This issue of WSDL MEP to SOAP MEP mapping for errors was analyzed in great details by the invaluable Dave Orchard (then at BEA) in 2005: http://www.pacificspirit.com/blog/2005/04/05/underlying_protocol_is_a_completely_leaky_abstraction I have not read anything more complete on this topic. Unfortunately, there are more questions than good practices in Dave's article. We are in 2008 and the problem is still confusing. What should we do, us, poor user souls ? In the end, we shall all REST :-) -- Jacques Talbot ________________________________ De : public-soap-jms-request@w3.org [mailto:public-soap-jms-request@w3.org] De la part de Eric Johnson Hi Mark, Perhaps I'm missing something here? One aspect of the one-way SOAP MEP is that you *cannot* return a fault. Mind you, if you're using WSDL with an MEP that has an input message and fault message, wouldn't an application map that two a two-way SOAP-MEP, or use WS-Addressing to indicate a return address? -Eric. Mark R Maxey wrote: I don't fully understand the abstract SOAP MEPs and how it applies to this document. Can someone explain how a receiver of a one-way SOAP/JMS message can return a fault back to the sender when the JMSReplyTo is required to not be set? Thanks, Mark Maxey
Received on Friday, 8 August 2008 07:51:01 UTC