Small Quadratic Growth Patterns

[dani]

So, with the discovery of the 22-cell quadratic growth, my testing of the switch engine pair search space, I think it would be best if this stuff had its own thread, instead of cluttering the Cordership Discussion Thread, which I will continue to do if any more interesting stuff pops up that isn't quadratic.

The important stuff about the last few designs can be found from here down.

I just tried a potential design for a 23-cell quadratic growth pattern based on the newer SEP breeder, but unfortunately no blonks in the blue area can make it form:

Code: Select all

x = 140, y = 140, rule = LifeHistory
140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B
$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$
140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B
$140B$140B$140B$140B$140B$140B$140B$140B$140B$93BC46B$91B2C47B$90BC
49B$90BC49B$90BC49B$140B$140B$140B$140B$140B$140B$100BC39B$98B2C40B$
99BC40B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$
140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B
$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$98BC41B$98BC41B$
98BC41B$99B2C39B$99B3C38B$99B2C39B$140B$140B$140B$140B$140B$140B$140B
$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$
140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B$140B
!

I'm going to run a 4-cell search on the original design to see if there's any alternate 22-cell quadratic growths, just out of curiosity.

[dani]

Search done. It found this alternate 22-cell quadratic growth where the tub is in a different place, but nothing with the other 5 compatible 3-cell patterns:

Code: Select all

x = 170, y = 216, rule = B3/S23
168bo$167bo$167b3o141$9bo$8b3o$10bo40$48bo$47bobo$48bo23$2b3o$o$bo$2bo
$bo$o!

[dani]

New breeder! Produces a block-laying switch engine every 4992 generations:

Code: Select all

x = 90, y = 103, rule = B3/S23
52bob2o3bo5bobo$58bo2b3o3bo$53bo5b4obob2o$52bo3b2obo3bob2o$52b4o9bo$
55b2obo4b3obo$52b6ob3ob3o$54b2o5bo2b3o$54b3o2b2o3bo2bo$53b3obo7bo$54b
2o3bo2b2o3bo$52bo3b2o3bobobo$57b5o3b3o$53b2obo3b4o2bo$52bob2o3bob5obo$
54b2ob2o2bob4o14$79bo$80b5o$79bobo2bob3o$82bo6bo$79bo3b2o4bo$78bobo3bo
4bo$79bo6b3o32$2ob3ob3o2b2o$obo2b4o4bo$obob4obo4b2o$5o2bo5b2o$2bo3bo2b
2ob3o$ob3ob4o3b2o$ob2ob2o6bobo$bob3o2b2obob2o$3ob2obobob3o$3o2bob3o2bo
b2o$obobob3o5b2o$b2o3bo4b3obo$o2bobo3b6o$5bo5bo2bo$2o6b2o2b4o$bob2obob
obo4bo14$32b3o$26bo2b3o2bo$25bo8bo2b2o$25bo5bobo5bo$25b2ob2o2b2ob2o2bo
$26bo7bobob2o$35bo!

I couldn't find a blueprint for a potential 21-cell quadratic growth out of it, because the only working R+glider produces a GPSE facing away, meaning they can't interact. I'll try fiddling with it later.

[dani]

I was in a rush this morning so I couldn't figure out why it wasn't working, but I finally got a blueprint for it after a little time:

Code: Select all

x = 300, y = 250, rule = LifeHistory
.C$2.C$3C58$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$
90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.
210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$
90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.
210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$
90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.
210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$
90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.
210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$
90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.
210B$90.171BCBC36B$90.170BC39B$90.171BC2BC35B$90.173B3C34B$90.210B$
90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.
210B$90.210B$90.210B$90.210B$90.53B3C154B$90.56BC153B$90.56BC153B$90.
210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$
90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.
210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$
90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.
210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$
90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.
210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$
90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.
210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$
90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.210B$90.
210B!

When I plugged that area into apgsearch, it produced this 21-cell quadratic growth after about ten minutes:

Code: Select all

x = 266, y = 164, rule = B3/S23
bo$2bo$3o122$221bo$221b2o19$261bobo$260bo$261bo2bo$263b3o14$143b3o$
146bo$146bo!

The search is still young, so I'll edit if it finds any more placements.

EDIT: Colour in any one of the 28 3-cell islands here for a 21-cell quadratic growth variant.

Code: Select all

x = 1100, y = 250, rule = LifeHistory
.C399.C399.C$2.C399.C399.C$3C397.3C397.3C58$90.210B190.210B190.210B$
90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.
117BD92B$90.210B190.210B190.117BD92B$90.210B190.210B190.117BD92B$90.
210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$
90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.
210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.122B
3D85B190.210B$90.126BD83B190.210B190.210B$90.126B2D82B190.210B190.
210B$90.210B190.210B190.210B$90.210B190.126B3D81B190.210B$90.210B190.
210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B
190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.
210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$
90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.
210B$90.210B190.210B190.151BD58B$90.210B190.210B190.151BD58B$90.210B
190.210B190.151BD58B$90.210B190.210B190.210B$90.210B190.210B190.210B$
90.210B190.210B190.144BD65B$90.210B190.210B190.144BD65B$90.210B190.
210B190.144BD65B$90.210B190.210B190.210B$90.210B190.106B3D101B190.
210B$90.115BD94B190.210B190.210B$90.115B2D93B190.210B190.210B$90.210B
190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.
210B190.210B190.210B$90.210B190.164B3D43B190.210B$90.210B190.210B190.
210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B
190.119BD90B$90.210B190.210B190.119BD90B$90.210B190.210B190.119BD90B$
90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.
210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B
190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.
210B190.210B$90.210B190.210B190.210B$90.131BD78B190.210B190.210B$90.
131B2D77B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.
210B$90.210B190.210B190.137BD72B$90.210B190.210B190.137BD72B$90.210B
190.210B190.137BD72B$90.210B190.210B190.210B$90.210B190.210B190.210B$
90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.
210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B
190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.
210B190.210B$90.210B190.210B190.210B$90.210B190.155B3D52B190.210B$90.
171BCBC36B190.150B3D18BCBC36B190.171BCBC36B$90.170BC39B190.170BC39B
190.170BC39B$90.171BC2BC35B190.171BC2BC35B190.171BC2BC35B$90.151BD21B
3C12BD21B190.173B3C34B190.173B3C34B$90.151B2D35B2D20B190.210B190.210B
$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.190B3D17B
190.210B$90.210B190.210B190.210B$90.210B190.210B190.156BD53B$90.154BD
55B190.167B3D40B190.156BD22BD30B$90.154B2D54B190.210B190.156BD22BD30B
$90.210B190.210B190.179BD30B$90.170BD39B190.210B190.171BDBD36B$90.
170B2DBD36B190.210B190.171BDB2D35B$90.173B2D35B190.174B3D33B190.171BD
B2D35B$90.210B190.210B190.174BD35B$90.53B3C154B190.53B3C124B3D27B190.
53B3C154B$90.56BC153B190.56BC153B190.56BC153B$90.56BC153B190.56BC153B
190.56BC153B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B
190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.
210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$
90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.
210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B
190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.
210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B
190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.
210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$
90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.
210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B
190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.
210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B
190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.
210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$
90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.
210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B
190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.
210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B
190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.
210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$
90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.
210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B
190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.
210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B
190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.
210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$
90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B190.
210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.210B
190.210B$90.210B190.210B190.210B$90.210B190.210B190.210B$90.210B190.
210B190.210B$90.210B190.210B190.210B!

There's a part in the 2nd blinker set where two blinker placements are right next to eachother and look like a toad - only fill in one at a time.

EDIT2: Adam mentioned on the Discord that this breeder is only formed by two raw GPSE's with nothing special inbetween, so that may explain the trove of timings.

I tried a 10+10 method for a potential 20CQG but it just wouldn't budge. The timings are even more precise than I thought, and the closest I was able to get with my collection of predecessors was a pair with one facing the wrong direction, and the ash timing way off in some direction.

[dani]

you could try patterns that put any kind of switch engine at the right place/phase, and pray that the messy starting reaction converts both into GPSEs

This interesting idea came from Discord, and I spent the past day and a half organizing my files, and sadly did not find any results. I know of 27 10-cell switch engine predecessors, most from some 2 posts by simsim.

Two close calls:

Code: Select all

x = 1249, y = 174, rule = LifeHistory
112.A991.A$112.A991.2A$112.A990.2A$110.2A$110.A$110.A5$A$A$.3A19$78.
3A9$1002.A$1000.2A$1001.A5$1013.A$1012.A$1012.A$1009.3A109$1246.2A$
1246.A.A$1246.A10$144.3A$144.A$145.A!

However, it hit me that I don't think simsim (or anybody for that matter) has covered the 7-cell methuselah + 3-cell blinker/preblock space, so I'm going to be running many searches to see if there's predecessors in there. 7 cells has a lot more unstable objects so I'm hopeful.

EDIT: Oops, I actually think he did. One of the predecessors I used above even uses a 7+3. I'll run some methuselahs I don't see represented in the data set.

[dani]

Switch Engine Pair Breeder #4: Emits a GPSE each 6528 generations:

Code: Select all

x = 110, y = 76, rule = B3/S23
79b2ob4ob2o$79bo2bobo$85b3o$80bo5b3o$79bob2o2bob2o$79b2o2b2o3bo$80bobo
b4o$79bob2obo3bo$81bobob4o$79bo3b2o2b2o14$103b2o$102b2ob2o$107b2o$105b
o3bo$106b3o18$b2o3bo2bo$o3b2ob3o$2o2b6o$bo3b2obo$4obobo$o4b2obo$2obob
2ob2o$b4o2b3o$3b2ob4o$b4o2bo13$28bo$29bo$19bo10bo$18b3o11b2o$18bo2bo3b
o6b3o$21b2o2b3o7bo$21bo4b9o$20bo6b7o$29b3o!

No 20-cell 10+10 equivalents, as usual. There was this pretty close call:

Code: Select all

x = 197, y = 393, rule = B3/S23
b2o$2o$bo29$16bo$15bobo$15bo$14bo305$192bo$193b2o$196bo$194b2o$195bo
47$185bo$185bo$185bo!

[dani]

I thought we were done with new breeders, but it seems not. Switch Engine Pair Breeder #5, generates a BLSE every 26112 generations:

Code: Select all

x = 104, y = 67, rule = B3/S23
72bo2bo2b2o$70b2o6bo$73b2o$70bobo2b3obo$74b2obo$70bob4o2b2o$70b2o2bo2b
2o$70b2o2b2o2b2o$70bo2bob5o$70bo4bobo13$98b2o$97bo2bo$91bo4bo$89bo2bo
2bo$89bo2bo2bo4b2o$89bo5bo4bob2o$90bobo4b3obo$90bobo7$3o2b2o$2b2obo$2o
3bo$2o4bobo$2b2o3bo$o5b4o$o5b3o$o3bobo$3b3o2b2o$bo2bo3b2o14$26b2o$22bo
$22b5obo$20bo3bo2b3o$21bobo$21bo8bo$30bo$30bo!

This time, 10+10 *almost* budged. This pair of predecessors creates a high period puffer but sadly not the breeder:

Code: Select all

x = 162, y = 190, rule = B3/S23
94bo$93b3o$93bo12$127bo$125b2o$124bo$125bo26$o$bo$3o142$160b2o$159b2o$
161bo!

I have a couple ideas on how I can programmatically explore more of the 20-cell space that I'm going to run soon.

[dani]

6G Quadratic growth:

Code: Select all

x = 120, y = 363, rule = B3/S23
95bobo$96b2o$96bo138$117bobo$118b2o$118bo3$114b3o$116bo$115bo69$22b2o$
23b2o$22bo71$92b3o$94bo$93bo69$2o$b2o$o!

Based off #4 in this thread. The previous record for quadratic growth was 7G with switch engine ping pong. I figured that at least one of these puffers could yield a 3+3 synthesis due to a lot of them being made of just two plain switch engines.

EDIT: If you haven't seen, Rocknlol got a Breeder #1 ancestor to fit into a 23x13 bounding box. Check it out here.

[dani]

A 20-cell quadratic growth from my latest experiments:

Code: Select all

x = 97, y = 33, rule = B3/S23
94bo$92bobo$94b2o6$88b3o11$96bo$95b2o8$3bob2o$2bo3bo$bo$bo$obo!

It forms a never-before seen SEP Breeder #6, which produces a BLSE every 11136 generations.

The bounding box is remarkably small. It's a 13.6x+ reduction from 21CQG.

[calcyman]

A 20-cell quadratic growth from my latest experiments:

Code: Select all

x = 97, y = 33, rule = B3/S23
94bo$92bobo$94b2o6$88b3o11$96bo$95b2o8$3bob2o$2bo3bo$bo$bo$obo!

It forms a never-before seen SEP Breeder #6, which produces a BLSE every 11136 generations.

The bounding box is remarkably small. It's a 13.6x+ reduction from 21CQG.

Congratulations! Do you have a name for the pattern? It needs a specific name in order to continue this mod-3 tradition:

26: wedge-grow
25: 25-cell quadratic growth
24: 24-cell quadratic growth
23: switch-engine ping-pong
22: 22-cell quadratic growth
21: 21-cell quadratic growth
20: ???

[dani]

Ahhh I'm not really sure, I've never named a pattern. The only feature I can think of that distinguishes it from the other patterns in the series is that it has that unnamed 9-cell island on the left that turns into H+SE. Just so this post has some content, here's a few variations on that cluster for increased trivial variants of the 20-cell quadratic growth, along with the alternate R-pentomino grandparents that can be swapped in:

Code: Select all

x = 17, y = 35, rule = B3/S23
3bob2o6bo2bo$2bo3bo5bo2b2o$bo9bo$bo9bo$obo7bobo6$2b2ob2o5b2o2bo$6bo8b
2o$bo9bo$b2o8b2o$o9bo6$5b2o9bo$3bo2bo6bob2o$b2o8b2o$bo9bo$o9bo$2bo9bo
5$3bob2o6bo2bo$bo4bo4bo3b2o$bo9bo$2o8b2o$2bo9bo!

If I feel like there's a decent shot I can knock another cell off soon, would it be worth waiting until the time comes to name that one, or is the mod 3 special?

EDIT 04/08/2022: added another equivalent to the above RLE

[squareroot12621]

[…]Congratulations! Do you have a name for the pattern? It needs a specific name in order to continue this mod-3 tradition:

26: wedge-grow
25: 25-cell quadratic growth
24: 24-cell quadratic growth
23: switch-engine ping-pong
22: 22-cell quadratic growth
21: 21-cell quadratic growth
20: ???

AGH! Why is there already a pattern named Double X?! I'm going to call it "Growing Double X."
I'm sure someone can find a better name than I can.

Edit: Maybe "Breeder 20" would work.

[77551enpassant]

I suggested a name yesterday in the naming proposals thread. "Double trouble" would fit.

[yujh]

(sorry for yet another non-constructive post) I suggest it to be called 20 cell quadratic growth.

[NickGotts]

Many congratulations on this work, dani! When simsim beat the 26-cell record I'd established with Wedge-grow, I made some attempts at beating the 23 mark established by switch-engine ping-pong, but with no progress.

Could you post your 27 10-cell switch engines?

[dani]

Many congratulations on this work, dani! When simsim beat the 26-cell record I'd established with Wedge-grow, I made some attempts at beating the 23 mark established by switch-engine ping-pong, but with no progress.

Could you post your 27 10-cell switch engines?

Thank you!! }:)

Yeah, none of them are new:

Code: Select all

#C By Michael Simkin
x = 47094, y = 1081, rule = B3/S23
57bo999bobo940b3o1082bo2008bo905b3o1023b3o998bo2024bo947b3o3028bo969b
2o4052bo946bo3004bo1992bo8000bo998bobo997bo1017bo1072b3o997bo8999bo$
57bobo997b2o943bo1080b2o2009b2o1931bo999bo2024b2o3977bo968b2o4052bobo
943b2o3006b2o1990bo8000b2o998b2o997bo1015b3o1073bo998bo8999b3o$57b2o
999bo944b3o1076bo2010b2o1933bo998bo6003bo969bo4054bo944bo3009bo1988bo
7999b2o1000bo998b3o1017bo1071bo997bo9000bo$3083bo4944b2o11027bo3951b2o
13010bo2070b2o$23010bo13010bo3$20013bo$20012bo$20012bo$20009b3o2$
19001bo$19000b3o$19002bo2$5003bo$2067b2o2932b2o$2068bo2931bo$5001bo$
36000b3o6$10000bo$10001b2o$10000bo$10000b2o$10000bo$15016bo$14003b3o
1009bobo10006bo$14003bo1011bo10007bobo$11023b3o2974b3o1011bo10005b3o2b
o$11025bo$11026b3o15$23000bo$23000bo14000bo$23000bo13999b2o998bobo$
37001b2o997b2o$38000bo9000bo$7001bo39999b2o$7000b3o39997bobo$7002bo
997b3o$8001bo$3001bo3001bo1996bo$3000b2o3002b2o$3001b2o3000bo$6003b2o$
6003bo27$2bo$b2o998b3o$obo997bo$1000bo65$33143b2o1000bo$33143bobo998b
2o$33143bo1000bobo997b3o$35144bo$35145bo837$b3o1996b3o33014bo11984bo$o
2001bo33013b2o11983bo$o2002b3o33010bo11983b3o$35019b3o5$47079bo$47079b
3o$47080bo7$2067b2o$2068bo$35000bo$35000bo$35000bo57$9bo$8b2o$7bobo!

The only "new" one is from someone on the Discord (I think IgnacyJ?) noticing that one of them could be rewound by a tick and still be 10 cells.

[NickGotts]

Interesting! There must be some there I wasn't aware of - I haven't kept up with simsim's postings - but there are also some you don't have. The six below only differ trivially, but they are all composed of a single cluster, which none of your 27 do, and I think one was the first 10-cell infinite growth pattern found, by Paul Callahan back in 1998.

#CXRLE Pos=44549,27387
x = 1374, y = 985, rule = B3/S23
12bo660bo694bo2$12b2o659b2o693b2o2$14b3o658b3o692b3o2$15b3o658b2obo
690bob2o$16bo660bo694bo970$bo660bo694bo2$2o659b2o693b2o2$2b3o658b3o
692b3o2$3b3o658b2obo690bob2o$4bo660bo694bo!

I'll do a more systematic comparison, though my guess is that you have all the others I know of - most are from from simsim.

I have a project to do an absolutely complete and systematic survey of all patterns up to 10 cells, which I was reporting on here: viewtopic.php?f=2&t=3604, until I restarted paid work in July 2020. I hope to get back to it soon.

[NickGotts]

I see you (dani) added a comment to the "Systematic Survey of Small Patterns" on Nov.30th 2018. I discovered after my last posting on that thread that I might have overlooked some multi-cluster patterns with fewer than 10 cells, started work on redoing that part of the search, then got a job (which I had not intended to do, having retired some time ago, but the job was Covid-related). My employment has ended, but I still have related work to do. Once that's finished, I'll get back to the survey.

[mvr]

Here are the 10-cell r + pi BLSE seeds, I haven't seen these recorded anywhere.

Code: Select all

b2o$2o$bo31$35bo2bo$36b3o!
o$2o$obo45$49bo$48bo$48bo$48b2o!
43bo2bo$44b3o52$o$2o$obo!
2o$o$o$bo7$12bo$12b3o2$13bo!
2o$bo$bo$o11$51bo$51b3o2$52bo!
o$3o2$bo45$48bo$47b2obo$48bo!
45bo2$44b3o$45bo51$o$3o2$bo!

Some can be advanced one generation and still be 10-cell:

Code: Select all

b2o$2o$bo46$49bo$48b2obo$49bo!
46bo2$45b3o$46bo51$b2o$2o$bo!
bo$2obo$bo8$13bo$13b2o$13bobo!
2bo$ob2o$2bo12$52bo$52b2o$52bobo!

[simsim314]

A 20-cell quadratic growth from my latest experiments

Very neat! I am very glad to see my old record breaking run from 26 by Nick Gotts to 23 quadratic finally broken.

I remember thinking how cool is that to break a record that held for 8 years. Now I see this record broken in about the same time (for exact dates see LifeWiki) to the knees of the custom small constellations apgsearch together with the new soup discoveries of switch-engines pair emitting a perpendicular line of switch-engines one after another (the typical switch-engine pair quadratic pattern). All that is new and done due to apgsearch. I needed to invent the switch engine ping pong that used 2 and not three switch engines. This required a very high periodic synchronization (this is why the pattern is so large). I am glad to see more solutions were found and our knowledge of the field is expanding rapidly. I believe that this record will not be broken that soon, as there is so much that can be done with 20 cells in CGOL. We are very soon will be able to prove we solved this problem completely.

Regardless of the discovery attribution, we all use calcyman's tools for our latest discoveries, so he deserves a big thanks for his effort for sure.

Congrats again!