This is an archived snapshot of W3C's public bugzilla bug tracker, decommissioned in April 2019. Please see the home page for more details.

Bug 6401 - Eventing: Notifications violates WS-I BP
Summary: Eventing: Notifications violates WS-I BP
Status: CLOSED REMIND
Alias: None
Product: WS-Resource Access
Classification: Unclassified
Component: Eventing (show other bugs)
Version: FPWD
Hardware: PC Windows XP
: P2 normal
Target Milestone: ---
Assignee: Doug Davis
QA Contact: notifications mailing list for WS Resource Access
URL: http://lists.w3.org/Archives/Public/p...
Whiteboard:
Keywords: hasProposal
: 6661 (view as bug list)
Depends on:
Blocks: 6430 6661 7207
  Show dependency tree / graph
 
Reported: 2009-01-14 00:05 UTC by Doug Davis
Modified: 2009-09-16 08:16 UTC (History)
2 users (show)

See Also:

Attachments

Description Doug Davis 2009-01-14 00:05:27 UTC 
WS-Eventing proposes that the event source define some output-only 
operations for the events it will send, as shown in the following example 
from the spec: 
<wsdl:portType name="Warnings" wse:EventSource="true" > 
  <wsdl:operation name="WindOp" > 
    <wsdl:output message="tns:WindMsg" /> 
  </wsdl:operation> 
</wsdl:portType> 

WS-I Basic Profile R2303 says the following: 
4.5.2 Allowed Operations 
Solicit-Response and Notification operations are not well defined by WSDL 
1.1; furthermore, WSDL 1.1 does not define bindings for them. 

R2303 A DESCRIPTION MUST NOT use Solicit-Response and Notification type 
operations in a wsdl:portType definition. 

Proposal:
Replace this with some WS-Policy that will allow an event source to 
advertise which events will can generate.
Comment 1 Robert Freund 2009-01-15 23:26:19 UTC 
2009-01-15: discussed and Gil has action to take up the output only issue at WS-I once more with feeling.
Comment 2 Robert Freund 2009-03-03 12:45:48 UTC 
2009-02-17 No member seems to be taking this up with WS-I
Comment 3 Robert Freund 2009-03-12 14:01:27 UTC 
members seem to agree that a practical solution is to shift to imput operations.
Action-48 on Gil, Wu, Dug
Comment 5 Robert Freund 2009-04-28 21:19:21 UTC 
Action-61
Comment 6 Robert Freund 2009-06-23 11:56:59 UTC 
*** Bug 6661 has been marked as a duplicate of this bug. ***
Comment 7 Robert Freund 2009-06-23 11:58:20 UTC 
2009-06-11 Issue 6661 is merged with this issue.
Text from 6661 is as follows:
Appendix I of the WS-Eventing Member Submission, "Service Metadata for
Eventing", is incomplete in a number of areas. Firstly it does not indicate how
an Event Sink is expected to derive the binding details for the Notifications.
Will the Notifications be sent as RPC/Encoded, Doc/Literal? Will the
Notifications be sent using SOAP 1.1 or SOAP 1.2? Secondly, the relationship
between the "endpoint reference for the event source" and the annotated
portType is unclear. One might think that the "endpoint reference for the event
source" is equivalent to the wsdl:port that implements a binding of the
annotated portType, but this conjecture is not supported by any text or
example.

"Service Metadata for Eventing" is also technically incorrect in that it
requires Event Sources to process a WSDL containing output-only operations in
order to generate the code necessary to correctly process and dispatch
Notification messages which is a violation of BP 1.1 R2303. While one could
argue about the relevance of WS-I profiles to infrastructure-level protocols,
it is clear that creation of Notification handlers is, in this context, an
application-level activity as the type and structure of Notifications is driven
by the application and not WS-Eventing.
Comment 13 Robert Freund 2009-09-02 08:38:13 UTC 
2009-09-01 resolved with comment 12