Joss: Need to draft guidelines for authorship

[arfon]

Things to discuss/support:

• Who should be listed as an author (do all committers to a repository count?)
• How to handle authorship between releases of a software package?
  
  
  
  • Suggest publish major versions only (point authors to semantic versioning)
  
  
  • Suggest that authors between major versions should be only those authors that have contributed to to the delta
  
  
• What happens after acceptance?

[kyleniemeyer]

Recording a comment from our conversation (with @danielskatz) from earlier: should there be a way to allow discussion of implementation details or underlying theory (e.g., equations)? Or should that be mandated in the software documentation?

Some suggestions/options:

• Leave implementation/theory details in documentation.
• Keep "main" section of paper empty, but allow appendices.
• Allow "methodology" or "implementation details" (e.g.) section in paper

[kyleniemeyer]

(meta: hopefully this was the appropriate issue for that comment)

[danielskatz]

I think this is more who we should recommend be listed as an author. And I don't know the answer, other than that it is the submitter's decision in the end. I would say anyone who makes a significant contribution should be an author, but the submitter still needs to decide what significant means.

[kyleniemeyer]

Ah... yeah, totally wrong issue. I'll find a better place for it.

[labarba]

RE. who should be listed as author ... I agree that this is a decision for the submitters. However, we may want to add guidelines on at least the requirement that any submission should be approved by all co-authors, and all contributors.

[jkahn]

Addressing the questions:

authorship between multiple releases

and also

what happens after acceptance

This raises the same questions I did here:

Minting a new Zenodo snapshot & DOI before reviewing seems like wasted effort, since (I would hope) nearly all reviewers are going to request changes (even if minor) that will require a new Zenodo snapshot.

Can reviewers bless a particular tag or revision in the target repo, and have authors or JOSS take the zenodo snapshot/doi after the review process is complete?

I get that this might not be convenient for non-github users, but it seems like the publication of a new release might warrant a version number upgrade as well, etc etc.

[arfon]

Minting a new Zenodo snapshot & DOI before reviewing seems like wasted effort, since (I would hope) nearly all reviewers are going to request changes (even if minor) that will require a new Zenodo snapshot.

@jkahn - this is a good point. My original motivation for requesting the archive was that if the author has a very quick review (i.e. no modifications) then there's nothing left to do for the author post-review for the publication process to be completed.

That said, it does seem like a significant fraction of submitted software thus far has required a new DOI to be issued at the end.

I _thought_ that Zenodo supported the concept of linking to the 'latest' DOI for a particular archive but I don't seem to be able to craft a URL to return this for any of the stuff I have archived in Zenodo. @lnielsen is this possible? That or versioned DOIs 😉

[labarba]

I agree that minting the DOI before review is wasted effort — I was thinking the same thing, seeing the issue threads develop.

And perhaps it's nice for the author to have an "accepted without changes, now go get your DOI, congrats" moment.

[jkahn]

@labarba -- that's about as close as any academic gets to a victory lap.

[cMadan]

I've been working on submitting a paper to JOSS and have also been thinking it makes more sense to get the DOI after the reviewers sign off (and as part of a 'all done, congrats!')

[arfon]

My only concern with getting authors to archive/issue an DOI after acceptance is that it makes some of the automated document preparation harder (Crossref metadata, compiling the PDF etc) but I can think of some work arounds for this.

[arfon]

This is WIP is #137