[quicwg/base-drafts] QUIC Ossification (#2496)

[martinduke <[e-mail]>]

There is a long conversation happening in the mailing list about QUIC version number ossification:
https://mailarchive.ietf.org/arch/browse/quic/?gbt=1&index=ytjKHacss9s3i3nualU2tbU9Jco
I'm adding it here to bring in others and consider it on the agenda for a WG meeting.

I won't try to summarize everyone's position from the thread. Given the status quo, it would be easy for someone to write a dumb middlebox that drops QUIC version > 1, or indeed version > 2. This appears to have happened with TLS v1.2. Given all of our effort to counter ossification, it would be tragic if we fell down on this rather simple hurdle.

There is disagreement as whether obfuscation is valuable or not. Though I have no answers, I personally would like to explore a few questions:
1) Are there any changes to VN we can make to reduce the chance of ossification happening?
2) If ossification happens anyway, do we have strategy to do something to overcome it?
3) If we are going to end up with some more obscure mechanism, does it make sense to lock in on bytes 2-5 of every long header to be 0x00000001 in perpetuity?

-- 
You are receiving this because you are subscribed to this thread.
Reply to this email directly or view it on GitHub:
https://github.com/quicwg/base-drafts/issues/2496