Conway's Game Of Life

This is just yet another implementation of Conway's Game Of Life written in Typescript.

It primarily started as a little excercise with Javascript Generator using a functional approach.

Concepts

• Cell is a [number, number] tuple denoting coordinates on the grid.
• Neighbors is a function with parameter cell: Cell calculating the neighbors of the given cell.
• Generation consists of all cells: Cell[] which are alive, and its epoch: number.
• Generations is a Generator yielding Generations either for direct iteration with for...of, or for sequential consumption with its next() function. If you do not want to continue, i.e. do not care about any values, optionally pass false to next().
• World is a factory function with parameters cells: Cell[] for the initial Generation and epochMax?: number which creates and returns the corresponding Generations.
• Game is a factory function with parameters Neighbors which creates and returns a World.

The defining parameter of a Game is the way how neighbors for a Cell are calculated by means of a Neighbors function. Most importantly it indirectly determines the available coordinates which a Cell can occupy, i.e. if the grid of the Game is infinite, finite (but unbounded) or bounded.

There are helper factory functions creating World functions with the some default Neighbors functions. The default Neighbors calculate Neighbors for a Cell in the literal sense, i.e. the Cell[] ((usually of length 8)) directly adjacent to the given Cell:

• bounded with paramater max: Cell defining the maximum allowed x and y coordinates. The minimum [0, 0] is implicit. The game has hard (unconnected) bounds out ouf which no Cell can exist. This behavior is best conceptualized as rectangle or square.

• finite with paramater max: Cell defining the maximum allowed x and y coordinates. The minimum [0, 0] is implicit. The game has connected bounds. A cell that would spawn outside any of these bounds (top, right, bottom, left) will spawn on the opposite side of the grid. This behavior is best conceptualized as a torus, (or) like a level in Pacman .

• infinite with no parameters. There are no boulds. Keep in mind that this especially means that a Cell can have negative coordinates. This behavior is best conceptualized as an infinite plane.

Typescript

This package is build with Typescript and comes with its own type definitions. Hense this project uses javascript generators only available since ES2015. So in order order to use within a Typescript based project make sure to set at least "target": "es2015" in your tsconfig.json.

{
  "compilerOptions": {
    "target": "es2015"
    // ... and more
  }
}

Basic usage example

// Typescript example with "module": "commonjs" in tsconfig.json

import { games, Cell } from '@limpirium/game-of-life';
// const world = games.bounded([10, 10]); // or
// const world = games.finite([10, 10]); // or
const world = games.infinite();
const glider: Cell[] = [
  [2, 1],
  [3, 2],
  [1, 3],
  [2, 3],
  [3, 3],
];

// direct iteration
let generations = world(
  glider,
  1000, // set a maximum of 1000 epochs or iteration will probably run forever
);
for (let generation of generations) {
  console.log(generation);
}

// manual, sequential access
generations = world(glider);
console.log(generations.next()); // -> {value: {cells: [...], epoch: 0}, done: false}
console.log(generations.next()); // -> {value: {cells: [...], epoch: 1}, done: false}
console.log(generations.next()); // -> {value: {cells: [...], epoch: 2}, done: false}
// to stop pass `false` to `next()`
console.log(generations.next(false)); // -> {value: {cells: [...], epoch: 3}, done: true}
console.log(generations.next()); // -> {value: undefined, done: true}

Todos / next steps

• check in browser
• write some tests
• stop generation if there are no changes from one generation to the other
• provide a library of some famous initial Cell[]

Ideas

• possibility for custom birth/death conditions
• custom Neighbors functions and explore how this effects the Game Of Life. Haven't tried it or checked for research myself yet, but may be interesting 🤔 or maybe not 🤷