| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| | |
|
| |
|
|
| |
Accidental octal values...
|
| |
|
|
|
| |
Generate .xscreensaver as if saver is allowed, then when we check what
the user selected in vmchooser, patch .xscreensaver if necessary.
|
| | |
|
| |
|
|
|
|
|
| |
This turned out to be a little too agressive; We run some async jobs on
session close that were killed by logind. Defer the process cleanup
until idle-daemon sees that the user doesn't have any more open
sessions.
|
| | |
|
| |
|
|
|
| |
Get rid of old, brittle script in pam session close
that would do wonky stuff to kill stray user processes.
|
| |
|
|
|
|
| |
vmchooser's sessionstart sets an environment var to tell us whether we
should enable the screen saver for a particular user. Honor this var and
write or delete the flag file idle-daemon depends on.
|
| | |
|
| |
|
|
|
|
|
|
|
|
| |
Keep existing schedule entries (we have --min-idle 7200) but add entries
for 23:00 and 4:00 to non-empty schedules. We might skip a tightly
scheduled entry due to --min-idle, e.g. an exam lasts until 19:30 and
the schedule says poweroff at 20:00. We'll miss that entry if the user
didn't log out before, since two hours didn't pass yet.
If the schedule was empty however, do nothing, since it might be desired
that the machines keep running until they die.
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
| |
vmware ungrab code has to fake mouse cursor movement, resetting PC idle
time :-( - Now we can't record the PC idle time the moment the screen
saver turned on anymore, since it will be reset at that point.
|
| |
|
|
|
|
|
| |
In case a shutdown/reboot/... is manually triggered (i.e. via web
interface), the according timestamp won't be in the idleaction hint file
for the xscreensaver hack. Update the timestamp in case a reboot is
imminent, so the screeensaver can update the countdown.
|
| | |
|
| | |
|
| |
|
|
|
|
| |
If a user unlocks the session and locks it again within two minutes,
the idle timeout will not be reset. Even worse, this also happens when
you manage to logout and in (as a different user) within two mins.
|
| |
|
|
|
|
|
| |
The screensaver info is off by 5 minutes for reboot/shutdown, since
there's an additional delay of 5 minutes if someone's logged in.
Always add those 5 minutes, since the message will only ever be seen if
someone is actually logged in.
|
| | |
|
| |
|
|
|
| |
This should qield more accurate feedback, especially now that we have
the xscreensaver hack giving user feedback.
|
| | |
|
| |
|
|
|
| |
Utility similar to xprintidle. If xscreensaver is currently active and
locked, it will print the unix timestamp of when the lock was activated.
|
| |
|
|
| |
Closes #3364
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
| |
If the wakeup is scheduled exactly at the time a scheduled
shutdown is, the cronjob for the shutdown would not run, as
the client is still about to wake up at that time.
If we wake up two minutes early, the cron daemon should have
enough time to fully wake up, notice the clock jump, and then
finally trigger the scheduled shutdown.
|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
............................................________
....................................,.-‘”...................``~.,
.............................,.-”...................................“-.,
.........................,/...............................................”:,
.....................,?......................................................\,
.................../...........................................................,}
................./......................................................,:`^`..}
.............../...................................................,:”........./
..............?.....__.........................................:`.........../
............./__.(.....“~-,_..............................,:`........../
.........../(_....”~,_........“~,_....................,:`........_/
..........{.._$;_......”=,_.......“-,_.......,.-~-,},.~”;/....}
...........((.....*~_.......”=-._......“;,,./`..../”............../
...,,,___.\`~,......“~.,....................`.....}............../
............(....`=-,,.......`........................(......;_,,-”
............/.`~,......`-...............................\....../\
.............\`~.*-,.....................................|,./.....\,__
,,_..........}.>-._\...................................|..............`=~-,
.....`=~-,_\_......`\,.................................\
...................`=~-,,.\,...............................\
................................`:,,...........................`\..............__
.....................................`=-,...................,%`>--==``
........................................_\..........._,-%.......`\
...................................,<`.._|_,-&``................`\
|
| |
|
|
|
|
|
|
| |
This used to work around a bug in the xserver where it would always
reset the idle time whenever the screen goes to standby or suspend.
This does not happen anymore on a recent system, so xprintidle would
erroneously add the timeout values to the idletime, resulting in
incorrect high results when the screen is turned off.
|
| |
|
|
|
|
|
| |
github misidentifies lots of includes in its code stats and
syntax highlighting. Try to make it a nicer place by giving
it a hint via that shebang. It's actually useless in files
that are being sourced, but it doesn't hurt either.
|
| |
|
