Someone mentioned in a comment over at http://www.25lines.com that Math.random() should be outlawed.
At first I didn’t agree, but there’s truth in the fact that there are much more interesting patterns to be gained by abandoning random and getting unpredictable coordinates otherwise. Here I’m using a Pickover strange attractor.
Hmm, I didn’t agree with abandoning Math.random() either, but seeing this makes me reconsider. I’m going to have to revisit the bit-101 strange attractor posts, which at the time were a bit over my head. Thanks Keith.
While visualisation of these attractors can be very attractive (e.g., http://www.complexification.net/gallery/machines/henonPhaseDeep/ ) I cannot see how they can be applied as generally or abundantly as a source of pseudo-random numbers. I agree that random() should not be used as prolifically and arbitrarily as it is to generate variety etc., but it is still fundamental to many interesting, especially dynamical, processes (e.g., initialising turing patterns.)
hmm, it’s either white noise, or it’s pink noise, and white noise is very very unusual in nature, if it even exists – I guess a white noise could be for example intergalactic background radiation for example? would that qualify? If Math.random() supposedly should qualify as white noise – then all you need to do is to do is take Math.random()-Math.random() and you get something totally different as opposed to white noise, Right? An attractor is something else, isn’t it? an attractor is random space? and sure, in a way more natural, but it is not more random? is it? it is a defined random, and a defined random are in my opinion always more beautiful. like the sounds of the waves or the sound of a waterfall… but it doesn’t beat random. – random is random and nature is nature.