How to convince Professor to use GPL

Question

I work in the research group and over the last year I have written library which I would like to be released under GPL terms. First I must convince my employer/professor that GPL is a good thing.

He is not opposed to releasing software and source per se. His concerns seem to be:

• plagiarism for commercial gain
• involvement with legal issues
• issues of ownership

I know GPL addresses all those issues. Unfortunately I do not have very good convincing skills.
If you were in similar position, advocating free software, how did you convince someone?

this is the company that is primary concern: http://en.wikipedia.org/wiki/GAUSSIAN#License_controversy. I do not know much about them but they seem somewhat aggressive.

Answer 1

The GPL is a copyright license; it's not the same as "public domain". The original author still owns the copyrights to the code, and the license controls how it can be incorporated into other projects. If someone violates the terms of the license (for example, by putting the code in a non-GPL project), there are conceivable legal remedies -- though i'm not aware of anyone actually suing for GPL violations yet and winning.

Answer 2

You seem to be confusing GPL and open source. GPL is one of many open source licenses, and -- at least in my opinion -- the worst of all possible open source licenses, because it is "copyleft"; anything that uses GPL code must, itself, be released under a similar license. If you want to convince your professor to release code as open source, consider promoting a more permissive license such as the New BSD License, the Simplified BSD License, the MIT License, or the Apache 2.0 License. Open source simply means that the code is freely available to read, although it is common for open source projects to make the code and program freely available to use, modify, and distribute (with caveats).

Given that you are trying to convince your professor to make the code open source, I would argue that open source is incredibly beneficial to academia; by releasing your code as open source, others can build on and extend your work, rather than having to reinvent the wheel, thereby increasing innovation.

As for plagiarism, if you have your code hosted in an online repository with suitable copyright notices, then if someone copies your code, the snippets that they used should be quite easy to find with a Google search. As for the matter of commercial gain, these licenses do not prohibit commercial gain; however, it would be a great thing if a company found an interest in the code, because most likely their needs will change, and then they will turn to you to add feature XYZ that they need. In fact, open source projects with permissive licenses are far more likely to receive industry backed financial support than a GPL-based or non-commercial license (take, for example, the Apache foundation). Such industry backing can help to fuel further innovation, while still keeping the code free for everyone else.

As to ownership, an open source license merely grants others the right to distribute (and, depending on the license, modify) the code, subject to certain constraints given in the license. It does not affect who owns the code. The authors of the code, in authoring the code, become and remain the code's sole copyright owners.

As for legal issues, most companies don't want to pay the huge fees or suffer the negative publicity of a lawsuit for copyright infringement or violating the terms of a software license agreement. Usually this will never even be an issue. If someone does infringe, the copyright owners (meaning you and whoever else wrote the code with you) can exercise their authority as copyright owners by suing a person or company for infringing on the terms of the license. This is a highly unusual situtation, though.

Answer 3

When I worked for a university, I asked my research professor to let me share my program with other people under the GPL. After I explained what it was, he was all too happy to say yes. The big points for him were:

1. He spent a lot of time writing grant proposals to get money from the government. Apparently, to the people in government agencies in charge of handing out research money, "open" anything is music to their ears. It means they only have to fund it once, not once per university! Be sure to list everybody else who uses it. Then when they're reading your grant, they'll think "If we fund university A, it's like we're funding universities A, B, C, D, and E!".

2. There aren't that many labs that do what we did, and (mostly by historical accident) they all used their own programs, with their own algorithms, file formats, and so on. Sharing data and comparing results wasn't impossible, but it wasn't as easy as it could have been. Sending everybody a copy of our program made this 10 times easier. Being the only user of a program, and having one programmer in the world understand your system, is a risk. If everyone is using your software, and has the source code, there's less risk of catastrophe.

How I'd respond to the issues you raise:

• "plagiarism for commercial gain" -- no different than if you publish any other work; it's still copyrighted

• "involvement with legal issues" -- again, copyright is still copyright; the GPL is nice because in addition to the law, you'll have every geek on the internet on your side :-)

• "issues of ownership" -- I didn't really discuss this, but I would have been happy to sign over ownership to the lab or the university or whatever; I think it's common for the organization to officially own whatever you create while in their employ, and the GPL doesn't change this

Answer 4

Better propose LGPL or MIT or BSD license - they all count as open source as well but carry lower risk of litigation since they are actually more permissive and more premissive use means much lower litigation chances whihc I believe is the core and legitimate concern.

Theoretically speaking, you or your professor could be forced by a regent or a dean to enter a litigation over a GPL violation if he can show that it's winnable and has any kind of crazy personal reason to push for it. With the other 3 he has vistually no chance to do that since they essentially just allow everything as long as any use recognizes you as the author.

Answer 5

From the link you've given, you might find him tough to convince. The Licensing Controversy you linked to is that if the two of you release software that competes with Gaussian's software, then Gaussian might refuse to licence their software to him, and he could lose access to one of his research tools.

The GPL can't do much about that -- because it's not about the license on your software but on Gaussian's software. So if that is the sticking point, then the argument isn't really about the GPL at all. It's whether releasing your software will bring him, yourself, and the university enough benefit (reputational, career, or even financial) to be worth the risk of Gaussian saying "sorry prof, but we're no longer willing to license our software to you".

Answer 6

You've had a lot of good replies already. But, I've a few further suggestions. First, to address the concerns that you mentioned:

plagiarism for commercial gain

The GPL is, as others have noted, a copyright licence, and thus provides the same level of legal protection against use of the code not authorized by the license as do proprietary licences and the strategy of not letting the code out of the lab under any license. In practice, the level of protection is greater than at least the "don't let it out" strategy. Under the later, most likely only the University's lawyers will care to pursue plagiaristic use. Under the GPL, the University's lawyers will be joined by an army of geeks, some with law degrees and a professional interest in defending the terms of the GPL.

A few of your respondents suggest a more permissive licence would help. I'd suggest that if your Prof's concern is with others taking the code in question and making profit on it, the GPL with its "viral" terms is a better bet. The GPL certainly allows others to redistribute code for profit so long as they respect the terms of the GPL. But, since those terms require that they make the code available under the same terms as which they obtained it, the economic forces tend to drive the price for the code to 0.

Releasing code under the GPL leaves only the same opportunities for (legal) commercial gain as does publishing a journal article describing some method or process that can be exploited commercially.

involvement with legal issues

Again, the GPL is a copyright licence and so provides the same level of legal protection as does any other licence or the strategy of not distributing at all. Again, use of the GPL provides, in practice, a larger pool of resources upon which to draw than do proprietary licences or not distributing at all.

issues of ownership

Again, the GPL is a copyright licence and has no effect on the ownership of the code. It remains owned by the holder of the copyright.

Some other points:

BusyBox and Westinghouse

You might cite the recent (2010-08-03) decision against Westinghouse for their violation of the terms of the GPL with regards to BusyBox to show that the GPL has teeth and addresses the point of "plagiarism for commercial gain." Westinghouse was ordered to pay triple damages, legal costs, and to surrender all remaining televisions that used BusyBox in violation of the GPL. It seems to me that this case ought to go a long way to address the concerns, especially if you are located in the US.

Peer Review

Releasing the code that goes into the production of research results seem essential if academia is to continue to have meaningful peer review. With the code available, others can inspect the entire process that leads to the results. Without the code, part of the research method is inside a black box. (Some sort of "clear source" licence would achieve this end, too, but don't mention that ;-))

Standing on the shoulders of giants

Academic research is a cumulative affair, where researches extend each others results for the common good. To keep code secret is to cause other researchers to reduplicate work pointlessly and to needlessly slow down the communal advancement of knowledge.

Answer 7

plagiarism for commercial gain

There will never be any physical protection from this happening. Relying on Law 'XYZ' for any actual protection is fruitless in this context. Take a look at the never ending issue with Piracy.

Bottom line:

You can't steal something that's free :)

involvement with legal issues

Thats why we have Lawyers...

issues of ownership

I'd think of it this way:

When we die; beyond our death, the only lasting thing we leave behind is our knowledge.

If you acquire some knowledge and never pass it onto others, then what is your legacy?

Answer 8

I think it's common for the organization to officially own whatever you create while in their employ, and the GPL doesn't change this

As far as i know if you are employed, and write some code your employer owns copyrights to it by default