Open Stack

Excerpts from Chris Dent's message of 2017-07-17 11:32:28 +0100:
> On Sun, 16 Jul 2017, Jeremy Stanley wrote:
> 
> > The incentives at play in free software are mostly still the same
> > incentives which drive other typical psychological/sociological
> > patterns in human society. As is the case with just about anything,
> > if you want something done right (or even at all sometimes) in
> > OpenStack you need to do it yourself... and that's why SIGs probably
> > need to be at least 50% comprised of people who will actually be
> > implementing the things the SIG members want to see, or else those
> > things ultimately won't happen.
> 
> I agree 100% with your conclusion that sigs need to (at least
> eventually) include the people who will be doing implementation.
> 
> However, I think it is disingenuous to expect that free and/or open
> source software physics plays a large part in the development of
> OpenStack. Not simply because most of the developers are paid (well
> paid in the grand scheme of things) to be here, nor merely because
> many find they must leave when they no longer have an employer who
> will sponsor their presence, nor because significant chunks of the
> developers do not actively work (although some do) on clouds on a
> day to day basis (so couldn't legitimately be said to be scratching an
> itch), nor because the ability to have influence and make change is
> often highly reliant on being willing and able to show up day in and
> day out (not just on those days when you have an itch), nor because
> many of the people who do have some license to choose what they do
> in this environment choose not what they want to but what they feel
> must be done. It is all of those things and more.
> 
> OpenStack is a commercial enterprise to collaboratively develop a
> suite of software with an open license. And that's great.
> 
> But it means the physics of getting things done are not as simple
> and hackneyed as people like to claim. I think it is perfectly
> legitimate for people to form a group around a special interest,
> for that interest to be important, and in the absence of available
> people to work on it seek assistance finding those resources.

I do, too. That's not how I read the proposal or other similar
proposals that basically boil down to some group coming up with an
idea and looking for an authority to tell people to implement it
(at different times, I've seen variations of the board, the TC, the
PTL, even a new "innovation TC").  My apologies to Melvin if I
misread things, this topic makes me a bit edgy.

We have been successful when contributors collaborate to build the
things they need.  We have been successful when we find ways for
users with needs to collaborate on resolving those needs with
contributors.  We have been less successful without the collaboration
component.

> With luck, because of the apparent obviousness of importance of the
> need, people will gather around it. Sometimes that doesn't
> happen. That's going to be normal because of the different
> physics, so we need additional strategies.
> 
> Things like community goals and the top 5 list driven by the TC can
> help. But sometimes the needs are not visible to them and the needy
> group is left out in the cold and has unanswered requests.

I still don't like that term "unanswered." Are we really not replying
to these things at all?  Are you using "unanswered" to mean "not
implemented"?

> In the past those requests have sometimes been directed to PTLs and
> the TC. Projects are already overwhelmed and have been so for
> years, playing a constant game of catchup or desperately trying to
> stay afloat in the face of decreasing resources. In our current
> commercial environment going to the projects is probably the wrong
> place to go. Should they go to the corporate members of the board?
> 
> We've heard arguments that OpenStack will not be able to rely on
> corporate largesse in the future and needs to come up with ways to
> be more driven by users and operators. That's all well and good and
> something we should strive for for other reasons than economics.
> While corporate entities are still making a pile of money off
> selling OpenStack-based products and services, I think there's still
> plenty of room for SIGs to make demands of the board.

I still think before we start trying to solve the problem with some sort
of fundamental change to our community, we should look at the details of
the past cases and generalize from there.

Doug