Open Stack

Sean Dague wrote:
> On 06/21/2017 01:10 PM, Michał Jastrzębski wrote:
>> One of key components which, imho, made SIGs successful in k8s is
>> infrastructure behind it.
>>
>> When someone proposes an issue, they can tag SIG to it. Everyone in
>> this SIG will be notified that there is an issue they might be
>> interested it, they check it out and provide feedback. That also
>> creates additional familiarity with dev toolset for non-dev sig
>> members. I think what would be important for OpenStack SIGs to be
>> successful is connecting SIGs to both Launchpad and Gerrit.
> 
> I think this is a key point. The simpler tools that github has, which
> require that you build a workflow based on tags outside of the tools,
> actually enables the effectiveness here.
> 
> Does k8s community currently have the same level of operators that
> aren't developers participating as OpenStack?
> 
> I wonder if we're going down this path, if some kind of tooling like
> standard tags for issues/patches should be added to the mix to help gain
> the effectiveness that the k8s team seems to have here.

For Launchpad/Storyboard we could totally reuse tags. Absence of tags in
gerrit bites us here (and in other places too). I know that was a
planned feature, does anyone have updated status on it ?

> I also think it's fine to rebrand WG to SIG, but we should also be
> honest that it's mostly a rebrand to consolidate on terminology that k8s
> and cncf have used that people find easier to understand so it's a way
> in which openstack is not different than those. Consolidating on terms
> isn't a bad thing, but it's really a minor part of the workflow issue.

It's both a consolidation and the signal of a change. If we continued to
call them "workgroups" I suspect we'd carry some of the traditions
around them (or would end up calling them new-style WG vs. old-style WG).

> It might also be a good idea that any SIG that is going to be "official"
> has the requirement that they write up a state of the sig every month or
> two with what's done, what's happening, what's next, and what's
> challenging. At a project the scale of OpenStack one of the biggest
> issues is actually having a good grasp on the wide range of efforts, and
> these summaries by teams are pretty critical to increasing the shared
> understanding.

++

Previously in the thread, we mentioned the need to clean up if SIGs are
no longer alive -- that regular reporting could be a good indicator of
liveness.

-- 
Thierry Carrez (ttx)