[km]

that's neat, but it would be much more interesting if the terrain was randomly generated at the start. as it is, we'll only see a few good solutions that are local maxima. the mutation function miiight be able to kick it out of that, if there is a true maxima for this terrain.

still, version 2 needs random terrain so we can see it evolve unique solutions. probably wont happen though, this looks like someone's final project for an evolutionary algorithms class (though i would give this project an A)
_________________

vi) RPGs (Role-Playing Games)
For adolescents; half-formed personalities roaming (in packs) in search of identity.