Ouch. Perl has to compete against Common Lisp for a number crunching exercise? That's going to make it
tough. Plus the test seems way under specified. They don't mention how many iterations to run the algorithm. They don't mention if compile times are included ('cause you could certainly create a common lisp macro to do all the work at compile time). Oh, and you can pretty much guarentee that C will win the grand prize. Or am I misreading it? Are there three grand prizes? Overall the contest seems like quite a mess.