Re: Software as society (was: Re: Documentation Standards was Re: [ox-en] UserLinux)
- From: "Niall Douglas" <s_fsfeurope2 nedprod.com>
- Date: Sat, 13 Dec 2003 09:34:43 -0000
On 12 Dec 2003 at 21:28, Casimir Purzelbaum wrote:
- "practicality" -- doesn't the desire to "make the world a
better place" have mostly practical reasons? Is there no
practice of GPL-ing?
A pipedream is different to a realistic plan. I view the GPL as a
pipedream because it's not realistic. Why? Because there are clearly
superior alternatives available.
Also the mantra behind the GPL is riddled with logical typing errors
eg; that forcing anything linked with GPL code to reveal its source
as preventing exploitation of your code. If that were the case, why
hasn't BSD code been far more ripped off than it has?
The GPL is a lock-in - it permanently limits your future options -
and is thus kin to Microsoft-style methods of control and
exploitation. Business wants to avoid lock-ins which is why it's
diversifying off Microsoft - creating new options, but I'd doubt if
Microsoft will be going away any time soon.
It's interesting to compare Microsoft and Intel. Virtually no one
worries about being locked-in to Intel hardware though that wasn't
always the case. Before software became portable, even Microsoft
diversified Windows NT to run on four architectures only to remove
all but Intel again. I'd imagine something like that will happen with
Microsoft eventually as it has with Intel and IBM.
- cannot see GPL as the "de facto software license" -- this seems
to be related to the business point of view we were speaking of
Not just that - the GPL is bad for software just like closed-source
is generally bad for software.
- Why do you think, the GPL was bad for the quality of software?
(just because it is not as attractive for businesses
'investments' as other licenses are? -- business-point-of-view
again? Or do you have other points? (Pointers appreciated))
The single best short-term way to improve the quality of software is
to create more reuse of it so that less reimplementation of
functionality happens. The GPL prevents everyone reusing GPLed
software. Therefore, it's inferior to say the LGPL which has no such
As I've mentioned in previous posts, I don't see much step-change
innovation happening in volunteer-based projects. Since the GPL isn't
very commercialisable, it keeps the door mostly shut on introducing
capital which has a much better chance of creating step-change
innovation as mutual consent is no longer required. I take the view,
as supported by reams of evidence from biology, that regular step-
change innovation is the best way to improve quality in the long-
term. Here I don't mean quality in terms of less bugs but rather its
usefulness and applicability as an engineering solution.
However, I recognise that it is as popular as it is because the
legal support for software is so lousy so the GPL looks attractive
on the surface. Patching up a leaky tire is one approach - far
better is to replace the tire.
(Sorry to say that, but this reminds me of your statements about
fantasy and practicality and of their relationship to reality ;-)
All the things the GPL states it wants are admirable, thus people
don't see the negative consequences - though RMS is on record as
saying that he doesn't much like copyright law as applied to software
and without copyright, the GPL wouldn't be enforceable.
We need a systematic reworking of the legal support for software and
indeed all digitally representable information. What we need is
pretty clear - how to achieve it without removing the traditional
patterns of domination, control and exploitation is far harder. Once
again, western global economic hegemony is preventing our world from
moving on :(