🚀 go-pugleaf

Here's a weird idea, prompted by contemplation of the fact that Bill Gates,
if he were feeling generous, could buy every programmer in the United
States a copy of BSD386, and still have money left over.

friedman@gnu.ai.mit.edu (Noah Friedman) writes:
>afford the slightly higher costs the next time you purchase GNU software,
>please consider it---the FSF could use the money.

Has the FSF ever considered a temporary compromise of its principles in
the short run in order to better achieve long term results?  In other
words, sell its software to raise money for development of free software?

Consider, for example, this fantasy universe, in which something like
the above was done early:

	It's the late '70s.  While most of us who hack on "real" computers
dismiss the new microcomputers as amusing toys, RMS realizes that he can
do a lot of good with a lot of money, and realizes that microcomputer
software is going to be big.  Thus is born GNU-Basic for the 8080.  Soon
to follow is GNU-DOS for IBM.  These are sold as commercial, supported,
binary-only software.

	Ten years later, RMS is worth about $5,000,000,000, and GNUSoft
dominates the industry.

	Now that he has money, RMS can do free software in a big way.
Interest and investment income bring in about $500,000,000/year for
RMS.  He takes this and hires 2500 programmers at $100,000/yr each
to write free software.  Another 500 or so people are hired at $100,000/yr
each to document and support the free software.  In addition, up to
2000 GNUGrants of $100,000 each are given each year to programmers not
included above to develop free software (i.e., if you want to develop
something neat, you submit a proposal and perhaps receive a GNUGrant).
In ten years, the only significant non-free software is software that
organizations develop for their internal use.  Everything else is free.

OK, back to our universe.  It's probably too late for anyone to become
a software billionaire, but suppose FSF operated in a for-profit manner
for several years, and then switched back to free-mode?  Would the net
result, say, 20 years from now be better or worse?

--Tim Smith