> But -- and this is the crucial point -- that doesn't mean we shouldn't strive to be better than we are now.

Agreed! And I also agree that we are really making progress. But we're far from a world where people can crank out provably correct code (let alone the proofs).

> I'm not sure why these mundane observations receive such attention in computer science. Maybe because with software the model and the system look so similar.

Excellent point, and yes, I think that is the problem. Modeling the world in code is not much different from... er... modeling the world in code :-)