On Mon, Feb 18, 2008 at 09:08:00PM +0000, Ben Pfaff wrote:
I think that the glib typedefs for int and other standard types from C89
are
a stupid idea that should never have seen the light of day. There is some
sense in gboolean, since C89 didn't have a boolean type. I think the rationale is that gtk/glib has many different language bindings. Thus, whilst in the case of C, gint is a simple typedef, in other languages it may not be so simple. The idea is, that if one sees "gint" in code (in any language) one knows that it's a type that's supported by glib (in any language). So, I'm trying to stick to the glib identifiers whereever they're not going to be used outside of GUI code. This would, for instance, make it easier for somebody who wanted to take the GUI and port it to (say) python. Also, I understand there are some doxygen like tools for gtk+ that benefit from using the g types. J' -- PGP Public key ID: 1024D/2DE827B3 fingerprint = 8797 A26D 0854 2EAB 0285 A290 8A67 719C 2DE8 27B3 See http://pgp.mit.edu or any PGP keyserver for public key.
signature.asc
Description: Digital signature
_______________________________________________ pspp-dev mailing list [email protected] http://lists.gnu.org/mailman/listinfo/pspp-dev
