Re: [TLS] WGLC for "Deprecating TLSv1.0 and TLSv1.1"

[Peter Gutmann <[e-mail]>]

Martin Rex <[e-mail]> writes:

>BEAST is an attack against Web Browsers (and the abuse known as SSL-VPNs), it
>is *NO* attack against TLS 

That actually applies to an awful lot of recent attacks on TLS - they're
attacks that rely on web software that's actively cooperating with the
attacker, not attacks on TLS per se.  Similar issues affect numerous attacks
on CMS (branded as S/MIME in email) and OpenPGP, they require mail software
that actively cooperates with the attacker.

For any new attack on a protocol like TLS, you really need a three-stage
summary of what's vulnerable:

1. Web-based use of TLS: Pretty much everything.
2. Non-web-based use of TLS: Very little.
3. Non-web-based with a few basic mitigations (EMS, EtM): Nothing, or close to it.

Peter.