Re: Pybattery - battery statusbar applet with candies
 

Thanks, that did the trick. I don't know why but none of the versions have ever saved the settings for me. Only a minor annoyance.

Re: Pybattery - battery statusbar applet with candies
 

0.3.5-rc1
- ADDED: autoupdate for almost all updatable items
- FIXED: Some minor bugfixes
- WARNING: Candidate version with quite large cpu consumption

It is neither a stable version nor recommended update.
I just want to get some feedback about improvements and cpu consumption.

Re: Pybattery - battery statusbar applet with candies
 

0.3.5-rc2
- ADDED: autoupdate for almost all updatable items
- FIXED: Some minor bugfixes
- ADDED: Light sensor support for N810
- UPDATED: Acceptable CPU consumption now

It is neither a stable version nor recommended update.
I just want to get some feedback about improvements and cpu consumption.

Re: Pybattery - battery statusbar applet with candies
 

If I replace my original Nokia battery with Mugen, is the battery estimate then still right? What if I replace them back? Or back and forth?

Re: Pybattery - battery statusbar applet with candies
 

It's an interesting question. No testing and feedback are available yet. It all depends on HAL and battery module. I hope, they can get the things right.
So it would be really great, if you test it ;)

Re: Pybattery - battery statusbar applet with candies
 

Since it uses the same estimate as the Nokia applet, it doesn't give valid times remaining. % capacity is largely correct, but it doesn't know capacity is over 2x as much hours.

Re: Pybattery - battery statusbar applet with candies
 

Added to Fix-It for 0.3.5

Re: Pybattery - battery statusbar applet with candies
 

Is 0.3.5 going to use its own algorithm for estimating runtime left, then? Or am I missing something?

Re: Pybattery - battery statusbar applet with candies
 

That is what I'm doing right now, but this is hard to do objectively because the usage paterns differ. :(

Re: Pybattery - battery statusbar applet with candies
 

I'm still thinking about own time left algorithm (TLA)... On the one hand it would great not to deal with Nokia's time left stuff anymore, on the other hand... On the other hand AP is splitting into 2 packages:
- Advanced Power Monitor (APM) (a daemon (still written in Python, but C looks more and more attractive), which gets all the info and uses D-Bus to notify about changes and get values). Mainly for cross-project and cross-application usage. I want to keep APM lightweight as much as it can be using Python for such things.
- Advanced Power (GUI's statusbar item).
That splitting is my #1 at the moment. To my mind implementing a statistics based TLA is the thing to be done, but I'm not sure, if it gonna be in APM-0.3.5, may be in APM-0.3.6.
In fact ideas about other possible algorithms are welcomed ;)