Re: gpe-config-sleep(gpe-isgui) in CVS

--- Moray Allan <moray_at_sermisy.org> wrote:
> On Mon, 2002-08-19 at 14:42, Erik Karlin wrote:
> > I started implemented reading/writing the backlight values directly, as you
> can
> > see from the #ifdef READ_BACKLIGHT stuff. Since I don't have the kernel
> > includes yet, I could not even compile that stuff to test it. If/when I get
> the
> > includes sorted out, I can make those changes in gpe-config-sleep and
> > gpe-conf/sleep if you like.
>
> Erm, why do you think that reading directly from the hardware in the
> configuration program is neater than calling the (setuid) bl? Even apart
> from concerns with privileges, it would seem nicer to me if people
> didn't have to arrange to have kernel headers around to compile it.
> Also, it's easier for us if we leave the responsibility to work out how
> to read the brightness on future hardware with the distribution, rather
> than having to implement it ourselves.
>
Yea, I suppose. I'm certainly all for passing off responsibility..hehe.
And, I haven't been able to easily obtain the kernel headers either, which is
why I haven't turned it on. I am doing all this stuff _on_ my ipaq.

I put it in because it was only a few lines of code. Most of bl.c is parsing
the command line. I figured that if there was a change in the future for
handling the backlight, i'd either have to recode the "system" call to bl or
recode the "ioctl" call to the backlight...6 -- 1/2 dozen thingy.

In the end, it doesn't bother be one way or the other.

-Erik

__________________________________________________
Do You Yahoo!?
HotJobs - Search Thousands of New Jobs
http://www.hotjobs.com
Received on Mon Aug 19 2002 - 16:40:04 EDT