RE: WS-I and RPC

> As Roy says;
>
> "In order for SOAP-ng to succeed as a Web protocol, it needs to start
> behaving like it is part of the Web. That means, among other things,
> that it should stop trying to encapsulate all sorts of actions under an
> object-specific interface."
>  -- http://lists.w3.org/Archives/Public/www-tag/2002Apr/0181

I am reading this totally wrong, so please ignore what I say.

If I have an object representing purchase order and an object-specific
interface for that purchase order (get/set) I should not strive to
encapsulate all sorts of actions under an object-specific interface and
expose this object as a Web service.

(In this particular case the service would have HTTP bindings w/o SOAP, so
it would be doing HTTP GET/PUT)

Rather I should have a resource that provides a set of operations which may
access (read/write) one or more objects, but by itself does not map to any
one object. I would then let the resource touch which objects it deems
necessary based on the operation that needs to be performed, e.g.
sendMeInvoiceDetails or cancelPreviousRequest.

(And in this particular case I assume sendMeInvoiceDetails does not return
an invoice but actually sends it asynchronously over SMTP or maybe by fax)

arkin


>
> FWIW, SOAP 1.1 and 1.2 are "SOAP-ng" - but people still use them like
> they're SOAP 1.0, which isn't.  Roy's comments were more an indictment
> of current practice than the specs themselves.
>
> MB
> --
> Mark Baker.   Ottawa, Ontario, CANADA.        http://www.markbaker.ca
> Web architecture consulting, technical reports, evaluation & analysis
>