[License-review] License Committee Report - for board meeting of 2013-11-06 [NASA, EUPL, and Tidepool]

[Richard Fontana]

On Thu, 26 Jun 2014 08:10:36 -0400
"Tzeng, Nigel H." <Nigel.Tzeng at jhuapl.edu> wrote:

> Or simply to ignore it since your corner case is highly unlikely to
> ever occur in the real world.  The number of non-NASA contributors to
> a NOSA project that is not another government agency or contractor
> for a governmental agency approaches zero. 

I don't believe OSI should approve/disapprove licenses based on
assumptions about which particular (few) entities are likely to use the
license. 

> In any case, which open source requirement is violated even if the
> existing wording is unchanged?  A question you've sidestepped twice.  

You mean OSD planks? I can come up with arguments for why the provision
as it exists could in practice be used to discriminate against fields of
endeavor, but the problem mainly lies outside the OSD. I'm no
arch-anti-proliferationist but this license clearly implicates the
policy against license proliferation, so it has to be scrutinized
closely for that reason. We should not be quick to approve new licenses
that seem to have problematically-drafted provisions even if the
problems are not mentioned in the OSD (e.g., excessive vagueness,
problems of comprehensibility, inconsistency across provisions). 

If this were an entirely new provision in a new license submitted by a
non-lawyer with no affiliation with a prestigious institution, I am
sure it would be criticized.

The awkward situation here is that the existing OSI-approved version of
NOSA contains a substantially similar provision. To me, personally, that
doesn't dispose of the issue. 

What bothers me particularly here is that Bryan says that this
provision is *required* to be in this license by virtue of the NASA
enabling statute. That is a really strong statement -- I understand it
as a contention that NASA would somehow be violating its own statutory
mandate if this provision weren't in the license. I don't know if that
argument was made back when the original version of NOSA was approved.
As some indication that this is dubious I point to the following
(clearly non-FOSS) NASA license, which contains no such provision:
https://fedoraproject.org/wiki/Licensing/NasaCDF

Then there are things like this:
https://review.openstack.org/gitweb?p=openstack/gantt.git;a=blob;f=gantt/__init__.py;h=b8044faecee0a314ab7144a06d1cab6ae5e54988;hb=092bd857cc1aa504c782214f2993b12ad2d7cbe8

If (as may be the case, based on what I know about OpenStack) that
should not be read as NASA granting a license under the Apache License
2.0, we presumably had NASA signing the OpenStack US Government CLA:
https://wiki.openstack.org/wiki/GovernmentCLA

If I'm not mistaken that special CLA was prepared by Rackspace because
NASA said it couldn't sign the regular CCLA. However it too is a
software license containing no counterpart to the anti-endorsement
clause in NOSA.

- RF