Re: Commit messages

[Paul Eggleton <paul.eggleton@linux.intel.com>]

Hi Armin,

On Sun, 09 Oct 2016 20:29:39 akuster808 wrote:
> On 10/09/2016 05:35 PM, Paul Eggleton wrote:
> > I've been going through the commits in preparation for our release notes,
> > and unfortunately what I'm seeing is that the number of empty or
> > incomplete commit messages has noticeably increased over previous
> > releases. Commit messages are vitally important to providing a readable
> > history, not just in support of preparing release notes, but also if
> > you're trying to track down a regression or any other reason you might
> > want to go back and find out what has been done and most importantly
> > *why*.
> 
> Also helps in determining if a commit needs to be backported to stable
> branches.

Indeed.
 
> > Here are our commit message guidelines, which have been in place for some
> > 
> > time:
> >    http://www.openembedded.org/wiki/Commit_Patch_Message_Guidelines
> > 
> > I know everyone's busy, but please take 5 minutes to read these if you
> > haven't recently and try to keep them in mind when writing commit
> > messages. No commit is too trivial for a complete commit message.
> 
> Agreed. We also need more folks reviewing incoming patches to help
> spread the load.

That is indeed true.

> The other options is to should start rejecting commits that do comply to
> the guidelines.

Actually I forgot to mention one thing in my email - it's been a long time 
coming, but we should soon have a set of tests automatically run on incoming 
patches, which would immediately flag these kinds of basic things up before a 
human needs to look at the patch. More to come on that soon.

Cheers,
Paul

-- 

Paul Eggleton
Intel Open Source Technology Centre