HTTP binding for WSDL 1.2

In the HTTP (not SOAP) binding for WSDL 1.2, [1] recommends when the
HTTP method is POST,

(a)     If there are > 1 uncited input message parts, then the document
root of the request body is an element whose local name and XML
namespace is the operation {name} and {target namespace}, respectively.
Each of the uncited input message parts are serialized as children.

(b)    If there are == 1 uncited input message part, the document root
is the single part, and there is no 'wrapper' based on the operation.

(c)     If there are == 0 uncited input message parts, the HTTP body is
empty.

 

The same rules apply to all output message parts in the HTTP response
body (independent of whether the request method was GET or POST).

 

The specification and implementations would be simpler if the document
root was always constructed from the operation {name} and {target
namespace} when there are > 0 uncited input message parts. That is, 

(a)     If there are > 0 uncited input message parts, then the document
root of the request body is an element whose local name and XML
namespace is the operation ....

(b)    (cut)

(c)     If there are == 0 uncited input message parts, ...

 

Any concerns if we make this simplification?

 

--Jeff

 

[1] http://lists.w3.org/Archives/Public/www-ws-desc/2003Mar/0068.html