J'ai mis quelques idées dans le proto-article (publi-03-transformers).
C'est encore fouilli et il y a encore plein d'idées à rajouter. J'ai
essayé de mettre de l'ordre dans ces idées mais je ne pense pas encore
que ca soit le bon ordre.
Autrement Eelco m'a répondu sur la mailing list stratego-dev :
> > If you can tell me what you think of it...
>
> I think this is very useful. Could you apply these changes to the
> subversion repository directly?
--
Valentin DAVID
valentin(a)lrde.epita.fr
Akim Demaille <akim(a)epita.fr> writes:
> > Quand est-ce qu'on se la fait ?
>
> mardi. Avant, je ne peux pas :( Et vous, c'est possible ?
Ya les partiels, a partir de 18h00 c'est possible.
--
Valentin DAVID
valentin(a)lrde.epita.fr
> Hi Clement,
> Thanks for your quick reply. We do need this as soon as possible,
> otherwise we will have to get on and build it ourselves - which I
> think is a waste of time since you have already done so. Pending
> an official release, is there any possibility of getting a private
> release sooner? Of course I could sign a non-disclosure agreement
> if that would help.
Hi!
Our work is under GPL, so there is no such problem (unless you plan
to release closed source software based on our grammar, in which case
we need to talk).
We would be most happy to participate to your effort, but you ought to
know that:
1. developing this grammar is a major investment. I don't think you
underestimate it, but I wish to emphasize this even further. It
took us 2 years, and yet it is not finished (but few people and
little time, agreed).
2. we used Stratego to disambiguate the parse trees (which _have_ to
be ambiguous in the case of C++), not ASF. ASF people said we
could migrate, but that's not our current wish.
3. There is still some effort to spend to finish the disambiguation,
but this effort will be discontinued because of main developers are
students, and they have to leave for an internship from October to
at least December (both included)
4. The mere size of the grammar, and the modularity issues that it
imposes, led us to improve the tools themselves (extensions to SDF,
modifications to Stratego, etc.). This sole task does require a
real investment, with a steep learning curve...
5. have you read the Transformers report?[1]
All these issues lead me to ask you whether you would have some
proposal for internships. This would certainly be a major speedup for
your project, since both Clément and Valentin are fluent in SDF,
Stratego, _and_ C++ and its grammar. Plus, if they don't get an
internship that would keep them onto Transformers, the project will be
frozen for the next four months; that would even hinder our technology
transfer.
Of course, we too are very much interested in finding internships that
would help the Transformers project...
I hope I am not hitting on you too much, but actually this internship
issue is becoming urgent, and your message enables hope...
Regards,
Akim Demaille
[1] See the (beta) web page
http://www.lrde.epita.fr/cgi-bin/twiki/view/Projects/Transformers
which should very soon include this report...
