On Thu, 2005-01-20 at 10:21 -0800, Greg KH wrote: > On Thu, Jan 20, 2005 at 10:02:24AM -0800, Sean Hefty wrote: > > Greg KH wrote: > > > > >Personally, I think it's a stupid thing to try to license this code in a > > >dual way, as any port someone is going to have to do to get this code to > > >work in another os will be almost a complete rewrite in the end > > >anyway... > > > > I think that companies want to be able to make derivative works without > > needing to make the derivative open source, versus porting it to > > another OS. > > And then run that derivitave work on a Linux GPL kernel? Hah, good luck > with your lawyers if you try to do that. And good luck trying to work > around symbols that the openib code is using that are marked > EXPORT_SYMBOL_GPL(). > > Why do people try to do such stupid things, haven't the IB members > learned from the past... > > Remind me to _never_ send in an openib kernel patch if this is the > reason why the license is what it is. > The idea was for folks to be able to take the OpenIB code, under BSD, and port it to OSes other than Linux. I agree that having a BSD only stack running on Linux would be silly and should be avoided. In the end I only care about Linux. I think performance should drive what Linux features OpenIB uses. If it improves performance then it's worth using.
- Matt _______________________________________________ openib-general mailing list [email protected] http://openib.org/mailman/listinfo/openib-general To unsubscribe, please visit http://openib.org/mailman/listinfo/openib-general
