Legal Analysts Examine SCO Claims Against GPL
Aug 22, 2003, 15:00 (9 Talkback[s])
(Other stories by Steven J. Vaughan-Nichols)
No-Size-Fits-All! An Application-Down Approach for Your Cloud Transformation
By Steven J. Vaughan-Nichols
The SCO Group's CEO Darl
McBride may have thought he was pulling a public relations coup on
August 18th at his company annual trade show, SCO Forum in Las
Vegas, by showing slides with obscured lines of code purporting to
show that SCO Unix code has been stolen and placed in Linux.
Instead, according to open source advocate Bruce Perens, McBride
ended up with egg on his face.
Perens, with aid from other open source advocates showed
conclusively in his masterful paper, Analysis of Linux
Code that at least two code samples has been "twice released
under the BSD license: once by Unix Systems Labs (AT&T), and
again by Caldera, the company that now calls itself SCO."
Linus Torvalds, Linux's founder, said in a
ZDNet Australia interview, that "So one code snippet was from
pretty much original Unix -- and yes, Caldera released the old Unix
code itself back when they still remembered that they made all
their IPO money off Linux, which is interesting partly because it
shows how SCO has been lying all along: they said several times how
they are talking about SVR4 code, not 'old Unix' code, and now they
show old Unix code on their slides."
A SCO spokesman insisted though that the code shown at SCO Forum
was from a current Unix licensee and showed SCO-owned 32-bit Unix
System V code, not the 16-bit 'ancient'
Unix code that had been released under Ransom Love's regime for
personal and non-commercial use.
He also said, in response to open source's leaders request that
they be freely shown the offending code that all of Non-Uniform
Memory Access (NUMA), IBM's Journaled File System (JFS), and
Silicon Graphics' journaling file system (XFS)-this last is a new
SCO claim-and the majority of Symmetric MultiProcessing (SMP) code
in Linux was in violation of SCO's copyright. And, he reiterated
that over a million lines of code in Linux 2.4 and above was taken
from SCO's System V and derived works.
Open Source Inititive President Eric S. Raymond has put out
own report on the leaked code, a report that stipulates that
SCO may indeed have reason to believe that the disputed code is
owned by them. But that same report also concludes that since SCO
knowingly released their code under the GPL, all of this ownership
discussion is moot.
SCO has now taking the legal tack of deriding the GPL itself.
Which seems an odd step, given that SCO has already contributed its
own code to GPL both by virtue of its employees, such as Christoph
Hellwig, Kernel Engineer Unix/Linux Integration for Caldera's
Germany subsidiary, Caldera Deutschland GmbH was working on combining
Unix and Linux, and, of course, released its own Linux
distributions. SCO has responded that "Section 0 of the GPL
specifically states that the code must be contributed and that SCO
never did this." Instead, SCO just "distributed the code, which is
very different from contributing the code."
Does this argument hold water? Thomas C. Carey, chairman of the
business practice group at Boston-based Bromberg & Sunstein, LLP, an IP
litigation and business law firm, thinks, "Perhaps he means to say
that SCO did not execute a copyright assignment. But if SCO made
the Unix System V code available to customers with the GPL included
in the file, then it granted third parties the right to make an
unlimited number of copies and to make derivative works."
Frankly, he thinks that, "Rather than quibble over semantics
(distribute versus contribute), SCO would be better off to simply
argue that its publication under the GPL was inadvertent. I think a
court would be sympathetic to that argument, unless SCO was
negligent in its publication."
John Ferrell, founder of Carr & Ferrell, LLP, a
Silicon Valley intellectual property and corporate law firm,
believes that "The fact that SCO is denying contribution seems to
be setting up the defense that SCO was mistaken as to what they
were licensing, and likewise those licensing the software were
mistaken as to what they were receiving (Unix code rather than
Linux code). This contract of 'mutual mistake' in some
jurisdictions would permit either party to void the contract."
Ferrell also says, "There is an interesting 19th century case
about a pregnant cow named Rose that was mistakenly thought to be
barren and sold at a very low price. When upon delivery the cow was
found to be pregnant, the seller called the deal off. The Supreme
Court of Michigan in this case decided that since both parties were
mistaken as to the true nature of the goods (barren cow vs.
pregnant cow), and the contract could be voided by the seller.
Although still good law in some jurisdictions, courts in general
are very reluctant to undo contracts."
This also presumes that the GPL is a bi-lateral contract. The
GPL, though, is written as a unilateral contract, so the Rose's
case, which established the doctrine of mutual mistake, may not be
Who's right? Only the courts, if SCO ever get that far, can
decide. For now, though, the open source community is certain that
SCO has only damaged their case instead of helping it.