Open Stack

One of the highly reported issues was that when test discovery failed,
testr showed no useful details about the issue.

The root cause of this was an interaction between test *listing*,
which the standard library doesn't formally support (its on my list to
add it there) and discovery, which only knew how to report failures as
pseudo-tests.

I recently fixed that in the standard library (so it will be in Python
3.5) - but unittest2 0.8.0 has it backported for older Pythons, and
using testtools 1.2.0 or newer will export the needed data for the
import errors to be identified even when not running the tests. A
subunit point release will be needed in order to consume that data,
which will probably be 1.0.0.

I'll send a followup when everything is in releases.

Cheers,
-Rob

-- 
Robert Collins <rbtcollins at hp.com>
Distinguished Technologist
HP Converged Cloud