Re: [RFC] laptop-detect being installed just on useful architectures

On Friday 08 February 2008, Otavio Salvador wrote:
> > 1) You did not answer my question whether laptop-detect really supports
> >    sparc. If it does not, we should not install it there (and
> > laptop-detect should not be built for sparc).
>
> It does, if kernel does.
>
> Currently laptop-detect checks dmi and if there're batteries[1] on the
> system, in those cases, it guess it's a laptop.

I still have some doubts whether it really makes sense to install it. I expect 99.999% of sparc systems that get Debian installed will be either workstations or, even much more likely, servers. Have you ever actually _seen_ someone with a sparc laptop?

I doubt that it is really justified to install laptop-detect for the remaining 0.001%, especially considering that quite a few of the packages in the laptop task are x86-centric anyway.

I really do not see any benefit here and would suggest we at least not install it by default for sparc in pkgsel until we actually see a user with a sparc laptop.
And even then I think that we could probably test fairly easily _inside_ D-I whether it is a sparc laptop or not by checking proc or the output of 'prtconf' instead of installing it by default for all sparc systems. Sparc has much better basic hardware reporting than x86 does.

> > 3) The laptop test in tasksel should be modified to _first_ test
> > whether laptop-detect exists or not. As it is now it will barf errors.
>
> I found it in two places and changed both them.

1. The test checks whether laptop-detect is executable twice as which itself already does that.
2. The test will still produce output.

Instead of:
+ if [ -x "`which laptop-detect 2>/dev/null`" ] && \ you should do:
+ if which laptop-detect >/dev/null && \ (Yes: >/dev/null and not 2>/dev/null)

-- 
To UNSUBSCRIBE, email to debian-boot-REQUEST@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org


application/pgp-signature attachment: This is a digitally signed message part.