FSF is creating a problem that never existed!
Posted Oct 2, 2006 22:18 UTC (Mon) by nim-nim
In reply to: FSF is creating a problem that never existed!
Parent article: Busy busy busybox
> The kernel is under a GPLv2-only license, and as such outside of the
> reach of the FSF
So no one dragged the kernel developpers in the debate, they could work with the FSF on its new project, or ignore it altogether.
I see your point on non-kernel code, but you know, I don't find claiming a position of authority as a core kernel developper to discuss the problems of other projects much more tasteful than what the evil things FSF is supposed to have done.
> Have i understood your point correctly, that you deny that the FSF has
> unilateral legal power to change the contribution dynamics of that huge
IANAL so I won't speculate on the value in international IP law of a license preamble. Especially if you add a vague "contribution dynamics" to the mix (though a judge would probably agree that any person following Linux development knew perfectly well what "GPLv2 or later" meant after the numerous "GPLv2-only" clarification messages of Linus)
Moraly, as far as I know the GPL v3 is perfectly consistent with the GNU manifesto, so should the FSF relicense code which has been assigned to them to the GPLv3, I don't see who could honestly complain (of course legally the FSF can relicense it to whatever it wants to)
Practically, for code which hasn't been assigned to them the FSF can at most argue they're allowed to create a GPLv3 fork, and this fork won't fly unless :
1. they convince the major existing contributors
2. they write enough useful new code
(true both on the enforcing license and maintaining code front)
And this is mostly how big decisions are taken in FLOSS forks anyway. The orbital laser power of the "GPLv2 or later" clause seems greatly overrated to me (except for simplifying the management of past or minor contributions). The basic rule doesn't change a FLOSS fork which pisses of its contributors is a dead fork.
On the dynamics front, should the FSF write a pleasing license, it will certainly impact the contribution dynamics of both new and existing projects. IIRC one of the "GPLv2-only" clarification messages Linus wrote argued that :
1. he didn't blindly trust the FSF but
2. the code churn of the kernel was sufficient for relicensing, should most of the developpers agree with the new license
So I certainly do not agree with the idea the GPLv2-covered codebase is so big it's impossible to pull a GPLv3, even for the Linux kernel.
Nevertheless what ultimately counts is the quality of the actual GPLv3 text. Arguing about things the FSF may or may not have said (and never enacted) behind the scenes is pointless. There's an easy way out for everyone to get a good GPLv3 out of the door. The FSF already pulled its baby past the point where a little flaming could kill it.
to post comments)