Re: [quicwg/base-drafts] Improve Large Server Certificate Scenario (#3784)

[Jana Iyengar <[e-mail]>]

(I'll invoke the @mcmanus for his thoughts, since he wrote that article.)

At a high level, we should be working on reducing the handshake size, for two reasons: reduced packet loss, and reduced network load. Fewer packets in the handshake is generally good. Compression is aligned with this outcome, and it is not a steep hill. Simply inflating the client's hello goes in exactly the opposite direction.

I would suggest that simply moving to QUIC without cert compression was not expected to reduce handshake latency. It was always a critical part of gQUIC, and was always considered a part of the equation for realizing QUIC's potential.

That is how I read the article -- that to realize QUIC's promise, cert compression is necessary. Perhaps it might be useful to mention this somewhere in the TLS draft?


-- 
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/3784#issuecomment-649895802