Open Stack

Josh,

We did have the discussion at summit.  We aren't even going to have a "big
spec".  We are going to run as bare minimal of a spec as possible.  The
reason I brought up spec additions is because Kolla has become so large,
managing the project using current techniques is not optimal.  I don't
like process but in the future, the PTL will need an easy on-ramp to
managing Kolla output.

I agree with all of the risks you point out related to usage of specs.  I
am hopeful we can make this is a super lightweight process without the
nitpicking that occurs in other projects.  That said, the rollcall vote
specs I expect to be as painful as they currently are (by design).

Regards
-steve


On 7/12/16, 4:03 PM, "Joshua Harlow" <harlowja at fastmail.com> wrote:

>Something for consideration to make the specs process not to painful and
>one that I think (?) glance pioneered is to have a 'bigger spec' and a
>'smaller spec' template.
>
>https://github.com/openstack/glance-specs/blob/master/specs/lite-specs.rst
> 
>(smaller)
>
>https://github.com/openstack/glance-specs/blob/master/specs/template.rst
>(bigger)
>
>Just my 2 cents, don't bog down people to much with just big specs and
>nit picking and trying to develop the whole feature in the spec (cause
>that will off-put new people and senior people and ...).
>
>-Josh
>
>Michał Jastrzębski wrote:
>> Hey guys,
>>
>> Since our project matured, we decided that we should have a discussion
>> regarding our spec process.
>>
>> https://etherpad.openstack.org/p/kolla-N-midcycle-specs
>>
>> Currently we do specs for most critical things, and they require
>>majority vote.
>>
>> We want to introduce another way, to enable non-nuclear specs.
>> To summarize our discussion so far:
>>
>> 1. Specs will require only 2 * +2 by default
>> 2. Specs sit at minimum 2 weeks in gerrit before first +2 arrives, so
>> other people will have time to look at it
>> 3. Any core can require a rollcall vote - then it becomes rollcall vote
>> 4. If nobody calls for rollcall vote, after 2 weeks spec can be merged
>> with normal 2 * +2
>>
>> Thoughts?
>> Michal
>>
>> 
>>_________________________________________________________________________
>>_
>> 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
>
>__________________________________________________________________________
>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