Re: [Techteam] Major power concern

[C. Scott Ananian]

On Thu, Oct 2, 2008 at 7:58 PM, Chris Ball [EMAIL PROTECTED] wrote:
 From devel@:
Test 1 - default 765 image
Estimated time = ~2 1/2 - 3hrs

Test 3 - default 711 image
Estimated time = 4-5 hrs

 We'll need to set up tests with 765/767/711 tomorrow to see whether this
 is true.  We should suspect #8711, but Pia says she didn't launch any
 activities.

I'm not convinced that the testing methodology is sound: the machines
should be fully charged, with the same battery, and times taken until
fully discharged (until poweroff).  Trusting the battery's charge
monitor circuit's percentage is just Wrong, and extrapolating that to
a run time is Doubly Wrong.  You're just compounding estimates, which
are based on the battery's own rough knowledge of previous
performance.

Also, there was some suspicion that hovering over the battery icon in
767 could itself be leading to greater battery consumption -- the
popup menu for the battery is not present in 711; it would be best to
eliminate that uncertainty as well.
 --scott

-- 
 ( http://cscott.net/ )
___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel

Re: [Techteam] Major power concern

[Richard A. Smith]

C. Scott Ananian wrote:

 I'm not convinced that the testing methodology is sound: the machines
 should be fully charged, with the same battery, and times taken until
 fully discharged (until poweroff).

A 10 minute olpc-pwr-log using a delay 30 should be a reasonably 
accurate measurement.  If the Wh readings kicked out by the 
process-pwr_log.py are similar then the time can be increased.

-- 
Richard Smith  [EMAIL PROTECTED]
One Laptop Per Child
___
Devel mailing list
Devel@lists.laptop.org
http://lists.laptop.org/listinfo/devel