But something you said is exactly the basis for SCO's suit against IBM:

>>> ... certainly it isn’t right to use someone else’s code and distribute it for free if the origional creator didn't want it used that way.

SCO says IBM took its intellectual property (the SMP code from UnixWare) and contributed it to the 2.4 kernal (under GNU terms, making it everybody's to use) w/o SCO's permission. If SCO can demonstrate that (which they haven't yet), it's absolutely wrong and IBM owes SCO big-time.

Actually, that is as far as I know incorrect. Basic SMP support was developed several years ago and really got off the ground after Caldera sent Alan Cox one of the early SMP machines. Yes, the same Caldera that is now The SCO Group. Caldera also worked for a several years on making UnixWare and Linux more compatible.

IBM's contributions have been mentioned often before, amongst which we find JFS, NUMA, and RCU. None of these actually exist in UnixWare, but they were developed by Sequent and later IBM on a kernel that was based on the SysV codebase. Now NUMA is (at the moment) pretty useless for the average user, you really have to get into the massive multiprocessor systems like IBM mainframes.

JFS is a bit superfluous, the kernel already has 3 perfectly fine journalling filesystems. ext3 allows for easy migration for ext2 systems. reiserfs has btree directories and is fast for massive file creation/deletion. XFS is simply impressive. In my eyes, the only interesting part of JFS is for existing AIX users that want to plug their harddisks into a linux box.

So that leaves RCU, which only got merged into the kernel during the 2.5 development kernels. So if 2.4 infringes, then this can't really be the offending code that has 100's of identical lines (yet).

The SCO/IBM lawsuit really seems to be about a contract dispute, who owns the code that was developed by Sequent/IBM. SCO claims that they have distribution (copy) rights as a result of the code being developed as a/on a derived work of the SysV codebase.

For the rest they are throwing up various smokescreens, most of these seem to be anticipating on actually winning the contract dispute.

If they win, they get distribution rights to some parts of the code. And if they do not agree to distribution under the GPL, the 2.4/2.5 kernels will become unlicensed and the exclusive right to copy is with (all) the original authors/owners of the code base. Now those guys can decide whether you can or cannot copy on an arbitrary basis.

So run-time binary licenses are just plain stupid, your binary kernel is unlikely to even contain any NUMA or JFS code, and running a kernel is not violating any distribution rights. How you got the kernel possibly was not legal, but that is really the problem of the distributor (RedHat/SuSe/kernel.org).

Anyone who buys an SCO license clearly agrees that he knowingly violated the copyright of all kernel developers during distribution (aka. bought stolen goods). As a result each kernel developer can independently sue his ass off for copyright infringement. But that's the only thing the SCO license gives you as there is no such thing as 'a right to execute a program' only 'a right to copy/distribute a program'.

My guesses as far as the 'hundreds of lines of identical lines of code'. Possibly example Intel/AMD/chipset initialization, standard textbook functions (allocators), or BSD or Sysv4 derived code. Or code copied from Linux when Caldera was working on their Linux/UnixWare integration strategy. Besides, there probably are few people who have access to UnixWare, but pretty much everyone has complete access to BSD/Linux code.

Finally, it might just have nothing to do with the actual IBM suit which seems to be more focused on obtaining rights to code that isn't even part of UnixWare. Perhaps they are blowing smoke to inflate stock prices, maybe they are trying to distract the opposing lawyers so that they don't focus all of their defense on the actual issues at stake.

This is just my uninformed opinion