Re: [hybi] Multiplexing in WebSocket

[Ian Hickson <ian at hixie.ch>]

On Tue, 27 Oct 2009, Greg Wilkins wrote:
> Ian Hickson wrote:
> >> I think the IETF should be considering a bidirectional solution for 
> >> all web clients - not just for HTML5 JS clients
> > 
> > What's wrong with TCP, BEEP, or Jabber?
> 
> Once upon a time there was the internet and IP was God (even though Al 
> Gore invented the internet and he really should have been God).
> 
> But then somebody invented the Web (was it Bill Gates?) and now port 80 
> is God.
> 
> The Web runs on top of the Internet, but restricts all the wonderful 
> features of IP to a one way port 80.
> 
> TCP, BEEP, or Jabber are really great Internet protocols, but they don't 
> generally work on the web.  Jabber (aka XMPP) has to fall back to BOSH 
> on HTTP in order to span the web.

I'm not really sure what you mean by "the web" here.

If your requirement is for a protocol that can tunnel other bydirectional 
protocols over HTTP, then WebSocket isn't it. We should decouple the 
requirements for WebSockets (which were listed in one of the first 
messages to this list) from the requirements for a tunnel-over-HTTP 
protocol, because they aren't compatible requirements.

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'