Re: Should we restore manually maintained ChangeLogs

[Paul Eggert]

On 03/09/2016 08:04 AM, Eli Zaretskii wrote:
> > Cc: address@hidden, address@hidden, address@hidden
> > From: Paul Eggert <address@hidden>
> > Date: Tue, 8 Mar 2016 23:58:07 -0800
> >
> > > > My own experience is otherwise. For the kinds of development I do, I rarely see 
> > > > ChangeLog screwups now, whereas I used to see them routinely.
> > > With or without git-merge-changelog?
> > Without.
> Then I understand why your experience is so negative.

This appears to be implying that if I had installed git-merge-changelog 
and configured Git to use it, I would have had a better experience. I 
don't see how this would be so. As I said, I rarely do merges, so making 
merges work better wouldn't help me.

> > > Please describe the details of your proposal.
> > For the more-traditional approach, apply the attached patch to emacs-25, and
> > merge it to master.  Other branches can pick it up as needed.
> Didn't we consider this approach, and decided that having ChangeLog.2
> was better?

Yes and no. We came up with Emacs's current approach as a compromise. I 
would have preferred the approach taken by Guile (no ChangeLogs at all; 
just use 'git log' or whatever). Second-best would have been Coreutils 
(ChangeLog files in tarball, mainly to keep ChangeLog-file fans happy; 
they're not used by people doing active development). The approach 
currently taken by Emacs is a further compromise, which I think nobody 
likes.