It would also be very interesting to hear from David on machine representations of scientific theories and their conflicts. It seems pretty clear that Special Relativity, for example, could have emerged considerably sooner had there been available a way to more systematically identify and examine assumptions, alternatives and consequences. What’s less clear to me is whether the same is true of current research into a) fundamental laws, b) emergent behavior.

This culminated in one of the first models of CSP which was incorporated in a programming language as the rendezvous construct of Ada.
http://en.wikipedia.org/wiki/Communicating_sequential_processes

With Ada, one can take any deterministic Petri Net and with very little effort, write the semantics directly as rendezvous behavior between tasks. A concurrent task or thread model is necessary as the foundation to any Petri Net implementation. Since tasks are built into the language, along with a built-in synchronization mechanism, that’s why Ada works as a good model. Say what you want about the language, but this was a case of some computer scientists who actually wanted to see if they could directly implement some theoretical concepts into a mainstream language.

BTW, in the Unified Modeling Language (UML), Petri Nets are often referred to as Activity Diagrams. These are the brainchild of Grady Booch, who started off as an Ada guru, and spun those off from Booch Diagrams.

I guess my point is that there is a rich history on working Petri Nets into software programming paradigms. Occam is also an early concurrency model, which may have predated Ada.

The other point of entry is to study the history behind languages with co-routines and threads. Lots of these have library support and it wasn’t directly built into the language.

Also consider languages with guarded Horn clauses such as have been implemented with concurrent versions of Prolog or with the Actor model.
http://en.wikipedia.org/wiki/Actor_model_later_history
A horn clause is basically a trigger for firing a Petri Net transaction.

I am personally doing lots with Prolog these days, which is a good match for the Semantic Web.

But John sees it right. The Azimuth Project is a (the!) serious c21st internet academy, where we should meet as real persons.

So you can pick up a pattern of avoiding using spaces if they’re only for aesthetics. This isn’t a good reason for choosing web-account names, but it is a reason.