Let's correct some facts and assumptions in Dr. Gary Pisano's Expert Report on behalf of SCO in the SCO v. Novell trial, shall we? Why? First, it's fun. Geek fun, granted, but that is who we are. Second, I see what I view as mistakes, and I surely don't want people to be confused or for errors to go down in history unanswered. If the testimony is excluded, which is the relief that Novell is asking the court for, it won't be answered in the trial. So, for historians, and for our own enjoyment, here is Groklaw's answer to the expert report of Dr. Gary Pisano, part 1. Feel free to meaningfully and politely respond to his report in your comments, and I'll glean everything, add my own 2 cents, and we'll do a second, color-coded version of the report, with our rebuttal interspersed. To aid us I've done as text Novell's memorandum in support of its motion for a Daubert hearing to disqualify Dr. Pisano (not just some of his testimony), and then Exhibit A, his submitted report, and a snip from his deposition, in which he tries to say an online survey he didn't do is reliable because Laura DiDio of Yankee Group did it. It's also why SCO would be wise not to oppose this motion. I'm sure the Novell lawyers would make mincemeat of him, should he take the stand, because although he is apparently an expert on business administration, what he doesn't know about Linux and Unix is a lot, by my reading, unless he is selectively cherry picking only materials that match SCO's story. So, let's assume good faith and help him out.
To get us started, here are a few things I noticed off the top of my head:
- He says he can't find any other factor for the decline in interest in SCOsource, other than Novell's public claim that it retained the copyrights. Seriously? He can't find *any* other reasons? I believe, if he had checked SCO's SEC filings and financial teleconferences, which he lists as references, he'd have found some. Or just read Rob Landley and Eric Raymond's Halloween IX: It Ain't Necessarily So, which they published soon after SCO sued IBM in March of 2003, listing so many, many reasons not to believe that there was any danger from SCO's claims.
-
Pisano claims that but for Novell's statement on May 28, 2003 that it didn't transfer copyrights to SCO under the APA, SCOsource would have taken off. He dates a decline from that date in interest in SCOsource. But he references:
Gartner, "SCO's Threat to Sue Linux Users Serious but Remote," Weiss, G. May 21,
2003.
That was written prior to Novell's statement on May 28. People lacked interest in buying protection for a remote conceivability.
-
Pisano says millions of Linux enterprise folks would have purchased SCOsource by the end of 2007 ("I believe that users with high
levels of concern about mitigating the risks associated with IP infringement would have acquired
a license in the four years between the announcement of the program and the end of 2007"), but he wrote that in May of that year, and in August of 2007, just three months later, Judge Kimball ruled on summary judgment that SCO didn't get the copyrights. Clearly, there were plausible reasons to think SCO was wrong in its claim, quite apart from Novell's words, particularly for anybody following the litigation closely, like on Groklaw, which he scrupulously avoids mentioning. It's odd, but with SCO's motion to make sure no witnesses mention Groklaw's coverage, it will leave the only journalists being mentioned the ones Dr. Pisano lists, like Maureen O'Gara and Dan Lyons. Do you consider them unbiased? If not, and I'd agree with you, why is their biased coverage all right to mention, and no one is allowed to mention Groklaw, not that Groklaw is biased, in my opinion? In every way, I see SCO gives evidence of wanting to speak without anyone getting to rebut. None of that extensive rebuttal coverage in the media is chronicled by Dr. Pisano chronicles, that I could see.
- He also references private interviews he conducted, I gather, with Darl McBride, Chris Sontag, and a number of HP employees. Is that kosher? I mean, is he introducing it as evidence through a back door? No one can cross examine the unnamed HP employees, for example, so I can't see how it's acceptable to introduce it.
- He opines that the market for indemnification stemmed from Novell's public claims, because they reduced the perceived risks of offering such indemnification:
82. First, if SCO's ownership of the UNIX copyrights were settled, then the risk of offering a program indemnifying purchasers from liability to SCO increases significantly. A provider of indemnification, like any provider of insurance, essentially is taking a bet on an uncertain liability. The indemnifier profits if the risky event in question does not occur. The indemnifier's risks increase with the likelihood of the risky event in question. In this case, the "risky event" is SCO being able to enforce its intellectual property rights. Economic reasoning suggests that a provider of insurance (here, an indemnification) will not offer such insurance in the absence of uncertainty. If the event in question has no probability of occurring, there will be no demand for the indemnification. In contrast, if the event in question is certain to occur, there will be no supply of indemnification. If SCO's property rights were certain, then potential
indemnifiers would not have found it profitable to offer indemnification. They would be, in essence, taking on a certain liability. By making SCO's property rights uncertain, Novell's actions decreased the perceived risks for a vendor of offering indemnification for its customers. As a result, the number of companies that began offering indemnification increased. In the "but for" world, where SCO's ownership is certain and risks of being found liable for infringement are higher, these vendors would not have been likely to enter the market. However, I can tell you without fear of contradiction that OSRM didn't offer indemnification on that basis. It did a code comparison along with other due diligence, and no matching copyrightable-by-SCO code turned up. And by the way, you'll notice the lack of anything to shake a stick at in the IBM litigation, after all these years of discovery. So they were right not to worry, I deduce, or at least rational. We'll see about the court system, but the view in the trenches has always been that in any fair courtroom, SCO would lose.
- And he doesn't mention the GPL at all as a factor, even though, as I earlier pointed out, it was central. Dr. Pisano therefore imagines that 19-45% of all Linux users would have taken a SCOsource license, if Novell hadn't spoken out. Dr. Pisano doesn't factor in the GPL, but the marketplace did understand it and did factor it in. Frankly, a lot of the "evidence" he so carefully collects as references in his footnotes and documents list were things the Linux community viewed as deliberate FUD, and we just laughed at it. Remember SCO's the "GPL is UnConstitutional" silliness? We were falling down laughing. We knew that many of the things we were hearing were simply ridiculous. If you know Linux's history and UNIX's history, SCO's claims were at the center of a truly hilarious performance, as we viewed it, and we didn't take the claims seriously, even if we thought some of the principals were creepy scary. We *knew* there was no Minix code in Linux. We knew Linus would never steal anybody's code. And we knew SCO's description of the development process wasn't true. And we noticed SCO refused to show the code, and when it lifted its skirt to show just an ankle, so to speak, it was even funnier, because it was not their code they showed. It was like they were the Keystone Kops. So when SCO would make their claims, it was funny. You don't get scared of things that make you laugh out loud. We expected SCO to lose from day one, because of the GPL, even if they could persuade a jury that they should get the copyrights, because they can't *do* anything with those copyrights. They released Linux under the GPL repeatedly. They like to pretend they are not Caldera, just Santa Cruz, but that's not true. So we expected them to lose. My very first article about SCO was titled, "SCO Falls Downstairs, Hitting its Head on Every Step," and SCOsource was already announced, the first version, by that time. His numbers are then, in my view, mistaken. I'm trying to be polite.
Novell lists a number of other flaws that it believes disqualify Dr. Pisano as an expert on the matters he is offering to speak to, in the memorandum in support. But honestly, if they put him on the stand, it'll be a hoot. Maybe not for Harvard, which ought to be mortified, frankly, in my opinion. For example, Dr. Pisano
also references, in his report and in his deposition,
the Yankee Group report, "Indemnification Becomes Open Source's Nightmare and Microsoft's Blessing", also Exhibit B to Novell's memo, which is a funny title too, because it didn't turn out like that at all. That paper is available all over the Internet, for example here. Oh, look. Microsoft put it on its website [PDF]. What a coincidence. The title makes it sound like deus ex machina, rather than a Microsoft fueled-production number. And the report includes the amazing admission that the odds of getting sued by SCO were really low, like the odds of getting audited by the IRS. Can she actually prove that statement with a study of some kind, I wondered, or is that a guesstimate?
She suggests indemnification, which is exactly what the marketplace offered and users accepted from folks like Red Hat and HP and Oracle and Novell, which proved to be a fine avoidance strategy to any SCOsource license. People are not likely to prefer getting protection from an enemy rather than a friend. It was a Sunbelt Software survey, a Microsoft partner, by the way. "The Yankee Group 2004 Windows, UNIX & Linux Comparison Survey", meaning for all we know, it was the Windows users in the survey who were so interested in indemnification. I seriously doubt any Linux folks would go to Sunbelt Software to take a survey. Maybe as a lark. I remember this Yankee Group report too, because it misquoted OSRM and Ms. DiDio made statements to the press that I viewed at the time as libel. That article takes you to some of the many times Ms. DiDio got things wrong, in case historians wish an antidote to Dr. Pisano's naive acceptance of everything Yankee Group. And I remember researching Sunbelt's methodology, which Dr. Pisano thinks so highly of. They do online surveys, or arrange for them to be done, which theoretically anyone can respond to, I gathered. There are no phone calls on the online survey. It's just online, and it's the honor system, from all I could determine. Hey. If it's on the Internet, it means it's true, right? Who'd ever lie on the Internet? In fact, here are the questions for another such Yankee Group survey, on Windows v. Linux TCO. Look at the bottom of the page. It tells you that you too can do such an online survey by clicking on the link to VoVici. I guess that means Groklaw could run a survey on how many Fortune 1000 companies think SCO is full of it, and thought so prior to Novell saying it owned the copyrights, and not SCO. Why, yes. Yes we could. And you could all answer for the companies you work for, without mentioning it to your boss, or just pretend to be Steve Ballmer. Why not? On the Internet, no one knows you're a dog. Just kidding. Don't ever pretend to be Ballmer. It might stick, and then where would you be? I can't believe any expert would defend such survey methods. Defend them? He builds his castle around them, deciding from this survey, this Internet poll, how SCO's damages should be calculated as a result of Novell letting the world know it was of the opinion that it owned the copyrights. And he does so despite not knowing anything at all about their methods. Check out his deposition.
Should Novell really, truly have kept that a secret, that they were of the opinion they retained the copyrights? If you answer no, that they had a duty to reveal it, then why were they sued? Because SCO is SCO, and they can find folks willing to help them. At $600/hour, I noticed in his report. Here's the Yankee Group disclaimer at the end of its report he so loves: The Yankee Group believes the statements contained in this publication are based on accurate and reliable information. However, because our information is provided from various sources, including third parties, we cannot warrant that this publication is complete and error-free. The Yankee Group disclaims all implied warranties, including, without limitation, warranties of merchantability or fitness for a particular purpose. The Yankee Group shall have no liability for any direct, incidental, special, or consequential damages or lost profits. This publication was prepared by the Yankee Group for use by our clients. See what I mean? He relies on Yankee Group, but they say, don't rely on *us* whatever you do -- we relied on third parties, and who knows, really, what they did? It's interesting but don't use it for "a particular purpose". Like in a lawsuit, to "prove" something, you mean? He also in his deposition says he doesn't know the precise methods used, but he believes the survey is reliable, because that's what they do for their living, so they wouldn't do anything to undermine their credibility. Then he compares it to Toyota, of all unfortunate choices at this moment in time. "They know how to make cars. These guys know how to do surveys." I mean, just unfortunate. Exhibit D to Novell's memorandum is a report on how to do surveys properly, Reference Guide on Survey Research by Shari Seidman Diamond, and rather than do it as text, here's a link to where you can find it on the Internet, and if you search by title, Google will do it as HTML for you.
I don't know which lawyer did this deposition of Dr. Pisano, but whoever it was did a wonderful job. Notice the page numbers on the deposition. It starts with asking Dr. Pisano about methods and reliability on pages 64 and 65. Then they apparently went on to some other topics, and then wham, on page 212 the lawyer is hammering away again, asking in several different ways about why Pisano trusts the reliability of the survey and whether he knows anything at all about the methodology behind the survey, and the lawyer is successful in getting him to admit that he doesn't know anything about that. He merely "believes" it must be reliable. That is not enough for a court of law. Or, phrasing it more carefully, it's not supposed to be enough, as Novell points out in its memorandum. Dr. Pisano references himself to prove a point, which made me smile. So if you want to get a snapshot of his thinking, check out the title of his paper,
"Profiting from Innovation and the
Intellectual Property Revolution [PDF]. The title says it all. This is not to mock. I've no doubt at all that Dr. Pisano is well qualified to speak on the subjects he is an expert in. And if I met him at a dinner party, I'd probably enjoy talking to him about his areas of expertise, and on that I'd listen respectfully, knowing that he knows more about those areas than I do. But I'd probably give him an earful about what he doesn't know about SCO, Linux, Unix, and the GPL. We'll be noting some technical issues in his report, too, like his description of servers, but I'll save that for part 2 of our rebuttal.
This is all soooo long, I'll give you some links so you can hop around to each document, depending on which one you want to read, without having to scroll:
[ Pisano's Expert Report, Exhibit A ]
[ Novell's Memorandum in Support of Daubert Hearing Motion ]
[ Pisano Deposition, Exhibit C ]
***************************************
EXHIBIT A
Brent O. Hatch (5715)
Mark F. James (5295)
HATCH, JAMES & DODGE, PC
[address, phone, fax]
David Boies (admitted pro hac vice)
Robert Silver (admitted pro hac vice)
Edward Normand (admitted pro hac vice)
BOIES, SCHILLER & FLEXNER LLP
Devan V. Padmanabhan (admitted pro hac vice)
DORSEY & WHITNEY LLP
[address, phone, fax]
Stephen N. Zack (admitted pro hac vice)
BOIES, SCHILLER & FLEXNER LLP
[address, phone, fax]
Stuart Singer (admitted pro hac vice)
BOIES, SCHILLER & FLEXNER LLP
[address, phone, fax]
Attorneys for Plaintiff, The SCO Group, Inc.
IN THE UNITED STATES DISTRICT COURT
FOR THE DISTRICT OF UTAH
THE SCO GROUP, INC.,
a Delaware corporation,
Plaintiff/Counterclaim-Defendant,
vs.
NOVELL, INC.,
a Delaware corporation,
Defendant/Counterclaim-Plaintiff.
_____________________________________
EXPERT REPORT AND
DECLARATION OF GARY PISANO
Civil No.: 2:04CV00139
Judge Dale A. Kimball
Magistrate Brooke C. Wells
DECLARATION AND EXPERT REPORT OF GARY PISANO
IN THE UNITED STATES DISTRICT COURT ....................... 1
I. ASSIGNMENT ....................... 2
II. QUALIFICATIONS ....................... 3
III. SUMMARY OF CONCLUSIONS ....................... 4
IV. BACKGROUND ....................... 5
A. The Evolution of UNIX....................... 5
B. SCO's Relationship with UNIX....................... 7
C. The Effect of Linux on the Market for UNIX ....................... 10
D. Novell's Statements Regarding the UNIX Copyrights ....................... 13
V. THE POTENTIAL MARKET FOR SCO'S SCOSOURCE LICENSE ....................... 19
A. Estimated Number of Potential Customers ....................... 19
B. Demand for Linux IP Indemnification ....................... 23
C. SCOsource Adoption ....................... Rate 31
D. Alternative Vendors Would Not Have Been a Viable Option Absent
Novell's Slander ....................... 32
i. HP's Indemnification Program ....................... 36
ii. Red Hat's Indemnification Program ....................... 36
iii. Novell's IP Indemnification ....................... 37
iv. Oracle's IP Indemnification ....................... 38
E. The Future Market for SCOsource Licenses ....................... 39
1
I. ASSIGNMENT
1. I have been retained to estimate the market demand for and number of SCOsource
Intellectual Property Licenses for Linux (also known as a "right to use" or "RTU" license) that
the SCO Group would have sold had that business not been damaged by Novell's alleged slander
of title, breach of contract, and unfair competition.
2. I calculated the number of lost license sales by estimating the potential market for
the SCOsource product and SCO's penetration of this market had SCO's ownership rights of
Unix System V copyrights not been challenged by Novell. To arrive at my estimate, I analyzed
the industry conditions facing SCO during the relevant time period, and applied my expertise in
technology strategy, innovation, and industry analysis to the computer software and server1
industries and to the relevant intellectual property and licensing issues in this case.
3. In analyzing competitive forces and market conditions, I rely on my professional
training in economics and business. There is a substantial body of work in my field on the
economic and business theory and analysis of high-technology industries, as well as the
2
competitive strategies employed within such contexts.2 This literature is widely accepted in the
management and economics fields and is taught in economics departments and business schools
across the United States. It is regularly relied upon by businesses, business consultants, and
federal agencies.1 For the purposes of this report, I assumed liability. In my professional
opinion, the relevant damage period extends from the date of first slander, May 28, 2003, to the
end of trial in approximately 2007. Thus, I examine the market for SCOsource licenses during
this period assuming that SCO rightfully owns the UNIX copyrights and as if Novell had never
challenged SCO's ownership rights.
4. In the course of my analysis, I reviewed substantial information, which
documents and materials are identified in Exhibit B of this report. I reserve the right to
supplement this report should additional information come to light or be produced in this matter.
II. QUALIFICATIONS
5. I hold a B.A. degree in economics (with distinction) from Yale University and a
PhD in business administration from the University of California, Berkeley.
6. I am the Harry E. Figgie, Jr. Professor of Business Administration at Harvard
Business School. I currently serve as the co-chair of the Technology and Operations
Management unit. I have taught MBA, doctoral-level, and executive-level courses
in Technology and Operations Management, Operations Strategy, the Management of Innovation,
and Competition and Strategy since joining the Harvard faculty in 1988.
3
7. During that time, my research and teaching have focused on technology and
operations strategy, process and product development, vertical integration, outsourcing, and
technology licensing. I have consulted for a number of corporations in a range of high-technology industries, advising senior managers on issues of technology strategy, operations
strategy, product and process development, and competitive strategy.
8. I have published six books and authored many book chapters within the strategy,
operations, and technology areas. In addition, I have published numerous articles in leading
professional journals, including The Harvard Business Review, Management Science, Decision
Sciences, Strategic Management Journal, Administrative Science Quarterly and others. I have
developed, authored, and published over 100 papers, presentations, case studies, and other course
materials related to the technology, operations, strategy, innovation, product development, and
technology licensing areas. My curriculum vitae, which includes a list of my publications, is
attached as Exhibit A.
9. I have received numerous awards and distinctions for excellence in research and
publishing, and served as an editor of the technology management journal, Research Policy.
10. In the last four years, I have only filed one other expert report, in the SCO v. IBM
litigation. To this date, I have not testified at trial in that matter, although I have given
deposition testimony.
11. My billing rate in this case is $600 per hour.
III. SUMMARY OF CONCLUSIONS
12. After applying the principles and methodologies from my field of study to the
facts of this case, I reached the following conclusions:
(a) The potential license market for this analysis includes Linux 2.4 and Linux 2.6
4
Environment shipments. These are the Linux versions that SCO contends
infringe its UNIX copyrights. For reasons I detail later in this report, I
believe this is a conservative estimate of the size of the Linux Market.
(b) SCO would have sold its SCOsource license to between 19% and 45% of the
Linux Market, but for Novell's statements regarding copyright ownership.
(c) Based on my knowledge of the industry and my research and analysis in this
case, I am not aware of any causes for SCO's loss of those SCOsource
license sales other than Novell's conduct during the relevant time period.
(d) Given the lapse of four years since Novell's initial statement, and the
changes in SCO and the market during that time, a court decision now
vindicating SCO's ownership of the UNIX copyrights would not enable SCO simply to recover the SCOsource sales in the future that it has lost due to
Novell's slander.
IV. BACKGROUND
13. SCO's damages in this case were caused by Novell's multiple statements
contending that Novell, and not SCO, owned the UNIX copyrights. In order to analyze the
impact of Novell's statements on SCO, it is important to understand the industry context for the
events at issue. This section contains a brief description of the evolution of UNIX, SCO's use of
UNIX, the rise of Linux and SCO's response to that development in the marketplace. Finally, I
address the series of Novell's statements disputing SCO's ownership of the UNIX copyrights,
and their timing in relation to the SCOsource program.
A. The Evolution of UNIX
14. UNIX is an operating system, like Windows, DOS, or (later) Linux. The origins
of UNIX can be traced back to AT&T. In 1969, AT&T began development of UNICS, later
5
renamed UNIX. AT&T initially licensed UNIX to end users, such as universities and
researchers, where its use spread rapidly in the 1970s. UNIX became popular because it could
run on any hardware platform; it also enabled software programs to run more easily across many
hardware platforms.4 These features allowed customers to use UNIX on hardware from many
different vendors, and gave greater flexibility in choosing software applications to run on the
hardware and operating system.5 UNIX was considered an "open," "cross platform" operating
system (where "open did not mean that the source code was freely available, as in "open source,"
but rather that UNIX provided greater flexibility, in contrast to "closed" systems that effectively
locked-in users).
15. Prior to 1983, AT&T licensed UNIX primarily to universities and researchers,
along with the source code, for teaching purposes as well as internal use, but in the latter case,
such use was limited to specific Central Processing Units ("CPUs"). At this time, customers did
not have the right to re-distribute UNIX. Instead, to run UNIX on more than one machine or
CPU, a customer was required to obtain additional licenses.
16. Starting in 1983, AT&T sought to capture the commercial value of UNIX by
licensing UNIX to large computer system vendors such as Sun, HP, SGI, and IBM. These
licensees typically specialized and bundled their modifications and enhancements of UNIX with
their own proprietary hardware. AT&T's approach included a separate licensing agreement
6
(required in addition to their traditional customer licensing agreement) that allowed the licensee
to become a "distributor" of UNIX. Put differently, licensees could distribute their own
modified and enhanced versions of UNIX to customers in binary form, but could not distribute
the source code outside the license framework.
17. Licensees had to pay a significant upfront fee for this right to distribute UNIX.6
Additionally, licensees were required to pay for each binary copy they distributed; these
payments were in the form of a royalty.
18. AT&T formed a separate entity, UNIX System Laboratories (USL), retaining
majority ownership until 1993, when it sold its interest to Novell.
B. SCO's Relationship with UNIX
19. SCO's predecessor in interest, the Santa Cruz Operation, Inc.("Santa Cruz"), was
incorporated in 1979 as a UNIX system porting and consulting company.7 The company shipped
its first product, SCO XENIX System V, a packaged version of the UNIX operating system, in
1983. In 1985, the company introduced its first operating system for the 32-bit Intel
microprocessor environment, SCO XENIX 286, and followed with SCO XENIX 386 in 1987.8
By early 1989, nearly 3,000 applications ran on SCO's UNIX.9
20. In September of 1995, Santa Cruz purchased the "UNIX System source
technology business", "including core intellectual property" from Novell Inc., which Novell had
7
acquired from AT&T's UNIX System Laboratories.101112 21. Qn or about May 4, 2001, Santa Cruz sold its Server Software and Professional
Services Divisions to Caldera Systems, Inc. This transaction transferred the UNIX and
UnixWare rights to Caldera Systems. In August 2002, Caldera changed its name to The SCO
Group Inc.
22. An important element of SCO's strategy was to focus on developing and extending
the UNIX operating systems for use on the Intel Architecture. Historically, most "flavors" of
UNIX, the different UNIX operating systems offered by companies with System V distribution
agreements, ran on proprietary RISC (Reduced Instruction Set Chip) microprocessors. Because
these chips are designed to integrate the processor with the specific hardware and the specific
UNIX operating system from a single manufacturer, a given manufacturer's "flavor" of UNIX
could only be used on that manufacturer's hardware. Because SCO's UNIX products were
designed to run on Intel Architecture-based computers, and because multiple hardware vendors
8
provided server systems based on the Intel Architecture, SCO's products offered customers the
ability to choose their hardware vendor separately from choosing their operating system.13 23. With the acquisition of Novell's UnixWare in 1995, SCO extended its role as the
pre-eminent provider of the UNIX operating system for Intel Architecture-based servers.14 By 1999. SCO held 80% share of the UNIX on Intel market.15 As depicted in the following table, worldwide shipments of servers based on Intel Architecture approximately tripled from 1995-1999.16
9
Figure 1: Hardware Server Shipments, Intel architecture, Worldwide, (1995-2005)
[See PDF for chart. Source: Gartner.17 ]
C. The Effect of Linux on the Market for UNIX
24. In the late 1990's, IBM was losing market share in the mainframe market, losing share in the UNIX on RISC market, and was the third place provider for servers running Windows on Intel.18 Looking to reverse this slide, IBM chose to embrace Linux, which, at the time, was a popular operating system for web serving but not viewed as capable of handling the mission-critical applications required for an enterprise-ready operating system. IBM committed resources, including engineer resources and software code, to make Linux enterprise-capable. IBM claims they committed over $1 billion to the Linux effort.19 The features and functionality of Linux improved. As a free UNIX clone, offering similar functionality and features as UNIX
10
did without the price of UNIX, Linux quickly captured the UNIX on Intel opportunity that SCO
was positioned to leverage.
25. In 2003, SCO discovered that portions of its protected intellectual property had
been included in Linux by IBM, and possibly by others as well. This issue regarding the
presence of SCO's intellectual property in Linux became the basis for litigation currently
pending between SCO and IBM.20 SCO further learned that Linux infringed its UNIX
copyrights.21
26. Qn January 22, 2003, faced with a dramatic decline in its server operating
business due to Linux, SCO announced the creation of a new division, SCOsource, chartered
with expanding the licensing of its copyrighted UNIX code outside of SCO's operating system,
including for use on Linux. SCO also announced the retention of Boies, Schiller & Flexner,
LLP, to assist in research and advise the company on its intellectual property. SCOsource
introduced the SCOsource license to enable customers to use Linux without infringing on SCO's
copyrighted code.22 The SCOsource licenses primarily took two forms: (1) a UnixWare source
code license to developers; and (2) a right to use license (RTU License) for commercial end-users of Linux.23 Both licenses contained a covenant not to sue, stating that the licensee would
not be exposed to liability for the use of SCO's intellectual property in Linux. These licenses
enabled customers to use Linux without infringing on SCO's copyrighted code.
11
27. On March 7, 2003, SCO filed a legal action against IBM claiming
misappropriation of trade secrets, tortious interference, unfair competition and breach of
contract. SCO alleged that IBM tried "to improperly destroy the economic value of UNIX,
particularly UNIX on Intel, to benefit IBM's new Linux services business."24 SCO claimed that
IBM was required by contract to keep SCO's UNIX software code confidential, and prohibited
from unauthorized distribution or transfer, including to Linux.
28. During the early months of the SCOsource program, SCO signed expansive
licensing agreements with Sun Microsystems and with Microsoft. The Sun agreement was
executed on February 25, 2003.25 The Microsoft agreement was announced by SCO on May 19,
2003.26 Through a series of amendments and licenses, these two agreements ultimately
represented over $25 million in revenue to the SCOsource division in 2003.27
29. On May 12, 2003, SCO sent a mailing to "Fortune 1000" Linux users stating that
"Linux infringes on our UNIX intellectual property and other rights."28 SCO also reminded
Linux users that "Linux distributors do not warrant the legal integrity of the Linux code provided
to customers. Therefore legal liability that may arise from the Linux development process may
also rest with the end user."29 These letters were the precursor to SCO's expansion of its
SCOsource licensing program.30
12
D. Novell's Statements Regarding the UNIX Copyrights
30. On May 28, 2003 SCO released its earnings for the second fiscal quarter, and
announced several positive developments:
During the quarter ended April 30, 2003, the first two licensing agreements
related to our SCOsource initiative, our division for licensing and protecting the
Company's UNIX intellectual property, provided the Company with $8.8
million in cash and added $6.1 million to gross margin. There are over 6,000
source code licensees of our UNIX operating system, and we believe the
SCOsource initiative will continue to gain momentum as we pursue
enforcement ofthe Company's intellectual property rights," said Darl McBride,
president and CEO.
McBride continued, "These positive quarterly financial results, including net
income for the first time in the Company's history, have strengthened our
balance sheet and financial position. Our increased cash balance and working
capital has positioned the Company for its launch of SCOx, our web services
strategy, and will provide us with other opportunities to drive growth in future
quarters."
"We expect that revenue for our third quarter, ending July 31, 2003, will be in the
range of $19 million to $21 million. These projections anticipate revenue
contributions of approximately two-thirds from our operating system platforms
and one-third from our SCOsource initiative," said McBride.31
31. On the same day of SCO's announcement, Novell issued a public letter claiming
that Novell, not SCO, owned the copyrights to the UNIX code SCO claims is in Linux.
Specifically, Novell stated:
Importantly, and contrary to SCO's assertions, SCO is not the owner of the
UNIX copyrights. Not only would a quick check of the U.S. Copyright Office
records reveal this fact, but a review of the asset transfer agreement between
Novell and SCO confirms it. To Novell's knowledge, the 1995 agreement
governing SCO's purchase of UNIX from Novell does not convey to SCO the
associated copyrights. We believe it unlikely that SCO can demonstrate that it
has any ownership interest whatsoever in these copyrights.32
13
32. This was the first public indication that there was any question regarding SCO's
ownership of the UNIX copyrights. This date, May 28, 2003, marks the beginning of the
damages period from Novell's statements regarding the ownership of the UNIX copyrights.
33. In the week that followed Novell's statement, SCO directed Novell's attention to
Amendment No.2, a contractual document between Santa Cruz and Novell regarding ownership
of the UNIX copyrights. In response to this document, Novell issued a press release agreeing
that Amendment No. 2 "appears to support SCO's claim that ownership of certain copyrights for
UNIX did transfer to SCO in 1996."33 In a letter to SCO that same day, Novell continued to
characterize SCO's responsive claims against Novell as "absurd and unfounded."34
34. Despite its qualified public admission regarding Amendment No. 2, within weeks
Novell reasserted its claim to copyright ownership. On June 26, 2003, Novell notified SCO that:
Upon closer scrutiny... Amendment No. 2 raises as many questions about
copyright transfers as it answers. Indeed, what is most certainly not the case is
that "any question of whether UNIX copyrights were transferred to SCO as part
of the Asset Purchase Agreement was clarified in Amendment No.2" (as SCO
stated in its June 6 press release).35
This letter precluded SCO from assuring prospective SCOsource customers that Novell no longer
asserted ownership over SCO's copyrights.36
35. On July 21, 2003, SCO announced that it had received U.S. copyright
registrations for UNIX System V source code.37 SCO also indicated that it would offer licenses
14
to support run-time, binary use of Linux for all commercial users of Linux based on kernel
version 2.4.x and later.38
36. Novell responded to the announcement of SCO's copyright registration by
reiterating its claim of copyright ownership, stating "SCO's claim to ownership of any
copyrights in UNIX technologies must be rejected, and ownership of such rights instead remains
with Nove11."39
37. On August 5, 2003, SCOsource formally announced the availability of the SCO
Intellectual Property License for Linux, and that SCO would be meeting with commercial Linux
users to present the details of this "right to use" ("RTU") license. The run-time license permits
the use of SCO's intellectual property in binary form as contained in Linux.40 Within weeks of
this announcement, a Fortune 100 company, Computer Associates, requested from SCO and
purchased a SCO RTU license.41
38. On October 2, 2003, Novell filed for (and ultimately received) eight UNIX
copyright registrations.42 In this process, Novell declared under oath to the United States
Copyright Office that, "it retains all or substantially all of the ownership of the copyrights in
UNIX ..."43
39. On November 4, 2003, Novell announced its agreement to acquire SuSE Linux,
and emphasized that this acquisition "affirm[s] Novell's commitment to promoting the open
15
source model and developer community", and "will be an important step in Novell's efforts to
accelerate enterprise adoption of Linux."44 At the same time, Novell announced that IBM was
investing $50 million in Novell to help facilitate this purchase.45
40. SCO sent a second mailing to commercial Linux users on December 19, 2003
stating that "use of the Linux operating system in a commercial setting violates our rights under
the United States Copyright Act."46 SCO indicated that "No one may use our copyrighted code
except as authorized by us ... Once you have reviewed our position, we will be happy to further
discuss your options and work with you to remedy this problem."47
41. On December 22, 2003, Novell publicly reasserted its claim that SCO does not
own the UNIX copyrights, publicizing that Novell had also applied for and received UNIX
copyright registrations, and disclosing its correspondence with SCO on the issue purportedly to
show that "SCO has been well aware that Novell continues to assert ownership of the UNIX
copyrights."48
42. On January 13, 2004, in conjunction with closing its acquisition of Linux vendor
SuSE, Novell further asserted its copyright ownership claims by announcing that Novell would
be offering a new indemnification program for qualifying Novell Linux enterprise customers.
According to Novell, this new indemnification would provide "a measure of protection against
16
potential copyright infringement claims."49 Novell also reiterated that its copyright registrations
confirmed that it retained ownership of UNIX copyrights.50
43. Novell's statements attacking SCO's copyright ownership continued into the
following year. Qn March 16, 2004, Novell's Vice Chairman Chris Stone emphasized Novell's
position to an audience of potential SCOsource customers when he addressed SCO in his
keynote speech at the Open Source Business Conference, saying "You didn't invent Linux. Or
intellectual property law. We still own UNIX."51 This comment was also widely reported in the
IT press.
44. Novell continues to assert today that it owns the UNIX copyrights.
45. For Novell, the entity that had sold the UNIX copyrights to SCO, to publicly
reject SCQ's ownership claims was a virtual guarantee that customers would be reluctant to
purchase a license from SCO, particularly as the SCOsource licenses granted a right-to-use the
very IP now being claimed by Novell. There is a substantial body of economic literature on
intellectual property and markets for know-how that explains the impact of Novell's actions on
the SCOsource business.52 In broad terms, a firm possessing intellectual property can capture
economic rents through two basic business models: it can embed the intellectual property into
some end product which it sells directly to users (a product revenue model) or it can license the
intellectual property to other producers who, in turn, will embed it in end products of their own
17
(a licensing royalty model). With SCOsource, SCO was pursuing a licensing royalty model.
Economic theory of licensing and intellectual property is very clear on the point that the
licensing royalty model will only be viable if the licensor has well-defined and well-protected
intellectual property. Without such protection, the innovator faces a serious disclosure risk.
Disclosure risk refers to the risk that, once disclosed to would-be licensees and actual licensees,
the unprotected intellectual property will be copied and disseminated to non-licensees. The
potential for uncontrolled distribution, in turn, attenuates the incentive of would-be licensees to
pay for a license (since they can presumably acquire the IP for free).
46. In this case, it was impossible, ex-post, for SCO to restrict access to their
technology as users already had access to it in the form of Linux. Because would-be licensees of
SCO's intellectual property already had access to that property via Linux, their willingness to
pay for a license is a function of their belief that SCO owned the rights and would enforce those
rights. Because SCO had already alerted potential customers through letters and other public
statements about its intention to enforce its intellectual property rights, the market was aware of
SCO's willingness to enforce its rights. Novell's statements interfered with the potential
customers' willingness to pay by creating doubt regarding SCO's ownership rights and thus, its
ability to enforce those rights.
47. The following sections describe my calculations of the effect Novell's statements
had on SCO's market for licensing their UNIX intellectual property on Linux. Throughout my
analysis, I used highly conservative assumptions in order to present the lowest bound of the
damages likely suffered by SCO.
18
V. THE POTENTIAL MARKET FOR SCO'S SCOSOURCE LICENSE
48. In this section of my report, I estimate the number of SCOsource licenses that
SCO would have sold but for the actions by Novell at issue in this litigation. I calculate this
number based on the total number of Server Operating Environments that SCO contends infringe
its intellectual property, the demand among those users for protection from their infringement,
and the impact of Novell's statements on that demand. I also evaluate, absent Novell's
statements, SCO's ability to compete for these customers with other indemnification providers,
and whether these competitors would have been significant competition in the but for world.
A. Estimated Number of Potential Customers
49. I conservatively estimate the total potential market for the SCOsource Right to
Use license during the damages period at 7.4 million servers. This estimate is based on a
calculation limited to North American servers using versions of Linux with code that SCO
claims infringes its UNIX copyrights from 2001 to 2007.
50. I considered data regarding the number of Linux shipments starting in 2001
because the first release of Linux that SCO claims contained its intellectual property occurred in
January 2001.53 Accordingly, that install base (consumers with Linux 2.4 shipments beginning
in 2001) would be potential SCOsource customers at the onset of the damage period in 2003.
Moreover, I have included in my estimate only the Linux server operating environment, although
the infringing versions of Linux can be run in both client and server operating environments. A
'server operating environment is multi-user, multi-tasking, and multi-platform, designed primarily
for commercial businesses with the need to manage shared data and applications for multiple
19
users.54 A client operating environment generally is an operating system for a desktop,
supporting applications, database, and Web-browsing functionality.55 My understanding is that
SCO primarily targeted commercial users of Linux for SCOsource licenses, i.e. those using the
server operating environment.56 Accordingly, I focused exclusively on shipments of Linux
server operating environments from 2001 through 2007 in calculating the size of the potential
market for SCOsource licenses.
51. To estimate the total number of Linux SOE shipments (shipments of Linux for
use in a server operating environment, or SOE), I reviewed IDC Server Operating Environment
reports covering the years 1998 to 2007. IDC is an industry-leading provider of market
intelligence and advisory services for the technology sector and has the most comprehensive
practice for operating systems in general, and Linux specifically. They are generally relied upon
by IT professionals and business executives to make decisions regarding technology purchases
and business strategy, and I believe their estimates of market sizes and market and technology
trends are reliable. In each report, IDC can revise its estimates for earlier years, reflecting
improved data and analysis or changing taxonomies of earlier data over time. As such, my
standard methodology when using data from IDC reports is to use the data for any given year
from the most recent report covering that respective year.57
52. Linux server operating environment deployments fall into one of two categories:
revenue-generating shipments (i. e., customers who purchased their Linux operating system
20
through a distributor such as Red Hat or a hardware vendor, such as HP), and non-paid
deployments (i. e., customers who obtained a free version of Linux, possibly through a direct
download from the web). Both types of deployments include infringing code and therefore are
potential customers for a SCOsource RTU license.58
53. IDC reported SOE revenue shipments for each year from 1998 to 2007, but nonpaid shipment data was not available for all years. However, since 2000, IDC has estimated the number of nonpaid versus revenue-generating (paid) SOE shipments.59 IDC reported SOE revenue shipments for each year from 2001 to 2007, and non-paid shipment for the years 2002-2007. The table below shows IDC's reported and projected Linux SOE shipments (both revenue and non-paid) from 2001 through 2007.60
Table 1. Linux SOE Shipments Worldwide, 2001 - 2007 (in 000's)61 [ Source IDC, see PDF]
21
Accordingly, between 2001 and 2007, an estimated 15.0 million Linux server operating
environments shipped worldwide.
54. To more accurately define the size of the SCOsource license market, I made
several adjustments to the 15.0 million figure derived above from IDC's data. First, I adjusted
the 2001 data to account for shipments that may have been shipments of earlier versions of
Linux, which, for these purposes, SCO did not target with its RTU license. For this adjustment, I
estimated the percentage of shipments for 2001 reasonably attributable to Linux 2.2 and
subtracted them from the potential market.62 Second, it was necessary to subtract any Linux
SOE shipments that could be attributed to SCO, since SCO Linux customers were already
indemnified by SCO. This number is not significant, particularly as SCO stopped selling Linux
in May 2003.63 I calculated a ceiling for potential SCO Linux shipments based on the smallest
individual vendor numbers reported by IDC with the result of no more than 28,000 shipments,
and likely many fewer.64 I confirmed with SCO that the actual number of their shipments
probably was much lower than 28,000, but definitely was not more than that amount.65 Taking a
conservative approach, I subtracted the full 28,000 from the total number of shipments
representing potential license.
22
55. Finally, I further reduced the total Linux market to include only shipments for
North America.66 SCO made SCOsource available internationally on a country-by-country
basis.67 Linux use data is available by region, and North America is the only region with full
availability. While I fully expect that SCO would have generated SCOsource revenue outside
North America, limiting the data set to North America is the most conservative estimate possible.
The table below details the results of these adjustments.
Table 2. Linux SOE North American Shipments (Less 2001 Adjustment and SCO Linux Adjustment) (millions) [See PDF for table]
56. Based on the above analysis, I estimate that from 2001 to 2007, the total size of
the North American market for the SCO RTU license is approximately 7.4 million units.
B. Demand for Linux IP Indemnification
57. In my opinion, but for Novell's actions, SCO would have been able to sell
SCOsource Right to Use licenses to between 19% and 45% of the 7.4 million Linux deployments
described above. This opinion is substantiated by market research and my understanding of the
industry.
58. Indemnification is an important issue for technology users in general, and the
users of enterprise servers in particular. In the case of indemnification, a critical issue affecting
demand is a customer's risk tolerance. Enterprise server owners are highly risk averse with
respect to the potential costs and consequences of intellectual property litigation.
23
59. Enterprise servers run mission critical software and are expected to be highly
reliable, available and serviceable (an industry metric known as RAS). Commercial users place
substantial emphasis on minimizing the amount of risk associated with running a server, its
operating system and mission critical applications. There are two types of risk facing a user of
an operating system: operational risks and intellectual property risks. Operational risks have to
do with the operating system becoming unavailable to the user. Intellectual property risks have
to do with a user's potential liability if the operating system contains unlicensed intellectual
property. A user of an operating system that contains unlicensed code, for example, faces
multiple potential legal risks, including: the risk of time- and resource-consuming litigation; the
risk of disruption of service due to the need to replace infringing code; and the risk of damages
being assessed. Intellectual property risks may include operational risks as well, since
intellectual property infringement may result in a functional system becoming legally
unavailable to a user (because code needs to be replaced, or an injunction prohibits use, etc.).68
60. In order to minimize intellectual property risk, commercial users prefer to be
indemnified for their information technology. Accordingly, prior to 2001 and the introduction of
Open Source software (and Linux specifically) to the enterprise server market, it was standard
practice for operating system vendors to include indemnification to their users. Users of
enterprise servers paid licensing fees for operating systems and software vendors provided
indemnification that ensured any intellectual property liability would be borne by the vendor, not
24
the end-user. QpenServer, UnixWare, Windows, AIX, and Solaris all require license fees and all
provide indemnification to their users. 69
61. With IBM's support of Linux in 2000 and the launch of Linux 2.4 in January
2001, enterprise server customers increasingly adopted Linux. However, Linux was not
distributed with indemnification provisions or other protection from intellectual property risks.
This would not have been viewed as a significant risk until 2003, when SCO announced its
finding that Linux infringed SCO's copyrights. SCO's concerns and public position on its UNIX
rights alerted Linux users to their intellectual property risks. These risks naturally stimulated a
demand for some type of protection, either through indemnification or a license. For instance,
Novell's CEO at the time has testified that the question of copyrighted UNIX code in Linux "was
a perceived problem" in the market, and that "the market was spooked with regard to that - the
allegations that were being made by SCO ..."70
62. The effect of SCO's claims on the demand for protection was noted in the
contemporary media: "Intellectual-property protection of open-source software has moved to the
forefront in the computing industry as the result of matters such as the SCO Group's ongoing
attack on Linux."71 This trend remained strong into 2005, as did the demand for indemnification.
"Software companies increasingly are dangling the offer of intellectual-property liability
indemnification in front of customers concerned about protecting themselves against the threat of
25
lawsuits. Some may wonder whether the hubbub over software indemnification is just much ado
about nothing, but chief information officers know better." 72
63. Further, the market was aware of SCO's intent to protect its intellectual property
and to offer solutions to enable commercial users to use the infringing code in Linux. When
SCO initially introduced SCOsource in January 2003, there was heavy media coverage.73 The
Microsoft and Sun deals were signed in the first five months of SCOsource activity, and other
good opportunities were in the SCOsource sales pipeline. Jeff Hunsaker testified that, "[I]n
2003 ... That's when we introduced SCOsource. We had some good opportunities in building a
pipeline. And then the Novell issue came and, wham, it dropped almost in half. .."74
64. Novell's assertions that it owned the UNIX copyrights severely impacted
customers' demand for SCOsource licenses. As the public confidence in SCO's ownership
decreased, so did the demand for SCO's licenses. Qne of the barriers to a well-functioning
market for IP licensing is an inability to enforce the IP rights being sold.75 When SCO attempted
to sell the SCOsource license it found that the "Novell copyright dispute was continually thrown
back in our face."76 The dispute with Novell "caused customers to back away from any interest
26
in the IP agreements."77 Describing the impact of Novell's assertions on SCO's sales attempts,
Jeff Hunsaker testified that:
Well, it's caused tremendous consternation with all of our customers, and it's
caused a lot of concern. It's clearly had an impact on our day-to-day dealings
with our customers such that, you know, many of them have questioned us now.
"Well, there was never a question that you all owned the copyrights in UNIX and
UnixWare. You've been selling it to us for many years. Now all of the sudden
Novell comes out and states that they still retain the copyrights. So which is it?"
So it's had a huge impact. Not on a few of our customers; pretty much, I would
say, across the board. I mean, it has made an impact and caused a lot of
uncertainty, fear, doubt, and has cast a pretty grim shadow over some of the --
over our business.78
65. Just as the media had provided widespread coverage of SCOsource initiatives,
Novell's statements received extensive media attention, continuing into 200479 The National
Retail Federation, for example, issued a 2004 public statement claiming that Novell's actions
raised "serious questions regarding whether the SCO Group ever gained legal ownership to Unix
System V."80 The resulting confusion implicated every attempted SCOsource sale.81 "It's as if
someone came to Novell and said, 'By the way, you don't own NetWare anymore. We own the
copyrights for NetWare."82
66. By the time Chris Stone addressed the Open Source Business Conference in
March 2004, the issue of the UNIX IP ownership was pervasive, as he testified: "[T]his was a
big issue and it was discussed and there were questions in the audience and it was on everyone's
27
mind. I couldn't go anywhere without someone asked [sic] this question." 83This statement was
widely reported and repeated by the media.84
67. Industry press and analyst reports noted the futility of purchasing SCO's RTU
license given the uncertainty regarding SCO's copyright ownership. For example, after Novell's
statements claiming copyright ownership, potential customers shifted to a "maybe later" mode," delaying purchasing a license until there is a resolution of the copyright ownership issue. 85 The magic question is..., when should you start seriously looking for shelter [from SCO's legal action]? My answer? Probably on the day that a judge announces that he agrees with SCO's interpretation of what it acquired from Novell.86
68. My conclusion that market demand was impacted by Novell's statements is further supported by the fact that many of SCO's potential customers specifically referred to the dispute with Novell as a reason for not pursuing a license with SCO. Regal Entertainment, for example, specifically highlighted Novell's statements as a reason they declined to pursue further license negotiations with SCO.87
The repeated reasoning was "How can SCO sue enterprises using Linux for the misappropriation of its intellectual property if it hasn't established beyond a
shadow of a doubt that it owns the intellectual property in question?"88
69. Above, I estimated the total size of the Linux server operating environment in North America. SCO contends that all 7.4 million servers identified in that analysis were
28
running infringing versions of Linux. Thus, all 7.4 million servers in North America were
candidates for the SCOsource license. To identify the "but for" penetration of that market by
SCOsource, I considered a number of industry studies examining Linux users' preferences for
intellectual property protection. Since all of these studies were conducted after Novell's
statements, they reflect demand already diminished by Novell's claims to ownership of the
UNIX copyrights, and thus represent a minimum measure of the demand that could have been
realized in the absence of slander. Users willing to obtain indemnification were those most
concerned with the risks of IP litigation. This set of users would thus have been the most likely
purchasers of SCOsource Right to Use licenses. The level of interest in indemnification post-Novell's statement represents a lower-bound on the interest since Novell's statements. Novell's
statements cast SCO's ownership rights into question and would have reduced demand for
indemnification. There is no plausible economic rationale for Novell's statements to have
increased demand for indemnification. Thus, I use the level of demand post-action as an
absolute minimum on the likely demand prior to Novell's statements.
70. Several industry analyst reports attempted to gauge the level of demand for Linux
IP protection. In May 2004, a Forrester study estimated that 22% of respondents were interested
in buying Linux from a vendor who had a Linux indemnification program, with an additional
36% that were "unsure".89 While the 22% represents a lower bound of demand, in my opinion,
some fraction of the 36% that identified themselves as "unsure" would have been buyers of SCO's RTU license in the absence of uncertainty regarding SCO's ownership of the IP.
71. Six months later, a November 2004 Yankee Group Report showed that 19% of
respondents were highly concerned about Linux indemnification (8% considered Linux IP
29
indemnification a "top priority," while an additional 11% were "very concerned" about Linux
indemnification).90 An additional 26% indicated that they were "somewhat concerned" about
Linux indemnification.91 Again, it is worth stressing that these responses reflect post-slander
opinions clouded by uncertainty of SCO's intellectual property rights. Had the slander not
occurred and damaged perceptions about SCO's ownership of UNIX copyrights, it is very
possible that these 26% of respondents who listed themselves as still "somewhat concerned"
(despite the slander) would have been buyers of the SCOsource license. Including all 26% as
SCOsource customers (in addition to the 8% who considered IP protection a top priority and
11 % who identified themselves as very concerned) suggests an upper bound of the range for
demand at 45% of users.
72. Finally, in a March 2005 Yankee Group study of Linux users, approximately 20%
of respondents said they planned to purchase third-party indemnification to protect their Linux
assets.92, 93 This study did not identify any groups of users other than those who were planning to
purchase protection.
73. Again, because all these reports gauged demand after Novell had called SCO's
ownership of the UNIX copyrights into question, I believe that the lower range numbers are
lower than they would have been if SCO's ownership of the UNIX copyrights had not been
30
clouded and many of those on the fringe or undecided would be more interested and included as
part of the demand. 94
74. Considering all of the above evidence of market demand, I estimate that between
19% and 45% of the Linux market would purchase a SCOsource license in the absence of
uncertainty regarding SCO's ownership of UNIX copyrights. Applying the minimum 19% rate
to the Linux Market calculated earlier translates into 1.478 million lost SCOsource RTU
licenses. In the following table, I illustrate the lost SCOsource licenses under different
assumptions over the estimated penetration range of 19% to 45%:
Table 3 North American Linux SOE Shipments Available to SCO, and Potential
SCOsource Licenses, 2001-2007 (000's) [See PDF for Table]
75. Because the slander must have dampened expressed interest in indemnification
and because the ownership issue hindered SCO's ability to proceed with enforcement actions, it
is my opinion that absent Novell's statements, SCO probably would have experienced demand at
above the 19% threshold.
C. SCOsource Adoption Rate
76. But for Novell's statements attacking SCO's ownership of the UNIX copyrights, I
believe that SCO would have sold between 1,404,000 and 3,325,000 server licenses by 2007, as
reflected in Table 3 above.
77. As explained above, the relevant Linux SOE shipments span from 2001 to 2007.
Servers that SCO contends required a SCOsource license in order not to infringe SCO's rights
31
can be separated into two classes, depending on whether or not a SCOsource license was
available when they acquired their Linux SOE. I grouped potential servers by calendar year,
evaluating the group of servers from 2004 to 2007 (after SCOsource), and those acquired during
2001 to 2003 (the install base at the time of the SCOsource launch).
78. With respect to the first group, those that shipped from 2004 to 2007, I concluded
that, but for Novell's actions, the SCOsource sale would occur in the same year as the SOE
shipment. Corporate IT purchasers normally acquire licenses for operating systems, enterprise
servers, and other software at the time of purchase and installation. Accordingly, I believe that
these IT purchasers would have acquired SCOsource licenses at the time they purchased their
Linux SOE. In other words, SCO would earn the sale of a SCOsource license at the time of the
SOE shipment.
79. For the install base in place when SCO launched the SCOsource program, there
would be a different buying pattern since the SCOsource license was not necessarily available at
the time of their purchase.95 I believe that this group would complete their purchases of
SCOsource licenses by 2007. While I can't determine the exact timing of when each of these
users would have acquired the protection of a SCOsource license, I believe that users with high
levels of concern about mitigating the risks associated with IP infringement would have acquired
a license in the four years between the announcement of the program and the end of 2007.
D. Alternative Vendors Would Not Have Been a Viable Option Absent Novell's
Slander
80. Following Novell's attacks on SCO's UNIX copyright ownership rights, other
vendors began offering alternative forms of Linux indemnification to their customers. The
32
longer that the dispute between Novell and SCO endured, the more programs were initiated that
compete (to some extent) with SCOsource RTU licenses. Each vendor offers different coverage,
with varying prerequisites that customers must meet in order to be eligible for the
indemnification.
81. In my opinion, these alternative programs would not detract from SCO's
realization of 19% -45% of the Linux Market in the "but for" world. Absent Novell's
statements, SCO would have been the provider of choice for the customers calculated above,
while the other indemnification vendors who entered the market would not have been viable
options for two primary reasons: (1) the broader the protection offered by these programs, the
greater their potential liability once SCO's ownership rights are clear, and the less likely they are
to enter the market; (2) the narrower the protection offered by the alternative programs, the less
they offer a complete substitute for a SCOsource license, and the less likely they are to compete
with SCO. This section addresses both of these issues.
82. First, if SCO's ownership of the UNIX copyrights were settled, then the risk of
offering a program indemnifying purchasers from liability to SCO increases significantly. A
provider of indemnification, like any provider of insurance, essentially is taking a bet on an
uncertain liability. The indemnifier profits if the risky event in question does not occur. The
indemnifier's risks increase with the likelihood of the risky event in question. In this case, the
"risky event" is SCO being able to enforce its intellectual property rights. Economic reasoning
suggests that a provider of insurance (here, an indemnification) will not offer such insurance in
the absence of uncertainty. If the event in question has no probability of occurring, there will be
no demand for the indemnification. In contrast, if the event in question is certain to occur, there
will be no supply of indemnification. If SCO's property rights were certain, then potential
33
indemnifiers would not have found it profitable to offer indemnification. They would be, in
essence, taking on a certain liability. By making SCO's property rights uncertain, Novell's
actions decreased the perceived risks for a vendor of offering indemnification for its customers.
As a result, the number of companies that began offering indemnification increased. In the "but
for" world, where SCO's ownership is certain and risks of being found liable for infringement
are higher, these vendors would not have been likely to enter the market.
83. At the time SCO announced the SCOsource division and began offering licenses
for its IP in Linux, there were no providers of such protection for intellectual property in Linux.
As the dispute with Novell escalated, so did the number of alternative indemnification providers.
The following table captures the chronology of the alternative indemnification programs.
[See PDF for table]
34
84. By creating uncertainty regarding SCO's ownership rights, Novell reduced the
apparent risk of offering an indemnification program and created a market for indemnification
(as an alternative to the SCOsource RTU license). This significantly altered the competitive
landscape for SCO in selling its SCOsource licenses. For example, in the case of HP, there were
negotiations with SCO regarding providing indemnification to HP's customers under a bulk
license. This arrangement would have been beneficial to both companies. However, HP did not
conclude this Linux deal with SCO and eventually offered a form of indemnification to its
customers without any liability protection from SCO.
85. Second, where an indemnification program offers less than complete coverage
(e.g., caps on total damages, replacement risks not compensated, downtime not considered), the
program is less likely to compete directly with SCO. In order for an indemnification program to
be a true substitute for SCO's RTU license, it would need to indemnify the user against litigation
expenses, damages, and costs for replacing infringing code, including labor and downtime.
SCO's Right to Use license mitigates all of these risks, and as such, each of the alternative
sources of Linux indemnification offers protection that differs significantly from the security of
SCO's Right to Use license. I conducted a detailed analysis of each of the indemnification
programs available in the market and the scope of their protection. As shown below, each of
these programs entails significant limits on the protection offered and do not constitute viable
substitutes for SCOsource licenses.
35
i. HP's Indemnification Program
86. HP placed a number of restrictions on customers to qualify for the program that
substantially distinguish the program from SCO's RTU license. The HP restrictions include
prohibiting customers from using or modifying the source code (which restricts one of the primary benefits of Open Source software). Additionally, customers must purchase a Linux
distribution directly from HP or purchase HP copies of a Linux distribution from an HP
authorized reseller. Customers also have to run the Linux operating system solely on HP
hardware, thus substantially restricting the value of hardware independence associated with the
Intel architecture. Finally, customers are required to purchase standard or premium service
contracts for Linux support from HP; the least expensive qualifying contract is $1,431 per server
per year.96
87. Other than requiring customers to purchase software support from their hardware
vendor, the most significant HP restriction is the lock-in to HP's hardware. One of the biggest
values of running UNIX (or Linux) on the Intel platform is hardware independence. Reasonable
customers would prefer mitigating the risks associated with intellectual property in Linux while
still maintaining the freedom of choice in selecting hardware that previously was inherent in
Linux.
ii. Red Hat's Indemnification Program
88. Prior to Novell's December 22, 2003 statements claiming ownership of the UNIX
copyrights, Red Hat launched a limited program called Open Source Now to defray legal costs of
Linux developers. In January 2004, after Novell's release of all the communications with SCO
regarding UNIX copyright ownership, Red Hat launched its Open Source Assurance program.
36
The program allowed Red Hat, at its discretion, to modify any software found to be infringing, or
replace the infringing component with a non-infringing component if there is an IP issue. This
program did not indemnify users against lawsuits or damages. As Yankee Group states, "It
doesn't promise to compensate customers for downtime, incompatibilities or legal expenses in
the event of an IP dispute."97 Substantially later, in November 2006, Red Hat added a limited
indemnification clause to the Open Source Assurance program. However, the program still does
not provide any relief to its customers who may experience increased labor costs or downtime in
association with finding that part of their Linux code infringes protected intellectual property.
As such, although Red Hat's indemnification coverage has increased as Novell's statements
persisted, even Red Hat's new indemnity clause does not mitigate many of the risks addressed by
SCO's right-to-use license.
iii. Novell's IP Indemnification
89. Novell launched its Linux Indemnification Program on January 14, 2004, shortly
after Novell completed the acquisition of SuSE and re-publicized its copyright ownership claims.
Novell's indemnity program does not cover any SUSE licenses shipped prior to January 14,
2004. Additionally, Novell has considerable perquisites to qualifying for participation in the
program: customers must meet the $50,000 annual minimum purchase requirement for licenses,
upgrades, and updates within the year preceding a claim against their indemnified Linux product;
maintain current upgrade protection for the indemnified product; maintain a current Premium
Service Level contract; and register with the indemnification program within 10 days of
obtaining the covered product.
37
90. The most significant difference in Novell's indemnification program is that
Novell limits the indemnification to 125% of the fees from licenses, upgrades and updates, and
caps damages liability at $1.5 million.98 This amount also does not cover downtime potentially
associated with litigation or a finding of infringement. Novell's own documentation lists several
industries and the costs associated with an hour of downtime. They range from ATM fees at $12
- $17 thousand to brokerage firms at $5.6 - $7.3 million.99 In other words, some of the most
substantial risks from which users of enterprise servers most need protection are not covered by
Novell's indemnification program.
iv. Oracle's IP Indemnification
91. On October 25, 2006, Oracle announced that it would start distributing and
offering support services for Red Hat's version of Linux. Oracle's service agreement includes
indemnification for end-users and has no minimum purchase requirements or damage caps. It
does, however, have a termination provision that allows Oracle to terminate indemnification if
they are unable to modify the material to be non-infringing or acquire a license for the infringing
material in a commercially reasonable manner. In essence, Oracle will take over a customer's
legal defense and pay any damages awarded in court, but will not compensate for a customer's
losses related to downtime, replacement, etc. Moreover, if Oracle exercises their option to
terminate the indemnification, the customer is required to resolve the infringement without
Oracle's assistance.
92. Extremely limifed coverage and many restrictions minimize the value of the
competing indemnification programs; but for the uncertainty Novell created regarding SCO's
38
ownership rights SCO's RTU license would have been the most reasonable choice for the
customers seeking IP protection. The absence of any alternative vendor providing coverage that
would compare favorably with the assurance of a SCOsource RTU license is evidence of the
significant cost of providing full coverage (including downtime associated with changing
infringing code or licenses required as a result of ongoing use of infringing code). That cost
would be exponentially greater in a world where SCO's ownership ofthe UNIX copyrights is no
longer disputed. Moreover, the risk of less than complete coverage that consumers are willing to
bear in the status quo would be less bearable to risk averse consumers if there were no question
that SCO, and only SCO, had the right to license copyrighted UNIX code. The result would be
near-elimination of competition - either competitors would opt not to enter the indemnification
market at all, or their plans would appear grossly inadequate in comparison to the certainty of a
SCOsource RTU license.
E. The Future Market for SCOsource Licenses
93. It is important to note that today's market is not the market of the "but for" world.
It is my opinion that, at this point, even a court's determination that SCO owns the UNIX
copyrights probably would not restore SCO to the position they were in before Novell's
statements. As a practical matter, SCO's ability to sell SCOsource licenses in the future is highly
uncertain. Nearly four years have passed since Novell's first attack on SCO's copyright
ownership, and technology markets change rapidly.
94. Moreover, currently there are public doubts about SCO's viability as a
company.100 Due in part to extensive media coverage, potential customers may also be aware
39
that SCO's financial resources today are far more restricted than they were in 2003.101 Concerns
about SCO's viability support the public perception that SCO is unable to enforce its IP rights
due to lack of resources, funding, etc.102 As explained above, when potential customers view
SCO as unable to enforce its rights, they are substantially less likely to purchase SCOsource
licenses.
95. Finally, Microsoft recently has begun making claims that it has intellectual
property rights infringed by Linux.103 Microsoft now asserts that Linux infringes 235 Microsoft
patents, and has not foreclosed the possibility of seeking royalties from users.104 During 2003-
2007, the prospect of SCO's copyrighted code being in Linux appears to have been the primary
threat to their IP that concerned Linux users seeking indemnification. Virtually all press or
analyst articles discussing software intellectual property indemnification highlighted the
importance of SCO's copyright claims to Linux.105 The presence of Microsoft as an additional
IP concern would significantly alter the future market for SCOsource licenses, which do not
indemnify against all threats, but only provide the user a right-to-use SCO's IP in Linux.
96. In short, based on my expertise in high technology industries and my research and
analysis in this case, I have no reason to believe that a court decision vindicating SCO's
ownership rights would allow SCO to recover its losses through SCOsource sales at this time.
40
97. I declare under penalty of perjury that the foregoing is true and correct.
Date Executed: May 25, 2007
City, State: Cambridge, MA
[signature]
GARY PISANO
___________________________
1 In my opinion, "servers" are distinct from "workstations" and "desktops." See "Server Operating Environment
and Software Platforms: 1999 Worldwide Markets and Trends," by J. Bozeman, IDC, May 1999, ("A server
operating system that supports server functionality, including the provision of application, database and Web
services to end users. These operating environments are multi-user, multi-tasking, multi-platform (microprocessor),
and mu1tivendor... A client operating environment (COE) is an operating system that supports application, database,
and Web-browsing functionality on a desktop system..") (emphasis added);
See also "Servers Worldwide Vendor Segmentation: 2000," by J. Hewitt, Gartner, June 8, 2001, p. 195 ("A server is
a computer system that has been configured with one or more of the following features: a multi-user operating
system or network operating system; large memory capacity, usually 32MB minimum; highly expandable disk;
generally a minimum of eight expansion slots; input/output (I/O) channels; used to do back-end processing for a
significant portion of the time it is online on the network. Therefore, all supercomputers, mainframes, midrange and
entry-level systems are counted as servers. PCs that are not marketed as servers but are used as servers will not be
counted.").
2 "Competitive Strategy: Techniques for Analyzing Industries and Competitors" by Michael Porter, 1980; see also
"Co-opetition" by Adam Brandenburger and Barry Nalebuff, 1996 (Complementors play a critical role in industries
like computers and software and thus are relevant for my analysis ofSCO's competitive position).
3See e.g., "Anticipating the 21 st Century: Competition Policy in The New High-Tech, Global Marketplace, A Report
of the Federal Trade Commission Staff," May 1996; see also "American Bar Association, Handbook on the
Antitrust Aspects of Standards Setting," 2004.
4 "Unity in UNIX," The Banker, 1 November 1995 ("UNIX was the first, and for a long time the only operating
system not tied to specific hardware architecture.")...
5 "The Art of UNIX Programming - Chapter 2 - History: Origins and History of UNIX, 1969-1995" by Eric Steven
Raymond, 2003, http://www.faqs.orgidocs/artu/ch02s01.html ("In 1973 Thompson and Ritchie finally succeeded in
rewriting UNIX in their new language. This was quite an audacious move; at the time, system programming was
done in assembler in order to extract maximum performance from the hardware, and the very concept of a portable
operating system was barely a gleam in anyone's eye.").
See also "Computer Wars: How The West Can Win In the Post-IBM World" by Charles Ferguson and Charles
Morris, p. 103 ("UNIX was written in C, and could be compiled for, or ported to, any processor... By the early
1980s, almost a quarter of DEC VAXs were running under UNIX rather than VMS. Almost all university VAXs
used UNIX rather than VMS ... ").
6
"The Success of Open Source," by Steven Weber, 2004, p. 39 (noting that licensing fees for UNIX "skyrocketed"
to $100,000 in 1988, and as high as $250,000 a few years later).
7 SCO Company History. http://www.sco.com/company/history.html.
g"lO-K Annual Report, The Santa Cruz Operation, Inc. For the fiscal year ended, September 30, 1996," p. 3 and n.3.
9 "UNIX is Everything OS/2 Wants to Be and More" by Martin Marshall, InfoWorld, March 6, 1989 ("There are
already 2,000 applications running on Sun systems and nearly 3,000 applications on SCO Xenix/Merged UNIX.")
8 "10-K Annual Report, The Santa Cruz Operation, Inc. For the fiscal year ended, September 30, 1996," p. 3 and n.3.
9 "UNIX is Everything OS/2 Wants to Be and More" by Martin Marshall, InfoWorld, March 6, 1989 ("There are already 2,000 applications running on Sun systems and nearly 3,000 applications on SCO Xenix/Merged UNIX.")
10
In the 1995 acquisition, SCO allegedly did not buy the royalty cash streams associated with System V distribution
agreements (contracts with companies like Sun, HP, IBM, SGI, etc.). My understanding is that the Asset Purchase
Agreement outlines a process where SCO collects those royalties and transfers 95% of them to Novell. SCO retains
5% of royalty amounts in exchange for managing the collection, accounting and transfer of those royalties.
11 "Milestones in The History of the SCO Group," http://www.caldera.com/company/history.html. May 9, 2006.
12 Amendment No.2 at ¶ A.
13
"Taking a Byte Out of Storage," Lehman Brothers Global Equity Research, February 22, 2002, p. 62 ("The
problem with standardizing on a single hardware vendor is that it entails a long term commitment with high
switching costs.").
14 "SCO Enjoys Revival Of Fortunes," Computer Business Review, January 3, 1995 ("SCO has become a major
force in determining UNIX's future."); see also "Readies UNIX Upgrade," InformationWeek, March 24, 1997, p. 1
("Today, SCO is not only the largest provider of UNIX systems on Intel, but it's also a UNIX powerhouse."); id., p.
1 ("SCO has become a major force in determining UNIX's future.").
15"NT vs. Unix vs. Unix - Unix standard unlikely, vendors seek Intel-based market share" by Gates, Lana, Software
Magazine, October 1997, ("The Santa Cruz Operation (SCO) is leading the pack with 80% of the Unix server
market on Intel, and 35% of the overall Unix server market, but it's under attack from Sun, Hewlett-Packard, and
IBM, which each hold between 12% and 16% of the market with proprietary RISC chips."); UnixWare 7 - Winning
ticket? - Santa Cruz Operation's operating system for Intel-based servers - Product Announcement, by Bournellis,
Cynthia, Electronic News, March 16, 1998, ("A devoted Intel partner since its inception, the Santa Cruz Operation
(SCO) holds 80 percent of the Unix server market on Intel.")
16 "A very large, tectonic shift has been occurring in the server market where companies are shifting servers
formerly based on relatively expensive, proprietary hardware and UNIX to Intel-based Servers which are extremely
low-cost to purchase." Deutsche Bank, Global Equity Research, The SCO Group, Inc., "A call (option) to arms",
October 14, 2003.
17 "Market Trends for 1996 Server Industry," Gartner, October 13, 1997; "Server Computing Worldwide Market Segmentation: 1998," Gartner, May 10, 1999; "Servers Quarterly Statistics Worldwide -- Database," Gartner, February 21, 2006.
18 IDC Worldwide Quarterly Server Tracker, February 21, 2006.
19 "IBM to spend $1 billion on Linux in 2001", by Wilcox, Joe, CNet News, December 12, 2000, http://news.com.com/2100-1001-24975.html
20
I submitted an expert report in that case opining on IBM's actions and its effects on SCO.
21SCO's discovery about its copyrights was a process; the first copyright infringement (the libraries) was found in
early 2003; additionally copyright infringements were discovered throughout 2003.
22 April 18, 2007 Interview of Darl McBride.
23 Initially, a third type of license allowed access only to SCO's UNIX System Shared Libraries for use with Linux;
this product evolved into (and was subsumed by) the general RTU license. See Jan. 22, 2003 SCO Press Release.
SCON0055153.
24
March 7, 2003 SCO Press Release. SCON0055173.
25 Sun/SCO Software License Agreement, SCO1287208-1287221.
26 May 19, 2003
SCO Press Release. SCON0079463 (Exhibit 8 to Jan. 24, 2007 Blake Stowell Deposition in SCO
v. Novell).
27Over the course of several quarters, Sun paid a total of approximately $10 million, and Microsoft paid $16.75
million.
28 SCON0024112-113.
29 Id.
30 April 18, 2007 Interview of Darl McBride.
31 May 28, 2003 SCO Press Release (emphasis added). SCON0055196.
32 May 28, 2003 letter from Jack Messman to Darl McBride (emphasis added). SCON0024115-17.
33 June 6, 2003 Novell Press Release. SCON0024123.
34 June 6, 2003 letter from Joseph LaSala, Jr. to Darl McBride. SCON0024122. This letter, along with other letters
Novell sent to SCO between May 28, 2003 and December 22, 2003, were publicly disclosed by Novell on December
22, 2003.
35 June 26, 2003 letter from Joseph LaSala, Jr. to Darl McBride. SCON0024148. (Emphasis in original.)
36 April 18, 2007 interviews of Darl McBride and Chris Sontag.
37 July 21, 2003 SCO Press Release. (Exh. 12 to Jan. 24, 2007 Blake Stowell Deposition in SCO v. Novell.)
38 Id.
39 August 4, 2003 letter from Joseph LaSala, Jr. SCON0024152.
40 August 5, 2003 SCO Press Release. SCON0055211 (Exh. 13 to Jan. 24, 2007 Blake Stowell Deposition in SCO
v. Novell).
41 SCON0048531-38.
42 See NOV000043025-43049.
43 NOV000043028, NOV000043031, NOV000043034, NOV000043037, NOV000043039, NOV000043042,
NOV000043045, and NOV000043048 (each with the same quoted statement).
44 Nov. 4, 2003 Novell Press Release. SCOR0000509-11.
45 Id.
46 SCO1690946-48.
47 Id.
48 Dec. 22, 2003 Novell Press Release at
http://www.novell.com/news/press/novell statement on unix copyright registrations.
49 Jan. 13, 2004 Novell Press Release. NOV000030412-15.
50Id.
51 Deposition Exhibit No. 1010, Feb. 6, 2006 Christopher Stone Deposition in SCO v. Novell.
52 "The Market for Know-How and the Efficient International Transfer of Technology", by Teece, David J, Annals
of the American Academy of Political and Social Science, Vol. 458, November 1981, See also "Profiting from
technological innovation: Implications for integration, collaboration, licensing, and public policy," by Teece, David
J. Research Policy, Volume 15, Number 6, December 1986, p. 285-305; "Profiting from Innovation and the
Intellectual Property Revolution," by Pisano, Gary, Research Policy, Volume 35, Number 8, October 2006, 1122-1130.
53 SCO's Intellectual Property License for Linux "applies to commercial uses of a Linux operating system that
contains a 2.4 or later version of the kernel." See SCO Intellectual Property License for Linux,
http://www.sco.com/SCOsource/description.html. Linux 2.4 was released in January 2001; the subsequent kernel
release, Linux 2.6, became available in December 2003.
54 See "Server Operating Environment and Software Platforms: 1999 Worldwide Markets and Trends," by J.
Bozeman, IDC, May 1999.
55 Id.
56 SCO first targeted large companies like Fortune 1000, later opening up licensing to small and medium businesses
and international companies.
57 One exception to my standard methodology is when calculating growth rates across data from different reports. In
such circumstances, such as in my report for SCO v IBM, it is more appropriate to take growth rate data from a
single report.
58 IDC reports anecdotal information from Linux vendors that customers buy one shipment but deploy that shipment
up to ten times. "Worldwide Linux Operating Environments 2005-2009 Forecast and Analysis: Product Transitions
Continue", IDC # 34390, p. 28. In such cases, the paid shipment user and the non-paid shipment user are the same.
Those entities should have the same demand for mitigating IP infringement risk for all its Linux shipments.
59 For example, a 2001 IDC study described a survey conducted by IDC that indicated that non-paid shipments were
approximately 40% of the total number of Linux shipments for the year 2000. See IDC report #24492, Linux from
Two Perspectives: The Relationship Between Software and Hardware Shipments, p. 14 ("Based on data obtained
through the September 2000 Linux study, IDC found that, on average, 40% of the copies of Linux the survey
respondents had obtained had come from a direct download operation."), ("Free copies that saw deployment as an
SOE accounted for 40% of the total SOE new license shipments during CY2000.").
60 IDC did not report non-paid shipments for 2001. To calculate a non-paid shipment estimate for 2001, I employed
an IDC study describing a survey conducted by IDC that indicated non-paid shipments were 39.5% of the total
number of Linux shipments for the year 2000. This ratio is more conservative than ratios for 2002-2007, so I
employed the ratio from year 2000 to arrive at an estimate for 2001 non-paid shipments (1,123 paid shipments
divided by 60.5% to arrive at total paid and non-paid time 39.5% to arrive at 734,000 non-paid shipments. (See IDC
report #24492, Linux from Two Perspectives: The Relationship Between Software and Hardware Shipments, p. 14
("Based on data obtained through the September 2000 Linux study, IDC found that, on average, 40% of the copies
of Linux the survey respondents had obtained had come from a direct download operation."), ("Free copies that saw
deployment as an SOE accounted for 40% of the total SOE new license shipments during CY2000.").
61 Source: IDC Report #s 27969, 30159, 32452, US05WP001972, 34599, 34390, 205385. 2001-2006 are IDC
estimates while 2007 numbers are IDC forecasts.
62 Linux 2.4 was the first version for which SCO's Linux Right-To-Use license was available. Red Hat Linux 7.1,
the first version to include the 2.4 kernel, was globally available April 24th, 2001. (Red Hat press release, April 16th,
2001, "Red Hat Announces Red Hat Linux 7.1 with 2.4 kernel"). Red Hat is the largest Linux OS vendor. SuSE,
the second largest vendor, started selling their Linux kernel 2.4 compatible edition on February 12, 2001.
(http://lwn.net/2001/0201/a/suse-7.1.php3). Thus, to creae a conservative estimate I attributed four months' worth
(33.3%) of 2001 shipments to kernel 2.2 and subtracted these from the 2001 shipments. By April 28th, 2001 the
Linux kernel had already advanced to release 2.4.4 (http://www.kernel.org/pub/linux/kernel/v2.4/).
63 This understanding is based on an interview with Darl McBride, CEO of SCO, and review of SCO's Linux
shipment data.
64 I estimated a theoretical ceiling for SCO Linux shipments by calculating the number of shipments for smallest
Linux vendor whose shipments are tracked by IDC, Red Flag Software Company. SCO's shipments must be at that
level or lower, else IDC would be tracking SCO's Linux shipments. Therefore I subtracted the number of Red Flag
shipments for 2001, 2002, and 5/12ths of the shipments for 2003, as SCO stopped selling Linux in May 2003.
65 April 18, 2007 Interview of Darl McBride. In fact, it was approximately 10,000. Id.
66 To calculate the North American share of total shipments, I averaged the North American share of server
hardware shipments from 2002-2005 (from IDC Server Tracker), and the North American share of Linux operating
environment revenue. I then applied a linear trend line to the average for these four years, and used the slope of this
line to project the North American share forward for 2006-2007, and backward for 2001.
67 "SCO Announces Worldwide Availability of SCO Intellectual Property License," January 15, 2004 SCO Press
Release. SCON0055254.
68 See Research in Motion and Vonage as two recent examples where IP infringement risks potentially became or
might become operational risks.
69 "Indemnification Becomes Open Source's Nightmare and Microsoft's Blessing", by DiDio, Laura, November
2004, ("Corporations that use proprietary software, such as Microsoft Volume Licenses, Apple Macintosh and the
various flavors of UNIX (e.g., Sun Microsystems Solaris), get indemnification protection.") and ("Indemnification is
a big-ticket item and is included as a standard component in proprietary software licensing contracts.")
70 April 14, 2006 Deposition of Jack Messman in SCO v. IBM at 240, 79.
71 CNET News, "This week in Linux news," November 19, 2004.
72 CNET News, "Perspective: How to fight against patent terrorism," January 6, 2005.
73 This story was covered by the New York Times, Cnet, Forbes, the Wall Street Journal, and many others.
74 March 30, 2007 Deposition of Jeff Hunsaker ("Hunsaker Dep.") in SCO v. Novell at 161.
75 "The Market for Know-How and the Efficient [nternational Transfer of Technology", by Teece, David J, Annals
of the American Academy of Political and Social Science, Vol. 458, November 1981, See also "Profiting from
technological innovation: Implications for integration, collaboration, licensing, and public policy," by Teece, David
J. Research Policy, Volume 15, Number 6, December 1986, p. 285-305; "Profiting from Innovation and the
Intellectual Property Revolution," by Pisano, Gary, Research Policy, Volume 35, Number 8, October 2006, 1122-1130; Arora, A. and A. Gambardella, 2001, Markets for Technology, MIT Press.
76 SCO v. IBM, Ex. 319 (Tibbitts Dep.) at 112-113.
77 SCO v. IBM, Ex. 324 (Hughes Dep.) at 182-183. See also SCO v. IBM, Ex. 308 (Langer Dep.) at 127 (Novell's
"public pronouncements questioning SCO's ownership of UNIX's IP negatively impact[ed]" ability sell licenses).
78 Hunsaker Dep. at 15 (emphasis added).
79 Id. at 16-17 ("the press just has jumped all over those statements... ")
80"Retail trade group calls SCO's claims baseless," May 5, 2004, http://news.com.com/2100-7344 3-5206729.html,
as accessed on April 10, 2007.
81 Id. (resulting confusion created "a very formidable challenge for every one of our customers.")
82Id.
83 Feb. 6, 2007 Deposition of Chris Stone in SCO v. Novell at 66.
84 See e.g., "Novell Exec: Don't Fear Open Source, Embrace It," Matthew Hicks, eWeek.com, March 16, 2004; "Novell plugs open source, dings SCO", Stephen Shankland, March 16, 2004, CNET News.com.
85 Tech Update, David Berlind, "Is Red Hat the canary in SCO's coal mine?", February 18, 2004. ("For the most part, the buzz is that most Linux using enterprises are in a wait and see mode. That may change in the event that a judge sides with SCO on what it purchased from Novell.")
86 David Berlind, Tech Update. "The SCO legal train: Know your options," February 18, 2004.
http://techupdate.zdnet.com/techupdate/stories/main/SCO_legal_train.html, as accessed on March 9, 2007.
87 Nov. 5, 2004 Deposition of Philip Langer in SCO v. IBM at 140-141/ SCON0067408-9.
88 David Berlind, Tech Update, "The SCO legal train: Know your options," February 18, 2004.
http://techupdate.zdnet.com/techupdate/stories/main/SCO_legal_train.html, as accessed on March 9, 2007
89 "Linux IP Litigation: Users Largely Unconcerned About SCO Suit, Indifferent to Indemnifications," by Fichera,
Richard, Forrester Research, Inc., May 14,2004. This data is from a survey of 36 North American companies.
90 "Indemnification Becomes Open Source's Nightmare and Microsoft's Blessing", by DiDio, Laura, The Yankee
Group, November 2004, p. 3. (Joint Yankee Group / Sunbelt Software, Inc. survey of 1,000 organizations.).
91 Id.
92 "2005 North American Linux and Windows TCO Comparison, Part I" by Didio, Laura, The Yankee Group, April
2005.
93 Because two of the three reports on the market demand for indemnification specifically mention that the survey
respondents were large companies (5,000+ employees for one and $1 billion in revenue for the other), I considered
whether the size of a company was likely to affect its demand for indemnification. A Gartner report ("User Survey
Report: Open Source and Linux Software Support Service, North America, 2006", July 24, 2006, p. 22) asked
customers the importance of open source contract components and broke the results down by company size. All size
groupings responded in a very tight range, with smaller companies (defined as having 1-99 employees) placing only
slightly less importance on indemnification.
94 Potential customers probably believed that if Novell were found to be the owner, they had no need to worry about
the copyrighted code in question as Novell had publicly indicated that believed Linux should remain free and
assured that they would not sue end-users.
95 I recognize that 2003 will contain some purchasers for whom the SCOsource license was available at the time of
their SCO purchase; however, this fact does not impact my conclusion as these purchasers would simply follow the
buying pattern described for the 2004-2007 group of customers.
96 Interviews of various HP sales personnel.
97 "Indemnification Becomes Open Source's Nightmare and Microsoft's Blessing," by DiDio, Laura, The Yankee
Group, November 2004.
98 Novell Linux Indemnification Program Flyer, hnp://www.novell.com/licensinglindemnitv/, as accessed on April
5, 2007.
99 Novell Premium Services, Services Brochure, p. 7.
100 See, e.g. Declaration of Darl McBride, May 18, 2007, ¶11 (noting public questions about SCO's viability and the
fact that these questions were not as prominent in 2003).
101 Id. at ¶13 (recognizing that, overall, SCO's finances are much more limited now than in 2003).
102 Id. at ¶¶1O-13.
103 See "Microsoft Takes on the Free World", Roger Parloff, Fortune, May 14, 2007.
104
Id.
105 IDC #32467, "Microsoft Augments Intellectual Property Indemnification," December 2004. ("In the wake of
the SCO-IBM intellectual property (IP) dispute, Microsoft's new IP indemnification policy effectively raises the
stakes for protecting software customers caught in the crossfire of vendor disputes."). I examined court cases to see
if there were any other intellectual property claims launched against Linux or any of its distributors and did not find
any. I also searched trade press and analyst reports for any other intellectual property holders making claims against
Linux users.
41
[EXHIBIT A is 9 pages of curriculum vitae. See PDF.]
EXHIBIT B
DOCUMENTS CONSIDERED
A. Discovery Documents
SC01750101.
SCO 1287208-1287221. Sun-SCO Software License Agreement.
SC01269005 -- 18. Sun-SCO Software License Agreement draft.
SC01605380 --93. Sun-SCO Software License Agreement, Clarification of License
Grant to UnixWare and QpenServer Drivers.
SCON0048531 -- 8. Intellectual property compliance license with Computer Associates
International, for SCO Unix rights.
SCON0048531 -- 38. Intellectual property compliance license with Computer Associates
International, for SCO Unix rights.
SC01300033 -- 46. Release, License and Option Agreement between Microsoft
Corporation and The SCO Group.
SCON0023874 -- 8. Release of Claims and Agreement between Hewlett Packard and
The SCO Group.
SCOI604204-8. Release of Unix Claims and Agreement between Hewlett Packard and
The SCO Group.
SCO1448557-63. Intellectual Property Compliance License for Linux between The SCO
Group and Questar, Inc.
SCON0024112-113. Communication between Darl McBride and Novell Inc. from 2002
to March 15, 2006.
SCON0024115-17. May 28, 2003 letter from Jack Messman to Darl McBride.
SCON0024122. June 6, 2003 letter from Joseph LaSala, Jr. to Darl McBride.
SCON0024148. June 26, 2003 letter from Joseph LaSala, Jr. to Darl McBride.
SCON0024152. August 4, 2003 letter from Joseph LaSala, Jr. to Darl McBride.
SCON0055153. SCO Press Release, January 22nd, 2003 "SCO Establishes SCOsource to
License Unix Intellectual Property."
SCON0055173. SCO Press Release, March 7th, 2003 "SCO Files Lawsuit Against IBM."
SCON0055189. SCO Press Release May 14th, 2003. "The SCO Group Signs Second
SCOsource License Agreement; Expects Net Income of $0.29 per Share for Second-
Quarter Fiscal 2003."
SCON0055l91. SCO Press Release May 14th, 2003. "SCO Suspends Distribution of
Linux Pending Intellectual Property Clarification; Announces Greater Focus on UNIX
and SCOx Strategy."
SCON0079463. SCO Press Release May 19th, 2003 "SCO Announces UNIX Licensing
Deal With Microsoft."
SCON0055196. SCO Press Release May 28th , 2003 "The SCO Group Reports Record
Quarterly Net Income."
SCON0055200. SCO Press Release June 6th, 2003. "SCO Confirms Copyright
Qwnership of Unix and UnixWare."
SCON0055204. SCO Press Release, July 21st, 2003 "SCO Registers UNIX Copyrights
and Offers UNIX License."
SCON0055211. SCO Press Release August 5th, 2003 "SCO Announces Intellectual
Property License for Linux."
SCON0077889. SCO Press Release, August 11 th, 2003. "The SCO Group Announces
Signing of Intellectual Property Compliance License for SCO Unix Rights."
SCON0055254. SCO Press Release, January 15th, 2004, "SCO Announces Worldwide
Availability of SCO Intellectual Property License."
SCON0039886-95. Powerpoint presentation SCOsource Announcement January 2003.
SCON0041017-9. SCOsource Q & A.
SCON0061292. SCOsource Licensing price list.
SCO 1616531. SCOsource Income Statement.
SC01778206. SCOsource pipeline projection for third quarter of 2004.
SCON0048221-3. Email from Jeff Hunsaker, April 1, 2003.
SCON0048495-6. Email from Jeff Hunsaker, June 19, 2003.
SCON0048488-9. Memorandum from Robert K. Bench and Mike Olsen, February 26,
2004.
SCON0050461-2. Email from Jeff Hunsaker, July 2, 2003.
SCON0040371-4. Operations Review Meeting notes, October 20, 2003.
SCON0043679-82. Confidence in SCO Unix Initiative Draft.
SC01534282-305. SCO News of the Day, March 2, 2004.
NOV000043025-43049. Novell Unix Copyright Registrations.
NOV000043028, NOV000043031, NOV000043034, NOV000043037, NOV000043039,
NOV000043042, NOV000043045, and NOV000043048.
NOV 000043055. Communication referencing Novell Press Release, May 28, 2003
"Novell Challenges SCO Position, Reiterates Support for Linux."
SCON0024123. Novell Press Release June 6, 2003 "Novell Statement on SCO Contract
Amendment."
SCOR0000509-11. Novell Press Release November 4, 2003 "Novell Announces
Agreement to Acquire Leading Enterprise Linux Technology Company SUSE LINUX."
NOV 000030412-15. Novell Press Release January 13, 2004 "Novell Completes
Acquisition of SUSE LINUX."
NOV 000030188-91. Novell Press Release, January 13, 2004. "Novell Supports
Enterprise Linux Customers with New Linux Indemnification."
NOV 000016694-721. Project Solstice Strategic Discussion, September 3, 2003.
SCONROOOOOO1-0000025.
SC01463062. IDC, "The Next Shoe Prepares to Drop: The SCO Group Preps for End-
User Litigation," IDC #30834, February 2004.
SCONOl14012-17. Letter from Merrill Lynch to SCO declining purchase of an IP
license, May 21, 2004.
SCON0054958-68. SCO July 21, 2003 Conference Call Transcript.
SCON0054970-78. SCO August 5, 2003 Conference Call Transcript.
SCON0041 046-57. SCO August 14, 2003 Conference Call Transcript.
SCON0054993-5001. SCO October 17,2003 Conference Call Transcript.
SCON0055003-14. SCO November 18, 2003 Conference Call Transcript.
SCON0041058-70. SCO December 22, 2003 Conference Call Transcript.
SCON0055030-46. SCO March 3, 2004 Conference Call Transcript.
NOV 000012533 - 48. SCO June 10, 2004 Conference Call Transcript.
.B. Depositions and Declarations
Deposition of Maureen O'Gara (March 23, 2007) in SCO v. Novell.
Deposition of Jack Messman (February 7, 2007) in SCO v. Novell.
Deposition of Chris Stone (February 6, 2007) in SCO v. Novell, including Exhibit 1010.
Deposition of Joseph LaSala (February 8, 2007) in SCO v. Novell.
Deposition of Blake Stowell (January 24, 2007) in SCO v. Novell, Exhibit 12.
Deposition of Christopher Sontag (March 16, 2006) in SCO v. IBM. SCON0037425-48.
Deposition of Greg Jones (January 26, 2007) in SCO v. Novell.
Deposition of Jack Messman (April 14, 2006) in SCO v. IBM.
Deposition of Jeff Hunsaker (March 30, 2007) in SCO v. Novell.
Deposition of Jeff Hunsaker (February 28, 2006) in SCO v. IBM. SCON0037926-8043.
Deposition of Ryan Tibbitts in SCO v. IBM, Exhibit 319.
Deposition of Philip Langer (November 5, 2004) in SCO v. IBM. Exhibit 308.
SCQN0067408-9.
Deposition of Larry Gasparro in SCO v. IBM. SCON0038167-205.
Deposition of Erik Hughes in SCO v. IBM, Ex. 324.
Deposition of William M. Broderick in SCO v. IBM. SCON0038985-6.
Deposition of Michael Paul Qlsen in SCO v. IBM. SCON0038502-11.
Declaration of Christopher Sontag, December 11th, 2006.
C.Books and Articles
Arora, A., Fosfuri, A., and Gambardella, A., Markets for Technology. MIT Press, 2001.
American Bar Association, "Handbook on the Antitrust Aspects of Standards Setting,"
2004.
Beale, Matthew. "IBM Moves on Linux," E-Commerce Times, May 26, 1999.
http://www.ecommercetimes.com/story/255.html.
Berlind, David, "Is Red Hat the canary in SCO's coal mine?", Tech Update, February 18,
2004.
Berlind, David, "The SCO legal train: Know your options," Tech Update, February 18,
2004. http://techupdate.zdnet.com/techupdate/stories/main/SCO_legal_train.html, as
accessed on March 9, 2007.
Berlind, David. "Defense funds: Taste great, but less filling for IT managers," Tech
Update, February 18,2004.
Berlind, David. "HP's Protection: SCO-only, but no dollar limit," Tech Update,
February 18, 2004.
Berlind, David. "Novell's protection: Covers more than SCO, caps damages, targets
enterprises," Tech Update, February 18, 2004.
Berlind, David. "Instead of indemnification, consider 'open source insurance'," Tech
Update, February 18, 2004.
Berlind, David. "What did SCO buy - Unix or the Brooklyn Bridge?" Tech Update,
February 19, 2004.
Berlind, David. "Microsoft Indemnifies. But does it really matter to you?", Tech
Update, November 10, 2004.
http://techupdate.zdnet.com/techupdate/stories/main/SCO_legal_train.html, as accessed
April 18, 2007.
Berlind, David. "Novell clarifies $40M payment to Microsoft as 'peace of
mind',"
ZDNet.com, November 22, 2006.
Bickel, 8., "JBoss Indemnification," The JBoss Matrix, as accessed at
http://blogs.jboss.comlblog/bbickel, March 1,2007.
Boumellis, C., "UnixWare 7 - Winning ticket? - Santa Cruz Operation's operating system
for Intel-based servers - Product Announcement," Electronic News, March 16, 1998.
Brandenburger, A., and Nalebuff, 8., Co-opetition. Doubleday / Currency, 1996.
Byfield, B., "EULAs, indemnification, and user protection," Linux.com, May 12, 2005,
as accessed from http://enterprise.linux.com, March 14, 2007.
CNET News, "This week in Linux news," November 19, 2004.
CNET News, "Perspective: How to fight against patent terrorism," January 6, 2005.
CNET News, "IBM to spend $1 billion on Linux in 2001,", December 12, 2000,
http://news.com.com/2100-1001-249750.html.
CNET News, "Ballmer repeats threats against Linux," February 20, 2007.
http://news.com.com/Ballmer+repeats+threats+against+Linux/21 00-7344_3-
6160604.html, as accessed on April 4, 2007.
CNET News, "Retail trade group calls SCO's claims baseless," May 5, 2004,
http://news.com.com/2100-7344_3-5206729.html, as accessed on April 10, 2007.
CNET News, "Experts say Microsoft's patent quest won't go far," May 16, 2007.
CNET News, "HP outlines Linux indemnity plan," Fried, I., September 24, 2003.
CNET News, "Red Hat scoops up Jboss," LaMonica, M., April 10, 2006.
CNET News, "Start-up launches Linux legal protection," Shankland, S., May 7, 2004.
CNET News, "Seeking certainty, CEO signs SCO Linux license," March 1, 2004.
CNET News, "Novell plugs open source, dings SCO," Shankland, S., March 16, 2004.
CNET News, "Microsoft agitates for open-source patent pacts," May 14, 2007.
Computer Business Review, "SCO Enjoys Revival of Fortunes," January 3, 1995.
Cushing, K., "Users left in the dark as SCO remains tight-lipped about evidence for
Linux lawsuits," ComputerWeekly.com, August 4, 2003.
Darrow, B. "Oracle to Support, Indemnify Red Hat Linux," CRN.com, October 25, 2006.
Deutsche Bank, Global Equity Research, "The SCO Group, Inc. A call (option) to arms,"
October 14, 2003.
Deutsche Bank, Global Equity Research, "The SCO Group, Inc A call (option) to arms:
appendices" October 14, 2003.
DiDio, Laura, "Indemnification Becomes Open Source's Nightmare and Microsoft's
Blessing," The Yankee Group, November 2004.
DiDio, Laura, "HP Linux Initiatives Get Game," The Yankee Group, October 6, 2005.
Didio, Laura, "2005 North American Linux and Windows TCO Comparison, Part 1," The
Yankee Group, April 2005.
Evers, J., "Do software users need indemnification?," Network World, November 29,
2004.
Ferguson, C. and Morris, C., "Computer Wars: How The West Can Win In the Post-IBM
World," New York Times Books, 1993.
Fichera, Richard, "Linux IP Litigation: Users Largely Unconcerned About SCO Suit,
Indifferent to Indemnifications," Forrester Research, Inc., May 14, 2004.
Fichera, Richard, "Is SCO Toast? All Signs Point to Yes," Forrester Research, Inc.
January 12, 2005.
Franklin, M. T., Alexander, W. P., Jauhari, R., Maynard, A. M. G., and Olszewski, B. R.
"Commercial workload preformance in the IBM POWER2 RISC System/6000
processor." IBM Journal of Research and Development, Vol. 38 No.5, September 1994.
FTC. "Anticipating the 21st Century: Competition Policy in The New High-Tech, Global
Marketplace, A Report ofthe Federal Trade Commission Staff," May, 1996.
Galli, P., "HP to Protect Customers From Linux Claims," eSeminars, September 24,
2003.
Gartner, "Market Trends for 1996 Server Industry," October 13, 1997.
Gartner, "Server Computing Worldwide Market Segmentation: 1998," May 10, 1999.
Gartner, "Servers Quarterly Statistics Worldwide -- Database, February 21, 2006.
Gartner, "Servers Worldwide Vendor Segmentation: 2000," June 8, 2001.
Gartner, "SCO's Lawsuit Sends a Warning to Linux IS Shops," Weiss, G., April 16,
2003.
Gartner, "Management Update: The Effects of the SCO Lawsuit on IBM AIX
Customers," Weiss, G., July 2, 2003.
Gartner, "What to Do During SCO's Protracted Lawsuit Against Linux," Weiss, G.,
March 23, 2004.
Gartner, "SCO's Suit Against IBM: An Update and Recommendations," Weiss, G., April
27, 2005.
Gartner, "SCO's Threat to Sue Linux Users Serious but Remote," Weiss, G. May 21,
2003.
Gartner, "Microsoft Plan to License Unix Won't Affect SCO's Lawsuit," Weiss, G., May
27, 2003.
Gartner, "SCO License Fees Would Hurt Linux Market," Weiss, G., July 23, 2003.
Gartner, "SCO's Legal Fees Could Jeopardize Its Software Business," Weiss, G.
November 19, 2003.
Gartner, "Novell Aims at Red Hat With SUSE Linux Indemnification Program," Weiss,
G., Enck, J., and Butler, A., January 16, 2004.
Gartner, "New Process for Linux Submissions Doesn't End Your Risks," Weiss, G., May
26, 2004.
Gartner, "Novell Embraces Open Source and Linux," Weiss, G., Enck, J., July 29, 2004.
Gartner, "OIN Will Promote Linux Innovation, but Raises Issues," Weiss, G., Enck, J.,
November 16, 2005.
Gartner, "Oracle's 'Unbreakable Linux' Will Threaten Red Hat," Weiss, G., Feinberg,
D., and Prentice, B., October 30, 2006.
Gates, L., "NT vs. Unix vs. Unix - Unix standard unlikely, vendors seek Intel-based
market share" Software Magazine, October, 1997.
Gerstner, L. J., Who Says Elephants Can't Dance. Harper Business, 2002.
Hewlett-Packard, "HP Care Pack Services," as accessed at www.hp.com.
Hewlett-Packard, "HP Indemnifies Customers Against SCO Lawsuit, Frequently Asked
Questions, June 2004" as accessed at www.hp.com.
Hicks, M., "Novell Exec: Don't fear open source, embrace it," eWeek.com, March 16,
2004.
IDC "Worldwide Quarterly Server Tracker," February 21, 2006
IDC, "Intellectual Property Protection for Software: Don't Assume You're Covered!"
IDC Analyst Connection, February 2005.
IDC, "Linux from Two Perspectives: The Relationship Between Software and Hardware
Shipments," IDC #24492.
IDC, "Worldwide Client and Server Operating Environments Forecast and Analysis,
2002-2006: Microsoft Extends Its Grip on the Market," IDC #27969, September 2002.
IDC, "Worldwide Client and Server Operating Environments Forecast and Analysis,
2002-2007: Microsoft Growth Sets Stage for SOE Dominance," IDC #30159, September
2003.
IDC, "Worldwide Client and Server Operating Environments 2004-2008 Forecast:
Microsoft Consolidates Its Grip," IDC #32452, December 2004.
IDC, "Microsoft Augments Intellectual Property Indemnification," IDC #32467,
December 2004.
IDC, "Worldwide Linux Operating Environments 2005-2009 Forecast and Analysis:
Product Transitions Continue," IDC #34390, December 2005.
IDC, "White Paper: Why Linux is Good for ISVs," IDC #US05WP001972, August 2005.
IDC, "Worldwide Client and Server Operating Environments 2005-2009 Forecast and
Analysis: Modest Growth Ahead, IDC #34599, December 2005.
IDC, "Worldwide Linux Operating Environments 2007-2010 Forecast and Analysis:
Linux Establishes Itself Aboard Client and Server," IDC #205385, February 2007.
Lehman Brothers Global Equity Research, "Taking a Byte Out of Storage," February 22,
2002.
Lyman, J., "HP's Linux Sales reach $2.5 Billion 2003," LinuxInsider, January 16, 2004.
Lyons, D. "IBM Refuses To Indemnify Linux Users," Forbes.com, August 5, 2003.
Lyons, D. "Holding Up Hollywood," Forbes, November 24, 2003.
Marshall, M., "UNIX is Everything OS/2 Wants to be and More," Info World, March 6,
1989.
McMillan, R., "Novell to indemnify SUSE customers," Computerworld, January 13,
2004.
McMillan, R., "SCO to take Linux licensing to SMB, international users, Some advise
waiting for resolution of legal issues before paying license fees," IDG News Service,
January 14, 2003.
McMillan, M., "HP to Indemnify Linux clients against SCO," IDG New Service,
September 24, 2003.
Microsoft, "Microsoft Intellectual Property Indemnification," November 10, 2004, as
accessed from www.microsoft.com. March 1, 2007.
NewsFactor Network, "Sun Revealed as SCO's Secret Licensee," July 11, 2003,
http://www.newsfactor.com/perl/story/21894.html.
Novell. "Novell Linux Indemnification Program Flyer,"
http://www.novell.com/licensinglindemnity/ as accessed on April S, 2007.
Novell Press Release December 22, 2003 "Novell Statement on UNIX Copyright
Registrations,"
http://www.novell.com/news/press/
novell_statement_on_unixIcopyright_ registrations.
Novell Linux Indemnification Program Addendum, Version 1.1, as downloaded from
www.novell.com.
Novell Corporate Price List, Product Availability and Price List, February 15, 2007. As
downloaded from www.novell.com.
Olavsrud, T., "HP to Indemnify Linux Customers," InternetNews.com, September 24,
2003.
The Open Group, "The Single UNIX Specification History & Timeline," 2004.
Oracle, "Oracle Enterprise Linux Services Agreement," as accessed from
www.oracle.com.
Oracle, "Top Five Facts About Oracle's Indemnification for Linux," as accessed from
www.oracle.com.
Oracle, "Oracle Unbreakable Linux FAQ," as accessed from www.oracle.com. February
15, 2007.
Oracle, "Oracle Unbreakable Linux," (PowerPoint Presentation), as downloaded from
www.oracle.com.
OSRM, Press Release, March 17, 2004, "Open Source Risk Management Unveils
Industry's First Vendor-Neutral Risk Mitigation Services," as accessed from
www.osriskmanagement.com.
OSRM, Press Release August 2, 2004, "Results of First-Ever Linux Patent Review
Announced, Patent Insurance Offered by Open Source Risk Management."
OSRM Press Release, December 2005, "Kiln and Miller launch Open Source compliance
insurance."
OSRM, Product Description Brochure, as downloaded from
www.osriskmanagement.com.
OSRM, "Global Collaborative Software and Development and Deployment in the Era of
Open Source: The Underwriters' Role," Hogg, M. R. and Egger, D., Enterprise Open
Source Journal, MarchiApril 2006.
OSRM, "Open Source Software IP Risk Audits: The Emerging Due Diligence Standard
For Technology M&A Transactions," Hogg, M. and Egger, D., Financier Worldwide,
Issue 44, August 2006.
Pisano, Gary. "Profiting from Innovation and the Intellectual Property Revolution,"
Research Policy, Volume 35, Number 8, October 2006, 1122-1130.
Porter, M. E. Competitive Strategy: Techniques for Analyzing Industries and
Competitors. Free Press / Simon & Schuster, Inc., N.Y. 1980.
Raymond, E. S., "The Art of UNIX Programming - Chapter 2 - History: Origins and
History of UNIX, 1969-1995," 2003.
Red Hat Press Release, April 16th, 2001, "Red Hat Announces Red Hat Linux 7.1 with
2.4 kernel."
http://www.redhat.com/about/presscenter/
2001lpresssevenone.html.as
accessed on May 14, 2007.
Red Hat Press Release, January 20, 2004, "Red Hat Announces Open Source Assurance
to Safeguard Customer Investment."
Red Hat, "Subscription Agreement," as accessed from www.redhat.com. March 1, 2007.
Red Hat, "Open Source Assurance FAQ," as accessed from www.redhat.com. March 1,
2007.
Red Hat, "Service Agreements: Service Level Agreements," as accessed from
www.redhat.com. March 1, 2007.
Renaissance Research Group, Special Situations Research, "The SCO Group, Inc."
Analyst Coverage, February 24, 2003.
Renaissance Research Group, Special Situations Research, "The SCO Group, Inc."
Analyst Coverage, March 6, 2003.
Renaissance Research Group, Special Situations Research, "The SCO Group, Inc."
Analyst Coverage, April 21, 2003.
Rosenbaum, Joseph. "Protect Thyself 101: A primer on indemnification," Tech Update,
February 18, 2004.
Salus, P., "UNIX at 25," BYTE, October 1994.
SCO Intellectual Property License for Linux,
http://www.sco.com/SCOsource/description.html.
SCO, "Milestones in The History of The SCO Group,"
http://www.sco.com/company/history.html. accessed August 9, 2006.
SCO Press Release, May 22, 2003. "The SCO Group Invites You to Join Its Second-
Quarter 2003 Results Conference Call,"
http://ir.sco.com/releasedetail.cfm?ReleaseID=109747.
Self, Karsten M. "SCO Teleconference Transcript," May 30, 2003.
http://zgp.org/pipermail/linux-elitists/2003-June/006387.html, as accessed on
February 15, 2007.
Sun Microsystems, "Solaris Subscriptions: What level of support is right for you?" as
accessed from www.sun.com. March 1, 2007.
SUSE press release January 25th, 2001, "SUSE Linux 7.1 pending release February 12,"
http://lwn.net/200l/020l/a/suse-7.1.php3.
Talmor, S., "Unity in UNIX," The Banker, November 1, 1995.
Teece, David J., "The Market for Know-How and the Efficient International Transfer of
Technology," Annals ofthe American Academy of Political and Social Science, Vol. 458,
November 1981.
Teece, David J., "Profiting from technological innovation: Implications for integration,
collaboration, licensing, and public policy," Research Policy, Volume 15, Number 6,
December 1986, p. 285-305.
Weber, S. "The Success of Open Source," Cambridge, MA: Harvard University Press,
2004.
Weiss, T., "Linux Users Not Worried About SCO Threats," Computerworld, July 23,
2003.
Weiss, T., "LinuxWorld: A defiant IBM says Linux indemnification is unnecessary,"
Computerworld, January 21, 2004.
ZDNet, "Critics pour scorn on Ballmer patent 'FUD' ," ZDNet.co.uk, November 18,
2004.
http://news.zdnet.co.uk/software/
0,1000000121,39174372,00.htm, as accessed on
April 4, 2007.
Zellen, Barry, "Hell No, We Won't SCO: Linux Users Indifferent To Indemnification,"
Enterprise Innovator, May 14, 2004.
"Index of /pub/linux/kernel/v2.4," http://www.kernel.org/
publIinux/kernel/v2.4/
D. Financial Statements
10-K Annual Reports, The Santa Cruz Operation, Inc. for the fiscal years 1996 - 2005.
1O-K Annual Reports, Red Hat Inc., for the fiscal years 2000-2005.
10-K Annual Reports, Novell Inc., for the fiscal years 2000- 2005.
10-K Annual Reports, Hewlett Packard Company, for the fiscal years 2002-2006.
****************************************
****************************************
Novell's Memorandum in Support of Motion for Daubert Hearing
[ back to top ] [ Pisano's Expert Report ]
[ Novell's Memorandum ]
[ Pisano Deposition ]
WORKMAN | NYDEGGER A PROFESSIONAL CORPORATION
Sterling A. Brennan (Utah State Bar No. 10060; [E-mail])
David R. Wright (Utah State Bar No. 5164: [E-mail])
Kirk R. Harris (Utah State Bar No. 10221; [E-mail])
Cara J. Baldwin (Utah State Bar No. 11863; [E-mail])
[address, phone, fax]
MORRISON & FOERSTER LLP
Michael A. Jacobs (Admitted Pro Hac Vice; [E-mail])
Eric M. Acker (Admitted Pro Hac Vice; [E-mail])
Grant L. Kim (Admitted Pro Hac Vice; [E-mail])
[address, phone, fax]
Attorneys for Defendant and Counterclaim-Plaintiff Novell, Inc.
____________________________
IN THE UNITED STATES DISTRICT COURT
DISTRICT OF UTAH, CENTRAL DIVISION
___________________________
THE SCO GROUP, INC., a Delaware
corporation,
Plaintiff,
vs.
NOVELL, INC., a Delaware corporation,
Defendant.
_________________________
Case No. 2:04CV00139
MEMORANDUM IN SUPPORT OF
NOVELL, INC.’S DAUBERT
MOTION TO DISQUALIFY
DR. GARY PISANO
Judge Ted Stewart
________________
AND RELATED COUNTERCLAIMS.
i
TABLE OF CONTENTS
Page
I. INTRODUCTION AND SUMMARY OF ARGUMENT
............................................. 1
II. ARGUMENT
................................................................................
...................................... 3
A. Dr. Pisano Cannot Offer Opinion Testimony Based on a Survey He
Knows Nothing About
................................................................................
............ 3
B. The Survey Does Not Measure Demand
................................................................ 4
1. Demand Cannot Be Measured without Reference to Price ........................
5
2. Demand Cannot Be Measured without Reference to Need ........................
6
C. Dr. Pisano’s Proxy Hypothesis Is Unfounded
........................................................ 7
D. Dr. Pisano’s Opinion as to the Size of the Market Is Inadmissible
under Rules 402 and
403........................................
...............
.......... 10
III. CONCLUSION
.....................................................
................................. 10
ii
TABLE OF AUTHORITIES
Page(s)
Cases
Bogacki v. American Machinary & Foundry Co.,
417 F.2d 400 (3d Cir. 1969)
..............................................
....................... 3
Claar v. Burlington N.R.R.,
29 F.3d 499 (9th Cir. 1994)
........................................................
........................ 9
Coastal Oil & Gas Corp. v. Garza Energy Trust,
268 S.W.3d 1 (Tex. 2008)
................................................................. 5
Commissioner of Internal Revenue v. Shattuck,
97 F.2d 790 (7th Cir. 1938)
.................................................
....................... 6
Crystal Semiconductor Corp. v. TriTech Microelec International, Inc.,
246 F.3d 1336 (Fed. Cir. 2001)
..................................................................... 5
General Electric Co. v. Joiner,
522 U.S. 136, 118 S. Ct. 512 (1997)
.................................................................................... 3
Loctite Corp. v. National Starch & Chemical Corp,
516 F. Supp. 190 (S.D.N.Y. 1981)
.................................................................... 4
Monolithic Power Systems, Inc. v. O2 Micro International Ltd.,
476 F. Supp. 2d 1143 (N.D. Cal. 2007)
...................................................................... 6
Porter v. Whitehall Laboratories, Inc.,
791 F. Supp. 1335 (S.D. Ind. 1992)
................................................................... 9
Pro-Football v. Harjo,
284 F. Supp. 2d 96 (D.D.C. 2003)
...................................................................... 4
Sheats v. Bowen,
318 F. Supp. 640 (D. Del. 1970)
................................................................... 3
The SCO Group, Inc. v. Novell, Inc.,
578 F.3d 1201 (10th Cir. 2009)
..................................................................... 7
United States v. Massey,
594 F.2d 676 (8th Cir. 1979)
............................................................................3
Statutes/Rules
Federal Rule of Evidence 402
....................................................................... 10
Federal Rule of Evidence 403
....................................................................... 10
Federal Rule of Evidence 702
..................................................................... 3, 10
Federal Rule of Evidence 703
................................................................... 10
ii
I. INTRODUCTION AND SUMMARY OF ARGUMENT
SCO sued Novell for allegedly interfering with its SCOsource licensing program,
by
which SCO has been trying to extract money from users of Linux operating
systems. According
to SCO, it would have made hundreds of millions of dollars from this program if
Novell had not
made known to those users that SCO does not have clear title to the copyrights
on which it was
threatening to sue them. SCO hired Dr. Gary Pisano to testify as to the unit
volume of lost sales
and Novell’s supposed causation thereof, as follows: -
“The potential license market … includes approximately 7.4 million Server
Operating Environment shipments”; and
- “SCO would have sold its SCOsource license to between 19% and 45% of the
Linux Market, but for Novell’s statements regarding copyright ownership.”
(Expert Report & Decl. of Gary Pisano [“Report,” reproduced as Ex. A hereto]
at ¶¶ 12(a), (b).)
Dr. Pisano takes his market penetration figures from this chart, which appeared
in a
November 2004 publication by The Yankee Group (reproduced as Ex. B hereto):
[Exhibit 1, Indemnification Concerns - see PDF]
1
Dr. Pisano arrives at 19% by combining the 8% for whom “[i]ndemnification and
product
warranties are a top priority” with the 11% who “are very concerned.” (Report at
¶¶ 71, 74.) He
gets to 45% by adding in the “somewhat concerned” respondents. (Id.) To this
point, there is
nothing very “expert” about Dr. Pisano’s opinion; anyone can read numbers off a
chart and add
them together.
Dr. Pisano then hypothesizes that the importance of indemnification, as measured
by the
survey, is “an excellent proxy for those who would purchase a SCO RTU [right to
use license].”
(Ex. C at 65:1-3.) Having posited that proxy relationship, Dr. Pisano concludes
that SCO would
have sold licenses to all respondents concerned about Linux indemnification,
thereby achieving
between 19% and 45% penetration of the 7.4 million deployment market, but for
Novell’s
publication of its rival claim to copyright ownership. (See Report at ¶12.)
As explained below, Dr. Pisano’s opinion regarding market penetration is
inadmissible
for at least three reasons. First, his reliance on an irrelevant survey that he
knows nothing about
is impermissible. Second, because the survey is entirely insensitive to price
and need, it cannot
reliably gauge demand. Third, the proxy relationship he assumes is untenable,
and ignores the
most obvious alternative explanation for Linux users’ decisions not to buy a SCO
license—viz.,
as rational actors they had no incentive to pay SCO for partial indemnification
that would have
been redundant of full indemnification they were already given for free.
The foregoing is directed to Dr. Pisano’s market penetration opinion. Without
that
opinion, Dr. Pisano’s other opinion—viz., “The potential license market …
includes
approximately 7.4 million Server Operating Environment shipments”—is
inadmissible under
Federal Rules of Evidence (“Rules”) 402 and 403 because the purported size of
the market, taken
alone, is irrelevant; and its introduction is likely to confuse the issues and
mislead the jury
regarding damages.
2
II. ARGUMENT
If … specialized knowledge will assist the trier of fact … a witness qualified
as an
expert … may testify thereto in the form of an opinion or otherwise, if (1) the
testimony is based upon sufficient facts or data, (2) the testimony is the
product of
reliable principles and methods, and (3) the witness has applied the principles
and
methods reliably to the facts of the case.
Rule 702. As set forth below, Dr. Pisano’s opinions are inadmissible because
they are not based
on sufficient facts or data, or the product of any reliable principle or method,
but are instead
either borrowed from unverified and irrelevant sources or founded on his own
ipse dixit. See
General Electric Co. v. Joiner, 522 U.S. 136, 146, 118 S. Ct. 512 (1997)
(“Nothing in either
Daubert or the Federal Rules of Evidence requires a district court to admit
opinion evidence that
is connected to existing data only by the ipse dixit of the expert. A court may
conclude that there
is simply too great an analytical gap between the data and the opinion
proffered.”).
A. Dr. Pisano Cannot Offer Opinion Testimony Based on a Survey He Knows
Nothing About
“In order for an expert to base his opinion on a study it is necessary that he
be able to
testify of his own knowledge as to the nature and extent of the source from
which statistics were
gathered.” Sheats v. Bowen, 318 F. Supp. 640, 644 (D. Del. 1970). Dr. Pisano
admitted that he
has no such knowledge. (Ex. C at 212:8–220:23.) Because Dr. Pisano has no
foundational
knowledge of the survey on which his proffered opinion testimony is based, that
testimony is
inadmissible. See United States v. Massey, 594 F.2d 676, 680 (8th Cir. 1979)
(“Concerning the
witness’ reference to the Canadian study there can be no doubt that the
foundation for such
testimony was insufficient. The witness testified that he did not know the
nature and extent of
the studies conducted from which the statistics were gathered.”); Bogacki v.
American Mach. &
Foundry Co., 417 F.2d 400, 407-08 (3d Cir. 1969) (“Venable was examined
exhaustively in
respect to the ‘Statistical Studies,’ and it became apparent that he could not
testify of his own
knowledge as to the nature and extent of the source from which the statistics
were gathered. It
3
follows … that Venable, lacking the necessary testimonial qualifications, could
not be permitted
to base his testimony on items of the ‘Statistical Studies.’”); Fed. Jud. Ctr.,
REFERENCE GUIDE ON
SURVEY RESEARCH 239 (reproduced as Ex. D hereto) (“The secondary expert who
gives an
opinion about the … interpretation of a survey not only should have general
skills and
experience with surveys and be familiar with all of the issues addressed in this
reference guide,
but also should demonstrate familiarity with the following properties of the
survey being
discussed: 1. the purpose of the survey; 2. the survey methodology, including a.
the target
population, b. the sampling design used in conducting the survey, c. the survey
instrument
(questionnaire or interview schedule), and d. (for interview surveys)
interviewer training and
instruction; 3. the results, including rates and patterns of missing data; and
4. the statistical
analyses used to interpret the results.”)
B. The Survey Does Not Measure Demand
The Federal Judicial Center’s REFERENCE GUIDE ON SURVEY RESEARCH cautions:
“Surveys not conducted specifically in preparation for, or in response to,
litigation … frequently
ask irrelevant questions.” (Ex. D at 237.) As explained below, that is precisely
what happened,
here. The survey from which Dr. Pisano takes his numbers is irrelevant to lost
profits because it
is insensitive to price and to whether respondents had an unmet need for
indemnification that
would have been met by a SCOsource license. Cf. Pro-Football v. Harjo, 284 F.
Supp. 2d 96,
144 (D.D.C. 2003) (“The survey … says nothing about whether the term
‘redskin(s)’ when used
in connection with Pro-Football’s football team disparages Native Americans. …
The survey is
completely irrelevant to the analysis.”); Loctite Corp. v. Nat’l Starch &
Chem. Corp, 516 F.
Supp. 190, 206 (S.D.N.Y. 1981) (“none of these surveys, which were conducted
prior to this
litigation and for marketing purposes unrelated to this controversy, tested the
single issue at
hand”).
4
1. Demand Cannot Be Measured without Reference to Price
Apparently, the question asked by the survey from which Dr. Pisano takes his
opinion
was: “rate the importance of Linux legal indemnification and product warranty.”
(Ex. B at 3.)
Respondents could choose one of four options: (1) “it is not a concern,” (2) “we
are somewhat
concerned,” (3) “we are very concerned,” or (4) “indemnification and product
warranties are a
top priority.” (Id.) Strikingly absent from the survey question is any reference
to price (or need,
as discussed in the next section).
“All markets must respect the law of demand,” according to which “consumers
will
almost always purchase fewer units of a product at a higher price than at a
lower price, possibly
substituting other products.” Crystal Semiconductor Corp. v. TriTech Microelec.
Int’l, Inc., 246
F.3d 1336, 1359 (Fed. Cir. 2001); see also Coastal Oil & Gas Corp. v. Garza
Energy Trust, 268
S.W.3d 1, 41 (Tex. 2008) (“we cannot repeal the law of supply and demand any
more than we
can repeal the law of gravity”). But the survey Dr. Pisano uses asked
respondents to “rate the
importance of Linux legal indemnification and product warranty” without even
mentioning price.
(See Ex. B at 3.) Whether or not the result is a valid measure of concern about
indemnification,
it cannot be a reliable guide to demand because it is entirely insensitive to
price.
Crystal Semiconductors, which affirmed remittitur of price erosion damages
awarded by
patentee’s expert had
simply multiplied the per unit price erosion by the number of units sold,
without making
allowance for the decrease in demand that would have accompanied a higher price,
according to
the law of demand. 246 F.3d at 1358. As the Federal Circuit explained, “[t]o
show causation
with reliable evidence, a patentee must produce credible economic evidence to
show the decrease
in sales, if any, that would have occurred at the higher hypothetical price.”
Because “Crystal …
presented no evidence of the elasticity of demand” or “make any estimates as to
the number of
sales it would have lost or kept had it increased its prices,” “Crystal did not
make a showing of
5
‘but for’ causation.” Id. at 1359. Because the survey Dr. Pisano appropriates
for his opinion in
this case is entirely insensitive to the inexorable law of demand, it likewise
cannot be a reliable
measure of demand. See also Monolithic Power Sys., Inc. v. O2 Micro Int’l Ltd.,
476 F. Supp.
2d 1143, 1155 (N.D. Cal. 2007) (“Mr. Bratic’s report does not account for the
law of supply and
demand. Mr. Bratic hypothesizes a royalty that would triple the average selling
price for MPS’
accused products, but he makes no allowance for the impact that increased prices
would have
had on demand.”).
Put another way, to translate Dr. Pisano’s estimate of penetration into damages,
it must
be multiplied by a price. The survey from which Dr. Pisano takes his estimate
might suggest that
1.4 million Linux users would accept indemnification if SCO were giving it away.
But that
survey cannot support any opinion that those users would pay $699, $100, or any
other amount
for it. See Comm’r of Internal Revenue v. Shattuck, 97 F.2d 790, 792 (7th Cir.
1938) (“It is a
matter of common knowledge that the value of any product or commodity, whether
it be wheat,
hogs or otherwise, is affected by the law of supply and demand.”).
Generic interest in indemnification, as measured by the survey from which Dr.
Pisano
takes his numbers, is not at issue in this case. The issue is lost profits,
which the survey cannot
help to measure because it is insensitive to price.
2. Demand Cannot Be Measured without Reference to Need
As Dr. Pisano concedes, a SCOsource license was not the only option available to
Linux
users who wanted indemnity. (Report at ¶ 83.) Several of the developers and
distributors of the
software for which SCO is trying to sell licenses indemnify their customers at
no additional cost.
(Id.) Moreover, such protection is superior to a SCOsource license because it
reaches all
intellectual property infringement claims, whereas the only protection afforded
by a SCOsource
license is from SCO’s copyright infringement claim. (Id.)
6
The survey Dr. Pisano uses apparently did not ask respondents if they already
had
indemnity. (See Ex. B at 3.) For all the survey reveals, the entire 45% for
whom
indemnification was a top priority or who were very concerned might already have
obtained
indemnity from some other source. Indeed, their concern might have led them to
obtain their
Linux from an indemnifying distributor. The survey simply does not say (and Dr.
Pisano
apparently does not know, see Ex. C at 212:8–220:23). Because the survey did not
ask
respondents if they already had indemnity, it can provide no insight into how
many of them
would have been willing to pay SCO for a license that would have been redundant
of any
indemnification they already had.1
C. Dr. Pisano’s Proxy Hypothesis Is Unfounded
The survey from which Dr. Pisano takes his numbers asked respondents to “rate
the
importance of Linux legal indemnification and product warranty.” (Ex. B at 3.)
To justify his
equation of the perceived importance of indemnification and warranty, on one
side, with but-for
purchase of a SCO license, on the other, Dr. Pisano reasons: “Users willing to
obtain
indemnification were those most concerned with the risks of IP [intellectual
property] litigation.
This set of users would thus have been the most likely purchasers of SCOsource
Right to Use
licenses.” (Report at ¶ 69.) Or, as he more succinctly phrased this line of
reasoning in his
deposition, the importance of indemnification as measured by the survey is “an
excellent proxy
for those who would purchase a SCO RTU.” (Ex. C at 65:1-3.) But as explained
below, even
assuming that the survey measures “[u]sers willing to obtain indemnification,”
and that those
users are motivated by concern “with the risks of IP litigation,” there is no
reason to think those
7
users would have been “likely purchasers of SCOsource … licenses.” Dr. Pisano’s
proxy
hypothesis is unfounded.
According to Dr. Pisano’s survey, “[b]efore SCO’s decision to initiate a $1
billion
copyright infringement suit against IBM in March 2003,” indemnification “was not
an issue.”
(Ex. B at 4.) “Linux indemnification [was] more of an issue late in calendar
year 2004,” when
the report issued, but not because of SCO. (Id. at 3.) Previously, “Linux and
open source users
thought little about indemnification …. because Linux and open source products
were not widely
deployed in enterprise environments.” (Id. at 4.) But by late 2004, when the
report came out,
“Linux is no longer a hobbyist’s toy; it’s becoming a commercial mainstream
product and is
assuming all of the advantages, risks and liabilities that accompany commercial
software.” (Id.
at 9.)
Linux’s penetration of the enterprise market meant there was more at stake for
individual
users, increasing their desire for indemnification. (See Report at ¶¶ 59-60.) As
the survey
recognized, that penetration also increased the risk that various opportunists,
not just SCO,
would try to extract royalties from the growing user base: “forget about SCO.
Imagine there is
no SCO lawsuit. … Should SCO lose, as many believe it will, there is nothing to
preclude other
individuals or organizations from filing similar lawsuits.” (Ex. B at 6.)
“Copyright and patents
are big business. With so much royalty money at stake, it’s foolish to think
that vendors will not
enforce their rights if an infringement threatens a significant revenue stream.”
(Id. at 9.) For
example, “[i]n August 2004, OSRM [Open Source Risk Management, Inc.] released
the results
of a commissioned survey that showed Linux infringed on 283 patents. IBM holds
approximately one-third of those patents and Microsoft holds 27.” (Id. at 8.)
IBM executives
“made no mention of when or if IBM would assert its rights over the GUI
[graphical user
interface] and LDAP [lightweight directory access protocol] portions of the code
– but the
possibility and the issue remain open.” (Id. at 9.)
8
Everyone expects SCO to lose. The threats against which users might seek
indemnity
come from elsewhere. But as Dr. Pisano concedes, SCO’s license addresses only
the (weak)
threat posed by SCO. (Report at ¶ 83.) As discussed above, it offers no
protection against other
copyright or any patent infringement claims. (Id.) Other indemnity providers,
such as the
developers and distributors of the software for which SCO is trying to sell
licenses, offer
protection that is superior because it reaches all IP infringement claims; and
that protection,
being free, is cheaper than SCO’s $699 license. (Id.) Dr. Pisano’s theory that
1.4 million users
would have paid SCO for partial protection that is redundant of full protection
they had already
obtained from other sources, for free, but for Novell’s assertion of ownership,
simply makes no
sense.2 At a minimum, Dr. Pisano’s analysis is defective because he has not
ruled out the most
obvious explanation for the behavior of those customers, viz., they are rational
actors. See Claar
v. Burlington N.R.R., 29 F.3d 499, 502 (9th Cir. 1994) (upholding exclusion of
testimony by
experts who had not “made any effort to rule out other possible causes”).
Dr. Pisano is the Harry E. Figgie, Jr. Professor of Business Administration at
the Harvard
Business School. (Report at ¶ 6.) But however prestigious his appointment, “[a]n
expert’s mere
guess or conjecture is properly excluded, because an expert is a conduit of
facts and not merely a
subjective speculator relying on stature alone.” Porter v. Whitehall Labs.,
Inc., 791 F. Supp.
1335, 1343 (S.D. Ind. 1992). There is no persuasive support for Dr. Pisano’s
critical proxy
hypothesis, or his corresponding theory of causation, other than Dr. Pisano’s
endorsement; and
whatever his stature, that does not suffice.
9
D. Dr. Pisano’s Opinion as to the Size of the Market Is Inadmissible under
Rules 402 and 403
As explained above, Dr. Pisano’s proposed opinion testimony regarding market
penetration is inadmissible because it does not meet the requirements of Rules
702 and 703.
Without that opinion, his proposed testimony about the size of the market is
inadmissible under
Rules 402 and 403. First, the size of the market, by itself, is not relevant to
anything. Second,
absent a proper damages analysis to which the size of the market is relevant,
evidence regarding
the size of the market is likely to prejudice Novell by confusing the issues and
misleading the
jury with respect to the calculation of damages.
III. CONCLUSION
Dr. Pisano’s opinion regarding market penetration does not satisfy the
requirements of
Rules 702 and 703; and without that opinion, his opinion regarding the size of
the market is
inadmissible under Rules 402 and 403. Thus the Court should not permit Dr.
Pisano to testify.
DATED: February 8, 2010
Respectfully submitted,
By: /s/ Sterling A. Brennan
WORKMAN NYDEGGER
MORRISON & FOERSTER LLP
Attorneys for Defendant and
Counterclaim-Plaintiff Novell, Inc.
1 Dr. Pisano tries to blame Novell for the existence of indemnity alternatives,
as well, theorizing
that Novell brought them into being “[b]y creating uncertainty regarding SCO’s
ownership
rights.” Novell did not create any uncertainty. At worst, Novell brought to the
attention of
SCO’s intended targets the fact that under a reading of the APA and Amendment
No. 2 which
the Tenth Circuit has ruled is plausible, Novell owns the copyrights SCO was
trying to license.
See SCO Group, Inc. v. Novell, Inc., 578 F.3d 1201, 1217 (10th Cir. 2009).
2Although not all Linux distributors offer indemnity, neither Dr. Pisano nor
the survey
disaggregate the customers of Linux distributors that do indemnify from the
customers of
distributors that do not.
10
*****************************************************
*****************************************************
*****************************************************
EXHIBIT C
[ back to top ] [ Pisano's Expert Report, Exhibit A ]
[ Novell's Memorandum in Support ]
Deposition of GARY PISANO, PhD
Date: July 27, 2007
Case: SCO v. NOVELL
SHARI MOSS & ASSOCIATES
[address, phone, fax]
... information that these -- that these respondents would have had.
Q They may have been influenced by SuSE, right, SuSE's statements that it didn't infringe?
A Absolutely.
The respondents would -- again, that's the real strength of this methodology, that you have these player -- you have these -- these respondents expressing an interest in, you know, whether or not they're going to purchase indemnification at that time.
So you have a series of -- you've got-- again let's take the 19 to 45 percent. You've got the residual of that, 81 to 55 percent, who have heard statements by SuSE. They've thought about issues around, you know, the infringement, et cetera, and they've decided that they're not going to be -- they're not concerned enough with the -- any of the issues.
But you have 19 to 45 percent who at that point are expressing concern and intentions to purchase indemnification.
64
And that's the proxy. It's an excellent proxy for those who would purchase a SCO RTU.
Q So what do you mean by the 55 to 81 percent are not concerned enough with any of the issues?
A That's the -- sorry. So they -- where that comes from, you have 19 to 45 percent in the surveys are showing this range of -- of concern from -- very concerned --
Let me just repeat that, because I think I mumbled the words.
What the surveys -- the 55 to 81 percent are those who said, you know, these IP issues are not a concern to us and/or we don't plan to buy indemnification.
Those -- they have taken into account all of the information available to them in making that -- expressing that opinion.
The 19 to 45 percent have also taken into account all the information concerning lots of factors, and they are expressing high degree -- you know, high degrees of concern and/or intentions to purchase indemnification, and they're a proxy for the
65
clients, I think -- well, I don't know if they're --
They're certainly giving it some degree of credibility. I don't think they would put it before their clients if they didn't think it was credible.
(Pause.)
Q Dr. Pisano, tell me what you know about the methodology for the Yankee Group 2004 survey.
A It was a survey of companies across a size -- it's actually laid out. I believe it's right in the report itself, a thousand -- a thousand customers -- sorry, a thousand users were contacted.
They were asked a variety of questions, a thousand organizations that they had -- that they had contacted.
I think this was an online survey. The respondents filled it out online, sort of a random sample.
MS. BORUCHOW: You're question is about 376, right? Exhibit 376?
MR. PERNICK: Yes.
THE WITNESS: Yes.
212
BY MR. PERNICK:
Q Is it your understanding looking at page 3 that a thousand organizations were surveyed?
(Witness read document.)
A Yes.
Q And do you see from the question in Exhibit 1, which is the box on page 3, it says that midsize or large organizations with 5,000-plus employees were asked to respond?
A Right, yes.
Q And is that in itself sufficient to convince you that this survey was reliable?
MS. BORUCHOW: Object to form.
A It appears to be a reliable survey. I have no reason to doubt it.
Q Well, aside from having no reason to doubt it, what do you base your conclusion that it's a reliable survey on?
A Again, this is an organization, Yankee, that does these kind of surveys routinely.
Companies rely on them for, you know, this kind of market research. That's what these guys do for -- for a living --, and it's a large sample.
213
(Pause.)
A It's reliable.
Q You say it's a large sample.
Do you know who responded, how many -- who at the various companies, anything like that?
A I don't have the details on -- on who responded.
Q I mean, to be clear, we know that a thousand companies were asked, I guess.
A Right.
Q We don't know anything about who responded, right?
A I'd have to sort of look at that. I don't recall offhand if I have the numbers on -- on that, if that was reported.
(Pause.)
Q Does it matter?
A Response rate? Yes, I mean the specific -- which specific players respond doesn't matter, you know, Company A or Company B.
It doesn't -- the individual companies responding doesn't matter, but --
Q What do you mean?
214
A Well, you said does it matter who responded.
So this goes out to a thousand organizations. And think about the companies, you know, 1 through a 1,000, you know, did Company No. 2 on the list respond or Company No. 3 on the list respond?
That doesn't matter for a survey.
Q What -- does it matter. --
I mean, you don't know how many responded when you're saying that doesn't matter?
A No, I didn't say that.
The overall response rate matters. I have -- I can't recall whether they report the response rate.
My presumption would be in a survey like this, in an organization like this, that they're getting certain high enough response rates; otherwise they wouldn't be --
You know, this is a company, again, whose primary -- they're in the primary business of doing these kind of surveys and, you know, and publish this kind -- these kind of data.
215
You can get actually -- there's surveys with very low response rates, which, again, as long as the responses are, you know, sort of randomly distributed, not biased on the response, you actually have a very robust study.
Q And how do you know in this case whether you have those conditions?
(Witness read document.)
A I -- don't know 100 percent for sure. Again, I'm relying on the fact that Yankee is a well-respected organization that --
You know, they -- they kind of make their -- they make their living really and their credibility in the market matters.
And so I'm going to, you know, believe that they are doing a good job on the -- on the surveys.
We often are forced to do that in research and in my own academic research where I don't have all the details of the surveys and response rates, but you look at the credibility of the -- of the source.
And, again, this is an organization that makes its living by doing response --
216
By doing these kind of surveys.
And so, you know, presumably, they -- they can't be out there, you know, publishing flawed studies.
Q Do you know what checks were employed to make sure that this survey was done on a sound basis, what kind of procedural mechanisms?
A I don't have details on that, no.
Q Do you generally?
A No.
I mean, there's a standard -- I mean, again, doing survey research, there's a very standard set of approaches that are -- that are kind of used in terms of, you know, sending it out, getting response rates, tracking.
It's -- it's -- it's not, you know -- making sure you record the responses correctly.
Again, a company like Yankee that does this kind of stuff would be really, I think -- you know, this --
You know, it's like Toyota making
217
cars. They know how to make cars. These guys know how to do surveys.
Q You're talking about in general, though?
You don't know anything about what Yankee did to conduct this survey, right?
MS. BORUCHOW: Object to form.
A I mean, I do.
They conducted a survey of a thousand organizations, you know, and asked them questions.
That's what they did for this survey. That's what you do in surveys.
Q Do you know if there was anything done to ensure that the respondents that these unknown, unnumbered entities had the authority and knowledge to answer for the companies?
MS. BORUCHOW: Object to form.
A I'm not sure I follow the question.
Can you --
Q Well, do you know if there were any checks employed to make sure that, okay, we're going to send this to Acme Company and we're going to make sure we get a response
218
from the person who's the one who should be answering?
A Again, in organizations like -- they routinely administer surveys. This is not the first survey Yankee's done. They list a whole bunch of other ones they've done.
So they have these things, have already kind of been worked out for who's getting them and how they're doing them.
They typically -- you know, there's -- you know, those kind of procedures are in place in organizations like this.
This is not an unknown -- Yankee Group is an extremely well-regarded, well known, you know, information provider in this space.
Q Do you know anything about what procedures, though, they employed here?
A I don't know the specific procedures, no.
Q Generally do you know?
A The, you know, standard survey procedures. This was done -- I can't recall -- I was looking through this so I could remember the
219
Specifics of whether it was done online and --
I -- I don't know the specific --
I mean, what kind of procedures are you --
Q I don't know. I'm asking you.
A Again, it's very standard -- standard kind of survey. Contact these organizations and -- and get responses.
Q Assuming there is some kind of standard survey procedure, how do you know it was employed here?
A Again, this -- Yankee group is a very well-respected, you know, research organization. They do these kind of things all the time. Their data is used widely by organizations.
It's a little like census data. I mean, you know, people use census data all the time in their research. Do they ask the questions, the details of the census? No. It's an understanding it's done -- done well.
Q Did you interview any of the respondents to this survey?
220
[ back to top ]
[ Pisano's Expert Report, Exhibit A ]
[ Novell's Memorandum in Support of Daubert Hearing Motion ]
[ Pisano Deposition, Exhibit C ]
___________________
|