Marc Goodner was amongst the first to comment on SSDL; he makes some good observations. But first… after Marc‘s suggestion, all the documents are now also available in HTML.
SSDL is indeed related to MEST. SSDL‘s design is based on the ‘message is the truth’ principle that governs MEST. More on this in the MEST paper which – now that SSDL is out – is our next step (we’ll be looking for the community’s valuable feedback soon). Please note that we are not advocating ownership of the ideas in the space; not at all. There have been many many smarter people than us who talked about this already. What we will be trying to do is to use a language similar to that used to describe REST in order to allow us to make informed comparisons between the different styles (REST, pub/sub, MEST, distributed objects, etc.).
The careful reader may notice that since WS-Addressing currently requires the wsa:Action information header to appear in all messages, we have a ssdl:msgref/@action attribute in the ssdl:msgref element. The value of the attribute is a URI which is used as the value of wsa:Action. If the attribute is omitted, ‘urn:ssdl:ProcessMessage’ is assumed 🙂
This doesn’t mean, of course, that there is an operation called ‘ProcessMessage’ at the ultimate receiver of the SOAP message. In the MEST and SSDL worlds there is no such ‘operation’ abstraction. Since WS-Addressing requires wsa:Action as the way to identify the semantics of the message, we thought it’d be a good idea to introduce a well-known URI that carries the generic ‘ProcessMessage’ semantics.
BTW… We do not treat SOAP as a transport protocol. SOAP is our transfer protocol. We do recognise the fact that SOAP uses underlying transport/application protocols for its transport requirements.
And a final word about Marc‘s comment on whether SSDL will make it as a specification/standard or whether the industry will adopt it. Please remember that this is an investigation into ideas; it’s a research mini-project. If someone from the industry thinks that this is a good idea, we’d be delighted. That hasn’t been the motivation though.
One response to “HTML versions available and SSDL’s relationship to MEST”