Re: todo-mode doc bug

[Stephen Berman]

On Wed, 27 Apr 2016 11:50:02 +0300 Eli Zaretskii <address@hidden> wrote:

>> From: Stephen Berman <address@hidden>
>> Cc: address@hidden
>> Date: Wed, 27 Apr 2016 10:19:55 +0200
>> 
>> >> There is no outstanding bug about this; my question was about whether it
>> >> is necessary to file one for this doc string change
>> >
>> > No need.  However, mentioning in the commit log the URL for the
>> > discussion of the issue should be considered.
>> 
>> This thread is the only discussion of this issue, so in the hope that
>> you agree with me that it isn't relevant to the doc bug per se and hence
>> wouldn't be appropriate to reference in the commit message, I've gone
>> ahead and committed the fix without such a reference.
>
> I often wonder why people need to assume something about what others
> think, instead of just asking them and waiting for an answer.  What's
> the rush?

Sorry.  I thought asking for further clarification of what you meant by
"discussion of the issue" was out of proportion to the significance of
this doc bug; in future I'll try to be less restrained.

> As it happens, I don't agree.  In this message:
>
>   http://lists.gnu.org/archive/html/emacs-devel/2016-04/msg00738.html
>
> you say more about the rationale for the fix than in the log message
> for the commit you just made:
>
>   Specifically, I noticed that the doc string of
>   `todo-show' is in part obsolete and incorrect -- in fact, it already was
>   when I initially committed this version of todo-mode to the Emacs
>   repository, because I had failed to update the doc string after making
>   the change in ea3ae33b5b94bf095c52f279887009e8fb71a9df prior to the
>   initial commit.  (In contrast, the Todo mode Info manual contains the
>   correct information.)
>
> Admittedly, I'm close to splitting hair here, but the references to a
> previous commit and to the manual are valuable and relevant
> information.
>
> That's why I generally advise to abandon doubt and simply include the
> reference in the log message, as no real harm could ever be done by
> doing that.

It didn't occur to me that you meant that specific message (it would
have be clear to me if you had referred directly to it rather than to
"discussion of the issue", but it's my fault for not asking for further
clarification.)  I guess the only repair it to revert my commit and then
make a new one with the reference.  Should I do that?

Steve Berman