[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: GPLv3 suggestion to solve KDE/QT problem and others

On Sat, Feb 19, 2000 at 05:23:43PM -0800, Joseph Carter wrote:
> On Sun, Feb 20, 2000 at 03:15:40AM +0200, Adi Stav wrote:
> > It appears that the easiest way to solve many license incompatibility
> > problems involving the GPL and other free licenses would be to add a
> > new version of the GPL, since that would not be difficult (unlike
> > rewriting huge projects) and because the people who actually have the
> > power to do that support our ends (unlike the KDE team and Troll).
> Flawed premise.  I would be offended if the next version of the GPL
> allowed people to use my software with stuff I didn't intend when I wrote
> it.  So obviously a change to the GPL which allows KDE to suddenly be a
> non-issue doesn't support my ends.

Hmm, this is an FSF trust issue. 

Let's see: if a developer did not trust the FSF, or wanted the license
of their software never to change, they could have licensed their
program as "This program can be redistributed under the terms of the
GPL version 2 and none other" (I wonder how much software is actually
licenses this way? Hmm). If you really didn't really want people to
use you software in any other way than the way the GPL described at
the time you read it, this is exactly what you should have done... A
new GPL version wouldn't be published to fix a typo.

By allowing the users to use any later version of the GPL at their
option you are effectively saying "I realize that the GPLv2 may have
flaws and problems that are not known now but may be found in the
future. I trust the FSF, when finding such a problem, to publish a new
and better version of the GPL, which could then be used with my old
programs. I trust the FSF to make sure that the new version of the GPL
still helps working towards the same ends as I understood the FSF was
working towards at the time when I applied the GPL to my program."

Such a problem was now found. The question now would be whether the
new version of the GPL can still be considered to be working towards
the same ends as software authors could understood the FSF to be
working towards at the time (say, from the FSF web site and
publications -- not from the GPL itself, as this is the part that
could be changed).

I say it would. The distinction the FSF makes in its writings is
between Free and proprietary, not GPLd and non-GPLd. In contrast, the
GPL itself does not admit the existence of a difference between Free
and non-Free software, only between GPLd and non-GPLd. This has
probably been done because it is easier to describe it this way in
legalese, and because of the copylefted nature of the GPL which
requires that code remains GPLd.

So a reasonable person would have to assume that newer versions of the
GPL would change to become more and more helpful in encouraging the
use of Free Software. There would be no reason to assume that the GPL
represents the true moral stand and goals of the FSF -- those were
described quite clearly in other places, and the GPL *IS* expected to
change over time.

> I want a solution to the KDE problem as much as anyone, but not if it
> means changing the GPL just for KDE.  No way.
Not just for KDE! It's a general problem... It's an issue that
concerns every linking between the GPL and other copylefted licenses
and so in a sense it's a GPL issue. Today you cannot link GPLd code
with the (free!) Qt library. Tomorrow GNOME will want to use Gecko
libraries better but will not be able to do so because of too much
non-GNOME GPLd code. And next year Microsoft will release all of their
device drivers under some nice, friendly copylefted Free license and
Linus won't be able to integrate them into the kernel because the
license might not be GPL-compatible.

I'm probably not helping my cause either, but no license is perfect
and the GPLv2 is no exception. The QPL is probably a lot less perfect
than most, and many other licenses have other issues, and still they
are Free licenses. The GPLv2's main problem is that it is HIGHLY
incompatible with other Free licenses. And this incompatibility does
not address ANY important issue. None whatsoever. It causes
incompatibility within our own software, which is supposed to be the
peak of interoperability and reuse, and it cause our community to

There was a reason why the FSF recommended allowing users to apply
future versions of the GPL, and this is exactly it. It's a safety
valve for use in case when the world changes and the old version of
the GPL is no longer as good.

The GPLv2 was last edited in 1991. How could the FSF know? The entire
problem is just a remainder from the days when the world consisted of
GPL, BSDL and proprietary. Now the world is different, and it's time
to readjust the GPL.

Now, if to think about it in a more practical manner, will most people
really not want their GPLd libraries to be linked to by non-GPLd Free
software? Evidence suggests not, as this is exactly what is happening
right now, with the KDE copyright law enfringement. None of them
complained about the use. Some complained (very justly) that they
weren't asked for permission but no-one said that they would have
refused if asked.

Also, if the FSF considers this suggestion seriously, it would
probably be discussed well enough before a new version actually
appears, and we'll be able to see exactly how much objection this
receives, and from whom. It may even be possible to include an
exception in the new version that would make the new version identical
to v2 when applied for software written by specific authors who

> -- 
> Joseph Carter <knghtbrd@debian.org>                 Debian Linux developer
> http://tank.debian.net   GnuPG key  pub 1024D/DCF9DAB3  sub 2048g/3F9C2A43
> http://www.debian.org    20F6 2261 F185 7A3E 79FC 44F9 8FF7 D7A3 DCF9 DAB3
> "What does this tell me?  That if Microsoft were the last software
> company left in the world, 13% of the US population would be scouring
> garage sales & Goodwill for old TRS-80s, CPM machines & Apple ]['s before
> they would buy Microsoft. That's not exactly a ringing endorsement."
>         -- Seen on Slashdot

Thank you for bringing up this issue :)

	- Adi Stav

Reply to: