SCO Drops Trade Secret Claim and Shows No "Infringing" Linux Code So Far: Eyewitness Account

Friday, February 06 2004 @ 03:57 PM EST

Contributed by: PJ

I have another eyewitness account. It's very detailed and the person tells me that wherever you see direct quotations, it's word for word. The notes are a bit choppy, due to the speed of creating them.

The big info is this: SCO has apparently dropped its trade secret claims against IBM and as for the copyright claim, it is melting. So far, according to IBM, SCO has shown not one line of "infringing code" and SCO, he said, "repeats their theory" that IBM put AIX code into Linux, and identified only 17 files and identified a few lines in those files.

How would that impact Linux end users, even if they could prove their legal theory?

(UPDATE: I have just heard from another eyewitness too, and he heard the same thing. He read to me the copyright infringement claim, and in fact it is only about IBM and only about their continuing to distribute AIX after the "termination" by SCO, not any prior copyright infringement.)

The other big news, to me, is that IBM seems not to want to drop the matter. My eyewitness says that they didn't ask for summary judgment. His impression was that they very much want to go forward. His impression also was that Marriott was very prepared and professional and that Heise looked amateurish in comparison. So that's two lawyers down. We'll see who shows up next time.

We'll have the Exhibits attached to the SCO Motion scanned and available by this afternoon or early evening and then we'll be able to confirm this news or clarify it.

Of course, SCO said they can identify millions or whatever if IBM will turn over the AIX code, all of it since the beginning of the world. But so far, it's 17 files. Count them. 17. And only a few lines. All related to their AIX-derivative code theory.

I'd say this was a very happy day. Of course, this is all coming in fast, and is subject to correction as we get more information, but I'm a happy girl so far. If any of you have some proof handy on HP contributions to Linux, that would be nice. Here are the notes. Excuse typos, if any.


Judge Wells, SCO v. IBM, Federal Courthouse, Rm 436, Friday 2/6/2004,
10:00 AM

Arrived to a packed courthouse (only holds about 30 people). I
recognized a few of the people, mainly in the back row (Blake
Stowell, Ralph Yarro). Attorneys for SCO included Mark Heise, Ryan
Tibbitts, and Brent Hatch. For IBM, attorneys were Dave Marriott and
Todd Shaughnessy.

At about 10:05, the attorneys were escorted to the judge's chambers
for a conference. During the ensuing minutes, Stowell was approached
by an AP reporter. She asked about the damages amount. Blake
indicated that it had been $3 billion, but "that was upped
yesterday", "amended yesterday to $5 billion", but it could be "much,
much more" because IBM continues to sell AIX.

The attorneys emerged about 10:30 AM, the judge followed shortly
after, and summarized what happened in the chambers, and that the
hearing would focus on two main areas:

1) request to go forward to hear arguments whether SCO complied with
the court order of 12/5, or if other action is required;

2) SCO motions for reciprocal discovery (from IBM).

Dave Marriott (IBM) went first: Said the "simple answer" is that SCO
has not complied, that the court ordered SCO to produce by 1/12, that
SCO has not provided all of the required documents, as shown by notes
from meetings and statements filed by IBM to the court yesterday.
Mr. Marriott claimed that about 40 "custodians" of SCO code were
supposed to be queried, but that documents from only about 1/2 (20 of
them) were provided to IBM.

The judge asked "If I find that they have not met their
requirements?", and Mr. Marriott asked that the judge order them to
do so by a specific date.

Mr. Marriott said SCO's complaint is that IBM took SysV code and put
it into Linux, but that in recent documents to the court:

1) SCO has abandoned that claim, and did not tie IBM acts to any
trade secret;
2) SCO failed to identify a single line of offending code;
3) that SCO then "repeats their theory" that IBM put AIX code into
Linux, and identified only 17 files and identified a few lines in
those files.

IBM asks in at least 7 places that SCO match Dynix code to SysV, and
SCO claims in their motion that IBM cannot give away their own (IBM)
code. Mr. Marriott asked, if the 17 files are not from SysV, then
IBM wants to know where they were derived from. Also quoted from
interrogatories #12 and #13, which is to identify what is in Linux
claimed by SCO, but SCO refuses to identify which, and SCO is
claiming that they may have additional rights which are also not
identified. SCO produced some invoices, but IBM contends they are
not enough. Mr. Marriott described as "Troubling" that SCO's CEO is
publicly saying "1 million" lines of code, and that IBM has a right
to know what they are and exactly where they are. He asked that SCO
be ordered to provide the information requested and tie it directly
to SysV and their derivatives.

Mark Heise stepped up next. Said that, as for compliance, last
hearing focused on interrogatories, and that SCO says they have
identified specific files. Claims that IBMs responses are:

1) "Not what we asked for " in questions
2) "Not applicable to the case" (NOTE: A number of people in the
back of the courtroom looked particularly puzzled at this point, and
looked to one another in a "Huh?" fashion)

"With respect to the overriding issue, that SCO failed to identify
line-for-line code copying", Heise claimed "that has not and is not
what the case is about". (Again, very surprised looks in the

Heise referred to the 1985 Agreement point 2.01, that "modifications
to SysV code must be treated as derivatives", and he claims that AIX
and Dynix are such derivatives. SCO feels that AIX and Dynix code
has been put into Linux, and that IBM admitted it publicly. Heise
made an analogy to "the first 10 rungs of a ladder", but the ladder
goes to step 20 now, and maybe step 16 has some issues with it.
Heise said that IBM "has not proven ownership of *their* code" and
that they must do so to show that it's OK to put it into Linux.
Heise gave a printout to the judge, and described it as showing
line-for-line ("in red") copying, and mentioned Async I/O and Scatter
Gather as two areas in question. Said they want IBM source code.
Judge said it is SCO's requirement to show: "this is about your
response, and compliance to the court order".

At that point, Heise said SCO cannot identify violations. The judge
said "The problem is, unless you identify those codes, then IBM is
not in a position to have a response. We're at an impasse, and the
case cannot continue with an impasse, that's why there was a court

Heise went back to the ladder analogy, saying "maybe rung 15 to 16"
might be involved, but they cannot identify the lines because SCO
doesn't have derivative IBM code. Heise then made comment (which
drew some audible "Huh?" responses from the audience), that
"Arguments of the case aren't appropriate at discovery." Went on to
claim that they have identified 400 million lines of Unix code and
300 million lines of Linux code affected, but also admitted that SCO
has not submitted everything required by the court order.

The judge asked: "How much time do you need, and if I order a strict
compliance, what is a reasonable date?" Heise said he anticipates 2
weeks more are needed, but would like 4 weeks.

Mr. Marriott stepped up again. He said reason they haven't provided
AIX code (but noted they have given Dynix code) was because of the
court order and they do not wish to defy the order. But went on to
say the question is SysV code in Linux.

The judge then asked if it is IBM's opinion that SCO is
"substantially in compliance" with the order, and Mr. Marriott said
no, and the interrogatories 1, 2, 4, 6, 9, 13 are examples of
shortcomings in the SCO responses, and that "the list goes on".
Judge said she would take that under advisement.

Judge then indicated that the hearing would proceed to the part about
motions by SCO to request discovery.

Heise spoke first: said "IBM hasn't provided AIX, and, prior to the
12/5 hearing, they had agreed to do so. Judge said that could be due
to the 12/5 ruling. Heise repeated that they had promised to give
such code before the 12/5 hearing and there is nothing from the
highest level of IBM execs either, but on 10/28 claims that IBM said
they would give a copy of Sam Palmisano's notes.

Went on to say that in terms of specifics to #2 and #3 (AIX and
Dynix), they have gotten nothing. Referred to the IBM CMVC (version
control) system, and that it would be easy for IBM to produce the
code, and that "they can get us the AIX code from one server". Also
said that "all contributions to Linux" are also known, and that IBM
admits to making a lot of contributions. Heise says the list of
contributors (which IBM provided) is not enough, and that IBM must
specify which persons gave which lines of code. He referred again to
Palmisano and a 10-page "move to Linux" document.

The judge then cut Heise short, and invited Mr. Marriott to speak.

Mr. Marriott said "SCO group has been handed 57 requests, and there
motion to compel addresses only 6". Said IBM gave info such as a
list of 8,000 people who worked on AIX and Dynix. Says SCO wants
every file in AIX, Dynix, and Linux made by IBM, because "SCO's
theory is that all code from AIX and Dynix inserted into Linux is
improper" yet SCO refers to only 17 files. Also pointed out that the
Linux source code is publicly available, and that SCO could compare
between their files and the Linux code directly, not by looking at
millions of lines of other code. Mr. Marriott also referred to a
Computer Associates decision as to how derived works are laid out
(Note: CA was sued by a Canopy company and Yarro was involved; it may
have been Altiris (?)).

He said that SCO says HP and Sun complied, but that they figured it
out without reference to Solaris and HP-UX code, and "How could they
do it? And make public announcements about it?" when those companies
had not provided source code. Said SCO didn't rely on production of
millions of lines of their code, and so IBM should not have to do so,
but that IBM *is* (his emphasis) prepared to produce and has already
given SCO Dynix source code, but not AIX code because it would
violate the court order. Marriott said IBM is prepared to give 3
million pages of code covering 232 products, and that IBM is prepared
to produce those documents, but not every draft of every version of
every file because that involved hundreds of products and thousands
of people. Said IBM is not hiding anything, but that SCO requested
"every conceivable contribution to Linux and Open Source". Then said
that the OSDL shows such contributions are public knowledge and that
SCO can get it easily, and that they have already accessed files, and
so asking IBM to do so would be unduly burdensome. Said Rule 26(b)2
says discovery "shall be limited if available from other sources with
less burden", and he quoted several cases.

The judge asked: Assume I'm going to order IBM (Note: I think she
meant SCO) to deliver, how long would it take IBM to review the
material. Marriott said they would do so within 14 business days of
lifting the stay, but that for 100 million lines of code, it would
take months.

(Note: Marriott quoted "Litchfield v. Spielberg" at this point, but
didn't explain what it is or else I couldn't hear him).

Heise spoke again: said "IBM has spent the better part of the day
saying AIX and Dynix arent derivative works". Said IBM can't say
it's not SysV derivative, and that they won't give SCO the source
code so they can prove it. Said there are over 1000 files from AT&T
in AIX. Also, that "As for HP, there is a fundamental difference in
that they haven't said they are contributing to Linux". Heise said
IBM is obligated to answer what they provided to Linux, but SCO
cannot say which of those things are covered by agreements.

Mr. Marriott then noted that HP has contributed to Linux and that
public records confirm that.