- From: Willy Tarreau <w@1wt.eu>
- Date: Mon, 20 Aug 2012 10:59:17 +0200
- To: Yoav Nir <ynir@checkpoint.com>
- Cc: Julian Reschke <julian.reschke@gmx.de>, "ietf-http-wg@w3.org Group" <ietf-http-wg@w3.org>
On Mon, Aug 20, 2012 at 11:50:18AM +0300, Yoav Nir wrote: > >> 4. The client also caches that this server supports HTTP/2.0. Next > >> time, it won't need the upgrade mechanism. > > > > How does the recipient then know that the message is HTTP/2.0? > > Remains to be seen what HTTP/2.0 will actually look like, but SPDY-00 is very > easily distinguishable from HTTP/1.x. Connections start with control frames, > and control frames have the high bit on the first octet set. HTTP/1.1 > requests tend to have ASCII characters at the start. I think a heuristic > approach to telling HTTP/1.x from HTTP/2.x is good enough for the servers. In fact those arguments were made for WebSocket too, and this went into a one-year long dabate with examples of how dumb intermediaries could be fooled etc... In theory I agree with you on the point above, but in practice we want to avoid annoying corner cases and more specifically we want to avoid wasting another year like we did when designing the HTTP/1.1->WS handshake :-) Cheers, Willy
Received on Monday, 20 August 2012 08:59:52 UTC