Re: [core] New Version: CoAP with Alternative Transports

[Bill Silverajan <[e-mail]>]

Hi Matthias,

 >
> Before I forget, here a comment on the URI format for alternative transports that arose during the Vancouver meeting:
> Since we all saw some interesting ways to encode the transport, I think we should make sure that we can actually encode such URIs in CoAP. Luckily, the Proxy-Uri option is a string and can hold any format, but we also have the separate Proxy-Scheme and Uri-* options...
>

Thanks, it's a very good point regarding current (and any future) option 
encoding.

We're taking care to ensure the syntax and encoding for the alternative 
transport URIs conform with RFC 3986 in terms of the scheme, host/port, 
path and query components. Eg in our draft -03, the last paragraphs of 
sections 2.2 and 2.3 highlight the encoding issues.

All the options you mentioned (except uri-port) are strings without 
applying percent-encoding. When composing and decomposing the URI 
to/from options, the algorithms in sections 6.4 and 6.5 of core-coap-18 
would also be helpful as a litmus test to determine which URIs for 
alternative transports we should settle upon.

Regards,
Bill