[License-discuss] Wrapping OSI licenses (UNCLASSIFIED)

[Tzeng, Nigel H.]

Marc,

I’m not necessarily advocating for a federal FOSS agreement but it strikes me that if one the agencies that has been forward thinking and generating GOSS code for public use as part of their strategy for nearly two decades under the NOSA license and likely has released one of the oldest GOSS codebases around (https://code.nasa.gov, https://github.com/chrislgarry/Apollo-11) believes it needs one that the OSI shouldn’t continue to be a roadblock for improving that already approved license.

Regards,

Nigel

On 2/20/18, 12:52 AM, "License-discuss on behalf of Marc Jones" <license-discuss-bounces at lists.opensource.org<mailto:license-discuss-bounces at lists.opensource.org> on behalf of marc at joneslaw.io<mailto:marc at joneslaw.io>> wrote:

Nigel,

I generally agree with your points. A comprehensive FOSS strategy does require considering a lot of different things including how you can acquire the rights to release the code as FOSS and determining what patent license works for your organization. That doesn't necessarily mean that the government needs a unique FOSS license/agreement.

You are right the government can not simply declare vendors release their code under a FOSS licenses. I did not mean to suggest that that is what DDS as doing. But it can contract for the required rights. To me it always seemed cleaner to acquire the right to release code under a FOSS license in a contributor agreement rather than having a FOSS license doing double duty. In the case of the government perhaps a clause in a procurement contract could serve that purpose of getting the government that right. Often the CRADA, FAR, and DFAR regulations allow the flexibility for the government to negotiate for those rights, even if they also mandate that the government at minimum get a certain set of government purposes rights.

The process for determining when patents should be licensed and what FOSS needs a software patent license at all, seems like it could take advantage of some of the processes of the programs you listed to determine what patents the government wants to license now. That might mean the government needs a new FOSS agreement, but perhaps it could in many cases also just grant a gratis patent license along with a existing FOSS license that does not grant any patent license, or, as you suggested, perhaps one of the existing FOSS licenses would provide an appropriate license.

-Marc

On Mon, Feb 19, 2018 at 3:25 PM Ben Hilburn <bhilburn at gmail.com<mailto:bhilburn at gmail.com>> wrote:
Ah, thanks for the clarifications! And yes, I agree, those are fair points.
Cheers,
Ben

On Mon, Feb 19, 2018 at 2:38 PM, Tzeng, Nigel H. <Nigel.Tzeng at jhuapl.edu<mailto:Nigel.Tzeng at jhuapl.edu>> wrote:

On 2/19/18, 8:38 AM, "License-discuss on behalf of Ben Hilburn" <license-discuss-bounces at lists.opensource.org<mailto:license-discuss-bounces at lists.opensource.org> on behalf of bhilburn at gmail.com<mailto:bhilburn at gmail.com>> wrote:
Not sure I'm following your argument, here? If a party has been contracted by the government to write code, as part of contract negotiations the government can require that the code be delivered as FOSS. Especially with the recent changes in the NDAA, the government is clearly trying to push acquisition officers to be more knowledgeable about these things.
My point was that there may be no contractor code at all and therefore there is no code under any sort of FOSS license, just public domain.  Depending on the existence of contractor developed code under a FOSS license to make the entire code base FOSS doesn’t work in this case.

The DDS policies posted online don't discuss patents much, aside from a bit in the license selection portion, "Our suggestions for permissive licenses are MIT, ISC, or BSD-3 unless patents are potentially involved in which case we suggest Apache 2.0 although the others work too." I have no idea how intra-government but inter-org patent licensing works, though, so I don't have anything to add to this piece of the discussion. It's worth noting, though, that the broader open-source community has long dealt with the same question, "what if someone unknowingly implements a patent and publishes it under the Apache license," problem that you raise here; I don't think it's unique.

The use of Apache 2.0 is problematic because it IS a fairly unique problem.  The issue is the USG as a single entity implies that a patent grant under Apache 2.0 provided by the ARL gives that patent away even if it was not created by the ARL but some other part of the federal government.

Your scenario is different where the developers implements a patent someone else owns.  They don’t own the patent so the patent grant under Apache is meaningless.

The only place that the broader open-source community has dealt with this issue is in the educational world which is why we have ECL v2.  Which is Apache with a patent grant only for those patents owned by the authors of the code.

_______________________________________________
License-discuss mailing list
License-discuss at lists.opensource.org<mailto:License-discuss at lists.opensource.org>
http://lists.opensource.org/mailman/listinfo/license-discuss_lists.opensource.org

_______________________________________________
License-discuss mailing list
License-discuss at lists.opensource.org<mailto:License-discuss at lists.opensource.org>
http://lists.opensource.org/mailman/listinfo/license-discuss_lists.opensource.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensource.org/pipermail/license-discuss_lists.opensource.org/attachments/20180220/1cd7b068/attachment.html>