Yesterday I took the train to Rehovot and visited the Weizmann Institute. I gave a talk in a local seminar and was hosted by David Harel's research group. An interesting work they are doing is on behavioral programming, which is a formal language independent model to specify reactive applications. They have been experimented with games, robots and toy helicopter. They are actually specifying event processing systems; I see a possible benefit in their approach that their formal model is a basis for validation which in event based systems is tricky, due to the temporal nature. I'll work with them on some use cases from our universe to learn more about their model. Can be interesting.
1 comment:
Hi Opher, please tell Harel to enhance Behavioral Programming by allowing b-threads to use composite event expressions to specify the "situations" they "request"/"wait"/"block" ;-) That would increase expressiveness of Behavioral Programming avoiding, for instance, the need to instantiate the same b-thread multiple times to "monitor" a "situation", as is done in the tic-tac-toe example for the SquareTaken b-thread that requires a separated instance to monitor each square...
Post a Comment