ESMValCore v2.8 release

[remi-kazeroni]

Hi @ESMValGroup/esmvaltool-developmentteam, the release process for ESMValCore v2.8.0 is ongoing!

Feature freeze

Friday, March 03.

Release team

Release manager: @remi-kazeroni
Co-release manager: @valeriupredoi

Release branch

v2.8.x

Release Status

Minor Release	Status	Milestone	Date	Docs	Release Tag	PyPi	conda-forge	docker	testing
v2.8.0rc1	Released	Core v2.8	03/03/2023	Docs	v2.8.0rc1	PyPi/v2.8.0rc1	anaconda	docker	webpage
v2.8.0rc2	Released	Core v2.8	17/03/2023	Docs	v2.8.0rc2	PyPi/v2.8.0rc2	anaconda	docker	webpage
v2.8.0	Released	Core v2.8	23/03/2023	Docs	v2.8.0	PyPi/v2.8.0	anaconda	docker	webpage

Feel free to ask general questions about the release in this discussion. I will open separate issues about the recipe testing, output comparison, selection of highlights for the Changelog, ...

Thanks in advance for your collaboration with the release of v2.8!

[valeriupredoi]

great overview/bird's-eye 🦅 - I'll have a go through these myself now, see if I have anything that I#m trailing the release by 🍺

[valeriupredoi]

I bumped Issues #1665 and #1800 to M2.9 - their PRs are bumped too, there is simply not much time to address them, but I will test the auto feedstock PR in a different package see if all works fine there

[remi-kazeroni]

Update on the feature freeze for the Core:

• Things have been delayed due to extra testing and work to merge Support wildcards in the recipe and improve support for ancillary variables and dataset versioning #1609

Remaining PRs

(see https://github.com/ESMValGroup/ESMValCore/milestone/14)

• Now that we have Support wildcards in the recipe and improve support for ancillary variables and dataset versioning #1609 in, we can focus on the last remaining PRs that were depending on it:
  • Combined offline and always_search_esgf into a single option search_esgf #1935
  • Fix issue with reading dates from files #1936
  • Add institute name used on ESGF for CMIP5 CanAM4, CanCM4, and CanESM2 #1937
  • Make extract_volume work with closed and mixed intervals and allow nearest value selection #1930 (if approved on time)

New urgent issue?

• Recipe runs sometimes fail due to multiple processes attempting to cleanup the same directory #1948

New Timeline

• Next Tuesday, 13:00 (GMT): Tech Lead Team Meeting to discuss pending v2.8 issues and discuss the release process
• Next Wednesday (March 1st, around noon): Feature freeze for the Core
• Around March 8th: feature freeze for the Tool. I will communicate about that in the Tool repository next Monday and go through the v2.8 milestone

[bouweandela]

It would be nice to add some overview information to the top post, as they do it in iris: SciTools/iris#4433, SciTools/iris#4522. We don't use a project, but could add a link to the milestone instead.

[remi-kazeroni]

Thanks for the suggestion! Please let me know what you think about the new top post.

[bouweandela]

Awesome, much more clear, thanks!

[Peter9192]

I noticed after #1609 the example notebook on https://esmvaltool--1609.org.readthedocs.build/projects/ESMValCore/en/1609/notebooks/loading-and-processing-data.html still uses the old add_ancillary and dataset.ancillaries. Should this be fixed before the release?

[remi-kazeroni]

Good find! It would be great to update these notebooks. Would you or @bouweandela have the time to update these notebooks for v2.8? I think it is not critical to have that ready for the feature freeze as we don't test notebooks but it'd be great to have that for the stable release.

[bouweandela]

I updated the notebook here: #1945

[remi-kazeroni]

The round of testing with v2.8.0rc1 does not seem to lead to too many errors, see overview here. Therefore, I'm aiming at including bugfixes into a rc2 which should be the (ideally/hopefully) the final release candidate. A lot of testing has happened already during the review of several PRs, reducing the testing load for this release 🍺

[remi-kazeroni]

Plans for next week would be to address/merge the issues/PR in the milestone v2.8. This includes the Changelog for v2.8 and various issues found during the first round of testing. If things go well, we could plan a second and last release candidate towards the end of next week (e.g. March 17). This would be also the time for the feature freeze of the Tool. The Tool release branch and Core rc2 would be used for one last round of recipe testing + output comparison against last release. If all goes well, we could then think of finishing the release in about 2 weeks from now 🤞

Milestone v2.8 (to finish)

• Draft changelog for v2.8.0 #1951
• Missing preprocessed OBS data when running recipe_bock20jgr_fig_1-4 #1962
• Update the institute facet of the CSIRO-Mk3L-1-2 model #1966
• empty ancestor list in recipe_esacci_lst #1967
  - [ ] group_ensemble_members fails when ensemble is given as a list #1964 (if agreement on this)

[bouweandela]

Looks good @remi-kazeroni! Could you update the table at the top with the next planned release candidate?

[remi-kazeroni]

Apart from finalizing the changelog for v2.8.0 (WIP in #1959), everything is ready to proceed to the release of the second (and final) release candidate for the Core v.2.8.0. The milestone is now empty. I am planning to make the release on GitHub and PyPi tomorrow,

[remi-kazeroni]

v2.8.0rc2 is now available on GitHub and PyPi, see top post

[valeriupredoi]

@remi-kazeroni you an absolute legend! I am heading over to the Changelog now

[remi-kazeroni]

Since no Core issues were noticed in the second round of testing (see ESMValGroup/ESMValTool#3127), @valeriupredoi and myself have decided to do the release of ESMValCore v2.8 tomorrow already: GitHub release and then on conda-forge. This will give us a few days to adjust the environment in the Tool, conda-lock files, ...

[remi-kazeroni]

The release of ESMValCore v2.8.0 is now completed! The package is available on all our platforms, see top post. Once the release of ESMValTool v2.8.0 is completed (mid-next week), the release of both packages will be announced through our communication channels. Stay tuned!

[bouweandela]

Thanks @remi-kazeroni and @valeriupredoi and the rest of the team! Great to see how well organized and clearly communicated this release was and how well you managed to engage the wider community.