On 9/6/06, JP Rosevear <[EMAIL PROTECTED]> wrote: > The problem is, I've seen no unequivocal declaration about gtk+ and glib > accepting these higher level abstractions, so perhaps matthias can > comment, because historically this has not been the case and is a > primary concern for me at least.
I think Federico and Havoc summed up pretty well how I think about these "higher level abstractions". There is nothing in principle the precludes the introduction of such abstractions in GTK+/GLib. Of course, there are factors that make this more work than simply copying code from libgnome or libgnomeui to glib/gtk: - We want APIs that can be implemented on other platforms as well, if at all possible - Some things that work fine in a desktop setting may cause problems in embedded situations (cf the stuck-with-2.6-for-cairo situation that we're still trying to rectify), and we may need to think about making certain features optional - The GTK+ core team is not necessarily growing at the same rate as the code base. It is pretty clear that GConf and gnome-vfs are the two major sore points in our library stack. They just live at the wrong spot in the stack. If we had nice configuration and vfs apis at the glib level, our library stack would look a lot nicer, and a lot of things could be done in less convoluted ways in GTK+. Of course, this is a lot of work, but I know that Alex is thinking about what needs to happen to gnome-vfs in this direction. Matthias _______________________________________________ desktop-devel-list mailing list [email protected] http://mail.gnome.org/mailman/listinfo/desktop-devel-list
