shokodemon

registered for this! this sounds like a good idea. my exams will be done around the end of july - i will read up about markov chains then and we can take a look at it. let's study the random gen code together and see what we can come up with. as to training, we'll see. maybe a half -trained markov is better than a non-trained random gen - and if the maps it makes are good, they can be fed to it in return.