The user needs to be able to look at the current board state. When the user clicks a space on the board, it needs to send a signal to all observers.

Also, the board UI needs to update whenever the board state changes. It should probably be subscribed to the engine so that it can receive messages.

Tasks

• The board needs to be shown during the game.
• The user should be able to click on spaces.
• When the board state updates, the board UI needs to update. (i.e. signals)

Done

The user needs to have a working board interface which can request updates and stays up to date with the rest of the game.

There needs to be an AI level which is totally stupid. All it would have to do is select a random available space.

Note that this does NOT involve creating UI/logic to select the Idiot AI. This is just the Idiot AI's game logic.

Tasks

• Create behavior which, when asked for a move, selects an open space via random selection.

Done

When playing against the Idiot AI, the AI needs to be functional, not cause any crashes, and play according to the specifications.

It might be cool to track wins and losses.

Or maybe do a best out of three, etc.

-player 1
-player 2

If the user exits a game in progress using the 'X' button, it should save the game state to a file. Upon loading the game, it needs to check for this file. If the file is found, the user should be prompted, asking whether they want to continue the previous game.

If the user does File -> Exit, the game shouldn't save anything.

Tasks

• If the user doesn't exit using File -> Exit, the game should be serialized to a file.
• Upon loading while the save file exists, the user should be asked if the file should be loaded.
• If the user declines, the file should be deleted.
• If the user accepts, the file should be de-serialized, and then the file should be deleted.

Done

The user should be able to resume an interrupted game if they want to.

Dummied out

This might be hard, depending on how we write it. If we do a simple TXT file, sure... but that's boring.

So we'll need to figure out how to open PDFs in a cross platform way... ech.

When it's the Mastermind AI's turn, it should select the best action based on the Minimax algorithm. This AI should never play sub-optimally.

This does NOT deal with UI for selecting the Mastermind AI. This is just the Mastermind's behavior during the game.

Tasks

• On the Mastermind's turn, it need to select the best move based on the Minimax algorithm.
• As part of the Minimax algorithm, it needs to have a "static heuristic" function. This will have to be implemented and thoroughly tested.
• Write detailed description of static heuristic function for customer to see.

Done

When playing against the Mastermind AI, the AI needs to be functional, not cause any crashes, and play according to the specifications.

Either when a player wins or when no more moves are possible, the game needs to end.

The UI probably could have some kind of notification, and then prompt to start a new game, or something like that.

Tasks

• The game should check after every turn (or after 5th turn, whatever; doesn't really matter.)
• The game should not continue after a player has won.
• The user should be allowed to start a new game.
• The game should end when the board fills up.

Done

The game should stop when a player wins or there is a tie. Also, all tasks are complete and relevant tests are written.

The board will be able to resize if the window is resized.

Don't focus on this unless we have extra time.

There needs to be a manual for the end-user to show how to use the program. This shouldn't assume any technical know-how, and needs to avoid jargon. Cheesy references might be good. Depends on how cheesy.

Tasks

• The manual needs to be drafted.
• Pictures should be added so that users who cannot read can still follow along.
• This should be a regular PDF.
• We need a scratch-and-sniff panel.
• Should have a blue fairy.

Done

We have a manual which has been proofread and matches up with how the program works.

When it is the player's turn, the player should be able to select a place to put their mark on an empty space. Obviously, they shouldn't be allowed to place a mark in a non-empty space.

Tasks

• Create behavior to alert the rest of the program when it's the human player's turn.
• Create UI to allow the player to click on an empty space.
• When a taken space is selected, either the program should ignore it or a warning message should be shown.
• Upon selecting a valid space, the player's turn needs to end.

Done

On the player's turn, they should be able to make their move without borking the program.

At the beginning of the game, the user needs to be able to select the kinds of players which will be playing. Currently, we have three possibilities: Human, Idiot AI, and Mastermind AI.

One of the requirements is that the humans need to be able to specify their own names.

We need to be able to select both player's modes. We can have two humans, two AI, or any combination of the two.

Questions

• Should be have UI options to swap this during a game? If so, do AI need to be nameable?

Tasks

• We need UI to select the player mode. There needs to be options for human, idiot, and mastermind.
• For the humans, user needs to be able to specify a name.
• And, of course, the user needs to be able to start the game once they're happy with the settings.

Done

There's an elegant UI for selecting player types and names which meets all of the requirements.