Fault contracts versus AcknowledgeType+Message

I wonder why have you chosen to return errors from the service layer to the client apps via AcknowledgeType + Message rather than custom fault contracts?  How could this approach address multi-lingual support?

Understood, thanks.

Michael:

I guess custom fault contracts could have been used.  
However, it is a technique that requires a fairly deep understanding of WCF and our focus is really on teaching common design patterns. 

As another example, we could have added extensive error handling (as we would in a production system), but that would also have diluted the learning experience for developers who are focused on learning pattern techniques (rather than error handling). 

I hope this helps.