We're in the process of revising the cbor diagnostic for https://datatracker.ietf.org/doc/html/draft-ietf-cose-merkle-tree-proofs
I still feel uncomfortable with how the commitment to the verifiable data structure is done in the protected header, and how the mutable / redactable proof types are referenced in the unprotected header. Different verifiable data structures have different proof types, for example binary merkle trees support inclusion proofs. I'm wondering if it would be better to make a top level unprotected header parameter that was a map, and then nest all the proof details under it, if we don't do this, we might see collisions on proof types, seems like we need to add another layer of nesting in the unprotected header. This would somewhat mirror how https://datatracker.ietf.org/doc/draft-ietf-cose-cwt-claims-in-headers/ creates a new map in the protected header. Instead of our current example output, we would have something like: protected header: { 1: -35, TBD_1: 1 // TBD_1 indicates a type of verifiable data structure, 1 indicates RFC9162 binary merkle tree. } unprotected header: { TBD_2: { // TBD_2 indicates verifiable data structure proofs 100: [ ... ], // 100 indicates inclusion proofs 200: [ ... ] // 200 indicates consistency proofs }, // ... other unprotected header params here... } This is what the current tooling we have outputs: Signed Inclusion Proof ~~~~ cbor-diag 18( / COSE Single Signer Data Object / [ h'a2013822...5f636838', / Protected header / { / Unprotected header / 100: [ / Inclusion proofs (1) / h'83020181...c6bf0202', / Inclusion proof 1 / ], 200: [ / Consistency proofs (1) / h'e5f5a481...c8ffd5e6', / Consistency proof 1 / ], }, h'', / Detached payload / h'6140da0f...419c8ec0' / Signature / ] ) ~~~~ Decoded Protected Header ~~~~ cbor-diag { / Protected header / 1: -35, / Cryptographic algorithm to use / 4: h'75726e3a...5f636838', / Key identifier / TBD_1: 1 / Type of verifiable data structure / } ~~~~ Regards, OS -- ORIE STEELE Chief Technology Officer www.transmute.industries <https://transmute.industries>
_______________________________________________ COSE mailing list [email protected] https://www.ietf.org/mailman/listinfo/cose
