[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: LablGL and LablGtk packages are in master incoming.



On Thu, Dec 07, 2000 at 13:16:55 +0100, Sven LUTHER wrote:
> let's put them in a -dev package, what do you think ?

 I don't know. The LablGtk provides an OCaml library. So, in some
 sense, it is just a "dev" package. Doing a dev package of that
 would be as doing a "dev.dev" package. And, moreover, we don't have
 a dev package for OCaml with the .h files: they are put directly
 inside the OCaml package. But maybe what I am saying is
 only nonsense...

> ideally, my c2caml tool should be able to regenerate them automatically ...

 Yea., 

> >  Note also that, sometimes, LablGtk uses heavily boxing for the
> >  pointers and then the macros are no more simple casts, but
> >  functions to box/unbox values.
> 
> Yea, i am familiar with this, like transforming a enum into a list of flags ?

 Something like that, yes.

> >  A last example is with polimorfic variants constructors: in the .h
> >  files I should find the C #definition for the constructors to use
> >  them in my bindings. Otherwise, I have to recompute them in my code.
> 

> They are auto generated by the var2xxx stuff, isn't it ?

 Yes, they are.

> >  So I think that the first step is to install also the .h and the
> >  second one is to convince Garrigue that they must be completed/rewritten.
> >  But this is unrelated to packaging.

> Ok, will do, could you tell me exactly what you need as .h, and i will but
> then into the package.

 All the four .h: ml_gdk.h, ml_glib.h, ml_gtk.h, wrappers.h
 Then, as soon as I have time, I will talk again to Garrigue to have
 also all the missing .h files.

					Regards,
					C.S.C.

-- 
----------------------------------------------------------------
Real name: Claudio Sacerdoti Coen
Graduate Computer Science Student at University of Bologna
Now on leaving at INRIA (Rocquencourt)
E-mail: sacerdot@cs.unibo.it
http://caristudenti.cs.unibo.it/~sacerdot
----------------------------------------------------------------



Reply to: