In the sound branch, I'm adding a new broadcaster for pygame.mixer. In the current setup, all broadcasters pull from the same output_queue from the demo that is currently running. That queue has no way of differentiating strings that are passed into it. This leaves the individual broadcaster to parse through the message and determine if it is an item to "broadcast".
Ideally, we should have alternate ways for a user creating a demo to control which broadcaster is receiving its info.