> Well, the code isn't really the goal -- it a concrete implementation of a 
> goal. A goal is more abstract, and usually allows many ways to implement it.

I think that if you're "thinking big" like FONC/VPRI are tasked to do, then the 
ideal is to be able to write down those abstract goals (and only those goals) 
and let the computer work out the best way to accomplish them. The trick is to 
build a system that's self-aware enough to be able to do that while also being 
trustworthy enough to be allowed to do it. :)

IMO, you can kind of see this thinking in the OMeta example of parsing IP 
packets simply by providing an ASCII diagram of a packet. That is seriously 
cool goal-specified programming right there where the only "program" is a 
picture of the goal itself! That's even better than listing out a bunch of 
goals in textural form, IMO.

l8r
Sean


_______________________________________________
fonc mailing list
[email protected]
http://vpri.org/mailman/listinfo/fonc

Reply via email to