[Linux-aus] SCO position, rationale and AUUG

[Greg 'groggy' Lehey]

On Thursday, 22 May 2003 at  8:56:49 +0800, Leon Brooks wrote:
> In a way, I'm surprised that Microsoft didn't just buy SCO/Caldera
> outright,

Maybe that's not their goal?

> but perhaps they have in mind waiting until SCO's value is less than
> that of sand in the Sahara and meantime making friendly overtures
> until the axe falls.

Take a look at their stock prices.  The stock has really rallied since
the lawsuits were announced.

> Yes, Michael Still, I think rolling AUUG in would also be a good
> idea, pursuant to which I've added their "talk" list to this
> message.

Note that it's a tossup as to whether the correct people are on talk at .
I don't know how many of our Board of Directors are.

> I value AUUG not so much for their press contacts (although those
> are good) as because they have a set of valuable points of view
> which are centred around different concerns to either SLPWA or LA's.

This is a good point.  It might be a good enough reason for AUUG and
Linux Australia to put out similar (but not contradictory) statements.
It would certainly increase press coverage.  The AUUG BoD is meeting
on Saturday, and this item is on our agenda.

> If no Australian Linux or FOSS organisations agree on the content or
> obviously want to take part, I'll make it a release from
> CyberKnights Pty Ltd. I would prefer to see a harmonious public
> chorus from our Linux, FOSS and business interests.

I don't see why CyberKnights shouldn't make a statement too.

> === PHRRRK, PHRRRK, IS THIS THING ON...? ===
>
> [Assorted organisations] are coming under increasing membership
> pressure to respond to accusations of code plagiarism from the Santa
> Cruz Operation (SCO).

I don't think that SCO is the Santa Cruz Operation any more.

> The extensive quality control systems which a patch must pass before
> it is accepted into the main Linux kernel would massively modify and
> almost certainly reject any incoming SCO UNIX code. The internal
> systems of SCO UNIX and Linux are quite different, so it would make
> little sense to try grafting code from one kernel into the other.

The second point is much more plausible than the first.  Recall that
SCO is saying exactly the opposite about quality control.  Before you
can say that Linux QC is better than SCO, you need to convince your
audience.  And that takes too long.

> The legal complaint filed by SCO against IBM contains many
> substantial errors of fact which leads [assorted organisations] to
> the conclusion that SCO have made a serious mistake in their
> evaluation of the situation, and that they should withdraw their
> suit against IBM.
>
> One error of fact centres on using Linux with large arrays of
> processors. The collaboration which led to Linux running on
> 64-to-512 processor systems such as SGI's Altix 3000 involved many
> people and companies - not just IBM - and resulted in Linux
> performance which dramatically outclasses SCO UNIX.

I think we need to look at where, if at all, SCO UNIX runs on
significant numbers of CPUs.  Sure, various versions of AIX, IRIX and
Solaris do, but you can be pretty sure that that has nothing to do
with the SCO code base.  But this is where it gets tricky: SCO could
say, for example, that the AIX development is based on SCO's code (not
true AFAICT), and that IBM's know-how (not necessarily code, recall?)
makes Linux better.  Turn that the right way, and just before you
strangle the last grain of truth, you might make a case which some
people would believe.

> SCO have freely published the source code in question (under the
> terms of the GNU General Public Licence (GPL) which protects the
> Linux kernel) for some time after filing the complaint, so even if
> the complaint once had any merit, it no longer does.

We don't know that.  They don't say how long.  They don't say that the
code which IBM supposedly introduced was actually in their distro.  It
might just have been in the development kernels.  Since they don't say
what it was, it's difficult to counter.

> Microsoft have recently purchased a UNIX licence from SCO to
> "improve the Unix compatibility of [...] Services For Unix", but
> while Microsoft would evidently like to stop shipping GPLed software
> itself for political reasons, their move does not validate SCO's
> charges.

This purchase of a UNIX license is really confusing.  By all accounts
Microsoft has various UNIX source code, so they must have had a
license.  After all, when SCO was effectively a part of Microsoft,
they wrote or at least maintained XENIX.

> We also note that Microsoft have dropped support for some of their
> software on SCO UNIX, but support those programs on Linux.

The point?  That even Microsoft doesn't believe in SCO's viability?

> Linux's future is bright and certain, underwritten and stabilised
> more by the political and technical freedoms embodied in the GPL
> than by the fickle to and fro of shareholders or the dangerous flame
> of personal or corporate ambition.
>
> It is quite safe to continue using Linux. This is true in terms of
> virus immunity and general reliability as well as in the legal and
> political senses.

Wearing my devil's advocate horns, this somehow doesn't ring true.
I'm not sure why, but I suspect that it won't be enough to counter
SCO's FUD.

You'll note that I've criticized what you have to say--I hope you'll
agree with the points I made--but I haven't said very much myself.
There's a good reason for that, and it's also the reason why AUUG
hasn't made a stand on the subject: I'm still trying to understand the
implications.  But we need to be very careful about what we do say.
Of one thing I am certain: this lawsuit *does* have the potential to
do serious damage to Linux.  Look at the effect on BSD of the AT&T
vs. BSDI lawsuit 10 years ago.  FreeBSD and NetBSD weren't even
involved in that lawsuit.

I do have a web page on the subject, at
http://www.lemis.com/grog/sco.html.  Note that these are my personal
opinions and have no relationship to any official AUUG standpoint.

Discussion welcome.

Greg
--
See complete headers for address and phone numbers
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 187 bytes
Desc: not available
Url : http://lists.linux.org.au/pipermail/linux-aus/attachments/20030522/9b61edb5/attachment-0001.pgp