Random number generation
From Maze
The RND function in 10 PRINT can be considered from a variety of perspectives:
- as a technical affordance of the BASIC language whose function can be understood
- as one out of a set of strategies for pseudo-random number generation, both in general and specific to platforms like the Commodore 64
- ...
Different conceptions of randomness:
- ramdomness as uniform distribution over all possibilities
- randomness as kolmogorov complexity of a sequence
- randomness as human perception (which is much closer to the kolmogorov complexity model than uniform distribution
How do pseudo-random number generators work? Examples of different random number generators, including the simple Galois linear feedback shift register generator used in the Ian's 6502 program and Michael's path program. Different pragmatic tests of randomness used to qualify the strength of different random number generators.
The RNG as a stand-in for complex decision making in art generators, game AI.
Examples of RNGs failing to be random enough. Examples of expressive computing would be particularly good.
Impact of RNG on the aesthetics of 10-print. Do weak RNGs with small cycle times (like an 8 bit Galois LFSR) change human perception of the pattern? What about strong RNGs with long sequences (a bunch of the same character in row - which should actually happen when sampling a uniform distribution and tends not to happen with weaker RNGs)?
What RNG is the c64 using? Historical reasons for using that one? Any discussions in the hobbyist computer scene about positive and negative properties of the built-in RNG?
