In the exported authenticators draft we claim that "The application layer protocol used to send the authenticator SHOULD use TLS as its underlying transport." This is of course natural -- why would you be using TLS authenticators if you were not using TLS -- but it seems that we would also benefit from saying what properties are actually *required* of the channel used to transport the authenticator. (Confidentiality? Binding to the key material of the TLS connection?)
-Ben _______________________________________________ TLS mailing list TLS@ietf.org https://www.ietf.org/mailman/listinfo/tls
