Re: GNOME 2.11/2.12 targeting GTK+ 2.8 (ie cairo based)

[mortenw gnome org (Morten Welinder)]

Owen,

the problem with gtk+/cairo in a stable setting is that it has only
seen testing and been designed for "hello world".  You need to consider
something bigger!

Try putting 600-1000 pango layouts on the screen at once.  GTK+ 2.6 is
already not very snappy with that.  Then make sure your X server is
not local so you get to suffer the roundtrips also.  I would like to
hear from you that this has been considered and perhaps even tested.
We still have not gotten anywhere near the screen update speed we had
in the 1.2 days.

Wearing my application developer hat, I would say that using GTK+ 2.8
for the next Gnome is crazy.  It isn't right to enlist the unsuspecting
users as guinea pigs just because you are having trouble getting people
to test.  The Cairo integration was done prematurely and now we suffer.

Glib 2.8, on the other hand, fine.  That's not a revolution.  It
might as well get renamed back to 2.6.whatever.

And btw., we still are not near a state where the file chooser is
acceptable.  The "save-as" mode really stinks while the "open" mode
is only marginally acceptable.  See
    http://blogs.gnome.org/view/mortenw/2005/05/25/1
for example.

Morten