I'm using speech dispatcher to make the house talk. I think it's using the festival backend, but I'm not sure. My code is here. It listens for HTTP POSTs and plays speech and music. Right now it only announces when wireless devices enter and leave the network, but there's more we'd like it to say. If you haven't been on the network for a while, and we have a song on file for your device, you'll get welcome music as well as a spoken introduction.

I wish I didn't have to write websound.py. There should be better sound servers out there that can do good timing and mixing for this kind of application. Maybe pd would be good. Sometimes the speech silently fails to work. I don't know what's going on yet.