- From: Martin Thomson <martin.thomson@gmail.com>
- Date: Mon, 30 Jun 2014 09:49:06 -0700
- To: Daniel Sommermann <dcsommer@fb.com>
- Cc: Yutaka Hirano <yhirano@google.com>, Jeff Pinner <jpinner@twitter.com>, "K.Morgan@iaea.org" <K.Morgan@iaea.org>, HTTP Working Group <ietf-http-wg@w3.org>, "C.Brunhuber@iaea.org" <C.Brunhuber@iaea.org>
On 30 June 2014 09:43, Daniel Sommermann <dcsommer@fb.com> wrote: > What's the right order to tackle this issue? Can we remove END_SEGMENT and > then patch the websockets proposal, or do we need to have a story for > websockets first? I think that we have to resolve the issue, either: a) A new protocol (like websockets) can be layered on top of/integrated with HTTP/2 without changing the application negotiation token, as long as the properties of the protocol are compatible. That is, intermediaries that support HTTP/2 can carry websockets. b) A new protocol needs a new application negotiation token.
Received on Monday, 30 June 2014 16:49:33 UTC