DEEEP was a prototype of modular, peer-to-peer informal learning software. It used multicast networking to automatically discover other instances of the application on the local network. The participants could then collaborate in real-time to explore topics such as Moiré patterns and L-systems.