Re: [quicwg/base-drafts] Flow control for post-handshake CRYPTO messages (#1834)

[David Benjamin <[e-mail]>]

I don't think that's an accurate reduction.

TLS stacks usually impose ad-hoc limits tighter than the maximum 2^24, sometimes as a function of the maximum certificate size the caller configures. It's unfortunate that it's ad-hoc but different consumers do indeed tolerate different certificate sizes. It's plausible to do that analysis per-message and to do it per-flight since the set of messages in a flight is different.

Post-handshake is very different. The pattern of messages is no longer fixed. Rather it is a property of how the application uses TLS. Indeed the QUIC spec even calls out applications choosing to send tickets at various points during the connection. Thus TLS-level limits, ad-hoc or spec'd, are not sufficient in this case. QUIC needs to do its share of the work.

(There's a reason I dislike post-handshake messages so much. :-) People always forget this complexity.)

-- 
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/1834#issuecomment-427228821