I have a question related to faults in wsdl 2.0. As I understand now that any operation could have infault and outfault, I need to know the reason for this as there is an attribute named messagelabel that can specify whether it is an in or out. Also I don't understand the reason for the different fault propagation ruleset as my limited mind can't see any practical uses.
Practical use of fault propagation may be as follows:
System A sends request to system B which pass through a mediating system of some kind.
When system B generates a fault as the result of a request, it may be of interest to the client from which the request originated to receive the fault.
Thus the mediating system may be configured to propagate any faults from system B to system A, from which the request originated.
Hope this at least shows an example of when fault propagation can be useful.