W3C home > Mailing lists > Public > xml-dist-app@w3.org > April 2006

Re: Draft Simplified SOAP One-way MEP

From: <noah_mendelsohn@us.ibm.com>
Date: Tue, 25 Apr 2006 17:58:17 -0400
To: "David Orchard" <dorchard@bea.com>
Cc: xml-dist-app@w3.org
Message-ID: <OFED52B302.3F6F7F56-ON8525715B.0077DF17-8525715B.00793243@lotus.com>

I finally got to take a look at these.  I'm afraid they're a bit too 
simple for my taste.  In particular, I can't quite convince myself that 
they normatively require anyone to send anything, though certainly a 
sympathetic reader would get the idea.  At the very least, I would give 
the instructions in prose, e.g.:

The scope of a one-way MEP is limited to the exchange of a message between 
one sending and one receiving SOAP node.  The sending node MUST send the 
SOAP Message provided in 
http://www.w3.org/2003/05/soap/mep/OutboundMessage to the node identified 
as http://www.w3.org/2003/05/soap/mep/ImmediateDestination.  The sender is 
not responsible for reliably detecting whether transmission succeeds or 
fails, but the sender SHOULD fault in a binding specific manner if it 
descovers that transmission is in fact unsuccessful. 

The receiving node MUST determine whether a given message has been 
received successfully, and if so, MUST process the received message in 
http://www.w3.org/2003/05/soap/mep/InboundMessage according to the (2.6 
SOAP processing model).  Determination of success by the receiver MAY be 
conservative, I.e. the receiver may in exceptional circumstances treat as 
erroneous or lost  a message which is received intact (typical reasons for 
making such decisions might include shortage of buffer space, network 
interface overruns, etc.).  Receivers MAY fault in a binding-specific 
manner if some particular message is declared in error (note, however, 
that in many cases where receipt is unsuccessful, information identifying 
the message or its sender may be unreliable, in which case there may be 
little if any value in reflecting a message-specific fault.)

I'm not sure the above is quite right, but it makes clear I think that 
even when there are no state machines, it's important to cover the details 
and the edge cases. 

Noah Mendelsohn 
IBM Corporation
One Rogers Street
Cambridge, MA 02142

"David Orchard" <dorchard@bea.com>
Sent by: xml-dist-app-request@w3.org
03/30/06 08:04 PM
        To:     <xml-dist-app@w3.org>
        cc:     (bcc: Noah Mendelsohn/Cambridge/IBM)
        Subject:        Draft Simplified SOAP One-way MEP

I attach an HTML and xmlspec version of the SOAP One-way MEP done in 
simplified state transition-less style.  I like this much better than the 
"complex" style, done at 
[attachment "entitiesedcopy.dtd" deleted by Noah Mendelsohn/Cambridge/IBM]
[attachment "one-way-mep-simple.html" deleted by Noah 
[attachment "one-way-mep-simple.xml" deleted by Noah 
Received on Tuesday, 25 April 2006 21:58:32 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 22:01:29 UTC