On Mon, Jun 02, 2014 at 11:55:33AM -0400, Russell Bryant wrote:
On 05/29/2014 06:08 PM, Dan Smith wrote:
>> I guess what I am driving at is that the process of creating a tag in
>> the OpenStack client projects occurs at pretty arbitrary points in
>> time based on the needs of other OpenStack projects that want to set
>> requirements on them rather than anything relating to the needs of
>> downstream distributions such as RDO or RHELOSP. Because no other
>> OpenStack project needed the particular functionality (and fixes)
>> added to python-novaclient in the last three months no new tag was
>> requested nor created. In this case it means we're missing some 84
>> odd commits made since the 2.17.0 tag was created.
>
> Yup, exactly. We've had features that were cross-project that had client
> changes required. We'd push the change into nova, then push into
> novaclient, tag the novaclient, update requirements for the other
> project, push that change, etc.
>
> On the other hand, we also have "huh, we haven't done a client release
> in a while" moments. For examples like nova events, instance groups,
> etc, I think it makes plenty of sense to stay current on the client
> packages.
>
>> Given this what I'm wondering is if there is any reason we shouldn't
>> move to a model where we rebase the python-*client packages to the
>> latest git commit at each milestone (J-1, J-2, J-3, RC, GA),
>> regardless of the existence of a tag, to ensure we are always picking
>> up the latest changes?
>
> Assuming proper testing of course, +1 from me.
Agreed. It really should be safe to use from any commit. That's
certainly the intention. As Dan mentioned, we're completely
non-disciplined about when a client actually gets tagged. Practically
that's just about getting something uploaded to pypi occasionally, and
that's not terribly relevant for us anyway.
--
Russell Bryant
What about the point that the CI runs on what gets uploaded to pypi
and therefore the random commits are not as thoroughly tested?
Dave