bug#25105: M-p in diff-mode jumps too far

[Tino Calancha]

On Fri, 13 Jan 2017, npostavs@users.sourceforge.net wrote:

> Tino Calancha <tino.calancha@gmail.com> writes:
>
> > > I merged the bugs, and dropped #25400 from the cc list, because I'm
> > > getting duplicate emails everytime.
> > Thank you.  Sorry for the duplicated e-mails :-(
>
> Hmm, I'm still getting duplicates.  Maybe I shouldn't have merged?  Oh well.
Could be because i answered the e-mail to you and CC to Bug#25105? If you
are registered to that bug you might get 2 e-mails.
> > > I think the commit summary could be a bit friendlier, otherwise it looks
> > > good to me.
> > Thanks.  I agree: if i read my own commit after some months, probably
> > i will not fully understand it.
> > I propose following more verbose one:
> >
> > Fix Bug#17544
>
> I was thinking more about this summary line.  And maybe add some higher
> level explanation before the ChangeLog entry about the problems being
> solved?
How about copy verbatim the comentary from Dima in 2c8a7e5? He explained
very clearly the problem, and i am solving the same thing.
Mention to the regressions should must appear in the previous commit, 
where i do the actual revert.