bimandre@icarus.cs.kuleuven.ac.be (Andre Marien) (01/30/91)
Recently, I saw a nice piece of code in a program used to benchmark prolog systems. I am not going to copy it here, just give a similar piece of code. (no, this is not about benchmarks, although that would be interesting) (no, this is not artificially created) p(T) :- T =.. [F|AL], length(AL,AR), q(F/AR,AL,R) . q(+/2,[X,Y],R) :- R is X + Y . q(+/1,[R],R) . q(*/2,[X,Y],R) :- R is X * Y . (I may have the order of the arguments to length wrong; the reason I found this beauty was that I was curious what length/2 was used for) As long as people use prolog this way when they try it out (which was NOT the case for the programer who wrote this, I am afraid) prolog has no chance of being accepted by the rest of computer science. Suggestions for remedies are invited. Andre' Marien bimandre@cs.kuleuven.ac.be