Slashdot Mirror
Universities, the GPL and Patents?
nonlnear asks: "I'm about to finish a PhD in Mathematics and am starting to realize that I am not a big fan of my university's policy about inventions, patents, software, and the like. The gist of it is: you invent while working here => we own everything => we will patent everything. I am planning on a career in academia, but am very conflicted about this way of doing things. What Universities out there will allow me to publish (otherwise patentable) software under the GPL?"
Just because something is patented doesn't mean it can't be released under the GPL.
The GPL is a licence for granting specific permissions for people to use what *you* (still) own.
In particular, the GPL (as opposed to the LGPL) only allows non-commercial software to use your copyrighted/patented work, so you can still make a good business licensing it to commercial users.
That said, the current version of the GPL is great for licensing copyright, but somewhat murky on more complex patent issues in a global market, and that is what is being improved in GPL version 3.
The easiest way to force something under the GPL or other copyleft licenses is to make a derivative work from GPLed code. So consider using the GNU Scientific Library or something similar as your base. Your University will most likely not make you rewrite it and, if they have a legal department, will most likely not ask you to violate the GPL.
For a good piece on GPL in academia, see Releasing Free Software if you work at a University by Richard Stallman.
I'm pretty sure the University's position is a popular one. Even Google does it. That's why they give their employees that 20% personal project time.
Good question. Jennifer Washburn might know (about US institutions, at least).
I'm in the process of writing a report for the University of Toronto and recommendations for releasing research as open source. I was told not to even concentrate on convincing them to do it, they're already leaning towards that, but rather to make recommendations on licences. As far as I can tell, no department at U of T says that they own research, and profs and grad students generally have a free hand to release their research as they want.
U of T is also home to the Knowledge Media Design Institute, which is a huge proponent of Open Source. This year they ran a lecture series called Open Source | Open Access which was entirely on the place of open source within the academic community. They're also offering grants to students to work on open source software!
I'm not sure how good the math program is here, as the maths frighten me. From walking around campus, I do know that we have something called the "Fields Institute for Mathematics", which seems very official and such not. Give it a look, there are worse places to be than downtown Toronto.
Give me an e-mail if you want some more info on U of T
Sleep is for the weak!
Our lab has been facing some patent/open source issues in the last year. I was pleasantly surprised to find that the University of Calgary's IP policy makes a lot of sense -- if you invent it, you own it. Ownership of the IP is to be agreed upon by the creators, with disputes resolved by the university. The U generally prefers agreements with a fair split between the primary creators... ie the supervisor should expect at most an even share. Often the students who actually create the IP get a larger share than their supervisor. The university has a system where they claim a percentage if they help you patent the invention, but you're free to release it to the public domain if you wish.
Even (or especially) if the actual implementation sucks. Just so long as it's enough to break the patent. Pick someone on the other side of the planet. If you're in Norway, pick someone in Brasil; if Japan, maybe Zimbabwe tickles your fancy.
Then when a patent is effectively unenforceable due to prior art (give it at least a month or so), you tell the university that since the code cannot be patented, you will release it under the GPL, never letting on that you discussed the concept with J Random Khazakstani or whomever, admin of the patentsuck.sf.net project. Then "merge" your project with patentsuck.sf.net once the Uni has accepted that fait accompli in order to avoid duplication of effort.
Got time? Spend some of it coding or testing
That's ridiculous and really messed up that a university would have a policy like that. The whole idea of a university is the dissemination of knowledge, not locking it up in legalities.
On further research, it looks like I was wrong. According to U of T's Copyright policy the university holds the copyright to anything "created by an Author in the course of the Author's employment by the University." On the other hand, "For the purposes of this Policy, research and instruction, or the creation of instructional Works, including Instructional Software, undertaken by members of the University's Teaching Staff or librarians shall not be deemed to be made or undertaken in the course of their employment by the University." This leaves me thoroughly confused, which is about par for the course.
However, in another twist in this dramatic story " Computer Software that is not Instructional Software will be deemed to be an "Invention" under the Inventions Policy, and the rights and obligations with respect to such Computer Software and the disposition of revenues therefrom shall be in accordance with the Inventions Policy." This Invention Policy says that the university essentially everything you make in your office.
So, we're no different than anyone else, but damn if we don't have the best student union in the greater Canadas.
Sleep is for the weak!
They kept you fed while doing this and now you're surprised they don't want something back?
We haven't gone the way of the USA. Yet...
Not always, and really, not often. Most large universities and corps will clame ownership of all your thoughts on the assumption that even if you do some work at home, you're using knowledge gained at work as well. In any case, it will be in your contract.
"Who are in control, they are not in control of anything - they don't even control themselves!" - Glen Beck
The University of Waterloo in Canada generally makes a selling point of the fact that staff, faculty, students, RAs, etc retain ownership of any copyrights or patents that stem from their research.
"What Universities out there will allow me to publish (otherwise patentable) software under the GPL?"
All of them. It's called publishing your research, something they also require. Write it up and send it to a journal. They require that you turn over copyright of the article to them. Make the code available under GPL and write the URL into the article. If the university wants to fight about it, let them fight with the journal and in the process make it clear to all their employed academics that publishing their work will get them problems. Won't happen.
You can also team up with someone from a different university and publish collaboration, and let the universities fight for the rights whether they have the same or different policies.
Also, they cannot prove you invented something while there unless you say so and document it as such for them. I have been involved in research with others, including works in progress and undeveloped ideas, for quite some time. I can honestly tell them that I can't honestly sign such a thing without an equally binding document that says my previous work and collaborations, no matter how much or little completed, don't fall under their 100%. I've successfully used this in commercial settings, and they're much more stringent than universities.
"I may be synthetic, but I'm not stupid." -- Bishop 341-B
Interesting that you bring up student unions. In Chicago, apparently it is politically incorrect to call them student unions. As a result, my University renamed our student unions from "Chicago Circle Center" and "Chicago Illini Union" to "Student Center East" and "Student Center West".
Needless to say, everyone ignores the official names.
Staying on topic, I would like to point you to DJB's page on this issue:
http://cr.yp.to/patents/tarzian.html
Apparently my University adopted a similar policy in 2003 and this prompted DJB to come up with patentable intellectual property for them: things like a "coin operated elevator" and a "soap saver dish". Beautiful.
My other car is first.
OMG! It's raining! W's Fault! W's Fault!
This is University policy, not government policy... I guess that's a little tough for you to understand.
Very right.
I'm not sure about patents, but IIT in Chicago does allow BSD licencing of software written on their paycheck. GPL/LGPL is a bit too dificult for them to get their head around so far... We DO release little stuff with the GPL|LGPL but big stuff seems to go BSD (if its going open source at all).
The situation the poster describes is, I think, atypical. In my experience (and I've been faculty for 22 years) virtually all universities concede to their faculty the copyright on what they write as part of their research or teaching. The only situation I can think of in which the university gets the copyright is when a faculty member writes something at the behest of the university. For instance, if you're one of a number of people who write a policy manual, the university will normally hold the copyright on that. But it is unheard of in my experience for the university to hold the copyright on research publications by faculty.
Patents are another story. Policies vary quite a bit. A common one is that the university has the right of first refusal. If they turn it down, its yours. If they decide they want it, you get a certain percentage of the profit.
Software, naturally, is sort of in between, being something written and typically protected by copyright, but at the same time more in the nature of a "thing" or a "product". In practice, lots of software is released under the GPL or BSD licenses. Roughly speaking, the less obvious commercial value the software has, the more closely it is tied to your research, and the smaller the group of people who work on it, the more likely you'll be able to release it freely. If you write a compiler for the cool new dysfunctional language you've designed, all by yourself, the odds are they'll never even notice, much less care. If, on the other hand, a bunch of people create something that looks like a product and looks like a money-maker, the university may take a different view of things. If they can argue that you've used a lot of university resources (other than your research time), that will make it look more like it belongs to them.
People have sometimes gotten into trouble with big projects like this. Stephen Wolfram left Cal Tech when he got into a fight with them over the ownership of SMP, the symbolic math program that he had written, essentially the predecessor to Mathematica.
My impression is that unless you get involved in things like SMP that look really attractive as products, releasing software is generally not a big problem, though some places will want to use a BSD-type license instead of the GPL. I'd be curious to see if anyone else on Slashdot knows of problems of this sort arising in practice.
No need to be confused. I'm pretty sure that what they're saying is that somebody who writes something as part of his or her job, that is not research and not teaching material, is engaged in writing a work for hire, so the university gets the copyright. In other words, they're pretty much exempting faculty. The copyrights they are talking about are the ones on the stuff written by Human Resources people or the person hired to write a history of the university or something for the alumni magazine.
The obvious answer is to check if there is a similar project already GPL'ed and then work on that rather then inventing something new.
In that case the university would have to tell you to stop working on the project which is unlikely.
If there isn't you might consider doing the first parts at home and making that part GPL and then continuing on it at work later on.
Copyright law is much much much older than W.
It is government policy. The University can only make such claims to the intellectual property because the courts have upheld their silly intellectual property clauses.
While the Constitution guarantees rights to authors and inventors, there are endless laws on the books which do nothing but protect the ways in which an organization can extort those rights from the author or inventor.
Back asswards.
fast as fast can be. you'll never catch me.
You can release the source code of a patented algoritm, but you have to release it after the patent become effective. Patents laws require taht there is no prior art, not even by yourself. So you must patente before releasing the code.
this post contain no useful information, no need to mod it down
... then change institutions. Here (in the midwest at a Research I school), the university takes a pretty good appraoch to the problem and typically works with you (the inventor) to do a 50-50 split on things that you have come out with while working on their network/machines/time/etc. This seems pretty sane to me, as everyone wins - YOU may or may not have had the time and legal muscle to get a patent in and approved, and THEY wouldn't have had the invention in the first place.
...
When going for either P&T or hiring, ASK about these issues, and if you don't like the answer, tell them and move along.
Hope this helps a bit
Ah yes, but the stated purpose of granting those rights (through the artificial legal fiction of pseudo-property) was to promote the advancement of the Arts and Sciences. Seems like an author who wants to use the GPL to insure distribution of the sourcecode (with the right to modify same) is taking a path that maximizes the promotion of the advancement of Arts and Sciences. It is in fact, putting the Science back into Computer Science.
Just because something is patented doesn't mean it can't be released under the GPL.
That's at best misleading. You cannot release something under the GPL if someone else holds the patent on it and doesn't explicitly permit that usage. And if you yourself hold the patent on it, you have to give the recipients a transferable license, otherwise the license isn't GPL even if you call it that.
In particular, the GPL (as opposed to the LGPL) only allows non-commercial software
Quite to the contrary: the GPL encourages you to develop commercial software, but it forces you to make the licenses transferable. That happens to have the effect that some business models don't work well for GPL'ed software, but that's a side effect not a goal. All things being equal, RMS likes programmers getting paid well and companies making money with software, he just doesn't like them imposing restrictions on users of the software.