1939) (“Whether an alleged waiver is express or implied, it must be intentional.”). “Waiver requires the voluntary and intentional abandonment of a right which, but for the waiver, would have been enforceable.” Gen. Motors Acceptance Corp. v. Clifton-Fine Sch. Dist., 85 N.Y.2d 232, 236 (1995) (internal citations omitted). To relieve itself of its breaches of contract, IBM has the burden of proving that SCO intended to waive IBM’s breaches, even while drawing all factual inferences in SCO’s favor. IBM’s motion does not remotely carry this burden.
Far from evincing a knowing intent to relinquish its rights under the UNIX license agreements, SCO has diligently sought to protect and enforce those rights since almost immediately after it learned of IBM’s intention to breach its obligations. Within weeks of IBM’s announcement that it intended to “obliterate UNIX” by exploiting its AIX expertise in Linux development, SCO notified IBM of SCO’s intention to terminate IBM’s license agreement due to IBM’s breach. After delivery of the termination notice in accordance with Section 6.03 of the license agreement, SCO attempted to meet and confer with IBM, but IBM failed to make any effort to cure its breaches during the two-month grace period provided in SCO’s termination letter. Accordingly, effective June 13, 2003, SCO terminated IBM’s System V licenses. This same process played out with respect to the Sequent license agreement, which was terminated effective July 30, 2003. SCO then promptly filed the present suit alleging, among other things, IBM’s breach of both the IBM and Sequent agreements.
IBM’s argument thus necessarily reduces to the claim that SCO impliedly waived the contractual rights it has diligently sought to protect and enforce in this lawsuit. That argument fails on the law and the facts. It has long been the law in New York that questions of implied waiver are never appropriate for summary judgment: “where a waiver is not express, but found in the acts of a party, summary judgment is not appropriate. An implied waiver is invariably a matter of intention and, therefore, an issue of fact.” McDarren v. Marvel Entertainment Group, Inc., No: 94 CIV. 0910 (LMM), 1995 WL 214482, at *5 (S.D.N.Y. Apr. 11, 1995) (internal citations omitted) (Exh. B); accord Gary v. Peckham, 468 F.2d 1241 (10th Cir. 1972) (“The question of implied waiver (and there is no contention that there was an express waiver) is an issue of fact which is to be decided only after an evidentiary or factual hearing.”).
In addition, even the facts that are available now, before discovery has closed, plainly belie IBM’s suggestion that SCO impliedly waived IBM’s breaches by distributing and making available Linux after SCO filed its suit. See IBM Contract Mem. at 76. It is illogical to suggest, as IBM does here, that a clear and unequivocal intent to waive certain rights can be inferred from acts taken after the filing of suit to protect those rights; and IBM has not located a single case to the contrary.
Moreover, IBM relies for its motion on an inaccurate presentation of the facts; for example, contrary to IBM’s claims, SCO did not make Linux available to the general public on its website after it filed this suit. In accordance with its preexisting contractual obligations to its customers, to whom SCO owed continued upgrades, maintenance, and service, SCO made certain Linux products available on its website to its customers only and on a non-public (and later password-protected) basis. Once SCO learned that non-customers had been able to access the products despite the password protection, it promptly removed all Linux products from its website.
Further factual development will reveal that IBM’s “waiver” argument is simply another attempt by IBM to exploit its own improper conduct. IBM cannot plausibly claim that SCO’s intentions with respect to its rights were unclear at any point IBM identifies. Instead, IBM asserts that its own impermissible actions in breach of the license agreement created a situation whereby SCO had only two choices when it learned of IBM’s breach: abandon its customers and its Linux business overnight – and inflict the maximum amount of damage on those customers and itself because of IBM’s violation or immunize IBM for its long, undisclosed course of improper conduct. IBM demands the right to put SCO to this impossible choice even though there is nothing SCO could have done to undo IBM’s basic violation at the time in question, and IBM even asks the Court to grant it this immunity as a matter of law. No precedent remotely supports such a perverse result, which would effectively permit the fact and extent of IBM’s wrongful conduct to absolve IBM from responsibility for that very conduct.
C. Even on Summary Review, the Facts That IBM Omits Preclude
Summary Judgment on IBM’s Eighth Counterclaim
IBM also now moves for summary judgment on its Eighth Counterclaim, which alleges that SCO has violated IBM’s copyrights in code that IBM contributed to Linux. In support of that motion, IBM contends that if SCO provides a release of potential claims based on its own software to a client who also uses Linux, SCO has somehow sold Linux in violation of the terms of the General Public License (“GPL”) or the Lesser General Public License (“LGPL”) (together, “the Licenses”) that pertain to Linux. This contention fails upon even brief review.
The GPL and LGPL both state, in their preambles (which IBM does not quote):
“When we speak of free software, we are referring to freedom of use, not price. Our General Public Licenses are designed to make sure that you have the freedom to distribute copies of free software (and charge for this service if you wish).” IBM’s Eighth Countercl. SJ Mem. Exhs. 26 & 27 at 1 (emphasis added).
Under the Licenses, derivative works made from works covered by the Licenses must “be licensed as a whole at no charge to all third parties under the terms of this License.” Id. ¶ 2. IBM quotes only the second half of this phrase, removing the “as a whole language,” and thus attempting to ignore the distinction between charging for a license and charging for other related products or services. In additional language that IBM fails to acknowledge, the Licenses both go on (in the same paragraph) to draw that distinction:
“it is not the intent of this section to claim rights or contest your rights to work written entirely by you; rather, the intent is to exercise the right to control the distribution of derivative or collective works based on the Program [in LGPL, ‘Library’]” Id. (emphasis added).
The GPL and LGPL thus clearly provide that while collecting “license fees” for Linux as a whole is not permitted, collecting other fees and specifically protecting one’s rights in one’s own intellectual property is.
IBM ignores this distinction and alleges that SCO has “collected, and attempted to collect, royalties and licensing fees from Linux users in excess of the fees permitted by the GPL and LGPL.” IBM’s Eighth Countercl. SJ Mem. ¶ 66. IBM makes that statement without any factual support. The Licenses contain no limitation on what fees can be charged for related products and services, and IBM does not offer any testimony or other evidence concerning any such limitations.
IBM instead bases its claims on two kinds of conduct. First, IBM relies on SCO’s efforts to enforce or resolve potential claims arising out of its intellectual property claims, including legal actions, press releases, and letters to users of Linux. See IBM’s Eighth Countercl. SJ Mem. ¶¶ 66(a)-(d), (h)-(j), (l)-(n). None of those activities could be considered “collecting a license fee or royalty” for Linux, and IBM points to nothing in the Licenses (nor is there anything) that either expressly or even impliedly forbids press releases, letters, public statements, or even legal actions aimed at enforcing a party’s intellectual property rights.
Second, IBM disingenuously claims that SCO has sold licenses for Linux to several companies. See IBM’s Eighth Countercl. SJ Mem. ¶¶ 66(e)-(g), (k). In addition to torturing the language and meaning of those licenses, IBM omits to mention that the agreements IBM cites contain releases of claims because they settle SCO’s potential legal claims against the licensee. Id. ¶ 3. Discovery (and a review of the “licenses” complained of by IBM) will show that all of them follow this same general form and none even purports to be a license for Linux.
In paragraph 66(f) of its brief, for example, IBM states: “On October 14, 2003, SCO sold Leggett & Platt, Inc. an ‘IP Compliance License’ for Linux. (Ex. 33.).” IBM thus quotes the term “IP Compliance License,” but not the word “Linux.” That is because Exhibit 33, an invoice for the license in question, neither covers Linux nor contains the word Linux anywhere in the document. IBM simply grafts that additional language onto the document to try to secure summary judgment on its own counterclaim before any discovery can be taken.
IBM’s reliance on a license that SCO sold to Questar is similarly inappropriate. That license, even as quoted by IBM, is one explicitly “to use SCO IP Rights . . . concurrent with run-time use of the Linux Operating System.” IBM’s Eighth Countercl. SJ Mem. ¶ 66(g) (ellipses in IBM original). IBM thus admits that this license does not even purport to be a license for Linux. The agreement defines “SCO IP Rights” (further language IBM declines to acknowledge) as “SCO’s intellectual property rights in any and all past, current or future versions or portions of the SCO’s software products commonly known as UNIX System V and/or UnixWare” concurrent with run-time use of Linux. The definition explicitly exempts Linux from its definition: “SCO IP rights shall not include any right to copy, distribute, modify or alter Linux Software.” IBM’s Eighth Countercl. SJ Mem. Exh. 34 ¶ 1.11 (emphasis supplied). The foregoing points illustrate how there is not merit to IBM’s suggestion that this agreement or any other evidences SCO’s attempt to “collect royalties and licensing fees” for Linux.
III. IBM MAY NOT FORCE ADJUDICATION OF ITS ALREADY PREMATURE
AND FACT-INTENSIVE DISPOSITIVE MOTIONS WHILE ALSO REFUSING
TO PRODUCE PREDICATE DISCOVERY NECESSARY FURTHER TO
DISPROVE THE PRECISE THEORIES OF THOSE MOTIONS
IBM not only is pressing its fact-intensive dispositive motions no less than six before the cut-off for fact discovery in the Court’s Amended Scheduling Order, but also expressly seeks to defend that tactic on the grounds that SCO could by now have taken depositions of material witnesses in this case. See Letter from B. Hatch to T. Shaughnessy dated Aug. 26, 2004 (Exh. 30), and Letter from T. Shaughnessy to B. Hatch dated Aug. 27, 2004 (Exh. 31). Yet IBM has taken every step to preclude SCO from obtaining even rudimentary discovery necessary for SCO to proceed with such depositions in a remotely complete or efficient way, and has done so while simply ignoring core theories of SCO’s claims (Part A, below).
SCO shows below, by way of example, how IBM has precluded – and how it has been in IBM’s interest to preclude – SCO from obtaining the discovery necessary to permit SCO to depose the material witnesses in this case regarding IBM’s interpretation of the license agreements (Part B, below). SCO further shows that the discovery IBM has failed to produce would itself bear directly on the merits of IBM’s contract motion (Part C, below).
A. IBM Brings Its Dispositive Motions in a Discovery Context
That Shows the Motions to Be Plainly Inappropriate
IBM seeks the adjudication of its motions in a context in which the Magistrate Judge granted (on March 3) SCO’s initial motion to compel adequate responses to certain initial discovery requests and required IBM to provide specified discovery that it had refused to provide and to supplement deficient responses, but IBM has yet to comply with that Order. IBM’s failure forced SCO to renew its earlier motion to compel – simply to secure compliance with the Court’s prior Order. See SCO’s Mem. in Support of Its Renewed Motion to Compel (July 6, 2004) (“Compel Mem.”) & Reply Mem. in Support of Renewed Motion to Compel (Aug. 26, 2004) (“Reply Compel Mem.”); see also Disc. Mem. and Reply Mem. re Disc.
The discovery SCO continues to await is necessary in order for SCO to undertake the following crucial steps to uncover more evidence contradicting the bases for IBM’s recent dispositive motions:
To use the information from that rudimentary discovery to identify the more targeted discovery necessary to permit SCO to move forward most efficiently under the Court’s Scheduling Order;
To use the rudimentary (and certain targeted) discovery to depose material witnesses, and to do so with as much information relevant to those witnesses as is practicable, so as not to waste the opportunity; and
To use information obtained in the foregoing ways to oppose the fact-intensive dispositive motions IBM is now pressing.
Unless discovery is permitted to occur in this order, SCO will be prejudiced by having to run the risk of wasting depositions (including potentially not being permitted to re-open depositions) before SCO has obtained the predicate discovery needed to complete them and make them useful.
SCO should be permitted to take the foregoing steps, and to use all of the other opportunities afforded to a litigant in the midst of fact discovery, during the time the Court permitted for such purposes in the Amended Scheduling Order. In any complex case, a litigant reasonably is entitled to the discovery that is relevant to the subject matter of a deponent’s testimony – that is why depositions typically await the production of substantial fact discovery. In this case, the Court also has imposed constraints on the number and time of the depositions the parties may take. In addition, since the Court imposed those constraints, IBM had added its multiple, sweeping counterclaims to the litigation – and yet refuses to increase the number of depositions available for each side. SCO therefore has reasonably awaited IBM’s long-delayed production of material plainly relevant to the depositions of the material witnesses in this case.
By refusing to permit that process to go forward, IBM has precluded SCO from obtaining discovery on numerous issues, including those bearing directly on IBM’s recent dispositive motions. In its July 9 opposition memorandum, for example, SCO demonstrates that it has not had the opportunity to uncover even a substantial amount of the evidence with which SCO would oppose IBM’s motion for summary judgment on its Tenth Counterclaim. See SCO’s SJ Opp. Mem. re IBM’s Tenth Counterclaim at 71-88.
Similarly, in support of its motion for summary judgment on SCO’s contract claims, IBM mischaracterizes SCO’s contract claims as resting entirely on the proposition that AIX as a whole and Dynix as a whole are modifications of, or are derived from UNIX System V. See, e.g., IBM Contract Mem. ¶ 62; see also Part II, above. IBM thus disregards its failure to produce predicate discovery relevant to SCO’s actual claims. SCO has made clear that IBM would violate the license agreements by contributing to Linux any material that included any non-literally copied UNIX source code or any UNIX method or concept. See, e.g., Reply Mem. re Disc. at 8-9. That is an interpretation of the license agreements that is much narrower than the one IBM now derides.
IBM then seeks to support its motion by alluding to SCO’s discovery responses and contending that SCO “fails to offer any evidence that IBM’s purported contributions to Linux (such as RCU, JFS, EVMS and AIO) – extracted from AIX and Dynix and standing alone – are modifications or derivative works of any code in UNIX System V,” IBM’s Contract Mem. at 52. What IBM does not tell the Court, however, is that IBM has blocked the discovery SCO needed to have provided the specific responses whose absence IBM now seeks to exploit. SCO has repeatedly demonstrated to the Court that only upon IBM’s production of specified discovery can SCO undertake reasonably and efficiently to identify the evidence showing that IBM has made contributions to Linux from UNIX derivatives and modifications. See Disc. Mem. at 3; Reply Mem. re Disc. at 7-15. At the same time, when SCO tries to obtain the evidence to support that narrower interpretation of the agreements – the identity of source code files contributed to AIX and Dynix by IBM and Sequent programmers – IBM tells the Magistrate Judge that the discovery is “wholly irrelevant.” IBM’s Mem. in Opp. to SCO’s Renewed Motion to Compel (Aug. 4, 2004) at 9.
If all IBM had contributed to Linux was its own, independently created source code, IBM’s foregoing strategy would seem to be unnecessary. More to the point, SCO’s mischaracterization of SCO’s contract claims and plain disregard for SCO’s discovery requests exemplify the impropriety of IBM’s dispositive motions.
B. IBM Has Failed to Produce Basic Discovery That Is a Necessary
Predicate to the Depositions of Material Witnesses Regarding SCO’s
Contract Claims and IBM’s Contracts Motion
IBM’s failure to produce long-sought and rudimentary discovery has precluded SCO from deposing two principal classes of material witnesses regarding SCO’s contract claims and IBM’s contracts motion.
AIX and Dynix Programmers. Over a year ago, SCO asked IBM to identify the programmers who have made contributions to AIX and Dynix and their specific contributions. In addition, the Court ordered IBM to produce that information over five months ago, in its March Order. Yet IBM has failed to provide any of that information. See Disc. Mem. and Reply Mem. Re Disc.; Compel Mem.; and Reply Compel Mem.
IBM initially claimed that the information was contained in the AIX program itself, but SCO showed in its Renewed Motion to Compel that IBM’s claim was not true – and IBM no longer argues otherwise. SCO further shows in its Reply Memorandum in support of its Renewed Motion to Compel that IBM now takes every step, however contradictory and incredible, to prevent the use of IBM’s CMVC System and Sequent’s RCS, which store all of the information regarding the precise contributions of each person to AIX and Dynix:
IBM first argued that it would take it “many, many months” to use CMVC and RCS to determine the programmer identities and contributions – notwithstanding that IBM is one of the largest computer companies in the world, and has marketed CMVC to its customers as a tool capable of easily producing the information SCO seeks; and
After SCO submitted material demonstrating that IBM could use CMVC and RCS without any significant burden to determine the programmer identities and contributions in a matter of days, IBM moved to strike the declaration demonstrating those facts – but now concedes that it would take only “weeks” to use CMVC to produce the requested information, and make no claim at all about the time it would take to use RCS to produce the information.
See Reply Compel Mem. at 6-13. SCO further shows in its opposition to IBM’s motion to strike that (among other things) in a case of this complexity and magnitude, a production that can be accomplished in a matter of weeks simply cannot, as a matter of law, constitute an “undue burden” sufficient to avoid production under Rule 26(b)(iii). See SCO’s Mem. in Opposition to IBM’s Motion to Strike the Declaration of Chris Sontag (Aug. 26, 2004) at 4-10.
IBM has also failed to produce early versions of AIX, which discovery is necessary to permit SCO to trace the progression of UNIX System V material through AIX and Dynix. See, e.g., Disc. Mem. at 4-14; Reply Mem. re Disc. at 7-20; Compel Mem. 4-8; Reply Compel Mem. at 10-13. IBM does not dispute that it has failed to provide the information. Without that information, SCO could not depose the AIX and Dynix programmers in a remotely complete way. SCO is plainly entitled to ask programmers questions about the code they have contributed to AIX and Dynix; for example, without the early versions of AIX, SCO cannot sensibly depose any programmer about the contributions he made to AIX – because SCO will not even have the contribution in hand.
IBM thus has literally precluded SCO from deposing AIX and Dynix programmers, by refusing even to identify those programmers; and has effectively precluded SCO from deposing those programmers (even if they had been identified), by refusing to produce the evidence of their contributions to AIX and Dynix, or even all of the relevant versions of those programs.
Material Contract Witnesses. In refusing to produce the information showing even the identities of the programmers as well as their contributions, IBM obviously is going to great lengths to prevent SCO from securing these basic depositions. IBM’s intransigence in these respects not only is remarkable on its face – sixibiblio months into a case about software development, SCO still cannot obtain such rudimentary information – but also is highly prejudicial in several ways. Apart from SCO’s obvious inability to effectively depose the programmers themselves absent the additional discovery, the long duration of IBM’s refusal to provide the foregoing information is directly proportional to the utility of that information with respect to SCO’s ability to take other, key contract depositions. IBM’s own motions highlight how that is true.
The bulk of IBM’s contracts summary judgment motion does not concern a pure legal reading of SCO’s license agreements, but rather addresses the extrinsic, factual evidence that IBM offers to support its interpretation of those agreements. IBM presents the declarations of several witnesses who testify that the agreements are quite narrow in scope (such that AT&T would have had more protection from the copying of UNIX if the company had not entered into the agreements at all). The only two witnesses who have been subjected to cross-examination on that issue, however, previously testified – when they were still employed by AT&T, when AT&T’s interests were still at stake, when AT&T was enforcing the agreements, and before their contact with IBM or its counsel – that the same agreements are quite broad in scope. In alluding to the reasons and justifications for the broad scope of the agreements, those witnesses testified to the importance of the programming process by which the licensees would be developing products modified and derived from UNIX. See Part II.A.4, above.
In preparing to take the depositions of the declarants on whom IBM relies, SCO is entitled first to take the depositions of the AIX and Dynix programmers, to show that the foregoing risks inherent in the programming process were also present in the programming processes at issue in this case. In other words, IBM improperly seeks to preclude SCO from cross-examining IBM’s summary judgment declarants with the very information that is the express basis of the contradictory sworn testimony of the only two declarants who were cross-examined regarding the scope of the license agreements. SCO’s ability to present contract deponents with such information concretely (for example, in the form of programmer testimony) is crucial under the circumstances – where as much as twenty years have passed since the agreements were drafted and enforced, but witnesses are now third parties who have neither the incentive nor wherewithal to review their prior testimony or contemporaneous documents bearing on the current testimony, and whom IBM has no incentive to provide with such information prior to their testimony. SCO’s capacity effectively to depose the foregoing contract witnesses will be a function of the extent to which SCO can concretely show and remind these witnesses of the types of risks that concerned AT&T and underlie the license agreements. At this point, SCO could not take any of the foregoing depositions without taking a substantial risk of having to re-open the deposition to ask further, highly relevant questions arising out of documents produced subsequent to the depositions.
There is therefore nothing to be gained, and much to be lost, in accelerating the adjudication of fact-intensive dispositive motions no less than six ahead of the fact-discovery cut-off and without even the basic discovery that IBM’s own motions show to be relevant. The depositions of the two declarants noted above, Otis Wilson and David Frasure, are cases in point. In March and April of this year, IBM produced approximately 670,000 pages of documents to SCO. IBM thereafter noticed the depositions of Messrs. Wilson and Frasure, but not of any of the other witnesses who have given declarations in support of IBM’s motions. SCO objected to the timing of the depositions, pointing out (among other things) that IBM had only recently made its large production of documents. See SCO’s Expedited Motion and Mem. for a Protective Order (June 3, 2004) & Reply Mem. in Support of Motion for Protective Order (June 5, 2004). Only after the depositions had been taken did SCO find that IBM’s production of documents included the deposition transcripts of Mr. Wilson’s and Mr. Frasure’s flatly contradictory sworn testimony excerpted above, and thereby understand why IBM had scheduled the depositions of those particular witnesses when it had. Although SCO reserved its right to re-open the depositions (and plainly is entitled to do so), obviously the more efficient way to have proceeded with the depositions would have been to permit SCO adequate time to receive and process the information relevant to those witnesses’ testimony.
C. IBM Has Failed to Produce Basic Discovery That Will Itself
Bear Directly on SCO’s Contract Claims and IBM’s Contracts Motion
The multiple categories of documents and information that SCO has long sought not only are necessary predicates to the depositions, but would also bear directly on SCO’s contract claims and, therefore, IBM’s contracts summary judgment motion.
AIX and Dynix Programmer Identities and Contributions. In deposing the AIX and Dynix programmers (whom IBM has yet to identify), SCO would:
take the depositions of the programmers that IBM claims produced the entirely “homegrown” code (which IBM tells the Court is all that it wants to protect);
show in those depositions that these programmers were able to, and did in fact, engage in non-literal copying, compounded many times over by repeated reliance on IBM and Sequent’s initial appropriations and subsequent versions based on them; and
show through these depositions that it would be difficult, not easy, for IBM or Sequent to demonstrate what was entirely “homegrown” (given that the programmers were tasked with using AT&T innovations to facilitate the delivery of IBM and Sequent products to market).
If the programmers were going to testify that their contributions did not reflect any copying from UNIX, IBM’s refusal to provide the requested information would be unnecessary.
Earlier Versions of AIX. SCO seeks to demonstrate that IBM’s particular contributions to Linux are modifications or derivatives of UNIX System V. SCO is attempting to do so by tracing the development of IBM’s particular Linux contributions from UNIX System V code through AIX or Dynix and into Linux. IBM has largely frustrated SCO’s efforts to obtain discovery to support that theory. IBM is well aware of SCO’s efforts. SCO has sought production expressly for the purpose of showing this progression. See p. 24 n.5, above. Indeed, in the same letter IBM cites for the proposition that SCO claims it need not trace any particularized development, SCO stated: “SCO, based on the information currently available, has attempted to identify the specific lines of code from UNIX System V from which IBM’s contributions from AIX and Dynix/ptx were derived.” Letter from B. Hatch to T. Shaughnessy (Apr. 19, 2004) at 2 (Exh. 33). But IBM has refused to provide particular versions of AIX and Dynix, omitting steps in the development chain which prevents SCO from tracking the progression of SCO’s proprietary material from version to version and finally into Linux.
CMVC and RCS Information. IBM also continues to refuse to produce other documents SCO has long sought that would bear directly on IBM’s interpretation of the license agreements. IBM continues to refuse to produce, for example, the AIX and Dynix revision information available on the CMVC and RCS systems, including programming notes, design documents, and white papers.
SCO explained in its July 12 memorandum regarding discovery that those materials will provide SCO with extensive information about IBM’s programmers and their involvement with AIX and Dynix. Those materials may contain admissions concerning the meaning of, and limitations imposed by, the license agreements; admissions regarding IBM’s liability for breaching those agreements; and admissions that the development of AIX and Dynix depended on UNIX System V. Even under IBM’s narrow theory of the contract, the materials are relevant because they would permit SCO to test the validity of IBM’s claim that at some point AIX and Dynix ceased to be modifications or derivative works “based upon” UNIX System V and were no longer subject to the restrictions of the license agreements. Even more basically, those documents would logically include programmers’ acknowledgements of the constraints they faced upon exposure to UNIX materials.
SCO has sought the foregoing materials since well before the date (May 28) that SCO submitted its opening memorandum regarding discovery to the Magistrate Judge. SCO specifically sought the CMVC and RCS information as soon as it could do so – that is, upon IBM’s failure in late March of this year to produce any version of AIX. Only after SCO reviewed IBM’s limited production at that time was it able properly to demonstrate to the Magistrate Judge why SCO was then plainly entitled to receive the CMVC and RCS materials. Accordingly, to the extent that SCO continues to litigate to obtain those documents, necessary to permit SCO to depose the material witnesses regarding the contract issues in this case, SCO does so as a consequence of IBM’s delays and inadequate responses to discovery.
SCO reasonably expects, moreover, that the Magistrate Judge will order IBM to produce the foregoing CMVC and RCS materials. In addition to the relevance of those materials to SCO’s contract claims, the materials are directly relevant to IBM’s Ninth Counterclaim, in which IBM asks the Court to declare that “IBM does not infringe, induce the infringement of, or contribute to the infringement of any SCO copyright through the reproduction, improvement, and distribution of AIX and Dynix.” IBM’s Second Amended Counterclaims ¶ 167. In thus seeking a clean bill of health for the entire history of AIX and Dynix, IBM itself puts directly at issue the systems designed to store those histories and to permit others most efficiently to examine those histories.
IBM should not benefit from slowing the timing of its discovery in this way, while accelerating the timing of its dispositive contracts motion no less than six ahead of the discovery schedule. If IBM is to rely on the supposed absence of evidence that it has contributed to Linux code copied from UNIX, it should not be permitted to do so even as it holds back the discovery that would include such evidence.
For all of the above reasons, SCO respectfully requests that the Court grant its motion to enforce the Court’s Amended Scheduling Order and to defer consideration of IBM’s summary judgment motions until after the end of the Court’s fact-discovery period.
DATED this 8th day of September, 2004.
HATCH, JAMES & DODGE, P.C.
Brent O. Hatch
BOIES, SCHILLER & FLEXNER LLP
Robert Silver, Esq. (admitted pro hac vice)
Stephen N. Zack (admitted pro hac vice)
Mark J. Heise (admitted pro hac vice)
Attorneys for The SCO Group, Inc.
See also Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 255 (1986) (the district court may “deny summary judgment in a case where there is reason to believe that the better course would be to proceed to a full trial”); 6 James Wm. Moore et al., Moore’s Federal Practice § 56.15(6) (2d ed. 1989) (same).
The discovery hearing before the Magistrate Judge was scheduled for September 14, the day before this Court’s hearing on the pending motions relating to the Tenth Counterclaim, only because IBM rejected any of the several earlier dates that the Court had provided for that hearing.
SCO also has previously shown that expert testimony is necessary for the Court to consider and resolve the issues of “substantial similarity” underlying SCO’s copyright claim and IBM’s copyright defenses. See SCO’s July 9 Mem. at 63-65. Accordingly, the parties would have to brief for the Court, after fact discovery, the issues surrounding the parties’ expert testimony.
Because responding to IBM’s motions in full at this juncture would undermine the specific purpose of this motion, SCO does not purport herein to oppose IBM’s dispositive motions on the merits or on the basis of Fed. R. Civ. P. 56(f), and reserves the right to do so depending on the resolution of the instant motion.
In its July 12 Reply Memorandum re Discovery, SCO expressly delineated its two theories under two separate section titles. See Reply Mem. re Disc. at 11, 13. SCO explained that “One way for SCO to prove that IBM breached the License Agreements is for SCO to demonstrate that versions of AIX and/or Dynix/ptx are derivatives of System V within the meaning of the Agreements, and that IBM disclosed, exported, or published those derivatives.” Id. at 11 (emphasis added). But SCO sought production of discovery related to, among other things, the “particular” and “precise” contributions of Dynix programmers to Linux, expressly for the purpose of supporting its second theory of liability: “SCO will prevail on its breach of contract claims if it can demonstrate that IBM in fact disclosed protected ‘methods or concepts.’ The most efficient and straightforward way for SCO to discover if IBM made such disclosures is through access to Dynix/ptx programmers and their particular contributions.” Id. at 14 (emphasis added). SCO repeatedly clarified its goal of learning “the identity of Linux contributors” and “specifics about their own Linux contributions” in support of this second theory of contract claims. Id. (emphasis added).
SCO has sought production expressly for the purpose of showing this progression: “SCO, based on the information currently available, has attempted to identify the specific lines of code from UNIX System V from which IBM’s contributions from AIX and Dynix/ptx were derived.” Letter from B. Hatch to T. Shaughnessy dated Apr. 19, 2004, at 2 (Exh. 33).
For more than a year, IBM has refused to provide SCO with the discovery necessary to trace the progression of UNIX System V material through AIX or Dynix into IBM’s particular Linux contributions. IBM has refused to provide particular versions of AIX and Dynix, omitting steps along the development chain which prevents SCO from tracking the progression of SCO’s proprietary material from version to version and finally into Linux. SCO has previously raised these very issues: “To trace the foundations of AIX source code, and to have a complete understanding of Dynix’s source code origins, SCO must have access to all interim and final versions of AIX, Dynix and ptx; to programmer notes and design documents that reveal the work behind the revisions to the programs; and most importantly, to all revision control systems that track changes to AIX and Dynix, thus exposing the sources of IBM’s current AIX and Dynix code an revealing what portions of UNIX System V made their from AIX and Dynix into Linux.” SCO Mem. Regarding Discovery, dated May 28, 2004, at 6. IBM has also refused to provide SCO with each programmer’s particularized contributions to AIX and Dynix, preventing SCO from deposing the very programmers who could identify, as the writers of AIX and Dynix, what role UNIX System V played in the development of each portion of AIX and Dynix source code.
See, e.g., Gates Rubber Co. v. Bando Chem. Indus., Ltd., 9 F.3d 823, 844 (10th Cir. 1993) (setting forth the test for “substantially similarity” with respect to computer programs); SAS Institute, Inc. v. S&H Computer Sys., Inc., 605 F. Supp. 816, 830 (M.D. Tenn. 1985) (finding copyright violation on the basis of substantial similarity of program’s “organization and structural details,” and specifically rejecting the proposition that such similarity could only be demonstrated in “specific lines of code”); Whelan Assocs., Inc. v. Jaslow Dental Labs, Inc., 609 F. Supp. 1307, 1321 (E.D. Pa. 1985) (noting the substantial similarity of two programs in holding that “direct evidence of copying is not required”); Atari, Inc. v. N. Am. Philips Consumer Elecs. Corp., 672 F.2d 607 (7th Cir. 1982) (“When the alleged infringing work is not a duplicate of the copyrighted work, a determination of whether protected elements have been taken requires a close analysis of the differences and similarities between the two works.” (citing E. Kitch & H. Perlman, Legal Regulation Of The Competitive Process 665 (2d ed. 1979))); see also Warner Bros. Inc. v. Am. Broadcasting Prods., 654 F.2d 204 (2d. Cir. 1981) (“it is well settled that copying may be inferred where a plaintiff establishes that the defendant had access to the copyrighted work and that the two works are substantially similar.”); Sobhani v. Radical.Media, Inc., 257 F. Supp. 2d 1234, 1238 (9th Cir. 2003) (holding that defendant’s commercials, similar to plaintiff’s commercials, were unlawful derivative work).
SCO sets forth additional examples of such prior testimony given by IBM’s declarants (which is flatly inconsistent with their more recent declarations) in Part II.A.4, below.
Indeed, SCO intends to file, at the appropriate time, a motion for summary judgment on its contract claims, in which SCO will show that the plain, unambiguous meaning of the license agreements in this case support SCO’s position. Of course, the goal of efficient case management would best be served by consideration of all summary judgment motions concerning the contract interpretation issues in this case at the same time, following the close of fact discovery.
Section 1.04 of the license agreements defines “SOFTWARE PRODUCT” as follows: “materials such as COMPUTER PROGRAMS, information used or interpreted by COMPUTER PROGRAMS and documentation relating to the use of COMPUTER PROGRAMS.” The Section also refers to the Schedule accompanying each license agreement for a listing of the “Materials available from AT&T for a specific SOFTWARE PRODUCT.”
The license agreements contain a New York choice-of-law provision. See Exh. 13 § 7.13, Exh. 14 § 7.13
This language was included in the original IBM and Sequent license agreements, as well as in the IBM Side Letter.
This plain language reading of the IBM side letter is only reinforced by AT&T’s subscription newsletter, $ echo, on which IBM erroneously attempts to rely for its misguided argument. See IBM Contract Mem. at 63. The April and August 1985 editions of that newsletter simply announced – as the sections quoted by IBM make clear – that AT&T would be revising its standard agreement to address the ownership of modifications and derivative works prepared by licensees. See Exhs. 10, 11, as with the IBM Side Letter Agreement, the newsletter never suggested in any way that the clear and explicit use and disclosure restrictions in the license agreements would be diminished or eliminated.
The IBM Side Letter Agreement amended Section 7.06(a) of the IBM software license agreement to read, in relevant part, as follows:
“LICENSEE agrees that it shall hold SOFTWARE PRODUCTS subject to this Agreement in confidence for AT&T. LICENSEE further agrees that it shall not make any disclosure of such SOFTWARE PRODUCTS to anyone, except to employees of LICENSEE to whom such disclosure is necessary to the use for which rights are granted hereunder. LICENSEE shall appropriately notify each employee to whom any such disclosure is made that such disclosure is made in confidence and shall be kept in confidence by such employee. Nothing in this agreement shall prevent LICENSEE from developing or marketing products or services employing ideas, concepts, know-how or techniques relating to data processing embodied in SOFTWARE PRODUCTS subject to this Agreement, provided that LICENSEE shall not copy any code from such SOFTWARE PRODUCTS into any such product or in connection with any such service and employees of LICENSEE shall not refer to the physical documents and materials comprising SOFTWARE PRODUCTS subject to this Agreement when they are developing any such products or services or providing any such service.” Exh. 12 at 4 (emphasis added).
16 To support its claim that its interpretation of the IBM Side Letter should also impact on the Sequent agreement, IBM even goes so far as to rely on contract language that was not included in either of the IBM or Sequent agreements. See IBM Contract Mem. at 54 (citing Exh. 39, a license agreement between AT&T and The Santa Cruz Operation).
17Accord State v. Home Indem., 66 N.Y.2d 669, 671 (1995) (the trier of fact, not the court on summary judgment, must resolve ambiguity in contract); Wells v. Shearson Lehman, 72 N.Y.2d 11, 27 (1988) (the court’s determination that the contract is ambiguous “necessitates extrinsic evidence” and operates to “bar summary judgment”); Aubuchon Realty v. Fid. Nat’l Title Ins., 743 N.Y.S.2d 626, 628 (App. Div. 2002) (court cannot interpret contract as a matter of law if “there is ambiguity in the terminology used in the instrument and resolution of that ambiguity depends upon extrinsic evidence”) (emphasis in original)); Ruttenberg v. Data Sys., 626 N.Y.S.2d 174, 175 (App. Div. 1995) (where the court determines the contract to be “ambiguous and the intent of the parties becomes a matter of inquiry, a question of fact is presented which cannot be resolved on a motion for summary judgment”); Chem. Bank v. Flushing Sav. Bank, 536 N.Y.S.2d 442, 446 (App. Div. 1989) (emphasizing that where “the intent of the parties to a contract is not clear from the language of the contract itself must be determined by disputed evidence or inferences outside of the written agreement, we have not hesitated to deny or to reverse a grant of summary judgment”).
18 The other cases IBM cites in support of its proposition that a court can find summary judgment when extrinsic evidence is exceedingly one-sided are out-of-jurisdiction cases that fail to address the critical shortcomings of the summary judgment motion at issue in this case – filed on the basis of a wholly incomplete record, before the close of discovery, and in reliance on untested affiant declarations. See p.33, below. In Compagnie Financiere de CIC et de L’Union Europenne v. Merrill Lynch, Pierce, Fenner & Smith, Inc., 232 F.3d 153, 161 (2d Cir. 2000), as in Moncrief (discussed above), the court’s decision rested on the case’s unique procedural posture; the Second Circuit emphasized that because the summary judgment motion at issue “followed a bench trial, the trial record was before it.” The Court also observed that neither party had suggested the existence of any additional evidence not already submitted at trial: “The parties confirmed at oral argument that the record was complete with regard to extrinsic evidence supporting their interpretations.” Id. (emphasis added); see also, e.g., Farmland Indus. v. Grain Bd. of Iraq, 904 F.2d 732, 736 (D.C. Cir. 1990) (motion for summary judgment was filed more than six weeks after the close of discovery).
19 In addition, IBM’s reliance on the doctrine of contra proferentem is misplaced. Contra proferentem “is applied only in cases where there is an ambiguity in some terminology of the policy [contract] which cannot be resolved by the presentation of extrinsic evidence.” Alfin, Inc. v. Pac. Ins. Co., 735 F. Supp. 115, 121 n.5 (S.D.N.Y. 1990). Accordingly, courts have often rejected the application of the doctrine in the summary judgment context. See Schering Corp. v. Home Ins. Co., 712 F.2d 4, 9-10, (2d Cir. 1983) (refusing to apply contra proferentem because “the party opposing summary judgment propound[ed] a reasonable conflicting interpretation of a material disputed fact”); Liszt v. Karen Kane, Inc., No. Civ.A.3:97-CV-3200-L, 2001 WL 739076, at *10 (N.D. Tex. Jun. 26, 2001) (Exh. A) (“The interpretive rule of contra proferentem therefore runs into a conflicting, and stronger, rule in this context – that the existence of a fact question as to an ambiguous contract precludes summary judgment.”); Commonwealth Ins. Comp. v. Stone Container Corp., No. 99 C 8471, 2003 WL 260713, at *2 (N.D. Ill. Jan. 31, 2003) (Exh. C) (“We will not apply the rule of contra proferentem on summary judgment when there are substantial questions of fact as to the intent” of the parties). Moreover, several cases suggest that the doctrine of contra proferentem may be completely inapplicable to the contracting circumstances of this case, which indisputably involved arms-length deals between sophisticated and well-represented parties. See 67 Wall Street Co. v. Franklin Nat’l Bank, 37 N.Y.2d 245, 249 (1975) (holding that contra proferentem benefits parties “who had no voice in the selection of [the agreement’s] language”) (emphasis added); Hertzog, Calamari & Gleason v. Prudential Ins. Co. of Am., 933 F. Supp. 254, 258-259 (S.D.N.Y. 1996) (noting that “the parties were in parity with each other” and thus “the contra proferentem maxim has no proper office to perform.”); Eley v. Boeing Comp., 945 F.2d 276, 280 (9th Cir.1991) (“Although contra proferentem is strictly applied in the interpretation of insurance contracts, it is not automatically or universally applied to other contracts, especially those that result from arms-length bargaining by parties of equal power.”)
20 AT&T’s UNIX licensing agreements employed substantially similar terminology designed to protect the same rights and privileges vis-a-vis AT&T’s various licensees. See Pfeffer Decl. at 2 (Exh. 27) (“Although not every single such license agreement was identical, they generally contained substantially similar terms and reflected the same intent, especially insofar as those agreements concerned protections for the licensor’s intellectual property and other proprietary rights in UNIX.” (emphasis added)). IBM’s own witnesses concede as much. See, e.g., Wilson Dep. at 259 (Exh. 15) (“The terms and conditions were pretty much set in a boilerplate, and any negotiation was usually just clarification to determine which software product someone needed. So there wasn’t a specific negotiation with individual licensees that would be any different than the boilerplate standard agreement.” (emphasis added)). Accordingly, the witnesses’ testimony and writings concerning substantively similar (and often identical) language in other UNIX licensing agreements is indisputably relevant to the interpretation of the particular licensing agreements at issue here.
21 Mr. Frasure’s counsel, James G. Szymanski, has served in several in-house legal capacities for IBM, and has represented IBM in numerous notable cases. See Exh. 17. Aside from Messrs. Wilson and Frasure, IBM’s counsel have entered into attorney/client relationships with several former employees of AT&T and USL – companies that are SCO’s predecessors in interest to the UNIX code and business – thereby precluding SCO from even talking to these individuals (who are, in effect, its former employees) about the very issues that IBM claims are not disputed.
22The license agreement at issue in the BSD case had defined “LICENSED SOFTWARE” consistently with the definition of that term in the IBM and Sequent agreements – in relevant part, as “all or any portion of the computer programs, other information and documentation prepared by LICENSEE as a modification or derivative work based on any of the materials so listed or furnished” in the schedule attached to the agreement. Exh. 20 (Educational Agreement dated July 1, 1983, at 1).
23 See, e.g., Gilbert Frank Corp. v. Fed. Ins. Co., 70 N.Y.2d 966, 968 (1988) (holding that party seeking to establish waiver must demonstrate a “clear manifestation of [the other party’s] intent” to waive contractual rights); Civil Serv. Employees Ass’n v. Newman, 450 N.Y.S.2d 901, 903 (App. Div. 1983) (“a waiver must be clear, unmistakable, and without ambiguity”).
24SCO has explained: “To trace the foundations of AIX source code, and to have a complete understanding of Dynix’s source code origins, SCO must have access to all interim and final versions of AIX, Dynix and ptx; to programmer notes and design documents that reveal the work behind the revisions to the programs; and most importantly, to all revision control systems that track changes to AIX and Dynix, thus exposing the sources of IBM’s current AIX and Dynix code and revealing what portions of UNIX System V made their from AIX and Dynix into Linux.” Disc. Mem. at 6.
25 In its First Request for Production of Documents (June 24, 2003), SCO sought “All documents that identify any person at IBM and Sequent who had access to AIX source code” to “Sequent Dynix source code.” Exh. 32. In its Second Request for Production of Documents (Dec. 4, 2003), SCO sought “All source code for AIX and Dynix since 1985 including all instructions, information used and all documentation relating to the use of AIX and Dynix, including but not limited to, all development notes.” Exh. 33.