to what extent is the gnu project philosophical?

[Andy Wingo]

Greetings, comrades :)

In the context of the recent discussions about what it means to be GNU,
how GNU should be organized, and about the virtues and risks of building
a more bottom-up governance structure for the GNU project, I started
wondering a bit more about the nature of GNU.

I think we all recognize that one of Richard Stallman's strong points is
a kind of strength of principle; of stubbornness, and of an ability to
put together inspiring, coherent arguments for free software.

However in the intertwined history of GNU and of the FSF, it's never
been quite clear to me when this work corresponded to GNU, and when it
corresponded to the FSF.  There is
https://www.gnu.org/philosophy/philosophy.en.html, hosted on gnu.org,
but administered by the FSF, which RMS was also the head of.  Indeed
some of those articles are written by people affiliated with the FSF but
not with GNU, for example the excellent
https://www.gnu.org/philosophy/when-free-software-isnt-practically-superior.html,
written by Benjamin Mako Hill, who was an FSF director and now AFAIU is
still a voting member, but not a GNU maintainer.

So my question is: is GNU fundamentally about producing coherent,
empowering free software systems, or is it fundamentally about
developing and propagating an inspiring, liberatory philosophy?

Of course the two of these exist in a kind of dialectic; one without the
other is not effective at writing a new history.  But they are different
kinds of work.

The answer to this question bears upon the future organization of the
GNU project.  If you consider GNU to be essentially a kind of moral
beacon, then it's less important how much and what kind of free software
you're producing.  On the other hand if you think that GNU needs to
focus on software production, then you might be willing to focus on the
practice and the product of GNU, without so much focusing on its
practitioners.

Some people argue that the ultimate strength of GNU is in the moral
rectitude of RMS, and to an extent they are right.  I am sure RMS
inspired all of us to join GNU, even if there are currently diverging
perspectives on how GNU should run in the future.  But do we need to
have the same kind of purity as RMS to continue the work of GNU?  If the
work of GNU is fundamentally philosophical, then perhaps yes -- maybe no
developer who uses a smartphone is suitable to be a part of GNU
decision-making, as someone who willingly accepts the compromise to use
a system having non-free software perhaps shouldn't be trusted to
expound the vision of a world in which all software respects the user's
freedoms.

                              * * *

If you will forgive the military metaphor, for my part I have always
seen GNU to be to the FSF what the People's Liberation Army of Namibia
was to the South West Africa People's Organisation: the armed wing of a
liberatory people's party.

The realm of ideas pertains to the FSF: theory, organization, advocacy,
and so on.

GNU, on the other hand, is about action in the software domain: the
construction of an ever-growing software commons, putting the theory of
the FSF into practice, and lending validity to the FSF's work.

If my description corresponds to what other think -- your thoughts
welcome! -- then the problem of identifying who is capable of
participating in the governance of GNU is made much clearer.  All you
need is a history of producing free software and a will to continue to
do so.

GNU and the FSF would remain closely linked regarding questions of what
would be nice to build and what must not be built, from the perspective
of enhancing software freedom, but the set of people that might be good
at organizing GNU might not be the same as those organizing the FSF.

Thoughts?

Yours in free software,

Andy