Maximilian Roos

Excellent @DanielGoman ! (not to use xarray to tout my personal projects — but re the doctest failures — running https://github.com/max-sixty/pytest-accept over them will auto-fix them, if that's helpful)

Great! If you add an `xfail` we can merge...

> Hi @max-sixty , what is a `xfail` ? https://docs.pytest.org/en/6.2.x/skipping.html — it marks the test as "expected failure". > Also, the issue turned out to be bottleneck-bug , ref: pydata/bottleneck#393...

@RubendeBruin it seems my guidance was bad — the test failures suggest that because this is a segfault `xfail` doesn't work, because the whole program crashes. So instead, let's leave...

You're right to remove the CET & UTC... the Google Cal should update for the relative timezones if people use that.

For sure @raybellwaves !

@noahjahn — thanks for sharing. FWIW we have an almost identical setup in GitLab (no ` DOCKER_DRIVER: overlay2` but I don't think that should matter) — unfortunately the caching is...

FWIW this is the config I use for IPython, to get cursors: ```python 3 │ import sys 4 │ from prompt_toolkit.key_binding.vi_state import InputMode, ViState 5 │ 6 │ 7 │...

If anyone has a workaround for this, that would be great — I'm planning on using `sed` or similar... Thanks in advance.

Ah, good point. We removed `parenthesized_expr` in favor of `(...)` always being an inline pipeline. I remember thinking that was a good change! Though it does mean we now have...