[trusty] light-locker-settings may prevent xfpm to apply power source dependent monitor power management

DPMS sleep/standby/off time is not the same as blank time and xfpm1.2 didn't handle blanking at all. We added that in 1.4, but that came out too late for Trusty. Backporting those features is not really an option, even though trusty will be around for quite a bit (simply because our workforce is too limited for that – feel free to help out and propose patches).

> DPMS sleep/standby/off time is not the same as blank time and xfpm1.2 didn't handle blanking at all.

Thats my point exactly: If xfpm is running light-locker-settings should not apply any screen blank settings, because xfpm does not handle it.

I've attached a patch which makes light-locker-settings set dpms standby/off instead of screen blank/dpms standby, if xfpm is running.

Since no service handles the blank times with the switch away from XScreensaver to light-locker, we decided that light-locker-settings should take care of it until xfce4-power-manager was ready. So the current solution in trusty makes perfect sense. You could argue that the DPMS settings conflict and that this one should be removed though. So far there have been no complaints about this duplication though, so dropping that feature seems a bit intrusive for an SRU.

The attachment "dont_set_screen_blank.patch" seems to be a patch. If it isn't, please remove the "patch" flag from the attachment, remove the "patch" tag, and if you are a member of the ~ubuntu-reviewers, unsubscribe the team.

[This is an automated message performed by a Launchpad user owned by ~brian-murray, for any issues please contact him.]

> You could argue that the DPMS settings conflict and that this one should be removed though.

dpms don't conflict, because they are synced between xfpm and light-locker-settings.

> Since no service handles the blank times with the switch away from XScreensaver to light-locker, we decided that light-locker-settings should take care of it until xfce4-power-manager was ready.

Here lies the problem: Since xfpm does not handle screen blank, one can easily break the power source dependent monitor pm of xfpm. Consider following situation: User sets up ac timings in xfpm to 10/15 and battery timings to 7/10. While on battery he decides to reduce this timings and uses light-locker-settings to set them to 3/4. Back on ac he wonders why his screen goes black after 3 min. He checks light-locker-settings and xfpm and both say 10/15.
xset q shows, that xfpm did switch the timings after the power source has changed, but since it does not handle screen blanking it is still set to 3.

I guess the question here is, do we want (incomplete) screen blank gui support for crt monitor users (afaik on lcd monitors screen blank, dpms standby/suspend/off are all the same), or a consistent monitor pm with xfpm.

As I already said, backporting those features is not really an option, even though trusty will be around for quite a bit (simply because our workforce is too limited for that – feel free to help out and propose patches).

The main problem the dependency chain (xfce4-power-manager 1.4 needs a newer version of light-locker-settings etc.), so bottom line: from my pov we'll have to live with what we got as Trusty is in bugfix mode, that's how I understand Ubuntu's release model.

Just to be sure there is no misunderstanding, I was talking about light-locker-settings 1.2.1 and xfpm 1.2.0 (both from trusty). The patch is for those versions. If you decide to apply it, we will loose a gui to set screen blanking (it will be turned off), but we will gain better consistency with xfpm and eliminate some confusing configuration results.

Yeah, I know. Losing control over blanking isn't really an option, as it defines the timeout for light-locker's timed-locking feature.

Are you sure? If I use lock session when screensaver is activated, delay: 10s and xset s 100 dpms 10 0 120. Session locks after 20s.
Same with xset s 0 dpms 10 0 120.
So the first pm event seems to start the timer.

Yes, it's the first event. But the blank timeout is set to 10mins by default, so it's not entirely unlikely that this will trigger (unexpected and - at least via GUI - uncontrollable) timed locking.

Removing the ability to control the blank time means we'd have to disable blanking altogether.

> But the blank timeout is set to 10mins by default ...

Unless there is something hard coded into light-locker, this shouldn't be an issue: ll-settings will set two timings only and will disable (=0) the other ones. With xfpm running: xset s 0 dpms t1 0 t2, and without: xset s t1 dpms t2 0 0.