IBM's Memorandum in Opposition to SCO's Motion to Compel Discovery -- as Text

Friday, November 21 2003 @ 08:11 PM EST

Contributed by: PJ

Here, as promised, is IBM's Memorandum in Opposition to SCO's Motion to Compel Discovery. It is available as a PDF here. The clerk has put up a notice that December 5th is confirmed as the day for oral arguments on IBM's Motion to Compel, which, if accurate (and we haven't always found the clerk's records to be so) would indicate that today's conference with Judge Wells did not resolve the issues. So, it's on for December 5.

This document, however, relates to SCO's later Motion to Compel Discovery. I do not know if it was raised today at the conference, or if it will be raised at a future date. We'll learn those details in due time.

The heart of IBM's argument is this sentence: "A party may only obtain discovery that is relevant to the claims or defenses in a lawsuit." They are pointing out that you can only ask for things pertinent to the actual case, and since SCO refuses to identify what the case is about, it prevents IBM from being able to evaluate if their requests are legitimate or not. SCO shouldn't be allowed to blindside them, they are saying. This is a powerful argument, because by refusing to identify the actual code, IBM is being deprived of a basic legal right relating to discovery.

One other point you will notice: SCO has complained that IBM didn't provide info about third parties. IBM says the parties agreed they would not ask each other for that third-party information in discovery; however, as long as it is reciprocal, they would be happy to alter course and provide it. Clever. They know that SCO has more to lose than to gain by such an agreement. And don't skip the footnotes. In this document, a lot of the meatiest arguments are found there.


Alan L. Sullivan (3152)
Todd M. Shaughnessy (6651)
Amy F. Sorenson (8947)
Snell & Wilmer L.L.P.
[ street address ]
[ city, state, zip ]
Telephone: [ number ]
Facsimile: [ number ]

Evan R. Chesler (admitted pro hac vice)
Thomas G. Rafferty (admitted pro hac vice)
David R. Marriott (7572)
[ building ]
[ street address ]
[ city, state, zip ]
Telephone: [ number ]

Attorneys for Defendant/Counterclaim–Plaintiff
International Business Machines Corporation








Civil No. 2:03cv0294

Honorable Dale A. Kimball

Magistrate Judge Brooke Wells

Defendant/Counterclaim-Plaintiff International Business Machines Corporation("IBM") respectfully submits this Memorandum in Opposition to Plaintiff/Counterclaim-Defendant The SCO Group, Inc.'s ("SCO") Motion to Compel Discovery.

Preliminary Statement

SCO has moved to compel responses to six of its discovery requests. The motion is misguided and should be denied. As an initial matter, much of what SCO seeks IBM has already agreed, or is willing, to produce as soon as is practicable. More fundamentally, SCO's motion should be denied because most of what it seeks is irrelevant, overly broad and unduly burdensome, particularly in light of SCO's continued refusal to particularize its claims. Unless and until SCO tells us what this case is about, IBM should not be required to collects tens of millions of ages of documents and gather information that is very likely irrelevant to the case.



SCO's Document Request No. 11 seeks all contributions made without confidentiality restrictions by IBM to Linux. SCO's Interrogatory No. 5 asks IBM to "[i]dentify all IBM or Sequent personnel that work or worked on developing source code, derivative works, modifications or methods for AIX, Dynix and Linux, specifying for each person their precise contributions to each". In its motion to compel, SCO complains that IBM has refused to produce meaningful discovery relating to its Linux contributions and its development work regarding AIX and Dynix. Specifically, SCO complains that IBM has (1) failed adequately to identify all persons at IBM or Sequent who have ever worked on developing AIX or Dynix and their precise contributions (see Pl. Mem. at 12-13); and (2) refused to identify and produce all of its contributions and development work in Linux (see id. at 9, 13). SCO's complaints are baseless. [1]

First, in response to the request that IBM identify the personnel who have worked on developing AIX and Dynix and the precise contributions of each, IBM identified more than 8,000 individuals who may have, or may have had, access to AIX or Dynix source code, or to defect or fix records. Based upon our investigation to date, we believe that the persons who have or have had such access are the persons who worked on developing AIX and Dynix.

SCO's request that IBM identify the precise contribution that each of these thousands of employees made to AIX or Dynix (over the decades of their development), however, is overly broad and unduly burdensome. AIX and Dynix are complex software programs, comprised of a myriad of files and individual lines of code, and parsing out exactly which individuals did what to every piece of code in the programs is a huge undertaking. It makes no sense to require IBM to chronicle the history of every line of code in AIX and Dynix, when very little (if any) of it will ever have anything to do with case. What makes sense, we submit, is for SCO to identify the particular lines of AIX or Dynix source code that IBM is alleged to have misused and for IBM then to provide SCO with the development history of that code to the extent possible.

Second, SCO's request that IBM identify (and produce documents constituting) all of its contributions to Linux is similarly broad and unduly burdensome. IBM has made a lot of contributions to Linux, all of which are, by definition, publicly accessible. Largely because the composition of Linux is a matter of public record, IBM does not maintain a central repository comprising all of its contributions. To identify all its contributions to Linux, IBM would have to either (1) search Linux for the term "IBM" or the names of IBM contributors (which we have provided to SCO); or (2) interview each of the hundreds of persons who we believe may have made a contribution to Linux. IBM should not be required to search Linux for its contributions when SCO can do that itself (as it purports to have done). Nor should IBM be required to interview hundreds of employees in an effort to reconstruct a record to which SCO already has access. [2]

Moreover, it makes no sense for IBM to have to identify and produce all of its contributions to Linux when SCO has represented publicly that it knows the "offending" code. When SCO tells us that information, it may make sense to require IBM to interview the individuals who made contributions relating to the lines of code that SCO identifies. We are not there yet, however, since SCO has not identified a single UNIX file or line of code that IBM is alleged to have misused or misappropriated.

IBM is not required to undertake a broad and burdensome effort to collect documents and information that may or may not be relevant to the case. See Leucadia, Inc. v. Applied Extrusion Techs., Inc., 755 F.Supp. 635, 637 (D. Del. 1991) (granting defendant's motion to postpone discovery until after defendant concluded its discovery regarding plaintiff's alleged trade secrets). [3] Nevertheless, IBM has proceeded with discovery as best it can based upon our understanding of SCO's allegations. Indeed, we have already collected and produced a significant quantity of documents and volume of information responsive to Document Request No. 11 and Interrogatory No. 5. [4]

IBM should therefore not be required to make any further disclosures of this kind until SCO provides specifics about its case. A party may only obtain discovery that is relevant to the claims or defenses in a lawsuit. See Fed. R. Civ. P. 26(b)(1). The only way to determine whether the information SCO seeks is relevant to this case -- and therefore subject to discovery -- is for SCO to do for IBM what it has long been doing for third parties: disclose its supposed evidence and tell us in a meaningful way what its claims are about. It is not possible properly to determine what is fairly discoverable without knowing more specifically what trade secrets or confidential information is at issue. Requiring IBM to provide the discovery SCO seeks would not only be unduly burdensome but it would also be counter to sound principles of case management, which require SCO first to define the scope of the case.


SCO's Document Request Nos. 2 and 3 seek all versions or iterations of AIX and Sequent Dynix source code, modifications, methods and/or derivative works from 1999 to the present. SCO's motion to compel responses to these requests should also be denied.

IBM has already agreed to produce source code for all releases of AIX and Dynix distributed within the time frames specified in SCO's requests. ( See Exhibit A.) These source code files alone represent the equivalent of at least an approximately 3 million pages of documents.

The additional documents that SCO is apparently seeking here -- "all versions or iterations" of AIX and Dynix source code -- are estimated to represent the equivalent of more than 40 million pages. [5] Producing these documents would impose undue burden on IBM, particularly as there is no reason to believe that any of this code is at issue in the case. Once SCO identifies with particularity the alleged trade secrets or confidential information that IBM is alleged to have misused, IBM will be able to engage in a more manageable and reasonable search for any development history of any particular file. [6]

SCO's suggestion that IBM is deliberately dragging out the process of providing third-party notices or obtaining third-party consents is without merit. Promptly upon receipt of SCO's discovery requests, IBM began the process of determining what steps, if any, were required to notify or obtain the consent of third parties before producing the documents SCO seeks. Substantial time and energy have been devoted to this exercise. [7] Although the process has not yet run its course, we expect shortly to begin producing the documents at issue, absent the objection of a third party. [8]


By its Interrogatory No. 2, SCO (1) seeks the names and addresses of the persons who are believed or known by IBM to have knowledge concerning the issues in this lawsuit and (2) requests that IBM specify the subject matter about which the person has knowledge. SCO moves to compel supplemental answers to this interrogatory on the grounds that IBM did not name any third parties in response to SCO's request, and further, that it neglected to name certain other IBM employees, officers or directors that SCO believes have relevant knowledge.

IBM limited its response to Interrogatory No. 2 to IBM employees because it believed the parties had agreed not to identify any third parties in their responses to interrogatories calling for the identification of persons with knowledge. SCO apparently has a different understanding of the parties' agreement, despite the clear correspondence between the parties. [9] So long as the parties' obligations are reciprocal, IBM does not object to supplementing its answer to this interrogatory to include third parties believed to have knowledge about the issues in this lawsuit.

Further, IBM did not neglect to include any IBM officers and directors in its discovery responses. We do not believe that IBM's CEO, or any IBM directors, have knowledge concerning the issues of this lawsuit. We assume SCO does not intend Interrogatory No. 2 to call for the identification of anyone who is familiar with IBM's business or is generally aware of SCO's case.


SCO's Interrogatory No. 4 asks IBM to identify all persons who have, or have had, access to UNIX source code, AIX source code and Dynix source code and the precise materials to which each person had access. SCO complains in its motion that IBM has refused to provide contact information for each of the more than 8,000 persons who have, or had, access to AIX, Dynix or UNIX System V source code, and that IBM has not identified any former IBM or Sequent employees who had historical access to the Dynix source code. SCO's complaints are without merit.

First, SCO's request that IBM provide contact information for the more than 8,000 individuals identified is unduly burdensome, and there is no good reason to require IBM to provide such information. It would take IBM more than 500 hours to compile this information, assuming (conservatively) it takes no more than 4 minutes per person to assemble. Moreover, SCO does not require such information in order to "evaluate the list of names [IBM] provided" as it claims. (Pl. Mem. at 11.) There is nothing to "evaluate". SCO asked IBM to identify persons with access to source code, and we did. [10]

Second, although it is correct that Attachment C to IBM's response to Interrogatory No. 4 is limited to persons at IBM with current access to Dynix code, this is not because IBM was withholding information from SCO. IBM provided all of the information it could find and assemble after a reasonsable search -- the business records that IBM maintains in the ordinary course of its business do not identify persons with historical access to Dynix code. We continue to try to identify persons who used to (but no longer) have access to Dynix code, and we will amend IBM's response to this interrogatory at appropriate intervals, as information becomes available.


For the foregoing reasons, IBM respectfully requests that the Court deny SCO's Motion to Compel Discovery.

DATED this 19th day of November, 2003.

Alan L. Sullivan
Todd M. Shaughnessy

Evan R. Chesler
David R. Marriott

Attorneys for Defendant/Counterclaim-Plaintiff
International Business Machines Corporation

Of counsel:

Donald J. Rosenberg
Alec S. Berman
[address, etc.]
Attorneys for Defendant/Counterclaim-Plaintiff
International Business Machines Corporation

[1] In support of its motion, SCO purports to characterize IBM's contractual obligations to SCO. As we have stated in previous submissions, IBM does not agree with SCO's interpretation of the contracts at issue.
[2] SCO argues that IBM should be required to disclose all of its contributions to Linux, irrespective or whether they are publicly available, on the ground that only contributions actually incorporated into Linux, as opposed to all contributions, are publicly available. That is false. Contributions that are not accepted into any Linux kernel or Linux distribution are publicly accessible, for example, at sites such as and Furthermore, while it is possible that a proposed contribution to Linux that was not actually incorporated into Linux might not be publicly accessible, any such contribution is hardly relevant here. SCO could not have been injured (in a cognizable way) by a contribution that was ignored.
[3] See also Xerox Corp. v. International Business Machines Corp., 64 F.R.D. 367, 371 (S.D.N.Y., 1974) ("[Plaintiff] should be able to identify in detail the trade secrets and confidential information alleged to have been misappropriated by [defendant]. Clearly until this is done, neither the court nor the parties can know, with any degree of certainty, whether discovery is relevant or not; and it is doubtful whether [plaintiff] can undertake a meaningful discovery program".); Microwave Research Corp. v. Sanders Assocs., Inc., 110 F.R.D. 669 F.R.D. 669, 673-74 (D. Mass. 1986) (halting broad discovery of defendant where plaintiff listed products and methods that it claimed embodied trade secrets rather than the trade secrets themselves).
[4] IBM has produced more than 120,000 pages of documents, many of them relating to contributions to Linux. IBM continues to produce responsive documents -- both from its central files and from individuals' files -- on a rolling basis as they are located, reviewed, and processed. To date, IBM has produced documents from at least 90 different different sources.
[5] SCO's motion also seeks to compel docouments outside the scope of its document requests. For example, SCO asks that IBM be required to produce development and design methods of AIX and notes relating thereto and development and design methods of Dynix and notes relating thereto. ( See Pl. Mem. at 5, 7.) Document Request Nos. 2 and 3 do not seek such notes.
[6] SCO makes much of the fact that IBM has stated that it intends to produce the "base operating system" for AIX and Dynix. In fact, we are not limiting our production of AIX or Dynix to the base operating system.
[7] To respond to SCO's discovery requests, IBM has had to identify and review a substantial quantity of documents, including the documents requested in SCO's Document Request Nos. 2 and 3. In addition to being reviewed for responsiveness and privilege, each of these documents has to be reviewed to ensure that it is not produced without a required third-party notice or company-specific and a contract-specific inquiry. Upon locating a document requiring third-party notice or consent, we have followed the process set out in each contract. Most of the contracts require IBM to send written notice and a request for consent to the third party. Absent objection by the third party, and in accordance with Paragraph 10 of the Protective Order, IBM will be able to produce the documents at issue.
[8] Despite SCO's objection to IBM's providing notice and/or seeking the consent of third-parties, SCO recently notified IBM that it too has withheld documents from production on these very grounds. Similarly, SCO has failed to produce the very kinds of documents that it now seeks to compel IBM to produce.
[9] On September 15, 2003, IBM informed SCO that it would identify IBM officers and employees believed to have knowledge about the issues in this lawsuit and that it expected to amend its understanding as soon as practicable. (See Exhibit B.) Thereafter, both parties confirmed our understanding that the parties would indemnify only their employees, officers, and directors. (See Exhibits C, D and E.) SCO advised IBM that it "disagree[d] with the scope of the list of witnesses that [the parties] agreed to exchange" only after IBM submitted its supplemental responses. (Exhibit F.) Counsel for SCO then stated that he would "address [the issue] in greater detail in my correspondence later this week" ( id.), but SCO filed this motion to compel instead.
[10. If SCO had met and conferred with IBM before seeking to compel a response to this interrogatory, we would have suggested that, rather than having IBM compile thousands of addresses (few of which SCO is likely ever to require), the parties reach a compromise pursuant to which IBM would provide SCO with contact information for up to 100 persons on an as-needed basis (such as if and when SCO determines that it actually has reason to contact a listed person).