Please see https://www.ietf.org/id/draft-camwinget-tls-use-cases-00.txt
for some example use case scenarios impacted by encrypted SNI.
As Ekr subsequently pointed out, it would be useful to make a
distinction between conformant and non-conformant use case scenarios
(which we plan to do in the next version).
Thanks
-- Flemming
On 11/13/17 8:16 PM, Bret Jordan wrote:
What I think I am more worried about right now is jumping in to
designing a technological solution before we know and understand what
is going to break and is a solution going to actually solve the
perceived problem(s) or make them worse. Technological changes do not
always make things better.
Open Questions:
1) Is encrypted SNI the best solution to address the perceived problem(s)?
2) Do we fully understand the problems we are trying to solve and
understand the best way of solving them?
3) Will this make things better or worse for the majority of use-cases?
4) Does it incur so much collateral damage that it hurts the average user?
5) If we make it client opt-in (which seems like a fundamental
requirement), does this single out the client for extra scrutiny by a
well funded threat actor or nation state?
Just some food for thought
Bret
Sent from my TI-99/4A
PGP Fingerprint: 63B4 FC53 680A 6B7D 1447 F2C0 74F8 ACAE 7415 0050
_______________________________________________
TLS mailing list
TLS@ietf.org
https://www.ietf.org/mailman/listinfo/tls
_______________________________________________
TLS mailing list
TLS@ietf.org
https://www.ietf.org/mailman/listinfo/tls
