20

I recently presented a paper in a top-tier conference in a computer engineering field. I did that work as a research assistant under a Professor but now I am working in the industry. Two different researchers, let's call them John and Sam, wish to work on an extension to this work. But both need help from me since they feel I will be able to solve their problem quickly based on my experience. John is a PhD student with whom I had worked with and was the second author of the paper. Sam is a PhD student, in another university, who I met at the conference. Since both have a request for help and both are working on the identical extension of the work, who should I help out?

Background: I originally planned to work on my paper alone along with my advisor. After I was done with about 80% of the work, I met John (who is a Phd student under my advisor) and decided to collaborate with him because he too was working on a similar problem (he at that time was working on the extension itself). John didn't contribute directly with the work (I was sole author of all code and did all analysis with help my the advisor) but he helped me with conference selection and writing of the paper. As John was sponsored by a company, having his name as an author meant that now all the ownership of the work laid with the company, at least that was what I was told by John and the advisor. This meant they had all the rights to the code (though I never signed away my rights explicitly). My advisor is particular about legal issues and hence don't want me to make the code open-source. But since the company was not interested in using the code and I didn't want my work to go to waste, I have made the code open-source, unknown to my advisor or John.

Question: I wish to maximize the use of my work (since it is my first research work). So should I continue to help both John and Sam who are both working on the same topic? My concern is that one of them will have wasted their time if the other is successful in publishing a paper first.

I am more confident about Sam's skills (based on his past papers and John doesn't have much experience in this field) but John has indicated to me from the start that he was more interested in the extension than the work that I did in the paper.

My other option is to reveal to both of them about each other so that they can possibly collaborate. But since I have told Sam that my code is open source, Sam can possibly tell John this information which might irritate my former advisor. I would like to avoid this because I might require future favors such as a letter of recommendation should I choose to pursue PhD.

Edit: I have open-sourced the code via GitHub (changed the repository from private to public). I still haven't added any license (so legally by default it is not open-source as in free to use but anyone can browse through it) so I still retain all rights. Basically, I used the wrong terminology, the code is public (instead of open-source). I haven't talked with any representative from the company, only my advisor and John have been in contact with them. From what John tells me, the company is no longer interested in the project and I think John's funding is also stopping because of this.

Based on the answers, I am going to ask Sam and John to work together (at least tell both of them about each other) and ask Sam not to reveal about the public code. I am still not convinced about making the code private again because this being a computer engineering field (applied field), just based on paper, it will be hard for anyone to replicate the work (the evidence is that both John and Sam need my help).

  • 18
    "I have made the code open-source, unknown to my advisor" - this was an extremely unwise decision, which is already starting to come back and bite you. Never do this again. What will you tell your supervisor if he finds out and asks whether you plan on open-sourcing code you write for your employer, too? – Stephan Kolassa May 9 at 11:45
  • 6
    You are allowed to publish your own code. The code itself does count as literature, to which you as the author own the copyright. John or the company from whom he was sponsored need licensing rights to publish your code. Nevertheless the company can own the publishing rights to the paper or the fundamental method of your code in form of a patent. But considering the fact that you, the main author and creator of the intercultural property, never signed any working contract or licensing agreement the company does not have any rights or only have the rights to Johns contributions. – GittingGud May 9 at 11:54
  • 14
    Both comments above are rubbish. At least in the United States, the ownership of software written as part of work (including being a graduate student, whether paid or not) resides with the employer (the university) or with any other entity with which the university may have made an agreement in return for money (e.g., the company that sponsors the research). So you can put your code on a website and put an open source license on it, but that's legally not your decision to make and has no legal validity if you don't have the approval of the university. – Wolfgang Bangerth May 9 at 12:57
  • 6
    @WolfgangBangerth: Yet in many cases, the university has a general policy that the copyright in such work is automatically assigned back to the student or employee who created it. (The University of California has such a policy, for example.) So in at least some cases, it is the student's decision to make, after all. I really don't think it's possible to make a blanket statement one way or the other. – Nate Eldredge May 10 at 0:16
  • 4
    @WolfgangBangerth: Well, my reading of policy.ucop.edu/doc/2100003 doesn't seem to allow for such: "ownership of copyrights [of scholarly and aesthetic works]... shall reside with the originator" unless certain specific exceptions apply. I don't see how the university could change its mind unless they modified the policy in advance. Now, there is a separate policy for patents, which the university does get a share of if they want. – Nate Eldredge May 10 at 2:44
123

Work with both John and Sam on the same paper. You all have something to offer, so pool your efforts and work together.

  • 32
    ... and to increase the impact of your joint work. – usr1234567 May 9 at 8:38
  • This is the right suggestion IMHO, because if they both work independently on the exact same thing, not only will there be some/a lot of wasted effort, but they are also likely to get into some conflict over who gets the credit, or - one of them will publish and the other one will be stuck with nothing to show for his work. – einpoklum May 9 at 19:29
  • @usr1234567: A larger group of authors is not a guarantee of a higher-impact paper IMHO. – einpoklum May 9 at 19:30
  • 5
    +1 Also OP's name should be on both works. How do you explain to an editor that your name is on two manuscripts on the same subject, but with different coauthors? Split personality? ;-) – Karl May 9 at 19:40
  • 12
    This is so obviously the right answer it's almost embarrassing that it needs to be said... – Phill May 10 at 2:18
10

This open-source issue is really knotty. My first concern is your legal footing -- make sure you do not have any legal liability for making that code open source. You may need to consult with a lawyer.

Second, John is an author of your paper. So, this is joint work, regardless of who did what. Now that it's published, anyone can do anything with the published work -- but anything unpublished (e.g., code) could be considered IP (see previous paragraph). Even beyond the legal issues, this is joint work, so leaving John to work on his paper while you help Sam scoop him is rather ethically questionable. Even if John and Sam were both total strangers, it would be reasonable to either put them in contact with each other, or tell one that you've decided to work with the other on the same problem.

  • 1
    Consult with your university's legal department, since they own the code.. And communicate with your ex advisor about what their conclusion is. Take down the open-source code if you need to. Then you can proceed with collaborating with your two interested parties, as adults do, instead of trying to pit them against each other – George M May 9 at 17:10
  • @GeorgeM Unless (a) the OP signed a contract giving the university rights to her code, or (b) the OP was paid to write the code (would make it a "work for hire"), the OP owns the copyright on the code and can legally do with it as she pleases. [Doesn't mean some people won't be angry about it...] – nerdfever.com May 11 at 20:02
  • Consult with your university's legal department, since they may well own the code... I don't think nerdfever that you have any further secret information that would allow you to be so definite, especially since we aren't even certain of the country much less the university – George M May 13 at 18:43
  • Well, I think OP should be careful -- telling the university's legal department that they illegally published university-owned code on the internet could end poorly (though I can't say definitively that this was illegal, of course). I think that's the real issue here, sort of a shame that the top-rated answer ignored the most important point. – cag51 May 13 at 19:01
5

The "open source" issue would probably have been easily solved, except you have now made the situation worse by going behind people's backs.

I'm coming from the "industry" side of this sort of situation, and from long experience it's unlikely the company actually wants the code. They want the ideas expressed in it, to implement their own way.

John has apparently got the message the "IP is important" but he isn't the person in the company making the decisions, or signing off the sponsorship money! It would have been much better to deal with whoever is managing the sponsorship, and get a proper agreement in writing if necessary.

It's very unlikely the company would want to prevent the university doing further research, especially since it appears they are getting this code "for free" and weren't aware of it when they set up John's original sponsorship. In fact they might even sponsor more research.

But from the industry side of the table, the one thing that really p*sses me off is when people start trying to play silly games with IP - and "putting something in the public domain" behind my back seems like exactly that, a silly game. If the IP "belongs" to the sponsoring company or to the university, that's fine either way. both sides can do business starting from that point if they are both willing partners. But when one partner suddenly takes unilateral action, that sends a message that they are no longer a willing partner.

In the short term - well, from an industry point of view you don't expect every research project to pay for itself, so if this one doesn't produce any useable end product (and code with an open source license might not be useable by the company!) that's not going to surprise anyone.

But in the longer term, if there are several universities who are possible places to sponsor research, and one of them has a track record of playing IP games, guess which one I'm less likely to recommend, next time around...

2

You should all do a collaborative study. John and Sam should have shared first authorship or Sam first and John second, with you providing guidance and consultation on the entire project.

If you are specific about the open-source nature of the code, discuss this issue with both John and Sam. Do this, before going to your legal advisors.

As John was sponsored by a company, having his name as an author meant that now all the ownership of the work laid with the company, at least that was what I was told by John and the advisor.

I'm sorry but this seems very shady to me and I think John is playing you for a fool or John himself is a fool. If you are the author of the code, it is your code. If the code is present on a private repository (is it?) in the cloud, it establishes that you are the creator and contributor to the codebase. The contribution by author part in the paper establishes the contribution by both You and John.

Unless you sign some sort of agreement relinquishing your rights to the code, it does not become IP belonging to a third-party. Furthermore, you may have already signed away whatever rights you had to the code when you became a student at your university. Go back and check all the papers they made you sign back then. In such a case, the code belongs to the university.

Also, it is worthwhile to check the licenses of the dependencies you may have imported and also the licenses of the language that you used to write the code. In many cases, the licenses will stipulate (check Apache) that usage of this software means that you will have to preserve and maintain the license in your codebase.

Do all this before you speak with a legal advisor.

  • 1
    Why do it before talking with a legal advisor? Do you expect a legal advisor to advise halting further research? – Vaelus May 9 at 15:07
  • No, I do not. But in my experience, speaking to the institutional legal advisor almost always means that the project will get mired by a lot of red tape, which means progress will be slowed down. – FoldedChromatin May 10 at 8:07

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.