Re: When can we stop writing ChangeLogs?

[Florian Weimer <fweimer at redhat dot com>]

* Carlos O'Donell:

> On 10/3/19 9:37 PM, Gabriel F. T. Gomes wrote:
>> On Mon, 30 Sep 2019, Florian Weimer wrote:
>> 
>>> * Siddhesh Poyarekar:
>>>
>>>> On 30/09/19 2:36 pm, Florian Weimer wrote:  
>>>>> I'm looking for something far simpler here—a script that submitters and
>>>>> reviewers can run to see what's going to be committed, and some level of
>>>>> project consensus that this is how patches should be posted.  
>>>>
>>>> What kind of scripting are you thinking of?  pwclient should be able to
>>>> download a patch from patchwork and apply it, giving you an indication
>>>> of what a patch is going to look like.  
>>>
>>> The patch isn't the problem.  It's the commit subject, author and commit
>>> message that I'm after.
>> 
>> Supposing that patches created with git format-patch are posted, which can
>> then be applied with git am --scissors, do you foresee any problems with
>> Reviewed-by statements and friends?  I mean, are you trying to match
>> [exactly] the patch on the mailing list with the patch that has been
>> actually committed.
>
> The patch *should* conform to git am --scissors for cutting out non-patch
> parts.

To be clear here, the problem is not the scissors processing.  What I
find very confusing is that scissors does not drop anything.  It is just
a boundary between the commit message (taken verbatim) and the part
which is parsed as a patch.  This means that the part of the scissors is
not actually ignored by git am; it gets passed to the patch parser in
git apply, and what happens then entirely depends on the patch parser
(which may or may not match what the human eye sees).

Thanks,
Florian