Python Cannot be Killed

Cliff Wells LogiplexSoftware at earthlink.net
Fri Jun 20 13:50:17 EDT 2003


On Thu, 2003-06-19 at 19:05, Alexander Schmolck wrote:
> Cliff Wells <LogiplexSoftware at earthlink.net> writes:
> > 
> > Anyway, my point being that no matter how one might feel about
> > Stallman's views, I don't think anyone can cast any real doubt on his
> > integrity.  
> 
> Not that my aim is to cast any real doubt on Stallman's integrity, but why
> can't I?

Because spreading FUD is evil.  You'll be forced keep one pinky near
your mouth whenever you say "one million dollars".

> > If the GPL changes, it will be to bring the letter in line with the spirit,
> > not to strip authors or users of their protections or rights.
> 
> While it seems indeed unlikely that the FSF will *maliciously* modify the GPL
> in a way that doesn't reflect the aims of authors of GPL'ed software, it is
> far less of a stretch to assume that the FSF could make changes that some
> authors would feel rather unhappy about.
> 
> Indeed the probability of the FSF changing licenses in such a fashion seems to
> be 1, judging by recent FDL issues:
> (<http://lists.debian.org/debian-legal/2003/debian-legal-200304/msg00256.html>)

Yes, this sort of thing is to be expected.  I also expect that it will
eventually be worked out in a reasonable fashion (FWIW, I side with the
Debian crew on this one).  If it isn't, then there is always the
recourse of forking emacs again <wink>

> > Yes, it *is* a matter of trust.  But then so is running any software you
> > didn't personally write, trusting users not to copy your software
> > illegally or the clerk at the store to give you the right change. 
> > Despite appearances, the world runs on trust.  
> 
> Let me remind you that I brought this up in the context of a zealot's claim
> that only the GPL "guarantees anything", whereas other licenses don't, which
> is patently false, even in the wider sense that BSD/MIT style licenses require
> faith in others' whereas GPL doesn't.

Okay, context accepted.  I only vaguely read that individual's post
because it seemed a bit incoherent.  I do agree with him that for the
most part, if you are concerned about protecting your code from being
usurped by a commercial company, the GPL offers more protections than
the BSD/MIT licenses (I think this is what he was saying <wink>), but
also agree that zealotry is best left out of the discussion.

> Me, personally, I'd under most circumstances rather place my trust on my
> licensees under the simple terms of a license like MIT which I both fully
> control and understand than on RMS's/the FSF's ability to judge what will be
> right for me, my code and the world in general.

The problem I have with the BSD-style license is exemplified by Mac
OSX.  Apple acquired a new OS gratis, and yet still won't provide even a
Quicktime player for other OSS systems (including BSD itself).  I find
that aggravating.

Consider also the recent discovery that several commercial routers are
running embedded Linux and include enhancements that have not been
released back to the kernel developers.  Since Linux is GPL, the kernel
developers have recourse to obtain those improvements, if they desire. 
The BSD license would have left them helpless.

> No doubt there are plenty of cases where the GPL is a good choice, but I find
> that sort of advocacy that tries to convince you that only the GPL can be
> "trusted" and promotes "freedom" a bit irritating.

Understood, and I tend to agree.  For libraries, utilities and the like,
I think the BSD-style license is fine.  For larger projects (like the
Linux kernel, GCC, etc), the GPL seems more appropriate.  The dividing
line for me would be how I would feel if some company used my code in
their proprietary system.  For small stuff, who cares.  Larger projects
need more protection.

> > For various reasons, I happen to trust people in the Open Source world more
> > than others and will continue to do so until proven wrong.
> 
> I appreciate your trust ;)

=)

Regards,

-- 
Cliff Wells, Software Engineer
Logiplex Corporation (www.logiplex.net)
(503) 978-6726  (800) 735-0555






More information about the Python-list mailing list