On Sat, 10 Apr 2021 at 00:35, Michael Thomas <mike@xxxxxxxx> wrote:
On 4/9/21 4:26 PM, Phillip Hallam-Baker wrote:
It is only a 'three packet handshake' if you ignore the off path interactions with the DNS service. The timeout on DNS tens to be rather smaller than that most would be comfortable with for crypto.I don't see why it can't be long lived, but even normal TTL's would get amortized over a lot of connections. Right now with certs it is a 5 message affair which cannot get better. But that is why one of $BROWSERVENDORS doing an experiment would be helpful.
When I was designing Certificate Transparency, Chrome ruled out any side channel communications requirement during handshake. Given that DNS is required anyway, perhaps this would be different. However, the other problem is introducing DNS as a trust root - the DNS hierarchy is considerably less secure than CAs were even before CT but now it's really a very poor option in comparison.
Could be fixed with DNS Transparency, of course.
Mike
On Fri, Apr 9, 2021 at 7:17 PM Michael Thomas <mike@xxxxxxxx> wrote:
I wrote a blog post about how something like DANE could be used instead
of certificates for the TLS handshake to get it back to the original 3
packet handshake. I know that isn't news to a lot of people, but the
interesting part is that a Google could perform an experiment to see how
well it works in real life just like they did with Quic and Spdy. Since
Quic is all about making setup time faster, it seems like a pretty
reasonable experiment since it would cut out 2 packets generally
speaking since DNS can be cached.
https://rip-van-webble.blogspot.com/2021/04/quic-elephant-in-room.html
Mike
