***************************************

Brent O. Hatch (5715)
HATCH, JAMES & DODGE
[address, phone]

Robert Silver (admitted pro hac vice)
BOIES, SCHILLER & FLEXNER LLP
[address, phone]

Stephen N. Zack (admitted pro hac vice)
Mark J. Heise (admitted pro hac vice)
BOIES, SCHILLER & FLEXNER LLP
[address, phone]

Attorneys for Plaintiff

________________________

IN THE UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF UTAH, CENTRAL DIVISION

____________________________

THE SCO GROUP, INC.

Plaintiff,

v.

INTERNATIONAL BUSINESS
MACHINES CORPORATION,

Defendant.

__________________________

PLAINTIFF SCO'S REPLY MEMORANDUM
IN SUPPORT OF ITS MOTION TO
AMEND THE SCHEDULING ORDER

__________________________

Case No. 2:03CV0294DAK

Judge Dale A. Kimball
Magistrate Judge Brooke C. Wells

______________________________

Plaintiff/Counterclaim-Defendant The SCO Group, Inc. ("SCO") respectfully submits this reply memorandum in support of its motion to amend the scheduling order.

PRELIMINARY STATEMENT

This is a complex intellectual property dispute turning on, among other things, technical issues about confidential software programs. SCO's requested amendment to the scheduling order would provide for a discovery plan that deals with the substantial practical problems and issues that have arisen, while only extending the trial date by five months. This limited amendment is necessary, in part, because IBM did not begin to produce crucial discovery, including any versions of its confidential operating system known as AIX - a program concededly central to this case - until late March of this year.(1) IBM still has not produced critical discovery that SCO needs, and SCO is following the procedures established by the Magistrate Judge in March to obtain this discovery. Nevertheless, the record shows that SCO has acted with full diligence and has been found to have complied with its obligations in "good faith."

Yet, IBM proposes to begin depositions immediately and complete 78 of them in the 45 business days that remain before the discovery deadline - an obviously impossible schedule that would force SCO to take its depositions without the written and documentary discovery (or the ability to prepare based on discovery) that is routinely available to any plaintiff in the present type of case. Even SCO's proposed schedule would entail an average of 10 depositions per month, many of them highly technical, for eight-straight months around the United States and possibly abroad - an aggressive schedule under any view, but one that seeks to complete the remaining work as quickly as practicable while maintaining SCO's opportunity for a fair adjudication of its claims.

IBM opposes SCO's motion on the dramatic, although generalized, grounds that SCO's present request is actually part of an "improper scheme" to impose "fear, uncertainty, and doubt" on the industry. To prove this "scheme" IBM says SCO is responsible for the delay at issue because SCO has not timely responded to all IBM discovery and because SCO filed a claim that it later discontinued.

IBM conversely argues that IBM did not create any need for any additional time when (1) IBM itself filed a broad patent counterclaim and later dropped it; (2) IBM itself deferred or decided not to provide discovery, including discovery as basic and central as its own AIX and Dynix code; and (3) IBM itself added numerous counterclaims to a schedule that concededly was not and could not have been designed to accommodate such claims.

The record demonstrates that each of IBM's contentions is wrong. For example, throughout its brief IBM accuses SCO of "misconduct," "malfeasance," and "disregard of two Court orders."(2) This pervasive invective asks this Court to simply assume that IBM's discovery (and merits) positions have already been adjudicated to be proper, and that SCO's positions have already been adjudicated to have been wrongful. Although IBM places dispositive reliance on these assumptions, the record contradicts each one:

• IBM tells this Court that "SCO still refuses to identify the specific code from UNIX System V that IBM is claimed to have misused" in connection with its breach of license and other claims. In fact, SCO has identified a great deal of specific code that SCO expressly contends IBM misused in violation of its contract with SCO (See Section I below);
• IBM repeatedly asks this Court to assume that SCO - because it asserted it needed other versions of IBM programs to respond to certain IBM discovery - has contemptuously acted in "disregard of two Court orders." In fact, the Magistrate Judge found the very opposite - that SCO in dealing with discovery has acted in "good faith" to comply with the Court's Order. March 3, 2004 Order at 3 (See Section II below);(3)
• IBM asks this Court to assume that SCO was sanctioned with a stay of discovery because it was unable to support an element of its trade secret claims and withdrew that claim. In fact, the record plainly shows that the discovery stay was imposed because each side claimed it needed more discovery from the other - and because the Magistrate Judge believed that SCO, as plaintiff, should go first. (See Section II below). Under IBM's logic, IBM should also have been sanctioned because it also withdrew a patent counterclaim after it had been pending for a comparable period of time;
• IBM asks the Court to assume that none of its conduct caused substantial delay. In fact, IBM repeatedly promised production of AIX and Dynix source code in 2003, but forced SCO to move to compel that production based on IBM's continuing failure to actually produce the necessary source code. After the Magistrate Judge ordered that production, SCO finally obtained its first access to any usable version of the essential AIX source code in late March of this year (See Section I below); and
• IBM asks this Court to assume that its refusal to provide the remaining requested versions of its programs has already been held to be proper - and that SCO has been contemptuous in claiming that it needs such material to respond to certain IBM discovery. In fact, to the contrary, the Magistrate Judge's March 3, 2004 Order establishes a now ongoing procedure to give SCO an opportunity to demonstrate that very need - a procedure that would make no sense and not exist if IBM's repeated characterizations of SCO were accurate. (See Section II below).(4)

Just as IBM mischaracterizes SCO's discovery conduct and SCO's theories of the case, it similarly mischaracterizes public statements made by SCO about the case and about the status of discovery generally. For example, IBM falsely states that a Vice President of SCO publicly reported that "SCO's strategy in this case is not to `put everything on the table at the start but instead [to] bring out arguments and evidence piece by piece."' (See IBM Br. at 5). Contrary to IBM's representation, the statement made in a German magazine was not a statement about SCO's "strategy in this case," but rather was an explanation of "the particular way litigation is conducted in the United States," as opposed to Europe, and a clarification as to why confidentiality agreements are needed in the United States prior to viewing evidence.(5)

I. IBM'S DISCOVERY CONDUCT HAS CAUSED SUBSTANTIAL DELAY AS A MATTER OF UNDISPUTED RECORD FACT

The undisputed record contradicts IBM's remarkable claim to this Court that it "has never delayed in providing SCO with the discovery to which it is entitled."(6)

A. AIX and Dynix Code

IBM has substantially delayed production of its AIX and Dynix source code, which is at the heart of this case.(7) One of SCO's core claims is that IBM breached its license agreements when it contributed restricted AIX and Dynix source code to Linux. Without access to this source code, SCO would have to rely on nothing more than IBM's public pronouncements that it had contributed AIX and Dynix source code to Linux - statements that SCO had access to before filing this case and which supplied a basis for institution of this action. IBM, of course, would certainly contend that such statements, without more, were insufficient as proof of a breach of its agreement. Production of this source code will enable SCO to identify the specific files and lines of AIX and Dynix that IBM contributed to Linux and to continue the complex and technically demanding analysis necessary to identify all of the instances of IBM's irrefutable copying from AIX and Dynix into Linux.

To prove these matters, almost a year ago, on June 24, 2003, SCO requested IBM to produce "all versions or iterations of AIX and Dynix source code, modifications, methods and/or derivative works thereof' from 1999 to the present.(8) Although IBM conceded on August 13, 2003, that its AIX and Dynix code was relevant and agreed to produce versions it selected,(9) IBM failed to produce even a single line of code from either AIX or Dynix between June 24, 2003 and December 4, 2003. On that date, which was the day before the hearing on SCO's motion to compel production of the source code, IBM finally produced two CDs containing limited versions of Dynix. IBM still did not produce a single line of AIX code at that time.

IBM claimed in August 2003 that it had to delay production of this code due to a need to obtain third-party consents, but IBM waited until October 2003, before even seeking these approvals. After hearing argument on SCO's motion to compel, on March 3, 2004, Magistrate Judge Wells, as part of an overall Order lifting a temporary discovery stay, also issued a specific numbered directive requiring that IBM finally produce at least some versions of AIX code (and additional Dynix code).(10) On March 4, 2004, almost nine months after SCO originally requested its production, SCO finally received limited versions of AIX and additional Dynix source code so that SCO could begin to conduct the necessary code comparisons to prove its case. Even then, however, as discussed below, the source code was first produced in a format that was unusable (and that IBM knew or certainly should have known would be unusable). This is certainly inconsistent with IBM's representation to this Court that IBM "has never delayed in providing SCO with discovery" in this case.

B. Documents

IBM's delays were not limited to its source code. IBM did not produce documents from several of its top executives and key personnel in response to SCO's initial request for responsive documents. For example, Sam Palmisano, one of the architects of IBM's Linux strategy and the current CEO of IBM, was not listed as a source of documents. As a result, as of March 3, 2004, when Magistrate Judge Wells granted SCO's motion to compel, IBM had not produced a single document from the files of Mr. Palmisano or many of the key senior IBM officials responsible for overseeing IBM's Linux operations - some of the central witnesses in the case. Incredibly, among the documents IBM did not produce was a ten-page memo to IBM's senior management concerning IBM's decision to embrace Linux. This document has since been ordered to be produced per the March 3, 2004 Order. As detailed in SCO's motion to compel leading up to that Order, SCO was aware of this document only because IBM provided a copy of it to The New York Times.(11)

IBM needlessly delayed other discovery. When SCO produced CDs, it did so with the necessary document breaks to facilitate electronic review, IBM did not. When SCO brought this defect to IBM's attention in October 2003, IBM did nothing. Again, this is not the conduct of a party that "has never delayed in providing SCO with the discovery to which it is entitled."(12)

C. Witnesses

IBM's approach to witnesses has not been different. Rather than identify the witnesses with knowledge about the subject matter of the Complaint, IBM responded both over-inclusively and under-inclusively, providing a list of over 7,200 current and former IBM employees only while listing no third parties. IBM's unnecessarily lengthy list of internal employees significantly excluded IBM's CEO Sam Palmisano and Irving Wladawsky-Berger, key executives who are intimately involved in the Linux project in which IBM contributed AIX and Dynix to Linux which is central to SCO's case. Moreover, IBM did not even provide the most basic information concerning the 7,200 witnesses it did identify - not even telephone numbers and addresses.

On March 3, 2004, Magistrate Judge Wells ordered IBM to provide the requested information for "1,000 representative witnesses" taken from IBM's list of 7,200. However, when SCO sent a request for the identity of all persons on Exhibit E to IBM's earlier answers to interrogatories and added a list of 81 additional witnesses, IBM balked. It refused to provide the contact information for over half of these additional witnesses claiming they did not appear on IBM's lists of witnesses even though their names were drawn from IBM's own source log of people from whom IBM was producing documents.

D. IBM's Claim That It May Withhold Most Versions Of The Source Code At Issue Has Not Been Upheld And Is The Subject Of A Now Pending Review Process Expressly Established By The Magistrate Judge

IBM contends that SCO has engaged in "misconduct" because SCO has taken the position that it needs access to further versions of AIX and Dynix in order to respond to portions of IBM discovery. But it is remarkable for IBM to ask this Court to assume that SCO has committed "malfeasance" simply because it has disagreed with IBM, its adversary, in this litigation. Certainly the Magistrate Judge has expressly said that this is not the case. Rather than validate all IBM's refusals to provide further versions of AIX and Dynix in the Court's March 3, 2004 Order (as IBM contends), the Magistrate Judge has done the very opposite: the March 3, 2004 Order expressly establishes a now ongoing procedure to give SCO an opportunity to demonstrate its need for the remaining versions of AIX and Dynix. Specifically, in addition to ordering IBM to finally produce at least some limited versions of the source code (so that SCO by late March 2004 at least had more than zero versions), the Order sets up a procedure pursuant to which "SCO is to provide additional memoranda" explaining its need for the versions IBM refuses to provide at all. The "Court will then consider ordering IBM to produce more code from AIX and Dynix." (See March 3, 2004 Order, at 4).(13) If IBM were correct that SCO was engaging in "misconduct" for disagreeing with IBM, then this entire procedure established by the Magistrate Judge, of course, would make no sense.

II. IBM'S DESCRIPTIONS OF SCO'S DISCOVERY CONDUCT ARE DEMONSTRABLY INACCURATE

A. IBM's False Claim That SCO "Still" Will Not Identify The Code That SCO Claims That IBM Misused In Violation of its License

According to IBM, SCO has engaged in "discovery misconduct" and "malfeasance" because, "in disregard of two court Orders, SCO still refuses to identify the specific code from UNIX System V that IBM is claimed to have misused...." (See IBM Br. at 2-3, 11). IBM's invective is inaccurate. SCO has in fact provided voluminous materials to IBM showing the precise restricted code that IBM took from AIX and Dynix and contributed to Linux in violation of its contractual promises.(14)

IBM's claim that SCO has not identified code IBM misused is nothing more than a request that the Court simply assume that IBM is right about a central, contested merits issue in the case. IBM entered into a contract, i.e., a license, which required it to maintain "confidential" and not to transfer "in whole or in part" any UNIX source code it received under the license, as well as any "modification" or "derivative works based on" that software. (See, e.g., Software Agreement, ¶¶ 2.01, 7.10). Since IBM's versions of UNIX, which are known as AIX and Dynix, are - by definition - "modifications" and/or "derivative" works based on UNIX System V, the code contained in either AIX or Dynix must be maintained "confidential" and "may not be disposed of, in whole or in part." (Id. at ¶¶ 2.01, 7.10). Therefore, IBM's public statements that it "dumped" substantial amounts of code from AIX and Dynix into Linux, which is available for free download on the internet, confirm a clear, material breach of IBM's license. SCO must prove that IBM contributed code from AIX and Dynix to Linux to establish liability.

IBM has a different contract interpretation. According to IBM, even though both AIX and Dynix were developed from UNIX System V and include thousands of lines of original UNIX System V code, unless SCO can show that the specific lines of code that IBM contributed to Linux were copied from UNIX System V, SCO cannot establish that IBM breached its contract. IBM's position ignores the fact that it has license agreements with SCO - and the further (also basic) fact that the express terms of such agreements place the owner of the licensed intellectual property in a better position than it would have been in without the agreement. For example, SCO could have sued under the copyright laws to prove literal copying without the agreement, yet IBM says that is all that SCO can do even with the agreement (in fact, on IBM's view of the need to show literal copying, SCO has even fewer rights with the agreement than without it).(15)

Many of the discovery disputes in this case have arisen from this fundamental disagreement. SCO needs to inspect the versions of AIX and Dynix, (including logs) that have existed in any form during the period commencing in 1985 and continuing to the present to support the preparation of its jury case, to support alternative theories of liability, to rebut IBM's argument, and - in the first instance - to respond to IBM's discovery requests.(16)

IBM claims that SCO has no such need, that SCO could respond to all discovery without this additional information, and that SCO is guilty of "malfeasance" for disagreeing with IBM. IBM does not disclose that the Magistrate Judge has already set up a specific process to consider SCO's "good faith" efforts to comply with all discovery without such information and SCO's explanation of its need for that information. That process is now ongoing and may well itself not be completed - given IBM's own resistance to providing discovery - until sometime later this summer, which approaches the current August 4, 2004 discovery deadline IBM seeks to maintain. IBM's opposition to SCO's motion to amend the scheduling order asks the Court to preempt and render meaningless the Magistrate Judge's specific and detailed procedure for deciding this central issue.

IBM's repeated claims that SCO has engaged in discovery "misconduct" are directly contradicted both by the Magistrate Judge's recent "good faith" finding and the record that details SCO's extensive discovery productions in this case. IBM has not hesitated to attempt to impose burdens of the most extraordinary type - for example - two recent discovery requests whose responses are being served on IBM today (interrogatory nos. 15 and 16), would by their express terms require 20 person years of effort to prepare a response.(17) But disagreement on what is appropriate in discovery or providing information in a substantially less burdensome fashion without adversely affecting legitimate discovery needs is not "discovery misconduct." Nevertheless, as set forth in detail in the Notice of Compliance and the Declaration of Ryan Tibbitts filed with the Court dated January 12, 2004, and SCO's April 19, 2004 supplement and Declaration of Chris Sontag dated April 19, 2004, SCO has exhaustively complied, and continues to comply, with its discovery obligations in this case. Indeed, in responding to IBM's numerous discovery requests from August 4, 2003, to date, SCO has produced 241 CDs containing approximately 1.5 million pages of documents, and has produced all of the source code for its products in its possession, which exceeds over 700 million lines of code demanded by IBM.

B. IBM's Claims About SCO's Trade Secret Count

IBM repeatedly emphasizes SCO's long-since voluntarily dismissed claim for trade secrets.(18) On page 4 of its Memorandum, IBM repeats its remarkable charge that SCO somehow acted improperly in bringing a trade secret claim that it later dropped. SCO's conduct was not improper. Rather, it is a commonplace occurrence in litigation; claims are dropped in favor of stronger claims. Indeed, IBM itself has done so when it dismissed its own patent counterclaim in this case eight months after IBM filed that claim. Of course, by that time, SCO had undertaken considerable effort to obtain publicly available information to defend itself by showing IBM's inequitable conduct in obtaining the patent in question.

IBM inaccurately points to the Magistrate Judge's imposition of a "stay" as some type of punitive measure or sanction against SCO. (See IBM Br. at 6). Again, the record contradicts this assertion. As previously noted, the parties had a dispute concerning which party should provide discovery first, given claims by each party of need from discovery from the other as a precondition to proceeding. SCO maintained it needed IBM's AIX and Dynix code prior to providing its analysis of that code for the reasons described above. Faced with deciding who should proceed first, the Magistrate Judge required SCO to respond and "postponed all other discovery," so that SCO could produce the discovery it was capable of producing.(19) (See December 5, 2003 Hearing Transcript at 4). As IBM also fails to disclose, the Magistrate Judge also gave SCO the opportunity to file an Affidavit detailing why it could not provide discovery if (and to the extent) it was unable to absent further information from IBM. Nothing in the record supports IBM's claim that the stay of discovery was a punitive measure or sanction against SCO.

In sum, the record, including the Magistrate Judge's finding of good faith and the now ongoing Magistrate-established-procedure, make clear that SCO has acted with diligence. IBM's inaccurate invective should not deprive SCO of its right to engage in effective discovery on its claims.

III. IBM DOES NOT DISPUTE THAT ITS NUMEROUS COUNTERCLAIMS WERE NOT CONTEMPLATED BY THE SCHEDULING ORDER AND ADDED A SUBSTANTIAL BURDEN

It is SCO's position that the addition of 14 counterclaims to a schedule that could not have been designed to accommodate them created great burdens that ultimately contributed to SCO's need for an amendment of the scheduling order. IBM does not dispute that the original scheduling order was based solely on the complaint, and that it was designed without the benefit of any of IBM's counterclaims - in fact, without even any discussion of such claims. Nor does IBM dispute that these claims entailed very substantial incremental work.

Instead, IBM relies almost entirely on the argument that the impact of its 14 counterclaims should not count because "almost all of its counterclaims" have been pending for approximately nine months. (See IBM Br. at 6). This observation misses the point, which is that the original scheduling order -- which concededly did not and could not have anticipated those 14 counterclaims - already contemplated the allocation of that same nine month period to other purposes. A single nine-month period cannot be counted twice. For example, while SCO has moved to sever the complex patent counterclaims (and plainly needs to detach them from this core contract action), the fact that these broad patent claims have been pending during this nine-month period - which was designed to be devoted to other purposes - has unavoidably contributed to SCO's present need for more time.

IBM also uses its Opposition to focus on its admittedly "recently-filed" declaratory judgment copyright counterclaim. That counterclaim seeks a declaration that none of any of IBM's "Linux-related" activities or uses infringes any SCO copyright. (See IBM Br. at 12). IBM contends that this counterclaim does not add broad, new and unmanageable issues to the present case.(20) This is in fact one of the central issues presented by SCO's motion to dismiss or stay IBM's Tenth Counterclaim, which IBM added to the case on March 29, 2004 in filing its Second Amended Counterclaims. Like the discovery issue on which the Magistrate Judge has directed a now pending adjudication, the issue IBM presents here is the subject of a distinct submission and that separate issue will therefore be fully briefed for the Court.

IV. SCO HAS DEMONSTRATED GOOD CAUSE TO AMEND THE SCHEDULING ORDER

At the time the June 20, 2003 scheduling order was entered, this case consisted of affirmative claims by SCO and no counterclaims by IBM. As such, the scheduling order contemplated that the bulk of document discovery would be completed by early Fall 2003 and that the parties would have more than eight months to conduct iterative discovery and depositions of fact witnesses. The amendment SCO seeks will result in a similar schedule, despite the fact that since the scheduling order was entered, IBM has asserted 14 separate counterclaims.

Assuming the bulk of document discovery is completed by the Fall of 2004, SCO's proposed amended schedule again will provide parties with eight months to conduct iterative discovery and depositions. Indeed, SCO's proposed amendment sets forth a deposition schedule that assumes an average of 10 depositions per month, for eight months, throughout the United States and potentially abroad. Many of these depositions will involve highly-technical subject matters. SCO's proposed schedule is thus not a relaxed schedule that involves unnecessary time - it is plainly itself a highly aggressive schedule. By comparison, the current schedule of 78 depositions in 45 business days is simply impracticable, if not impossible. Such a schedule makes no sense (and, to SCO's knowledge, is unheard of) outside of contexts such as hostile tender offer litigation.

Besides the impossible deposition schedule, another illustrative hurdle facing SCO is that, contrary to IBM's recent supplemental filings, SCO still has not received responses to information requested in June 2003. For example, SCO still has not received the information it requested regarding the identification of the precise contributions of each of the hundreds or thousands of persons involved in the creation of AIX. IBM's April 19, 2004 supplement makes the incorrect statement that the precise contributions of the individuals identified can be ascertained "from the products themselves." This is not true regarding AIX. For AIX, the source code does not contain this information; rather, it is maintained by IBM on a central server that identifies all modifications made to the AIX code as well as who made the modifications and when they were made. SCO has repeatedly requested this information be produced and IBM has refused.

SCO also must contend with a variety of other complicating factors. For example, SCO needs other (historical) versions of IBM programs to demonstrate alternative bases for showing derivation from UNIX System V. IBM has claimed that SCO must make such showings, but IBM also refuses to provide SCO with the discovery needed to make such showings. As noted, the Magistrate Judge has provided SCO with the opportunity to demonstrate the nature of its need for the information IBM refuses to provide. SCO has now provided a detailed memorandum demonstrating this need. SCO did so after making every effort to do without such other versions of the IBM programs, which, as described in these submissions, SCO was ultimately not able to do until it received at least some versions of the code in March.(21)

There are, of course, other hurdles and impediments SCO has incurred and would face under IBM's proffered schedules. But SCO has established specific, concrete bases that are independently sufficient to show why it needs the additional time requested. Given those concrete and detailed showings of legitimate need and prior diligence by SCO, IBM's pervasive reliance on claims about SCO's "ulterior" motive in seeking this amendment is plainly inadequate as a matter of legal argument and is otherwise improper.

CONCLUSION

For all of the foregoing reasons, SCO respectfully requests that the Court amend the Scheduling Order as detailed in SCO's initial memorandum.

[signature of Brent O. Hatch]
Brent 0. Hatch (5715)
HATCH, JAMES & DODGE
[address, phone]


Robert Silver
BOIES, SCHILLER & FLEXNER LLP
[address, phone]

Stephen N. Zack
Mark J. Heise
BOIES, SCHILLER & FLEXNER LLP
[address, phone]
Attorneys for Plaintiff

CERTIFICATE OF SERVICE

I hereby certify that I caused a true and correct copy of the foregoing PLAINTIFF SCO'S REPLY IN MEMORANDUM IN SUPPORT OF ITS MOTION TO AMEND THE SCHEDULING ORDER be placed in the United States Mail first class postage prepaid on the 28th day of May, 2004 to the following:

Donald J. Rosenberg, Esq.
[address]


Evan Chesler, Esq.
David R. Marriott, Esq.
Cravath, Swaine & Moore LLP
[address]

Alan L. Sullivan, Esq.
Todd M. Shaughnessy, Esq.
Snell & Wilmer L.L.P
[address]

1. AIX is a modification or derivative work based on UNIX System V that IBM licensed from SCO's predecessor.

   
   

2. SCO's task of correcting IBM's pervasive deployment of invective, and showing how the record contradicts such claims, unavoidably requires more space than would have been needed if the focus could have been maintained on the parties' respective views as to the practical issues raised by the present motion.

   
   

3. Where SCO cannot cite to specific documents in the court files, SCO relies on correspondence between counsel which is in IBM's possession. SCO will have this correspondence available at argument should the Court desire to inspect it.

   
   

4. As a general matter, the picture IBM presents bears little relation to what is actually happening in discovery: IBM is over 1000 times the size of SCO, yet it has produced far less information and documents than SCO. IBM has also had six months to analyze SCO's production while SCO has had less than half that time to analyze IBM's production. Is it any wonder that IBM is resisting a motion that will provide SCO with an adequate opportunity to prosecute its case and defend against IBM's counterclaims?

   
   

5. IBM similarly mischaracterizes the status of discovery generally. For example, although it firmly argues that its three patent claims (which SCO has moved to sever) should remain part of the case, it represents to the Court that all document discovery in the entire case is essentially complete and that depositions will require no more than two months to complete. In fact, as to IBM's patent claims, IBM has not produced a single document. Indeed, SCO's response to IBM's document request is not due until mid June of this year.

   
   

6. IBM Br. at 13.

   
   

7. IBM claims that "SCO does not need any AIX and Dynix source code for its analysis" because SCO has stated that Sun Microsystems ("Sun") and Hewlett Packard ("HP") have not violated their licenses by contributing their versions of UNIX to Linux without access to Sun or HP's modifications or derivative works of UNIX System V (See IBM Br. at 14-15). This excuse for not producing AIX and Dynix is incorrect. Unlike IBM, HP, for example, has never publicly proclaimed it has contributed its version of UNIX to Linux and, in fact, HP has represented specifically that it has not done so.

   
   

8. See SCO's First Request for the Production of Documents and First Set of Interrogatories, dated June 24, 2003 (Document Requests No. 2 and 3).

   
   

9. See IBM's Responses and Objections to SCO's First Request for Production of Documents and First Set of Interrogatories, dated August 13, 2003.

   
   

10. As further discussed in Section II below, the Magistrate Judge did not indicate that she was staying discovery as a sanction against SCO. Rather, she "postpone[d]" additional discovery by either party until SCO could complete the discovery of which it was capable before considering SCO's motions to compel. (See December 5, 2003 Hearing Transcript at 4). When the Magistrate Judge's March 3, 2004 Order not only lifted the stay generally, but also added a specific, numbered directive granting relief requested in SCO's motion to compel, it also ordered IBM to produce the central source code which it had failed to produce prior to that date.

    
    

11. This document, and others like it, were produced to SCO only after Magistrate Judge Wells ordered its production. Thus, its omission from IBM's production could not have been the result of a clerical error.

    
    

12. IBM Br. at 13.

    
    

13. SCO has worked hard, and expended substantial time and resources, in an effort to find ways of answering certain of IBM's discovery requests without access to the non-public information IBM has long declined to provide. But SCO has been unable to completely do so. SCO has now submitted (along with other discovery responses) the memorandum contemplated by the March 3, 2004 Order explaining its need for the requested information.

    
    

14. SCO has provided dozens of detailed examples of the code IBM copied from AIX and Dynix into Linux and discusses the significance of these contributions. When IBM provided code in March, SCO identified further instances of IBM's contributions of AIX and Dynix to Linux. (See, e.g., Exhibits B and C to SCO's April 19, 2004 discovery responses).

    To provide this detailed information and otherwise comply with its discovery obligations, on August 4, 2003, SCO produced approximately 50 CDs containing source code for System V and UnixWare, identified dozens of SCO products, set forth its product development history by providing beta projects and source code and detailed its involvement in Linux in response to IBM's First Set of Discovery. On October 23, 2003, SCO supplemented its response to IBM's First Set of Discovery by producing approximately 54 CDs containing additional source code, as well as source code files and headings identified within the Linux 2.4 and/or Linux 2.5 kernel. SCO also then responded to IBM's Second Set of Discovery by producing various agreements, communications and documents, including an additional 54 source code CDs for various SCO products.

    On January 12, 2004, SCO produced approximately 82 CDs containing additional source code for various products, executive files, press and media documents, Linux sales invoices and SEC financials. In addition, SCO identified specific lines of source code that IBM contributed to Linux. SCO's ability to identify additional lines of code was hampered by IBM's refusal to provide AIX source code. When IBM finally produced additional versions of AIX and Dynix code pursuant to the Court's March 3, 2004 Order, SCO identified additional examples of improper copying of AIX and Dynix into Linux in SCO's April 19, 2004 response.

    
    

15. To first get the benefit of the agreement for itself and then place SCO in a position of effectively having no agreement, IBM is (not surprisingly) forced to take extreme and implausible positions. For example, even though the agreement expressly states that IBM may develop "modifications" or "derivative works based on" the licensed UNIX System V, and even though IBM told the world that AIX was such a modification, IBM now refuses even to admit that the very same source code is a derivative or modification of UNIX. Similarly, the protected subject matter of the agreement is defined as the unitary "software product," and while the agreement allows IBM to create modifications and derivative works based on that "software product," it requires IBM to restrict its uses of those modifications and derivatives in all of the ways that uses of the original "software product" would be restricted.

    IBM will therefore, for example, be forced to argue that its source code, AIX, which it always treated as a single "software product" and sold to customers as a single "software product," is in fact not a single "software product" after all, but instead an aggregation of countless stand alone "sub-products," all somehow "tied" together in ways IBM has not previously explained to its customers.

    
    

16. This discovery would serve several legitimate and indeed critical purposes. It would allow SCO to show additional transfers of AIX and Dynix to Linux. It is essential to allowing SCO to show the derivation of Linux from System V through the evolution of intermediate versions of AIX and Dynix. Each of these are ways to win under the contract, and SCO will establish other ways. SCO is entitled to the proof it needs to proceed under these (and other) theories. Even more immediately, SCO requires access to the requested information simply in order to respond to IBM's discovery demands - especially in the fact of IBM claims that SCO has engaged in "misconduct" because it has not yet responded to portions of those discovery demands.

    
    

17. IBM's multipart Interrogatory Nos. 15 and 16 would require SCO to provide the genealogy - over a 20 year period - for each of over 240,000 lines of source code on a line-by line basis. At 10 minutes per line, it would take someone over 20 years to respond to these two interrogatories.

    
    

18. As a purely factual matter, IBM's interrogatories, for example, were not confined to trade secrets; rather, the questions asked for "trade secrets and any confidential or proprietary information," which also would encompass items IBM was required to maintain in confidence pursuant to the license agreement and which SCO has identified.

    
    

19. Specifically, Magistrate Judge Wells began the hearing stating that it was her intention to require SCO to respond and "otherwise postpone all other discovery until such filings have been [made] and compliance has been achieved." (See December 5, 2003 Hearing Transcript at 4). The March 3, 2004 Order then reflects Magistrate Judge Well's conclusion that SCO had acted in "good faith" in seeking to comply despite the problems SCO had initially raised in December 2003 as impediments to compliance - the problems created by IBM's own discovery conduct and decisions. Thus, Magistrate Judge Wells -- in the same March 3, 2004 Order lifting the stay that "postpone[d]" discovery - also included a specific, numbered directive requiring IBM to finally provide at least some versions of its AIX program and established a procedure allowing SCO to show why it needed other versions (one of the issues SCO had raised in December 2003).

    
    

20. It is just common sense that IBM's claim would add new, broad and unmanageable issues to the case. IBM's claim by its terms implicates every single use of Linux. IBM is an end-user of Linux, as are a great many (perhaps countless) other entities. IBM's declaratory counterclaim would therefore - among other things - add the broad set of issues common to that very large group. SCO has not sued IBM in its capacity as an end-user of Linux. It has sued IBM because of facts specific to IBM, not because of facts that IBM shares in common with countless other entities.

    
    

21. IBM's complaints about SCO thus are self-contradictory. IBM says that SCO could have done without these other versions and, for example, could have responded to IBM discovery without these other versions. Indeed, IBM contends that SCO was effectively in contempt for disagreeing with IBM. But IBM also did not provide SCO with any versions until March, 2004, and SCO could not try to make the detailed and specific showing of its need for these other versions - which it has now submitted - until it received at least one version of AIX from IBM.