[Development] Cherry picking to replace a change set

[Oswald Buddenhagen]

On Mon, Sep 02, 2013 at 08:16:17AM -0700, Thiago Macieira wrote:
> On segunda-feira, 2 de setembro de 2013 12:55:38, Oswald Buddenhagen wrote:
> > On Sat, Aug 31, 2013 at 03:08:56PM -0700, Thiago Macieira wrote:
> > > Of course, each commit must stand on its own and be self-contained (it has
> > > to compile and should hopefully pass all tests). If you have to choose
> > > between atomicity and self-containment, prefer self-containment.
> > 
> > atomicity implies self-containment. it goes both ways. you can submit
> > neither quarks nor molecules.
> > http://qt-project.org/wiki/Commit_Policy 8.1 is pretty clear on that.
> 
> That defines what atomic is. It doesn't say that the commit must compile and 
> pass all tests if the rest of the commits in a topic are ignored.
> 
in fact, point 4 of the commit policy is pretty clear on that matter. it
is absurd to remove function (specific to the scope of the commit) from
the definition of atomicity.
also, the policy does not know a "topic" concept, for good reasons. you
cannot use topics (or branches which you intend to merge, for that
matter) as an excuse for violating the policy.
at the very most you can temporarily introduce chunks of dead code if it
does not affect the function of configurations which are expected to
work at all times. but even that is a stretch and should be done only
for big changes.