Re: Moving forward with HTTP/2.0: proposed charter

On 03/08/2012, at 3:52 PM, James M Snell <> wrote:
> I want to just clarify that the proposed language does leave the room open for potentially non-backwards compatible changes to be made within HTTP 2.0. Specifically, the way I read this, it should be possible for us to tunnel a 1.1 message through a 2.0 request, as well as reasonably translate a 2.0 message into a 1.1, but the charter also allows for the introduction of mechanisms within 2.0 that do not necessarily translate directly into 1.1 so long as those are clearly spelled out within the specification (In particular I'm thinking of things like the binary optimized header encoding, UTF-8 support, etc). I see nothing in the draft spec language that rules non-backwards compatible changes as being out of scope. If that's correct, then I'm +1 on the revised charter.

It does not prohibit optimisations, but the combination of the requirements for supporting current use cases, current extensions and 1.x<->2.0 does constrain things quite a bit. 

Any such scheme is going to have to balance the concerns and gain consensus; e.g., telling people that new headers have to be defined twice (as HTTP/1.x and HTTP/2.x) probably isn't going to fly, because of the burdens that places on new headers as well as intermediaries.


Mark Nottingham

Received on Friday, 3 August 2012 21:05:43 UTC