A New Glider Found For Conway's Game of Life

An anonymous reader writes "Conway's Game of Life is now forty two years old, but it continues to inspire as well as being the basis of an actively researched field, with computer scientists now announcing they have found a new form of the famous 'glider' pattern (once suggested by Eric S Raymond as the insignia of computer hackers) that runs over a so-called Penrose universe."

Comments

Life enthusiast

[Apocryphon]

"Life enthusiast Adam Goucher has discovered ...."

About time! I've always found the terms pro-life and pro-choice too politicized and constraining.... I'm a life enthusiast!

not life as we know it

[Dave Whiteside]

it's not a new glider in the game of life , but a glider in the Penrose tiled universe - inspired by Conways Game of life...

the article need to be read

it is seriously cool though

Re:not life as we know it

[Thud457]

not life as we know it

I'm a doctor, not a tile layer, Jim!

Re:not life as we know it

[uigrad_2000]

I'm pretty sure I attempted a hexagonal game of life. It's the first thing people would think to try after discovering Conway's original version.

The problem with the hexagonal version is that each tile has only 6 neighbors, as opposed to 8 in Conway's version. This reduces the complexity so finding interesting patterns is a lot more difficult. The way around this is to add more states.

After reading the article, it sounds like one researcher theorized that a stable glider could not be found for the Penrose tiling, and offered $100 to anyone who did. Some other friends of his found an answer, but had to "cheat" by expanding the number of states (for a given tile) from 2 to 4.

It is kind of cool, or would be if they actually showed the 4 states and the exact rules. Since they decided to leave the technical explanation out, it's a rather uninteresting article. It's not really slashdot worthy, in my own humble opinion.

not the same

[eyenot]

They didn't prove anything except that by increasing the complexity of 'Life', they can force some kind of complex behaviour that would have been impossible for the simpler version we're all more familiar with. They changed the rules from 'alive or dead' tiles to '00 01 10 or 11' tiles. There are two different rhomboids in the Penrose tile universe they're playing in, so it seems to make sense that you will find some sufficiently complex means of navigating it if you observe two bits at once.

I think it should have been couched differently: Penrose universe NOT non-repeating, given a sufficiently complex, self-changing pattern to look for.

Re:not the same

[ceoyoyo]

You have it backwards. By increasing the complexity, namely by making the pattern of tiles non-repeating, it was believed things like gliders were impossible. That turns out not to be the case.

The existence of the glider has nothing to do with whether the universe is non-repeating or not. Penrose universes are mathematical constructs that are proven to be non-repeating.

Inspiration

[David89]

Stuff like this toinspires me everyday

As expected...

[vlm]

Its creation is an achievement because gliders were previously thought to exist only in regular cellular automata, such as the most famous one, the Game of Life

On wikipedia that would get flagged as weasel words (or the whole article deleted for non-notoriety). Who thinks gliders should only exist in regular automata? If anything my opinion is that modern automata thought was the other way around, expecting them to exist.

Note that gliders are not rare or unusual in automata. Some of the first original researchers thought that only gliders/spaceships that exist lived only in Conways GoL but further research a long time ago showed they're ridiculously commonplace in other rulesets. As seen below. So the tone of this discovery is more accurately described as "much as we suspected, but never bothered to prove, until now" rather than the stereotypical serendipitous discovery tone of "that result looks weird, WTF, who ever would have guessed"

This is separate from the penrose tile thing, which I don't follow. It might, or might not, be the case that a glider in the very specific ruleset of penrose tiles is a hard problem. But in the wide universe of all rulesets, gliders/spaceships and stuff seem very widespread. As a general rule if a ruleset is terminally boring then it definitely does not have gliders, but if its not terminally boring then almost all of them have either chaotic and/or glider-like behavior.

http://www.ics.uci.edu/~eppstein/ca/

".... I have investigated whether gliders exist in many semitotalistic rules similar to Life, where the behavior of a cell depends only on its own state and the number of live neighbors. The results show that the existence of gliders is commonplace ....."

http://uncomp.uwe.ac.uk/genaro/rule54/glidersRule54.html

".... We displayed all gliders of Rule 54 including two new glider guns (also extensible) ... "

Rule 54 has nothing to do with the famous rule 34. Well I guess there are self replicating patterns in CA rule 54 which could be interpreted as pr0n by another one dimensional cellular automata, I guess.

Re:As expected...

[sFurbo]

Gliders are commonplace on repeating grids. According to TFA (and this makes sense), it was thought that they could not be made on non-repeating grids. After all, which direction should it follow? How to make sure it can even exist in the place it will move to?

However, I feel that by allowing more types of tiles, it should be clear that it was possible. For example, with four types of cells, you could have
"front of glider" (becomes "back of glider")
"back of glider" (becomes "not glider")
"side of glider" to keep the rest in check (keeps status unless in contact with "back of glider", when it becomes "not glider")
"not glider" (becomes "side of glider" if in contact with one "side of glider" and one "front of glider", becomes "front of glider" if in contact with two "side of glider" and no "back of glider")

This seems to be what they have done.

The word you are looking for is cellular automaton

[Hentes]

This not Conway's game of life.

Re:slashdot worthy

[TaoPhoenix]

I dunno, at the rate which we aggressively avoid reading TFA's, maybe it is Slashdot worthy - someone who really is interested will deep-click to the meaty theory on the web.