[License-review] For Approval: OpenMDW License Agreement, versions 1.1 (OpenMDW-1.1)

[横井雄太]

Josh, Mike, Sado-san, Pamela, Richard, and all,

Josh's latest message helped me isolate what I think may be the threshold
question in this discussion.

Mike's response to my earlier question clarified that OpenMDW-1.1 does not
make its termination triggers symmetrical with all of the rights it grants.
Database-right and trade-secret claims do not trigger termination; patent
and copyright claims do. Mike explained that this boundary reflects the
litigation threats the drafters considered realistic for openly published
Model Materials.

I think Josh is asking a logically prior question, however: even if
copyright litigation is a realistic threat to model publishers, what
open-source principle makes copyright litigation an acceptable termination
trigger in the first place?

Perhaps the cleanest way to test that is to strip away every broader
feature of Section 5.

Suppose a license provided only this:

- a licensee brings a copyright infringement claim against the licensor;
- the claim alleges that the licensed material itself infringes copyright
owned by that licensee;
- the claim concerns only that licensed material, not training conduct,
unrelated material, or some other dispute;
- filing or maintaining that claim terminates only the copyright permission
granted to that claimant for that same licensed material;
- every other grant remains intact.

Would that narrowly matched copyright-for-copyright termination provision
itself be incompatible with open-source principles?

If yes, then the line seems categorical: copyright litigation cannot
properly serve as a defensive termination trigger, even where the asserted
right, licensed subject matter, counterparty, and terminated grant are all
tightly matched.

If no, then copyright as a trigger is not itself the decisive problem, and
the actual OpenMDW-1.1 provision can be examined separately along at least
three axes:

1. Scope of consequence:
   Section 5 terminates all rights and grants, rather than only the grant
corresponding to the asserted right.

2. Scope of counterparty:
   The triggering lawsuit may be against "any person or entity," rather
than only the licensor or provider of the relevant Model Materials.

3. Timing:
   Termination occurs upon filing or maintaining the claim, before the
merits of the infringement allegation have been determined.

This also seems to sharpen the nexus-and-proportionality point raised by
Sado-san.

Nexus can tell us how closely the claim must relate to the licensed subject
matter. Proportionality can tell us whether the consequence is too broad.
But neither necessarily answers the threshold question Josh identified:
whether copyright litigation may legitimately be used as a termination
trigger at all.

So rather than asking whether Section 5 as a whole is too broad, I think
the more diagnostic question is:

Where is the first point at which the provision crosses the open-source
boundary?

If the minimal hypothetical above already crosses it, then the broader
features of Section 5 are secondary.

If it does not, then identifying which additional feature crosses the line
would make the disagreement much more concrete and may also make the
eventual OSI decision easier to articulate as a reusable principle rather
than as a conclusion specific to OpenMDW.

I am not arguing for either outcome here. I am trying to identify the
smallest proposition on which the disagreement actually turns.

I would be very interested in how others see that boundary.

Best,
Yuta Yokoi
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.opensource.org/pipermail/license-review_lists.opensource.org/attachments/20260904/e98de840/attachment-0001.htm>