Re: Is it time to drop ChangeLogs?

[Eric Abrahamsen]

Nikolaus Rath <address@hidden> writes:

> On Mar 08 2016, Eric Abrahamsen <address@hidden> wrote:
>> Nikolaus Rath <address@hidden> writes:
>>
>>> On Mar 07 2016, Stefan Monnier <address@hidden> wrote:
>>>>> When I submitted my first Emacs patch, I was astonished when I was asked
>>>>> to re-submit with my commit message essentially duplicated in the
>>>>> ChangeLog.
>>>>
>>>> I can't remember ever asking this kind of duplication.  We used to
>>>> duplicate the ChangeLog-info in the ChangeLog file and in the commit
>>>> message (tho we stopped doing it a year ago), but we never asked for it
>>>> to be duplicated in the email messages that submits a patch.
>>>>
>>>> So I have the impression that I'm misunderstanding you.  Could you
>>>> describe more precisely the kind of duplication you're talking about?
>>>
>>> No, you understood me correctly. However, I've just looked this up again
>>> and it turns out I was actually submitting a patch to Gnus, not
>>> Emacs. It was eventually committed by Eric Abrahamsen to the Gnus git
>>> repo, who first asked me to re-submit my patch with a commit message
>>> that was formatted like a Changelog entry, and to duplicate the message
>>> in the ChangeLog itself. It seemed to me that he considered that a
>>> questionable practice himself, but said it was obligatory because Gnus
>>> closely followed the Emacs standards here.
>>>
>>>
>>> Unfortunately that discussion never made it into the bug
>>> (http://debbugs.gnu.org/cgi/bugreport.cgi?bug=20956), so I'm
>>> paraphrasing from memory. But I'm pretty sure I'm remembering correctly,
>>> because I was so surprised by this.
>>
>> Not duplicating it in the email, just commit message + ChangeLog
>> duplication. This may be getting conflated with me asking for the commit
>> to be emailed as a git-format-patch attachment, which was just because I
>> was lazy (and it also preserves the author/committer distinction).
>
> Actually, I was quite happy about that. I very much prefer if patches
> that I submit also have me as the author of the commit.

That was my thinking, though as Eli and others point out, it's usually
done differently.

> (Eric, I think my last email sounded a bit like telling on you. I'm
> sorry about that, I think I should have written it differently (or asked
> you first). I'm very grateful for the time you took to help me with
> those patches, and for patiently explaining the rationale for the
> format).

No worries, I didn't read it that way! I'm very much not a maintainer,
just someone who occasionally patches Gnus when Lars is on vacation, so
it's been a learning experience for me, too.