Open Stack

On 01/03/2017 07:30 PM, Emilien Macchi wrote:
> I've noticed some TripleO core reviewers self approving patch without
> respecting our review policy, specially in tripleo-quickstart.
> 

This is slightly misleading. To me, self-approving is +2/+A on your own
patch.

What has been going in tripleo-quickstart is different though. We have
allowed cores to +A a patch from another core with only a single +2.
That is against the policies laid out in tripleo-specs[1,2]. However,
following those policies will effectively make it impossible for cores
of tripleo-quickstart to get their own work merged in anything
approaching a reasonable amount of time.

This is because there are currently only 3 cores reviewing
tripleo-quickstart with any regularity. So the policies below mean that
any patch submitted by a core must be reviewed by every other core. I
think it has actually been a full month since we even had all 3 cores
working at the same time due to holidays and PTO (currently we only have 2).

If we want to apply the policies below to quickstart, I get it... they
are after all the agreed on policies. I think this puts moving CI to
quickstart this cycle at a very high risk to complete though, which also
means getting container CI is also at risk.

[1]
http://specs.openstack.org/openstack/tripleo-specs/specs/policy/expedited-approvals.html#single-2-approvals
[2]
http://specs.openstack.org/openstack/tripleo-specs/specs/policy/expedited-approvals.html#self-approval