OK--I think I understand now. My concern was not about the initial racks or predictability, but the limit on "independent" iterations of a simulation. In fact, for the statistical simulations that I run (unrelated to quackle, I'm afraid), reproducibility is key, so you have to make sure you know or can retrieve the seed at every step (presumably not an issue for quackle except a handful of academics). As long as quackle is drawing a new seed at the start of every iteration based on a nearly random cpu state, depending on the time in milliseconds taken to compute various things, with unknown other processes running, the period for simulation iterations should be near infinity and you would never have a situation where running 10^z iterations is the same as running a sim with 10^(z-1) iterations ten times in a row and stacking them, right?
On 9/12/07, Graham Toal <[EMAIL PROTECTED]> wrote: <snip> Resetting the seed to an 8-bit value before running each sim would be > disasterous, > for example. You'ld only ever simulate 256 games. That doesn't > happen in practise > because you don't ever reset the seed. <snip> > However from the point of view of bias, all this is irrelevant. It > only would matter > if quackle were playing against a human for something valuable (money or > reputation) and someone went to the bother of mounting a cryptographic > timing attack against it.
