[openstack-dev] [tc][docs] documenting openstack "constellations"

Doug Hellmann doug at doughellmann.com
Wed May 2 21:28:40 UTC 2018

Excerpts from Graham Hayes's message of 2018-05-02 20:15:14 +0100:
> On 02/05/18 20:11, Doug Hellmann wrote:
> > Excerpts from Zane Bitter's message of 2018-05-02 11:38:55 -0400:
> >> On 01/05/18 16:21, Doug Hellmann wrote:
> >>> Excerpts from Andreas Jaeger's message of 2018-05-01 21:51:19 +0200:
> >>>> On 05/01/2018 04:08 PM, Doug Hellmann wrote:
> >>>>> The TC has had an item on our backlog for a while (a year?) to
> >>>>> document "constellations" of OpenStack components to make it easier
> >>>>> for deployers and users to understand which parts they need to have
> >>>>> the features they want [1].
> >>>>>
> >>>>> John Garbutt has started writing the first such document [2], but
> >>>>> as we talked about the content we agreed the TC governance repository
> >>>>> is not the best home for it, so I have proposed creating a new
> >>>>> repository [3].
> >>>>>
> >>>>> In order to set up the publishing jobs for that repo so the content
> >>>>> goes to docs.openstack.org, we need to settle the ownership of the
> >>>>> repository.
> >>>>>
> >>>>> I think it makes sense for the documentation team to "own" it, but
> >>>>> I also think it makes sense for it to have its own review team
> >>>>> because it's a bit different from the rest of the docs and we may
> >>>>> be able to recruit folks to help who might not want to commit to
> >>>>> being core reviewers for all of the documentation repositories. The
> >>>>> TC members would also like to be reviewers, to get things going.
> >>>>>
> >>>>> So, is the documentation team willing to add the new "constellations"
> >>>>> repository under their umbrella? Or should we keep it as a TC-owned
> >>>>> repository for now?
> >>>>
> >>>> I'm fine having it as parts of the docs team. The docs PTL should be
> >>>> part of the review team for sure,
> >>>>
> >>>> Andreas
> >>>
> >>> Yeah, I wasn't really clear there: I intend to set up the documentation
> >>> and TC teams as members of the new team, so that all members of both
> >>> groups can be reviewers of the new repository.
> >>
> >> +1. What would be the reviewer criteria? Majority vote for adding new 
> >> constellations and 2x +2 for changes maybe? Or just 2x +2 for 
> >> everything? Just to clarify, since the TC reviews usually operate pretty 
> >> differently to other code reviews.
> >>
> >> thanks,
> >> ZB
> >>
> > 
> > This is just documentation, so 2x +2 is what I had in mind.
> > 
> > Doug
> > 
> Is it just docs though? I thought the the point of constellations
> was for a curated set of projects per use case?

The point of constellations is to help users find the components
they need to achieve their goals, by giving them smaller versions
of the OpenStack component map to navigate. It's the sort of thing
a product management group would produce as part of the product
documentation, if we had such a group.

If someone disagrees strongly about the components of a particular
constellation, some items can be listed as alternatives or a
completely separate constellation can be created to support the
different use case. Neither solution requires the TC to, as an
overall governing body, make a decision on for the entire community
using our most strict consensus model for voting.

If we set it up in a way that only the TC can make decisions about
the contents of constellations, then we'll never produce a useful
number of them.


More information about the OpenStack-dev mailing list