Re: Significance of the GP licence.

[Alan Mackenzie]

In gnu.misc.discuss David Kastrup <dak@gnu.org> wrote:
> Alan Mackenzie <acm@muc.de> writes:

>> Quite simply, that it is the GPL itself which is the main reason for
>> the popularity of Linux amongst the people who write it.

> Well, that's half of the story.  Linux has been written to support a
> preexisting GNU userland.  And that userland has a tradition of being
> popular and freely available quite before Linux.

What is the reason for that popularity (amongst developers), if it's not
the GPL.  GPL vs. BSD license was one of the few big differences between
the projects way back then.

> And BSD became freely available only some time after GNU/Linux.

Yet how does that explain why Linux is so much more popular amongst
developers than a BSD kernel?  BSD became freely available at a very
early stage of the development of GNU/Linux, early enough to catch up on
its merits.

> The GNU userland is unpopular among BSD developers because, well, they
> are BSD developers.  And because their kernel of choice already comes
> in one package with a userland.

The BSDs include some GNU stuff, just as GNU/Linux includes some BSD
licensed stuff.

> So quite a lot of popularity of GNU/Linux comes from GNU, and not
> necessarily just because GNU is GPLed.

Would you argue that GNU would have become just as popular (amongst its
developers), had it been licensed under something like the BSD licence?
I would doubt that very much.

>> If, for some currently inconceivable reason, Linux was relicenced
>> under what you call a "free as in freedom" licence, many developers
>> would cease development.

> The Linux kernel developers tend not to be all too religious about
> licensing.  Well, they do, but they call their religion pragmatism.

That pragmatism being that they can get on with development without
bothering too much about the licence, which they know they can trust.
How happy would these folks be about being unpaid hackers for, e.g.,
Apple?

>> This might leave a mere rump, scarcely larger than the groups which
>> maintain the BSD kernels.

> I doubt it.

Well there's little prospect of that experiment taking place, thankfully.

-- 
Alan Mackenzie (Nuremberg, Germany).