SCO offers UnixWare licenses for Linux [LWN.net]

SCO offers UnixWare licenses for Linux

Posted Jul 21, 2003 17:36 UTC (Mon) by stark (guest, #13095)
In reply to: SCO offers UnixWare licenses for Linux by dwalters
Parent article: SCO offers UnixWare licenses for Linux

You're really missing something critically important here.

Hypothetically :

If I submit code through LKML to Linus and it gets into the
kernel, and I say "hey, I wrote this, it's cool and works
wonders on the thingamabob" and everyone's happy, and it's
all GPL licensed and everyone uses it, then by your definition
everything is fine and OK.

This is, obviously, patently false, ignorant, and misleading.

The problem is that there is a trust layer in that I said
"I wrote this" when I may, in fact, have stolen it from SCO.

Then the GPL licensed code contains tainted material and it
is wrong (morally, ethically, and legally) to distribute that
GPL licensed code base to the world.

So if IBM submitted code to the Linux kernel that IBM stole from
SCO then IBM has committed fraud and the Linux kernel must remove
that code and nobody is allowed to use that code without SCO's
permission.

So you should be saying "how come SCO is charging me insurance to
use something without proving that it belongs to them?"

I don't think they could win a court case against an end user
unless they can prove that the end user knew the source code
was tainted in some way, so until they release a code fragment
we should all be safe :)

The problems that people _should_ be harping on are :

1 - Does SCO actually own the copyright? If not, then IBM
simply stole it from someone else, so this isn't entirely
a good thing.

2 - Is SCO harping about specific code or IP? If it's IP that's
vaguely defined, then of course we can bash them for being so
frivolous, but if it's a specific line of code then there's
an issue. The problem of "why haven't they released the
code fragments as proof?" is a separate issue.

3 - Has the Linux kernel ever, from any source, not just IBM/SCO,
taken any source code for which we're unsure of the original
source? If so, we should fix that ASAP! An OpenBSD style code
audit would work wonders for the Linux Kernel source tree!

4 - I seem to recall early on in this SCO claiming that the source
code wasn't in the kernel.org sources, but in other sources
that IBM helped develop (I immediately thought of Red Hat's
Advanced Server product.)
Perhaps they are right, and there is a line of code, but it's
not in the kernel and it's only in some vendors' modifications
to the kernel, only in one specific distribution.

In the end, I'm reserving judgement till the code fragments in
question are identified and the issue is resolved. There's simply
too much acrimonious commentary going on here (and seeing as my
employer is under Chapter 11 right now, I don't need any more
acrimony :)

(Log in to post comments)

SCO offers UnixWare licenses for Linux

Posted Jul 21, 2003 19:51 UTC (Mon) by brouhaha (subscriber, #1698) [Link]

1 - Does SCO actually own the copyright? If not, then IBM simply stole it from someone else, so this isn't entirely a good thing.

What??? If SCO doesn't own the copyright on the code in question, it does NOT follow that IBM stole anything.

For instance, IBM almost certainly owns the copyright to the read-copy-update (RCU) code. This code was developed by Sequent who was acquired by IBM. SCO claims that their contract with IBM gives SCO ownership of any modifications to Unix that IBM develops. But this is based on an interpretation of the contract clause that is very unlikely to be upheld by a court, IMHO. There's a remote chance that SCO might have some limited rights to RCU, but not exclusive ownership.

SCO offers UnixWare licenses for Linux

Posted Jul 23, 2003 19:47 UTC (Wed) by cpm (guest, #3554) [Link]

Hypothetically;

The trust layer is always a variable.
Yes, the straw man "you" could have stolen
the code and included it for review and
had it later adopted.

Be we are not talking about a straw man,
we are talking exactly about IBM.

Nothing SCO has said yet about this
so-called case has made any sense.
Everytime and I mean everytime it looks
like they might get pinned down they
change the story, changing the rules.

Can you come up with an example of a developer
who would have been the type of person
who did this?

Yes, SCO has stated that it is the Kernel
code they are talking about, then they
have stated it is not the kernel code.

If it isn't the kernel, it isn't Linux.
Linux is the kernel.

does SCO own the copyright? What copyright?
The whole history of the "Unix wars" is
so convoluted that the very few who understand
them have a very difficult time explaining
them.

As to whether SCO is harping about lines
of code, or IP, well, that changes. Depends
on which formal complaint.

Yes, it is acrimonious.

SCO has gone from how many to how many in
how long? What do they do now? Peddle licenses
and lawsuits. They are dead. I think it's
not just a little disingenious to think
that the whole MS/SCO license agreement a
back room deal to try to harm the entire
linux movement.

It's a very old trick, "Let's see 'em deny it".

Works really well in the press and doesn't cost
much.

SCO is nothing but big bosses and lawyers anymore.
This whole "IP" thing is so perverse it's sick.