Re: Debian i386 freeze
Manoj Srivastava wrote:
>>>"Oliver" == Oliver Elphick <olly@lfix.co.uk> writes:
...
> I think you misunderstand the argument. The license that is
> being used does give users of the binaries (if any binaries are
> distributed) the right to modify everything that is needed to create
> the binary -- and this includes the interface definition files
> (headers) from QT.
The kde licence cannot give any such right, because kde are not the
authors of Qt. kde don't claim any such right, no-one imagines they
either claim or have it. This is really not sensible!
>
> That violates Troll licence, really, for QT; this is teh
> reason we may not distribute KDE binaries.
>
> Oliver> The GPL (applied by kde) can only apply to kde's own code,
> Oliver> not to anyone else's. If it is necessary to link some
> Oliver> non-free software, in the knowledge of kde (which is the
> Oliver> case), they must necessarily have given consent to do this by
> Oliver> the very fact of releasing their software and inviting the
> Oliver> world to use it. In addition, one of their developers has
> Oliver> specifically given consent on this mailing list.
>
> Umm, this is streatching it. The right to modify QT headers is
> not theirs to give; a pure GPL implies it for binaries; we, if we
> distribute the binaries, are equally at fault.
But it is evident to (almost) everybody that this is not pure GPL. Either
kde regard Qt as part of the normal distribution, or they implicitly
waive this particular requirement of the GPL. Whichever is the case,
their actual position, and therefore the terms of the actual licence,
are crystal clear. [Written licences are there to help, but they are
not definitive. Oral statements and behaviour are equally valid,
provided that they can be proved.]
>
> Why can't all this implied knowledge be written down, up
> front, by the KDE folks? Consent on the mailing list is not as good
> as written consent in the package/license it self (I do not want to
> argue, in court, whether the person sending an informal mail message
> actually represented the developers, or not).
How would you prove it came from the developers? What sane court would
entertain this case for half a minute?
> Oliver> People keep losing sight of the fact that the licence is
> Oliver> applied by the authors of the software. It gives the right
> Oliver> to use the software under the conditions stated. The licence
> Oliver> is not a contract; it is completely one-sided. It cannot be
> Oliver> enforced against a supplier by a recipient, because it is NOT
> Oliver> a contract. It can only be enforced by the software authors.
>
> We can only distribute the binaries if we meet the terms of
> the license. One of the terms of the license is that modifications of
> *all* the source code is permitted.
with the [implied] proviso that it does not apply to Qt.
> If that violates the Troll
> license, then our distribution of the binaries is illegal.
given the proviso, it doesn't
>
> We do not know if the authors shall never sue us, for indeed,
> if they were so committed, they would have amended the license.
I think they are totally unable to understand what you are going on about.
>
> Oliver> The purpose of the GPL is to enable software authors to keep
> Oliver> their software free. It contaminates any derivative works,
> Oliver> so that they in turn have to be under the GPL. However, it
> Oliver> cannot work backwards, otherwise Motif would now be GPL'd
> Oliver> too!
>
> Motif, on Solaris and other Unices, is deemed system software,
> that is, it is exected to be on the system as shipped. In which case,
> you can link GPLed software. You are right, though, I do not think we
> can supply Motif linked software, since Motif is not shipped with
> Linux.
Again, any packages built with Motif can be shipped, unless they are
derived from GPL'd software whose authors refuse permission for it to
be linked with Motif. This is because of the implied term in the author's
licence (GPL with implied modifications) that it is OK to do so.
>
> Oliver> This is where the frustration comes in - people do not seem
> Oliver> able to understand that the licence, whether GPL or
> Oliver> otherwise, is wholly under the control of the authors.
>
> Oh, absolutely: they have the rights; and seems to me those
> rights prevent us from distributing binaries, since we, the
> distrivbutors, have to follow the license.
But we follow the real licence, not the one that you insist on; any
author who really meant what you try to make them mean would need
to be committed to an asylum!
>
> Oliver> It can be varied or withdrawn at any time. This is inherent
> Oliver> in the authors' rights under the Berne Convention. I think
> Oliver> that the convention might well allow an author to withdraw
> Oliver> software from free use, even if it had previously been put
> Oliver> under GPL. The author might publish an announcement
> Oliver> withdrawing all licences, and this would (I think) be legally
> Oliver> binding; he would be unilaterally withdrawing a right that he
> Oliver> had unilaterally granted. Of course, he would have trouble
> Oliver> enforcing this, because of the difficulty of ensuring the
> Oliver> announcement was seen by everyone affected.
>
> Umm, no. I think not. You may be able to put out a new
> version under a different licence, but if licenses were allowed to
> change from instant to instant, then licensees would have no
> protection at all, and the whole system would be useless.
Article 9:
(1) Authors of literary and artistic works protected by this Convention
shall have the exclusive right of authorizing the reproduction of these
works, in any manner or form.
Article 12
Authors of literary or artistic works shall enjoy the exclusive right of
authorizing adaptations, arrangements and other alterations of their works.
The Convention does not say that rights once granted must be maintained.
They are entirely within the author's power, so I believe that it is
indeed possible for an author to withdraw a GPL licence.
> Oliver> (If the author had supplied software for a consideration, he
> Oliver> would not be able to withdraw the licence to his customer,
> Oliver> because then there would be a contract enforceable by the
> Oliver> customer against the author.)
>
> I think that money changing hads has little to do with whether
> there is a contract or not. At least, not under US law, afaik. So
> mocrosoft money is under a license, which you can't break, even if
> there has been no change of money.
I would be surprised if US law had diverged so much from English law.
Consideration is of the essence of a contract. If there is no consideration
there is no contract. A licence is a licence, not a contract.
So, you can have a contract to grant a licence, but the licence itself
is not the contract.
--
Oliver Elphick Oliver.Elphick@lfix.co.uk
Isle of Wight http://www.lfix.co.uk/oliver
PGP key from public servers; key ID 32B8FAA1
========================================
"For God so loved the world, that he gave his only
begotten Son, that whosoever believeth in him should
not perish, but have everlasting life." John 3:16
--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: