[openstack-dev] The danger of capping python-*clients in core projects, and forbidding it in the future

Julie Pichon jpichon at redhat.com
Thu Jul 11 14:51:25 UTC 2013


Hi Dirk,

"Dirk Müller" <dirk at dmllr.de> wrote:
> > Indeed. The whole idea behind a single release channel for python client
> > libraries was that you should always be running the latest, as they
> > should drastically enforce backward compatibility.
> 
> Well, backward compatibility can be tricky when it comes to test.
> We've for example recently had an issue where the newer keystoneclient
> broke mocking in tests. It is debateable if tests are part of the
> backward compatibility or not.
> 
> See for example https://bugs.launchpad.net/horizon/+bug/1196823

This is arguably a deficiency of mox, which (apparently?) doesn't let us mock properties automatically.

> This is currently also preventing me from being able to get a change
> on stable/grizzly past gating checks (which stumble on exactly this
> "regression").

The grizzly backport for the stubs has been merged [0] so patches should be unblocked now.

Regards,

Julie

[0] https://review.openstack.org/#/c/35309/



More information about the OpenStack-dev mailing list