Programming the Internet


I spoke at the BYU Computer Science colloquium today. The title of my talk was Internet Programming. I borrowed quite heavily from some of Rohit Khare's work and some of my own thoughts in that area. As an example that many people can relate to, I also included a discussion of Clemen Vasters' Content Pipeline idea. Here is a copy of my slides. I'm not entirely happy with the presentation yet. You still get a lot of blank stares from some folks. Even so I had some great discussion afterwards that produced the following ideas:

  • Kent Seamons and I talked about the use of the trust negotiation work that he and his students are doing as a way of creating trust between parties involved in a service exchange.
  • Kevin Seppi and I talked about using Baysian techniques to analyze these kinds of systems for performance.
  • Kelly Flanagan brought up the idea of using an ALIN device to duplicate a transaction stream and using one stream to drive the production service and the other stream to drive a QA service or one in development. You could even have a process on the ALIN device that compares the result streams and produces a report of the differences in content and performance.

In short, almost anything you'd develop or measure on an operating system, you could do on ALIN devices. That leaves a lot of room for creative thinking.