Tracking power leackage in my freerunner

[rixed at happyleptic.org]

-[ Thu, Jan 28, 2010 at 11:48:01AM +0300, Paul Fertser ]----
> Your figure is good. #1024 fix saves you about 4-8mA, a considerable
> amount during suspend-to-ram, but you won't notice the difference if
> you keep the device in "idle" but not suspended state.

Ok, make sense.
So I checked charge_now during suspend (measuring before and after
suspend of course) and it seams that the battery is then discharging
at the rate of 23mA, which leads to a total of approx 50 hours.
Much less than expected. :(