From: Junio C Hamano <gitster@pobox.com>
Date: Wed, 11 Nov 2020 16:39:05 GMT
Subject: Re: [PATCH v2 06/26] test: completion: add run_func() helper
Message-ID: <xmqqsg9f4zme.fsf@gitster.c.googlers.com>
In-Reply-To: <CAMP44s0XA6QjOZTJaC5CK9Rp9ySfoH9_rJu-AoEGgprstzprfw@mail.gmail.com>

Felipe Contreras <felipe.contreras@gmail.com> writes:

> But even with no other reason for it, the patch stands on its own.
>
>> > +run_func ()
>> > +{
>> > +     local -a COMPREPLY &&
>
> This is the line that was smuggled in. It should be part of a separate
> patch, since this is behavior change.
> ...
> Do you want me to add: "In two places we generate an output that
> didn't exist before, but nothing ever reads it." ?

That would be very friendly to readers who may later wonder why the
change was made, yes.

In any case, I am not a shell-completion person, so even if I said
"yes that would make the patch perfect", that would not count as
much ;-)