ACPI time remaining

[David Zeuthen]

Richard Hughes wrote:
> Isn't it the best to use in this situation? I could see how a running
> average (or difference of squares) over N might be a better (even though
> rate is discrete) w.r.t processor load, but isn't that overkill?

Yeah. We can always refine it though.

>>You need to poll for battery status on ACPI anyway, so just put it in
>>the main hald code for now.
> 
> I thought the addon would use the acpi event interfaces as discussed
> earlier? I can knock up a quick and dirty addon to do this if you so
> desire.
> 

At least on my IBM T41 laptop I don't get any events when battery charge 
status changes. Btw, I could actually open /proc/acpi/event even though 
acpid was using it. Fun.

So, saturday I wrote an addon for getting the ACPI events and kicking 
hal to Rescan() the device but I haven't committed it yet. Will do soon.

> You've already added battery.remaining_time.... if you mean for the other
> stuff then it's here:
> 
> http://hughsie.no-ip.com/write/hal/battery-rate.patch
> 

Hmm, I don't like that you changed battery.remaining_time from seconds 
to minutes, that doesn't make much sense to me :-). Please include the 
patch for the spec along with the patch for the code.

(btw, please attach patches to the mail so I and others don't have to 
wget it or use a browser to view it. Thanks.)

Cheers,
David
_______________________________________________
hal mailing list
hal at lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/hal