So, to follow up on this, I’m going to announce the 2015 tournament in early August. Everything will be the same except for the following:
Random-length rounds rather than fixed length
Single elimination instead of round-robin elimination
More tooling (QuickCheck-based test suite to make it easier to test bots, and some other things)
Edit: I am also debating whether to make the number of available simulations per round fixed rather than relying on a timer.
I also played around with a version in which bots could view each other’s abstract syntax tree (represented as a GADT), but I figured that writing bots in Haskell was already enough of a trivial inconvenience for people without involving a special DSL, so I dropped that line of experimentation.
So, to follow up on this, I’m going to announce the 2015 tournament in early August. Everything will be the same except for the following:
Random-length rounds rather than fixed length
Single elimination instead of round-robin elimination
More tooling (QuickCheck-based test suite to make it easier to test bots, and some other things)
Edit: I am also debating whether to make the number of available simulations per round fixed rather than relying on a timer.
I also played around with a version in which bots could view each other’s abstract syntax tree (represented as a GADT), but I figured that writing bots in Haskell was already enough of a trivial inconvenience for people without involving a special DSL, so I dropped that line of experimentation.