Status: adopted 2026-08-27, while this project had zero outside contributors.
That timing is the point. A patent policy adopted before anyone has contributed binds everyone who ever does. One adopted afterwards has to be agreed retroactively by people who have already given you their work, which in practice means it never happens. Every recognised standards body requires a policy of roughly this shape before it will consider a document, so a specification without one is not adoptable however good the engineering is.
This is not legal advice, and nobody here is a lawyer. It is a plain-language policy modelled on how IETF and W3C handle the same problem. Before anything in this repository is relied upon commercially, have counsel read this file.
Unchanged from the day this project started, and stated first because it is the part implementers most need:
No patent search has been performed. No essentiality claim is made. No patent licence is granted by this project beyond the contributor grants described below.
Power beaming is a patent-dense field: retrodirective arrays, rectenna topologies, beam-forming control. It is entirely possible that implementing parts of this specification requires a licence from a third party. This project has not investigated that and is not in a position to. Implementers should take their own advice.
Nothing below changes that. What follows governs contributions, not the specification’s freedom to operate.
[IPR-1] Anyone contributing to this repository SHALL disclose any patent or patent application they, or the organisation they contribute on behalf of, know to be essential to their contribution.
[IPR-2] Disclosure is required for patents the contributor actually knows about. This policy does not require a patent search, and no contributor is expected to conduct one.
[IPR-3] A disclosure SHALL be made before or at the time the contribution is
submitted, by opening an issue titled IPR disclosure or by a note in the pull request.
Recorded in IPR-DISCLOSURES.md.
[IPR-4] Where a contributor cannot disclose — because the patent is unpublished, or their employer forbids it — they SHALL say that a non-disclosable interest exists rather than stay silent. A stated gap is workable. A silent one is not.
This is the shape of IETF BCP 79: disclose what you know, no search obligation, and the disclosure is public. It works because the cost of complying is near zero and the cost of being caught not complying is severe.
[IPR-5] By contributing, a contributor grants a royalty-free, non-exclusive, worldwide, irrevocable licence, to anyone implementing this specification, under any patent claim they own or control that is necessarily infringed by implementing their contribution as it appears in the specification.
[IPR-6] The grant extends only to the contributor’s own contribution, not to the specification as a whole and not to any third party’s patents.
[IPR-7] The grant may be terminated, as to a party, if that party brings a patent infringement action alleging that this specification or an implementation of it infringes their patents. This is the defensive-termination clause of Apache-2.0 §3, applied to the specification text as well as the code.
[IPR-8] A contributor unwilling to make this grant SHALL say so before contributing. Their contribution will be declined rather than silently accepted on unclear terms.
Most standards bodies accept RAND, meaning reasonable and non-discriminatory terms: a contributor may charge for a licence provided they charge everyone comparably. It is the compromise that gets large patent holders to participate.
Royalty-free is stronger, and it is chosen here for a reason specific to this project’s situation: there are no contributors yet. The only person bound by it today is the author, who holds no relevant patents, so it costs nothing to commit. Adopting the stronger term now and holding to it is worth more than adopting the weaker one and hoping.
If an organisation with a genuine patent portfolio later wants to contribute and cannot
make a royalty-free grant, that is a conversation worth having, and it would be handled by
amending this policy in public rather than by quietly accepting the contribution.
GOVERNANCE.md governs how such a change would be made.
IPR-DISCLOSURES.md records every disclosure received.
It is currently empty, and that emptiness is a dated statement rather than an oversight. It means: nobody has told this project about an essential patent. It does not mean none exists. See §1.
The two are separate and both apply.
| Copyright | Patents | |
|---|---|---|
| Specification prose | CC BY 4.0 (LICENSE-TEXT) |
[IPR-5] |
| Schemas, examples, tooling | Apache-2.0 (LICENSE-CODE) |
Apache-2.0 §3, plus [IPR-5] |
Apache-2.0 already carries an express patent grant for code. CC BY 4.0 carries none. It is a copyright licence and says nothing about patents at all. [IPR-5] exists mainly to close that gap on the prose half, which is where the normative requirements live and therefore where an essential patent would actually bite.
It binds contributors. It cannot bind anyone who has never contributed, which is everyone who currently holds a patent in this field. A third party’s essential patent is entirely unaffected by anything written here, and no policy any standards body has ever written changes that either.
The honest summary: this makes the specification adoptable by a body, and makes contributions safe to build on. It does not make implementation safe. Nothing can, short of a search this project has not done.