Open access papers are usually published under a CC-BY license. Does this license also apply to the supporting information? Especially if there is code that has an extra GPL or CC-BY-SA license attached?
I am not a laywer. If the attached code includes an explicit license, I would not just assume that the license the paper was published under applies to the code as well. Especially if the GPL is involved.
It really depends on exact terms of the particular journal. For example, here’s what JOSS says:
Copyright of JOSS papers is retained by submitting authors and accepted papers are subject to a Creative Commons Attribution 4.0 International License.
Any code snippets included in JOSS papers are subject to the MIT license regardless of the license of the submitted software package under review.
Note that this does not affect the license of the linked software package.
There may be other open access journals that redistribute the supplementary materials directly and offer their own terms for those (at the express agreement of the copyright owner as a condition of publication).
Someone(s) hold(s) the copyright and they’re the ones who can grant you the ability to copy and reproduce it (or not!). If the original authors maintained their copyright, then you can always just ask them directly if the terms from the journal are unclear.
It is specifically this [1] article. I have translated the code to Julia [2] and originally wanted to put the MIT License on it but found later that there was a GPLv3/CC-BY-SA license on the code in the supporting information.
Under the Article there is this:
Terms & Conditions
Most electronic Supporting Information files are available without a subscription to ACS Web Editions. Such files may be downloaded by article for research use (if there is a public use license linked to the relevant article, that license may permit other uses). Permission may be obtained from ACS for other uses through requests via the RightsLink permission system: http://pubs.acs.org/page/copyright/permissions.html.
I read it like the CC-BY license should also apply to the supporting information.
[1] https://pubs.acs.org/doi/10.1021/acsmeasuresciau.1c00054
[2] SRDFilter.jl/src/SRDFilter.jl at master · gdkrmr/SRDFilter.jl · GitHub
CC licenses cannot, or at least should not, be applied to code, see Frequently Asked Questions - Creative Commons, “Can I apply a Creative Commons license to software?”
License on the paper and the supplementary material covers the published text and form of the paper. For a license on any code that might be related to a paper, you would have to contact the authors. In an academic context, there might also be fair use considerations that could allow you to use published code. For that, you should get advice from a legal professional; your institution may have an office to deal with questions like this.
In this particular case, the code in the zip file linked from the paper clearly indicates that the code is licensed under GPLv3. The legally safest thing to do is to release your translation under GPLv3. You can also contact the authors and get permission from them to translate the code and release the result under the MIT license. That’s what I would do (since I dislike the GPL). Lastly, you could also ignore the license. When exactly a translation is or is not a derivative work isn’t fully settled, as far as I understand. The more you deviate from the original in structure, the more likely it is simply a new implementation of “math”. I also doubt that the authors of that paper have any interest in suing. These or other “fair use” questions depend on your specific jurisdiction. So make your own common-sense judgement on what kind of legal risk you are taking on, or consult a lawyer. (In case it needs to be said: IANAL, so this is not legal advice)