Torvalds licensing ignorance

Posted Mar 22, 2007 10:51 UTC (Thu) by Wol (subscriber, #4433)
In reply to: Torvalds licensing ignorance by bug1
Parent article: The Torvalds Transcript (InformationWeek)

It's actually quite easy to relicence a project ... particularly when it has a high code churn like Linux.

All it takes is Linus to say "I'm relicencing my code as 'v2 or v3'", and to demand that all new code is licenced the same.

Then you contact all the other "v2 only" people and ask them to relicence their old code.

Finally, you wait (probably not that long) for all the "v2 only" authors that couldn't or wouldn't relicence to be written out of the system. And if you publish their names I'm sure a janitor project would rapidly spring up to do that ...

Bingo. The kernel is now "v2 or v3" right through, and can be changed to "v3 only".

Cheers,
Wol

Torvalds licensing ignorance

Posted Mar 22, 2007 16:57 UTC (Thu) by viro (subscriber, #7872) [Link] (4 responses)

Sure, I'll even help you to start that list: Al Viro <viro@[anything]>
There's your first entry, now you have something to rewrite. Quite
a bit, actually, so by the time you are through with that you'll
probably have more list entries submitted. Have fun.

Torvalds licensing ignorance

Posted Mar 22, 2007 17:22 UTC (Thu) by bronson (guest, #4806) [Link] (1 responses)

Hang on, Al. I agree that the GPLv3 as of the last draft is really overreaching. But what if someone writes the license that should have been the GPLv3? A respin, not a ground-up rewrite. Call it GPLv2.1. It fixes up a few ambiguities in GPLv2, clarifies some language for international courts, but is otherwise is substantially exactly the the same as the GPLv2.

Would you be against relicensing your code to that?

Torvalds licensing ignorance

Posted Mar 22, 2007 17:41 UTC (Thu) by viro (subscriber, #7872) [Link]

I don't believe that FSF will go for that, at least not until they
*really* give up all hope for v3 being feasible. And probably not
even then, due to perceived loss of face. IOW, it's even less likely
than relicensing of the kernel. Sure, if we run into real trouble
with v2 (i.e. serious failure in court somewhere) we'll have to do
something that would fit your description. But that's (a) unlikely,
as far as we know and (b) it would be v2 + explicit modifications,
not FSF-blessed v2.1.

Torvalds licensing ignorance

Posted Mar 23, 2007 4:48 UTC (Fri) by ldo (guest, #40946) [Link] (1 responses)

Why do I get the feeling somebody thinks he's indispensable?

Torvalds licensing ignorance

Posted Mar 23, 2007 6:08 UTC (Fri) by viro (subscriber, #7872) [Link]

Not at all. So, are you volunteering for rewrite? Of course it's
doable, just... will take some work.