Re: [hybi] Extensibility mechanisms?

[Greg Wilkins <[e-mail]>]

Ian Hickson wrote:

> For all intents and purposes, the server is everything under control of 
> the end point authority. 

Even the smallest deployment will have to deal with
ISPs, Data centres or even home routers, which are not under
the authority of the application developer.


> The spec explicitly
> states that conformance requirements are essentially "black box" 
> conformance requirements. It doesn't require the server to use a single 
> processor; the TLS, Web Socket, and subprotocol handshakes could all be 
> handled by completely different machines on different coasts, so long as 
> the end result is indistinguishable from what the spec requires.

End points are always going to be assembled from
multiple "boxes" black, grey and white, probably
from different vendors, that interoperate using
open standards.    These "boxes" may just be language
libraries, or ISP provided firewalls or full
corporate DMZ layers.

If interoperability is to be achieved, the standard
cannot end at the first box.

So back to the original objection I raised.  If
an intermediary sees a HTTP connection upgraded to
websocket, then it should be able to make the
reasonable assumption that protocol it will
subseqently see will be that described in the
websocket specification.   Allowing arbitrary
framing by sub protocols will break that
reasonable assumption.


regards