Slashdot Mirror
SGI's Letter to the Linux Community
_Upsilon_ writes "SGI has released a letter to the Linux community in response to SCO's recent threat to revoke the UNIX licence for Irix. The letter mentions that they inadvertently did submit some System V code into the Linux kernel, that has since been removed (and some more in the process of being removed). The article points out that the code fragments in question had already been released into the public domain as well."
"found brief fragments of code matching System V code in three generic routines (ate_utils.c, the atoi function " they're complaining about copied atoi code?!?!?! exactly how many variations of converting ascii to integers could there be? And why would anyone care about something so simple? Unless it was a FUD screen...
SGI admitting to having contributed some System V code to the Linux kernel brings to mind a question I've had for years: what are the chances that in a big closed source project (like Windows for example), that some developer hasn't used some open source code at one point or another? How do you protect against this?
Is there a process to audit big companies code? MS threatens me with audits to check my license compliance, can I audit them to check that no open source code is in their products?
And for that matter, which license would win? If GPL'd code was found in a product like Windows, would Microsoft be forced to open source the entire thing?
"Following this occurrence, we continued our investigation to determine whether any other code in the Linux kernel was even conceivably implicated. As a result of that exhaustive investigation, SGI has discovered a few additional code segments (similar in nature to the segments referred to above and trivial in amount) that may arguably be related to UNIX code. We are in the process of removing and replacing these segments."
It would be nice if they (IBM, SGI) ran the ENTIRE Linux/GPL code base through a tokenized comparison with their reference SCO Unix trees. Thus GPL community could start identifying and removing any suspect code NOW, rather than wait for a trial outcome.
It's legally valid because you didn't licence your car from Ford -- you purchased an object. Things like cars and toasters aren't (yet) subject to licencing (just wait til your toaster has software on it though... betcha the box has a EULA).
You did, however, licence your linux software from the developers who wrote it and if they stole code, you, as a licencee, may be responsible, not for damages, but for payment of licencing fees on the code that was misappropriated.
We could argue all day about wether software *should* be considered identical to a physical object, but at present it is not. You don't buy software, you buy software licences, always.
So in summary, yes, it *could* stand up in court, not as a liability issue, i.e. damages for past illegality, but as an issue of paying a licence for the use of SCO IP. This presupposes that they can prove that Linux actually carries any such IP and that the version of the kernel you happen to be using contains any of it. The first is highly questionable and the second will be moot about a week after any infringing code segments are actually identified.
(the latter case, incidentally, does nothing to help IBM in their contractual dispute which *is* a liability issue and *is* about past damages. If infringing code is found, even if the code is removed from the linux sources, those responsible for the misappropriation, presumably IBM, will be held liable for damages. However, users should be free and clear, though linux's reputation would be substantially tarnished.)
[i am, of course, not a lawyer, but i'm pretty sure my understanding is correct.]