Short sprint on power management

[Thomas Pfeiffer]

Hi,
whenever I try to actually use my PA tablet, I painfully notice that 
power management on PA is currently pretty much broken, conceptually and 
as a result practically.
Currently there are three pretty serious bugs related to this in Bugzilla:
https://bugs.kde.org/show_bug.cgi?id=304877
https://bugs.kde.org/show_bug.cgi?id=310377
https://bugs.kde.org/show_bug.cgi?id=310506

The conceptual problem underlying these bugs is
- We have three ways to set energy-related options:
  - Settings -> Screen
  - Battery Plasmoid
  - "Energy Saving" (the Desktop KCM, can only be launched via search
    in the Launcher)
- The one which currently always trumps any others (regarding both 
automatic locking/suspend/shutdown and screen brightness) is Energy 
Saving, the one that is - rightfully - hidden from the user.
- This has the consequence that any changes users make using the methods 
which are visible to them (Battery Plasmoid and Settings) have no effect 
or are reset automatically after a short while

Another shortcoming is that the Screen module isn't very well thought 
through:
- "Turn off the screen" and "Sleep" don't make any sense as separate 
options in PA because after the screen is locked, the device goes to 
sleep anyway (turning off the screen as well, of course), so they can 
just be removed
- Imo it does make sense to have different settings for AC vs. battery 
power and low power on a tablet as well.
  - Maybe tablets are not connected to AC a lot, but especially when 
doing big updates, it makes sense to plug it in and especially during 
that, you don't want it to go to sleep every few minutes as it 
interrupts the process every time
  - When the battery is low, it does make sense to dim the screen

To summarize, our current power management is in desperate need of 
improvement, and this is something that affects every user. So if we 
spend a sprint on Alarms, which some users probably won't even need, no 
matter how good they are, we should spend one on power management as well.
Unless the "battling power management setting methods" problem proves to 
be very complicated to solve, a 1-week-sprint should suffice for this.

What do you think?