30

I have used github code for my research paper. But that code comes with no license. I have asked author to add license to code but he has not replied so far. Will this be a problem for me in future? I have not intentions for owning that code. I have just used that code without any changing and i would like to cite that repository link in my paper giving credit to original developer. Please give your valuable suggestions

2
  • 19
    From an academic standpoint, I think citing the github repository and/or any related papers in your paper is actually quite sufficient. You’re not taking credit for the software - you’re acknowledging the creator, citing a link to the original source, and giving credit where credit is due.
    – J. Tylka
    Jan 5 '20 at 15:54
  • 1
    @J.Tylka your comment is encouraging but all other links and answers have scary legal language. I have also contributed (a little) in the code to remove some bugs. But still the developer have successfully ignored my request to add license in the repo. Luckily, I have found another repo under MIT License by original creator of algorithm. Let's see what happens Jan 6 '20 at 13:48
35

This topic is treated in this answer on Open Source Stack Exchange. I reproduce it here because it might also be relevant for Academia. The Open Source.SE answer points to GitHub's "No License" page:

If you find software that doesn’t have a license, that generally means you have no permission from the creators of the software to use, modify, or share the software. Although a code host such as GitHub may allow you to view and fork the code, this does not imply that you are permitted to use, modify, or share the software for any purpose.

Your options:

  • Ask the maintainers nicely to add a license. Unless the software includes strong indications to the contrary, lack of a license is probably an oversight. If the software is hosted on a site like GitHub, open an issue requesting a license and include a link to this site. If you’re bold and it’s fairly obvious what license is most appropriate, open a pull request to add a license – see “suggest this license” in the sidebar of the page for each license on this site (e.g., MIT).
  • Don’t use the software. Find or create an alternative that is under an open source license.
  • Negotiate a private license. Bring your lawyer.

Note in the second line that you do not even have permission to use the software. Unfortunately, this means that, in order to be safe, you should either wait for the developers to reply to you and add the license, or write your own code.

16
  • 5
    Adding a license to an unlicensed repository may be a complex task. Only the copyright holder can license the code, which means that every contributor to the repository must individually agree to license the code under a new license. If there's only one contributor, then it's easy, as you only need that person. If there's more than one contributor it may be anywhere from easy to impossible to add a license to the repository. The maintainers can add a license that all future contributions are bound by, but they don't have the right to add-/re-license existing code which they didn't author.
    – Makyen
    Jan 5 '20 at 18:23
  • 8
    Wondering - what is the legal reason for not being able to use (i.e. run) the code?
    – usul
    Jan 5 '20 at 21:07
  • 17
    @Kaz: I don't find that argument convincing, since essentially nothing in your comment changes if there is a license note. Jan 6 '20 at 10:27
  • 12
    @Makyen Perhaps an ironic comment considering we are on SE, which has changed licenses without consent from copyright holders multiple times...
    – user115829
    Jan 6 '20 at 14:07
  • 6
    @usul: Good question, but too broad by far for a comment. Code is widely protected by copyright, but running code is treated differently in different jurisdictions. In some, there's an implied license to use if legally obtained (e.g. from Github), which would be sufficient. Additionally, some jurisdictions have the notion of fair use, which then usually includes academic research, but again this is not universal.
    – MSalters
    Jan 6 '20 at 14:40
11

There's two different things to consider:

  • Academic honesty - for this a citation will be fine. The emphasis here is on not claiming credit for the ideas of others, and by citing the repository appropriately you've done this. There's a slight concern here that if the code turns out not to be licensed in a way that's possible for people to use then it's harder for people to reproduce your results, but that's no different to making use of a proprietary tool (which is considered acceptable).
  • Copyright/licencing/"are you legally allowed to use the code". I suspect the answer is "not really, but practically you will most likely get away with it". I also think this isn't really on-topic for this site. If it turns out you can't the consequences will likely be having to pay some compensation rather than an academic problem.

Probably the best solution would be to reimplement the algorithm in the repository yourself and cite the original source, thereby giving yourself the best chance of being legally fine while fulfilling your commitment to honestly acknowledging your sources. Note that this isn't the same as mechanically retyping their code - you should be aiming for a more high-level reimplementation.


Small addendum: it might be worth considering the source of the code. My experience is that a small-scale programs written for academic work (e.g. a data analysis algorithm from a mainly experimental group) are published without a license and without anyone ever really thinking about a license. The authors are generally very happy if the code is used because it shows someone's interested and gets a citation, but aren't hugely interested in formal licensing (which might involve painful interactions with university IP departments or similar). If the code comes from this kind of source then they'd probably be happy for you to use it (with a citation). This obviously doesn't provide legal clarity, but might give some re-assurance.

3
  • 1
    I disagree with respect to the academic honesty opinion. If the user doesn't have license to the code, this could be viewed as piracy, and there are policies that define piracy as an academic ethics issue: bvu.edu/policies/academic-honesty for example. Jan 6 '20 at 19:44
  • Granted, that is unlikely, but it is possible. Jan 6 '20 at 19:44
  • 1
    @ScottSeidman I don't completely disagree - more generally I many organisations probably view "obey local law" as a professional standard. However, my feeling is that most academic employers would be fairly non-judgemental given the circumstances described (assuming there isn't anything dodgy OP hasn't disclosed, of course). But I could see that if someone sued their institution over it then it could have consequences for them.
    – DavidW
    Jan 6 '20 at 19:57

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.