should detect ekiga (voip) activity as activity and not lock screen

I've forwarded the issue upstream: http://bugzilla.gnome.org/show_bug.cgi?id=330237

libwnck adds a NET_WM_STATE_FULLSCREEN status apps can use to prevent the screensaver from lockong while in fullscreen mode (gnome-screensaver respects this setting), i think something like NET_WM_STATE_LOCK_OVERRIDE with the same functionallity should be added to libwnck to make it possible for the screensaver to determine that an app is trying to prevent the locking (indeed ekiga could abuse the fullscreen state as a workaround)

ekiga apparently already has a dbus handler for gnome-screensaver which is just not enabled

as you can see in the upstream bug, the dbus code in ekiga is not anywhere yet regarding this problem

If the VOIP call doesn't have video, why don't you want the screen etc. to turn off?

Because that would lock the screen and force me to reenter my password which would be inconvenient when I need to note something down in the middle of the conversation.

Hi Tollef,

I don't mean to labour the point but I can't see the justification in inhibiting for a voice-only call. I agree completely for video calls. If I am jotting things down on my computer, or stepping someone through something, then it shouldn't disable (and in your case lock) the screen. In those situations, however, it wouldn't come on because I would be using the computer enough to prevent it.

If, when I am on a voice call, I leave the machine and wander around the room for 10 minutes then the situation is no different to one where I go to use a normal phone, or do anything else. I may want to come back and make notes on my computer after 15 minutes but that would always involve the hassle of entering a password if I have chosen to have the screensaver lock the screen.

Obviously, inhibiting the screensaver when the user is in a long voice call (and not using the screen) would raise the possibility of screenburn. Keep in mind, however, that inhibiting the screensaver also inhibits the diming and powering off of screens belonging to laptop users concerned with saving power.

I disagree about handling voice+video different from voice-only calls. I'm in both cases using the computer and I think this is similar enough to the case of watching a movie (which inhibits the screensaver from starting). Note that I might not be jotting down stuff all the time and I have a very short lock timeout (1 minute) which combined with a long password means unlocking the screen actually is a significant time cost in the middle of a conversation. You could obviously say that those are silly settings, but to me as a user, I'm using the computer when sitting in front of it and talking.

About the wandering around the room for ten minutes; I have a cord about 1.5m long so it'd be a very small walk in my case.

Screen burn is mostly a problem for CRTs and then you are talking about much, much, much more time than what you'd get from sitting in a phone call. And if you're worried about that, you can always just switch on the screensaver yourself.
Same goes for laptop power consumption. (Actually, I think dimming the screen would be fine, but not turning it off or locking it.)

> (Actually, I think dimming the screen would be fine, but not turning it off or locking it.)

Note that, with the current GPM, the screen does dim between when the session is idle and when the screen is blanked. Of course, this doesn't happen if the screensaver is inhibited because the session never becomes "idle". That means that, because voice-only calls inhibit the screensaver, the screen will not dim. Currently, the inhibit options are limited to display and computer sleep, AFAIK.

this has been fixed upstream now.

Should be fixed with 3.2.0 in jaunty.