Re: Missing changes in merges from emacs-25 to master

[Eli Zaretskii <eliz@gnu.org>]

> From: Stefan Monnier <monnier@iro.umontreal.ca>
> Date: Tue, 22 Mar 2016 12:30:52 -0400
> 
> >   . Phillip then said "git pull", followed by "git push".
> 
> That's a problem in itself, since it leads to annoying commit diffs that
> repeat other things, and it changes "git log" ordering in confusing ways
> (makes it seem like mainline's development is Philip's and when happened
> on master during this time was done on a branch).

Indeed.  But sometimes this cannot be avoided, because no matter how
quick you are, there's always a chance that someone else pushes
between your local commit and your push.  The probability of this
could be brought down significantly by "git pull" before the local
commit, but it cannot be eliminated altogether.

> >   . When John merged all of that to master, bb0cd31 was skipped.
> >     Skipping it omitted all the 9 commits that were merged by bb0cd31.
> > So the question (to John) now becomes: why was bb0cd31 skipped?
> 
> Probably because it's a merge commit, so its commit message contains
> "merge" which is in bzrmerge-skip-regexp.  We should probably remove
> "merge" from bzrmerge-skip-regexp (and use "backport" as the main magic
> keyword) or at least refine it to only match things like "don't merge".

You mean, gitmerge-skip-regexp.  Yes, I think that's the reason.