Commit 7b159b55d6398d2d240817276ae27ed68f11deda breaks ABI?

[Juan Pablo Salazar Bertin]

David Zeuthen <davidz <at> redhat.com> writes:

> 
> 
> On Wed, 2007-08-29 at 21:45 +0100, Matthew Garrett wrote:
> > The change to make GetBrightness calls return unsigned values breaks
> > g-p-m on 0.5.9, since it now assumes that values are unsigned and breaks 
> > with the signed ones. Could this be reverted rather than breaking 
> > visible ABI? Right now g-p-m effectively breaks on any released version 
> > of hal, which is less than ideal.
> 
> Yeah, it would be bad to break ABI. I can't remember why we did this and
> the thread the commit message points to
> 
> http://gitweb.freedesktop.org/?
p=hal.git;a=commitdiff;h=7b159b55d6398d2d240817276ae27ed68f11deda;hp=3734b53baf
d98672db52ce55678f80136a5af84c
> 
> yields no clues.  Anyway, I've reverted the patch for now. As a side
> effect g-p-m no longer complains with "Sleep problem" when I suspend via
> g-p-m (or close the lid) and FWIW brightness controls still work. Anyone
> know what made me commit that patch? Anyone? Bueller? Thanks.
> 
>      David
> 
> 


Read this thread: http://news.gmane.org/find-root.php?
group=gmane.comp.freedesktop.hal&article=8564

It was decided that it was better to change everything to return UINT32 
instead of INT32; otherwise you break it for the apps that are currently using 
it.