🚀 go-pugleaf

Many software developers would like to take GNU software and insert
it into proprietary products.

Sometimes they simply demand we let them sell our work, as if they
were entitled to it.  Or, less blatantly unreasonable, they say our
work is not "really free"--because it requires them to respect
freedoms of others which they would rather be free to deny.  (You
might as well say that a society is not "really free" unless people
are free to own slaves.)  Of course, the software these people write
isn't "really free" or free at all, but somehow the same standard
doesn't apply to them.

These statements are annoying but I doubt they convince very many readers.

Others go about it differently.  Instead of denouncing the FSF 
distribution terms as an ethical issue, they try to convince people
that other proprietary software developers (other than the one who is
speaking) would never use the software on our terms.

Of course, what other proprietary software developers will or won't do
is not really of great concern to such a person.  Rather, his hope is
to persuade us free software developers to let him do exactly as he
pleases--to ask nothing of him in the way of respect for the freedom
of the users.

Predicting that nobody will use the GNU libraries under our current
terms (whatever those happen to be) is a natural way of doing this.
While he can't prove these predictions, no one can disprove them
either.  Even if they are based on exaggeration, some people may
believe them.  And he has nothing to lose by saying them.

Thus, we were recently offered the prediction that no one would use
LGPL libraries if they have any requirement at all, no matter what it
might be, because managers would refuse to consider any concession
however light its practical cost.

Actually, there are reasons to think that many proprietary developers
(even if not all) will use our libraries under the LGPL.  One is that
many of them have received it enthusiastically.  Another is that the
requirement of offering tapes of object files on request is comparable
to the requirement for distributing GNU utilities in binary form.
Large companies have already accepted that requirement.

Of course, I can't prove this prediction of success any more than
others can prove their predictions of failure.  The sensible policy
for free software developers is not to try to predict, but rather to
observe the actual results which will happen over the next year or
two.  It will be much better to base our future plans on what people
actually do than on what someone would have us think other people will
do.