How to properly use zenodo DOIs for derived software?

+3 votes
323 views
asked Sep 8, 2016 in Open Science by Markus Ankenbrand (65 points)
edited Sep 8, 2016 by Markus Ankenbrand
When connecting GitHub to Zenodo to get DOIs for software, the list of authors is automatically determined by the contributors to the repository. This is a great thing. However it gets tricky when working on a fork of another project. The contributors to the original project are still listed as authors. Which I think is the right thing to do as they deserve the credit. However they might not approve of the fork or they might not even know. Also the one who forked the project might have never met the original creators.

However citing the project with the Zenodo doi would lead to a author list which contains original authors and contributors to the fork intermixed. Implying that they worked together on the project (and all approve of it and know about it).

Asking the original authors for their permission is the obvious thing to do. However what can be done if they do not answer or even refuse to be listed? As the list of authors is derived from GitHub by Zenodo there is no way to remove contributors.

What can be done to properly use Zenodo with a forked software project?
commented Sep 8, 2016 by Markus Ankenbrand (65 points)
For a real world example: I (Markus Ankenbrand) recently forked the Phinch project (phinch.org) by Holly Bik et. al. and created the derived project Blackbird (https://github.com/molbiodiv/Blackbird). Here is the Zenodo page: https://zenodo.org/record/61721
It creates the following citations:
"Shujian Bu, Markus Ankenbrand, Nick Yahnke, and Holly Bik. Blackbird: Version 1.2.0. doi:10.5281/zenodo.61721, September 2016"
However I have never met any of the others, nor do I know if they approve of Blackbird.
So I don't feel like I can ask people to cite using the Zenodo doi.

1 Answer

+3 votes
answered Sep 20, 2016 by Markus Ankenbrand (65 points)

In case that the forked software project adds enough novelty to be considered a research product on its own, it should be alright to just list all the contributors to the fork and not those to the original project.

Citing the "Software citation principles" article (Smith AM, Katz DS, Niemeyer KE, FORCE11 Software Citation Working Group. (2016) Software citation principles. PeerJ Computer Science 2:e86 https://doi.org/10.7717/peerj-cs.86), section about derived software:

In the case of one code that is derived from another code, citing the derived software may appear to not credit those responsible for the original software, nor recognize its role in the work that used the derived software. However, this is really analogous to how any research builds on other research, where each research product just cites those products that it directly builds on, not those that it indirectly builds on.

Transferring this principle to the question at hand the way to go is removing the contributors to the original project from the list of contributors for the fork. The new version of Zenodo released last week (Sep 12, 2016, https://www.openaire.eu/zenodo-relaunch) features improved GitHub integration. So it is now possible to adjust the list of contributors.

Any other opinions are welcome.

Welcome to Open Science Q&A, where you can ask questions and receive answers from other members of the community.

If you participated in the Open Science beta at StackExchange, please reclaim your user account now – it's already here!

e-mail the webmaster

...