The good question is, having implemented something like this, let me explain how this works:
Connections
A live web application is not really live - it just received a persistent request; which means that it still works exactly the same as the “normal” Rails application, except that clients do not close the connection (therefore, why are you interested in opening another port).
- , . , Rails ( )
"chat", . ( Ajax) -
, Rails ActionDispatch,
ActionController:: Live, , . (, resque Redis), .
, - , "" "" Rails-. Rails - , , - nodeJS socket.io ( , chat.yourapp.com), Rails
- , . , , , , , "" . , , .
, Pusher
-, Rails Redis ( )
, Rails (http://yourapp.com/chat), Rails (http://yourapp.com/chat/send) ( )