Re: [hybi] workability (or otherwise) of HTTP upgrade

On 7 December 2010 11:53, Mark Nottingham <mnot@mnot.net> wrote:
> I fully agree we need to unblock this discussion and ship a protocol.
> I'm trying to understand why people are digging in their heels on a
> design that's supposed to be helping server deployment, when
> AFAICT the server folks are telling them it's not workable.


>From my point of view, the only things in play that are not workable
are the unframed bytes in the current draft and the bogus hosts in
adam/erics proposal.

But from a server implementers point of view, both CONNECT and
GET+Upgrade+Hello are workable handshakes, so I'd really like to see a
draft based on either of them, but with the unframed bytes and space
encoding strangeness removed, maybe also with the robust framing
(invert FIN bit) and Hello frames used.      Either one has hurdles to
overcome, but they can hardly be worse than what we currently have.

My preference is to come up with a draft based on GET+Upgrade+Hell,
but the browser vendors appear more included to go with a CONNECT
bases solution, so if that's what we need to do to get rid of the
deployed -76 implementations, then I think it is a good next step.
The -76 implementations violate HTTP spec to a much greater degree
than the expressed concerns about CONNECT


cheers

Received on Tuesday, 7 December 2010 11:09:49 UTC