Hi Andreas, Andreas Rottmann <[EMAIL PROTECTED]> writes:
> You should probably also have a look at the Scheme-Level FFI of PLT > Scheme[1], and my reimplementation for Scheme 48[2]. Thanks for the links, that's the kind of think I had in mind. Perhaps you'd like to port S42 to Guile as well? :-) BTW, there's one small potential issue with the dynamic construction of function calls like this: it's almost impossible to catch API/ABI inconsistencies beforehand. For instance, if one wraps `strcmp' from Scheme as if it were, say, "char *strcmp (int, double, float)", it's impossible to catch the problem before actually calling the wrapped `strcmp'. Symbol versioning could help in some cases, but I don't think it's readily usable (and probably not portably). Conversely, good old C wrappers (be they generated or not) would most likely catch such problems at compilation-time. Any thoughts about this? Thanks, Ludovic. _______________________________________________ Guile-devel mailing list [email protected] http://lists.gnu.org/mailman/listinfo/guile-devel
