Add single DOI field to CITATION.cff

[patrick-austin]

Changes:

• APA: Sturniolo, S., Liborio, L., Chadwick, E., Murgatroyd, L., Laverack, A., Mudaraddi, A., & Muon Spectroscopy Computational Project. (2022). pymuon-suite (Version v0.2.3) [Computer software]. https://github.com/muon-spectroscopy-computational-project/pymuon-suite
• BibTeX:

@software{Sturniolo_pymuon-suite_2022,
author = {Sturniolo, Simone and Liborio, Leandro and Chadwick, Eli and Murgatroyd, Laura and Laverack, Adam and Mudaraddi, Anish and {Muon Spectroscopy Computational Project}},
license = {GPL-3.0},
month = {8},
title = {{pymuon-suite}},
url = {https://github.com/muon-spectroscopy-computational-project/pymuon-suite},
version = {v0.2.3},
year = {2022}
}

To:

• APA: Sturniolo, S., Liborio, L., Chadwick, E., Murgatroyd, L., Laverack, A., Mudaraddi, A., & Muon Spectroscopy Computational Project. (2022). pymuon-suite (Version v0.2.3) [Computer software]. https://doi.org/10.5281/zenodo.7025643
• BibTeX:

@software{Sturniolo_pymuon-suite_2022,
author = {Sturniolo, Simone and Liborio, Leandro and Chadwick, Eli and Murgatroyd, Laura and Laverack, Adam and Mudaraddi, Anish and {Muon Spectroscopy Computational Project}},
doi = {10.5281/zenodo.7025643},
license = {GPL-3.0},
month = {8},
title = {{pymuon-suite}},
url = {https://github.com/muon-spectroscopy-computational-project/pymuon-suite},
version = {v0.2.3},
year = {2022}
}

This is using the "concept" DOI (should resolve to the most recent version of the software). It's also possible to use the specific versioned DOI instead. Undecided on which would be better to use?

[joelvdavies]

From what I can gather when citing software it should be specific to the version used, so I would guess its best if the doi is specific as well as then it links directly to the source repository in zenodo for that version.

On another note it seems the identifiers section is where the DOI should be, but I cant currently see why it wont render them. citation-file-format/citation-file-format#318

Update Despite what it says there, their repo also uses doi and github docs also have both so I think using doi directly is fine

[patrick-austin]

From what I can gather when citing software it should be specific to the version used, so I would guess its best if the doi is specific as well as then it links directly to the source repository in zenodo for that version.

In cases like Galaxy (the wrapper is explicitly tied to a specific version) I agree. There would be cases where I think the concept would be more valuable (e.g. on the main website, since that's not tied to a specifc version and it means you don't have to update it). But a citation is a lot more relevant to the former use case than the latter.

On another note it seems the identifiers section is where the DOI should be, but I cant currently see why it wont render them. citation-file-format/citation-file-format#318

Turns out they've just done the same thing (duplicating one of their DOIs from identifiers) as the PR for that issue:
citation-file-format/citation-file-format@f5b4716
Specifically it's the versioned DOI, so lets go with that? There's discussion of the fact ruby can handle it, and there's an open issue to do it for GH, but right now I think this is the best we can do.

[joelvdavies]

From what I can gather when citing software it should be specific to the version used, so I would guess its best if the doi is specific as well as then it links directly to the source repository in zenodo for that version.

In cases like Galaxy (the wrapper is explicitly tied to a specific version) I agree. There would be cases where I think the concept would be more valuable (e.g. on the main website, since that's not tied to a specifc version and it means you don't have to update it). But a citation is a lot more relevant to the former use case than the latter.

I see, yes, I think since the former is more likely the versioned one is best since anyone taking the citation from github will automatically get the current one and on zenodo you will still be able to see the other versions available anyway.

On another note it seems the identifiers section is where the DOI should be, but I cant currently see why it wont render them. citation-file-format/citation-file-format#318

Turns out they've just done the same thing (duplicating one of their DOIs from identifiers) as the PR for that issue: citation-file-format/citation-file-format@f5b4716 Specifically it's the versioned DOI, so lets go with that? There's discussion of the fact ruby can handle it, and there's an open issue to do it for GH, but right now I think this is the best we can do.

Yes I just saw that as well 😄