[License-review] Request for approval of new "MGB 1.0" license

[McCoy Smith]

A few thoughts, in my personal capacity.


On 9/24/2025 5:26 AM, Barksdale, Marvin via License-review wrote:
> Thanks all for the early review.
> Agreed, there are several formatting issues that may have been caused 
> by exporting to plain text that have corrupted the submitted version, 
> including added and missing characters. Although I am withdrawing this 
> submission, in anticipation of further review I would like to like 
> flag the areas where we are in alignment and hopefully clarify the 
> areas where we are not:
I assume you're going to do another revision to the license to address 
some of the concerns raised with a least the way the patent grant is 
articulated? If so, I won't get into details on that grant other than to 
briefly respond to your discussion below.
>
> 1.
>      Regarding the questions around intent behind MGB 1.0's disclaimer
>     of obligations : "For clarity, no patent license is granted by a
>     Contributor for infringements caused by: (i) your or any other
>     party's Derivative Works, or (ii) the combination of the Work with
>     anything other than the Contributor Version."
>
>     Although we originally intended to add clarity to the initial
>     grant through utilizing an approach approved by the OSI in other
>     licenses such as MPL 2.0, I agree with the findings that this
>     language became contradictory in its adoption to the MGB 1.0
>     use-case.  I believe striking this clarifying clause and
>     integrating Contributor Version in the Contribution definition
>     does not interfere with MGB 1.0's core approach.
>
Obviously, there are OSI approved licenses that use Contributor version 
and Contribution to mean different things (example: 
https://www.mozilla.org/en-US/MPL/2.0/ or 
https://www.gnu.org/licenses/gpl-3.0.en.html) . I think the problem here 
was to try to use those predecessor licenses' language in a license 
(Apache-2.0) that doesn't formulate it that way. You might want to just 
make sure you're adjusting the patent grant in Apache-2.0 in a way 
that's consistent with those predecessor licenses' grants. MPL-2.0 has a 
grant that contemplates both definitions. Perhaps use that one instead?
>
> 1.
>
>      Regarding MGB 1.0's Patent Grant : "Subject to the terms and
>     conditions of this License, each Contributor hereby grants to You
>     a (perpetual*), worldwide, non-exclusive, sublicensable,
>     no-charge, royalty-free, irrevocable (except as stated in this
>     section) patent license, under any patent claims owned or
>     controlled by the Contributor, to make, have made, use, offer to
>     sell, sell, import, and otherwise transfer the Work and Derivative
>     Works, but only to the extent such patent claims claim inventions
>     embodied in [their Contribution(s)]."
>
>     The MGB 1.0 patent grant's intent is to grant rights under patent
>     claims controlled by Contributor to the Work and Derivative Works
>     to the extent such claims claim inventions embodied in their
>     Contribution (in the form it exists immediately after the
>     Contributor submits it.) Compared to the Apache 2.0 Patent Grant
>     to "patent claims licensable by such Contributor that are
>     necessarily infringed by their Contribution(s) alone or by
>     combination of their Contribution(s) with the Work to which such
>     Contribution(s) was submitted."
>
>     Apache 2.0’s grant of all licensable patent claims that are
>     necessarily infringed by the contribution or by combination of
>     their contribution with the original work opens the door to
>     downstream infringement via later acquired contributor-controlled
>     claims.  The effect here is similar to elements the GPL 3.0
>     license, which includes a grant to the contributor's essential
>     patent claims:  all patent claims owned or controlled by the
>     contributor, whether already acquired or hereafter acquired, that
>     would be infringed by some manner, permitted by this License, of
>     making, using, or selling its contributor version, but do not
>     include claims that would be infringed only as a consequence of
>     further modification of the contributor version.
>
I'm not sure this explanation makes sense. GPL is designed to make clear 
that any grant applies to subsequently-acquired patents ("whether 
already acquired or hereafter acquired,") even if those patents were not 
owned or controlled by the licensor at the time the 
contribution/contributor version was made. But you say that Apache-2.0, 
upon which you based your draft, has the problem of "open[ing] the door 
to downstream infringement [I assume you mean patent grants] via later 
acquired contributor-controlled claims. Which is it? If you want to 
carve out those claims from the grant, I think you should be explicit 
about it, and explain why it is you think carving out those claims is an 
improvement or necessary for certain licensors.
>
> 1.
>
>     The intent of MGB 1.0 is for the contributor to grant the patent
>     claims necessary to use the contributed software, not to, for
>     example, infringing claims later acquired as a consequence of
>     further modification (e.g. new infringement and the entire
>     portfolio exposure referenced by Lawrence Rosen in “Open Source
>     Licensing – Software Freedom and Intellectual Property Law"). Thus
>     this goal is accomplished even after removing the disclaimer of
>     combination language (see: 1. above) causing several issues.
>
Is there a specific section of Larry's book that you're referencing? In 
the discussion of OSL & AFL he says "Those patent claims are available 
to both the Original Work and Derivative Works. This is not a license to 
embody those patent claims in independent works." Are you trying to 
exclude "independent" (i.e., non-derivative) works, or something else? 
If you don't want to have subsequent modifications made by downstream 
licensees to be be licensed, I think you need to be clear about that and 
explain what sort of downstream modifications you are trying to exclude. 
Apache-2.0 (as well as other more recent OSI licenses) is fairly clear 
that it only applies to their own Contribution, or that Contribution "or 
by combination of their Contribution(s) with the Work to which such 
Contribution(s) was submitted." Are you trying to exclude the 
combination grant, or both the Contribution grant *and* the Combination 
grant? I think you need to be clear about that. If you want to exclude 
the Contribution grant, this license probably has the problem of not 
conveying all rights (or requiring payment for exercising rights) per 
OSD 1. If it is the Combination grant, and you want to expressly 
disclaim that grant, I think you might be running up against the state 
of the law of patent exhaustion as articulated in Quanta/LGE and progeny.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensource.org/pipermail/license-review_lists.opensource.org/attachments/20250929/2a05345e/attachment.htm>