New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

cvxopt is gpl licensed #486

Closed
samuelstjean opened this Issue Nov 28, 2014 · 4 comments

Comments

Projects
None yet
3 participants
@samuelstjean
Contributor

samuelstjean commented Nov 28, 2014

I mentionned this on another thread, but nobody noticed/responded, and I saw that Elef wanted to release by next week on the mailing list. Well, so what should we do with the module? The code says that it must use cvxopt for the constrained optimisation and just exit if it's not installed, but there was another time where it was working with it I guess?

Else it can be dropped or the license changed. which I wouldn't mind since I'm sitting on some stuff that can't be merged 'cause of license issues.

@arokem

This comment has been minimized.

Member

arokem commented Nov 28, 2014

I don't think the license is an issue. We're just importing cvxopt, not rewriting it.

@arokem arokem closed this Nov 28, 2014

@arokem

This comment has been minimized.

Member

arokem commented Nov 28, 2014

I don't think the license is an issue. We're just importing cvxopt, not
rewriting it.

On Fri, Nov 28, 2014 at 2:38 PM, Samuel St-Jean notifications@github.com
wrote:

I mentionned this on another thread, but nobody noticed/responded, and I
saw that Elef wanted to release by next week on the mailing list. Well, so
what should we do with the module? The code says that it must use cvxopt
for the constrained optimisation and just exit if it's not installed, but
there was another time where it was working with it I guess?

Else it can be dropped or the license changed. which I wouldn't mind since
I'm sitting on some stuff that can't be merged 'cause of license issues.


Reply to this email directly or view it on GitHub
#486.

@matthew-brett

This comment has been minimized.

Member

matthew-brett commented Nov 29, 2014

I think it not clear what the license situation is when BSD licensed module 'dipy' imports GPL licensed module 'cvxopt'. See for example : http://en.wikipedia.org/wiki/GNU_General_Public_License#Linking_and_derived_works

If you think that dynamic linking against GPL code subjects you to the GPL (see link), then a Python module foo.py that loads a GPL library in an extension needs to be under the GPL, but I don't know what the situation is for a Python module bar.py that imports foo.py.

But, in any case, our BSD license is compatible with the GPL, so that, if we do become subject for the GPL by linking to a GPL library, our code can be deemed to be GPL, but remain BSD while it does not link to the GPL library. So I agree, I don't think we need to change our license to use cvxopt.

@samuelstjean

This comment has been minimized.

Contributor

samuelstjean commented Nov 29, 2014

Well, depends on the inerpretation, but apparently anything loading glp code and using it should also be gpl licensed. As an alternative, there are lib which create scripts ot only interface with those parts and save .npy files, which are loaded afterward by the rest of the library

Anyway, according to this https://softwarefreedom.org/resources/2007/gpl-non-gpl-collaboration.html#x1-50002.3 at the end of the paragraph, it seems that the project could be dual-licensed, in the sense that it would be BSD as a whole, but everything using cvxopt and using a file contained in that would be gpl.

Edit : The gpl also seems to plainly state that anything using gpl code is de facto gpl licensed https://www.gnu.org/licenses/gpl-faq.en.html#IfLibraryIsGPL

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment