Hi there!
The RPC/Literal project I'm working on wants me to define faults that only contain 2 elements, faultcode and faulstring. Further, they want me to define these elements on my own according to a specific proprietary system. I tried this two different ways.
When I did so by having my functions throw a SOAPFaultException, I was easily able to have the system throw my own defined faultcode and faultstrings, but it also insisted on including the faultactor and a detail attribute as well.
I tried to get around this by defining a custom exception with only a faultcode and a faultstring element and having it throw this. When I did this, it included my custom faultcode and faultstring, but also included a separate faltcode (env:client) and its own faultstring (the name of the custom exception I had thrown). This was for obvious reasons not the ideal either.
Is there any way to have a function return a fault like that with JUST a faultcode and faultstring that I can define? Or any other way to have a function like that that would normally return one kind of complex datatype but would instead return the fault datatype in certain instances? Attention to this would be HIGHLY appreciated.