Hi,

I have a very general question that I hope someone can help me to answer.

The basic question is what type of trust needs to be established between a publisher and subscriber.

Let's imagine that we are on a system that is set up to "accept" requests from random outside servers but is not set up to "initiate" requests to random outside servers. I.e., a system behind a strict firewall.

It seems to me that if the publisher is a secure server, then it will have a hard time triggering the onMessage method of a random subscribers. The reason I think this is because I guess that the onMessage must be making an Remote Method Invokation on the untrusted subscriber system.

So, basically, can you have a publisher that is behind a strict firewall?