[plug] SCO publishes code references, case sucks, big surprise

[Leon Brooks]

    http://www.groklaw.net/article.php?story=20040207061940758

They've basically identified files containing code written and 
contributed by IBM and containing references to RCU, NUMA and JFS.

They've dropped copyright and trade secret claims, and are now down to 
under 1000 lines, *none* of which was written by The SCO Group or any 
of its ancestors.

Their entire case now rests on contract violation by IBM, and the terms 
of IBM's contract quite clearly say that they're allowed to do what 
they've done. Santa Cruz' contract with Novell also gives Novell the 
right to waive any of old-SCO's claims, either by directing old-SCO to 
do so, or by doing it themselves (both of which Novell have done).

Here's a classic snippet: in this piece of C code, they are claiming 
that the name "RC_PLOCAL_usertrap" proves that this was leaked from 
Sequent code, even though "cpu_number_map" and "smp_processor_id" are 
Linux-specific and kind of inevitable:

    RC_PLOCAL_usertrap(cpu_number_map(smp_processor_id()))++;

To further add to the confusion, this one word is used to cover 
"copying" into kernel/i386/trap.c *and* 8 lines of init/main.c in two 
groups.

Now the big question: will their stock go up on Monday because they 
released something tangible, or down because the stock traders can hear 
their techies laughing themselves into a stupor?

Cheers; Leon