[TLS] The consensus is half-mistaken (Re: Confirming consensus for ALPN)

[Nico Williams <[e-mail]>]

On Fri, Mar 15, 2013 at 4:45 PM, Eric Rescorla <[e-mail]> wrote:
> At today's TLS WG f2f meeting we had rough consensus to adopt
> ALPN as the solution to the upper-layer protocol negotiation problem
> posed to us by the HTTP WG.
>
> if you have an opinion on this topic and were not at the meeting,
> please send a message to the mailing list and/or the chairs by
> Friday March 29.

I believe that privacy protection for as much as possible of the
application layer negotiation is highly desirable.  In this sense I
think the consensus for ALPN is mistaken.  I do believe NPN is too
complex, and in that sense the consensus for ALPN is correct.

Can we find a middle of the road?

My compromise proposal is based on providing privacy protection for
the record type in TLS (post-CCS, of course) -- a feature that should
be desirable anyways and separately (e.g., to hide re-negotiation from
non-MITM middleboxes).  Such a feature would allow a post-handshake
non-application-type record for optimistic application protocol
negotiation (zero extra round trips in the optimal case, one more
round trip in the pessimal case).

Nico
--