IBM Lawyer Hired Away by Microsoft
and a Second Analyst Unconvinced
by SCO "Proof"
I saw this little notice on Law.com:
"After handling IBM's intellectual property matters for nearly 30 years, Marshall Phelps Jr. has jumped to Microsoft Corp., where he will oversee more than 11,000 trademarks and 3,000 U.S. patents as vice president and deputy general counsel for IP."
He was VP for IP and licensing at IBM. Hmm. How convenient. Do you suppose he knows anything that might be of value to anybody rooting for SCO?
Another interesting tidbit: An analyst in Germany who looked at SCO's proof code and is talking, saying that their lawyer forgot to ask him to sign an NDA, says, I think (my high school German is a lot rusty), that he can't be sure if there is any substance to SCO's claim. One reason is there's no way to verify if the UNIX code is actually System V, because the code was pulled out of any traceable record, and all dates, even on the comments, had been removed. Also, it appears to him that while there were similarities in functionality, there were differences in implementation for some of it.. One section seems to be identical, but who knows which came first, the chicken or the egg? He also noted some differences. My German isn't up to this task, so take a look yourself at the German version here. Google's English translation is here. And here is how Sherlock translates it, with the usual somewhat comical results:
"I had today the possibility of looking at me the loading code paragraph. By a mistake on the part of the representing law office my colleague and I, in contrast to the 7 other assigned one, did not have to sign which received today insight, a discretion assertion. Completely contrary to the examiners of the company Microsoft, who must protect silence obviously even opposite their own superiors, and only the internal audit department in relation to report to refund may. Now, to the code: Under notarial supervision 46 pages A in each case one half code from Linux (to a large extent printed out Posts directly from the Linux Kernel mailing list) and the other half of listings of SCO were submitted. Whether it actually concerns thereby SysV sources, is like that not to be reconstructed, since they tore from the connection. Interesting is also, which all dates were removed from both, even from the comments. The coming AR are actually in parts identically, even some jokes are alike on both pages. Remarkable is however, which is nevertheless quite different to to most corresponding places the source code which is before coming arene. The fundamental structure of the complained of functions is similarly, the concrete implementation however nevertheless quite differently. Variable ones and function names are different, Schlefien are structured differently, conditions run over chain queries or bit pattern. All in all only one can be surely said: The functions supplied by the respective code paragraph are often alike, which was to be expected in addition, from the front in here. In the concrete implementation there are not however so many differences that a proof of the same origin will become difficult, although reliably not possibly. Crack point is however a function of the scheduler, which is on a length of scarcely 60 lines actually identical up to small differences. Here there is also a whole number of corresponding comments. Comparable similarity indicates only one routine of the store management, with which there are comments however only in the Linux version at all. Whether alone from these two agreements however a judicious proof can be created, only one lawyer with Sichrheit can estimate. The vague similarities in other sections hold I for, since for both the same standards to reason were finally situated, certain an agreement thus to expect insufficiently are. Which concerns however the same comments for however different source areas, I can make myself on it no Reim. That would have to be again examined in any case more exactly, in particular also with available dates. Because with it a violation of the clock jack right could only be proven at all. Which concerns the discussion around from SCO/Caldera themselves under the GPL sold section of Linux, then must be said, which had to decide so far still no court on the right security of the GPL. If this should be determined however, what is not safe, then SCO can consult only the sections from Linux to the comparison, which did not publish it and in their development or advancement it was not involved. Also this regard I as a difficulty in the coming procedure. Since however the original, ungepatchten Linux sources, but only modifications, which inserted most diverse Distributoren, were not attacked in any case is to be still clarified whether these evtl. Right ones at the deplored paragraph have, either directly or indirectly e.g. by firm unions, transfers, "all inclusive" Deals u.ae. The chances on a procedure actually which can be opened are however not particularly high, since one agreed so far in most comparable cases out of court. This is however only my personally estimate, obligatory is in any case only the decision both pages of the representing lawyers or then possibly the responsible court."
If anyone can provide a better translation, please do.