https://bugzilla.gnome.org/show_bug.cgi?id=786693
--- Comment #11 from Bastien Nocera <[email protected]> ---
(In reply to Martin Blanchard from comment #9)
> (In reply to Bastien Nocera from comment #8)
> > It also looks easy enough to extend and add support for the 2 other
> > GtkSettings if you're interested.
>
> Hummm, isn't 'gtk-enable-animations' already implemented as a gsetting?
Yes, incorrectly so in the Wayland backend. In the X11 backend, animations are
force-disabled if the GSettings is false, then controlled by
gnome-settings-daemon, which will disable them if a software renderer is used,
or an external VNC client is connected.
> > ::: gdk/wayland/gdkscreen-wayland.c
> > @@ +70,3 @@
> > Why use a struct here, and not directly a timestamp? The struct will be
> > initialised to 0 when it's instantiated.
>
> I've kept the struct so that dbus settings are grouped, that make it clear
> that they'll be handle together.
That's fine. If a GTK+ reviewer doesn't like it, we can unbundle it.
> > @@ +713,3 @@
> > Why the guard here? Wouldn't you want to return the "0" instead of an error
> > here?
>
> You're right, the distinction between a somehow fake initial value and 0 is
> of little interest here.
Well, 0 is a valid value. This will break if somebody rewinds their clock back
to the 1970, but the value itself is of little interest, we could use an always
incrementing serial, but a timestamp will likely make races and debugging
easier.
--
You are receiving this mail because:
You are on the CC list for the bug.
_______________________________________________
wayland-bugs mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/wayland-bugs
