From: Karl Kurbjun <kkurbjun_at_gmail.com>
Date: Wed, 15 Sep 2010 00:28:20 -0600

On 09/15/2010 12:17 AM, Dominik Riebeling wrote:
> On Wed, Sep 15, 2010 at 7:34 AM, Karl Kurbjun<kkurbjun_at_gmail.com> wrote:
>
>> I think the intent of the message was clear. This is not a 100+ line
>> commit. It is not even a 10 or 5 line commit.
>>
> The amount of lines (or even characters) changed with a commit does
> not necessarily correlate with the description that is appropriate. A
> 100+ lines commit could justify a single line commit message only (I
> just need to remember the days when we had those "code police" commits
> which was completely sufficient. Of course those could have gotten
> comments like "fix indentation, remove tabs, and kill trailing spaces"
> or whatever but in such a case a 100+ lines commit wouldn't need more
> explanation). In contrast I can very well imagine a single line change
> that requires quite a bit of explanation since it fundamentally
> changes things (just think of some macro that is used all over the
> place).
>
>
> - Dominik
>

Agreed, but that is not the situation we are talking about. The
situation at hand is a semicolon error.

I have seen some of those "code-police" changes have far more subtle
effects that change how the code is compiled. That said, it is
relatively low risk which is why it doesn't justify a large description.

Similarly, I think the point could be made that adding a syntactical
semicolon is also low risk and does not need a detailed description.

If this was one of the scenarios you described I would agree that a
better message would be required but I still feel that this discussion
is a bit silly considering the circumstance.

-Karl
Received on 2010-09-15