Subject: Re: SCO Drops Linux, Says Current Vendors May Be Liable (fwd)
From: Brian Behlendorf <>
Date: Fri, 16 May 2003 06:03:41 -0700 (PDT)

On Wed, 14 May 2003, Matt Asay wrote:
> "Is SCO litigating itself into irrelevance?"
> By definition, one must be relevant before one can become irrelevant,
> through litigation or otherwise.  (-;
> There may actually be some validity to their complaint (wrt IBM
> abrogating trade secrets in order to beef up the Linux kernel).  It's in
> IBM's interest, but being in one's interest is not exactly conclusive
> evidence.  I guess a super-competent judge will make that determination.
>  Right.

For as much as I think the courts make foolish judgements from time to
time, I actually trust that most of the judges out there can differentiate
between potential motive and proof of wrong-doing.  If there is code in
the Linux kernel, and SCO can show it's unmistakably like code from their
codebase, and that the contribution to Linux came from an IBM employee,
IBM's in deep doodoo.  I don't think they'll have a case as strong as
that, so even though there's a motive there's no smoking gun and a judge
wouldn't find guilt.

> Regardless, I'm disinclined to believe that a judge is going to try to
> turn back history by finding against Linux in any way.  I mean, this is
> the court system that found Microsoft to be a monopoly and still refused
> to actually do anything about it.  What better way to actually enforce
> their antitrust judgment than by letting Linux well enough alone, so
> that it can beat up MS?

Again, as much as I think courts make bad decisions quite often, I tend to
think the courts do not play favorites or politics like that.  Maybe at
the very top, where the Supreme Court judges have their own political and
moral biases, does that matter.  At this level, I think they'll rule on
the facts, just as they issued countless findings of facts and judgements
in favor of MS over the last ten years.

As for SCO's RIAA-like warning to the Linux community, I think they'll
start getting heat from players claiming anticompetitive practices.
Again, if their claims have merit and IBM let SCO code leak into Linux, I
think the judge will find that IBM has to compensate SCO for the entire
current value of SCO's IP - but it would make it much harder for SCO to
similarly litigate anyone else in the future, as the IBM judgement will
probably have covered everyone else.

But that's worst-case.  SCO's market cap is $30M.  Why doesn't IBM just
buy them?  That would be far less than what IBM's legal expenses
will be on this.  Most likely it's because for IBM there's a principle at
stake (settle with one, and everyone else will attack you on frivolous
grounds too), and perhaps they've also done enough homework and auditing
to know that SCO just does not have a case.

> So, while SCO litigates itself into deeper and deeper irrelevance
> (ironically by tying itself to the one mast that the rest of the
> industry seems to be happily moving away from), I doubt that it's ever
> going to get much satisfaction from it all, but for a possible slap on
> IBM's wrist (if, in fact, IBM did anything wrong, which is by no means
> clear).

If there's a clear link - code verified as being SCO IP that was released
in violation of SCO's agreement - then IBM's in trouble and shame on them
for not keeping to their IP contracts.  If it were IBM on the other side
of the table claiming IP infringement, their lawyers would be ruthless.
$1B is a small fraction of the cash they have on hand (and a small
fraction of the benefit they've seen from Linux, too.)

Likewise if there are patent issues around this - patents that SCO owns,
licensed (even just mutually) to IBM, then IBM implemented in Linux code -
IBM owns so many patents that to be favorable to them when they violate
someone else's patents doesn't seem fair.  Perhaps that would lead to IBM
calling for a weakening of patent law, ironically enough.

What's more interesting is if there's a less clear link - code that *looks
like* or *embodies some ideas from* SCO source code that IBM licensed -
then you're going to find SCO asking the court to define a new litmus test
for when software is a derivative work.  I think that will be very heavily
swayed by IBM's pro-open-source stance, and the army of people IBM can
bring in to help make that case.  I would wager that such a result would
mean that many companies out there afraid of opening their own code, or
using third-party code (open or not), will be much less worried about the
potential legal risks of doing so.  Thus, open source, & the idea of
sharing code, wins.