Re: [TLS] a proposal for TLS 1.3

[Yoav Nir <[e-mail]>]

Hi Nikos. This is pretty much what httpbis did for HTTP/2. Only it did not take a year or two - only two IETF cycles. 

So a competition could be announced now, with proposals presented in London, and a decision reached before Toronto.

While there are many research groups, they don't do protocol work so much. 

The real tricky thing here is not designing a "new TLS", but co-existence with SSLv3, TLS 1.0, 1.1, and 1.2 without resorting to the kind of fallback mechanism where you try with TLS 1.3 (or 2.0, or even better - 4.0) and if you get a reset, try again with TLS 1.2, and if that doesn't work, go straight to SSLv3.

-----Original Message-----
From: [e-mail] [mailto:[e-mail]] On Behalf Of Nikos Mavrogiannopoulos
Sent: Tuesday, November 12, 2013 11:55 AM
To: [e-mail]
Subject: [TLS] a proposal for TLS 1.3

Hello,
 Having seen the current discussion for TLS 1.3, I have some proposal.
Since the changes requested are large, I'd suggest not to design TLS 1.3 in this working group. I'd suggest this working group to set the specs for the next TLS version (and better name it 2.0), and then announce a competition for the initial draft of the protocol. Then have the WG review the submitted proposals in a year or two, accept a winner, and the WG starts working on the winning draft.

I'm pretty sure there are many research groups that will be honoured to join such a competition and their results may outperform the results of the limited number of participants in the working group.

best regards,
Nikos