Game of Life: Features of This Implementation

The parent of this page is: Conway's Game of Life: a Java Applet with Autodetection of Patterns.

There are several other implementations of cellular automata (and Conway's Game of Life in particular) available in the Web (see references). So, why providing another one? There are a number of reasons to mention:

The present is a Java applet that auto-detects patterns (oscillators, gliders, and active patterns), reports their numbers, color-codes them (resulting in beautiful visual displays), and adapts its running behavior according to the patterns it detects.
It comes with the complete source code available, and an explananation of its data structures and algorithms. Feel free to get the source programs and modify them in any way you wish, or get ideas and make your own, smarter implementation!
It is not restricted to Conway's rules, but runs any other cellular automaton. You may define your own rules, or select from a list of proposed interesting "universes" of cellular automata.
Works in an "infinite" space, provided the number of patterns does not exhaust your computer's memory (and it takes millions of patterns for this to happen). The program does not store the space explicitly only the locations and sizes of patterns.
Provides basic pattern-design tools: insert gliders or other known patterns anywhere you wish, force symmetry by axis or by point (central symmetry), reflect, rotate, or translate patterns by given amount, and more. [This part is not implemented yet]
It is the only applet that combines all the above features, to the best of my knowledge.

OK, what about speed?

The features listed above do not come without a price. In particular, the ability to auto-detect patterns makes the present not the fastest possible implementation. If all you want is raw speed, you may take a look at some other programs listed in the references (specifically, I recommend Alan Hensel's superfast applet). However, I consider this program to be the first step towards writing a fully automated explorer of "universes" of cellular automata, so the emphasis was on the detection of patterns. On the other hand, this applet is not a terribly slow implementation it's faster than some other web-applets anyway: try running "Acorn" (select it from the "open file" button: ), which is easy to hand-type in other applications if they do not already include this simple 7-cell pattern, and compare its speed; or, try "Line puffer", and notice how the speed of this program remains nearly constant, in spite of the profusion of generated patterns, because it works only in active areas (you'll see them marked in red color, as opposed to magenta gliders and multi-colored oscillators). Many other programs slow down with "Line puffer"-type of inputs because they update the entire expanding space at each step.

Why the colors cannot be changed?

Colors in this program are not random but play an important role: they reveal the nature of the patterns that have been detected by the program. Thus,

Gliders are displayed in

magenta

Still lives (oscillators of period 1) in

deep blue

Oscillators of period 2 in

cyan

Oscillators of period 3 in

green

Oscillators of period 4 in

yellow

Oscillators of periods between 5 and 16 in various

hues of yellow

and

orange

Any other active patterns are displayed in

red

Hence, changing the colors of patterns would generally interfere with the visual display, so color-setting is not included as an option.

The above display shows a number of oscillators with periods from 2 to 16, arranged rectangularly at the border, and color-coded after their period has been detected. There is also a star-gun in the center, producing gliders (in magenta color) that manage to slip through the oscillators without hitting them. The "numbers" (in blue color) show the period of the oscillator to their right, and are oscillators themselves (of period 1, often called "still lives"). The above figure is a fully functional instance of the applet, and you can pause it and run it just like other instances of the applet in the rest of these pages.

Why was this applet created?

This program was created to address several issues:

To make a first step towards writing a fully-automated and "intelligent" explorer of universes of cellular automata.
To learn the Java programming language.
To explore the questions of emergence of complex and unpredictable behavior out of a simple set of deterministic rules.
For the sheer fun of it! (This was probably the most compelling reason!)

Back to the main page