FUNNYMAN2 wrote: February 11th, 2021, 9:58 am
you just test things together and see how it goes
That's not how patterns were found. (Unless that pattern was Kazyan's almost c/61)
YYDH wrote: February 11th, 2021, 3:52 am
I'm surprised at the wonderful patterns in the game of life.
How were those patterns discovered?
Are they all random discoveries? (I think not.)
What about those big and periodic patterns like OTCA metapixel?
How do we discover new patterns now?
Please let me know if you know anything about these questions.
Looking forward to your reply.
Wish you have a nice day!
PS:I'm not good at English so if some words were used inaccurately,
please don't put it on mind and I'm glad to correct.
Why are there suddenly four Chinese in this thread?
I want to attack this question from a different angle. I created a cellular automaton in Jan 2020 that is strikingly similar to Conway's Game of Life, and named it LeapLife. LeapLife is well-developed (in my opinion) and like 20% of all discoveries are discovered by myself. Methods in LeapLife cannot be directly applied to regular Life because they're done already or completely covered by another search already. But they can give inspiration or reflect how historical patterns in regular Life were found. (@goldenratio, I intentionally avoided "bluffing" it.)
0. I spend much time to learn Game of Life history.
1. The Gosper glider gun, along with the queen bee shuttle, was discovered by close observation of the R-pentomino, which evolves into a queen bee (without the blocks) and I believe Gosper tried all queen bee collision. There are 3 glider-producing collisions. One of them produces a glider that crashes into one of the queen bees, the other one produces a glider that didn't escape in time, so the only collision left is the gun now known as Gosper Glider Gun. I tried to apply this method, we've found a number of shuttles (Bookend shuttle, traffic light shuttle, Pi shuttle) but no luck making a gun. In contrast, most guns in OCA were found by rubbing two high period sparky oscillators, similar to
Period-33 glider gun. p22 and p24 glider guns were created, though no one managed to create a p19 glider gun with the p19 sparker.
2. Search program matters (and is the only practical way of, say, finding a ship with a new velocity), but they're the most useful when they first come out, because (usually )every search can only be ran once and we quickly run out of searches that can be completed in an reasonable time. (There are certain people in the Game of Life history who were crazy enough to run a search for a year, but as a middle school student I won't attempt that.) Small patterns are not necessarily known unless exhausive search is done - one such search, only partially completed (like 2/256), quickly found a new p12 spaceship. (Historically, the fireship was also discovered using an exhausive search.)
3. Soup based syntheses are low hanging fruits. They're really easy, you just take a soup, observe how did the object form, if it formed from the reaction of some small regions (LeapLife soup tends to become one large blob unlike Game of Life though) synthesise the small regions seperately. (If the spacing is tight, a one-glider-seed can be used to insert that region, see my octapole synthesis.) Unnatural syntheses is, on the other hoof, very hard that when someone completes one of them, we congratulate them many times. (That does not include still life syntheses
, they uses a method more similar to organic chemistry synthesis.) Typically, one builds a final step first as a proof of concept. Typically, the final step consists of a base still life (that should be similar to the object to be synthesised so it could be converted to it easily) and some activation gliders. Then people synthesise the base still life. I don't know how people actually attack this, because I haven't done one in LeapLife.
4. WLS/JLS (RIP wwei23) is very useful for most purposes, like aforementioned spaceship searching (it allows customizable searches like with a T shape, thus providing more variety than other programs like qfind) and unnatural synthesise (it can find reactions) and customized sparkers.
What pattern do you want to know about exactly?
