Open Stack

FWIW, from my perspective on the Project Navigator side, this format 
works great. We can actually derive the age of the project from this 
information as well by identifying the first release that has API data 
for a particular project. I'm indifferent about where it lives, so I'd 
defer to you all to determine the best spot.

I really appreciate you all putting this together!

Jimmy

> Thierry Carrez <mailto:thierry at openstack.org>
> April 5, 2017 at 5:28 AM
>
> Somehow missed this thread, so will repost here comments I made elsewhere:
>
> This looks good, but I would rather not overload the releases
> repository. My personal preference (which was also expressed by
> Doug in the TC meeting) would be to set this information up in a
> "project-navigator" git repo that we would reuse for any information we
> need to collect from projects for accurate display on the project
> navigator. If the data is not maintained anywhere else (or easily
> derivable from existing data), we would use that repository to collect
> it from projects.
>
> That way there is a clear place to go to to propose fixes to the project
> navigator data. Not knowing how to fix that data is a common complaint,
> so if we can point people to a git repo (and redirect people from there
> to the places where other bits of information happen to live) that would
> be great.
>
> Monty Taylor <mailto:mordred at inaugust.com>
> April 4, 2017 at 5:47 PM
> Hey all,
>
> As per our discussion in today's TC meeting, I have made a document 
> format for reporting versions to the project navigator. I stuck it in 
> the releases repo:
>
>   https://review.openstack.org/453361
>
> Because there was already per-release information there, and the 
> governance repo did not have that structure.
>
> I've included pseudo-code and a human explanation of how to get from a 
> service's version discovery document to the data in this document, but 
> also how it can be maintained- which is likely to be easier by hand 
> than by automation - but who knows, maybe we decide we want to make a 
> devstack job for each service that runs on tag events that submits a 
> patch to the releases repo. That sounds like WAY more work than once a 
> cycle someone adding a few lines of json to a repo - but *shrug*.
>
> Basing it on the version discovery docs show a few things:
>
> * "As a user, I want to consume an OpenStack Service's Discovery 
> Document" is a thing people might want to do and want to do 
> consistently across services.
>
> * We're not that far off from being able to do that today.
>
> * Still, like we are in many places, we're randomly different in a few 
> minor ways that do not actually matter but make life harder for our 
> users.
>
> Thoughts and feedback more than welcome!
> Monty
>
> __________________________________________________________________________ 
>
> 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

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstack.org/pipermail/openstack-dev/attachments/20170405/83852ab3/attachment.html>