🚀 go-pugleaf

Many people on the pro-FSF side of this discussion seem to be saying
that if one believes GNU libraries, such as G++, should be usable
without legal encumbrance, then one is just a wrong-headed software
hoarder.  I think this is perhaps too strong a stance.  In my own
view, the the libg++ policy sacrifices long-term goals of the FSF in
favor of a short-term moralistic stance.  I believe that the long term
goals of the FSF are best served by targeting specific
well-understoood software tools, and producing extremely high-quality
free replacements.  This tools will be at least equal with commercial
competition in terms of quality, and immeasurably superior in terms of
price.  Thus, the targetted products will be removed from the arena of
competition, allowing software producers to get on to the serious
business of actually advancing the state of the art.  I think this
strategy has is working marvelously for editors, and may yet work well
for C compilers.  I think the libg++ policy will prevent this strategy
from working with C++ compilers.  Though it is of course technically
possible to use G++ without libg++, in practical terms you can't have
one without the other.  Given the existence of competition from, for
example, AT&T, the benefits of libg++ aren't going to convince any
commercial enterprise to chuck the concept of software ownership.
Instead, they're going to by AT&T 2.0.  Thus, the libg++ policy helps
to keep commercial enterprises selling proprietary versions of software
available from the FSF in business.  Is this worth the moral purity of
the policy?

Well, I am not a free software zealot, or a zealot of any kind -- in
fact, I kind of think Bush is doing the right thing regarding
Lithuania.  Maybe I should get some more zeal about some things.  My
point here, though, is that the point is nowhere as cut-and-dried as
poeple are making it out to be.  In fact, in private communciations
Doug Lea has admitted to some mixed feelings about the current policy.

For what it's worth...

--
Dave Detlefs			Any correlation between my employer's opinion
Carnegie-Mellon CS		and my own is statistical rather than causal,
dld@cs.cmu.edu			except in those cases where I have helped to
				form my employer's opinion.  (Null disclaimer.)