Hexothello, a Speed Hack 2001 entry.

  The game is like Othello (Reversi), but played on a hexagonal grid.
  More moves are actually possible on a regular Othello board, but I
  didn't realise until after I started coding.

  The game is written in the Scheme language (a Lisp dialect), and
  runs in Runt, my Scheme interpreter.  That's why it's so slow.

  The AI is a very simple rule-based one, and only looks ahead one
  move by default.  But it's already slow enough.  This is my first
  attempt at an "AI", so don't expect to learn anything from it.  If
  you want to replace it, it's in `game.scm'.

  Everything was created during the Hack, except for:
    - the interpreter
    - a large part of the Allegro bindings
    - sort.scm, some sorting routines which came from SLIB
    - the samples (from www.sounddogs.com and other places)

Compiling

  - gcc to compile (other compilers will NOT work)
  - Linux or Windows, perhaps djgpp
  - 32 bit machine
  - top-down machine stack (e.g. Intel)
  - a Pentium-class machine, or faster machine when using the AI
    (development machine was a Pentium 233 MHz laptop)

  If you are using Mingw32 or djgpp, run "make MINGW32=1".  Unix
  people can just run "make", as it is the default.

  The Scheme interpreter may be unstable when compiled with djgpp, but
  I'm not sure.  I haven't tried under BeOS yet.

  If you are using Allegro 3.9.34 WIP or later under Linux, you may
  get strange crashes.  This is an Allegro problem.  Linking with the
  debug version or disabling assembly code in Allegro will help.

Notes

  Edit `init.scm' to change some things, such as the video mode.
  Changing to 8 bit video will require a lot of work though.

  You must run the executable from the current directory.  I'm too
  lazy to fix this.

Authors

  Peter Wang - Code and graphics
  David Wang - AI ideas

-- 
http://www.alphalink.com.au/~tjaden/ - tjaden@users.sourceforge.net