Frank Sorenson has taken the Harrop Declaration, striking out all the paragraphs IBM is asking to have struck from the record. What is left isn't much. And more significantly, there is nothing left that would by any stretch be considered a fact in dispute such as could defeat IBM's motion for partial summary judment.
Our thanks to JeR for transcribing the Harrop Declaration.
Brent O. Hatch (5715)
HATCH, JAMES & DODGE, PC
Stephen N. Zack (admitted pro hac vice)
Mark J. Heise (admitted pro hac vice)
BOIES, SCHILLER & FLEXNER LLP
Robert Silver, Esq. (admitted pro hac vice)
BOIES, SCHILLER & FLEXNER LLP
Frederick S. Frei (admitted pro hac vice)
Aldo Noto (admitted pro hac vice)
John K. Harrop (admitted pro hac vice)
ANDREWS KURTH LLP
Attorneys for Plaintiff The SCO Group, Inc.
IN THE UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF UTAH
THE SCO GROUP, INC.
INTERNATIONAL BUSINESS MACHINES CORPORATION
DECLARATION OF JOHN HARROP IN SUPPORT OF SCO’S OPPOSITION
TO IBM’S MOTION FOR PARTIAL SUMMARY JUDGMENT
Case No. 2:03-CV-0294 DAK
Honorable Dale A. Kimball
Magistrate Judge Brooke C. Wells
DECLARATION OF JOHN HARROP
1. I am a member of the law firm Andrews Kurth LLP, co-counsel for
Plaintiff The SCO Group, Inc. (“SCO”) in the
above-captioned action. Unless otherwise indicated, I make this
Declaration based upon personal knowledge.
2. I submit this Declaration in support of SCO’s Memorandum in
Opposition to Defendant/Counterclaim-Plaintiff IBM’s
Cross-Motion for Partial Summary Judgment on Its Claim for
Declaratory Judgment of Non-Infringement, dated May 18, 2004.
3. I explain below why, if IBM’s claim for declaratory
judgment of non-infringement (the “Tenth Counterclaim”)
remains in the case, SCO needs substantial discovery on the Tenth
Counterclaim in order to discover facts essential to justify
opposition to IBM’s Motion. The exhibits I refer to herein are
in the appendices submitted with SCO’s Opposition Memorandum.
4. In 1985, SCO’s predecessor-in-interest AT&T entered
into license agreements with IBM and Sequent (together,
“IBM”) that permitted IBM to access and use the
constituent source code of the UNIX computer operating system in
order to develop IBM’s own versions of the UNIX operating
5. The principal basis for SCO’s claims in this matter has
always been that the license agreements unambiguously governed
IBM’s right to use and transfer the source code in the
operating systems, called AIX, and Dynix, ptx, and Dynix/ptx, which
IBM developed with access to UNIX.
6. Until February 2004, SCO had not asserted any claim for copyright
infringement in this matter. Rather, SCO had asserted
breach-of-contract and tort claims arising out of IBM’s use
and transfer of UNIX material, including by contributing source code
from AIX, Dynix, ptx and Dynix/ptx into another computer operating
system, called Linux, developed in the 1990s.
7. SCO’s position has always been that in order to prove that
IBM breached the unambiguous license agreements, SCO need not prove
that any source code from any version of UNIX has been copied into
any version of Linux, and need not prove that IBM has violated any
SCO copyright. For example, because the license agreements do not
permit IBM to lease or transfer any parts of the AIX, Dynix, ptx,
and Dynix/ptx programs, IBM has violated the agreements by making
source code in AIX, Dynix, ptx, and Dynix/ptx publicly available.
8. SCO first alleged a claim for copyright infringement in its
Second Amended Complaint, filed on February 27, 2004. The primary
basis for that copyright claim is that IBM has continued to use and
distribute AIX, Dynix, ptx, and Dynix/ptx after SCO terminated the
9. SCO has not brought any claim that IBM contributed source code to
Linux in violation of any SCO copyright. SCO has not asserted here
any claim that any third party has contributed any source code to
Linux in violation of any SCO copyright. SCO has not brought any
copyright claim against IBM in this action in regard to any of
IBM’s numerous activities relating to Linux.
10. On March 29, 2004, however, IBM filed its Second Amended
Counterclaims and concluded in the Tenth Counterclaim: “IBM is
entitled to a declaratory judgment pursuant to 28 U.S.C. § 2201
that IBM does not infringe, induce the infringement of, or
contribute to the infringement of any SCO copyright through its
Linux activities, including its use, reproduction and improvement of
Linux, and that some or all of SCO’s purported copyrights in
UNIX are invalid and unenforceable.”
11. On April 23, 2004, SCO filed a Motion to Dismiss or Stay the
Tenth Counterclaim. SCO explained that the Tenth Counterclaim
“raises issues separate and apart from the primary breach of
contract and other direct claims and counterclaims in this
case.” (SCO Motion to Dismiss or Stay the Tenth
Counterclaim ¶ 2.) SCO explained that its Second Amended
Complaint “does not contain a claim against IBM for copyright
infringement arising out of its use, reproduction or improvement of
Linux.” (SCO Opening Mem. ¶ 3.)
12. SCO also asserted in support of its Motion to Dismiss or Stay
the Tenth Counterclaim that factual and legal issues raised in the
Tenth Counterclaim are already at issue in a pending federal action
that SCO brought before IBM filed the Tenth Counterclaim. In that
action (the “AutoZone action”), SCO had brought claims
under the Copyright Act on the basis of AutoZone’s use of
Linux as its operating system (that is, as an “end-user”
13. On May 18, 2004, IBM filed both its opposition to SCO’s
Motion to Dismiss or Stay the Tenth Counterclaim and its instant
motion for summary judgment.
The Nature and Timing of IBM’s Tenth
14. On June 9, 2004, further to SCO’s pending Motion to Stay
or Dismiss the Tenth Counterclaim and to IBM’s concession
during a hearing before the Court on June 8, 2004 that the Tenth
Counterclaim is a permissive counterclaim under Fed. R. Civ. P. 13,
IBM submitted a letter to the Court in which IBM withdrew its June 8
concession and argued that the Tenth Counterclaim is a compulsory
counterclaim. IBM made the same argument in its Supplemental
Memorandum in Opposition to SCO’s Motion to Dismiss or Stay
Count Ten, dated June 25, 2004.
15. I believe it is undisputed that the significance of the
distinction is that the Court may dismiss a permissive counterclaim
that would unduly complicate the litigation. Dismissal is the relief
SCO seeks in its pending Motion to Dismiss or Stay the Tenth
16. IBM’s Tenth Counterclaim is permissive because, as SCO has
asserted, it is not one that “arises out of the same
transaction or occurrence that is the subject matter of the opposing
party’s claim.” Fed. R. Civ. P. 13(a). SCO will address
the issue further in its Reply Memorandum in support of SCO’s
Motion to Dismiss or Stay the Tenth Counterclaim, which is due July
17. The inappropriateness of IBM’s Tenth Counterclaim and the
fact that SCO’s Motion to Dismiss or Stay the Tenth
Counterclaim is still pending support SCO’s opposition to
18. IBM argues that the question of the source code in Linux to
which SCO claims any “rights” was a subject of discovery
before IBM filed its Tenth Counterclaim and that for purposes of
opposing IBM’s Motion, SCO may not cite facts that were not
included in discovery responses SCO made in January and April 2004.
19. By bringing the claims it did, SCO has specifically avoided the
need for the broad and time-consuming discovery necessary to
determine (by way of example) the full scope of IBM’s numerous
activities relating to Linux, the source code thousands of third
parties had contributed to Linux, the origins of the source code
they contributed, the ways in and extent to which thousands of
end-users use Linux, and all of the other discovery necessary to
give SCO an opportunity to discover facts essential to oppose
IBM’s Tenth Counterclaim. (See paragraphs 41-75 below.)
20. The necessary and foreseeable effect of IBM’s Tenth
Counterclaim would be to introduce the substantial issues discussed
in paragraph 19 above into the discovery process. The discovery that
the Tenth Counterclaim foreseeably entails is inconsistent with
IBM’s repeated assertion that it is SCO who seeks
“delay” in this litigation.
21. In addition, any suggestion that SCO could have obtained the
facts essential to justify opposition to IBM’s Motion based on
the discovery to date is particularly unreasonable given that IBM
has not adequately or fully responded to discovery SCO propounded
over a year ago. (See paragraphs 76-90, below.)
22. Further, SCO has not purported to have identified in discovery,
nor has it certified that it has identified, all of the source code
in Linux to which SCO claims any “rights.”
Indeed, at the time that IBM propounded its discovery requests, the
question of the copyrights SCO has in source code in Linux was not
at issue in the litigation. At that time, there was no copyright
claim in the case at all; SCO had not even brought its narrow
23. It is true (as IBM asserts) that the Magistrate Judge ordered
SCO to respond to certain requests for discovery to which SCO had
objected, but it is not true (as IBM suggests) that the Magistrate
Judge ordered SCO to produce -- let alone on pain of entry of
summary judgment on a claim not yet brought -- anything more than
what SCO had reasonably been able to uncover at the time the
discovery responses were due.
The Scope of IBM’s Tenth Counterclaim
24. IBM further argues that, limited to its discovery responses in
January and April 2004, there are no facts SCO can present to create
any genuine issue of material fact and IBM therefore is entitled to
the declaration it seeks in its Tenth Counterclaim. SCO argues in
its opposition memorandum that IBM’s motion is nothing more
than a request for sanctions. Even if the Court finds otherwise,
IBM’s argument begs the question of the full scope of its
“activities relating to Linux,” which is an issue on
which SCO has taken little discovery.
25. In his declaration submitted with IBM’s motion for summary
judgment, Daniel Frye (the co-founder and present director of
IBM’s Linux Technology Center) enumerates several of
IBM’s activities relating to Linux:
-- IBM has contributed source code to Linux (Frye Decl. ¶ 5);
-- “IBM offers Linux training and support, applications
testing, technical advice and a hands-on environment in which to
evaluate Linux and Linux-based applications” (Id.
-- “IBM has many Linux-compatible offerings: mainframes and
servers that run Linux; memory solutions for Linux environments; a
broad range of Linux-compatible software offerings; services that
assist companies in deployment of Linux-based computing
environments, migration of database applications and data to Linux
systems, support for Linux-based cluster computing, server
consolidation, and a 24-hour technical engineering support
line.” (Id. ¶ 7);
-- “In connection with its Linux activities, IBM reproduces
Linux and makes Linux available to others, both in developing and
providing hardware, software and services for customers, and for
other, internal business purposes” (Id. ¶ 8); and
-- “many IBM employees -- particularly those who work in the
IBM Linux Technology Center -- use Linux as their platform for
day-to-day business computing, running office productivity
applications, developing software (including Linux itself), and
exchanging e-mail” (Id. ¶ 9).
26. SCO submits that certain facts are clear from the foregoing
description of (at least some of) IBM’s activities relating to
-- Mr. Frye makes no representation that he has identified
all of the activities relating to Linux in which IBM engages;
With respect to those activities he has enumerated, Mr. Frye does
not describe any of them in any significant level of detail;
respect to several of those activities Mr. Frye has enumerated, IBM
may be violating SCO’s copyrights in UNIX material whether or
not that material is in Linux.
SCO cannot obtain information on the
foregoing points without beginning and taking discovery with respect
to IBM’s Tenth Counterclaim (including, as just one example,
the deposition of Mr. Frye).
27. Discovery regarding all of IBM’s Linux-related activities
would be substantial and time consuming. There is no question that
IBM’s sweeping Tenth Counterclaim would necessarily pull into
this litigation those issues. That is one reason SCO believes the
Court should not exercise jurisdiction over the Tenth Counterclaim.
If the Court does exercise jurisdiction, however, SCO is entitled to
explore such subject matter to discover facts essential to oppose
the Tenth Counterclaim.
Satisfaction of Rule 56(f)
28. My understanding is, at least in cases where the parties have
taken discovery on the claim that is the subject of a summary
judgment motion, under Fed. R. Civ. P. 56(f), the non-moving party
must identify the probable facts not available and what steps have
been taken to obtain these facts, and must demonstrate how
additional time will enable him to rebut the moving party’s
allegations of no genuine issue of fact.
29. To justify opposition to IBM’s Cross Motion -- that is, to
contend that IBM is not entitled to a declaration of
non-infringement with respect to all of its activities relating to
Linux -- SCO must first be able to identify all of IBM’s
activities relating to Linux. Assuming a scenario in which IBM will
argue for an entitlement to a declaration of non-infringement with
respect to as many activities relating to Linux as possible. SCO
must have discovery to identify IBM’s activities relating to
Linux as a threshold matter.
30. As I explain below, if IBM’s Tenth Counterclaim remains in
the case, SCO should be entitled to begin and take substantial
discovery on the Counterclaim, pursuant to Rule 56(f). Where I lack
personal knowledge of the subject matter at issue, I refer the Court
to the Declaration of Christopher Sontag (“Sontag
Decl.”), Senior Vice President and General Manager of SCO,
submitted with SCO’s Motion for Continuance Pursuant to Rule
56(f), and the Declaration of Sandeep Gupta (“Gupta
Decl.”), Director of Engineering, The SCO Group, Inc., Murray
Hill, New Jersey, submitted with SCO’s Opposition Memorandum.
The Nature of the Uncompleted Discovery
31. SCO intends to seek evidence of facts essential to justify
opposition to IBM’s Tenth Counterclaim in several ways.
Evidence of line-for-line duplication of UNIX source code in Linux
source code is only part of the relevant evidence necessary for SCO
to oppose IBM’s Tenth Counterclaim.
32. There are inherent obstacles in identifying all line-for-line
similarities between two computer operating systems. Accordingly,
other means of identifying copying between computer operating
systems, and of the modification or derivation of source code from
one operating system into the source code of another operating
system, are crucial tools for discovering relevant evidence.
33. It is public knowledge that in contrast to UNIX, AIX, Dynix, ptx
and Dynix/ptx, the Linux operating system was not developed under
the control of any single entity or corporation. In 1991 a Finish
college student named Linus Torvalds began composing an operating
system. In his classes, Mr. Torvalds had been studying an operating
system that one of his professors (having received an educational
license to do so) based on and derived from UNIX.
34. Mr. Torvalds posted the material about the operating system on
the Internet for comment. The development of the operating system
thereafter became in effect a group project in which Mr. Torvalds
and his delegates made final determinations about which suggestions
from numerous third parties, many of whom are anonymous, to
incorporate. The kernel of the operating system that resulted came
to be known as Linux. According to IBM, IBM and thousands of third
parties have contributed source code to Linux.
Discovery Concerning IBM’s “Activities Relating to
35. For SCO to discover facts essential to its opposition to
IBM’s requested declaration, SCO needs to pursue discovery to
identify and determine the scope of IBM’s “activities
relating to Linux.”
36. SCO indisputably has not taken, and has not had the opportunity
significantly to take, discovery even to identify all of IBM’s
“activities relating to Linux,” let alone the nature and
extent of such activities.
37. IBM’s own public descriptions of the enormous extent of
its worldwide Linux-related activities reveals the broad scope of
discovery the Tenth Counterclaim would entail. For example:
-- As of a year ago, “The company has 250 developers working
on 29 separate Linux projects worldwide, according to Ken King,
director of technical strategy from I.B.M.’s software
group” (“No Concession from IBM In Linux Fight,”
New York Times, June 14, 2003 (Exh. 58)); and
-- On August 4, 2003, IBM issued the following press release:
“IBM Global Services offers the industry’s most
comprehensive portfolio of Linux consultive and support offerings,
from planning and design, to implementation and technical support.
IBM consultants skilled in Linux are available worldwide to help
customers design, build, enhance and operate their Linux
solutions” (Exh. 55).
38. In light of published reports, SCO reasonably believes that it
would be entitled to discovery regarding subject matter such as the
-- IBM is spending billions of dollars each year in an effort to
make Linux the world’s most popular computer operating system.
In 2001, IBM granted $1 billion for the vice president of technology
and strategy at IBM to build a Linux business.
-- In 2003 IBM’s Linux-related revenues grew 50% to more than
$2 billion. IBM’s mainframe hardware business grew 7% to just
over $3 billion; that growth is principally attributable to Linux,
which shipped on 20% of the mainframe support IBM delivered in 2003.
-- IBM is helping at least hundreds of third parties migrate their
computers off of other operating systems and onto the Linux
operating system. Since 2001, IBM has trained at least 3000
employees in Linux in order to launch the practice to help customers
migrate to Linux.
-- More that 12,000 IBM employees currently devote at least part of
their time to one or more activities relating to Linux. IBM uses
Linux in its own data centers. Linux now runs on more than 3400
servers inside IBM.
-- IBM is using Linux to help IBM win business. For example, IBM
generates substantial consulting fees for installing and customizing
Linux-based hardware and software for clients. IBM has been helping
companies transfer their software applications to run on a Linux
-- IBM has created 45 Linux technology centers in 12 countries,
where experienced engineers with backgrounds designing IBM’s
own operating systems, including AIX, contribute code to Linux.
(See “Kill Bill,” by Daniel Lyons, at www.
forbes.com (Exh. 52).)
39. Other published reports also indicate that IBM’s
Linux-related activities have grown and expanded and continue to do
so -- which, if true, would make it even more unreasonable to grant
IBM’s request for a declaration regarding all such
Linux-related activities even more inconceivable. For example:
-- IBM earned $1 billion in Linux-based revenues in 2002 -- more
than double its revenues for 2001 (“The Big Guys Latch Onto
Linux,” Business Week, March 3, 2003 (Exh. 38));
-- IBM believes that Linux is maturing into a software standard that
could be adopted more widely. IBM endorses analyses expecting the
Linux market to grow from $2 billion to more than $5 billion in 2006
(“IBM Clinches Security Certification for Linux,”
Forbes, Aug. 5, 2003) (Exh. 44));
-- IBM issued a press release on January 19, 2004 stating:
“IBM today announced new programs and supporting classes to
help Business Partners and customers move from the legacy Microsoft
Windows NT operating system to Linux, the fastest growing server
operating system in the word” (Exh. 45);
-- IBM stated in a press release issued on March 16, 2004:
“IBM today announced new partners, programs and incentives
that are helping to fuel Linux adoption and growth among
small-to-medium-sized businesses” (Exh. 49);
-- On June 6, 2004, the vice president of technology and strategy at
IBM was quoted as saying: “Linux is helping us win big
business,” and “If you become convinced that something
is going to happen whether you like it or not, you are far better
off embracing it” (“Kill Bill,” by Daniel Lyons,
at www. forbes.com/forbes/2004/0607/086_print.html (Exh. 52).)
40. There is little question that discovery regarding subject
matters such as the foregoing would be substantial and time
consuming. There is no legitimate question, however, that
IBM’s Tenth Counterclaim necessarily would pull into this
litigation such subject matter.
Discovery to Determine Contributions to Linux
41. As a result of how Linux evolved, there is no “road
map” that will allow SCO to trace the migration of UNIX code
into Linux. (Sontag Decl. ¶ 57.) One principal way for SCO to
discover at least some of the facts essential to oppose IBM’s
Tenth Counterclaim is to take discovery to determine who made
contributions of source code to Linux.
42. Discovery regarding who made contributions of source code to
Linux, and what contributions they made, is relevant to IBM’s
Tenth Counterclaim. SCO would seek to prove (for example) that IBM
had access to SCO’s copyrighted material and used the same or
substantially similar material in AIX, Dynix, ptx, and/or Dynix/ptx,
and/or contributed the same or substantially similar material to
43. To the best of SCO’s knowledge, there is no existing list
of all of the contributors of source code into Linux. To list Linux
contributors, SCO must review the Linux change log. In some cases,
the contributors are identified in the list by either full name,
e-mail address, or a single name. The change log contains incorrect
data, obsolete data, nicknames, and pseudonyms for authors. Further,
the change log is incomplete and does not list many Linux
contributors. SCO has initiated this review of the change log and
created a partial list of Linux contributors. (Sontag Decl. ¶
57). The list further demonstrates the need for additional
-- In order to seek to depose a contributor identified by e-mail,
SCO would have to assume the e-mail address is still current, send
an e-mail to the contributor, and hope that he or she responds, and
does so honestly.
-- SCO could not reasonably be expected to seek to depose a
contributor identified only by full name, as SCO has no idea where
the person lives.
-- Nor could SCO reasonably be expected to seek to depose a
contributor identified only by a single name.
44. SCO does not intend to depose thousands of contributors
worldwide to determine who made material contributions to Linux.
Instead, SCO seeks discovery to pursue the following, reasonable
-- Determine what third parties IBM has partnered with to develop
Linux and what work those parties have done. Many of these
arrangements are not in the public domain, particularly as to the
details of the partnering, such as which party makes what
contribution, the motivation for the contribution, and the starting
and ending code versions that resulted from the partnership. This
discovery will also help SCO identify specific code authors, who can
then be deposed.
-- Take discovery from Mr. Torvalds concerning his knowledge about
the contributors and contributions to Linux since its inception, and
the maintenance of any records about the development history of
Linux. Mr. Torvalds is expected to have detailed records of these
contributors and their contributions, material that is not publicly
-- Take discovery from individuals (some of whose identities SCO
knows) responsible for maintaining the Linux operating system
(so-called “kernel maintainers”). Kernel maintainers
take responsibility for approving and including patches for Linux
and should have a wealth of information on who has contributed what
code to the various Linux kernels over the years.
45. Several private groups also have made major contributions to
Linux, so SCO should also be permitted adequate time to identify and
take discovery from these entities.
Depositions of Contributors to Linux Are Essential
46. SCO seeks to discover facts essential to oppose IBM’s
Tenth Counterclaim by deposing the persons and entities that
contributed source code to any version of Linux. If company had
compiled Linux, for example, SCO should be permitted to depose the
principal employees who compiled the operating system. This is an
especially important form of discovery that will, SCO believes, lead
to the discovery of admissions of copying of (among other things)
source code, structure and sequence, and/or the preparation of a
derivative work by such contributors.
47. SCO has not had the opportunity to depose any of the
contributors of any source code into any version of Linux --
much less the major contributors of source code -- and therefore has
not had any opportunity to discover admissions highly relevant to
IBM’s Tenth Counterclaim. Indeed, SCO has not had the
opportunity to depose even the person (Mr. Torvalds) who is
acknowledged to have compiled the first versions of Linux -- and who
indisputably did so after having studied an operating system I
believe was expressly based on and derived from UNIX.
48. Nor has SCO had the opportunity to depose any of the
kernel maintainers. Mr. Torvalds and the kernel maintainers (and
there have been numerous such individuals since at least the
mid-1990s) are likely to be able to help identify who contributed
source code to Linux.
49. In addition, many corporations have made contributions to Linux,
and SCO needs to take discovery on certain of these companies to
determine the sources of their contributions.
50. SCO also needs to depose programmers who work for these
companies and made the contributions to determine the sources of
those programmers’ code contributions. Such programmer
depositions should enable SCO to streamline and prioritize and make
its investigation of substantially similar copying from UNIX to
Linux more efficient. A streamlined and prioritized investigation of
substantially similar copying will not be nearly as time-consuming
as a systematic line-by-line comparison would be. The programmers
might say, for example, that in some areas of Linux development they
relied on UNIX to a great extent and in other areas they did not.
Such testimony would enable SCO to focus on those areas the
programmers identified as relying on UNIX. Furthermore, this
discovery will show why the contributions were made and what
features the contributions relate to, and will allow SCO to trace
back from the Linux code to UNIX.
51. SCO has identified some authors of various portions of Linux
code from the Linux change log. (Sontag Decl. ¶ 57.) Those
authors should know the sources of their code and should be able to
provide information as to whether the code they contributed to Linux
was obtained from SCO copyrighted code.
52. Depositions of some of the foregoing individuals may permit SCO
to more reasonably to determine which of the numerous contributors
of source code to Linux to depose.
53. SCO needs to depose contributors to Linux and anticipate such
depositions will provide relevant evidence including, for example:
-- If such a person acknowledges that he or she has incorporated (or
induced, encouraged, or enabled others to incorporate) any UNIX
software into any version of Linux, that testimony would support
-- If such a person acknowledges that he or she has incorporated (or
induced, encouraged, or enabled others to incorporate) any software
based on or derived from any UNIX software into any version of
Linux, that testimony would support SCO’s opposition.
-- If such a person testifies that he or she contributed source code
into any version of Linux, and did so after having reviewed or
examined the UNIX software in any detail, that testimony would
support SCO’s opposition.
-- If (as noted above) such a person testifies that he or she knows
of someone who incorporated (or induced, encouraged, or enabled
others to incorporate) any UNIX software, or any software based on
or derived from any UNIX software, into any version of Linux, that
testimony would permit SCO to obtain additional admissible evidence.
54. SCO would also depose, for example, IBM principals regarding the
nature and extent of IBM activities relating to Linux. For example:
-- SCO would depose Mr. Frye, the co-founder and present director of
IBM’s Linux Technology Center, regarding the IBM activities
relating to Linux he enumerates in the declaration submitted with
IBM’s motion for summary judgment, including to determine
whether IBM engaged in activities relating to Linux not identified
in his declaration and what those activities are.
-- SCO would depose Irving Wladawsky Berger, vice-president of
technology and strategy at IBM, who has been described as “a
pivotal proselytizer of Linux inside the company” and is the
person at IBM who has proclaimed that “Linux is helping us win
business.” (“Kill Bill,” by Daniel Lyons, at
www.forbes.com (Exh. 52).)
-- SCO would depose Samuel Palmisano, formerly senior vice president
of IBM and now its Chief Executive Officer, who participated in
IBM’s decision in 2001 to provide Mr. Wladawsky Berger’s
group with $1 billion to build a Linux business. (Id.)
Depositions of Persons with Access to UNIX
55. Another principal way for SCO to discover evidence essential to
oppose IBM’s Tenth Counterclaim is to depose the persons and
entities that had access to UNIX, AIX, Dynix, ptx, and/or Dynix/ptx
software. (That class of persons of course may overlap with the
individuals described in paragraphs 46-54 above.)
56. SCO has not had the opportunity to depose (for example)
any of the persons employed by IBM or Sequent (which IBM
acquired) who had access to the UNIX software, nor any of the
persons at IBM or Sequent who participated in producing AIX, Dynix,
ptx and Dynix/ptx, respectively.
57. The depositions of (at least) the principal IBM and Sequent
employees who were permitted to and did access the UNIX software
prior to the advent of AIX, Dynix, ptx and Dynix/ptx will permit SCO
more reasonably to determine which of the individuals who had access
to UNIX to depose.
58. SCO has not had the opportunity to seek discovery from other
UNIX licensees about their contributions to Linux.
Examination of Multiple Versions of AIX, Dynix, ptx, and
59. To show that Linux code is substantially similar to UNIX code
requires a comparison of that code, which, as described below, is an
undertaking of great magnitude and complexity. SCO can significantly
streamline this effort by examining the lineages of AIX, Dynix, ptx,
and Dynix/ptx. By examining the source code in early and then
subsequent versions of AIX, Dynix, ptx and Dynix/ptx, SCO can relate
an existing version of AIX, Dynix, ptx, or Dynix/ptx code to UNIX
code. Assuming that Linux code is similar to AIX. Dynix, ptx, and
Dynix/ptx code, SCO can then prioritize its search effort to find
evidence of substantial similarity between UNIX and Linux code.
Without the ability to prioritize its search efforts, SCO may be
required to spend an enormous amount of time, on the order of 35
man-years, searching Linux code for evidence of copying. (Sontag
Decl. ¶¶ 15, 29-54; see also SCO’s Memorandum
Regarding Discovery, dated May 28, 2004 (Exh. 23).)
60. SCO seeks the following materials to prioritize its analysis of
copying of UNIX code into Linux, and so that it can rebut
IBM’s motion: (i) all version control system and bug-tracking
information (including documents, data, logs, files, and so forth)
for AIX, Dynix/ptx, ptx, and Dynix from 1984 to the present, and
(ii) source code and log information for all interim and released
versions of AIX, Dynix, ptx and Dynix/ptx from 1984 to the present.
(Sontag Decl. ¶¶ 35-36.)
61. The evidence SCO currently has -- a few versions of AIX that IBM
selected, Linux code, and System V code -- is insufficient to
prioritize and avoid lengthy analysis because IBM could have copied
System V code into early versions of AIX and Dynix and subsequently
modified in the later versions that SCO has. Tracing the derivation
of SCO-owned UNIX code from System V into the code’s current
form in Linux will be facilitated by SCO’s access to
IBM’s Configuration Management Version Control (CMVC) and the
versions of AIX, Dynix, ptx and Dynix/ptx. (Sontag Decl.
62. IBM has produced only later versions of AIX. IBM has not yet
produced the earlier versions of AIX (or of Dynix, ptx, and
Dynix/ptx). On that basis alone, SCO therefore has been
significantly hampered in its ability to discover relevant facts
essential to oppose IBM’s Cross Motion.
63. The following materials also are relevant to prioritize SCO
efforts to find evidence to rebut IBM’s motion: All design
documents, white papers and programming notes, created from 1984 to
the present. These materials provide a wealth of information related
to code development beyond that which can be found in the source
code testing, VCS and bug-tracking log. Design documents also list
authors of code whom SCO can then depose to help SCO prioritize its
search to find evidence of Linux code that is substantially similar
to UNIX code or obtain admissions. (Sontag Decl. ¶¶
64. Further, programming notes contain the thought processes of
individual programmers as they write and revise code sequences. For
example, programming notes often list changes made to code, and
sometimes list additional changes to consider. Thus, programming
notes provide detailed rationale for code changes and an indication
of how the code may change in the future. Programming notes also
list authors of code whom SCO can then depose to help SCO prioritize
its search to find evidence of Linux code that is substantially
similar to UNIX code. (Sontag Decl. ¶ 53.) Finally, depositions
of authors ma y lead to admissions of copying into Linux.
65. In addition, the examination of the lineage of any given code
sequences faces substantial obstacles as explained further below.
(Sontag Decl. ¶¶ 36-42.)
Comparison of Source Code
66. Another way for SCO to discover relevant facts to oppose
IBM’s Tenth Counterclaim is to compare the source code (i) in
UNIX and AIX, Dynix, ptx, and Dynix/ptx, (ii) in AIX, Dynix, ptx,
and Dynix/ptx and Linux, and (iii) in UNIX and Linux.
67. SCO has not been given a reasonable opportunity to complete any
of the kinds of comparisons necessary to uncover facts relevant to
SCO’s opposition to IBM’s motion for summary judgment.
In addition to the inherent limitations on SCO’s (indeed,
anyone’s) ability to compare source code within a reasonable
period of time, as explained above, IBM indisputably has not
produced any early versions of AIX, Dynix, ptx, or Dynix/ptx source
code so that SCO could compare those with the source code in Linux
and SCO’s copyrighted UNIX code and streamline the analysis
SCO’s Reasonable Expectations Regarding Discovery
68. The kinds of discovery I have identified can be reasonably
expected to provide probative evidence regarding whether IBM’s
Linux activities infringe SCO’s copyrights. I will now
describe SCO’s expectations in this regard.
Recognition of Potentially Infringing Material in Linux
69. The record demonstrates that many individuals familiar with
Linux recognize that source code therein may infringe SCO’s
copyrights. Two examples are noteworthy.
70. In an article dated March 3, 2004, for example, the person
regarded as the developer of Linux, Mr. Torvalds, grudgingly
acknowledged with respect to the issue of whether Linux infringes on
SCO’s copyrights: “The only thing that makes any ounce
of sense is their claims about somebody using (Unix) System V
libraries.” (Exh. 64.)
71. In an article dated November 29, 2003, “Linux kernel
maintainer” Andrew Morton commented as follows on this
litigation, specifically in reference to “the XFS and JFS file
systems, which were originally developed under a Unix license and
then ported over to Linux”: “‘SGI did develop it.
It could be [SCO] has a legitimate case there, not technically, but
on the letter of the law.’” (Exh. 56.)
Comparison of Source Code
72. In addition to the foregoing, SCO reasonably expects that
further comparisons of source code will permit SCO to present
evidence in opposition to IBM’s Tenth Counterclaim. Examples
of facts from discovery to date that show copying of material from
UNIX into Linux include (i) substantial similarity of the
Read-Copy-Update (“RCU”) routine in Linux to a routine
in UNIX; (ii) copying of UNIX System V init (SYS V init) code in
Linux version 2.6; (iii) substantial similarity of the user level
synchronization (ULS) routines in Linux and similar routines in
UNIX; (iv) copying of SCO’s UNIX System V IPC code in Linux
2.4.20; (v) copying of SCO’s copyrighted UNIX “header
and interfaces” in Linux; and (vi) copying of SCO’s UNIX
Executable and Linking Format (ELF) codes in Linux. (Gupta Decl.
¶¶ 3-86.) The foregoing evidence demonstrates copying from
UNIX into Linux -- and is probative even if SCO is not seeking to
assert copyright in the foregoing material. SCO has not retained a
testifying expert on copyright issues; SCO has filed only a
relatively narrow copyright claim in this action and did so only in
February 2004. Such an expert would testify to the relative
importance of the foregoing materials in Linux.
Depositions of Contributors to Linux
73. SCO reasonably expects that depositions of Linux contributors
will reveal or lead to the revelation of facts relevant to
SCO’s opposition to IBM’s Motion. By way of example, Sam
Palmisano, then senior vice president of IBM and now its Chief
Executive Officer, has publicly described Linux as “a
community developed version of UNIX.” (“I.B.M. to Use
Linux In Software For Internet,” New York Times, Jan.
10, 2000 (Exh. 48).) SCO reasonably expects that the depositions of
individuals who (like IBM) acknowledge at the outset that they have
participated in the development of a “version of UNIX”
are likely to provide testimony that would demonstrate (for example)
that as a derivative of UNIX under the copyright laws, the use,
copying and/or distribution of Linux infringes SCO copyrights.
Depositions of Persons with Access to UNIX, AIX and/or Dynix
74. The depositions of persons and entities that had access to UNIX,
AIX, Dynix, ptx, and/or Dynix/ptx would permit SCO the opportunity
to discover facts essential to oppose IBM’s Tenth
Counterclaim. SCO anticipates the depositions may be relevant in
numerous ways. For example:
-- If such a person acknowledges that he or she induced, encouraged,
or enabled others to incorporate any UNIX software, or any software
based on or derived from any UNIX software, into any version of
Linux, that testimony would support SCO’s opposition.
-- If (as noted above) such a person testifies that he or she knows
of someone who induced, encouraged, or enabled others to incorporate
any UNIX software, or any software based on or derived from any UNIX
software, into any version of Linux, that testimony would permit SCO
to obtain additional admissible evidence.
-- If such a person testifies that he or she contributed source code
to any version of Linux, that testimony would support SCO’s
75. SCO reasonably expects that it will discover through such
depositions facts essential to oppose IBM’s Cross Motion. By
way of example:
-- SCO expects AIX engineer Dave Kleinkamp to testify that the
open-source community uses methods and concepts from AIX in order to
debug and improve the performance of Linux;
-- SCO expects Dynix engineer Paul McKenney to testify that he
participated in the development of the first iteration of RCU with
Dynix engineer Brent Kingsbury, who at the time was intimately
familiar with the RCU routine in UNIX from which SCO believes the
Dynix RCU routine derives, and that he has contributed to Linux
specific functions (such as symmetric multiprocessing and locking
techniques) that SCO believes are derived from UNIX;
-- SCO expects Dynix engineer Brent Kingsbury to testify that he was
involved in the design discussions regarding RCU and had previously
authored the design documentation for the version for the UNIX
operating system (4.2 MP) that included the RCU routine;
-- SCO expects Dynix engineer Gerrit Huizenga to testify on key
similarities between Dynix and Linux with respect to the overall
design and specific configuration of hardware and software (that is,
architecture), from which testimony SCO also expects to identify
IBM’s core areas of interest in UNIX and Linux; and
-- SCO expects Dynix engineer Jack Vogel to testify that when IBM
created its “Nifty Fifty” list of ideas on how to
improve Linux (Exh. 46), IBM intended to make those selections in
significant part on the advice of persons who had had access to and
substantial experience working with UNIX and AIX.
SCO’s Efforts to Obtain Evidence and Why They Were
IBM’s Failure to Produce Basic Discovery
76. In light of the procedural posture of this case, the parties
reasonably have not taken discovery on IBM’s Tenth
Counterclaim. (See paragraphs 14-23 above.) What is more, to
date SCO has been unable to obtain discovery relevant to its own,
long-standing claims in this case. Such discovery would permit SCO
to take further discovery, significant portions of which would bear
on IBM’s Tenth Counterclaim.
77. SCO served interrogatories and requests for production on IBM
over one year ago. Because the request came early in the case, it
sought the most basic types of discovery -- the elements on which
further discovery is based and without which further discovery could
not effectively proceed (for example, the identity of potential
witnesses, basic agreements, relevant correspondence, and AIX,
Dynix, ptx, and Dynix/ptx source code).
78. IBM’s responses were incomplete and in many instances
non-responsive. SCO sought to resolve the matter with IBM, but found
that this effort only resulted in further delay, ultimately leaving
SCO no other option but to move to compel IBM to respond.
79. On March 3, 2004, the Court granted SCO’s motion and
ordered IBM to comply, requiring IBM to provide specified discovery
that it had refused to provide and also to supplement deficient
responses. IBM, to date, has still failed to comply with the
80. IBM’s failure forces SCO to now renew its earlier
motion to compel -- simply to secure compliance with the
Court’s prior Order. (See Memorandum in Support of
Plaintiff’s Renewed Motion to Compel, dated July 7, 2004,
attached as Exh. 25; see also SCO’s Memorandum
Regarding Discovery, dated May 28, 2004, attached as Exh. 23
81. In short, over one year after SCO made discovery requests of the
most basic type, SCO has been forced to move to compel production
not once, but twice, the second time simply to pursue enforcement of
relief that this Court has already expressly ordered.
82. Until SCO receives from IBM the basic discovery requested, which
the Federal Rules contemplate will occur at the outset of the case,
SCO remains handicapped in that it cannot use such initial discovery
as the basis for more targeted discovery, including discovery
pertaining to IBM’s Tenth Counterclaim.
83. The Magistrate Judge ordered IBM, for example, to supplement its
response to SCO’s Interrogatory 5, which sought the identity
of “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.” Rather than providing the
requested information, IBM referred SCO to its earlier produced list
of over 7,000 names and stated that, to the extent readily
determinable, the contributions of these persons can be discerned in
the cases of AIX, Dynix, ptx, and Dynix/ptx in the “products
themselves.” In addition to not complying with the
Court’s Order to fully identify those persons, IBM’s
statement is not accurate.
84. As to AIX, the precise contributions of the thousands persons
referenced in IBM’s answers do not appear anywhere in the AIX
product itself. None of the numerous AIX files that SCO has reviewed
specifies the precise contributions of any of the 7,200 named
individuals. In fact, the AIX product does not appear to identify
any of the authors of the code, much less what each person
85. IBM’s refusal to provide this basic information severely
prejudices SCO’s defense of all of IBM’s counterclaims.
Had IBM properly answered this question submitted over one year ago,
SCO would have known the precise contributions of each person to
AIX, which in turn would have allowed SCO to take depositions of
significant authors of AIX, which in turn would have provided direct
evidence relating to IBM’s duplication, modification, and/or
distribution of material in UNIX in which SCO holds copyright.
86. Similarly, in the fall of 2003, SCO first requested contact
information relating to employees, witnesses, and other individuals
identified in IBM’s interrogatory responses. When IBM refused
to provide the necessary contact information, SCO filed it motion to
compel and this Court issued its order requiring IBM to provide the
requested contact information for up to 1,000 potential trial
witnesses. After the Court Order was entered, SCO sent IBM a request
for the identity of selected witnesses, including a list of 81
persons taken from IBM’s discovery responses. IBM refused to
provide the necessary information for 49 of the 81 persons, claiming
it had no obligation to do so.
87. There is no good faith basis for IBM’s refusal to provide
contact information when those same individuals were identified by
IBM as witnesses from whom it had retrieved documents to support its
case. It is rather incongruous to claim that these people have
relevant information when IBM wants to collect it, but are not
witnesses when SCO wants to find out what they know. It must also
follow from IBM’s contention that the Tenth Counterclaim
arises out of the same transactions or occurrences as SCO’s
claim that IBM believes the listed witnesses would provide testimony
relating to the Tenth Counterclaim. IBM continues to ignore
SCO’s request and, to date, has not provided the contact
88. Over a year ago, 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. IBM failed to produce even a single line of code from
either AIX or Dynix between June 24, 2003, and December 4, 2003. On
December 4, 2003, which was the day before the hearing on
SCO’s motion to compel production of the source code, IBM
produced two CDs containing limited version of Dynix. IBM did not
produce a single line of AIX code at that time.
89. 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 stay, also issued a specific numbered directive
requiring that IBM finally produce at least some version of AIX code
(and additional Dynix code). 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 source code comparisons. (Even then, the
source code was first produced in a format that was unusable, and
that IBM knew or should have known would be unusable).
90. Comparison of AIX and Dynix source code with source code in UNIX
and Linux, as explained above, 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, as I explain below, to identify all of the instances of
IBM’s copying from UNIX into AIX and Dynix and into Linux.
Inherent Limitations on the Review of Source Code
91. In addition to IBM’s failure to produce basic discovery,
there are inherent limitations on the ability of any company or
person (or computer) to compare source code between computer
operating systems. Those limitations have precluded SCO from
completing the source code comparisons necessary to obtain facts
essential to justify opposition to IBM’s motion for summary
92. Notwithstanding the following limitations, SCO has undertaken
the tasks detailed in Mr. Sontag’s Declaration, and the
numerous tasks listed in SCO’s Memorandum Regarding Discovery,
dated May 28, 2004, regarding the comparison and tracking of source
code. (See Exh. 23.) Mr. Sontag addresses these issues in
detail in his attached declaration. Among the important factors
bearing on the evidence SCO has to date are the following.
93. A comparison of all of the source code in one computer operating
system with all of the source code in another computer operating
system could not be performed manually for purposes of any
litigation. A representative example makes the point. The Linux
kernel version 2.4, comprises approximately 4 million lines of code.
The UNIX System V 4.2 MP kernel comprises approximately 3.4 million
lines of code. There are numerous versions of UNIX, AIX, and Dynix.
Assuming even nominal times for reviewing this much code, as much as
35 man-years may be expended looking for evidence of copying UNIX
code into Linux. Clearly, some means for streamlining the review is
needed. Given the foregoing facts, as well as other significant
limitations on the use of automated search tools (described below),
SCO and its experts have not sought to undertake any wholesale
comparison of the source code in any two computer operating systems.
(Sontag Decl. ¶¶ 15, 18-23.)
94. Automated search tools cannot remove this burden completely
because they have very significant limitations. The tools are
designed to find lines of code that are identical or nearly
identical in every detail, and they perform that function well. SCO
has sought to modify and improve the tools to locate lines of code
that are not identical but are nearly identical, but the tools have
not performed that function well. Ultimately, the automated tools
simply assist a programmer to locate blocks of code that might have
similarities. The programmer must then visually review the code in a
difficult and labor-intensive process. Often this review is only
possible if each version of the code can be reviewed to follow the
changes from one version to the next. (Sontag Decl. ¶¶
95. In the face of the foregoing limitations, SCO and its engineers
have sought to compare the source code in UNIX System V with source
code in AIX/Dynix and Linux by making only educated guesses about
where similar source code may appear in the systems being compared.
An example of such an approach is to start by comparing the names of
the files in the operating systems. SCO has considered the structure
of the operating systems being compared and has compared like
components with each other (for example, compared filesystems with
filesystems, inter-process communication with inter-process
communication, program loading with program loading, and the like).
These comparisons of course represent only a very small fraction of
the total number of comparisons that could be made among the
numerous versions of the UNIX, AIX/Dynix, and Linux operating
systems. (Sontag Decl. ¶¶ 14-23.)
I declare under penalty of perjury that the foregoing is true and
July 8, 2004
CERTIFICATE OF SERVICE
Plaintiff, The SCO Group, hereby certifies that a true and correct
copy of DECLARATION OF JOHN HARROP IN SUPPORT OF SCO’S
OPPOSITION TO IBM’S MOTION FOR PARTIAL SUMMARY JUDGMENT
was served on Defendant International Business Machines Corporation
on the 9th day of July, 2004, as follows:
BY HAND DELIVERY:
Alan L. Sullivan, Esq.
Todd M. Shaughnessy, Esq.
Snell & Wilmer L.L.P.
Evan R. Chesler, Esq.
Cravath, Swaine & Moore LLP
Donald J. Rosenberg, Esq.