Thread for basic non-CGOL questions

[tommyaweosme]

i assume empty emptiness means the oscillator-like things it leaves, like plow world but slightly different.

[Empty Emptiness]

Welcome to the forums!

Why does B3/S0145678 have so many oscillators?

The question is a bit unclear; what are you asking about? Random soups grow without leaving any well-separated oscillators. On the other hand, cellular automata such as Conway's Life (B3/S23) have infinitely many distinct oscillators.

Code: Select all

x = 10, y = 10, rule = B3/S0145678
bobob2ob2o$ob4o3bo$bobo2b4o$obobo3bo$bob2o4bo$ob2o2b3o$ob2o2bob2o$bob
4o2bo$bo2b5o$2o5bobo!

Code: Select all

x = 47, y = 11, rule = B3/S0145678
25bob2obo$16bo8b6o$16bo6b10o$7b2o6b3o6b8o7bo3bo$2bo4b2o5b5o6b6o6b3o4b
2o$2o5b2o5b5o4bob6obo4b3o6bo$7b2o6b3o7b6o5b3o$16bo7b8o$16bo6b10o$25b6o
$25bob2obo!

[Empty Emptiness]

i assume empty emptiness means the oscillator-like things it leaves, like plow world but slightly different.

nope

[Banananananan]

Difference between Corderoid and cordership?
Also: can a b1 rule not explode?(any b1, isotropic, or generations)

[confocaloid]

Difference between Corderoid and cordership?

A Cordership is a spaceship that moves diagonally with simplified speed c/12 (the actual period is higher), and is based on the switch engine from Conway's Life.
An example of a Cordership is the 2-engine Cordership. Not counting Life (B3/S23), there are also known Corderships in B3/S236e. By allowing three or more cellstates, or by allowing higher-range weighted neighbourhoods, it's possible to find other cellular automata with Corderships.

A corderoid is any engineered spaceship that is entirely made out of interacting instances of some moving engine. There might be several instances of a puffer engine, arranged and phased in such a way that their interaction removes all the ash and the pattern becomes a spaceship. Instead of a puffer engine, there could be an unstable object that would not continue to move by itself.

[b-engine]

Difference between Corderoid and cordership?
Also: can a b1 rule not explode?(any b1, isotropic, or generations)

1. Corderoids are a family of spaceships that consists of stabilized puffers or unstable "moving" objects. Corderships are a subset of corderoid that consists of corder engines.
2. There's at least a non-explosive B1e rule with soups that stabilize into photons, or occasionally oscillator or even a gun:

Code: Select all

# You can find this in Request a Rule to be "Tamed" thread
#R B1e2cek3cjkry4jrw5cny6ck7c/S2kn3-aejr4knqty5ejk6e
!
#C [[ RANDOMIZE ]] 

EDIT by dvgrn: This post used the "xrule" short form, but as of build 1147, LifeViewer no longer supports this as a valid RLE header. Would suggest switching to the "#R" syntax that I've edited in here.

[tommyaweosme]

me have a new format for yall to check out!

Code: Select all

r=(rule)
(rle)
c (settings)

everything is optional (except rle)
if rule is emited, its life
if settings is emited, nothing happens
all problem am solve! (just need update to lifeiewer for that happen)

[hotdogPi]

Do you have shares in a competitor to Golly or something?

[tommyaweosme]

r= is much easier to type

you know what, just r

so rlife and rb3s35 are rule referring

and czoom16 is [[ ZOOM 16 ]]

edit: i JUST NOTICED that empty emptiness's b3s0145678 kills still lifes that are in regular life. aka, things that die in one generation in this rule are still lifes in life. take advantage and find small sparks. convert them into life, and get big still life.

maybe useful for heshel condits

[confocaloid]

edit: i JUST NOTICED that empty emptiness's b3s0145678 kills still lifes that are in regular life. aka, things that die in one generation in this rule are still lifes in life. take advantage and find small sparks. convert them into life, and get big still life.

maybe useful for heshel condits

As written, this sounds like a shortcut that would allow finding big still lives without actually searching for big still lives, which would indeed be truly remarkable. Unfortunately as far as I can tell that doesn't really make the problem easier; how do you find things that die in one generation in B3/S0145678?

[tommyaweosme]

by using a spark searcher. its easy to find sparks, more than still lifes

[confocaloid]

by using a spark searcher. its easy to find sparks, more than still lifes

It's not too hard to find still lives using software. For example, I just found a large still life in Life using Logic Life Search:

Code: Select all

x = 20, y = 20, rule = B3/S23
bo3b2o3bob2ob2ob2o$obo2bobo2b2obobo3bo$bobobobo5bo2b2o$3bobob2o4bo3b2o
$2bobo5b2obobo3bo$2bobob4obob2ob4o$b2obo2bo8bo$o2bobo6b3o3b2o$bobob2o
4bo2b2o3bo$2obo3bo2bobo5bo$o2bo3bo3bo5bo$2b2obob4o4b3obo$4b2o4bobobo3b
2o$3o3b2o3bob2obo$o2b2obo2bobo3b2ob2o$3bo2bob2o2b2o3bobo$4b3o3bobo4bo$
b3o3b2obobobo3bo$bo2bo2bobo3b2o2b2o$3b2o3bo!

# ./lls -b 20 20 -s p1 -p '>150'
Getting search pattern...
Done

Preprocessing...
Done

Width: 22
Height: 22
Duration: 2

Number of undetermined cells: 400
Number of variables: 3791
Number of clauses: 124166

x = 22, y = 22, rule = B3/S23
bbbbbbbbbbbbbbbbbbbbbb$
bbobbboobbboboobooboob$
bobobbobobboobobobbbob$
bbobobobobbbbbobboobbb$
bbbboboboobbbbobbboobb$
bbbobobbbbboobobobbbob$
bbboboboooobobooboooob$
bboobobbobbbbbbbbobbbb$
bobbobobbbbbbooobbboob$
bboboboobbbbobboobbbob$
boobobbbobbobobbbbbobb$
bobbobbbobbbobbbbbobbb$
bbbooboboooobbbbooobob$
bbbbboobbbbobobobbboob$
booobbboobbboboobobbbb$
bobboobobbobobbbooboob$
bbbbobboboobboobbbobob$
bbbbbooobbbobobbbbobbb$
bbooobbboobobobobbbobb$
bbobbobbobobbboobboobb$
bbbboobbbobbbbbbbbbbbb$
bbbbbbbbbbbbbbbbbbbbbb!


Total solver time: 0.0927116870880127

[tommyaweosme]

rlifesrc is quicker with that i think

[confocaloid]

(i really want to win highest lifespan on tori b3/s23 by placing a really high period glider tracks with lcm)

What is the lifespan of this pattern?

Code: Select all

x = 40, y = 40, rule = B3-ry/S23-ky
19bo$18bobo$18bobo$19bo4$27bo$28bo$26b3o2$11b2o$10bobo$12bo2$37b2o$36b
o2bo$b2o34b2o$o2bo$b2o4$30bo$30bobo$30b2o2$13b3o$13bo$14bo7$20bo$19bob
o$19bobo$20bo!

Context:

On 2022-09-28 at 05:09:53 UTC, haukeworpel wrote:

In b3-rys23-ky there is a glider-beehive collision where the beehive is shifted five cells and the glider is reflected 90 degrees. This "pathological" pattern is four of these reactions arranged in a square that expands with every collision.

[confocaloid]

What is the smallest pattern by bounding box that contains every B34567 transition and every S1234567 transition? I was able to get 15 by 13.

Code: Select all

x = 15, y = 13, rule = B/S012345678History
.4B.2BABA2BA$.3ABA2BA2B2AB$.ABABA2BABABAB$.BAB2ABABA5B$.4BA2BA2B2A2B$
.BABA2BAB2AB2AB$.7AB3AB2A$2BA3BAB2A2BA2B$BAB5AB2ABA2B$6B5A3BA$.2B2AB
4ABAB2A$2.B2A3BA2BAB2A$2.BA4BA4B2A!

I think you need to include some additional details/explanations about exactly what you are trying to do, before your question becomes clear, unambiguous and answerable. [...]

I meant neighborhood; not transition. I completely forgot that that was a word and fell back onto "transition" because I couldn't think of the right word. That is on me.

I can see several possible ways to interpret the question (i.e. it's still ambiguous). Depending on the interpretation, one can determine several different lower bounds (I don't know whether these lower bounds are attainable):

Isotropic conditions?

A cell can be in one of two states ("off", "on").
A 3-by-3 configuration of cells is said to be strange iff one of the following conditions is met:

• (a) the middle cell is "off", and the number of "on" neighbours is between 3 and 7;
• (b) the middle cell is "on", and the number of "on" neighbours is between 1 and 7.

What is the smallest bounding box area of a known pattern (including all specified cells, both "on" and "off") that contains some orientation of every strange 3-by-3 configuration?

Up to rotations and reflections (one can look through the table),

• there are 10+13+10+6+2 = 41 strange configurations due to (a),
• there are 2+6+10+13+10+6+2 = 49 strange configurations due to (b),

so the pattern must contain at least 41+49 = 90 distinct fully specified 3-by-3 configurations. If the middle cells of those 90+ configurations lie inside a m-by-n rectangle, then the bounding box of the pattern is (m+2)-by-(n+2).

With positive integer m, n, what is the lowest possible value (m+2)(n+2) such that mn >= 90? Assume m >= n; then

• if n = 1 then m >= 90, (m+2)(n+2) >= 276
• if n = 2 then m >= 45, (m+2)(n+2) >= 188
• if n = 3 then m >= 30, (m+2)(n+2) >= 160
• if n = 4 then m >= 23, (m+2)(n+2) >= 150
• if n = 5 then m >= 18, (m+2)(n+2) >= 140
• if n = 6 then m >= 15, (m+2)(n+2) >= 136
• if n = 7 then m >= 13, (m+2)(n+2) >= 135
• if n = 8 then m >= 12, (m+2)(n+2) >= 140
• if n = 9 then m >= 10, (m+2)(n+2) >= 132

This gives a lower bound 132 on the bounding box area.

Neighbour-counting conditions?

A cell can be in one of two states ("off", "on").
A neighbour-counting condition with range-1 Moore neighbourhood can be fully specified by telling the state of the middle cell ('B' = "off", 'S' = "on") and the number of "on" neighbours of the cell.
The twelve conditions B3, B4, B5, B6, B7, S1, S2, S3, S4, S5, S6, S7 are said to be strange, and there are no other strange conditions.
What is the smallest bounding box area of a known pattern (including all specified cells, both "on" and "off") that contains every strange condition?

A lower bound on the bounding box area is 30, via the same reasoning as before. (To demonstrate every strange condition at least once, the pattern must contain at least 12 distinct fully specified 3-by-3 configurations. If the middle cells of those 12+ configurations lie inside a m-by-n rectangle, then the bounding box of the pattern is (m+2)-by-(n+2). The possibilities are n = 1, m >= 12, bb >= 42; n = 2, m >= 6, bb >= 32; n = 3, m >= 4, bb >= 30.)

I think the following pattern provides an upper bound of 42:

Code: Select all

#C green (LifeHistory state 1) = specified on
#C red (LifeHistory state 4) = specified off
#C Neighbour-counting conditions, left to right: B7, B6, B5, S5, S6, S7, S4, B4, B3, S3, S2, S1.
x = 14, y = 3, rule = LifeHistory
4AD2A2DA4D$A3D4A2D3AD$4AD3ADA4D!

Anisotropic conditions?

A cell can be in one of two states ("off", "on").
A 3-by-3 configuration of cells is said to be strange iff one of the following conditions is met:

• (a) the middle cell is "off", and the number of "on" neighbours is between 3 and 7;
• (b) the middle cell is "on", and the number of "on" neighbours is between 1 and 7.

What is the smallest bounding box area of a known pattern (including all specified cells, both "on" and "off") that contains every strange 3-by-3 configuration?

Distinguishing rotations and reflections, there are 218 strange configurations due to (a) and 254 strange configurations due to (b), so the pattern must contain at least 218+254 = 472 distinct fully specified 3-by-3 configurations. If the middle cells of those 472+ configurations lie inside a m-by-n rectangle, then the bounding box of the pattern is (m+2)-by-(n+2). Enumerating possible widths as before gives the lower bound 567 (if n = 19 and m = 25).

[d/dx]

Can someone with more computing power figure out if this grows quadratically?

Code: Select all

x = 98, y = 99, rule = B36/S23
$74b3o$74bo2bo$74bo3bo$75bo2bo$76b3o43$23b3o$22bo2bo$21bo3bo$21bo2bo$
21b3o9$14b3o$13bo2bo$12bo3bo$12bo2bo$12b3o16$15b3o$14bo2bo$13bo3bo$13b
o2bo$13b3o!

[tommyaweosme]

looks like yes

[confocaloid]

looks like yes

The single sentence reply "looks like yes" neither answers the question "Can someone with more computing power figure out if this grows quadratically?" in the preceding post, nor otherwise contributes something valuable to the discussion. Did someone run the pattern (how long? what were the results?) Was there some analysis of the evolution of the pattern?

I guess I could just use the report feature (forum rule 4) instead of replying; decided to reply instead. Both the quoted single-line reply, and this reply to that reply, could just as well not be here, without any negative consequences for the accumulated knowledge about Life/CA.

[d/dx]

There was some analysis of the pattern, by me, around 150.000 generations. There appears to be a replicator puffer heading northwest, but it is VERY dirty. I'm not sure if its stabilised state produces replicators by itself.

[silversmith]

I ran the pattern for ~1.5 million generations, and there's no sign that it stabilized into something quadratic. The area of interaction seems too chaotic for anything like a space filler, while the replicators which are produced by the main puffers tend to collide with debris and eventually form a bomber backend which causes them to shrink over time, or at least not grow linearly.

QuarterSize.png (14.25 KiB) Viewed 6167 times

[unname4798]

me have a new format for yall to check out!

Code: Select all

r=(rule)
(rle)
c (settings)

everything is optional (except rle)
if rule is emited, its life
if settings is emited, nothing happens
all problem am solve! (just need update to lifeiewer for that happen)

Code: Select all

XXYYTRRRrleC
XX - Width
YY - Height
T - Bounded grid type
RRR - OTCA rulestring
C - Settings
All numbers are in base 64.
RRR and C are optional.
If there is no RRR, the rule would be Life.
If there is no C, default LV settings are used.

[tommyaweosme]

XXYYT should be optional

[unname4798]

XXYYT should be optional

T is optional, I just didn't mention it being optional. As for XXYY:
XX is pattern width.
YY is pattern height. (Both have the max value of 4096 and are required)
, is required to separate RLE from X, Y, bounded grid type and rule. It is also used for separating RLE from commands.

[tommyaweosme]

00 works just fine, so i dont know why it isnt optional

[Banananananan]

How do Map transitions work? What do all the letters mean? It looks confusing.