[License-review] License Committee Report - January 2017

[Richard Fontana]

On Mon, Jan 09, 2017 at 08:08:26AM -0800, Josh berkus wrote:
> On 01/08/2017 07:52 PM, Richard Fontana wrote:
> 
> > Upstream Compatibility License v1.0 (UCL-1.0)
> > =============================================
> > 
> > Submission: https://lists.opensource.org/pipermail/license-review/2016-October/002856.html
> > 
> > Comments: Questions were raised about conformance with the
> > nondiscrimination policy of OSD 5. It was suggested that the license
> > be redrafted so that all downstream modifications are licensed under
> > the Apache License 2.0 (rather than just upstream licensors receiving
> > a copy under the Apache License 2.0).
> > 
> > Recommendation: At request of license submitter and others, OSI to
> > provide general guidance by commenting on whether a license that
> > privileges one class of licensees by giving it greater permissions
> > relative to other licensees conflicts with OSD or should be
> > discouraged or disapproved for non-OSD policy reasons.
> 
> Again, there was follow-up on this where we suggested an improvement to
> the structure of this language, and the submitter was waiting on
> feedback from the committee that that improvement would make it more
> acceptable.  They never got it.

Here's my understanding of the situation:

Nigel proposed UCL, a license which is basically: "OSL 3.0, except
that upstream licensors get downstream modifications under the Apache
License 2.0". Some comments suggested that this asymmetrical licensing
feature conflicted with OSD5. Others argued that it would be incorrect
to ground a policy objection to such a feature in the OSD.

I believe you and Nigel proposed that the OSI board provide some
general guidance on whether asymmetrical features like that which is
the main distinctive feature of UCL are inherently in conflict with
the OSD (or alternatively should not be approved because of some
policy rationale extrinsic to the OSD).

I believe you proposed an alternative approach under which all
modifications to code originally released under the (copyleft) UCL
would be licensed under the Apache License 2.0. This still seems to
have some 'asymmetry' but not to the same degree as the original
UCL. I think Nigel said he was willing to consider implementing this
approach but would wait for feedback from the OSI.

So the immediate action is for the OSI board to consider whether to
provide some guidance about the relevance of asymmetrical licensing
features in proposed licenses. The fact that a 'less asymmetrical'
approach to designing the UCL has been proposed has been passed on to
the OSI board.

Richard