Add CITATION.cff

[joshmoore]

Created using https://citation-file-format.github.io/cff-initializer-javascript/#/

Open questions/issues:

• list all spec authors here as well as in the specs?
• list all versions here as well as in the specs?
• use the paper as a preferred-citation?
• apparently the tags in this repo need updating
• as does the license file

[github-actions]

Automated Review URLs

• Readthedocs
• render latest/index.bs
• diff latest modified

[joshmoore]

FYI from https://citation-file-format.github.io/

[will-moore]

This repo isn't really software. Are you using this repo if you create OME-Zarr with your own code? Should you cite this repo if you use ome-zarr-py?

[joshmoore]

This repo isn't really software.

Updated to "Specification" (but "type" must remain "Software" since the only other option is "Dataset").

Should you cite this repo if you use ome-zarr-py?

I wouldn't say it's a must.

[will-moore]

typo Nex-
Should the title include OME?

[joshmoore]

Typo fix pushed. Re: "OME" good question. It's hard to do the expansion, then. Maybe let's see how it looks when we start using it.

[joshmoore]

Re-opened to trigger RTD build.

[jo-mueller]

Coming back to this after reading issue #315 . Maybe a small statement in the Contributions guide would be nice for PR authors to add themselves there if they haven't yet?

[jo-mueller]

Ok, suggestion for attribution:

• Authorship: Since we'll break out the spec document into the ngff-spec repo authorship could simply go into the project metadata of the built docs.
• Citation: We currently point to the Zenodo view on the spec text as preferred citation. I think this works in general. At least it reproduces the current recommendation as per ngff.openmicroscopy.org
• License: The MyST metadata standard allows a few common licensing options to be embedded into the project's metadata. Maybe that would help? I think right now, the license contains references to a few different components of this repo. Maybe this could be simpler for only the spec text?
• Tags: A tag in the ngff-repo would then be authoritative and do the following things:
  • Trigger a Python script that builds a citation.cff from the metadata of the spec documents
  • Trigger the upload to Zenodo with the correct author information

Open question

• Should all authors be displayed in the document's frontmatter?
• How should they be listed? Alphabetically based on the last name?

Next steps

@joshmoore was so kind to print me a list of all contributors to the spec. I would ping all of them here and ask them to add their name/affiliation/orcid/github to this branch over at ngff-spec under ./ngff_spec/myst.yml

@dstansby @joshmoore @will-moore LMKWYT :)

[will-moore]

Where do we have citation instructions? Ah - I see https://ngff-spec.readthedocs.io/en/latest/citing.html but nothing from https://ngff.openmicroscopy.org/.
I guess the zenodo link will need updating as part of what's triggered by a new Tag?

Don't know about "frontmatter" or where/how it's displayed?
Listing Alphabetically based on the last name seems best, apart from maybe Josh (as Editor)

[jo-mueller]

but nothing from https://ngff.openmicroscopy.org/.

Yes. I could see this page becoming more of a hub for all sorts of resources beyond just the specification. I think it makes sense to move citation information along with the spec if someone wanted to cite the spec itself.

I guess the zenodo link will need updating as part of what's triggered by a new Tag?

Or maybe it could be removed entirely? For once, readthedocs shows you the version of the document you are looking at anyway. Secondly, the link to Zenodo could point to the latest version and simply say: "All past versions of the spec are available on Zenodo (link)"

It's probably possible to find out the DOI on Zenodo on build and insert it in the markdown documents but maybe that's overkilling it ^^"

Don't know about "frontmatter" or where/how it's displayed?

Essentially, every markdown document in the page gets a header-like part (the frontmatter), the content of which can be defined from the project metadata (the myst.yml) or overwritten in every document by inserting this at the top:

---
title: Whatever
foo: bar
---

If nothing is displayed, then the result would look something like this example.

Listing Alphabetically based on the last name seems best, apart from maybe Josh (as Editor)

👍

[jo-mueller]

Hi @joshmoore @will-moore I figured that it's possible to simply add people as authors in the jupyter book metadata and then autogenerate a CITATION.cff from this, see ome/ngff-spec#25. If you think that's a suitable way to do it, I'd merge that branch over there into the ngff repo and ping all contributors to add themselves as authors.

[jo-mueller]

PR to work on: ome/ngff-spec#31

May supersede this PR.