Canonical kernel contributions (was: Linux on the desktop...)

["Jarod Wilson" <user=74@nabble.blu.org>]

 On Wed, 2008-07-16 at 11:40 -0400, Jarod Wilson wrote: 
> On Wed, 2008-07-16 at 09:46 -0400, Daniel Feenberg wrote: 
> > 
> > On Tue, 15 Jul 2008, Ruben Safir wrote: 
> > 
> > >> standpoint, I really have no clue what they do in other areas as far as 
> > >> upstream contributions. However, I do have insider knowledge[*] that 
> > >> they do NOT send kernel contributions upstream to Debian or anyone else. 
> > >> Canonical's "fixes" and "enhancements" are more or less considered 
> > >> "secret sauce" and a competitive advantage. 
> > >> 
> > > 
> > > 
> > > That is a violation of the GPL 
> > > 
> > > 
> > The GPL only obliges Cannonical to supply source to anyone they supply 
> > with binaries. The enhancements are of course available to kernel 
> > developers, if they choose to use to read the Ubunto source. 
> 
> ...which most developers simply don't have the time to do. People are 
> supposed to submit patches to the lkml for review and inclusion, not let 
> them linger in some place where developers have to regularly poll and 
> poke to see what's new and interesting. If everyone did that, the kernel 
> would be in a sorry state, it just doesn't scale. 
> 
> > Anyone who has browsed the kernel mailing list can understand why a 
> > developer might choose not to submit patches - most submissions are met 
> > with insults from very possesive maintainers. 
> 
> Simply not trying because it might be hard is a cop out. But I don't 
> even think that's the case here, I believe its policy. Really, getting 
> stuff into the kernel isn't *that* hard. Sure, its not as simple as 
> "here, take this, trust me, it works", but it *can* be done. I'm no 
> uber-hacker or anything, and yet, there are more patches in Linus' tree 
> attributed to my name in the past year than there are attributed to 
> Canonical in 5 years.