We are working on a paper that is heavy in both theory and numeric. The numerical code was exclusively developed by a subset of the authors (i.e. some of the authors have no made a single contribution to the implementation of such code).

The code will be published in Github and in Zenodo to get an associated DOI of it for citation. When doing so, one has to declare the authors of said code. Should all authors of the paper be included or just those that exclusively contributed to the code?

  • Arguments I see in favor: The code came from the developed theory and would have never been written without extensive discussions among the authors of the paper.

  • Arguments I see against: These authors do not even have a Github username and are not listed as contributors in the repository. Even if they join Github, the repository history explicitly shows who contributed to it.

  • What is the reasoning behind obtaining separate DOIs for the code and the paper? It is quite normal in many fields to simply publish a paper, and ask people to cite that if they use the code. – nabla Nov 13 '20 at 8:09
  • 2
    The DOI is to reference to the repository that contains the paper. It is useful when one wants to actually find the code (say, if write down an url address in the paper, this address might dissapear at some point). In this regard, authorship of the published code is irrelevant, but I am not sure. – pancho Nov 13 '20 at 8:33

It is an interesting situation. Clearly the 'authors of the code' have the copyright (and the repository shows who actually contributed). But the purpose of the code was to address a question broadly, as demonstrated by the broader group on the paper. For the purposes of a DOI (which does not indicate IP ownership on the code) I think having the same author list as the paper is acceptable, but not necessary.

As something else to think about, would the code have been developed outside the context of the project leading to the paper? Did the non-code-writing collaborators have no intellectual input to the code?

As an aside, if the repository URL disappears, then the DOI link will die as well.

  • Thank you for your answer. The code is a direct conclusion of the paper and there are no other works remotely close to it, therefore, the code wouldn't have developed outside that context. The non-code-writing collaborators had input in the sense that the main paper discusses technical aspects as convergence and stopping criteria, which are employed by the algorithm. Regarding URL, Zenodo is supposed to host forever the repository with its url/DOI. I am not aware of a any safer option. – pancho Nov 13 '20 at 15:59
  • @pancho - as far as I can tell, having been on networks from the early 1980's, the only things that are 'forever' are the very things you wish would disappear before your children find them... – Jon Custer Nov 13 '20 at 22:47

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.