I was playing with the Py3 tests for a week and:
- The testrunner is brought down by some failure categories
- All testruns will have failures in them so I cannot yet map isolation failures
There’s a py3 wrapper:
The current main benefit over zope.testrunner is the ZSERVER_PORT allocation for the Robot tests.
And there is an isolation failure test wrapper, but for this to actually yield anything useful, one would have to have
bin/test -t .EmptyLayer run without errors or failures first.