The current version of sniffle differs fundamentally from the project that Dr. Littman and I initially envisioned. One can trace the evolution of our thinking in my project log, which is included in an appendix. A more succinct summary of our early thinking can be found in our earliest project description, available at http://www.cs.princeton.edu/~tpkelly/webrc.html.
We began with an needlessly narrow vision of a simple, practical solution to a particular problem. Many of Dr. Littman's lab demonstrations involve delicate arrangements of cumbersome devices. It is inconvenient to disassemble and reassemble the demonstrations for one-time showings, e.g. to classes of students. It is also impractical to crowd a large class into a small laboratory in order to present a demonstration. Dr. Littman would like to set up his experimental apparatus in his lab and show it to audiences anywhere without much fuss. To achieve this goal we decided to use two independent systems, one to control an I/O board which in turn would control the actual lab demo, and another to handle video feedback.
It gradually dawned on us that what we really wanted was a general-purpose tool that would allow us to interact with an I/O board in a PC at the lab from anywhere in the world. We realized that we could wrap more general capabilities into the system, e.g. the ability to communicate with a digital oscilloscope via a serial connection. This was the insight that turned a one-time throwaway ``gee-whiz'' demonstration project into an extensible, reusable remote control system.