What I'm finding disturbing this morning is that none other than Bradley Kuhn of the Software Freedom Law Center has apparently added his voice to the chorus of those claiming that "Microsoft violated the GPL!" What's disturbing is that Bradley certainly knows better.
Stephen Hemminger, the Vyatta engineer who initially turned up the issue (and no one is claiming that there was no potential compliance issue with these drivers) denies that there was a violation. Microsoft has likewise denied that there was a violation (but, of, course, Microsoft is Satan Incarnate...)
Just to remind the folks who should know better (and enlighten the folks who have not actually read the GPL), a violation will occur under the following set of circumstances:
- Microsoft distributes a binary copy of a "program" (in this case, the Hyper-V drivers), which includes GPL-licensed code, to a third-party, sources not included.
- The third party makes a request to Microsoft for the sources corresponding to the binary they received.
- Microsoft fails to provide those sources, in a timely fashion, for no more than the cost of creating the media and shipping it to the requester.
The SFLC has absolutely no business making such a claim unless they're prepared to support it, and the only way they can do so is by producing the third-party mentioned above, something which has so far not happened. The SFLC should issue a clarification immediately: they've called Microsoft liars, which is about par for the course, but they've also called Stephen Hemminger a liar, which is completely unreasonable.
A FURTHER CLARIFICATION FOR THOSE WHO DON'T
READ COMMENTS BEFORE POSTING THEIR OWN
Microsoft could have failed to comply by either the sequence of events noted above, or by failing to include an appropriate notice of how to get sources. There are assertions (and I don't personally know for a fact that there isn't an appropriate offer buried in the wad of stuff one gets with a Windows Server 200 license, I don't have one) that no such offer was tendered by Microsoft.
Assuming, for the sake of the discussion, that this was indeed the case, then the copyright holder of the code being infringed (so far unidentified) could indeed have cause to complain about a violation—since the copyright holder is the only one in a position to actually do anything about it—but it would be difficult for him-or-her to start complaining after the sources had been published and submitted, and the code was brought into compliance with the license's terms.
Which seems to be exactly what Bradley's doing, albeit at second hand. Usually this sort of rhetoric gets saved for when the SFLC files an actual suit, but it's being dragged out here when such action was avoided, thanks to Microsoft's deciding in this instance to play by the rules. That the SFLC is doing this strikes me as either grandstanding or ambulance-chasing after the case has already been settled out of court. My comments about "sore winners" still stand.
directhex has pointed out that what Microsoft evidently did in the Hyper-V drivers is to have a GPL-licensed driver shim, which pulled in a hitherto-proprietary binary blob at boot time, no different than what Nvidia and ATI do with their drivers. About which no one is screaming "Violation! Violation!", even though neither Nvidia nor ATI appear to have corrected those situations...
Which of these things is not like the others...? This seems to be some sort of a trick question...
READ COMMENTS BEFORE POSTING THEIR OWN
Microsoft could have failed to comply by either the sequence of events noted above, or by failing to include an appropriate notice of how to get sources. There are assertions (and I don't personally know for a fact that there isn't an appropriate offer buried in the wad of stuff one gets with a Windows Server 200 license, I don't have one) that no such offer was tendered by Microsoft.
Assuming, for the sake of the discussion, that this was indeed the case, then the copyright holder of the code being infringed (so far unidentified) could indeed have cause to complain about a violation—since the copyright holder is the only one in a position to actually do anything about it—but it would be difficult for him-or-her to start complaining after the sources had been published and submitted, and the code was brought into compliance with the license's terms.
Which seems to be exactly what Bradley's doing, albeit at second hand. Usually this sort of rhetoric gets saved for when the SFLC files an actual suit, but it's being dragged out here when such action was avoided, thanks to Microsoft's deciding in this instance to play by the rules. That the SFLC is doing this strikes me as either grandstanding or ambulance-chasing after the case has already been settled out of court. My comments about "sore winners" still stand.
AND SOME NEW INFORMATION,
CASTING FURTHER DOUBT ON THE WHOLE THING
CASTING FURTHER DOUBT ON THE WHOLE THING
directhex has pointed out that what Microsoft evidently did in the Hyper-V drivers is to have a GPL-licensed driver shim, which pulled in a hitherto-proprietary binary blob at boot time, no different than what Nvidia and ATI do with their drivers. About which no one is screaming "Violation! Violation!", even though neither Nvidia nor ATI appear to have corrected those situations...
Which of these things is not like the others...? This seems to be some sort of a trick question...