Re: Serious design flaw in STARTLS documents

This all seems fine to me, except.

>A server MUST NOT return the STARTTLS extension in response
>to an EHLO command received after a TLS handshake has completed.

(That's the "TLS" extension, by the way...) I would add before that:
=====
Both the client and the server must know if there is a TLS session active.
A client MUST NOT attempt to start a TLS session if a TLS session is
already active.
=====
That is, even if the server doesn't say TLS, the client might issue a
STARTTLS command. This covers both sides of the attempt to do
TLS over TLS.

--Paul Hoffman, Director
--Internet Mail Consortium