making the tablet during charging time was one of the reason why I began working on the clock; in some of the previous version was present a couple of graphs with the battery and the light levels (then removed because to have access at that infos I always used weird os calls and godly permissions) but everything should be pretty ready for having some data to visualize, study or store. the Pygame lib (that the clock use for the graphics) have some sound support (that i prefered not use in the clock because was damn heavy on the cpu) that can be use for a specific application (but I should check if can access the tablet mic). Ill take a look because I need to fill a bit of empty space when the application dont run fullscreen.

@jolouis
-he swears about once a day that he's not a programmer-
many time a day: I know my limits
In case you'll have some result with that c app we should try to find a way to visualize something: just give me some numbers

I read somewhere about an app for iphone which had the same thing; monitoring your state of sleep and wakes you up when it is the best time to do that. Instead of a wristband accelometer or a mic you put your phone under your mattress and it senses your movement with its accelometers.

It would be cool if someone developed an app for maemo like this

During REM sleep the body does not move (only rapid eye movement -hence the term) -in fact the body is paralyzed and prevented from moving (this stops you acting out your vivid dreams which occur during this phase of sleep). It is during non-REM sleep that the body can move (and does so roughly every 20 mins or so for postural adjustment so you don't get bed sores). This is the deepest phase of sleep. The only reliable way to tell REM and non-REM sleep is to have scalp electrodes that pick up the very different patterns of brain waves associated with these two states of sleep. So monitoring movement will not give you a REM-aware alarm clock I'm sorry to say. If you used a bluetooth link to a device that could measure your brain waves, then you would get somewhere...

actually those apps aren't working very well. you can get approximately same results when you choose timeframe and let the app wake you up at random time.

there is no real way for the phone to figure in which phase of sleep you are into. don't know about bt heartbeat stuff but without it there is no difference between accelometer/mic data and using /dev/random

New version of Sleep Analyser allows for setting alarm time and sets alarm when it detects your movement:
http://maemo.org/downloads/product/M...sleepanalyser/