Re: CR20

Proposal 3 in my mail makes no reference to CR17 because it is not
justified on the same grounds. It simply states that (for the HTTP/SOAP
binding) the anonymous URI means what we all agree it means: that a
response message is to be sent in the payload of the HTTP response. We just
need to say so explicitly and to clarify that this definition implies that
you can only use it in the [destination] property of response messages. I'd
rather clarify what we all agree on, since it works well, than go on a
hunting expedition to find new semantics at such a late stage in the
process.

Paco




                                                                                                                                        
                      David Hull                                                                                                        
                      <dmh@tibco.com>          To:       Francisco Curbera/Watson/IBM@IBMUS                                             
                                               cc:       public-ws-addressing@w3.org                                                    
                      02/06/2006 06:19         Subject:  Re: CR20                                                                       
                      PM                                                                                                                
                                                                                                                                        




Francisco Curbera wrote:

      1. Restrict the defaulting of To to the anonymous URI to response
      messages
      only. This is fully consistent with the resolution of CR17, that
      restricts
      the defaulting of ReplyTo (to the "anonymous EPR") to request
      messages.
      Advantages: consistent approach to the use of defaults for optimizing
      the
      synch request response patter, but leaving other potential issues
      unmodified.

The proposal 3 currently on the table builds on the same notion of
consistency, but I don't think this notion necessarily follows in either
case.  In the case of reply-to: we limited the handling of anonymous to
request messages because reply-to: only has a natural meaning at all in
request messages.  That's not to say that one couldn't define further
meanings, just that there's no single obvious way to do so.

>From the resolutions of CR17 and i067/i068/cr 15, it seems pretty clear
that all we're saying is that an anonymous response endpoint occurs in the
context of a SOAP request-response MEP, it means to use the response
message of the MEP.  We don't disallow it in other circumstances (whatever
those may be).  Note also that we handle SOAP 1.1/HTTP and SOAP 1.2
/everything separately.

The most consistent way to extend this to [destination] would be to state
what anonymous [destination] does mean in a given set of circumstances.
There doesn't seem to be any controversy that for a response message,
anonymous [destination] would mean use the response message of a request
response MEP.  Saying this doesn't disallow other meanings elsewhere, any
more than talking about reply-to: in requests disallowed anything else.

Beyond that, we might say that anonymous [destination] for requests means
the destination URL of a SOAP 1.1/HTTP request, or that anonymous
[destination] for any SOAP 1.2 message is the ImmediateDestination property
if it's defined, or we might leave it undefined.

Whatever we say, to be consistent with the resolutions we've already
decided, we should handle SOAP 1.1/HTTP and SOAP 1.2/everything separately.

Received on Tuesday, 7 February 2006 14:30:20 UTC