Thread for your unsure discoveries

[praosylen]

Bookends as an eater:

Code: Select all

x = 79, y = 7, rule = B3/S23
2o18b2o9b2o9b2o9b2o9bo12b2o$o2bo16bo2bo7bo2bo3bo3bo2bo7bo2bo5bobo10bo
2bo$b3o17b3o2bo5b3o2bo5b3o2bo5b3o4bobo11b3o$8b2o16b3o7bo2bo8b2o10bo2bo
$b3o3b2obo10b3o8b3o2bo5b3o3b2o3b3o4bobo11b3o$o2bo5bo10bo2bo7bo2bo3bo3b
o2bo7bo2bo5bobo10bo2bo$2o18b2o9b2o9b2o9b2o9bo12b2o!

I don't know if this has already been found, but it seems as if it may be useful.
Edit: Here's another one:

Code: Select all

x = 9, y = 7, rule = B3/S23
2o5bo$o2bo3bo$b3o2bobo$6bo$b3o2bobo$o2bo3bo$2o5bo!

Edit 2: T-pentomino turns a pair of bookends into a 22-cell still life:

Code: Select all

x = 8, y = 7, rule = B3/S23
2o$o2bo2bo$b3o2bo$5b3o$b3o$o2bo$2o!

[praosylen]

This is so simple that I'm sure it's been found before:

Code: Select all

x = 11, y = 8, rule = B3/S23
2o$bo$bobo$2b2o5bo$8bo$8bo$8b2o$10bo!

[Scorbie]

I think I found a c/3 signal, but I'm not sure...
(EDIT: I found the pattern with dr2. Many thanks to Sokwe for modifying and uploading the drifter searcher!)
can it be made into a c/3 or maybe (3,1)c/9 signal? (compare gens 16 and 25)

Code: Select all

x = 16, y = 25, rule = dr2
3.8B$2.B2.A.A2.AB$.B4A.2A2.3B$BA4.A2.2A2.AB$BA2.A.A.A2.A2.B$BA.A.A.A.
A.2A.B$BA.A.A2.A.A2.AB$B2.A2.A.A.A.A.B$B3A.2A.A.A.A.B$B3.A2.A.A2.A.B$
.BA2.A.A.A.A.B$.B2A.A.A.A.A.B$.3B.A2.A.A.B$4.B.A.A.A.B$4.B.A.A.2AB$4.
BA.2A3.B$3.B.A3.3AB$3.B.A.2A4.B$3.B.A2.A.4AB$3.B2.A.A.A3.AB$3.BA.A.2A
3.A.B$3.B.A7.A.B$3.B2.7A2.B$4.5B6.B$9.6B!

I made the rule for viewing dr results. state 2 cells make adjacent cells stable. I made the rule so that I could look at the patterns without stabilizing their boundaries.

Code: Select all

# rules: 18
#
# Golly rule-table format.
# Each rule: C,N,NE,E,SE,S,SW,W,NW,C'
# N.B. Where the same variable appears multiple times in a transition,
# it takes the same value each time.
#
# Default for transitions not listed: no change
#
n_states:3
neighborhood:Moore
symmetries:rotate8
var a={0,1}
var b={0,1}
var c={0,1}
var d={0,1}
0,0,0,0,0,0,1,1,1,1
0,0,0,0,0,1,0,1,1,1
0,0,0,0,0,1,1,0,1,1
0,0,0,0,1,0,0,1,1,1
0,0,0,0,1,0,1,0,1,1
0,0,0,0,1,1,0,0,1,1
0,0,0,1,0,0,1,0,1,1
1,0,0,0,0,0,0,0,a,0
1,a,b,c,d,1,1,1,1,0
1,0,a,b,1,0,1,1,1,0
1,0,a,b,1,1,0,1,1,0
1,0,0,0,1,1,1,0,1,0
1,0,0,1,0,0,1,1,1,0
1,0,0,1,0,1,0,1,1,0
1,0,0,1,0,1,1,0,1,0
1,0,0,1,1,0,0,1,1,0
1,0,0,1,1,0,1,0,1,0
1,0,1,0,1,0,1,0,1,0

I'll make it into a life pattern when I have more time. I'll be thankful if somebody else helps me with that...

[Scorbie]

Oh, and this is known, right?

Code: Select all

x = 36, y = 14, rule = dr2
.6B$B6.B$B6A.2B$BA4.A.A.B$B2.A2.A.2A23B$B2A.A.A26.B$B3.A.27AB$B2A.A
24.A3.AB$.B.A.24A3.A.B$.B.A.A26.A.B$2.BA2.25A.A.B$3.4B24.A2B$7.25B!

[Sokwe]

I think I found a c/3 signal, but I'm not sure...

Unfortunately, it doesn't seem to be repeatable (at least not in a way that I could find by hand). Still, it's an interesting reaction. It might be useful at some point.

Oh, and this is known, right?

It's been known since at least 1971. It was in Lifeline volume 3 (page 17) and was apparently discovered by Wolfgang Ebenhoh. While I'm on the topic, Gabriel Nivasch has a nice page on light-speed signals.

[Scorbie]

I'll make it into a life pattern when I have more time. I'll be thankful if somebody else helps me with that...

Ugh, I came here again... I really should do other things...

Unfortunately, it doesn't seem to be repeatable (at least not in a way that I could find by hand). Still, it's an interesting reaction. It might be useful at some point.

Ohh, I see... Good choice to post it in the unsure discoveries thread.

Maybe I'll post some new rotors some time later.

By the way, I have some questions about dr2.

When I try to run it with different grid sizes, (by changing H# and W#,) Every rotor is UNKNOWN, and the rotor desctiption gets blurred....ugh. Any Ideas?

Code: Select all

Obviously, It's gotta be kidding me.
*****  Period 2 at gen 2
p2 r4 2x2 0@ @0	<- UNKNOWN
u4 r9 3x3 0@@ @@0 @0@	<- unknown
Full change counts: 0 1 2 4 2
Change counts: 0 0 1 3 1
Sizes: 0x0 0x0 1x1 2x2 1x1
Gen 0.  Rows 8 - 18.  Cols 33 - 43.
,
,
,,,o..oo
,,.o.o..
,,.o..o.o
,,.o.oo.o
,,..o....
,,oo..o.o
,,,,,oo.o
,
,

Also, could you tell me about the 'u' descriptions?? ex)u4 r7 3x4 ..2. 200B D.2. <- unknown

[Sokwe]

When I try to run it with different grid sizes, (by changing H# and W#,) Every rotor is UNKNOWN, and the rotor desctiption gets blurred....ugh. Any Ideas?

For some reason it seems to think that there are 4 rotor cells in the final pattern when there are only two. I am not sure why that is. What does your input file look like?

Also, could you tell me about the 'u' descriptions?

The 'u' descriptions are (I think) analogous to the 'p' descriptions, but the 'u' descriptions describe the entire history of the pattern. So if you have

Code: Select all

u4 r9 3x3 0@@ @@0 @0@   <- unknown

then at generation 4 the pattern was recognized as periodic, and 9 cells changed during the course of the pattern's evolution, all withing a 3x3 bounding box. The last piece describes the initial state of the pattern in the same way as the 'p' descriptions.

[Scorbie]

For some reason it seems to think that there are 4 rotor cells in the final pattern when there are only two. I am not sure why that is. What does your input file look like?

It was:

Code: Select all

H30
W80
c8
h4
w4
v127 1
r14 39
22
22!
skipstable
skipfizzle

I hope this doesn't hack your computer...

[Sokwe]

The problem seems to be solved if you replace H30 with H80. I suspect that Dean Hickerson's original program does not handle this properly when the H or W parameters are less than 80.

The 'H' and 'W' commands are really only intended to be used for symmetric searches. Basically, if you want the height of the symmetric pattern to be even, use H80. I don't think there would be any advantage to setting H or W less than 80. At each step, the program only checks around cells that were changed recently, so the overall size of the field doesn't affect the speed of the search.

[Scorbie]

I see! I wanted to make the grid size smaller because I was annoyed at dr doing a lot of calculations just for constructing known (5c/9, mainly) signals with different lengths and ends. (but orthogonal signals looked rarer and more exciting for me, that's why I made it W80)
(Adding a single cell to block the signal wouldn't work because we don't know where the signal will grow)
It's not a big problem, so I think I could do searches without changing the grid size. Thanks a lot for your help!

[praosylen]

May result in a synthesis of Octagon 4:

Code: Select all

x = 20, y = 20, rule = B3/S23
5bo$6b2o$5b2o3$17bobo$10bobo4b2o$6bo3b2o6bo$7b2o2bo$6b2o$12b2o$8bo2b2o
$bo6b2o3bo$b2o4bobo$obo3$13b2o$12b2o$14bo!

[NotAvailableCooper]

I found an orthogonal period 3 pushalong. Too bad no period 3 spaceship has a leading or sufficiently isolated side spark.

x = 7, y = 5, rule = B3/S23
2bo$bobo$o3bo$b2obo$6bo!

[Bjorn]

I'm not sure what to classify this as, but until someone suggests a better name I'm calling this a dirty MSS breeder. At first in this pattern's evolves nothing really interesting happens, until at generation 130,000 a sort of stationary puffer/replicator appears. As the pattern evolves more of these appear and some die off. I'm not sure if it is a true breeder or whether it eventually stabilizes.

Code: Select all

x = 5, y = 25, rule = B3/S23
o2bo$4bo$o3bo$b4o4$o$b2o$2bo$2bo$bo2$bo$2bo$2bo$b2o$o4$b4o$o3bo$4bo$o
2bo!

[codeholic]

It's definitely not a breeder. It's a one of multiple variants of a twin B-heptomino puffer, found in 1971.

And I see nothing spectacular at generation 130,000.

[Bjorn]

It's definitely not a breeder. It's a one of multiple variants of a twin B-heptomino puffer, found in 1971.

And I see nothing spectacular at generation 130,000.

Oops, wrong pattern
This should work a bit better:

Code: Select all

x = 110, y = 69, rule = B3/S23
9bobo$8bo2bo$7b2o$6bo$5b4o$4bo4bo6bo$4bo2bo5b2obo$4bo2bo5bo$5bo7bo$6b
4obo$7bo3bo$8bo6b6o$8b5o2bo5bo$15bo$8b5o3bo4bo$8bo9b2o$7bo3bo2bo$6b4ob
o$5bo7bobo$4bo2bo5bo$4bo2bo5b3o$4bo4bo$5b4o$6bo$7b2o$8bo2bo$9bobo11$5b
obo$4bo2bo$3b2o$2bo$b4o$o4bo6bo$o2bo5b2obo$o2bo5bo$bo7bo$2b4obo$3bo3bo
$4bo6b6o$4b5o2bo5bo81b2o5b4o$11bo85b2ob2o3bo3bo$4b5o3bo4bo79b4o8bo$4bo
9b2o82b2o5bo2bo$3bo3bo2bo$2b4obo88bo$bo7bobo83b2o8b2o$o2bo5bo84bo9bo2b
o$o2bo5b3o83b5o4bo2bo$o4bo90b4o3b2ob2o$b4o94bo4b2o$2bo$3b2o$4bo2bo$5bo
bo98b4o$105bo3bo$88b4o17bo$87bo3bo13bo2bo$91bo$87bo2bo!

Oh and it's around generation 130,000 (might be closer to 140,000) that it gets interesting.

[towerator]

Interesting, indeed, the pattern starts to emit self-replicating collisions.
I continued to run this, here's the result: a fourth "replicollision" (Do you have a better name?) appears at around 650K
And a fith at around 1M250K.
The third and the first replicollisions are damaged at aroung the same time and stop working
A sixth appears at around 1M600K, but quickly dies
A seventh then appears at around 1M750K
A eight at around 1M900K, quickly followed by a ninth that interacts a bit, and a tenth at gen 2M050K. The eigth dies at 2M100K, and ninth and tenth at generation 2M150K
An eleventh starts for a small time at 2M200K
A twelfth dwells at 2M250K, only to be destroyed too

I didn't find any periodicity yet, it is probably extremely long to come.
I don't know if the 4 great "replicollisons" survive indefinitely.

At 5M, there's still no sign of periodicity. But the increasing number of lines gowing on a finite number of gliders may result in a dry out.


EDIT: times said here are false because I looked at population...

[Bjorn]

Interesting, indeed, the pattern starts to emit self-replicating collisions.
I continued to run this, here's the result: a fourth "replicollision" (Do you have a better name?) appears at around 650K
And a fith at around 1M250K.
The third and the first replicollisions are damaged at aroung the same time and stop working
A sixth appears at around 1M600K, but quickly dies
A seventh then appears at around 1M750K
A eight at around 1M900K, quickly followed by a ninth that interacts a bit, and a tenth at gen 2M050K. The eigth dies at 2M100K, and ninth and tenth at generation 2M150K
An eleventh starts for a small time at 2M200K
A twelfth dwells at 2M250K, only to be destroyed too

I didn't find any periodicity yet, it is probably extremely long to come.
I don't know if the 4 great "replicollisons" survive indefinitely.

At 5M, there's still no sign of periodicity. But the increasing number of lines gowing on a finite number of gliders may result in a dry out.

Replicollision isn't bad. A thought I had would be to look to biology for a name, it seems to reproduce but only with help from an external source. Virus maybe. A name would be helpful as I've also seen a similar thing happen with a long line of glider-producing switch engines. I was also wondering if there could be some maximum number of replicollisions/viruses, as each replicollision/virus does seems to take a small chunk in the stream of gliders out. Even if it does dry out the concept of a new kind of breeder is still there.

[codeholic]

Looks like some kind of a crystal, or a parasite. http://pentadecathlon.com/lifeNews/2011 ... sites.html

[Bjorn]

Looks like some kind of a crystal, or a parasite. http://pentadecathlon.com/lifeNews/2011 ... sites.html

I was thinking of parasite, but wouldn't that also describe a feature of life without death?

[codeholic]

life without death

Do you mean infinite novelty?

[Bjorn]

life without death

Do you mean infinite novelty?

I might, I'm talking about B3/S012345678.

[biggiemac]

This is the base reaction being discussed? (For those who don't want to wait 130K generations)

Code: Select all

x = 24, y = 27, rule = B3/S23
5bobo$4bo2bo$3b2o$2bo$b4o$o4bo6bo$o2bo5b2obo$o2bo5bo11bo$bo7bo11bobo$
2b4obo13b2o$3bo3bo$4bo6b6o$4b5o2bo5bo$11bo$4b5o3bo4bo$4bo9b2o$3bo3bo2b
o$2b4obo3bobo$bo7bo4bo$o2bo5bo4bo$o2bo5b2o2b2o$o4bo5b2o$b4o$2bo$3b2o$
4bo2bo$5bobo!

I wonder if there is a way to perturb this slightly to afford a quadratic growth of these islands.

Edit: I overlooked the obvious fact that quadratic growth is not going to happen if there is only a linear supply of the gliders from the original puffer to grow from..

[towerator]

This is the base reaction being discussed? (For those who don't want to wait 130K generations)

Code: Select all

x = 24, y = 27, rule = B3/S23
5bobo$4bo2bo$3b2o$2bo$b4o$o4bo6bo$o2bo5b2obo$o2bo5bo11bo$bo7bo11bobo$
2b4obo13b2o$3bo3bo$4bo6b6o$4b5o2bo5bo$11bo$4b5o3bo4bo$4bo9b2o$3bo3bo2b
o$2b4obo3bobo$bo7bo4bo$o2bo5bo4bo$o2bo5b2o2b2o$o4bo5b2o$b4o$2bo$3b2o$
4bo2bo$5bobo!

I wonder if there is a way to perturb this slightly to afford a quadratic growth of these islands.

Edit: I overlooked the obvious fact that quadratic growth is not going to happen if there is only a linear supply of the gliders from the original puffer to grow from..

The "replicollisions" behave in a very chaotic way. At some later points, I saw one rising "spontaneously" detached" to the main "spine". The following mess killed an old one. There is a limited number of those on a pattern, and they seem to continuoulsy "battle" for space.

[Bjorn]

What if we found a replicollision that work on a sparser stream of gliders (or just a thinner field of derbies or both) that acted as sort of a transparent reaction (without deleting another glider)? It would replace the glider lost so it could grow quadratically. If something like this could be found it would be a brand new kind of breeder.

[codeholic]

What if we found a replicollision that work on a sparser stream of gliders (or just a thinner field of derbies or both) that acted as sort of a transparent reaction (without deleting another glider)?

If I'm not mistaken, there is no (trivial) Heisenburp reaction between gliders. I think I once did such a search with up to 3 gliders - but of course, it didn't cover all possible collisions.