Re: ocaml, QPL and the DFSG: New ocaml licence proposal.
On Sat, Jul 31, 2004 at 10:01:42AM +1000, Matthew Palmer wrote:
> On Fri, Jul 30, 2004 at 02:31:27PM +0200, Sven Luther wrote:
> > On Fri, Jul 30, 2004 at 07:53:42AM -0400, Walter Landry wrote:
> > > Sven Luther <sven.luther@wanadoo.fr> wrote:
> > > > On Thu, Jul 29, 2004 at 05:53:14AM -0400, Walter Landry wrote:
> > > > > Sven Luther <sven.luther@wanadoo.fr> wrote:
> > > > > > So this solves most of the issues, and we need to go through the QPL
> > > > > > 3b again, but upstream feels it is a reasonable clause, and would
> > > > > > like to keep it.
> > > > >
> > > > > I'm sure that anyone would love to have that kind of term in a
> > > > > license. It still feels non-free to me.
> > > >
> > > > Sure, but there is much less consensus about this one, so if a handfull of
> > > > people feel it is non-free, i doubt it will come into play.
> > >
> > > I would consider it a fee. It is even enshrined in US copyright law [1]
> > >
> > > The term "financial gain" includes receipt, or expectation
> > > of receipt, of anything of value, including the receipt of other
> > > copyrighted works.
> >
> > Ok, well. But we need to consider non-US law also.
>
> Let's, then. What does French copyright law define as a fee or financial
> gain?
>
> > > Since all copyrights flow to the originator, I can't help but see it
> > > as a fee for making modifications.
> >
> > Well, even if we see it as such, do we really want to declare this clause as
> > non-free ? After all it will simplify the administrative tasks involved in
> > havign upstream integrate changes back, and in general will be a win for free
> > software.
>
> It's not exactly rocket science to add a "some modifications (C) Foo Wombat"
> to each file touched by a patch submitted by Foo Wombat. With a decent
> revision control system it's even straightforward to identify when those
> changes get written back out again so you can remove the (C) notice.
>
> What you're talking about, I believe, is simplifying the ability of
> upstream to release proprietary versions of the software, requiring no
> explicit copyright assignment or permission grant. That is a little more
> difficult to judge as "a win for free software", except in the sense that
> upstream gets money to continue to develop "free" software. And yet there
> are no shortage of Free Software projects that have thrived without this
> clause.
Well, no, you cannot remove the existing copyright statement. Still the clause
3 deals with patches, and you can add your own copyright statement. I
understand there are the following issues at hand here :
A) people need to be able to add themselves to the copyright of files they
touch.
B) people need to be able to translate copyright statement that appear in
the about box of an interactive program.
C) if a whole file is removed, then it should be ok to remove the copyright
statement attached to it.
And that's the extent of it. Since we cannot modify the QPL anyway, would it
be enough to add an additional ecemption, or maybe simply an upstream
annotation saying that those are explicitly allowed ? I failed to write some
nice and concise wording eplaining the above.
> > Notice that the non-freeness involved here, is about the freedom to not
> > contribute back your changes, is this really something we want to defent ?
>
> Yes. Otherwise we end up with licence clauses requiring that all software
> you write must be freely licenced and publically distributed as a condition
> of some piece of software. Not contributing your changes upstream is
And we would accept those as free.
> something you want to protect just as much as your right to not distribute a
> work under a Free licence in the first place.
That you want to protect, i have no such interest, and i believe many in the
debian project would feel the same.
> Note that this statement shall not be construed in a way as to discourage
> copyleft. That is a notice from the original copyright holder saying "if
> you want to distribute bits of my work, you have to do equivalent things to
> your work". Compelling even *submission* upstream goes further, and
> compelling an all-permissive licence to upstream is well beyond the simple
> "share-alike" provisions of copyleft.
There is no submission upstream required here, just the right of upstream to
integrate the patches, if he comes to them anyway, into its software. And the
additional right to have it also integrated in the proprietary version of said
software, _AS LONG AS_ he also integrate it into the free part.
> > > > > > Also the first modification, well, i am not overly confident that it
> > > > > > is really needed, and i am sure my wording of it are abysmal, and i
> > > > > > ask for some help here in finding some nice and concise wording
> > > > > > which doesn't divert to much from the original. The old wording was :
> > > > > >
> > > > > > a. Modifications must not alter or remove any copyright notices
> > > > > > in the Software.
> > > > > >
> > > > > > And i changed it to :
> > > > > >
> > > > > > a. Modifications must not alter or remove any copyright notices
> > > > > > in the Software except by adding new authors.
> > > > >
> > > > > If I'm converting an interactive program to be non-interactive, I
> > > > > still can't remove a hard-coded copyright string that pops up in an
> > > > > "About" box.
> > > >
> > > > Bah. I doubt this is what was meant here, and i doubt this is going to be a
> > > > problem all over.
> > >
> > > If you don't think that is what is meant, then change the wording to
> > > say that (preferably, remove it). Otherwise it is just lawyerbait.
> >
> > Notice that i will have to add all this modification in a licence-patch why,
> > saying : The software is under the QPL, except ..., so the less change is
> > needed the less confusion it will be. I would much rather keep this one as is,
> > and concentrate at a later time to the change to another licence altogether,
> > maybe one of the upcoming CECILL family.
> >
> > Now, if you could propose a sane and not too involved wording for the above, i
> > and upstream would consider this. It should not exceed a few (preferably two)
> > lines though.
>
> Here's a nice short one:
>
> ""
Please, can we start to discuss seriously ?
Friendly,
Sven Luther
Reply to: