Re: GLADE and GPL (was: what DOES the GPL really say?)

[dweller@news.imagin.net (David Weller)]

In article <dewar.867445762@merv>, Robert Dewar <dewar@merv.cs.nyu.edu> wrote:
>Dave says
><<I've been following this thread for a little bit, and I'd like to toss
>in my opinion (what th' heck, right?).  The significant difference
>between the GPL and the ACL is much like the difference between
>capitalism and communism*.  The GPL assumes all users are sinners, and
>takes the approach of requiring the users to affirm, by usage of
>GPL'ed code, that they will not unfairly exploit the free software
>they are taking advantage of.  There's all sorts of legal stuff in
>>>
>
>This is complete nonsense, and frankly I am a little surprised at the
>level of misunderstanding (I would have thought Dave understood the
>philosophy behind the GPL better).
>
Not complete nonsense at all, your entire post captured exactly what I
meant.  I understand the GPL quite well, and in fact once strongly
considered using the same modified GPL that ACT uses on some of their
sources.  Unfortunately, I had _way_ too many people and companies say
they wouldn't touch my code if it was under GNU anything.  Sure, I
could have taken the time and effort to convince them otherwise, but
the ACL basically removed the "heavy" parts at the expense of risking
proprietary digressions (although it's not as bad as RObert points out
-- the ACL commonly referes to a "standard distribution" -- if you buy
a proprietary version that isn't derived from the standard
distribution, or if you download a release that's not a "standard
distribution", _you_ assume the risk).  All in all, I have enither the
time nor the energy to combat FUD.  It's hard enough to get the
components done in a reasonable amount of time!

Robert's thoughts are correct though, I understand the GPL/LGPL _very_
well :-)  My company frequently (and legally) uses GPL and LGPL
software on an almost daily basis without worries, fear, or paranoia.

>The business about the GPL assuming all users are sinners is fanciful
>stuff, but bears no relation to reality.
>
It wasn't meant to bear a relation to reality, it was a metaphoric
statement.  I think anybody could have realized that.

>Going back to the Booch components, Dave is of course free to choose any
>approach he likes for his work, but in practice the only difference between
>the ACL and the use of a broadened GPL such as is used by GNAT is that
>it makes it possible for someone to produce a proprietary version of
>these components based on Dave's work.
>
True, but the ACL places limits on how you can make a "proprietary"
version.  In particular, you must indicate it clearly enough such that
the obtainer cannot confuse what they are getting with the "standard
distribution".  The "proprietary" sense also comes from the fact that
a compiler vendor may replace the implementation with a "tuned"
version, but ensuring that the library is still
"specification-compliant".  

>proprietary version that you have to pay for and cannot get full source
>access, then we have lost something valuable.
>
>That's why I think it is unfortunate to use the ACL instead of the modified
>GPL for such projects, it seems freer, but can very easily lead to much
>less freedom.
>

Now who's speeading FUD? :-)

>Note another scenario which is even worse.
>
Scanario deleted:  Incompatible with the ACL.  Go back and reread
section 3.

>THe whole point of the GPL is entirely pragmatic. The idea is to make
>as much software as possible as freely available as possible, because this
>free availability benefits users.
>

The ACL supports those goals also.  There certainly would be less of a
need for the ACL if people were less paranoid about the GNU licenses.

>The idea of free software is to promote an environment in which the
>computing community can get its job done more effectively. The tremendous
>success of Linux shows that this idea can be a powerful one. More and more
>people are switching to using Linux, not because they want to join some
>polictical movement, but because it is the best technical tool for the job!
>

Well, Perl is doing quite well under the Perl Artistic License, so it,
too, is a model for success.  That's what the ACL is based on.  

I think Robert and I have VERY similar philosophies overall.