One has to be extremely generous when talking about complexity of the analog computers. Any argument with "friction" in it implies (as far as I see) that analog computer do not work (complexity wise) at all. It's like saying that IEEE number are not reals and so you cannot do any computation with reals on a computer (very true and at the same time total rubbish). One IMO more valid critique Fred's method would be "how do you pick the two cities in O(1)"? But then I rather marvel at those thingies and enjoy... Best, jj * Warren D Smith <warren.wds@gmail.com> [Sep 14. 2014 15:02]:
Your machine asymptotically does NOT exhibit constant time "lookup" to solve an A-B shortest path problem, that is where you are confused. The mass you are pulling on and the friction both grow with N...
So anyhow as I tried to demonstrate this machine really is worse than conventional algorithms...
_______________________________________________ math-fun mailing list math-fun@mailman.xmission.com https://mailman.xmission.com/cgi-bin/mailman/listinfo/math-fun