Re: [PATCH 5/6] power: pcf50633: properly reenable charging when the supply conditions change

[Pavel Machek <pavel@ucw.cz>]

On Thu 2009-11-05 02:45:19, Anton Vorontsov wrote:
> On Thu, Nov 05, 2009 at 01:24:31AM +0300, Paul Fertser wrote:
> [...]
> > > If you want to "top-up" the battery, perhaps that should be on
> > > explicit request, like something in /sys?
> > 
> > Well, probably but to me it seems like an unnecessary complexity. Also
> > since Li* batteries are dominating the market nowadays this
> > functionality would need to be introduced in the power_supply core to
> > provide userspace with a consistent interface. Anton, can you please
> > comment on this?
> 
> Power supply class doesn't have writeable properties (initially
> it was a 'battery monitor class'). Though. I'd happily merge
> the support for writeable properties. ACPI batteries would
> benefit too, i.e. they could export 'alarm' property (see
> drivers/acpi/battery.c).

Yep, writable properties would be nice for at least Spitz and
Openmoko. Both of them have slow and fast charging modes, and
sometimes user will need to override the default.

> Though, I'd want to apply this patch set as is (i.e. I prefer 
> the new behaviour, as it's really more 'user friendly'), and
> we could implement configurable option later.
> 
> Pavel do you have a strong objection against this patch, or
> you can live with it? :-)

I can live with it... but thanks for asking :-).

								Pavel
-- 
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html