Open Stack

On Thu, Jul 28, 2016 at 02:29:18PM -0400, Zane Bitter wrote:
> On 28/07/16 12:54, Jay Pipes wrote:
> >The TC has given guidance on this already:
> >
> >http://governance.openstack.org/resolutions/20160119-stackforge-retirement.html
> >
> >
> >"In order to simplify software development lifecycle transitions of
> >Unofficial and Official OpenStack projects, all projects developed
> >within the OpenStack project infrastructure will be permitted to use the
> >“openstack/” namespace. The use of the term “Stackforge” to describe
> >unofficial projects should be considered deprecated."
> 
> The word "project" has unfortunately had multiple meanings throughout the
> history of OpenStack (I think it's something to do with multitenancy this
> week, right?), so to be clear: when I say 'project' here I mean in the sense
> of 'team'.
> 
> So I believe it's quite clear that there are official projects with official
> repos and unofficial projects with unofficial repos, and that all of these
> repos are hosted in the openstack/* namespace. (Nobody in the thread has
> raised the namespacing as an issue AFAICT.)
> 
> What's not clear is whether official projects should have unofficial repos.
> I submit that if that _were_ clear then this thread would never have existed
> and we would all be happier :)

Well, that does happen today, just as a note:
https://git.openstack.org/cgit/openstack/ironic-staging-drivers

This is a bunch of out-of-tree drivers that a few Ironic developers
support, because Ironic is beginning to require third-party CI on
in-tree drivers.

As mentioned elsewhere, the Neutron stadium has many repos in and out of
the Neutron governance and the big tent.

So I'm curious, if we say "official projects should never have unofficial
repos", where that bar would be. Is it that the repo is not worked on by
the PTL? Some percent of core reviewers? Some percent of active
developers? Has the name in it?

If we make a rule that doesn't fit some unofficial repo, people will
either 1) work around the rule, or 2) put it on Github. Both of those
are (IMO) not any better for the community than having some unofficial
repo related to an official project.

(Oh, there's another worse outcome: the repo becomes proprietary
instead.)

// jim

> >The Fuel CCP repos are projects that are not official OpenStack projects.
> 
> Because of the aforementioned 'project' pun issue there's two ways of
> interpreting this. You may be saying that the repos are unofficial repos
> within the "Fuel" project (team), in which case the question of whether
> official projects should have unofficial repos applies.
> 
> Alternatively, you may be saying that the "Fuel CCP" project (team) is an
> unofficial project separate from the "Fuel" project (team), with it's own
> (naturally unofficial) repos, and that therefore the question of whether
> official projects should have unofficial repos is moot. In which case I
> think you at least have to forgive people for being confused ;)
> 
> >They are in the openstack/ git namespace because they use the common
> >infrastructure and there isn't any formal plan to have the repos join
> >the "official OpenStack projects" (i.e. the ones listed in the
> >projects.yaml file in the openstack/governance repository).
> >
> >Could they be proposed in the future as official OpenStack projects?
> >Maybe. Not sure, and I don't believe it's necessary to decide ahead of
> >time.
> >
> >Please stop using a marketing press release as some indication of what
> >the "intent" is for these repos or even that there *is* any intent at
> >this point. It's really early on and these repos are intended as a place
> >to experiment and innovate. I don't see why there is so much anger about
> >that.
> 
> My only interest here is to try to help two groups that are clearly not
> communicating very well to communicate better. TBH I don't think your
> response was as helpful to those ends as it could have been. Can we start
> again?
> 
> cheers,
> Zane.
> 
> >Best,
> >-jay
> >
> >On 07/28/2016 12:33 PM, Steven Dake (stdake) wrote:
> >>Doug,
> >>
> >>Zane's analysis is correct.  I agree with Zane's assessment that TC
> >>clarification can solve this situation.
> >>
> >>Regards
> >>-steve
> >>
> >>On 7/28/16, 9:15 AM, "Zane Bitter" <zbitter at redhat.com> wrote:
> >>
> >>>On 28/07/16 08:48, Vladimir Kozhukalov wrote:
> >>>>Fuel-ccp repositories are public, everyone is welcome to participate. I
> >>>>don¹t see where we violate ³4 opens². These repos are now experimental.
> >>>>At the moment the team is working on building CI pipeline and
> >>>>developing
> >>>>functional tests that are to be run as a part of CI process. These
> >>>>repos
> >>>>are not to be a part of Fuel Newton release. From time to time we add
> >>>>and retire git repos and it is a part of development process. Not all
> >>>>these repos are to become a part of Big tent.
> >>>
> >>>It seems to me that there are two different interpretations of what it
> >>>means for a repo to be part of the OpenStack tent, and that these
> >>>differing interpretations are at the root of the arguments in this
> >>>thread.
> >>>
> >>>The first interpretation is that repos listed as belonging to a team in
> >>>the governance repo are part of a deliverable that is released each
> >>>development cycle, and that the same team may also control other repos
> >>>that are not deliverables and hence not part of OpenStack. It's easy to
> >>>see how people could have developed this interpretation in good faith.
> >>>
> >>>The second interpretation is that the TC blesses a team; that the only
> >>>criterion for receiving this blessing is for the project to be "one of
> >>>us", which in practice effectively means following the Four Opens; and
> >>>that all repos which the team intends to operate in this manner, subject
> >>>to TC oversight, should be listed in the governance repo. It's also easy
> >>>to see how people could have developed this interpretation in good
> >>>faith. (In fact, I was following the big tent discussions very closely
> >>>at the time and this was always my understanding of what it meant.)
> >>>
> >>>The only additional thing needed to explain this thread is the
> >>>(incorrect) assumption on behalf of all participants that everyone has
> >>>the same interpretation :)
> >>>
> >>>Assuming everyone holds the first interpretation, the current
> >>>designation of the fuel-ccp repo looks completely logical and the
> >>>complaints about it look like sour grapes.
> >>>
> >>>Assuming everyone holds the second interpretation, the current
> >>>designation of the fuel-ccp repo looks like an attempt to avoid TC
> >>>oversight in order to violate the Four Opens while using the name of an
> >>>official project (and issuing press releases identifying it as part of
> >>>said official project), and the complaints look like a logical attempt
> >>>to defend OpenStack from at least the appearance of openwashing.
> >>>
> >>>I believe this entire controversy will evaporate if the TC can clarify
> >>>what it means for a repository to be listed in the governance repo.
> >>>
> >>>cheers,
> >>>Zane.
> >>>
> >>>__________________________________________________________________________
> >>>
> >>>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
> >>
> >
> >__________________________________________________________________________
> >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