There are now quite a few spec freeze exceptions out there.
Realistically, I don't think they will all get up, mostly because of
concerns about review bandwidth coming up to the end of the release

The proposed process from here is that I will wait until my morning
tomorrow (12 hours from now) to see what cores step up to support
freeze exception requests. Those which get two cores will get an
extension. For those which don't have enough cores signed on by
tomorrow, we'll let it sit for a couple more days before we time out
on them.


On Fri, Jul 18, 2014 at 8:44 AM, Michael Still <mi...@stillhq.com> wrote:
> At our weekly nova meeting just now I promised that I'd send an email
> here explaining our intentions with the spec freeze for juno.
> The proposed process works like this, but is subject to change:
>  - you request a spec freeze exception on the openstack-dev mailing
> list. Please do this in the next 48 hours.
>  - spec exceptions with sponsoring nova-cores will be given
> preference. I'd like to see at least two cores sponsoring because then
> they can drive the code review process, but if you don't have cores
> yet you can still propose an extension to the list and see who
> expresses interest.
>  - the freeze will then be discussed by nova-core and nova-drivers. We
> need to address several things here: how likely it is that the spec
> will be approvable soon; whether landing the proposed code will be
> disruptive to our juno release; and review bandwidth for the change.
>  - if a freeze is granted, it will be in the form of a seven day grace
> period to get the spec merged. This likely means a rapid iteration on
> the spec to get it to somewhere there is consensus. You are therefore
> agreeing to be available for rapid iteration on the spec.
> Thanks everyone for your patience as we learn how this process works.
> So far we've approved 85 specs for Juno, which I think is pretty
> impressive.
> Cheers,
> Michael
> --
> Rackspace Australia

Rackspace Australia

OpenStack-dev mailing list

Reply via email to