Most papers do not specify a separate license for the source code, so presumably it would be licensed under the same terms as the rest of the paper. Typically, this means that the copyright lies with the publisher. Then,

  • As a reader, do I need permission from the publisher to use source code contained in journal articles?
  • Would using the source code be considered fair use in the US? What about countries that do not have comparable laws?
  • As an author, what should I do if I want the code I publish to be freely usable by others?
  • related, but not a dupe academia.stackexchange.com/questions/10629/… – user7130 Dec 18 '13 at 4:34
  • are you talking about code snippets enclosed in the journal article, or are you talking about entire programs/scripts enclosed as supplemental material? – user7112 Dec 18 '13 at 6:57
  • It might seem odd, because your three questions are on the same topic, but this Q&A system works best if you just post one question at a time. (oh, and by the way, the assumption in your first sentence may well not hold in most cases) – 410 gone Dec 18 '13 at 12:59
  • A much more interesting question is: as an author, can I use the code I have published freely? As far as I understand, the way things work now with most journals, the answer is no. After the copyright transfer agreement, the code belongs to the publisher, like the rest of the paper, and there are significant restrictions to what I can do with it afterwards. – Federico Poloni Dec 18 '13 at 17:13
  • @FedericoPoloni even if the code is supplied as "additional material"? To be honest, I have never had the need to verify this. Another good reason to host the source code somewhere else, e.g., figshare, zenodo, github, ... – user7112 Dec 18 '13 at 18:32

I think the codes available in papers are provided as materials for several purposes, including

  1. Help reviewers and journal's board to efficiently evaluate the work and identify its technical merit.
  2. Enable readers to better grasp the algorithm/method proposed
  3. Enable researchers to utilize the code in their future research to avoid duplication
  4. Help (novice) researchers to play with the code for educational benefits
  5. Provide opportunity to peers to further improve the proposal

However, you may contact the author(s) if the want to use the code for

  1. Commercializing the algorithm/method
  2. Use the code for activities that generates money
  3. Register patent using available contents

In my institute, I see lots supervisor ask their master students to read recent papers and identify a small research problem to alleviate during their project work. Students either use the given code or contact the authors for code, which authors usually provide the code.

So, I think you don't need to explicitly state the code copyright in the paper. But, if you are really interested to do that, you can have a footnote in the first page allowing readers to utilize the code in full, if the EiC does not mind.

Hope it helps

  • 1
    I think your analysis of the purposes is spot-on. My question mostly concerns 3-5. If the code is included in the copyright transfer agreement with the journal, the publisher now owns the code. So contacting the authors would be useless, as they cannot legally give you any rights. – Mangara Dec 18 '13 at 19:24
  • There is something you need to consider. If the author gives copyright to the publisher, it doesn't mean himself does not have any right. No one can ever takes away the ownership from the author. The copyright the author gives to the publisher is to the version that they provide and still you have your own version (in most of the cases). In majority of journals, including IEEE and Elsevier, you can upload ur own pre-accepted file anywhere author likes, on his web page or other public repositories like Arxiv. If u r using code from the copyrighted file, cite it & this is what publisher needs. – Espanta Dec 20 '13 at 4:59
  • Those copyright agreements explicitly allow the author to post these versions. If they did not, it would most likely be illegal. – Mangara Dec 21 '13 at 11:18

Most of codes are free if you use it for research purpose, you just need to include right citations

If a code is commercial or GPL license then you need permission, for that you can contact the author directly and discuss the same

If you want your code to be freely available just make it open source licensed that's it.

Each country has different rules for patents and stuff, so be careful where the code is available etc. So discussion with author is the best thing to do. The license is usually specified on the website where you get the code from and also in the .cpp or .h files header.

  • for that you can contact the author directly My understanding is that journals typically require the author to sign a copyright transfer agreement. So the author does not actually own the code in that case, and cannot give me the right to use it. – Mangara Dec 18 '13 at 19:21
  • "Most of codes are free if you use it for research purpose" surely that will depend entirely upon how it is licensed? And how it is licensed is the point of this question. – Flyto Dec 19 '13 at 15:06
  • Thanks Flyto. I edited my response a little to include your comments. – Armin Mustafa Dec 19 '13 at 18:29
  • Mangara : Are you sure about this, According to me the code is owned by author and the paper or the idea is not, because the copyright transfer agreement is signed for the idea but not the code. – Armin Mustafa Dec 19 '13 at 18:33
  • 1
    I an not a lawyer, but I should perhaps also point out that AIUI, if the code is GPL-licensed then you specifically do not need permission (at least, not from a legal point of view. You'd still need to cite it from an academic-ethics point of view, and it would still be polite to talk to the author about it). Also, while you have stated that the license is found on the website where you get the code or in the file header, the question was asking about code included within articles. Not wishing to be harsh here - just a topic I'd like to see a good answer on :-) – Flyto Dec 20 '13 at 11:15

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.