DEBUG: Debug failing macpython wheels
might be because of a change in hypothesis
@phofl i've closed the nightly build issue. We may need an issue to unpin hypothesis if this PR gets closed.
Yeah we should probably open a new issue. Here or in the MacPython repo?
it does need to get unpinned in MacPython eventually, but there is not much visibility there, so an issue in pandas would be better. Do we know why we didn't see the same failures on pandas ci?
If this was an issue in test_round_sanity as suggested in https://github.com/pandas-dev/pandas/issues/47933#issuecomment-1205909274, we did xfail(strict=False) this test in https://github.com/pandas-dev/pandas/pull/47803
Still needed?