W3C home > Mailing lists > Public > public-ws-addressing@w3.org > January 2006

Metadata elements, understanding and targeted processing

From: Cahill, Conor P <conor.p.cahill@intel.com>
Date: Sun, 22 Jan 2006 12:25:06 -0800
Message-ID: <8AEB79DC01BE994D8DE3FD02FA5B475B029F16D2@orsmsx409>
To: <public-ws-addressing@w3.org>
In implementing a WS-A compliant messaging handler, I ran across an
issue with the interprestion of data within the Metadata element which
looks like it is not clearly defined by WS-A.
Elements within the <Metadata> element of an EPR may have targeted
entities who are supposed to process them and the issuer probably should
have the ability to document whether or not said targeted entity MUST
understand their processing requirements.  This, to me, is similar to
the S:actor and S:mustUnderstand  attributes defined by SOAP for SOAP
headers (although, obviously these are not headers while they are in the
Metadata element).
The <Metadata> child elements are similar to Soap headers from the point
of view of the entity that processes (dereferences) the EPR and while
WS-A doesn't (rightly so) define what the processing rules are for those
child elements, WS-A should provide a mechanism which the issuer of the
EPR can use to indicate to the targeted recipient(s) (including
intermediaries) who the issuer expects to to process the element and
whether or not the processing party must understand said element.
Without such capabilities, WS-A probably should make a statement that
any processor of the EPR MUST understand the entire content of the
<Metadata> section.
Received on Sunday, 22 January 2006 20:25:15 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:04:12 UTC