[plt-scheme] missing gsl scientific libraries

2008/8/19 Matt Jadud <jadudm at gmail.com>:
> i'm curious... has anyone ever looked at doing a MzScheme SWIG backend
> that emits CFFI code instead of creating C bindings? I've wondered if
> this would be a feasible approach from time-to-time, but never really
> investigated it.

It's been brought up from time to time on-list, but AFAIK no one has
actually done anything toward it. I think the difficulty is that the
bindings generated by SWIG would be platform-specific, as it would
expand all #ifdefs. In PLT-Scheme, you would really want to switch on
the value of (system-type) for platform-specific differences. The only
way I've thought of doing that with SWIG would be to generate flat
bindings for each system and do some automated diff on the results to
generate different branches of definitions.

    Henk