Well, the question is how the audio is getting to the user. As you said, WebPd requires FF4's audio API. If you want the actual Pd running on a server then you need a way of hearing the output. In my own pd-on-the-web project, PuréeData, I'll be using an mp3 or ogg stream. Latency will be an issue, but hopefully not a tremendously bad one.If i wanted to use Pd to generate sound for a web application, i'd have to keep an instance of Pd running on the server continuously, right? And keep all the patches i need open? Which means that if any patch crashes Pd there'll be no sounds anymore?
——t3db0t