Havoc Pennington wrote:
> Mark Leisher <[EMAIL PROTECTED]> writes:
> > Like one pair of functions (basically) to set/get values on an object, this is
> > another area where Xt was designed correctly IMHO. In Xt, all callbacks have
> > the same 4 parameters. In GTK terms the prototype would be:
> >
> > void (*callback)(GtkWidget *widget, GdkEvent *event,
> > gpointer call_data, gpointer client_data);
> There aren't that many variants for commonly-used signals, and you can
> always look it up.
>
> The advantages of the GTK way are numerous; one of the more important
> is that signals are also introspectable virtual functions.
The single-prototype-callback has a lot of merit, if you look at
realistic use-scenarios. My SDPGTK C++ library has to jump through a
lot of hoops to layer this type of paradigm on top of GTK+ (with the
ability to get at the "raw" event data, when needed), but it's clearly
worth it: in an interactive 3D modelling application with over 100
dialogs and literally thousands of different event-handlers, I only
count about 30 places in my code where I access the "raw" GTK+ callback
data at all.
If there was some way to provide an optional single-prototype-callback
to GTK+, I'm pretty sure library authors everywhere would have a field day.
Regards,
Tim Shead
[EMAIL PROTECTED]
_______________________________________________
gtk-list mailing list
[EMAIL PROTECTED]
http://mail.gnome.org/mailman/listinfo/gtk-list
