On Thu, 1 Nov 2001 goran.hultgren at bluefish.se wrote:

> Hi Andrew and all!
> 
> [Throwing myself back into the fray discussing licenses with a lawyer
> (ex?) - am I suicidal or what! :-)]

Yes, but don't let *that* stop you! :0

> "Andrew C. Greenberg" <werdna at mucow.com> wrote:
> > On Thursday, November 1, 2001, at 04:00  AM, goran.hultgren at bluefish.se 
> > wrote:
[snip]

> > Well, not to support the analogy, but there is little doubt that "free 
> > software" engineered so you can't do things with it isn't particularly 
> > free.

Actually, there seems to be *plenty* of doubt :) One thing that, IMHO,
that tends to get lost in these discussions generally, but that is of
special interest to users of monolithic images, is that ordinarily the GPL
tends to restrict *modification* and *distribution* type uses of
software. For something like a word processor, that's a very small
minority use. For programmers, that's a pretty bit use. Even then, using
GCC to write your programs doesn't push the GPL onto those programs.

As you, Andrew, have pointed out, monolithic images present a rather big
difficulty, since modification uses and "end user" use are not separate
categories (at least, in the eyes of the GPL).

This is, depending on your views, unfortunate.

> >  The use, indeed pedantic insistence by RMS, on using newspeak in 
> > lieu of English and reason to describe things proves this more clearly 
> > than anything else.

Hmm. While RMS gets in folks face, I don't think it's sheer newspeak, nor
do I think that there's a general lack of English and reason supporting
his uses. To my casual observation, he doesn't seem to be actively
revisiting that English and reasons in current disputes, but that's a
somewhat different problem.

> >  With all due respect, the authoritarian regime of 
> > the GPL,

Authoritarian *regime*? Andrew, clearly this is a bit of hyperbole at
best, and perhaps more likely it's own bit of newspeak.

RMS is certainly insistent, domineering, etc. but he doesn't always insist
on the GPL (though he strongly insists on GPL compatibility, sometimes
that amounts to the same thing). But RMS and the FSF don't exert much more
than moral/rhetorical force (plus, of course, the legal provisions of the
GPL). Compared with the monopoly force exerted by Microsoft (and the
standard licencing provisions of many many companies), this is hardly
*authoritarian*.

> > whose function is to constrain and limit what may be done with 
> > the software,

Well, *ideally* constrain and limit by preventing other constraints and
limits. If right to modify programs in my possession trumps right to
distribute without source, then a licence that preserves the latter is
preserving the *right* freedom. (Perhaps harder for the GPL, one might say
that the right to modify programs in my possession trumps the right
redistribute under an open source (ah, now *that's* newspeak :)) licence
of my choosing. That seems less appealing.)

> > as opposed to the Berkeley license, whose primary function 
> > is to pass the software along, shifting risk to the licensee, reserving 
> > the bare bones minimum obligation of acknowledgment, makes clear to all 
> > which is the free license.

No, not really. Again, it depends a lot on which freedoms are valued and
how the net effect is calculated. And freedom isn't the only measure as
one might want to calculate overall good. I *do* think that RMS is a bit
more focused on public good than on (libertarian style, at least) freedom,
so, in that sense, the use of Free often misleads. 

> > The fact that you can't use GPL in a monolithic image, unless you can 
> > relicense EVERYTHING ELSE IN THE IMAGE under GPL proves this clearly.

No, it doesn't. It proves, at the very most, that the GPL is broken with
regard to monolithic images. That RMS may well be *happy* with this
brokenness is a slightly different issue. Personally, I don't think there
*is* a true inherent notion of "derive" for software that one can
satisfactory put all the derivations on one side and all the "mere
uses" on the other. The linking criterion is espeically tendentious,
imho. But that seems to be a bug, not a matter of the intent
or "purpose" of the GPL.

Of course, I'm confused, in your arguments, as to whether we evaluate
intent or consequence when judging a licence "free". (I'm curious as to
this notion of *fuction*. I.e., the "function" of the GPL is... vs. the
"fuction" of the BSD is...) I think the distinction between a "buggy" free
licence and a non-free licence is significant.

Anyhoo.

Cheers,
Bijan Parsia.