Re: [gpm] Suspend/Hibernate interface

[Phillip Susi <psusi cfl rr com>]

On 3/8/2011 6:42 AM, Richard Hughes wrote:
> Right, it's expected the screensaver listens for the Sleeping()
> signal, and then it's got one second to lock the screen.

> I think for 3.0 we should remove the code in g-p-m that tells
> gnome-screensaver to sleep.

Why do you think this should be up to the screen saver instead of gpm?
Isn't gpm the place where such policy decisions are supposed to be made?

It looks like before, gpm emitted a Sleeping signal that took an
argument indicating whether it was suspend or hibernate, and now that is
done by UPower, but it has no argument.  It seems like UPower needs to
have the argument added back to the signal.

Also I'm still trying to figure out this gtk and dbus stuff.  I can't
quite figure out if you can directly bind the gpm control object
SLEEPING signal to UPower, or if you have to create a proxy object, and
register a callback that then re-emits the signal to the control object.
 I was trying to modify gpm so that instead of locking the screen and
then emitting the signal, that it would instead lock the screen when it
RECEIVES the signal from UPower.  That is when I noticed that UPower's
signal is missing the argument.