Open Stack

On 3/18/2016 5:46 AM, Markus Zoeller wrote:
> Matt Riedemann <mriedem at linux.vnet.ibm.com> wrote on 03/16/2016 09:49:06
> PM:
>
>> From: Matt Riedemann <mriedem at linux.vnet.ibm.com>
>> To: "OpenStack Development Mailing List (not for usage questions)"
>> <openstack-dev at lists.openstack.org>
>> Date: 03/16/2016 09:50 PM
>> Subject: [openstack-dev] [nova] Reminder to move implemented nova
>> specs from mitaka
>>
>> Specs are proposed to the 'approved' subdirectory and when they are
>> completely implemented in launchpad (the blueprint status is
>> 'Implemented'), we should move the spec from the 'approved' subdirectory
>
>> to the 'implemented' subdirectory in the nova-specs repo.
>>
>> For example:
>>
>> https://review.openstack.org/#/c/248142/
>>
>> These are the mitaka series blueprints from launchpad:
>>
>> https://blueprints.launchpad.net/nova/mitaka
>>
>> If anyone is really daring they could go through and move all of the
>> implemented ones in a single change.
>>
>> --
>>
>> Thanks,
>>
>> Matt Riedemann
>>
>
> Is there a best practice how to handle a partially implemented bp (with
> spec file)? For example [1] needs additional effort during Newton to
> finish it.
>
> References:
> [1] https://blueprints.launchpad.net/nova/+spec/centralize-config-options
>
> Regards, Markus Zoeller (markus_z)
>
>
>
> __________________________________________________________________________
> 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
>

John was just telling me about this yesterday. I guess one thing we can 
do is add a "(partial)" suffix to the title and mark the blueprint 
complete for mitaka, and then the idea is to create a new blueprint for 
newton for continuing the work, e.g. centralize-config-options-newton.

The idea being we show that something was completed in mitaka when 
you're looking at blueprints in launchpad for mitaka.

I'm generally OK with that approach, the thing I don't really like is 
when we have to re-propose specs and/or if there are dependent 
blueprints in launchpad. Because creating the new blueprint means you 
have to update the link in the spec when re-proposing it and you need to 
update all of the dependent specs in launchpad for the new newton spec. 
Maybe it's not a big deal, I can see benefits to either approach. 
Personally I don't like to consider a blueprint complete until it's 
actually complete, like has been the case with some of the cells v2 
blueprints we've re-proposed for newton.

With long cleanup efforts like objects and config options though, I can 
see how having release-specific blueprints is good.

Markus, so to answer your original question, :), I'd probably mark the 
existing bp as complete for mitaka and create a new 
centralize-config-options-newton blueprint.

-- 

Thanks,

Matt Riedemann