Somewhat random set of comments: - For all the reasons that it makes sense for GNOME to stick to it's published schedules, it makes sense for GTK+ to stick to it's published schedules. While we haven't always done a great job in the past (GTK+-2.2.0 was particularly bad), that doesn't mean that we have "license to slip".
There are certainly people who want to slip the GTK+-2.6 schedule
so we can get in introspection or libglade integration, but in
the end, slipping a few weeks or a month isn't going allow us to
do a really good job on those.
So, regardless of what the GNOME release team decides, I think
we should stick to the schedule we published.
- To be realistic, if GNOME-2.12 is released after GTK+-2.8,
most distributors are going to ship using them together.
So, the release team could decide to go with 2.6 even with
the current GTK+ schedule, but the result then would be less
testing of what people actually ship.
- For new APIs to be done well, we need language binding
author attention. If we decided to slip out the GTK+ schedule
a month or two and add more features, I'm not sure that those
features would get sufficient language binding author attention.
- While GNOME isn't going to be using many Cairo features for 2.12,
getting Cairo-1.0 and GTK+-2.8 out there early is important to
getting apps out using GTK+ and Cairo together for 2.14.
I think that a GTK+ with Cairo and small features is very realistic
on the schedule on gtk.org/plan/2.8.
One inherent difficulty with the way that GNOME is working these
days is with a new feature-release every 6 months, we don't really
give ourselves any time for .0 shakeout. A major code change
can easily take 6 months - after release - to shake out. By that
time, with our scheduling, we've released more major code changes.
Honestly, I expect some portability issues with Cairo after
1.0 is released. It's not going to work well on some set of older
X servers, and we'll have to work around those issues. But
people aren't going to find those issues until GNOME or a major
app is depending on Cairo.
Regards,
Owen
signature.asc
Description: This is a digitally signed message part
_______________________________________________ desktop-devel-list mailing list [email protected] http://mail.gnome.org/mailman/listinfo/desktop-devel-list
