notes on breakout at RDA11 on creating guidelines to help implement Software Citation, e.g. for conferences, journal editors, authors, reviewer, communities, organizations

[danielskatz]

22 March 2018

Attendees:

Neil Chue Hong Martin Hammitzsch Morane Gruenpeter Dan Katz Tom Bakker Sibylle Hermann

[danielskatz]

For conferences, @danielskatz has started work with some (e.g. ACAT), wants to do more than see what's common that can be pulled out

For journals, @npch noticed that journals have signed up to TOP guidelines, and maybe we could do something similar, with various levels (1 = general support, 2 = with general guidance about how journals / your journal support it, 3 = the journal fully supports it in all elements)

This in general is related to software citation principles, which say at a high level what to do, but not how to do it.

This could be tied to APA and similar style guides.

[danielskatz]

How do we actual write down these guidelines?

One set for conference, one for journals? One common?

Use examples (things to do), counter-examples (things not to do)? (examples sjhould be specific to the domain, ideally) Use rules? Use both?

guidelines should be generic (across domains) but also clear how to customize for a particular domain, and easy to customize (e.g., for a conference proceedings editor, for a journal publisher)

[danielskatz]

How far does this go? Do we work with APA to help them "improve" their software guidelines?

Then we (or the conferences, journals, etc.) next have to update their tools to support the guidelines, for example, the citation management systems, the templates for authors, etc.

[danielskatz]

Maybe we don't separate conferences and journals and societies, but should think about working with different disciplines, since disciplines often have conferences/journals/societies more integrated across their community than two conferences in different disciplines have things in common.

[danielskatz]

What are the guidelines for an author for citation? Do they have a range of quality (good, better, best practices?) Or is it more pragmatic? (Do the best you can, the reviewers should say if you did well enough)

And is this tied to reproducibility? For commonly used software, the version is not really important, but for next and rarely used software, the details are more important.

[danielskatz]

moved to https://github.com/force11/force11-sciwg/blob/master/meetings/RDA11-guidelines.md