The SciPy license says that it is "Copyright (c) 2003-2016 SciPy Developers", but you'd have to look in the git logs to figure out who those "SciPy Developers" are. That's good enough for some employers. For example, Google owns my recent contributions to NumPy but did not insist on adding a "Copyright Google" line to NumPy's license. On Fri, Mar 24, 2017 at 9:48 AM, Anne Archibald <peridot.faceted@gmail.com> wrote:

Not a lawyer, but currently the AUTHORS file indicates who owns the copyrights to scipy code. Perhaps a note on your line in the AUTHORS file to indicate that your contributions (on some dates) are owned by Los Alamos?

I would note in passing that many scipy contributors (including me) probably have or had similar constraints but are lucky enough that their employers don't care (or don't know). Many universities have rules about who owns intellectual property that can be surprising to their employees.

For many open source projects, there are enough contributors already that contacting everyone who owns the copyright is extremely difficult, and getting them all to agree to anything is almost impossible; thus large open-source projects rarely make such license changes, or do so with rather dubious legality by getting agreement from current and major past contributors only. I don't know that a national lab is any more likely to be trouble than an individual open-source contributor; at the least they should be easier to get a hold of.

Anne

On Fri, Mar 24, 2017 at 5:36 PM Tyler Reddy <tyler.je.reddy@gmail.com> wrote:

...

Hi,

I'm just wondering about the 'lowest friction' way to continue contributing to scipy when your employer owns the copyright to your open source contributions (but respects the license used by the project).

In short, Los Alamos would want me to put one line in one file that says they 'own' my contributions under the license of the project after Date xx/yy/zzzz. Hopefully that's a reasonable compromise (they originally wanted to put a line in each file & I said that definitely would not fly).

Not sure how annoyed people would be by that, or what specific file would be used, but they have been accommodating so if there's a request on the scipy end I could bring it up with them.

The biggest drawback I'm aware of is that changes to licenses may sometimes require (?) the involvement of the contributors (copyright holders)--if one such copyright holder is a national lab instead of a single individual, this might make the process a bit clunkier.

Best wishes, Tyler _______________________________________________ SciPy-Dev mailing list SciPy-Dev@python.org https://mail.python.org/mailman/listinfo/scipy-dev

_______________________________________________ SciPy-Dev mailing list SciPy-Dev@python.org https://mail.python.org/mailman/listinfo/scipy-dev

On Fri, Mar 24, 2017 at 9:48 AM, Anne Archibald <peridot.faceted@gmail.com> wrote:

Not a lawyer, but currently the AUTHORS file indicates who owns the

I would note in passing that many scipy contributors (including me)

copyrights to scipy code. Perhaps a note on your line in the AUTHORS file to indicate that your contributions (on some dates) are owned by Los Alamos? That would certainly work on scipy's end. We have plenty of space for each entry to include this note, either next to Tyler's name or under the Institutions subhead. https://github.com/scipy/scipy/blob/master/THANKS.txt probably have or had similar constraints but are lucky enough that their employers don't care (or don't know). Many universities have rules about who owns intellectual property that can be surprising to their employees.

For many open source projects, there are enough contributors already that

contacting everyone who owns the copyright is extremely difficult, and getting them all to agree to anything is almost impossible; thus large open-source projects rarely make such license changes, or do so with rather dubious legality by getting agreement from current and major past contributors only. I don't know that a national lab is any more likely to be trouble than an individual open-source contributor; at the least they should be easier to get a hold of. Yeah, we're never going to change the overall license both because it's already more trouble than it's worth (adding LANL makes no difference) and our current license is so permissive that there would be no real reason to in any case. -- Robert Kern