minimal closure usage

[Anthony Green]

On Tue, 2002-04-30 at 07:34, Sam Liddicott wrote:
> Perhaps I could contribute an (ugh) set of macros which taking standard
> types expand out to ffi_type stuff?

I think the mapping is pretty straight forward, with the exception of
char which may be signed or unsigned.

> Why ffi_type_sint and not ffi_type_int ?

sint == signed int.  I chose to use that for uniformity.  I didn't want
an ffi_type_char because char may be signed or unsigned depending on
your environment.  So I decided to be explicit about each type name.

AG