Re: [hybi] requirement draft as wg item

Greg Wilkins wrote:
> Mike Belshe wrote:
>   
>> * Requirement 11 refers to a "sub-protocol", but "sub-protocol" is not
>> defined. I am not meaning to be difficult, but I am not sure what does
>> or does not qualify as as sub-protocol.
>>     
>
> I agree and I also don't yet understand exactly what we all mean by
> subprotocol
>
> I'm not even sure there is a requirement for subprotocols.  It might
> just be a nice to have feature of a particular solution.
>   
It allows validation for applications, just like normal version numbers.

If you have an iPhone client talking quake/1.0 packets to a server, but 
the server is actually using an incompatible quake/2.0 packet syntax, 
it's useful to address that issue upfront as part of the handshake. 
(Plus allowing easy rejection of a quake/1.0 client trying to talk with 
an xmpp/2.0 server with a reasonable error message.)

You could always force the application to validate its own version in a 
separate handshake, but it's a universal need so has value here.

The "sub-protocol" really means the full protocol stack. "sub-" just 
meaning it's built on top of WebSockets, i.e. that the full protocol is 
one of the possible protocols that could be designed on top of WebSockets.

-- Scott
>
> regards
> _______________________________________________
> hybi mailing list
> hybi@ietf.org
> https://www.ietf.org/mailman/listinfo/hybi
>
>
>
>   