ChangeLog's: do we have to?

[Florian Weimer]

* Paul Koning:

>> On Jul 5, 2018, at 10:43 AM, Florian Weimer <fw@deneb.enyo.de> wrote:
>> 
>> * Aldy Hernandez:
>> 
>>> Can someone refresh my memory here, what are the remaining arguments for 
>>> requiring ChangeLog entries?
>> 
>> ChangeLog entries are part of the review, commit messages are not, so
>> you end up with surprises there.  At least that's what happens in
>> glibc.
>
> That issue could be fixed easily by having the proposed commit message
> reviewed.  That's approximately what we have today since the commit
> message is basically the change log entry, with perhaps another
> sentence or two.

We haven't been successful in implementing that as part of glibc patch
review.  To some degree, it's a bit of a chicken-and-egg problem
because “git am” tends to choke on ChangeLog patches (so we can't
really use it today) but even when it does not, I have yet to see a
posted patch where it produces the intended commit message. 8-(

Of course, your experience could be entirely different.