[Debian-eeepc-devel] eeepc-acpi-scripts integration: please test

[Luca Niccoli]

On 13 May 2010 21:54, Damyan Ivanov <dmn at debian.org> wrote:

>> It's not that they can't change things; the point is that it's their
>> responsibility, they deal with the consequences.
>
> This is also true for stuff under /etc.

Not really, of course if they misconfigure things it's their fault;
but keeping the system working if they have done thing that used to
work and would be broken by an upgrade is our duty.
I think we'll have to agree to disagree here; this weekend I'll try to
summarise my points and yours, and post a mail asking for a broader
input.
If you prefer to sum up your position yourself, that will obviously be
welcome :)
I want to make a point clear though, this is obviously orthogonal to
the compatibility with acpi-support, which doesn't depend on it.
I only happened to notice which pain is to deal with scripts to change
and move if they're in /etc/ while working on it, that's why I moved
them to /usr/ in my tree.

> There is no wireless toggling in e-a-s. Only notification. There was
> wireless un/plugging in the early days, but it is no longer necessary,
> including the current squeeze kernel.

Until 2.6.32 it is possible to configure out RFKILL_INPUT (possibly
even in later kernels), and we surely don't want to break toggling for
people with self compiled kernels.
So there is toggling, even if it's skipped in the default squeeze
kernel; moreover, the script wakes up wicd, that would not reconnect
otherwise when an interface is reactivated.
All this is going to acpi-support though (see #581312), without OSD
Cheers,

Luca