Re: [OT] Re: configure script

[David Kastrup]

eliz@is.elta.co.il (Eli Zaretskii) writes:

> On 13 Jun 2002, David Kastrup wrote:
> 
> > eliz@is.elta.co.il (Eli Zaretskii) writes:
> > 
> > > On 12 Jun 2002, David Kastrup wrote:
> > > 
> > > > The people involved in the GNU project don't get a say about this.
> > > 
> > > And your opinions--are they based on some poll of the GNU contributors 
> > > and maintainers?  Or are they just _your_ personal views?
> > 
> > My opinions are my own.
> 
> Then all we have is a "he said, she said" argument.  Richard has his 
> views, you have yours.  For all I know, most or all GNU maintainers could 
> support Richard, not you.
> 
> > I suggested to Stallman that he might come up with
> > a definition about who could be considered part of the GNU project,
> > so that it would be possible to do such a poll and thus clear up the
> > authority with which and for whom he was speaking.
> > 
> > His answer was that nobody has a say in this rather than he as the
> > person that put up the GNU project, so such a poll would be pointless.
> 
> If you really cared, you could start a poll yourself.

What I care about is limiting the damage RMS manages to inflict on
the GNU project.  Since RMS has stated in advance that the outcome of
such a poll could and must not change the way in which he feels fit
to fight for his cause, it would be an exercise in futility.  The
whole "Open Source" splitoff did not came about because people did
not care about Free Software: a certain indication of that is the
prevalent use of the GPL on such software.  It simply came about
because people became embarrassed to be working for the cause of Free
Software under Richard's banner and auspices.

Now Richard is not the one to concede one inch of his position for
mere "political" reasons.  And that has always been one of his
strengths.  For that reason I am trying to persuade him to assume a
less damaging stance by pointing out that his position could be
considered quite inconsistent even without considering the damage I
find it causes.

> > Stallman is, of course, entitled to his personal views, but I
> > would prefer if he labelled them as such.
> 
> Richard is the leader of the GNU project, to say nothing of his
> personal contributions.  He is also the president of the FSF, which
> works on the GNU project.  If anyone is entitled to speak on behalf
> of GNU, Richard is.

Sure thing.  But questions that remain: Is anybody entitled to speak
on behalf of GNU?  If so, with what authority?  Richard's answer is
that he nobody else is entitled to speak on behalf of the GNU project,
and that his speech is not to be encumbered by the opinions of any
contributors.  The next question is: what encompasses the range of
the GNU project that Richard is entitled to be speaking for?
Richard's answer to that is: simply anything that could be a
component of a GNU system.  So while he reserves the right not to ask
anyone about it, he assumes speaking with the authority and
legitimation of about every Free Software author or contribution
towards GNU systems, including the X window system, the BSD
networking utilities, the Linux and Mach kernels, the various system
utilities and so forth and so on.

> That doesn't necessarily mean you and I should agree with everything
> Richard says (you and others know how many times I disagreed with
> him on various issues), but swerving the argument to discussing
> whether he can or cannot say ``we'' is IMHO a waste of time.

The problem is that people stop taking Stallman seriously when he
indulges in petty infighting like this.  This would not in itself
matter so much if the fight for Free Software had already been won.

Unfortunately, it hasn't, and the vigilance and fight for freedom
never can be abandoned.  One of the next critical points I see coming
up involved the infrastructure in the toolchain.  At the current point
of time, GCC is the accepted compiler for both the Linux kernel and
the complete GNU systems being shipped with it.  Most GNU utilities
are pretty much agnostic to the compiler they are being compiled with,
and that is a good thing.  The Linux kernel isn't.  Because of its
prevalence, compiler manufacturers will start to support the necessary
GNU compiler extensions and provide alternatives where this is not
feasible.  Intel's compilers generate considerably better code than
GCC does.  It is only a matter of time until it becomes feasible to
compile complete GNU/Linux systems with proprietary compilers.  This
would mean that the resulting systems could not be modified and
bootstrapped without having to acquire the necessary proprietary
tools.  An awful step backwards from today's situation.  The solution
is to convince distribution maintainers not to switch to superior
non-free technology.  Only in that case, hardware producers like
Intel will keep an interest of keeping GCC more or less up to par.
And given the speed of evolution of today's hardware, manufacturer
support is important.

This involves decisions based on ideals instead of mere technical
merits.  And some of the things that Richard chooses to focus on are
not terribly suited to convince people that his ideals are suitable
for adoption if they lead to the results he comes to.

I would not want to compromise Richard's principles, but it is my
opinion that his chosen ways of fighting for them could be less
contraproductive.

It is precisely because I strongly care about Free Software and the
GNU project that it maddens me to see people getting alienated to
their causes for no reason I would deem compelling.

-- 
David Kastrup, Kriemhildstr. 15, 44793 Bochum
Email: David.Kastrup@t-online.de