Open joelposti opened 8 months ago
Would it make sense to have different kinds of attestations (credentials) with different revocation mechanisms?
If you present an attestation of your bank account number or a credential proving that you own a car with a specific license plate number, it doesn't make much sense to set strict requirements for non-traceability of the revocation, since the verifiers get identifying information as attribute values. The same, of course, applies to use cases where the user uses their PID for authentication.
The issuer of a credential could label the credentials they issue using a three-tier classification:
Thank you for bringing this to our attention. The requirement 6.3.4.7.3.ii will be removed in the next version of ARF (ARF v1.4). If you encounter any further issues, please don't hesitate to let us know.
Thank you! Looks like the requirement has been removed from ARF 1.4. This issue can be closed.
In section 6.3.4.7 of the ARF there is the following requirement:
Section 6.3.4.8 of the ARF presents a number of possible revocation mechanisms such as Attestation Status List and Online Attestation Status Protocol. It is difficult to see how the presented mechanisms, with the exception of the short-lived attestations mechanism and perhaps the OASP stapling, fulfill the requirement because they rely on the fact that the attestation has some sort of revocation-related identifier. That identifier can be used by the Relying Party for correlating the User.
Should the requirement be relaxed or dropped? Or should more advanced revocation mechanisms be used?
If more advanced revocation mechanisms are used it could be that the requirement 6.3.4.7.8 gets in the way next.