> How should inability to access key revocation lists impact the TLS handshake,
> if previous public keys and/or certificate hashes are not cached?
Nobody does revocation on the web, for some almost all encompassing definition
of nobody.
Instead, OCSP and OCSP stapling.
> I cannot see this in the standard. Considering that all one has to do is DDOS
> a certificate authority nowadays...
General PKI and key lifecycle issues are, properly, not part of the TLS spec.
/r$
_______________________________________________
TLS mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/tls
