Open Stack

On 07/30/2014 01:22 PM, Kevin L. Mitchell wrote:
> On Wed, 2014-07-30 at 09:01 +0200, Flavio Percoco wrote:
>> As a stable-maint, I'm always hesitant to review patches I've no
>> understanding on, hence I end up just checking how big is the patch,
>> whether it adds/removes new configuration options etc but, the real
>> review has to be done by someone with good understanding of the change.
>>
>> Something I've done in the past is adding the folks that had approved
>> the patch on master to the stable/maint review. They should know that
>> code already, which means it shouldn't take them long to review it. All
>> the sanity checks should've been done already.
>>
>> With all that said, I'd be happy to give *-core approval permissions on
>> stable branches, but I still think we need a dedicated team that has a
>> final (or at least relevant) word on the patches.
> 
> Maybe what we need to do is give *-core permission to +2 the patches,
> but only stable/maint team has *approval* permission.  Then, the cores
> can review the code, and stable/maint only has to verify applicability
> to the stable branch…
> 

We could also just do this by convention.  We already allow a single +2
on backports if the person who did the backport is also on stable-maint.

We could add to that allowing a single +2/+A if the person who did the
backport is on project-core, or if a person from project-core has given
a +1.

-- 
Russell Bryant