>+ DBusConnection* bus = dbus_bus_get(DBUS_BUS_SESSION, &error);
dbus_bus_get() will sometimes call
dbus_connection_set_exit_on_disconnect(TRUE), so that needs to be undone with
another dbus_connection_set_exit_on_disconnect().
We are actively trying to shorten the start-up path to reduce start-up time.
Blocking on DBus during creation of the first window wouldn't be helpful.
Is it feasible to initialize on receipt of an async reply?
Or I expect it would be possible to create a DBusPendingCall on window
creation, and then only block on it and steal_reply when it is really needed?
(In reply to Trevor Saunders (:tbsaunde) from comment #25)
> It's a bit of a usability problem
> since currently as the bug says firefox appears to be inaccessible in modern
> linux installs.
I thought GNOME 3 distros were at least providing a read-only GConf wrapper.
Comment on attachment 572700
patch with nits fixed
Can you provide function names and 8 lines of context with future patches, /developer. mozilla. org/en/ Installing_ Mercurial# Configuration
please? https:/
>+ DBusConnection* bus = dbus_bus_ get(DBUS_ BUS_SESSION, &error);
dbus_bus_get() will sometimes call _set_exit_ on_disconnect( TRUE), so that needs to be undone with _set_exit_ on_disconnect( ).
dbus_connection
another dbus_connection
>+ reply = dbus_connection _send_with_ reply_and_ block(bus, message, 1000, &error);
We are actively trying to shorten the start-up path to reduce start-up time.
Blocking on DBus during creation of the first window wouldn't be helpful.
Is it feasible to initialize on receipt of an async reply?
Or I expect it would be possible to create a DBusPendingCall on window
creation, and then only block on it and steal_reply when it is really needed?
(In reply to Trevor Saunders (:tbsaunde) from comment #25)
> It's a bit of a usability problem
> since currently as the bug says firefox appears to be inaccessible in modern
> linux installs.
I thought GNOME 3 distros were at least providing a read-only GConf wrapper.