Re: [RFC] Glib::filename_(to|from)_unicode => Glib->filename_(to|from)_unicode?

["A. Pagaltzis" <pagaltzis gmx de>]

* Ross McFarland <rwmcfa1 neces com> [2003-12-24 16:07]:
> this is mainly a question of uniformity of api

This point I can see; I would have commented on it in my previous
mail, but didn't know and didn't have the time to find out what
precedent, if any, has been set.

> there are many cases where inheritance would/could be necessary
> and if you're going to use the arrow syntax there then it's
> best to do it everywhere.

Agreed.

> while we're at it we probably ought to take out the prototype
> keyword unless its in there for a specific reason as we haven't
> used it anywhere else.  we could at least support both :: and
> -> from now on, and the only people broken would be new scripts
> with old rel's. old scripts with new rel's would be ok.

That would be ideal; I do have some doubts about how robustly
this can be done, however my only experience is doing this in
pure-Perl code (where it's tricky at best if you have to deal
with any even vaguely complex argument list and ends up
relatively fragile). I seem to remember though that there's a
Devel:: module or some such that allows to reliably detect which
way a function was invoked.

So with an eye on the API freeze deadline my vote is "fix it",
with an added plea to make sure invocation style detection works
robustly.

-- 
Regards,
Aristotle
 
"If you can't laugh at yourself, you don't take life seriously enough."