Your favorite OCA?

[iddi01]

Even though this is about OCA (other cellular automata), it's a chitchat topic, so i think it belongs better in here.

When posting (other than replying):
1. Post your favorite OCA, along with the reason, and (if possible) a link to its discussion thread.
2. Post any number of other ones that you like, and again with reason and preferably link.
3. (optional) Post how much you like Conway's game of life compared to OCA, and the reason.

Keep in mind that you might not have met your real favorite yet.
The later you post, the less likely you're gonna regret posting whatever that is as your favorite.
Posting here before trying at least 25 different OCA is generally not recommended.

Alright, here's my post (edit: changed the list of favorites yet again):
------------------------------------------------------
My top 3 favorite OCA:
#1 SlowLifeB0 (or some alternate name):
I like this mostly due to oscillators being very versatile, but also due to it having constructional potential, although the potential is uncomparable to...
#2 B3-n/S1e2-a3-e4e:
It's so full of useful reactions involving the t-ship that it's easier to construct stuff in it than any other 2-state rule. Also apgsearch runs quickly on it.
#3 B34e5-a/S234qtw:
It has four highly notable patterns with each one being enough to characterize a whole other rule.
#4 Seems like the only rule that has a Chinese name:
It's full of common high period oscillators, with a P240 being the most common object!

Opinion on CGoL:
I do like CGoL, especially before knowing OCA.
However after knowing many OCA i became increasingly un-interested about it, mostly because naturally-occuring oscillators and especially spaceships are quite un-versatile, but also because making a new "not useless" discovery is extremely hard.

[b-engine]

Hash1, Hash2, Honeybomb Life and B2in3-q4w/S23.

Hash1 and Hash2 have both Life like (or grounded Life like) and "Bosco-like" dynamics, while having a quite strange neighborhood (I haven't seen any other rule using hash neighborhood). Also, Hash2 has over thousands or even millions of unextensible still lifes. Also, both of them supports at least 1 type of agar.

Honeybomb Life is the first rule I've investigated in forums. It also have a big and high period oscillator which is obviously pretty common.

B2in3-q4w/S23 have very chaotic dynamics, yet not being explosive, means that soups typically last longer than in Life. It also have the most potential of containing an orthogonoid (I'm going to engineer a splitter using the traffic bomb reaction).

EDIT:
There're now 2 rules climbing onto my "favorite CA list".

B-Univ is the most complex CA I've ever made. It has hundreds of transitions (but not comparable to JvN29's thousands of transitions, which may be the most complex CA ever made) to allow it's universality.
All Bliptile technologies and patterns works here, as well as newer construction and transistors technologies. A universal computer constructor (B-UCC for the first one) is also possible.
(The poll shows that B-Univ is ranked second)

Ignorance supports all B3/S23This is my preference of calling standard Life.patterns, as well as neo-spartan technology using catalysts of opposite color, especially blue block and "block in pond", allowing new smaller oscillators and possibly spaceships. Many alien oscillators are found using nbsearch2a, which also insanely found a p19 and p43.

[iddi01]

Hash1, Hash2, Honeybomb Life and B2in3-q4w/S23.

In my first post:

Post your favorite OCA, along with the reason

If you don't add the reason, nobody will know how exactly you like those.

edit: That's right!

[H. H. P. M. P. Cole]

I have been mucking around in OCA for about one and a half years, way before I joined the forums. I don't have a favourite OCA per se, as different people have their own preferences in how to create a rule (since the rulespace is so large).

My favourite rule is Factorio, which is not surprising. I like it for a huge number of reasons. Keep in mind that almost all of these are subjective.

1. The rulestring is short, contrary to most INT CA explored these days.
2. The rule definition is also simple, and can be made even simpler by knowing that it is a totalistic CA.
3. There is a huge number of small objects (for some definition of small). I have stated in different threads that Factorio has 40 4-cell objects and 102 objects that have one 4x4 phase.
4. There are a surprising number of mathematical properties. For example, Width-1 and width-2 soups keep the same width, and R3,C2,S2,B3,NW0000000000000000010001110111000100000000000000000 and B3/S2V can be emulated.
5. The number four occurs many times in the rule and its objects.
- The smallest objects have four cells, and there are 40 (4x10) of them. Sadly these objects group in nine groups of three p2s, one group of six SLs, and one lonely c/1o, which is also the only known 1D pattern.
- there are four 1D infinite SLs
- there are four known width-2 still-lifes
- rows and columns in a 4x4 box can be permuted provided the bounding box does not expand in both permuted and original versions
- Durnak plasma is defined as an aperiodic agar/soup with density approximately 1/4, which has the property that it is extremely long-living.

This is a dealbreaker for me. Why? Because it's basically all I want in a CA, and even more, in one package.

EDIT (18/2/24): My second favourite rule is SoManyShips2, as it is figuratively and literally a rulegolfer's haven. I find that it is very amazing that such a rule, with so many ships, guns, linear growths, and technology even exists.

[Haycat2009]

1. B3/S2-i34cw has many common ash objects, including very large still lifes, 2 spaceships and many higher-period oscillators

2. Factorio has many interesting mathematical properties. Who does not like rules with interesting properties? The FWKnightship reaction is helping to prove omniperiodicity.

3. Grounded Life has a reburnable fuse that can be used for signal manipulation.

4. Lake Life is the most chaotic rule that I have ever seen that does not explode. B lasts 7700++ gens.

5. Determination has good tech and math proofs - a rare combination in an explosive rule.

[iddi01]

Actually, this topic is kind of a poll, and results are put in User:Iddi01/Most_popular_OCA.

edit: The threshold of 10 people has been reached! The results are out!

[HerscheltheHerschel]

Definitely Factorio, for all the reasons listed above.

[TYCF]

Can you choose a multi-state ruletable rule?

[confocaloid]

Can you choose a multi-state ruletable rule?

Why not? There are many interesting multistate and/or ruletable rules. (Generations rules are multistate too)

[tommyaweosme]

b-univ probablyviewtopic.php?f=11&t=6375
reason: its really cool and can do basically anything (its also probably turing complete)

[b-engine]

b-univ probably
reason: its really cool and can do basically anything (its also probably turing complete)

Yes. It's already Turing complete because Bliptile is and B-Univ simulates it. Several explicit examples of W110 simulator in Bliptile works in B-Univ. In other words, it's basically just Bliptile with constructional capabilities.
It's in the list of my favorite OCA too, even before it's explicitly made. I just forgot to mention it.

[tommyaweosme]

Yes. It's already Turing complete because Bliptile is and B-Univ simulates it. Several explicit examples of W110 simulator in Bliptile works in B-Univ. In other words, it's basically just Bliptile with constructional capabilities.
It's in the list of my favorite OCA too, even before it's explicitly made. I just forgot to mention it.

well i was gonna put omniperiodicworld but i realized that b-univ is very cooler then an omniperiodic rule that was made to be omniperiodic.

[b-engine]

well i was gonna put omniperiodicworld but i realized that b-univ is very cooler then an omniperiodic rule that was made to be omniperiodic.

Well, you may create (and label it as your favourite) a universal CA based on any CA known, like Codd or JvN29. You can use as may states as you want.
EDIT: You may modify the transitions or add states into OmniperiodicWorld, and see if it becomes universal (i.e. supporting Turing machine, W110 or cyclic tag systems).

[breaker's glider gun]

wire12allpermutations (and all permutations of that, like wire1allpermutations) because it gives you all the tools for making something awesome.

For example, I easily made this solid-state bit with an "on" and "off" input:

Code: Select all

x = 16, y = 10, rule = Wire1allpermutations
9A.4E2A$9.F$9.I.3EJS$10.I3$9J.4EJS$9.O$9.I.3E2A$10.I!

it can obviously made with "toggle" instead.
Or two could be made into a flipflop, like this:

Code: Select all

x = 20, y = 10, rule = Wire1allpermutations
10.I4.EI$9.I.3EI2.E2A$9.O6.I$9J.2EI2.I$13.2I$13.2I$9A.2EI2.I$9.F6.I$9.
I.3EI2.E2A$10.I4.EI!

One could also make arbitrary-length signals with it:

Code: Select all

x = 28, y = 10, rule = Wire1allpermutations
16.3E.3E$16.E.E.E.E$.9A.6E.3E.I$A9.F12.2EJS$A9.I.12E$A10.I$A$A$A$.27A
!

[TYCF]

(Maybe) StateInvestigator because it has many useful cell states.viewtopic.php?f=11&t=4009
Maybe LifeHistory or LifeSuper because it is useful and commonly used.

[iddi01]

[bump]

Note that recently, B3-n/S1e2-a3-e4e is quickly climbing the leaderboard of my favorite rules.

It is already my second favorite, and will likely become the first in the future.

I have already posted three posts mentioning it already, and the thread is already out!

[/bump]

[Mathemagician314]

I like B2-ac3-akr/S1c2eik3-eir4r8, because it has a lot of cool oscillators, spaceships (including a camelship), and a failed replicator.

I also like B3/S2-a358 because it's far enough away from Life to have a lot of unique objects but still retains some of the fun dynamics from Life.

[confocaloid]

I thought about mentioning some rules here, but eventually decided against it, because the whole idea feels too close to a "popularity contest", which is not really what the Life/CA thing is about.

(If one really wants to, then it is not too hard to understand which cellular automata / patterns / ideas someone here considers interesting, simply by looking through their forum posts.)

I think some of CA rules already mentioned here are not really noteworthy, besides the fact that a few people really wanted to make them "popular" for some reason. For example sometimes a rule has a more interesting variant, which was explored in depth. Being popular is different from being interesting or notable for some reason.

Why is it so hard to make a rule become famous nowadays?

I'd say if someone wants their rule to become "famous", they are probably missing the whole idea of the hobby. Most really interesting rules were never "famous" and will never become "famous", but of course that doesn't in any way prevent on-topic discussion of those rules / sharing discoveries and ideas.

[d/dx]

Langton's Loops.

I will not elaborate.

[eRroR_6o6]

Current favorite: tlife just because it's tlife

also 2nd place is B2cin3-i4eiz5i/S124a (probably because i (combined with apgsearch) did a lot on it(the link to the topic)
(can't really decide on 3rd or any other places)

Keep in mind this might change in the future.

[tommyaweosme]

#4 Seems like the only rule that has a Chinese name:

EDIT: i forgot the stupid quote

i think you forgot about 起泡蘇打水

EDIT: fixed typo

thats ok because its an obscure rule

[TYCF]

EDIT: i forgot the stupid quote

i think you forgot about 起泡蘇打水

thats ok because its an obscure rule

Did you mean User:Tommyaweosme/sparkling soda? (There was a typo)

[iddi01]

(Deleted a few hours ago due to wanting my 100th post to be important, but i'd better bring it back now)

i think you forgot about 起泡蘇打水

This thread was created before your rule, so it was the only rule with a Chinese name.

Also, no offense but could you stop creating user pages for your (non-notable) rules? They clutter up this page.

[confocaloid]

They clutter up this page.

That userpage says "No requirement of completeness or quality is made". Nothing prevents simply omitting links to "low-effort" userpages.

That said, a forum post in an existing thread is a better way to share some new CA, compared to going to the wiki and putting stuff there directly. Most CA aren't notable.