Improving the suspend quirks guessworking

[Victor Lowther]

On Tue, 2008-03-25 at 21:07 +0100, Danny Kukawka wrote:
> On Dienstag, 25. März 2008, Matthew Garrett wrote:
> > On Tue, Mar 25, 2008 at 01:23:18AM +0100, Danny Kukawka wrote:
> > > I think the current behavior is much more helpful than you think. The
> > > current behavior force the people to report problems and quirks back to
> > > us. And the benefit in this case is bigger for many users (because we add
> > > the quirk and it work then for hundreds users) then use some strange
> > > fallback workarounds.
> >
> > Why is this beneficial? If the user doesn't get in touch with us, then
> > that's because their machine works.
> 
> Because only one person need to get in touch with us and all other can benefit 
> from this report. I don't see any problem here. Such a default would let the 
> machine work with one driver or kernel version and it would break maybe with 
> the next, where is there the benefit?

heh.  Somehow, I do not think that relying on user feedback will work as
the tools currently stand.  

It would be good to have a script that collects all relavent information
when a suspend or hibernate fails and populates it in a remote database
someshere for further perusal (assuming we get the uer's permission,
etc).

Pertinent information would (at a minimum) include the output of lshal
along with all the information we can reasonably gather about the video
card and any loaded modules.

Once a significant number for reports start coming in, spotting patterns
(and having hard data to back everything up) will be much easier.

Ideally, this would be an opt-in once process for the end user.

> Danny
> _______________________________________________
> hal mailing list
> hal at lists.freedesktop.org
> http://lists.freedesktop.org/mailman/listinfo/hal
-- 
Victor Lowther
Ubuntu Certified Professional