3

In my university, professors in Computer Science decide by themselves the licenses for the code of their research group. This happens without consulting the university legal team (or the university, more generally). As an undergrad researcher, this strikes me as incredibly odd.

I have the following questions:

  1. Is this common practice?
  2. Is this good practice?
  3. How should a situation like this be handled?
    1. Is it reasonable to expect the software only to be published only if the contributors that did most of the work agree on it?
    2. The software may or may not have economic value. Is it reasonable to expect the professor to consult if he is allowed to distribute it under open-source licenses?

For more specific context, I have checked the percentages of ownership on patents. It turns out research Groups get <10% of the ownership. My university doesn't have any special rules for software.

I have no interest in my possible ownership of this software. However, it worries me that my group director is able to make this decision on his own.

5
  • 3
    You clearly have never dealt with lawyers trying to get a simple straightforward answer! And, often, the determining factor is the grant the work is performed under, or other non-university controlled aspects.
    – Jon Custer
    Mar 24 at 21:45
  • I have updated the question to clarify point 3. It is not clear to me the university is not involved. From my point of view, the software has a shared ownership between university, contributors and research group. This software may have economic value. Is it weird that I find unreasonable that the group director is able to open source this software without consulting the university directly?
    – Oca
    Mar 24 at 22:48
  • 1
    First, the university may not care and so doesn't want to get involved in such faculty research decisions. Second, to publish something, one needs the consent of all of its "authors". Other people may contribute and not rise to the level of "authorship". It is not outside the realm of possibility that the PI has consulted with the others who need to assent. Some paid/contract work, for example, doesn't result in authorship. It depends on who has contributed the "intellectual content", not those who spent a lot of hours working at the direction of the "creator".
    – Buffy
    Mar 24 at 23:11
  • I know for a fact the director has not consulted with any author the decision (myself included). His contribution is close to 1/4 of the total work involved.
    – Oca
    Mar 24 at 23:17
  • I feel like there's too many questions for one post here, especially the consulting one. Mar 25 at 5:01
5

The software developed by most academics has no commercial value. It is a research product, similar to a paper, and is usually shared with other researchers by hosting it on the researcher’s web page (as I do) or on GitHub or other repositories. There is usually no need for any sort of license, but some people will include a license to make clear what use by others of their code they see as acceptable.

There is also no need for the university to be involved in any of these decisions, just like the university doesn’t tell researchers what “license” to use for the papers they publish. I think your main source of confusion is related to thinking of research code as some kind of valuable property that needs to be protected. In 99% of cases that’s not the reality.

1

Patents and software licenses are completely different. Relatively little software is actually patented, and I'd suspect the university wants to be involved in any patent applications. University involvement in patents is also a way for a researcher to avoid the high cost of patenting anything (lawyers and such).

Patents need to be registered. Copyrights don't (US law, but it can vary elsewhere). Publishing a "creative work" implies a copyright for the creator.

A software license is usually about permitting some limited use of copyrighted materials, though patents can also be licensed. Some, but not all, universities want a say in software licensing. More want a say in patents.

Distinguish carefully between copyright, patent, and trade-secret (which doesn't apply here). All apply in different ways to IP.

In general (US, anyway) creative works get an automatic copyright. Only "inventions" are subject to possible patent. There is some controversy about whether a piece of software is an "invention", but in recent years it has often been considered so. This has led to some strange outcomes.

Currently (US again) copyright lasts a long time. Disney would like it to be infinite. Patents are more limited. This pushes some to prefer copyright when there is a choice.


A college or university that isn't heavily involved in research (teaching institutions) may have much less restrictive arrangements generally, since such questions arise less frequently - especially for patents.


The Public Domain article on Wikipedia has a general discussion of various aspects of IP and how long protections last. There are other, more specific, articles there, also.

2
  • I didn't intend my question to ask about patents. I meant that specific comment to draw an analogy between open sourcing code and releasing patents. If this software were to be a patent, the research group would have no say in the matter because it is a 10% shareholder. How come the director of the research group can open-source software with tens of contributors unilaterally?
    – Oca
    Mar 24 at 23:03
  • 1
    @Oca But if you draw an analogy with research papers (which are also subject to copyright) instead of patents, you get a very different intuition. Researchers regularly license or transfer their copyright to a third party as part of their work. The question of co-authors' consent is a separate question, but it is a question that also applies to research publishing in general.
    – jnanin
    Mar 25 at 9:01
0

Lawyers cost a lot of money. The university lawyers are probably busy trying to keep the university from being sued for any of several reasons and don't have time to go sort out legal review for a million software licenses. Outside counsel will eat your entire group budget as a retainer.

The cost of lawyers has to be balanced against the fact that most academic software is functionally valueless. Spending a lot of money and time to ensure you have the correct rights over something that is worth nothing is a good way to be an unsuccessful academic.

Finally, if you do have something of value and you spin it out to a startup, the university can get it's taste by taking a low-double-digit dilutive share or a couple percent non-dilutive (with the goal of cashing out around series A). So the school doesn't even have to pay a lot of attention to what you're doing until you try to turn it into a company.

Related to your edit- you might be able to refuse to allow the parts you've written get released by claiming copyright over your specific work, but that will probably destroy your current relationship with your group. They are also free to clean-room reimplement whatever you wrote and cut you out entirely. If you think it's valuable you'd be better off trying to go into business with the other people who control parts of the product, fighting them is a very bad idea. The university is not interested in getting involved until there's money on the table and will side with the professor they've invested in almost 100% of the time.

Also, if you've committed your code to a project with an existing license (like MIT) you've already released it under a permissive license and you can't revoke that.

2
  • I feel this answer touches on a lot of good points. I would like to point out that the cost of this software is easily in the tens of thousands of dollars. What strikes me as odd is the ability for the professor in question to render the asset virtually useless for the university. It is certainly something I will consider moving into grad school. And no, the software does not currently have a license of any kind (yet).
    – Oca
    Mar 24 at 23:50
  • @Oca It's worth remembering that a standard NSF IIS grant is a lot more annually then tens of thousands of dollars. That amount of money isn't enough to even really cover the costs of commercialization.
    – user133933
    Mar 25 at 0:57

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.