| |||||||||||
SCO's Evidence: This Smoking Gun Fizzles Out
Eric S. Raymond has made available his
analysis of SCO's case, delving into the similarities among various
versions of the malloc code, including the detailed differences between the
Linux and SVR4 versions of malloc/atemalloc. It also explains why the GPL
is *not* an applicable defense in this case.
(Log in to post comments)
SCO's Evidence: This Smoking Gun Fizzles Out Posted Aug 21, 2003 22:21 UTC (Thu) by BrucePerens (subscriber, #2510) [Link] I hope Eric realizes that the PowerPoint file is now available. He just did the two slides Heise photographed.bruce
SCO's Evidence: This Smoking Gun Fizzles Out Posted Aug 21, 2003 22:46 UTC (Thu) by FarcePest (guest, #3065) [Link] Fizzled out? Looks like it's smoking to me.
SCO's Evidence: This Smoking Gun Fizzles Out Posted Aug 21, 2003 23:21 UTC (Thu) by dwalters (subscriber, #4207) [Link] It would be interesting to see Eric's perspective on the other slides, in particular, the theory that if any code created by a Unix licensee ever touches Unix, SCO owns that code from then on, and can deny its creator the right to make use of it for any other purpose.This must really be the "meat and potatoes" of SCO's case against IBM (at least in terms of the sheer number of lines of code in question).
SCO's Evidence: This Smoking Gun Fizzles Out Posted Aug 21, 2003 23:01 UTC (Thu) by dwalters (subscriber, #4207) [Link] From Eric's analysis: More generally, SCO's case is, to say the least, somewhat impaired by the fact that SCO itself freely released all of the code of ancestral Unix. Under the conditions defined by the Ancient Unix license, issued by SCO/Caldera itself, the Linux kernel developers could copy the entire ancient Unix source tree into the kernel and SCO/Caldera would have no proprietary claim on the results whatsoever. Not quite.
Of course Eric makes several other very good points, such as the fact that most enterprise Linux installations, upon which SCO are seeking licensing fees, almost certainly don't contain this particular example of offending code. Eric points out many holes in SCO's case, but isn't it disingenuous to claim that "SCO/Caldera would have no proprietary claim on the results" in this instance?
SCO's Evidence: This Smoking Gun Fizzles Out Posted Aug 21, 2003 23:42 UTC (Thu) by rfunk (subscriber, #4054) [Link] He's updated it to version 1.1, which now addresses this issue.Basically it goes back to the ruling in the AT&T/Berkeley lawsuit.
SCO's Evidence: This Smoking Gun Fizzles Out Posted Aug 22, 2003 0:11 UTC (Fri) by dwalters (subscriber, #4207) [Link] OK, so any Linux code that was copied from 32V appears to be in the clear, even without attribution, because 32V is indeed the public domain (because of the pre-1996 copyright laws, which AT&T carelessly didn't pay enough attention to).Based on Eric's very thorough analysis, it's almost certaily true that the "slides 10-14" code in question was copied from 32V (and not System V, or BSD) into Linux 2.4.19. So it's quite possible that the person or company that decided to copy the malloc code from 32V to 2.4.19, did so because they believed the code was in the public domain. However, they had no business slapping an SGI copyright and a GPL notice on it (but that fact does mean that they misappropriated anyone's legitimate copyright). As an aside, this also means that Caldera really had no business releasing 32V along with all the other historical versions of Unix in 2002, because, according to the AT&T-vs.BSD ruling, it no longer belonged to them anyway ("Plaintiff has failed to demonstrate a likelihood that it can successfully defend its copyright in 32V"). I think this clears up any doubts that the "slide 10-14" claims are bogus, and the "slide 15" BPF code is a no-brainer, so that only leaves the "all your derivative works are belong to us" issue, as far as the PowerPoint presentation is concerned, and that seems to rest on some very shaky legal ground, as well as apparently incorrect claims about the development path of some of IBM's enterprise code (like JFS).
SCO's Evidence: This Smoking Gun Fizzles Out Posted Aug 22, 2003 1:02 UTC (Fri) by Arker (guest, #14205) [Link] It still contains this paragraph, which I don't believe is true: More generally, SCO's case is, to say the least, somewhat impaired by the fact that SCO itself freely released all of the code of ancestral Unix. Under the conditions defined by the Ancient Unix license, issued by SCO/Caldera itself, the Linux kernel developers could copy the entire ancient Unix source tree into the kernel and SCO/Caldera would have no proprietary claim on the results whatsoever. Perhaps it's technically true, I don't think they could claim the kernel because of infringement, but it's disingenuous to imply that the licenses are compatible if it's not.
Re: SCO's Evidence: This Smoking Gun Fizzles Out Posted Aug 22, 2003 7:54 UTC (Fri) by Ross (subscriber, #4065) [Link] I agree that the licenses (BSD+advert & GPL) aren't compatible. But it isthe GPL which says that no additional restrictions may be added. The BSD license does not forbid that. So the copyright holders which could legitimatly prevent distribution of software which is composed of both types of software are the copyright holders of the GPL'ed code. In this case, that would be the kernel developers and not SCO. (Once again, IANAL so don't consider this legal advice.)
|
|||||||||||
