Re: [quicwg/base-drafts] A fixed transport parameter profile is legitimate (#3429)

[Nick Harper <[e-mail]>]

My understanding of the argument is that we shouldn't force servers to condition their TPs based on the client's Initial packet (possibly including TPs). E.g. the server currently needs to condition its active_connection_id_limit TP based on whether the client's Initial had a zero-len SCID; PR #3426 relaxes that constraint.

Assuming we have some future sets of TPs for extensions (and these TPs are advertised values, not negotiated values) E_c on the client and E_s on the server: The server could always send E_s (so it doesn't have to condition its response), or the server could send E_c \cap E_s, and in either case the client will ignore values sent by the server not in E_c. For future extensions that have negotiated transport params, the server will have to condition those values based on what the client sent.

-- 
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/3429#issuecomment-582863797