@@ -14,10 +14,6 @@ well. planned improvements: - make the markov chain use probabilities, instead of just recording all - possible states and then picking one randomly. although this seems to work - decently already. - use the data gathered from the input level to build a level that is of different dimensions, like, say, WAY bigger. this could be interesting to actually play, although hacking into an emulator of the original game is README 374e834b9f3ce49d7fa7e3bf2b74346fd016c6f7 Nick Welch mack@incise.org http://github.com/mackstann/markovio/commit/a74a0c217520d131cb443248910c0212c760fef7 a74a0c217520d131cb443248910c0212c760fef7 2009-01-08T20:01:09-08:00 2009-01-08T20:01:09-08:00 actually it already does that 93f157c4ae4483b062f77a737479b66eb5b0e656 Nick Welch mack@incise.org