[License-review] ModelGo Attribution License, Version 2.0 FINAL CALL FOR COMMENTS

Wed Sep 23 05:56:24 UTC 2026

Hi Pam,

Your response is quite troubling to me. I read it as saying that the
> disparate treatment of copyright owners versus mere users of the software
> is intentional.

Thank you for being direct. I can see why my earlier response was
troubling. I was trying to explain why the original trigger was limited to
an action against the Licensor, but I did not fully consider what that
limitation would mean in practice. I also treated the question of
discrimination under OSD 5 as if it answered the separate concern you
raised about OSI’s review criterion.

I understand why you read it that way. The original wording did
intentionally draw a line based on who the licensee sued: it covered a suit
against the Licensor, but not a suit against another user. I did not mean
that copyright owners should be treated as a favored class. More
importantly, I should not have presented the fact that the same condition
applied to every licensee as a complete answer to the concern about its
effect.

As Joshua Gay described in the OpenMDW thread, the motive for the patent
> termination provision was because patents threatened the singular goal of
> the open source license, which is unfettered freedom to use the software.
> In other words, the open source license grants virtually unlimited right to
> use the software, but a patent aggressor can stop that use with a patent
> claim. The response was to withdraw the patent aggressor's patent (and
> sometimes copyright) license to use the software. So I find the intentional
> maintenance of the ability to sue a user troubling, given the historical
> basis for the patent termination clause.

I understand this rationale, and I also share your concern for protecting
the Open Source commons and preserving users’ freedom to use the licensed
materials. My earlier response focused on the fact that MG-BY grants rights
from a Licensor to its recipients, and I missed a case: a recipient could
later acquire a patent and use it to sue another user of the model (please
correct me if I’m wrong, patent law is not my area of expertise). Under the
original wording, that suit would not trigger termination because the
defendant was not the Licensor. The recipient could keep its MG-BY patent
license while asserting its patent against another user. My point that
model publishers and users do not usually contribute back upstream does not
answer that case.

When I described the narrower rule as a way to extend the commons, I meant
that it would avoid making MG-BY a general restriction on patent disputes
between users. I now see that this was incomplete: it could also leave
other users exposed to a patent aggressor who keeps the license. I agree
the patent-termination trigger should not depend on whether the defendant
is the Licensor. The revised wording below applies to an action against any
person or entity, while limiting the consequence to the patent license.

I also note that the one of the requirements for new licenses is ‘The
> license does not have terms that structurally put the licensor in a more
> favored position than any licensee.’ This does put the licensor in a more
> favored position than a licensee.

I agree this criterion is directly relevant. I still do not think the
clause is discrimination under OSD 5: it is triggered by a patent action,
not by a person’s identity, field, or type of use. But that does not
resolve your separate point. A condition can apply to all licensees and
still raise the question of whether it structurally favors the Licensor.
The revised trigger is broader because it applies regardless of who is
sued, but I understand that OSI may still need to assess the remaining
difference in how the clause applies to a licensee and the Licensor.

I also don't understand why you are characterizing this license as a
> ‘single-grantor license’ and distinguish it from Apache on that basis. All
> licenses are ‘single grantor,’ Apache included, that is, a single
> contributor is granting a license to their own contribution. If you are
> saying that the Apache license may be a grant from multiple licensors
> because it may be sublicensing from inbound licensors, the same is true of
> the ModelGo Attribution License, which is also sublicensable. So I'm not
> following why you think that your situation is, can be, or should be
> different from how the Apache license works.

You are right that “single-grantor” does not distinguish MG-BY from Apache,
and I should not have used it as the reason for different treatment. Both
involve grants from contributors, and MG-BY is sublicensable too.

For comparison, the current wording is:

5. TERMINATION

This License shall terminate immediately if You breach any material term
and/or condition of this License, or if You initiate any legal action
against the Licensor alleging that the Licensed Materials and/or Derivative
Materials infringe any patent worldwide. Sections 3, 4 and 6 shall survive
the termination of this License.

Here is the revised TERMINATION:

5. TERMINATION

This License shall terminate immediately if You breach any material term
and/or condition of this License. *The patent license granted to You under
this License shall terminate immediately if You initiate any legal action
against any person or entity alleging that the Licensed Materials infringe
any patent worldwide.* Sections 3, 4 and 6 shall survive the termination of
this License.

This keeps the existing rule that a material breach terminates the whole
License. For patent litigation, it removes Derivative Materials as a
trigger, applies regardless of who is sued, and ends only the patent
license.
I believe this revision strikes a better balance and may make the license
more attractive to potential adopters. The ModelGo licenses would not be
what they are today without the feedback from everyone in the OSI
community. Those discussions have also helped me continue to refine my
understanding of Open Source, and I’m sincerely grateful for that.

Best,
Moming

Pamela Chestek <pamela at chesteklegal.com> 于2026年9月23日周三 01:46写道：

> Hi Moming,
>
> Your response is quite troubling to me. I read it as saying that the
> disparate treatment of copyright owners versus mere users of the software
> is intentional.  As Joshua Gay described in the OpenMDW thread, the motive
> for the patent termination provision was because patents threatened the
> singular goal of the open source license, which is unfettered freedom to
> use the software. In other words, the open source license grants virtually
> unlimited right to use the software, but a patent aggressor can stop that
> use with a patent claim. The response was to withdraw the patent
> aggressor's patent (and sometimes copyright) license to use the software.
> So I find the intentional maintenance of the ability to sue a user
> troubling, given the historical basis for the patent termination clause.
>
> I also note that the one of the requirements for new licenses
> <https://opensource.org/licenses/review-process> is "The license does not
> have terms that structurally put the licensor in a more favored position
> than any licensee." This does put the licensor in a more favored position
> than a licensee.
>
> I also don't understand why you are characterizing this license as a
> "single-grantor license" and distinguish it from Apache on that basis. All
> licenses are "single grantor," Apache included, that is, a single
> contributor is granting a license to their own contribution. If you are
> saying that the Apache license may be a grant from multiple licensors
> because it may be sublicensing from inbound licensors, the same is true of
> the ModelGo Attribution License, which is also sublicensable. So I'm not
> following why you think that your situation is, can be, or should be
> different from how the Apache license works.
>
> Pam
>
> Pamela S. Chestek
> Chestek Legal
> 4641 Post St.
> <https://www.google.com/maps/search/4641+Post+St.+%0D%0A++++++Unit+4316+%0D%0A++++++El+Dorado+Hills,+CA+95762?entry=gmail&source=g>
> Unit 4316
> <https://www.google.com/maps/search/4641+Post+St.+%0D%0A++++++Unit+4316+%0D%0A++++++El+Dorado+Hills,+CA+95762?entry=gmail&source=g>
> El Dorado Hills, CA 95762
> <https://www.google.com/maps/search/4641+Post+St.+%0D%0A++++++Unit+4316+%0D%0A++++++El+Dorado+Hills,+CA+95762?entry=gmail&source=g>
> +1 919-800-8033 <(919)%20800-8033>
> pamela at chesteklegal.com
> www.chesteklegal.com
>
> On 9/21/2026 3:31 AM, Moming Duan wrote:
>
> Hi Pam and All,
>
>
> Thank you for walking through the scenarios. Let me try again with your
> own examples, and please tell me if I get any of this wrong.
>
> On my earlier point: that was my reading of the definitions (I'm not a
> lawyer), and I may have stretched it. I think your reading is the better
> one. Licensor is the rights owner whose grant is in Section 2.1. This also
> settles "the" versus "a" for me: for this clause there is one counterparty.
> This license has no mechanism for a distributor to pass on patent rights of
> its own; every patent right it grants comes from the Licensor. If someone
> downstream has patented their improvement, that grant happens outside this
> license. So far as this clause is concerned, there really is just the one.
>
> Your three cases, exactly as they operate:
>
> - C sues User A, who only runs the model: C keeps the license.
> - C sues User B, a distributor: C keeps the license.
> - C sues the Licensor: C loses the license.
>
> The license is one grant from one grantor, and the exchange inside it is
> bilateral. The Licensor grants the practice of the Licensed Materials and,
> by 2.1(a)(iii), of Derivative Materials too. The licensee's side of the
> exchange is not to turn a patent claim against the grantor. When the clause
> fires, the Licensor takes back the one thing that was given, from the one
> party who took it and then attacked the giver with it. The Derivative
> Materials part follows from the same place: 2.1(a)(iii) puts their practice
> inside the grant, so a claim that a Derivative infringes is a claim that
> the granted practice is unlawful.
>
> Apache-2.0 looks different because its shape is different. Its patent
> licenses come from many contributors, and any claim about the work, against
> anyone, even a cross-claim, ends the patent licenses You hold. A
> single-grantor license has no such machinery, so we chose the narrower
> trigger. You lose the license only by suing the grantor; nobody loses
> anything for enforcing a patent against another user. Under OpenMDW-1.1,
> termination can follow a copyright claim, a suit you voluntarily joined
> rather than started, or a claim brought against anyone at all. Under MG-BY,
> the clause fires on one thing only: a patent suit you yourself initiate
> against the Licensor. It reaches nothing beyond what our own grant covers.
>
> On OSD 5 itself. Everyone who holds the license holds it on the same terms
> and loses it the same way. I can't find a licensee who gets less than
> another licensee. The Licensor is not a licensee; it gives the license
> rather than holds it. Apache has the same asymmetry between a Contributor,
> who grants patent rights and can have them taken back, and a plain
> recipient, who grants nothing and so has nothing to lose. I haven't seen
> anyone call that discrimination. Our clause has that shape with one giver
> instead of many, and I don't see a classification of persons in it, which
> is what I understand OSD 5 to be about.
>
> To press one point, respectfully: I am not aware that OSI has ever held
> that a patent-termination shield must protect everyone to satisfy OSD 5.
> That would be a new principle, and if the review establishes it, I will
> revise the clause accordingly. Thanks.
>
>
> Best,
> Moming
>
> Pamela Chestek <pamela at chesteklegal.com> 于2026年9月21日周一 11:44写道：
>
>> Moming,
>>
>> I'm not following what you're saying. As I understand McCoy's point, it
>> is this:
>>
>> - Licensor is any contributor who added copyrightable content;
>>
>> - User A is someone who is using Licensor's version of the software but
>> has not contributed any copyrightable content and is not distributing the
>> software;
>>
>> -User B is someone who is using Licensor's version of the software, has
>> not contributed any copyrightable content, but is distributing the
>> software, who you are characterizing as a sublicensor of Licensor's rights;
>>
>> - User C is the person who owns the allegedly infringed patent.
>>
>> ===============
>>
>> User C sues a Licensor for patent infringement - the license terminates.
>>
>> User C sues User A for for patent infringement - the license does not
>> terminate because User A is not a Licensor of anything, they are just using
>> the software.
>>
>> User C sues User B for patent infringement - the license does not
>> terminate. First, characterizing User B as a sublicensor may not be a
>> correct assumption - why wouldn't User C say "no, I'm not a sublicensee of
>> User B, I'm a direct licensee of Licensor who I haven't sued, so the
>> license doesn't terminate"? But even assuming User B is a sublicensor, it
>> doesn't terminate the license. User B is a sublicensor, not *"the *Licensor,"
>> i.e., the "rights owner" per your definition, so the license doesn't
>> terminate.
>>
>> That is the discrimination that I understand McCoy to be describing. The
>> license only terminates if User C sues someone who contributed
>> copyrightable content, because those are the only people (or companies) who
>> have a license that can be terminated. Everyone else can still be sued for
>> copyright infringement.
>>
>> The use of the word "the" in "the Licensor" in the termination section is
>> problematic for another reason. The corpus may have many licensors. I think
>> the more natural reading is that only the license from the person being
>> sued terminates, not all of them. If it said "any legal action against
>> *_a_* Licensor alleging that the Licensed Materials and/or Derivative
>> Materials infringe any patent worldwide" would terminate all of them. But
>> that doesn't solve the problem described above with mere users, not
>> contributors, being the defendant in the patent suit.
>>
>> Pam
>> Pamela S. Chestek
>> Chestek Legal
>> 4641 Post St.
>> <https://www.google.com/maps/search/4641+Post+St.+%0D%0A++++++Unit+4316+%0D%0A++++++El+Dorado+Hills,+CA+95762?entry=gmail&source=g>
>>
>> <https://www.google.com/maps/search/4641+Post+St.+%0D%0A+++++++++++++++Unit+4316+%0D%0A+++++++++++++++El+Dorado+Hills,+CA+95762?entry=gmail&source=g>
>> Unit 4316
>> <https://www.google.com/maps/search/4641+Post+St.+%0D%0A++++++Unit+4316+%0D%0A++++++El+Dorado+Hills,+CA+95762?entry=gmail&source=g>
>>
>> <https://www.google.com/maps/search/4641+Post+St.+%0D%0A+++++++++++++++Unit+4316+%0D%0A+++++++++++++++El+Dorado+Hills,+CA+95762?entry=gmail&source=g>
>> El Dorado Hills, CA 95762
>> <https://www.google.com/maps/search/4641+Post+St.+%0D%0A++++++Unit+4316+%0D%0A++++++El+Dorado+Hills,+CA+95762?entry=gmail&source=g>
>> +1 919-800-8033 <(919)%20800-8033>
>> pamela at chesteklegal.com
>> www.chesteklegal.com
>>
>>
>> On 9/18/2026 2:58 AM, Moming Duan wrote:
>>
>> Hi McCoy and all,
>>
>>
>> OSD 5 says the license must not discriminate against any person or group
>> of persons. Here, Licensor is a role in the license, not a person. And the
>> license itself says how the role moves. The Section 2.1 grant is
>> sublicensable, and 2.2(a)(i) requires a copy of this License to travel with
>> every Distribution. Even the definition of "Licensed Materials", things
>> "Distributed by the Licensor", only really works if each distributor is the
>> Licensor for what it distributes. So a licensee who passes the model or a
>> Derivative Material on is granting the license onward. It stands as
>> Licensor to the people it supplied, and a patent claim by those people does
>> trigger termination.
>>
>> About your May 2025 conclusion, I want to be precise. You wrote then that
>> this termination provision was not an OSD violation but an uptake issue,
>> and that conclusion was about the Derivative Materials part, which is your
>> Q4:
>> https://lists.opensource.org/pipermail/license-review_lists.opensource.org/2025-May/005764.html
>> The OSD 5 question, whether the trigger should reach beyond the Licensor,
>> is new in this final call, and you yourself put it as "arguably". So I read
>> it as an open question rather than a settled one, and not something to
>> concede on my own.
>>
>> It's true that a user who never distributed anything, and sues another
>> licensee, does not lose the license. That is deliberate. The patent grant
>> has one source, and the promise not to assert runs to that source. A
>> dispute between two people who never granted each other anything is outside
>> that exchange.
>>
>> So my reading is that the clause, as written, does not discriminate
>> against any person or group, and I'd like to keep the current wording.
>>
>>
>> Best,
>> Moming
>>
>> _______________________________________________
>> The opinions expressed in this email are those of the sender and not necessarily those of the Open Source Initiative. Communication from the Open Source Initiative will be sent from an opensource.org email address.
>>
>> License-review mailing listLicense-review at lists.opensource.orghttp://lists.opensource.org/mailman/listinfo/license-review_lists.opensource.org
>>
>> _______________________________________________
>> The opinions expressed in this email are those of the sender and not
>> necessarily those of the Open Source Initiative. Communication from the
>> Open Source Initiative will be sent from an opensource.org email address.
>>
>> License-review mailing list
>> License-review at lists.opensource.org
>>
>> http://lists.opensource.org/mailman/listinfo/license-review_lists.opensource.org
>>
>
> _______________________________________________
> The opinions expressed in this email are those of the sender and not necessarily those of the Open Source Initiative. Communication from the Open Source Initiative will be sent from an opensource.org email address.
>
> License-review mailing listLicense-review at lists.opensource.orghttp://lists.opensource.org/mailman/listinfo/license-review_lists.opensource.org
>
>
> _______________________________________________
> The opinions expressed in this email are those of the sender and not
> necessarily those of the Open Source Initiative. Communication from the
> Open Source Initiative will be sent from an opensource.org email address.
>
> License-review mailing list
> License-review at lists.opensource.org
>
> http://lists.opensource.org/mailman/listinfo/license-review_lists.opensource.org
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensource.org/pipermail/license-review_lists.opensource.org/attachments/20260923/a5fd0308/attachment-0001.htm>