Hmm, all linux CI runs are failing, but not just on this PR, but on another one as well... wonder if there is some upstream issue.

The linux CI tests fail in pyvirtualdisplay, and that was recently updated, i.e. yesterday and the day before yesterday: https://pypi.org/project/PyVirtualDisplay/#history

It causes an issue in pytest-xvfb, already reported here:
The-Compiler/pytest-xvfb#22

We could either ignore it for now, awaiting the update in one of the above modules, or pin pyvirtualdisplay to version 0.2.5.

More info:

End of traceback:

  File "/usr/share/miniconda/envs/test-environment-3.6/lib/python3.6/site-packages/pyvirtualdisplay/abstractdisplay.py", line 313, in stop
    raise XStartError("stop() is called before start().")
pyvirtualdisplay.abstractdisplay.XStartError: stop() is called before start().

End of internal pytest error stack:

INTERNALERROR>   File "/usr/share/miniconda/envs/test-environment-3.7/lib/python3.7/site-packages/pytest_xvfb.py", line 46, in start
INTERNALERROR>     self._virtual_display.cmd.extend(self.args)
INTERNALERROR> AttributeError: 'Display' object has no attribute 'cmd'

@2xB figured it had to do w/ the virtual display, but hadn't had a chance to do anything more than glance at it... I'll read a bit more when I get the kids to bed and try and figure out a course of action....on one hand we should ignore, on the other hand, I hate red from the CI.

Closing/Reopening the PR to trigger a new CI run.

Looks great to me. Thanks @erikmansson !