[PATCH 2/3] ddx: Prepare xf86 to renounce input handling duties

Peter Hutterer peter.hutterer at who-t.net
Wed Sep 8 20:20:26 PDT 2010


On Wed, Sep 08, 2010 at 04:32:09AM -0300, Fernando Carrijo wrote:
> Peter Hutterer <peter.hutterer at who-t.net> wrote:
> 
> > On Tue, Aug 31, 2010 at 04:00:13PM -0300, Fernando Carrijo wrote:
> > > From: Fernando Carrijo <fcarrijo at freedesktop.org>
> > > 
> > > Signed-off-by: Fernando Carrijo <fcarrijo at freedesktop.org>
> > > Signed-off-by: Tiago Vignatti <tiago.vignatti at nokia.com>
> > 
> > I think I'd prefer to see these three patches squashed together.
> > Patches don't always have to be strictly divided by subsystems and
> > introducing a struct in one patch to remove it in the next one from another
> > part is a bit confusing. Better to just move the struct in one go, that way
> > the diff shows that it has moved.
> 
> Okay.
> 
> > In the case of these patches, the only real new parts seem to be the
> > makefile and includes, everything else is more-or-less cut/paste.
> 
> Right.
> 
> > Also, please add to the commit message _why_ you're doing this. You
> > referenced http://lists.x.org/archives/xorg-devel/2010-August/012204.html
> > in your header email but I can't find any argument for moving input handling
> > from the DDX to the DIX in there.
> 
> I'll strive to find the balance between what you guys consider under- and
> over-documenting. In advance, the changes brought in this series regard only
> the modularization of common functionality in order to avoid code duplication.

it's really hard to overdocument in a commit message. code comments can go
stale, a commit message cannot.
and unless something is really really obvious, describing "why" in the commit
message is always a good idea.

Cheers,
  Peter


More information about the xorg-devel mailing list