- From: <paul.downey@bt.com>
- Date: Mon, 17 Oct 2005 23:05:50 +0100
- To: <katy_warr@uk.ibm.com>, <public-ws-addressing@w3.org>
I was unclear the impact of adopting either proposal would have on our CR status - would we have to go back to Last Call? Also having recently talked with people tasked with securing Web services who worry about SOAPAction != wsa:Action != the 'action' being authorised, I'm inclined to keep the Status Quo. -----Original Message----- From: public-ws-addressing-request@w3.org on behalf of Katy Warr Sent: Mon 10/17/2005 10:17 PM To: public-ws-addressing@w3.org Subject: Issue i065 : Conflict between default action pattern and SOAPAction Here is a summary of what was said on the call re i065. Apologies if I missed of any points, but I'm sure someone can put me straight :o) Katy ISSUE: wsa:Action MUST equal SOAPAction but this is not possible in all cases (i.e. when wsa:Action is being gen'd by default Action pattern). PROPOSAL 1: In the absence of <wsaw:Action>, use SOAPAction in preference to defaultAction pattern. Pros + Keeps SOAPAction and wsa:Action the same which might be considered better architecturally Concerns: - Are implementations using the relative URIs for SOAPAction? (To investigate) - Makes the specification for generating wsa:Action more complicated PROPOSAL 2: In absence of <wsaw:Action>, use default Action Pattern (as stated currently). Relax the restriction that Action MUST equal SOAPAction (but recommend that they SHOULD equal) Pros: + Simpler pattern for wsa:Action generation Concerns: - Makes implementation more complex if the Action and SOAPAction differ (for example, which to dispatch on)
Received on Monday, 17 October 2005 22:05:59 UTC