Re: [TLS] Last Call: draft-ietf-tls-renegotiation (Transport Layer

[Marsh Ray <[e-mail]>]

<flameSuit state="on">
	I'm thinking this is the only big sticking point left for
	draft-ietf-tls-renegotiation?
</flameSuit>

Martin Rex wrote:
>
> We do *NOT* need more than one signaling mechanism, so the use of
> MCSV needs to be a MUST and the use of an empty TLS extension RI
> on the initial ClientHello needs to be a SHOULD NOT.

I for one would like to see the day (say in TLS 1.7) when we no longer
need to send the MCSV on every Client Hello. It is, after all, a hack.

In order for this to ever be possible, we need servers from the first
implementation of this patch to accept and work with either the
extension or the MCSV.

The MCSV was a concession to those who needed, on occasion, to not send
any extensions. It sounds like the great majority of servers do not
break when sent extensions.

Modern browsers are sending the SNI extension, no need for the hack
there. Let them use the MCSV only for the (sigh) fall-back logic.

- Marsh