It's Not A Bug, It's A Feature. (TM)
GTK+ 2.x uses by default its own input methods for character composition. Nevertheless, you can force GTK+ applications to use the Xlib composition mechanisms by setting GTK_IM_MODULE=xim among your environment variables.
But as far as I could notice, then you lose the Gtk+'s smart shortcut handling -- eg. if you've got Greek/Serbian Cyrillic/... map loaded into group1, and en_US in group2, it's enough that you press eg.
Control + alpha/cyrillic a
to initiate action Control+A (that being "latin A" :-).
So, this is not really a solution, especially not for Serbian folks who use latin and cyrillic interchangeably -- the same could be the case for some other languages.
Actually, I think that it's good thing that Gtk+ provides all the compose sequences itself, because it forces some kind of standardization. Still, it's bad because every simple change has to be compiled in order to be visible.
I'm not really sure about the reasons for this particular changes.
Btw, I'd prefer if Gtk+ (Pango, actually) supported rendering two or more separate combining characters with non-combining character as one glyph -- that's a basic Unicode thing :-)
Though, I'm getting way offtopic here.
Cheers, Danilo _______________________________________________ I18n mailing list [EMAIL PROTECTED] http://XFree86.Org/mailman/listinfo/i18n
