Address the alg language per the comment here

[OR13]

We may need to beef up the language about inclusion of alg as a required header parameter, but we could do that in a separate PR.

#11 (review)

[Sakurann]

+1

If only the proof property is needed for the chosen signature method (that is, if there is no choice of algorithm within that method), the alg header MUST be set to none.

I think alg = none should not be allowed. what is the point of using verifiable credentials if there is no integrity protection..

[OR13]

I think alg = none should not be allowed. what is the point of using verifiable credentials if there is no integrity protection...

I agree with you, but the VCWG decided to define VPs exactly this way.

If this document forbids alg: none in a VC-JWT, there are use cases that data integrity can handle that jwt cannot.

I think it's probably worth accepting the deviation there are explicitly forbid alg: none in the header of either a VC or a VP.

[Sakurann]

What are those use-cases? (Or where can I read them up?)

alg=none in VP-JWT does not make sense to me. What kind of information in a presentation is useful without a cryptographic holder binding that VP-JWT gives? Just send VC-JWT instead of an unsigned VP-JWT, no?

alg=none in VC-JWT also does not make much sense, but I am more willing to accept there might be issuers who want to reuse the syntax of VC-data-model (but than why not sign...)

[OR13]

alg=none in VP-JWT does not make sense to me.

https://www.w3.org/TR/vc-data-model/#presentations-0

^ basically how does vc-jwt support this use case... its ok if it doesn't... but better to be explicit about this.

alg=none in VC-JWT also does not make much sense

Agreed, also, its probably illegal... depending on your definiton of "external proof".

https://www.w3.org/TR/vc-data-model/#proofs-signatures

[OR13]

@Sakurann recent PRs elaborated on this in great detail, based on conformance to the normative requirements of the core data model... please review.

cc @mprorock @selfissued

Due to the core data model stating proof is optional, I will object to creating an incompatibility in functionality.

But I am happy to discuss in depth.

[OR13]

I'm marking pending close, on the offhand chance, folks doing the review, feel the current text is sufficient.

[OR13]

Marked pending close over 1 week ago, closing.