Single Comment

To Greg Strong and Fergus Duniho: I have not awaited, that CRC would cause that quick your experiments to create those starting positions. Because of that I have not yet published a procedure, which would show how Smirf managed that problem and how to standardize the number to position relation. It obviously is not well known, that it has been me, who invented and introduced the numbering schema for Chess960. Beside of that numeric relation I have worked out a two table look up scheme which is fitting on to one single page to be used in chess clubs to have all Chess960 starting positions immediately at hands. For CRC a solution is of course not that easy. But if you are interested in that, I could post code snippets, where those positions are created from a given number. When Smirf is starting, it quickly creates an array of valid position numbers by probing all 48,000 and filtering the valid into an array of some more than 20,000 entries, what overmore allows to index those array with a short int, which is good for some pseudo random number generators, selecting later a randomized valid starting position. Another solution could be to use 6000 char constants, where the mass of 48,000 bits could be used to encode the precalculated validity of the appropriate number.