Code review required for commits - formal Team vote.

[simo]

On Mon, 2012-10-15 at 08:38 +1100, Andrew Bartlett wrote:
> On Sat, 2012-10-13 at 11:41 -0400, simo wrote:
> > On Sat, 2012-10-13 at 09:06 +0200, Kai Blin wrote:
> > > On 2012-10-13 04:23, simo wrote:
> > > > On Sat, 2012-10-13 at 02:23 +0200, Kai Blin wrote:
> > > >> On 2012-10-13 01:13, Michael Adam wrote:
> > > >> 
> > > >>> comments: * I don't yet understand the reasons for "-1"s
> > > >> 
> > > >> My reason for a -1, as I've tried to explain on both my mail and
> > > >> on IRC is that I'd like to try this out first, just like we did
> > > >> for autobuild. Jeremy however proposes a policy change now, and
> > > >> says if we don't like it later, we can try to get another policy
> > > >> change done. That's not what trying things out should look like. 
> > > >> Apparently that makes me opposed to progress, but I can live with
> > > >> that.
> > > > 
> > > > Hi Kai, can you explain what 'try it out' means in this case ?
> > > 
> > > To quote from my previous mail:
> > > 
> > > """
> > > The way I see a trial working was that everybody tries to do reviews
> > > on patches sent to samba-technical, and everybody agrees to not push
> > > code directly. Basically, we do whatever the initial proposal was,
> > > without actually having it as a rule. And then, once we tried it for a
> > > while, we have a vote on whether we want to make the system mandatory.
> > > """
> > > 
> > > Jeremy said on IRC that he doesn't believe this is going to work, but
> > > it's still how I'd prefer this would be handled.
> > 
> > Fine by me, I've been basically doing this except for the most trivial
> > patches for the past 2-3 years, so from my point of view I am well into
> > the trial, I have no problem prolonging it.
> 
> Simo,
> 
> I'm confused by what you are saying you have done.  I don't think you
> are saying that you have not pushed any of your own patches in the past
> 2-3 years, but I just want to check.

No, what I did is that I obtained review for my patches, whether I push
them or not would make basically no difference, so that rule doesn't
worry me. I am confident my samba team mates can push as well as review.

I am also not welded on non-author push. If you have a clear acked-by
line sent to you for a specific patch, IMO you can push it yourself.
However I see value in reviewer-push as that assures the reviewer
actually applied the patch and didn't just look at it in the email, so I
am ok with that rule too if most people like it.

Simo.

-- 
Simo Sorce
Samba Team GPL Compliance Officer <simo at samba.org>
Principal Software Engineer at Red Hat, Inc. <simo at redhat.com>