Re: ocaml, QPL and the DFSG : QPL 3b argumentation.
On Fri, Jul 23, 2004 at 03:30:29PM +0100, Edmund GRIMLEY EVANS wrote:
> Sven Luther <sven.luther@wanadoo.fr>:
>
> > First point, this only applies to released software. Also let's see what the
> > trolltech annotation has to say about it, since it covers some doubt in the
> > language above :
>
> Firstly, I would think that the Trolltech annotation is irrelevant
Yes, you pointed that out already, and will aks the ocaml author about
including them or their own interpretation in future ocaml releases. Still, i
believe that these annotations, as the expressed interpretation of the QPL
authors would carry some weight in court.
> unless INRIA have publicly stated that it applies to their licence.
First to make one thing clear, ocaml is copyrighted by the Crystal Project of
INRIA, not the whole of INRIA as you seem to suppose.
> Secondly, even if they have done that it would probably only help to
> disambiguate the licence where it is ambiguous, and I don't think it
> is ambiguous in this case. Thirdly, I don't think the bit you quoted
> from the annotation out of context means what you think it means. So,
> to save time, I'd rather just ignore the annotation for now.
Ok.
> I think the effect of 3b is to give the initial developer the right to
> incorporate modifications into their code which they can subsequently
> licence in any way they like provided they also make it available
> under the QPL.
Seems a reasonable interpretation, i also agree with it.
> I think you agree with this really, so I don't know why you dragged
> the annotation into this.
To provide redundancy of interpretation.
> > The only way this would be considered as non-free is under the DFSG #1, when
> > you consider the fact of giving those right back to the upstream author a fee
> > or royalty. Ok, this can be argued, and probably will be in a subthread of the
> > corresponding topic, but my own position is that if we consider it a fee, it
> > must include a cost to M to fullfill it, and since M still keeps the whole
> > right to the patch he wrote, and in no way loses any of his rights to it, it
> > cannot really be considered a fee.
>
> I disagree with that argument.
>
> Firstly, I don't think the reference to "royalty of other fee" in DFSG
> #1 is meant to be interpreted narrowly; I think it's meant as an
> example of a restriction.
DFSG #1, sentence 2 : "The license may not require a royalty or other fee for
such sale."
Err, i have some doubt that this can be interpreted as an example of
restriction, meaning that there may be other restrictions which are not
acceptable also, and implicitly mentioned here. If this was the case, why not
say so explicitly, especially as the DFSG was recently cleaned up.
> Secondly, giving someone a right which they didn't already have does
> potentially "include a cost" because it prevents you from asking for
> payment in return for giving that right.
Ok. Agreed, but it is the only thing which you lose, in particular you retain
full rights to the modifications, including those to relicence it under some
other licence, or to apply it to other software, or to sell the modified
binary without giving any fee or royalty upstream. Hell, you could even ask
money from upstream, altough they would not be forced to give them back to
you.
Still, is that such an unacceptable compromise ?
> However, I think there is another argument for 3b not making the QPL
> non-free: M can choose to grant everyone a BSD-like licence for the
> modifications, and then the initial developer doesn't get any right
> they didn't already have.
Or he could do source/patch only distribution, and thus not distribute those
under the QPL. It would not be a DFSG free distribution though.
> Another way of stating the same argument: a licence that forces
> modifiers to licence their modifications to the initial developer is
> no worse than a licence that forces modifiers to licence their
> modifications to everyone, and the latter is arguably still free.
>
> I'm undecided, but I think I can just about accept 3b as consistent
> with the DFSG. Note that I'm not a DD, so my opinion is irrelevant.
> Only my arguments might count, if you choose to accept them.
I do. And the fact you are no DD doens't seem to have an influence here, only
an RM or ftp-master status may.
Friendly,
Sven Luther
> --
> To UNSUBSCRIBE, email to debian-legal-REQUEST@lists.debian.org
> with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
>
Reply to: