In a Hub-based architecture, should the hub have any concept of the meaning of messages?

In some messaging architectures the message broker is able to modify the messages it handles. This might imply at least some knowledge of what messages exist, or even their meaning. This is one possible solution to the question posed in ApplicationsMessagingInitialQuestionsEleven.

Not the semantics of messages intended for other clients. Given an extensible set of messages, it's not possible for the hub to know about all of them - so it shouldn't require to. What benfits could we gain from a more 'intelligent' messaging fabric? -- NoelWinstanley

---