pre-release for v2022.06.0

[keewis]

There's a few unreleased and potentially breaking changes in main, most importantly the index refactor and the new groupby using numpy-groupies and flox. During the meeting on Wednesday we decided to release a preview version to get feedback before releasing a full version, especially from those who don't run their tests against our main branch.

I am planning to create the pre-release tomorrow, but if there's any big changes that should be included please post here.

cc @TomNicholas

Edit: the version will be called 2022.05.0.dev0, which will ensure that e.g. pip will require the --pre flag to install it.

[headtr1ck]

If someone manages to fix #6624
That would be a good candidate.

[dcherian]

Doing a pre-release sounds good to me.

But we know of at least three regressions (tagged under "release-blocker"). Should we fix them first?

[headtr1ck]

This should be also merged: #6651

[keewis]

the cftime issues (#6624, #6613) look like they would be good to be fixed first, but since the remaining one (#6607) is also in 2022.3.0 I don't think we need to wait on that for the pre-release.

Something to remember, though, is that this is not a full release, so I don't think we need to have fixed every known bug / regression.

[dcherian]

The MVCE in #6607 does work on v2022.03.0. #6229 OTOH was broken earlier so technically not a regression

[keewis]

right, I misunderstood that MVCE. I still think that we don't need to wait on a fix to do the pre-release, though.

As far as I understand PEP440, to create a pre-release we just need add a rcN suffix (i.e. 2022.06.0rc0) or .devN (i.e. 2022.06.0.dev0. Do we need to update whats-new.rst for this?

[dcherian]

Do we need to update whats-new.rst for this

👍🏾 It'd be nice to point to "known regressions" to prevent a bunch of duplicate bug reports.

[kmuehlbauer]

I'm wondering if it would be possible to publish every commit to main on https://test.pypi.org/project/xarray/?

[keewis]

As far as I remember the only reason we don't do that already is that setuptools_scm creates nightly version strings that are not compliant to PEP440. The version scheme appears to be configurable, but I couldn't find documentation on how to do that. As such, if anyone can figure out how to create PEP440 (and thus PyPI) compatible development versions I think we can have a CI publish every commit on main to TestPyPI.

[keewis]

I just published the pre-release, it can be installed from PyPI:

mamba create -n <name> python=3.10 xarray
python -m pip install --pre --upgrade --no-deps xarray

(the --no-deps is important, otherwise we pull the pre-releases of other packages)

[TomNicholas]

Thanks @keewis !

I just tried making a PR in datatree depend on this release, but the CI environment installation fails. I'm not sure if that's my fault or a problem with the release 😕 Any ideas? (@malmans2 FYI)

[TomNicholas]

Nevermind, the answer was

it''s my understanding that the latest pre-release isn't published on conda-forge but on PyPI. Should it be put under the pip section

...
pip:
  - xarray>=2022.6.0rc0 

[dcherian]

I added pre-release install instructions to whats-new and scheduled a tweet for 45 minutes from now (giving RTD enough time to build)

[dcherian]

I'll issue the release once #6798 goes in.