Thomas Kluyver

What version of pytest are you seeing this with? Is there a full traceback of the error? The current code looks like fspath is optional: https://github.com/pytest-dev/pytest/blob/172c832cbdbfa407087ed6a895a298b48984da2a/src/_pytest/nodes.py#L633-L640

I think I'm looking at the newest version of that code. Can I ask you to check what version you're running? `pytest --version`

I don't think that was intended. I see there's also now a PR adding a failing test (#195). I'm a bit puzzled, though, because I thought this was up to...

Thanks for the detective work! That vaguely makes sense - IPython's `Out[n]` execution results are part of the history. But I don't see a good reason why telling it not...

That's very odd, I don't know why it would hang. If you interrupt it while it's hanging, do you get any traceback? There was a change to the logging setup...

> Do you have any ideas what I can do to further debug this issue? Not really, I'm afraid. Double check that it is definitely the log call hanging and...

On Windows I'd guess that the issue is trying to write to a nonexistant stderr; once some buffer fills up, the next write call will block. We've seen this before...

The suggested solution is to make writing to either stream flush the *other* one - so a write to stderr flushes stdout.

Yeah, the current behaviour is not actually wrong. But it can cause confusion (e.g. https://github.com/jupyter/notebook/issues/3070), so if it's practical to make it behave more like people expect, I don't think...

There are issues with putting the Spark web UI behind JSP - there's a (long) discussion on #57. At a minimum you need to configure Spark to tell it that...