Bart Massey, 2013-06-20

1. Build a move struct with a row and column index.

2. Your board will be a R x C array of "boolean". I will assume that this array is zero-based in this pseudocode. Initially, let R = 3, C = 4.

3. Write a function/method to initialize a board to all true.

4. Write a function to print a board (in ASCII, because graphics code takes forever). I used o and . as the board characters.

5. Write a function to make a move gotten from a human. Use whatever syntax is easy and makes you comfortable. Your function should return a valid Move struct. Don't overdo the error checking and beauty at this stage. My function took seven lines.

6. Write a function to make a move r, c on a board, by changing all squares that are at or below both r and c to false.

7. OK, here comes the interesting part...

    negamax(b, d):
      rx <- 0
      cx <- 0
      for r in 0..R-1
        for c in 0..C-1
          if b[r][c] ^ (r > 0 v c > 0)
            rx <- r
            cx <- c
            b' <- copy b
            makeMove(b', r, c)
            if negamax(b', d + 1)
              continue
            if d = 0:
              makeMove(b, r, c)
              print the move r, c
              print that the program is winning
            return true
      if d = 0 --- no winning moves, so just pick a loser
        makeMove(b, rx, cx)
        print the move rx, cx
        if rx = 0 ^ cx = 0
          print that the program has lost
        else
          print that the program is losing
      return false
   

8. Now you just need to write a main program that alternately calls negamax() and your human move() function until someone loses.