Re: A GNU Distribution

[Dave Crossland]

On 8 April 2012 16:32, Karl Berry <address@hidden> wrote:
>
> the biggest issue with making a
> GNU OS is simply that there has been no one with the time and energy
> needed to bring something into existence

For people to devote substantial time and energy to something requires
wage-amounts of money for those people, and that money has to come
from somewhere.

For the development of free software, the common perception is that
this money comes from paid work unrelated to the work of the developer
- that free software is distributed without charge by people who did
not charge anyone for their time to write it.

The most prominent free cultural project, Wikipedia, is made in this
way - very few people are paid to write encyclopedia articles.

But I believe that this perception is false.

Running the Wikipedia servers requires a small team of workers, and
the Wikimedia Foundation raises millions of dollars each year to pay
them for their time and energy, and for the resources required.

How, then, are people paid to work on writing free software?

Every time I have seen RMS give a public lecture, this topic comes up
in some form. These questions are often posed in a clumsy way, but if
the question is clear (maybe something like "Programmers need to be
earn a living, why doesn't that justify their proprietary restrictions
over us?" ;-) then RMS gives an equally clear answer.

My recollection of the themes in RMS' answers are that,

1a) He feels that a focus on business as the #1 value is misplaced and
the #1 focus should be on maintaining a free, just society, and

1b) how to do business is a minor detail in comparison to that, which
he has no interest in and nothing special to say about it;

2a) Many people earn a living writing free software but

2b) 'earning a living' is often a euphemism for getting really rich,
and they think they would not make it as rich with free software as
they would with proprietary software;

3) That if it is impossible to be paid to do something that is worth
doing, you should do it, and RMS would have waited tables rather than
work on proprietary software.

In talking to people in the audiences afterwards, I have found most
people discard point 1a because point 1b means that they do not have
practical, concrete ideas about how 2a occurs. Point 2a also gets
overshadowed by point 3, and people assume that it is impossible.

I think all these points are valid, and when the FSF paid people to
write strategically significant pieces of free software I think it was
easier for people to understand how 2a can occur.

(Probably when I feel I am done freeing fonts, I will make some
documentaries about how people do earn a living through free software
projects, and I'll be especially interested in the very small number
of people who have made it rich. Doing that while also making point 1a
may be tricky, but I think it is very important that people learning
software development know that it is possible to earn a living working
on free software - and that earning a living working on free software
is an attractive alternative to working on proprietary software in the
week and on free software at the weekend.)

For most software freedom activists, I believe their income does not
come from activity directly related to their software freedom
activism. It does comes from activity that is closely, but not
entirely, aligned with it; it comes from companies that promote the
open source philosophy. These companies employ people who participate
in free software activism, to write software that is available to the
public without charge under a libre license. The fiscal sponsors sadly
call the work 'open source,' but collaborating with such companies is
still okay, I think.

GNU participates in Google Summer of Code, for example, and I myself
am paid by Google to set up deals where Google pays font developers
around the world to make free fonts. Red Hat, Canonical, HP and many
other companies employ software developers full time to work on
publicly available free software.

Nothing about these money transactions is directly supporting
proprietary software. But such collaborations are problematic in that
the money earned by the companies that pays the individuals is earned
from proprietary software, or otherwise goes against the aims of the
software freedom movement.

In the above list of companies that distribute a lot of proprietary
software and a little bit of free software, the exception is Red Hat.
Last month Red Hat described itself in its public communication about
its recent historical earnings report as a "pure play open source
company" - by which I understood that it means that it does not
distribute any major pieces of proprietary software, any more (or
where it does, that is a vestige of an acquired company and it will
eventually become free software.) I think this is commendable policy
for a publicly traded corporation.

But solving some problems for the software movement and not creating
more to solve, does not mean actively working on the biggest
challenges of the software freedom movement. Red Hat makes software
that is useful for large businesses, and does not make software that
is useful for general purpose computing.

So. I believe that most free software written today is paid work. But
that work is not strategically significant to the software freedom
movement; and the movement does not have any explicit, coherent ways
of starting and funding projects that advance its aims strategically.

For example, a distribution of GNU.

Most people I know who do work on advancing the software freedom
movement's aims strategically earn wage-level amounts of money from
other sources. They see that while it may be impossible to be paid to
do something that is worth doing, they should do it without being paid
directly.

For example, I've read in the various books about the FLOSS movements
that RMS uses of financial investments and extremely low living costs
has allowed him to work on free software full time for many years now
without taking a paycheck from the FSF. Before that he did contract
programming on Emacs, and before that, sold copies of free software
(because this was before the public internet.)

Karl also found a way to retire early, and doesn't earn any money for
his work on GNU. (I am curious who is paid to work on GNU projects?
:-)

There is a good guide to the invest-and-live-cheap lifestyle model at
http://earlyretirementextreme.com although I personally don't have
experience with it; I inherited some money about 10 years ago that I
invested in gold. That enabled me to work on freeing fonts without pay
for several years, and the last year or so I have been paid wage by
Google Web Fonts project. Fonts is a area of free software included in
the GNU Manifesto but almost entirely neglected.

The FSF High Priority Projects List is an attempt to list some other
neglected areas, some strategically significant projects -
http://www.fsf.org/campaigns/priority-projects/ - and this shows very
little progress, year on year.

Of the 3 projects listed as 'completed' and taken off the list, there
is some hope in how to gather the time, energy and money in CiviCRM.
GNU PDF was removed because another project achieved the strategic
objective, unrarv3 was a small task taken on by an unpaid hobbyist. A
project for software to manage campaigns by a non profit, needed by
the FSF, was removed thanks to CiviCRM, which is a project that funds
its own development -
http://wiki.civicrm.org/confluence/display/CRM/Developing+with+the+CiviCRM+team

Of the GNU Free Distros List at
http://www.gnu.org/distros/free-distros.html, I am curious if anyone
here can explain how any of these projects work financially?

I haven't followed any of these projects closely in a few years, but
Karl says that all but Dragora have leveraged the sunk costs of
non-free distros. This makes them less relevant case studies than
ambitious non-distro free software projects that have become well
funded.

In conclusion, it seems to me that the FSF could add a GNU
Distribution project to its High Priority Projects List. To bring the
time energy and money needed by these projects to make them happen,
the FSF could more actively see funding for individuals who want to
make them happen, and help such individuals learn what is needed to do
fund raising for the FSF in order to pay for their positions.

Ideally this would become a feedback loop where enough funds are
raised to pay for the next priority project to progress.

Cheers
Dave