[openstack-dev] [Nova] The unbearable lightness of specs

Maxim Nestratov mnestratov at virtuozzo.com
Thu Jun 25 12:48:46 UTC 2015

25.06.2015 14:58, Sergey Nikitin пишет:
>             I've repeatedly stated that the fact that we created an
>             even smaller
>             clique of people to approve specs (nova-drivers which is a
>             tiny subset
>             of the already faaaar too small nova-core) is madness, as
>             it creates
>             an even worse review burden on them, and thus worsens the
>             bottleneck
>             than we already have.
>     I agree that the number of people who can approve specs is too
>     small at the moment. But relatively few people are actually
>     reviewing specs so building trust here is difficult. 
> I agree that the number of people who can approve specs is too small 
> at the moment too. But I disagree that we have few people who actually 
> reviews specs.
> For example I had a spec (approved for Juno and Kilo, but not approved 
> for Liberty) https://review.openstack.org/#/c/177112/ which has nine 
> +1 and zero +2. I implemented this spec a long time ago but I spent 
> about 2 months to reaprove spec in Liberty. And it's not reapproved yet.

I think it is a good example of improper approach to specs repository 
structure. Since specs are tied to release cycle, one should create 
specs for a new release cycle from scratch loosing all the history of 
considerations and discussion that happened in the past. It would be 
much better if corresponding to specs blueprints were responsible for 
tagging specs to particular release cycle, while unimplemented specs 
remained unmerged until their blueprints get status implemented. This 
would make it possible to reflect any differences that appear during 
code review in comparison with original spec.

> Especially strange for me is that not all nova cores have core status 
> in nova-specs repository. Now we have 14 nova cores and only 7 
> nova-spec cores. I think if a contributor worked hard to become a nova 
> core, he or she is competent to get a nova-spec core status.
> __________________________________________________________________________
> OpenStack Development Mailing List (not for usage questions)
> Unsubscribe: OpenStack-dev-request at lists.openstack.org?subject:unsubscribe
> http://lists.openstack.org/cgi-bin/mailman/listinfo/openstack-dev

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20150625/f6f7435e/attachment.html>

More information about the OpenStack-dev mailing list