Infrastructure

I think we should also consider use of the BSD license, which is significantly simpler than the Apache license, and grants wider latitude to others to use the licenses. One major difference is in the patent claim release in Apache 2.0, which might be useful (and is orthogonal to copyright).

Interestingly, although as a project we might decide on a license to use, each of the contributors works for an institution that will have its own licensing policies. Each contributor needs to be sure they have the authority to release their work under a license. For example, I work for UC, and all of my work is Copyright UC Regents, but there is a policy for academic grants that allows me to release open source (but UC still has copyright). When multiple contributors from different institutions modify files, the copyright status of the file/work becomes somewhat muddy (could either be copyright by the original contributor and others are modifying based on the license, or could be joint copyright on the file).

Another point: Metacat and Morpho are GPL, which has worked fine and has the benefit of keeping derived works available. But Kepler is BSD, because we didn't want the viral aspects of the GPL. Which means we can not include GPL code in Kepler -- in practice this takes a lot of attention to the license details as developers incorporate new libraries. Developing a good project-wide open-source licensing policy now would be useful.

Copying very relevant comment:1:#587 to user story #586. This task was specifically to evaluate the licenses currently appearing in the code and the libraries its dependent on (though the description is a bit misleading)

General solution is to work with Apache 2.0.

This process needs to be completed as part of hte prototype review.