Open Stack

On 03/13/2014 04:56 PM, Sean Dague wrote:
> On 03/13/2014 04:29 PM, David Kranz wrote:
>> On 03/13/2014 10:50 AM, Joe Hakim Rahme wrote:
>>> On 10 Mar 2014, at 22:54, David Kranz <dkranz at redhat.com> wrote:
>>>
>>>> There are a number of patches up for review that make various changes
>>>> to use "six" apis instead of Python 2 constructs. While I understand
>>>> the desire to get a head start on getting Tempest to run in Python 3,
>>>> I'm not sure it makes sense to do this work piecemeal until we are
>>>> near ready to introduce a py3 gate job. Many contributors will not be
>>>> aware of what all the differences are and py2-isms will creep back in
>>>> resulting in more overall time spent making these changes and
>>>> reviewing. Also, the core review team is busy trying to do stuff
>>>> important to the icehouse release which is barely more than 5 weeks
>>>> away. IMO we should hold off on various kinds of "cleanup" patches
>>>> for now.
>>> +1 I agree with you David.
>>>
>>> However, what's the best way we can go about making sure to make this a
>>> goal for the next release cycle?
>> Basically we just need to decide that it is important. Then we would set
>> up a non-voting py3.3 job that fails miserably. We would have a list of
>> all the changes that are needed. Implement the changes and turn the
>> py3.3 job voting as soon as it passes. The more quickly this is done
>> once it starts, the better, both because it will cause rebase havoc and
>> new non-working-in-3.3 stuff will come in. So it is best done in a
>> highly coordinated way where the patches are submitted according to a
>> planned sequence and reviewed immediately.
> So it's important that there is a full plan about how to get there,
> including the python 3 story for everything in requirements.txt and
> test-requirements.txt being resolved first.
>
> Because partial work is pretty pointless, it bit rots. And if we can't
> get to running tempest regularly with python3 then it will regress (I
> would see us doing an extra python3 full run to prove that).
>
> 	-Sean
Yes, and we are at the "top" of the tree in the sense that we depend on 
a lot of other packages but none (yet) depend on tempest.
It is not clear we will actually finish this in Juno.

  -David
>
>
>
> _______________________________________________
> OpenStack-dev mailing list
> OpenStack-dev at lists.openstack.org
> 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/20140313/0c50d5b2/attachment.html>