Re: [PATCH] Fix potential regressions in depcomp{3,5}.test.

[Stefano Lattarini]

On Wednesday 18 August 2010, Ralf Wildenhues wrote:
> * Stefano Lattarini wrote on Wed, Aug 18, 2010 at 08:35:20PM CEST:
> > On Wednesday 18 August 2010, Ralf Wildenhues wrote:
> > > Ideally, you'd commit this to a temporary branch based off of
> > > the patch that introduced the regression, and merge that
> > > branch into maint. If we follow this strategy consistently,
> > > then there one can simply merge the temp branch into all
> > > branches that have the parent.
> > 
> > Smart.  Would make another nice addition for HACKING ;-)
> 
> Already in HACKING, well, sort-of:
> 
> * For bug fixes of long-standing bugs, it may be useful to commit
> them to a new branch based off of the commit that introduced the
> bug, and merge this bugfix branch into active branches that
> descend from the buggy commit.
Ah-ah! *Now* this makes sense!  Though, it's not as clear as it could
be for "novices" (count me among them)...

What about this less terse and slighty extended formulation?

 * When fixing a bug (especially a long-standing one), it may be
   useful to commit the fix to a new branch based off the commit
   that introduced the bug.  Then this bugfix branch can be merged
   into all the active branches descending from the buggy commit.
   This offers a way to fix the bug easily and consistently.

I will propose a pach shortly in a new thread.
 
> > BTW, how should I call that temporary branch?
> 
> Dunno; whatever you like.  fix-depcomp-tests?  I name them fix-* in
> some projects.  It doesn't really matter for anything but the
> merge log entry.
Well, I'd rather avoid to have `merge branch foo' in the log :-)
I'll settle with the existing name of my branch:
`fix-decomp-tests-regression'.

Anyway, done the fixing, the merging, and the pushing.

Regards,
  Stefano