Re: [Gnu-arch-users] Re: <<< conflict markers

[Benoit Poulot-Cazajous]

Stefan Monnier <address@hidden> writes:

> > In fact, I think it would be quite handy to kill bad revisions by simply
> > tagging their ancestor into the current version, so you get this:
> 
> > patch-5 --commit---patch-6 (bad)      patch-7---commit---patch-8
> >      \                                /
> >       \---------------tag------------/
> 
> Neat!
> Too bad it doesn't work so cleanly if some other (non-bad) patch-7 was
> committed before you realized you wanted to kill patch-6.

It will also break badly when you replay from patch-4 (for instance), as tla
will apply patch-6...
What should tla do in such cases ?
Recognize that the path from patch-4 to patch-8 is
        patch-4 -> patch-5 -> patch-7 -> patch-8 ?
If my tree is already at patch-6, should "replay" apply patch-7 or say
"tree is already up to date" ?
IMHO, for "replay" to make sense in the mixed-tags-and-commits case, it should
compute the path to a goal revision (implicitly the latest revision of the 
current "branch") and follow it.

  -- Benoit