In a client-server application, both the clients and severs, who wish to receive/send redis messages, will be redis clients? And they will always send data to other redis clients through the redis server; the intermediary hop is unavoidable?
Would it be useful, or good practice, to ever write a client-server application whereby the application server logic was IN the redis server? Or to put it another way, would an application server ever embed a redis server object, thereby allowing it to communicate with its clients? My guess is this could be a poor decision.
Thanks for the assistance.