By the way, I think that the "seminar" that you guys had where you studied the climate models and coded them up as interactive browser software, for educational purposes, was great! I realize that we're having a lull in programming activity at this point, so de facto it is de-emphasized in today's Azimuth strategy, but I'm still holding on the hopes that, with some fresh blood, we may eventually be able to start another round of that seminar.
What I found when writing Petri Net Programming was that I _could_, without too much trouble, write a simulator for stochastic Petri nets, which uses the exponential distribution, etc., to figure out the next firing time, and I could "expain" how the code works, but there would be something fundamentally arbitrary about it, because it wouldn't have real meaning without the theory that is presupposed by such a simulator. That's why I think that to really get programmers interested in these projects, it will help to have an organized "educational package" for them, which I have started on, in a small way, in the blog articles that I have been writing.
If we had money, I would also recommend hiring an educational specialist to organize curriculum for specific audiences. John's vision of an ecosystem of people teaching science to each other holds a lot of sway for me, and we already have that, to some extent. But we're now only tapping into a small portion of the potential there. It would be nice if it just happened spontaneously, but some enzymes (staff) could really help it along!