Open Stack

Matt Riedemann wrote:
> How does the re-branding or re-categorization of these groups solve the
> actual feedback problem? If the problem is getting different people from
> different groups together, how does this solve that? For example, how do
> we get upstream developers aware of operator issues or product managers
> communicating their needs and feature priorities to the upstream
> developers?

My hope is that specific developers interested in a given use case or a
given problem space would join the corresponding SIG and discuss with
operators in the same SIG. As an example, imagine an upstream developer
from CERN, able to join the Scientific SIG to discuss with operators and
users with Scientific/Academic needs of the feature gap, and group with
other like-minded developers to get that feature gap collectively addressed.

> No one can join all work groups or SIGs and be aware of all
> things at the same time, and actually have time to do anything else.
> Is the number of various work groups/SIGs a problem?

I would not expect everyone to join every SIG. I would actually expect
most people to join 0 or 1 SIG.

> Maybe what I'd need is an example of an existing problem case and how
> the new SIG model would fix that - concrete examples would be really
> appreciated when communicating suggested governance changes.
> 
> For example, is there some feature/requirement/issue that one group has
> wanted implemented/fixed for a long time but another group isn't aware
> of it? How would SIGs fix that in a way that work groups haven't?

Two examples:

- the "API WG" was started by people on the UC side, listed as a UC
workgroup, and wasn't making much progress as it was missing devs. Now
it's been reborn as a TC workgroup, led by a couple of devs, and is
lacking app user input. Artificial barriers discourage people to join.
Let's just call all of them SIGs.

- the "Public Cloud WG" tries to cover an extremely important use case
for all of OpenStack (we all need successful OpenStack public clouds).
However, so far I've hardly seen a developer joining, because it's seen
as an Ops group just trying to make requirements emerge. I want the few
developers that OVH or CityCloud or other public clouds are ready to
throw upstream to use the rebranded "Public Cloud SIG" as a rally point,
to coordinate their actions. Because if they try to affect upstream
separately, they won't go far, and we badly need them involved.

Yes, it's mostly a rebranding exercise, but perception matters.
Hope this clarifies,

-- 
Thierry Carrez (ttx)