WS-Reliability vs. WS-ReliableConversation
Apart from some thoughts regarding loose coupling, David Chappell talks a bit about the need for reliable messaging over SOAP. I think he has a pretty strange view in one respect: He makes it sound as if it was Sun's and the other WS-Reliability authors' mistake that we now have two specs. In my view, it's just plain stupid that IBM and MS did not just express their problems with WS-Reliability and jointly turn it into a decent spec. But maybe I'm just dreaming ...