The new international initiative, Initiative for Open Citations (I4OC), that started this week with the aim of promoting the unrestricted availability of scholarly citation data, is a wonderful idea.
However, in order for this to be fully useful for software publications (meaning the publication of software itself, not publication about software or software papers), we also need to be able to add citations inside published software, as I discussed here about a year ago: “How should we add citations inside software?” (This is how software cites other works, which is different than the question of how software itself is cited, the subject of the FORCE11 Software Citation Working Group and the “Software Citation Principles“.)
I see two potential options, which somewhat overlap:
- Software publishing venues (e.g., zenodo and figshare) could create a mechanism for submitters to provide citation information when they submit software for publication.
- Software developers could add a file (perhaps called “REFERENCES”?) in their repositories. This file would contain a list of reference for the software in the repository, and ideally would be both human and machine-readable.
If 2 was implemented, the software publishers could ingest this file, making 1 easier.
To make this happen, we would need:
- The software community to agree that there is value is having references to other products (software, papers, data, etc.) as part of their work
- The software (and publisher?) community to agree on a standard way to record these references
- Software publishers to implement this as part of the submission/publication process
I think this is 100% feasible, and technically relatively simple.
If this makes sense, how could we move forward?
Are there other, better options?