Re: with-slots usage question
Miles Egan <[email protected]> Mon, 10 Nov 2003 11:30:12 -0800
| Newsgroups | gmane.lisp.clump |
|---|---|
| Message-ID | <1068492611.25524.49.camel@car> |
On Mon, 2003-11-10 at 10:56, Harley Gorrell wrote: > > but isn't it a problem if you try to define a generic > > function if there's already a regular function by that name? > > Yes it is. I find myself having to unbind 'defun's to > replace them with generic functions. I guess the ideal thing here from the programmer's point of view would be a language in which all functions are generic and the implementation is smart enough to optimize the ones that are never used polymorphically. The CL solution feels a little like the virtual/non-virtual distinction in C++. Maybe this is what you get when you try to add OO protocols to a language that wasn't originally designed with them in mind. How did Dylan handle this? -- Miles Egan <[email protected]> _______________________________________________ Clump mailing list [email protected] http://manly.caddr.com/mailman/listinfo/clump
signature.asc
(application/pgp-signature, 189 B)
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.3 (GNU/Linux) iD8DBQA/r+dDU4Jq/wH1PVERApw1AJ4zz2qYioehvWbp0+OAdk5JPnCRIgCgsjiP MxRLejen1yeJLXk4lavwVsg= =r+tl -----END PGP SIGNATURE-----