4 ms·
The comparison isn't entirely fair because there's no reason not to use TLS 1.3 on HTTP 1 or HTTP/2 connections. The 0-RTT advantage of QUIC is also available t
by jeroenhd 5y ago
The comparison isn't entirely fair because there's no reason not to use TLS 1.3 on HTTP 1 or HTTP/2 connections. The 0-RTT advantage of QUIC is also available to those protocols, which is one of the things the article claims make HTTP/3 faster.
The methodology section also doesn't say if 0-RTT was actually enabled during testing. I'd argue that any system or website with an admin interface or user account system should not enable 0-RTT without very strict evaluation of their application protection mechanisms, making it useless for many API servers and data sources. It's fine for static content and that can help a lot, but with the benefit of multiplexing I'm not sure how useful it really is.
- johncolanduoni 5y agoThis is true, though it’s worth mentioning that 0-RTT on TLS 1.3 + HTTP/2 still has one more roundtrip (for the TCP connection) than 0-RTT HTTP/3. Plus even if you don’t enable 0-RTT on either there’s still a one roundtrip difference in favor of HTTP/3, since it combines the (equivalent of) the TCP establishment roundtrip and the TLS handshake roundtrip. That’s the real advantage in connection establishment.
- lemagedurage 5y agoYes, TLS1.3 cuts out a RTT compared to TLS1.2, but QUIC cuts out another RTT by combining the TCP and TLS handshakes into one. For a full handshake: TCP + TLS1.2: 3 RTT TCP + TLS1.3: 2 RTT QUIC including TLS1.3: 1 RTT And for subsequent connections: TCP + TLS1.3: 1 RTT QUIC including TLS1.3: 0 RTT (but no replay attack protection)