LAID (was: python libs v lisp coolness?

[Andreas Rottmann]

mike420 at ziplip.com writes:

[ sorry for cross-posting, but I really think this is relevant on all
  newsgroups posted ]

> Joe Marshall <prunesquallor at comcast.net> wrote 
> in <he1qryba.fsf at comcast.net>:
>
>> Kenny Tilton <ktilton at nyc.rr.com> writes:
>> 
>>> I don't know.  Maybe I'm wrong. If Lisp is such a great language and
>>> people love it so much, how come they don't write any code?
>> 
[...]

> Lispers need to cooperate with others on libraries. Maybe a Consortium of
> Underdog Dynamically-Typed Languages needs to be started. Underdog 
> languages are more interested in libraries and cooperation.
>
> For each library, a language-agnostic interface description (in
> LAID language, of course) should be generated by porters, so that 
> later FFIs could be produced for each language automatically. This 
> is the opposite (and complementary to) SWIG: SWIG auto-generates dumb
> interface information and needs a lot of human post-work for each 
> specific language.
>
> "Are you guys having trouble with that GTK lib? - No, we got LAID!"
>
*g*

Actually, h2def.py does something like produce LAID (called defs
there). You basically scan the headers and adapt the resulting defs
(which are, bu coinicence, lisp/scheme data). It's used by the
guile-gobject GTK+2 bindings and the puthon GTK+2 bindings. We just
need some infrastructure/framework for cooperating on improving this
stuff and documenting the defs data format.

Regards, Andy
-- 
Andreas Rottmann         | Rotty at ICQ      | 118634484 at ICQ | a.rottmann at gmx.at
http://www.8ung.at/rotty | GnuPG Key: http://www.8ung.at/rotty/gpg.asc
Fingerprint              | DFB4 4EB4 78A4 5EEE 6219  F228 F92F CFC5 01FD 5B62