The plan is:
- Done: Get the HTML and CSS set up so that a board is represented
- Done: Hook into the board with some basic JavaScript and get the game board represented in code
- Done: Get the game logic stood up so that it can be played with two players
- Done: Refactor existing code to support AI features
- Ability to make plays on the game board then roll them back without affecting the view
- Ability to add and remove the click event handlers so that the player cannot affect the game board during the computer's turn
- Done: Get a basic first-pass "AI" stood up. I don't claim to understand all the intricacies of this game, so the first-pass will probably be incredibly easy to beat.
- In progress: Refine the AI so that it plays a little better. This may be tricky, as I'll have to develop some insights into the game as I go along.
- Technically, this goal is only complete when I have a set of rules that collectively lead to the game being solved. The game is currently in a fairly decent state, and should present a challenge to most human players.
What I will NOT be doing:
- I won't be focusing on the interface so much. This is more about getting the logic and functionality in place, as that is the challenge I am deliberately focusing on. I will, however, make it so the page is usable on mobile as well as desktop.
Things that I would like to do better compared to Tic-Tac-Toe:
- In Tic-Tac-Toe, I had places in code where I referred to cells on the game board by ID (0 through 8), while in other places, I used coordinate pairs (that is, [0,0] through [2,2]). While it is trivial to convert between the two, I would like to minimize the number of conversions. That is, I should be more thoughtful about the interface between the two "coordinate systems".
- I would like to structure my code better, mainly by being more modular (being more intentional about dependencies).
Special thanks to Nick for repeatedly defeating my scripted computer player and giving feedback.