This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.
WS-Eventing section 3.1 specifies wse:Identifier: "In some cases, it is convenient for all EPRs issued by a single event source to address a single Web service and use a reference parameter to distinguish among the active subscriptions. For convenience in this common situation, this specification defines a global element, wse:Identifier of type xs:anyURI, that MAY be used as a distinguishing reference parameter if desired by the event source." Three issues may be caused by removing <wse:SubscriptionManager> in <wse:SubscriptionEnd> per Issue (http://www.w3.org/Bugs/Public/show_bug.cgi?id=6397): 1) It may force a subscriber to create its own id and correlate it with the wse:Identifier returned by the event source. 2) It eliminates a common case where wse:Identifier alone can be used to correlate SubscriptionEnd and subscriptions; 3) It creates backward compatibility issue when implementations use wse:Identifier in SubscribeEnd message; Proposal: Instead of removing <wse:SubscriptionManager> from SubscriptionEnd, move it to the WS-Addressing [source] EPR of SubscriptionEnd message. This 1) preserves the wse:Identifier and 2) is consistent with the modeling of SubscriptionManager. The new message template looks like this: <s:Envelope > <s:Header > <wsa:From> <wse:SubscriptionManager> endpoint-reference </wse:SubscriptionManager> </wsa:From>? <wsa:Action> http://schemas.xmlsoap.org/ws/2004/08/eventing/SubscriptionEnd </wsa:Action> </s:Header> <s:Body > <wse:SubscriptionEnd > <wse:Status> [ http://schemas.xmlsoap.org/ws/2004/08/eventing/DeliveryFailure | http://schemas.xmlsoap.org/ws/2004/08/eventing/SourceShuttingDown | http://schemas.xmlsoap.org/ws/2004/08/eventing/SourceCancelling ] </wse:Status> <wse:Reason xml:lang="language identifier" >xs:string</wse:Reason> ? </wse:SubscriptionEnd> </s:Body> </s:Envelope>
All but 1 of wg on 2009-01-15 agreed to not open this issue as it was thought to be a re-opening of 6397 without enough new information to support reconsideration. http://www.w3.org/2009/01/15-ws-ra-irc#T21-42-02