Applications Messaging Standard

[Doug Tody]

On Thu, 8 Feb 2007, John Taylor wrote:

>> 
>>
>>>>   o	Be multi-protocol, with the same message content expressable
>>>> 	in at least two different wire protocols (XML/RPC, JSON,
>>>> 	OpenWIRE, STOMP, etc.).
>>> 
>>> Yes, but... err, why?
>> 
>> I addressed this already in my response to John's mail.  Aside from
>> possibly making things easier for clients, the main point is to separate
>> the semantic content of a message from the specific wire protocol used
>> for transport.  Otherwise one risks becoming locked in to a specific
>> technology, and you have something which is more an implementation than
>> a standard.
>
> I think that we're all agreed that separation of message spec from 
> infrastructure spec is a good thing.
> However, just because you _could_ send the message over an arbitrary wire 
> protocol, doesn't mean that we _have_ to include more than one wire protocol 
> as part of the standard.

True, and it might make sense for the actual standard to define only
one mandatory wire protocol which an infrastructure implementation
has to support, and which clients can assume is always available.

To ensure that the standard protocol is not tied to any one
wire protocol however, the design and probably the initial test
implementations will probably have to deal with at least two such
protocols.

 	- Doug