🚀 go-pugleaf

One example of an "acceptable compromise", in my view, would be
for FSF to officially offer a mechanism for licensing code for
derivative works that could be released without GPL.

A typical scenario:  a company would like to hire a programmer
or programmers to develop a text editor or database to their
own specifications.  It would clearly be an advantage to use
emacs as a starting point, but the self-perpetuating aspect of 
the GPL entails a "long-term" commitment.  If the investment of 
time and effort is going to be considerable and the end result
very different from the starting code, most companies will prefer
to begin with code that can be licensed outright.

This is a waste of effort for the company and a wasted opportunity
for FSF, which could make money to fund on-going development in 
other areas.  FSF would still have the option to refuse to license
code for purposes that clearly run counter to its interests, and 
they can choose to clearly limit the conditions under which the
licensed code can be re-used outside GPL.

Another example would be the case of an academic whose grant
is contingent on producing code that can be placed in the 
public domain, or be otherwise available to be modified 
according to the needs of the sponsor.  FSF could choose to
permit the academic to use FSF code in a derivative work in
exchange for a fair share of the grant money and perhaps
some kind of "no commercial use without permission" clause.

I'm not suggesting that FSF compromise on projects like 
non-redistributable ports of current software, but flexibility
on the use of FSF code for entirely new projects would get
more people working with FSF code and perhaps provide an
alternate source of income for FSF.

-- Bill (johnston@me.udel.edu)