Timetable for the release of CF-1.13

[davidhassell]

Hello,

At the CF workshop in September, it was agreed that the release of CF-1.13 will be sometime during the period Tuesday 8th - Friday 12th December 2025.

For a change to make it into this release it must therefore have completed its three week cooling-off by this time i.e. the proposal must be agreed to be complete and accepted by 23:59Z Monday 17th November 2025, and for there to be no new substantive objections in the following three weeks (editorial changes such as fixing spelling mistakes are still ok, though).

Many thanks,
David

[davidhassell]

Hello,

The deadline for starting the 3-week cooling-off period for getting features into CF-1.13 just been has been reached.

However ... there are a small number of very long-running issues which stand a very good chance of making it into CF-1.13, if we were to extend the deadline by 1 week: i.e. changes accepted by 23:59Z Monday 24th November 2025 and publication happening in the period Tuesday 15th - Friday 19th December 2025.

If there are any objections to this extension, please do raise them here very soon and we'll discuss how to proceed.

Note, that there will be no possibility of a yet further extension - i.e. CF-1.13 will definitely appear in 2025!

Many thanks,
David

[sadielbartholomew]

I'm happy with this, if it allows those long-standing issues to possibly make it into CF-1.13, given as you say the deadline won't be pushed back any further! 👍🏻

[cofinoa]

Hi all,

A quick update from the Zenodo side. The draft record for CF-1.13 is now created and the DOI for the upcoming release is already reserved:

DOI: 10.5281/zenodo.17801666
Draft record: https://zenodo.org/uploads/17801666
(accessible only to members of the CF Zenodo Community for review)

For reference, the citation for the new version will be:

Eaton, B., Gregory, J., Drach, B., Taylor, K., Hankin, S., et al. (2025). NetCDF Climate and Forecast (CF) Metadata Conventions (1.13). CF Community. https://doi.org/10.5281/zenodo.17801666

This should allow us to include the final DOI in the CF-1.13 documents as soon as the text is settled. Community members can already review the metadata in the draft record. I am still checking the safest way to manage draft access, but everything is in place for the release schedule.

Best,
Antonio

[JonathanGregory]

Dear Antonio @cofinoa

Thanks for creating the new record. I guess the list of authors is obtained from zenodo.json. Unfortunately this is not up to date with authors.adoc, as noted in conventions issue 579, where @DocOtak has suggested a possible explanation.

Best wishes

Jonathan

[ethanrd]

A quick tangential question ... If I want to site the CF Conventions in a non-version specific way, would I just drop the version number and use the overarching DOI (the one that always de-references to the current version)? Like this:

Eaton, B., Gregory, J., Drach, B., Taylor, K., Hankin, S., et al. (2025). NetCDF Climate and Forecast (CF) Metadata Conventions. CF Community. https://doi.org/10.5281/zenodo.14274886.

Thanks

[cofinoa]

Hi @JonathanGregory,

Yes, this behaviour is consistent with the current setup.

Just to clarify the model introduced in cf-convention/cf-conventions#579 (cf-convention/cf-conventions#580) :
authors.adoc is the single authoritative source for author metadata. Files such as zenodo.json, CITATION.cff, and about-authors.adoc are derived artefacts and are not meant to be edited manually.

The regeneration logic is already in place. There are two paths:

• Git hooks, which check and regenerate derived files, but are opt-in and only run if a contributor has installed them locally.
• The Makefile, which I would consider the preferred workflow. When the Makefile is used, the derived author artefacts are regenerated automatically from authors.adoc.

What happened here is that the document was updated, but the Makefile was not used and the hooks were not active, so the derived files were not refreshed. This is therefore an operational issue rather than a problem with the design itself.

As noted by @DocOtak, a CI-level check could help catch this situation automatically in the future. However, we probably need to further refine the editorial publication workflow before fully closing this part of the process.

Best,
Antonio

[cofinoa]

Hi @ethanrd,

Yes, if you want to cite the CF Conventions in a non-version-specific way, you should use the concept DOI (the overarching Zenodo DOI that always resolves to the latest published version), and omit the version number, as in your example.

Version-specific citations should use the version DOI (for example CF-1.13), while the concept DOI is appropriate when referring to CF in general.

Your citation example looks fine, and it can also be used at the DRAFT document stage as a placeholder, before a version-specific DOI exists.

Best,
Antonio

[ethanrd]

Thanks Antonio!

[cofinoa]

Hi all,

Just to close the loop on this discussion: the CF-1.13 final Zenodo record is now live: https://doi.org/10.5281/zenodo.17801666

At the same time, this release cycle has highlighted a few aspects of the editorial and publication workflow that could be improved or clarified going forward. Some of these are already captured in ongoing discussions and issues, for example:

• Discussion: CF Conventions editorial release workflow, including the Release Candidate stage:
  https://github.com/orgs/cf-convention/discussions/449

• Issue: Documentation build toolchain improvements and release metadata plumbing cf-conventions#632

  • Associated PR: Build toolchain and tests cf-conventions#622

• Issue: Align and standardise metadata in final CF documentation artifacts (PDF and HTML) cf-conventions#631

I think it would be good to treat these as process improvements rather than release blockers, and to revisit them calmly ahead of the next cycle.