[sword-devel] Closed source exploitation of open source works (a GPL loophole)

[Paul Gear]

Chris Little wrote:

> > > Further, since non-free
> > > software developers do not give back to free software, free
> > software should
> > > avoid assisting non-free software development by providing it
> > free work and
> > > finished tools.
> >
> > I think this is where you overstate the spirit of the GPL.  I
> > don't think the
> > point is to avoid helping proprietary software, but to benefit
> > free software.
> > Any detriment to proprietary software is incidental to the GPL's purpose.
>
> I don't state that detriment to non-free software is a goal of the GPL,
> simply that avoidance of assisting in the development of non-free software
> is.

I was equating the two.  Perhaps that was insufficient distinction.

>  Furthermore, you are incorrect in your assessment that my statement was
> overreaching.
>
> I quote below an excerpt from "Why you shouldn't use the Library GPL for
> your next library" by Stallman, from
> http://www.gnu.org/philosophy/why-not-lgpl.html, which I had in mind while
> making my statements.
> ...
> The point of the GPL is that those willing to play by the same rules of
> free, open source software are allowed to benefit from the work of others.
> Those who are unwilling are not.  If you want to argue this further, take it
> to Stallman.  I'm simply reporting his views, which are very clear.

OK - you win.  RMS wants proprietary software to suffer at the hands of free
software.

However, i think that:
a)  it's not likely to be an issue here, and
b)  the free (in the FSF sense) Bible software movement wouldn't be really
affected by the presence of a proprietary Bible package based on Sword anyway.

> ...
> Further, back to the question of whether this constitutes a loophole, I
> can't see how any of you can deny that it is.

I guess if you take as one of the aims of the GPL the prohibition of any
benefit to proprietary software resulting from free software, i guess it must
be.  However, i believe the GPL strikes a good balance as it is.

I know this is ground that has been gone over many times by many people, but
let me offer here my perspective on the GPL & LGPL.  (You can find a more
detailed expression of the philosophy behind my views at
<http://gear.dyndns.org/OpenBible/free_software.html>.)

For me, the primary benefit of the GPL (and LGPL, for that matter) is to keep
free software free.  It means that no one can take the package and turn it into
proprietary software just by saying so.  If proprietary software benefits from
it (as arguably it has from the proliferation of, say, Linux), that is fine
too.  If people are willing to pay for software, let them pay for it.

IMO, the only real difference between the GPL & LGPL is that the former is
targeted at applications, and the latter at libraries.  I do not believe that,
for example, BibleTime has to be GPLed, even though it depends on Sword, which
is GPLed.  The reason for this is that section 2 of the GPL applies to it.  It
is a separate work, and is distributed separately.  Therefore, it can use any
license it wants.  Is anyone aware of whether this has been tested in the legal
system?

This is supported by the following statement in section 2 of the GPL:

> Thus, it is not the intent of this section to claim rights or contest your
> rights to work written entirely by you; rather, the intent is to exercise the
> right to control the
> distribution of derivative or collective works based on the Program.

The above statement is fairly clear as to what the aim of the GPL was not.  It
seems you are right about RMS' desires, but i think he has changed them (or
perhaps just expressed them more intensely) since the GPL was written.

> ...
> Do we want to go with LGPL?

Personally, i don't see the need for the LGPL at all.  It seems to me that it
arose out of a perception that you couldn't write a proprietary application
using a GPLed library, which i think the above-quoted section of the GPL shows
to be false.  Again, does anyone know about any legal cases or other evidence
that might dispute this?

The LGPL says (in the preamble):

> The reason we have a separate public license for some libraries is that they
> blur the distinction we usually make between modifying or adding to a program
> and simply using it. Linking a program with a library, without changing the
> library, is in some sense simply using the library, and is analogous to
> running a utility program or application program. However, in a textual and
> legal sense, the linked executable is a combined work, a derivative of the
> original library, and the ordinary General Public License treats it as such.

It would seem that there is some sort of legal precedent (or at least the
suspicion of one) for applications that use GPLed libraries being regarded as
derivative works.  The only conceivable way such a situation could take place
is if they were statically-linked with the library.  By using dynamically
linked libraries (and thereby allowing the works to be distributed separately
as per the GPL section 2), all of these issues are bypassed.

As i mentioned previously, i disclaim any legal expertise in relation to these
licenses, but i think that it is a reasonable interpretation.

Paul
---------
"He must become greater; i must become less." - John 3:30
http://www.bigfoot.com/~paulgear