Re: [Caml-list] ocaml compiler license

[Roberto Di Cosmo <roberto@dicosmo.org>]

David, you are right: one cannot "change" the terms of a licence used
on a piece of code after releasing this piece of code under this licence.
But we are talking about "interpretation" of these terms, which is quite
a different story.

If you look at FSF/OSI/<whatever other external entity including me> to know
whether a git branch is fine as an implementation of "separate", you only get an
opinion... maybe a well respected opinion, but just an opinion that may become
the subject of debate later on. And avoiding debate is a desirable thing.

If Inria says publicly (e.g. on this mailing list): "for us it's fine to see
OCaml derivatives developed and published as separate git branches", then you
get the work owner's explicit statement that this practice is acceptable for the
only entity entitled to enforce the licence.  And that's the end of the story:
simple, clear, undebatable, no hassle.

Cheers

--
Roberto

On Fri, Mar 06, 2015 at 09:08:54AM +0000, David Allsopp wrote:
> Roberto Di Cosmo write:
> > Just one line on all this: as correctly hinted otherwise, the copyright
> > can only be enforced by the owner of the work (that is Inria in this
> > case); so, if a doubt arises about the meaning of some terms of the
> > licence arises, you should not look at what the FSF or the OSI or some
> > other lawyers say, but ask the owner of the work for clarifications.
> 
> Only as a legal aside, that is a *sensible* thing to do, but it is *not* how copyright works. Once you've released something covered by a licence, the terms are only clear as exactly what you've put in the licence. You don't get to say on later enquiry, "oh, I didn't think of that - no, you should interpret it this way". That's what courts (i.e. legal tests) are for and why licences should be written by lawyers. And at that point it *is* useful to look at OSI and FSF opinion, because what a court will take great interest in is legal precedent or, in its absence, what everyone else using those terms thinks.
> 
> 
> David
> 
> > Cheers
> > 
> > --
> > Roberto
> > 
> > 
> > On Fri, Mar 06, 2015 at 08:45:28AM +0000, David Allsopp wrote:
> > > Fabrice Le Fessant wrote:
> > > > From the QPL 1.0 license in OCaml sources:
> > > >
> > > > "... distribute your modifications, in a form that is separate from
> > > > the Software, such as patches."
> > > >
> > > > In GIT, the software itself is a set of patches, so it's not so
> > > > clear for me that if the modifications are another set of patches,
> > > > there are in a "separate form".
> > >
> > > This is why I said that it should be done in a separate branch. You're
> > happy with the idea that the file system (or the URL) provides separation
> > (i.e. one URL for the ocaml tarball, one URL for the patch file) - I'd
> > make the legal argument that git is simply a file system (when you remove
> > the porcelain, the plumbing *is* simply a file system, after all!) and a
> > git branch name provides the same separation as the two separate files.
> > >
> > > If one is going to argue that git branches do not provide separation,
> > then do we start talking in terms of needing the main sources and the
> > patch on separate harddisks; separate servers; separate internets? :o)
> > Indeed, with git clone {your-ocaml-repository} -b master --single-branch
> > you can clone the repository *without* downloading your patch branch at
> > all (i.e. the unmodified OCaml sources) so I would assert that git
> > definitely satisfies the "separate form" requirement.
> > >
> > > However, the QPL does not rigidly define "separate" and uses the woolly
> > term "such as" for its example, so that argument could only be resolved by
> > being tested.
> > >
> > > > Anyway, many people are distributing modified OCaml versions under
> > > > GIT, and INRIA has never complained about it, to the best of my
> > knowledge.
> > >
> > > Which is a much stronger argument, given that pull requests are
> > > allowed for the OCaml compiler, from public repos :o)
> > >
> > >
> > > David
> > >
> > > > --Fabrice
> > > >
> > > >
> > > >
> > > > On Fri, Mar 6, 2015 at 12:07 AM, David Allsopp
> > > > <dra-news@metastack.com>
> > > > wrote:
> > > > > Alexey Egorov wrote:
> > > > >> Hi Adrien,
> > > > >>
> > > > >> thanks for your reply.
> > > > >>
> > > > >> But wouldn't publicly avalable github repo with modified sources
> > > > >> violate license (just because non-patch/diff version is publicly
> > > > >> available)?
> > > > >
> > > > > IANAL, but if you put all your changes in a separate branch on
> > > > > your own
> > > > git repo (forked from the official Git mirror) then you are
> > > > literally storing a set of patches - the ability to download the
> > > > entire distribution as one whole (or even as a .tar.gz file, say on
> > > > GitHub) is just a convenience provided by the software, it's not
> > > > what you're actually providing.
> > > > >
> > > > >
> > > > > David
> > > > >
> > > > > --
> > > > > Caml-list mailing list.  Subscription management and archives:
> > > > > https://sympa.inria.fr/sympa/arc/caml-list
> > > > > Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
> > > > > Bug reports: http://caml.inria.fr/bin/caml-bugs
> > > >
> > > >
> > > >
> > > > --
> > > > Fabrice LE FESSANT
> > > > Chercheur en Informatique
> > > > INRIA Paris Rocquencourt -- OCamlPro Programming Languages and
> > > > Distributed Systems
> > >
> > > --
> > > Caml-list mailing list.  Subscription management and archives:
> > > https://sympa.inria.fr/sympa/arc/caml-list
> > > Beginner's list: http://groups.yahoo.com/group/ocaml_beginners
> > > Bug reports: http://caml.inria.fr/bin/caml-bugs
> > 
> > --
> > Roberto Di Cosmo
> > 
> > ------------------------------------------------------------------
> > Professeur               En delegation a l'INRIA
> > PPS                      E-mail: roberto@dicosmo.org
> > Universite Paris Diderot WWW  : http://www.dicosmo.org
> > Case 7014                Tel  : ++33-(0)1-57 27 92 20
> > 5, Rue Thomas Mann
> > F-75205 Paris Cedex 13   Identica: http://identi.ca/rdicosmo
> > FRANCE.                  Twitter: http://twitter.com/rdicosmo
> > ------------------------------------------------------------------
> > Attachments:
> > MIME accepted, Word deprecated
> >       http://www.gnu.org/philosophy/no-word-attachments.html
> > ------------------------------------------------------------------
> > Office location:
> > 
> > Bureau 3020 (3rd floor)
> > Batiment Sophie Germain
> > Avenue de France
> > Metro Bibliotheque Francois Mitterrand, ligne 14/RER C
> > -----------------------------------------------------------------
> > GPG fingerprint 2931 20CE 3A5A 5390 98EC 8BFC FCCA C3BE 39CB 12D3

-- 
Roberto Di Cosmo
 
------------------------------------------------------------------
Professeur               En delegation a l'INRIA
PPS                      E-mail: roberto@dicosmo.org
Universite Paris Diderot WWW  : http://www.dicosmo.org
Case 7014                Tel  : ++33-(0)1-57 27 92 20
5, Rue Thomas Mann       
F-75205 Paris Cedex 13   Identica: http://identi.ca/rdicosmo
FRANCE.                  Twitter: http://twitter.com/rdicosmo
------------------------------------------------------------------
Attachments:
MIME accepted, Word deprecated
      http://www.gnu.org/philosophy/no-word-attachments.html
------------------------------------------------------------------
Office location:
 
Bureau 3020 (3rd floor)
Batiment Sophie Germain
Avenue de France
Metro Bibliotheque Francois Mitterrand, ligne 14/RER C
-----------------------------------------------------------------
GPG fingerprint 2931 20CE 3A5A 5390 98EC 8BFC FCCA C3BE 39CB 12D3