2048

This project is aimed at crating a lightweight command-line implementation of the 2048 game (play2048.co) in the C language. The ultimate goal is to run the program on the TI-89 graphical calculator using TIGCC (tigcc.ticalc.org).

Building

For the TI-89

You can download the required tools from https://tigcc.ticalc.org. Under linux, the binaries of the latest beta (http://tigcc.ticalc.org/linux/tigcc_bin.tar.bz2) are used. The setup is as follows:

wget http://tigcc.ticalc.org/linux/tigcc_bin.tar.bz2 # Download the archive
mkdir tigcc; cd tigcc  # Create a director to unpack the archive and change to it
tar -xjf ../tigcc_bin.tar.gz  # You guess it
export TIGCC=$PWD # Set the TIGCC environment variable. This is required by the tigcc executable to locate other tools
cd /path/to/2048  # Change to the 2048 source directory

Then, in the source directory, the compiler can be invoked as:

$TIGCC/bin/tigcc -std=c99 -O2 run2048.c game.c main.c mem.c ui.c

Note that the output file will be named after the first source file on the command line. Thus, if the compiler is invoked as shown, the resulting TI-89 executable will be run2048.89z. This file can then be deployed to the calculator (or an emulator).

TIGCC also supports other calculators via the USE_V200 and USE_TI92PLUS preprocessor macros. Thus, if you would like to compile for these, simply invoke the compiler with the appropriate -D flag. Notice, however, that they are not officially supported and the experience may be degraded, for example, due to the different screen resolution.

Other platforms

Building 2048 is pretty straightforward, only CMake, make and a compiler toolchain are required.

cmake
make

This should produce the 2048 executable.

Internal representation of the grid

The data type used to represent the grid is essentially a unsigned int **. For a n×n grid, we have a n*n-sized array of unsigned ints with pointers to the first elements of each row. To give an example, for a 3×3 grid, this boils down to the following:

      grid
       ↓
      ┌──┬──┬──┐
      │  │  │  │
      └──┴──┴──┘
       ╷  ╷  ╷   
 ┌─────┘  │  └─────┐
 ↓        ↓        ↓
┌──┬──┬──┬──┬──┬──┬──┬──┬──┐
│  │  │  │  │  │  │  │  │  │
└──┴──┴──┴──┴──┴──┴──┴──┴──┘

Binary logarithms as values

Note that, in contrast to most other implementations, here the binary logarithms of the values are used. That means:

The lowest value is 1.
When merging 2 tiles with equal values, the value of the resulting is the value of the previous tiles incremented instead of the added values.
A value of 0 corresponds to an empty tile.