On 5/14/07, James Rayner <[EMAIL PROTECTED]> wrote:
> to be honest. I think it's a stupid idea.

I don't think it's stupid. I think that *ATM* it shouldn't be
implemented for the reasons you explained. I agree that
CONFIG_TIMER_STATS should never be in the official kernel, but I hope
that CONFIG_USB_SUSPEND will someday, when software problems are
fixed, make it.
In particular, I'd like to comment on the two points I don't fully agree with:

> - 1w is nothing. It's not worth saving given the above downsides, that
>   could even cost more battery

Here, have fun:
http://www.crhc.uiuc.edu/~mahesri/classes/project_report_cs497yyz.pdf :-)
(first result for "laptop power consumption" in google).

A brief summary: a modern laptop system consumes between 7.5 and 30
Watts (figure 3). One watt so is between 3.3 and 13.3% of the total
power consumption. This means that for an office/desktop use you
should gain *at least* 5% more battery life taking other precautions.
If we take 3 hours of full battery life (it could be even more on
recent systems), then we could extend it of 9 minutes. That, if not
much, is still something in terms of office productivity. The
minimum/maximum calculated on this back-of-an-envelope analysis is
6/24 minutes of work. Not bad, I think.

> - You could save the aforementioned 1w, just by unloading your usb
>   driver. Far better solution than a kludgy config option.

Yes. And you wouldn't charge your phone/ipod/whatever. I'd rather put
that 1W in my ipod than wasting it.


bardo

_______________________________________________
arch mailing list
[email protected]
http://archlinux.org/mailman/listinfo/arch

Reply via email to