Snowflakes (B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e)

[BlinkerSpawn]

A collection of cheap dot constellations would come in handy sometimes.
Also, I noticed the trapped-G reaction can be rearranged like this, opening one entire side for sparking for the other car-siamise-G, in case anyone wants to give that a try:

Code: Select all

x = 10, y = 20, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
bo$obo$3o$bo3$2bobo$2b3o3b2o$8bo$2bo5b2o3$3bo4$3bo$2b3o$2bobo$3bo!

[dani]

I've noticed two promising three-dot constellations. The left is called 'close tridots' and the right is called 'sparse tridots'

Code: Select all

x = 15, y = 5, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
o9bo2$3bo10bo2$o9bo!

The left is produced by an unperturbed T and, as such, mostly comes in pairs. However, the right is what happens when said T is perturbed in the mid-developmental stages, i.e.:

Code: Select all

x = 11, y = 3, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
o8b2o$2o7bo$o8b2o!

Now, here are two potentially useful reactions with sparse tridots and g:

Code: Select all

x = 7, y = 32, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
4bobo$4b3o4$o2$4bo2$o11$obo$3o4$o3bo4$2bo2$5bo!

As you can see, one reproduces the pattern rotated, and the other pulls it messily 2 cells back. I think this is pretty promising and other experiments with these constellations may lead to fun transparent catalysts like this:

Code: Select all

x = 5, y = 6, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
2o$bo2bo$2o3$2bo!

EDIT: Also, amazing work with the syntheses! I'm astonished by AbhpzTa's reduction(s), too.

[2718281828]

A collection of cheap dot constellations would come in handy sometimes

I have all 2G and 3G collisions (up to a certain radius due to infinite collision families, e.g.

Code: Select all

x = 16, y = 19, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
14bo$13bobo$13b3o$14bo8$2bo$bobo$b3o$2bo2$2o$bo$2o!

and also all other options, e.g. 2R, 3R, 2R+G, etc.)

However, the format is not very pleasant and with the in total 7148 3G collision files it gets quite messy.
Unfortunately my code works at the moment only for 3-spaceship collisions, but I am working on a 4-spaceship versions. It should work for Snowflakes quite well, as the reactions are usually very fast, even 5spaceship is thinkable.
Getting the 2G collisions requires 2 seconds, and all the 3G collisions less than 2 minutes. (In B3/S23 the 2G takes 8 seconds and the 3G a couple of hours)

Still, I attached (for the format the 27 2G collisions incl. the ash files, and bounded ash files[excluding escaping objects], short bash), such as all 3G collisions which results from the 2G collision that forms the T (the most interesting on) with the corresponding ash files.

[toroidalet]

sorry to interrupt the synthesis progress (and my lack of posts), but here's a c/70 tethered R rake:

Code: Select all

x = 32, y = 58, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
12bo$10b5o$10bo3bo$9b2obob2o$10bo3bo$10b5o$12bo4$14bo$14bo$14bo4$14bo$
13bobo$13b3o$14bo3$14bo$12b5o$12bo3bo$11b2obob2o$12bo3bo$12b5o$14bo2$
20bo$18bo$18b2o2$9bo6bo$11bo10bo$10b2o8bo$20b2o2$7bo$5bo$o$16b2o6bo6bo
$3bo2bo9bo12bo$16b2o6bo4b2o$o$5bo$7bo$7b3o5bo$9bo12bo$15bo4bo$20b2o4$
13b2o$13bo$15bo!

it can push a snowflake and any number of dots:

Code: Select all

x = 32, y = 58, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
12bo$10b5o$10bo3bo$9b2obob2o$10bo3bo$10b5o$12bo4$14bo$14bo$14bo4$14bo$
6bo6bobo$4b5o4b3o$4bo3bo5bo$3b2obob2o$4bo3bo$4b5o5bo11bo$6bo5b5o$12bo
3bo$11b2obob2o$12bo3bo$12b5o$14bo2$20bo$18bo$18b2o2$9bo6bo$11bo10bo$
10b2o8bo$20b2o2$7bo$5bo$o$16b2o6bo6bo$3bo2bo9bo12bo$16b2o6bo4b2o$o$5bo
$7bo$7b3o5bo$9bo12bo$15bo4bo$20b2o4$13b2o$13bo$15bo!

some dot pull reactions that eventually destroy the engine:

Code: Select all

x = 64, y = 58, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
31bo$29b5o$29bo3bo$28b2obob2o$29bo3bo$29b5o$31bo4$33bo$33bo$33bo4$33bo
$32bobo$32b3o$33bo3$33bo$31b5o$o30bo3bo$30b2obob2o$31bo3bo$31b5o$33bo$
63bo$39bo$37bo$37b2o2$28bo6bo$30bo10bo$29b2o8bo$39b2o2$26bo$24bo$19bo$
35b2o6bo6bo$22bo2bo9bo12bo$35b2o6bo4b2o$19bo$24bo$26bo$26b3o5bo$28bo
12bo$34bo4bo$39b2o4$32b2o$32bo$34bo!

the speeds are c/70d (2c/140d) and (1,3)c/210 ((2,6)c/420).

[2718281828]

sorry to interrupt the synthesis progress (and my lack of posts), but here's a c/70 tethered R rake:

Cool reaction.

I noticed a nice property of xq4_1f9c1263 (https://catagolue.appspot.com/object/xq ... 4iz5ar6i7e). It can evoke a glider from a snowflake similarly to the G, but in a stable manner. So it can pass a snowflake at any time to shoot a G:

Code: Select all

x = 21, y = 23, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
12bo$10b5o$10bo3bo$9b2obob2o$10bo3bo$10b5o$12bo3$4b3o$bo2bobo$3o3bo$o
2bob3o9$18b3o$18bobo!

EDIT1:
With an extra nice reaction, sending an R on the other side of the snowflake which is replaced, but xq4_1f9c1263 is turning into a standard flotilla:

Code: Select all

x = 53, y = 114, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
4$24bo$22b5o$22bo3bo$15b3o3b2obob2o$12bo2bobo4bo3bo$11b3o3bo4b5o$11bo
2bob3o5bo10$33bo$31b5o3b3o$31bo3bo3bobo2bo$30b2obob2o2bo3b3o$31bo3bo2b
3obo2bo$31b5o$33bo23$39b3o$39bobo2bo$39bo3b3o$38b3obo2bo26$39b3o$39bob
o2bo$39bo3b3o$38b3obo2bo25$39b3o$39bobo2bo$39bo3b3o$38b3obo2bo!

EDIT2:
A 3R flotilla/salvo can shift a snowflake 6 bits sideways plus shooting a glider.

Code: Select all

x = 39, y = 17, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
18bo$16b5o$16bo3bo$15b2obob2o$16bo3bo$16b5o$18bo5b3o$24bobo$24bo$11bo$
10b3o15bo$10bo2bo13b3o$26bo2bo!

[dani]

Didn't know about that ship, nice!

I found a stabilization to that one p2 agar:

Code: Select all

x = 132, y = 29, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
6b2o116b2o$7bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2b
o2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo$o4bo120b
o4bo$2bo2b3o2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2b3o2bo$b2obo
2bo116bo2bob2o$3bo6bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo6bo$b2o
bo2bo5b4o2b4o2b4o5b4o2b4o2b4o5b4o17b4o14b4o8b4o5b4o5bo2bob2o$2bo2b3o2b
o2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2b3o2bo$ob2obo120bob2obo$2bo
2b3o2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2b3o2bo$b2obo2bo5b4o
17b4o8b4o5b4o17b4o14b4o8b4o5b4o5bo2bob2o$3bo6bo2bo2bo2bo2bo2bo2bo2bo2b
o2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2bo2bo2bo2bo2bo2bo6bo$b2obo2bo116bo2bob2o$2bo2b3o2bo2bo2bo2bo2bo2bo2bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2bo2bo2bo2bo2bo2bo2bo2bo2b3o2bo$ob2obo7b4o8b4o5b4o8b4o5b4o17b4o14b4o2b
4o2b4o5b4o7bob2obo$2bo2b3o2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2b
o2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2b3o2bo$b2obo2bo116bo2bob2o$3bo6bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2bo2bo6bo$b2obo2bo5b4o8b4o5b4o8b4o5b4o17b4o20b4o20bo2bob2o$2bo2b3o2bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2b3o2bo$ob2obo120bob2obo$2bo
2b3o2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2b3o2bo$b2obo2bo5b4o2b
4o2b4o5b4o2b4o2b4o5b4o2b4o2b4o5b4o2b4o2b4o8b4o11b4o5bo2bob2o$3bo6bo2bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo6bo$b2obo2bo116bo2bob2o$2bo2b3o
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2b3o2bo$o4bo120bo4bo$7bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo
2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo2bo$6b2o116b2o!

EDIT: There's probably a breeder based on your component, it looks a bit tricky and I don't know how to insert R's like that:

Code: Select all

x = 198, y = 29, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
b2o20b2o20b2o20b2o20b2o$2bo21bo21bo21bo21bo$3o2bo16b3o2bo16b3o2bo16b3o
2bo16b3o2bo$6bo21bo21bo21bo21bo$6b2o20b2o20b2o20b2o20b2o$5b2o20b2o20b
2o20b2o20b2o4$12b3o19b3o19b3o19b3o19b3o$14bo21bo21bo21bo21bo$13b2o20b
2o20b2o20b2o20b2o3$104bo9bo9bo9bo9bo9bo9bo9bo9bo9bo$102b5o5b5o5b5o5b5o
5b5o5b5o5b5o5b5o5b5o5b5o$102bo3bo5bo3bo5bo3bo5bo3bo5bo3bo5bo3bo5bo3bo
5bo3bo5bo3bo5bo3bo$b3o19b3o19b3o19b3o19b3o9b2obob2o3b2obob2o3b2obob2o
3b2obob2o3b2obob2o3b2obob2o3b2obob2o3b2obob2o3b2obob2o3b2obob2o$3bo21b
o21bo21bo21bo10bo3bo5bo3bo5bo3bo5bo3bo5bo3bo5bo3bo5bo3bo5bo3bo5bo3bo5b
o3bo$2b2o20b2o20b2o20b2o20b2o10b5o5b5o5b5o5b5o5b5o5b5o5b5o5b5o5b5o5b5o
$104bo9bo9bo9bo9bo9bo9bo9bo9bo9bo2$16bo21bo21bo21bo21bo$17bo21bo21bo
21bo21bo$17b2o20b2o20b2o20b2o20b2o$16b2o20b2o20b2o20b2o20b2o$12b2o20b
2o20b2o20b2o20b2o$13bo21bo21bo21bo21bo$11b3o19b3o19b3o19b3o19b3o!

[2718281828]

EDIT: There's probably a breeder based on your component, it looks a bit tricky and I don't know how to insert R's like that:

Code: Select all

RLE

I think this 2R option is easier:

Code: Select all

x = 19, y = 15, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
11bo$9b5o$9bo3bo$8b2obob2o$3b3o3bo3bo$3bobo3b5o$5bo5bo3$3bo$2b3o$2bo2b
o!

There is also a stable 2 signal to three signal converter:

Code: Select all

x = 31, y = 28, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
2$11bo$13bo$12b2o3$12bo$10b5o$10bo3bo$9b2obob2o$10bo3bo$10b5o$12bo5bo$
16b5o$16bo3bo5bo$15b2obob2o2bo$4b3o9bo3bo3b2o$4bobo9b5o$6bo11bo2$10b3o
$10bobo$10bo!

All the snowflakes-interaction I found so far:

Code: Select all

x = 212, y = 243, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
17$109bo$75bo31b5o$73b5o29bo3bo25bo$73bo3bo28b2obob2o22b5o$72b2obob2o
22b3o3bo3bo23bo3bo$73bo3bo23bobo3b5o22b2obob2o$73b5o25bo5bo25bo3bo$75b
o5b3o51b5o$81bobo53bo$81bo19bo$68bo31b3o$67b3o15bo14bo2bo25b3o$67bo2bo
13b3o39bo2bobo$83bo2bo38b3o3bo$125bo2bob3o24$73bo$71b5o$39bo31bo3bo24b
2o$37b5o28b2obob2o24bo$37bo3bo9b2o13b3o2bo3bo3b3o17b3o88bo$36b2obob2o
7b2o14bobo2b5o3bobo106b5o$32b3o2bo3bo9bo14bo6bo5bo22bo85bo3bo$32bobo2b
5o10bo47b5o73bo8b2obob2o$34bo4bo60bo3bo72b3o2b3o3bo3bo$99b2obob2o71bo
2bobobo3b5o$100bo3bo85bo$100b5o$102bo$182bo2$177bo$176b3o2bo2bo$176bob
o$177bo2$102bo$101b3o$101bobo$102bo17$78bobo18b2o10b2o$32bo45b3o19bo
11bo$30b5o22bo40b3o9b3o$30bo3bo20b5o20bo$8bo5bo14b2obob2o19bo3bo18b5o$
6b5o3bobo7b3o3bo3bo19b2obob2o17bo3bo18bo11bo$6bo3bo3b3o7bobo3b5o14b3o
3bo3bo17b2obob2o15b5o7b5o$5b2obob2o14bo5bo16bobo3b5o12b3o3bo3bo16bo3bo
7bo3bo$6bo3bo40bo5bo14bobo3b5o11b3ob2obob2o5b2obob2ob3o$6b5o19b3o41bo
5bo13bobo2bo3bo7bo3bo2bobo$8bo21bobo22b3o41b5o7b5o$30bo24bobo43bo11bo$
8b3o$8bobo$10bo12$183bo$181b5o$143b2o36bo3bo$144bo35b2obob2o$142b3o30b
3o3bo3bo$175bobo3b5o$145bo31bo5bo$143b5o$143bo3bo6b3o27b2o$6bo49bo85b
2obob2o5bo29bo$8bo49bo35b2o47bo3bo6b2o28b3o$7b2o48b2o36bo47b5o$93b3o
49bo2$8bo5bo43bo$6b5o3bobo39b5o35bo$6bo3bo3b3o39bo3bo33b5o$5b2obob2o
43b2obob2o32bo3bo$6bo3bo39b3o3bo3bo28b3ob2obob2o2b2o$6b5o39bobo3b5o28b
obo2bo3bo3bo$8bo43bo5bo35b5o5bo$96bo$6b3o47b3o$6bobo47bobo$6bo2$75bobo
$75b3o2$77bo$75b5o$31bo43bo3bo$33bo40b2obob2o$32b2o35b3o3bo3bo28b2o$
69bobo3b5o29bo$71bo5bo29b3o$32bo$30b5o$30bo3bo42b2o31bo$29b2obob2o41bo
30b5o$30bo3bo44bo23b2o3bo3bo86bobo$30b5o69bo2b2obob2ob3o81b3o$32bo5bo
63bo5bo3bo2bobo$36b5o67b5o$36bo3bo5bo63bo$35b2obob2o2bo$24b3o9bo3bo3b
2o$24bobo9b5o$26bo11bo2$30b3o110b2o$30bobo111bo$30bo111b3o$191b2o$192b
o$145bo44b3o$143b5o$143bo3bo$142b2obob2o44bo$143bo3bo43b5o$143b5o43bo
3bo$145bo44b2obob2o$191bo3bo$191b5o$193bo$152b3o$152bobo20$20bo43bo$
13b3o2b5o34b3o2b5o$15bo2bo3bo8b2o26bo2bo3bo8b2o$14b2ob2obob2o6b2obo24b
2ob2obob2o6b2obo$18bo3bo8b2o29bo3bo8b2o$18b5o39b5o$20bo35bo7bo$54b5o$
54bo3bo$53b2obob2o$54bo3bo$54b5o$56bo3$55b2o$55bo$57bo!

[LaundryPizza03]

The average density from an unbounded random field is about 4.01%, based on a random 10000*10000 soup run for 633 generations.

[LaundryPizza03]

Most of the 300K+ pathologicals in this rule seem to be O(sqrt(t)) bounding box patterns, like the one mentioned in the first post. Some examples:

Code: Select all

x = 16, y = 16, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
b2o2b5ob2o2bo$o9b5o$3b4ob3ob4o$o3bob4ob2ob2o$bo3bo3bo$obo3b2ob2o2bobo$
2o3bo3bo2b4o$2obo2bo4b2o$4obo4b2o2bo$3b2o2b2obob2obo$bobo5bo5bo$ob3obo
2b2o2b2o$3o3bobo5b2o$2ob2o4bo4bo$2o2bob3o2b2o$obo2b4o4b3o!

Code: Select all

x = 31, y = 31, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
bo2b2o3b5obo$3o2b3o2bo2b2o$2b5ob3o3bo$bobob5obo2bo$5ob4o2bobo$5b7o2bo$
3o3bo2b3o3bo$ob3ob2ob2o2bo$o2bo2bo2bob2o2bo$2o3b3o3bo$3b2o3bo3b3o$4obo
2b3o2b2o$bob2o4b3ob2o$3bo3bob2o3bo$ob2o2bob4obo$2b3obobobobob3obobobob
ob3o$17bob4obo2b2obo$16bo3b2obo3bo$16b2ob3o4b2obo$16b2o2b3o2bob4o$16b
3o3bo3b2o$19bo3b3o3b2o$15bo2b2obo2bo2bo2bo$17bo2b2ob2ob3obo$15bo3b3o2b
o3b3o$16bo2b7o$16bobo2b4ob5o$16bo2bob5obobo$16bo3b3ob5o$16b2o2bo2b3o2b
3o$15bob5o3b2o2bo!

Code: Select all

x = 31, y = 31, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
3b3ob3o2bob3obo2b2o4b3o$o3bob6ob4ob2o2b7o$6o4bo4b3obob3o2bobo$2o2bobob
5o2b4ob2o2bo3bobo$b2obo4b3o2b3o2b4o3b5o$bo4bobo2bobo3bo2b2o2bo3bobo$bo
bobo3bobobobob2o3bo2bobobo$b2o5b4obob3o2bobo6b2o$3obob2ob2o2b3obo2bo2b
obobob2o$ob4o3b4ob5ob5ob2ob2o$3b3ob3obobobobobob3o2b4o$b2obo5bobo2b3o
2b2ob5obo$2obo2bo2bo2b2o3b3obo5bo2bo$2b2ob14obob4o3bo$5o2bobobobobobo
3b2o3bo2b2o$5ob6ob5ob6ob5o$2o2bo3b2o3bobobobobobo2b5o$bo3b4obob14ob2o$
o2bo5bob3o3b2o2bo2bo2bob2o$bob5ob2o2b3o2bobo5bob2o$b4o2b3obobobobobob
3ob3o$2ob2ob5ob5ob4o3b4obo$2obobobo2bo2bob3o2b2ob2obob3o$2o6bobo2b3obo
b4o5b2o$bobobo2bo3b2obobobobo3bobobo$obo3bo2b2o2bo3bobo2bobo4bo$5o3b4o
2b3o2b3o4bob2o$obo3bo2b2ob4o2b5obobo2b2o$2bobo2b3obob3o4bo4b6o$2b7o2b
2ob4ob6obo3bo$b3o4b2o2bob3obo2b3ob3o!

Code: Select all

x = 256, y = 1, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
2ob2o6b3o2bobob4ob2o2b8o3bob2o2bobobob5o2bo3bo6b7o5b4ob2obob3o3b2ob2ob
3o2b2obo4b2ob4o2b3ob5ob2o3b3ob2o3b2o2b2o2bo6b3o2bo5bo6bo4bo3b5o2b3o2b
2o4bo3bobo4bo2b2o2b5ob3obo3bo4bo2bob2obo!

Code: Select all

x = 31, y = 31, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
2bobobo3bobobobobobo3bobobo$2b2o4bo5b3o5bo4b2o$4obobo4b2obob2o4bobob4o
$b2obobo3b3ob3ob3o3bobob2o$o2b2o2b6obobob6o2b2o2bo$2bo6b3o2b3o2b3o6bo$
o2bo4bo3bo5bo3bo4bo2bo$2bobo7b2o3b2o7bobo$bo2bobobobo2b5o2bobobobo2bo$
4b2o5bo2bobo2bo5b2o$o2b3o2bo4bo3bo4bo2b3o2bo$3b3o3bo2bobobobo2bo3b3o$o
b3ob2o3bo2b3o2bo3b2ob3obo$2bo4b2obo3b3o3bob2o4bo$2ob3o2b2ob4ob4ob2o2b
3ob2o$b3obo2bo3b2o3b2o3bo2bob3o$2ob3o2b2ob4ob4ob2o2b3ob2o$2bo4b2obo3b
3o3bob2o4bo$ob3ob2o3bo2b3o2bo3b2ob3obo$3b3o3bo2bobobobo2bo3b3o$o2b3o2b
o4bo3bo4bo2b3o2bo$4b2o5bo2bobo2bo5b2o$bo2bobobobo2b5o2bobobobo2bo$2bob
o7b2o3b2o7bobo$o2bo4bo3bo5bo3bo4bo2bo$2bo6b3o2b3o2b3o6bo$o2b2o2b6obobo
b6o2b2o2bo$b2obobo3b3ob3ob3o3bobob2o$4obobo4b2obob2o4bobob4o$2b2o4bo5b
3o5bo4b2o$2bobobo3bobobobobobo3bobobo!

[2718281828]

Most of the 300K+ pathologicals in this rule seem to be O(sqrt(t)) bounding box patterns, like the one mentioned in the first post. Some examples:

The third example utilises a fast reflection reaction that I think we did not notice so far.
It allows for small period 15+4N , G-shuttles, the p15 version:

Code: Select all

x = 30, y = 47, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
3$13bo$9bo9bo$11bo5bo$10b2o5b2o2$13bobo9bo$13b3o7bo$5bo17b2o$7bo7bo$6b
2o5b5o2b2o4bo$13bo3bo2bo$9b2ob2obob2ob2o$10bo2bo3bo$4bo4b2o2b5o5b2o$
15bo7bo$6b2o17bo$7bo7b3o$5bo9bobo2$12b2o5b2o$13bo5bo$11bo9bo$17bo7$24b
o$22bo$4bo17b2o$6bo7bo$5b2o5b5o2b2o4bo$12bo3bo2bo$8b2ob2obob2ob2o$9bo
2bo3bo$3bo4b2o2b5o5b2o$14bo7bo$5b2o17bo$6bo$4bo!

Moreover, some more 3G-snowflake reactions:

Code: Select all

x = 157, y = 153, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
5$40bo$38b5o$33b2o3bo3bo3b2o$32bob2ob2obob2ob2obo$33b2o3bo3bo3b2o$38b
5o$40bo2$34bo11bo$36bo7bo$35b2o7b2o$39b3o$39bobo3$8bo$6b5o$b2o3bo3bo3b
2o$ob2ob2obob2ob2obo$b2o3bo3bo3b2o$6b5o$8bo4$8bo$7b3o$7bobo$8bo4$26bo$
24b5o$19b2o3bo3bo3b2o$18bob2ob2obob2ob2obo$19b2o3bo3bo3b2o$24b5o$26bo
6$26bo$25b3o$25bobo$26bo8$46bo$44b5o$39b2o3bo3bo3b2o$38bob2ob2obob2ob
2obo$39b2o3bo3bo3b2o$44b5o$46bo8$46bo$45b3o$45bobo$46bo5$72bo$70b5o$
65b2o3bo3bo3b2o$64bob2ob2obob2ob2obo$65b2o3bo3bo3b2o$70b5o$72bo10$72bo
$71b3o$71bobo$72bo3$97bo$95b5o$90b2o3bo3bo3b2o$89bob2ob2obob2ob2obo$
90b2o3bo3bo3b2o$95b5o$97bo3$95b3o$95bobo6$119bo$117b5o$112b2o3bo3bo3b
2o$111bob2ob2obob2ob2obo$112b2o3bo3bo3b2o$117b5o$119bo6$117b3o$117bobo
$141bo$139b5o$134b2o3bo3bo3b2o$133bob2ob2obob2ob2obo$134b2o3bo3bo3b2o$
139b5o$141bo6$140bo$139b3o$139bobo$140bo!

[2718281828]

We have quite a few different c/3-speed spaceships (strictly 2c/6 ships). But if I got it correctly, this one was not mentioned so far:

Code: Select all

x = 26, y = 9, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
11bo7bo$10b3o5b3o$11bo2b3o2bo$10b3obobob3o$11bobobobobo$14bobo$7b3o11b
3o$7b3ob3o3b3ob3o$8bo2bobo3bobo2bo!

It pulls two G's at the tail. It also has some tail spark.
There is also this more or less trivial interaction which removes the spark:

Code: Select all

x = 65, y = 51, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
4$27bo7bo16bo7bo$26b3o5b3o14b3o5b3o$27bo2b3o2bo16bo2b3o2bo$26b3obobob
3o14b3obobob3o$27bobobobobo16bobobobobo$30bobo22bobo$23b3o11b3o8b3o11b
3o$23b3ob3o3b3ob3o8b3ob3o3b3ob3o$24bo2bobo3bobo2bo10bo2bobo3bobo2bo5$
7b3o5b3o8b3o5b3o$7bobo2bo2bobo8bobo2bo2bobo$6bobo2b3o2bobo6bobo2b3o2bo
bo$7bo9bo8bo9bo$7b2o7b2o8b2o7b2o$10bo3bo14bo3bo$10bo3bo14bo3bo$9b3ob3o
12b3ob3o$10bo3bo14bo3bo$9bo2bo2bo12bo2bo2bo$10bo3bo14bo3bo$10bobobo14b
obobo$9b3ob3o12b3ob3o$10bo3bo14bo3bo2$12bo18bo4$11b3o16b3o$6bo5bo5bo6b
o5bo5bo$5b3o3bobo3b3o4b3o3bobo3b3o$11b3o16b3o$11bobo16bobo$7b2o3bo3b2o
8b2o3bo3b2o$5b2o4b3o4b2o4b2o4b3o4b2o3$7b3ob3ob3o8b3ob3ob3o$7bo2bo3bo2b
o8bo2bo3bo2bo$8bo7bo10bo7bo$8b2o5b2o10b2o5b2o!

[dani]

Said shuttle works with 3 gliders:

Code: Select all

x = 18, y = 23, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
9bo$5bo9bo$7bo5bo$6b2o5b2o2$9bobo$9b3o$bo$3bo7bo$2b2o5b5o$9bo3bo$5b2ob
2obob2o$6bo2bo3bo$o4b2o2b5o$11bo$2b2o$3bo7b3o$bo9bobo2$8b2o5b2o$9bo5bo
$7bo9bo$13bo!

Possible p17 gun engine:

Code: Select all

#C [[ STOP 17 ]]
x = 6, y = 13, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
2b2o$2bo2bo$o3b2o$2o3$bobo3$2o$o3b2o$2bo2bo$2b2o!

[2718281828]

Possible p17 gun engine:

Code: Select all

#C [[ STOP 17 ]]
x = 6, y = 13, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
2b2o$2bo2bo$o3b2o$2o3$bobo3$2o$o3b2o$2bo2bo$2b2o!

At least it gives a nice p17 which provides a quite unusual spark:

Code: Select all

x = 20, y = 13, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
8b2o$6bo2bo9bo$o5b2o3bo$2bo7b2o$b2o2$8bobo3$10b2o$6b2o3bo$6bo2bo$8b2o!

EDIT1:
This is close, but still not working:

Code: Select all

x = 23, y = 45, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
10b2o$8bo2bo$2bo5b2o3bo$4bo7b2o$3b2o2$10bobo3bo3$12b2o$8b2o3bo$8bo2bo$
10b2o5$10b2o$8bo2bo$2bo5b2o3bo$4bo7b2o4bo$3b2o2b2o$6bo11b2o$10bobo6bo$
18b2o$8bo$12b2o$8b2o3bo$8bo2bo$10b2o4$8bo2bo$6bo6bo$6b2o4b2o3$3bo6bobo
3bo3$6b2o4b2o$6bo6bo$8bo2bo!

The dot has to be inserted very quickly, somehow by oscillators which interact with some still lifes and potentially the G (as in second case) such that the dot has the right position. Alternative we could somehow side-shoot G/R's with period of 34 to insert the dot. If we have such a collision that works we can copy it with phase shift of 17 on the other side. A p34 gun should be feasible, but p17 is hard.
The third figure is a reduction of this oscillator, esp. this p17 sparker.

[2718281828]

I noticed a nice property of xq4_1f9c1263 (https://catagolue.appspot.com/object/xq ... 4iz5ar6i7e). It can evoke a glider from a snowflake similarly to the G, but in a stable manner. So it can pass a snowflake at any time to shoot a G:

Code: Select all

x = 21, y = 23, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
12bo$10b5o$10bo3bo$9b2obob2o$10bo3bo$10b5o$12bo3$4b3o$bo2bobo$3o3bo$o
2bob3o9$18b3o$18bobo!

An almost synthesis of xq4_1f9c1263:

Code: Select all

x = 16, y = 12, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
10b2o$10bo$10b3o$5b3o$2bo2bobo6b2o$b3obo8bo$o2bo5bo4b2o2$9bo2$6b3o$6bo
bo!

Maybe someone can finish it.
EDIT1:
The synthesis would likely allow as to build a N^(3/2) growth pattern. We take sqrt-guns for G/R's and build a sqrt-gun for xq4_1f9c1263 and a G sqrt-gun a couple of ticks later. Both spaceship pass a snowflake chain, as used in a breeder. This should give us linear times sqrt-growth.

[BlinkerSpawn]

...it's something, I guess:

Code: Select all

x = 21, y = 41, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
16bo$10bo9bo$12bo5bo$11b2o5b2o9$14bobo$14b3o2$5bo2bo$3bo6bo$3b2o4b2o2$
6bo3bo$o5bo3bo$6bo3bo2$3b2o4b2o$3bo6bo$5bo2bo12$11b2o5b2o$12bo5bo$10bo
9bo$16bo!

[toroidalet]

smaller p33 gun:

Code: Select all

x = 34, y = 38, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
19b2o$19bo2bo$18b2ob2o$20bo4$16b2o3b2o$17bo3bo$17b5o$19bo$b2o$2bo$o6bo
6b2o$3o12bo$2bo4bo6b2o4$18bobo$18b3o2$31bo$24b3o4b3o$14bo18bo$13b3o15b
o$13bobo15b2o$14bo2$11b2o3b2o$12bobobo$12b5o$14bo2$13bo$11b2ob2o$11bo
2bo$13b2o!

it seems unlikely that the double-r flotilla can be turned into the useful one because it's really hard to insert anything at all that close behind the ships, but here's a synthesis (cleanup not included):

Code: Select all

x = 37, y = 54, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
36bo$34bobo$34b3o8$31bo$29bobo$29b3o6$2o$bo$2o2$12b2o$13bo$12b2o$34bo$
33b3o$33bobo$34bo5$22b3o$22bobo$22bo6$17b3o$17bobo$17bo7$27bo$26b3o$
26bobo$27bo!

here's a really hard but possible synthesis for xq4_1f9c1263 (the hard part is the interaction between the snowflake predecessor and the glider):

Code: Select all

x = 8, y = 18, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
2o$b2o$bo3bobo$o4b3o2$bo4bo2$6bo5$6bo4$5b3o$5bobo!

[2718281828]

...it's something, I guess:

Code: Select all

x = 21, y = 41, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
16bo$10bo9bo$12bo5bo$11b2o5b2o9$14bobo$14b3o2$5bo2bo$3bo6bo$3b2o4b2o2$
6bo3bo$o5bo3bo$6bo3bo2$3b2o4b2o$3bo6bo$5bo2bo12$11b2o5b2o$12bo5bo$10bo
9bo$16bo!

Yes, it works for all periods 51+17N, this is the 51 version:

Code: Select all

x = 20, y = 33, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
12bo3bo3$14bo2$11bo5bo$14bo$10bo7bo4$4bo2bo$2bo6bo$2b2o4b2o$13bobo$3ob
2o7b3o$4b2o2bo$3ob2o2$2b2o4b2o$2bo6bo$4bo2bo4$10bo7bo$14bo$11bo5bo2$
14bo3$12bo3bo!

EDIT1:
Somehow related, p17 and p34 dot-hassler:

Code: Select all

x = 76, y = 51, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
2$52bo2bo3b2o$50bo6bobo2bo$50b2o4b2o3b2o2$50b2ob2o$8bo2bo3b2o30bo2bobo
bo$6bo6bobo2bo7bo2bo20b2ob2o8bo2b2o$6b2o4b2o3b2o5bo6bo34bo$14bo9b2o4b
2o18b2o4b2o10bo$50bo6bo9b2o$3bo4bobobobo3bo8b2ob2o20bo2bo$27bobobo2bo
24b2ob3ob2o$14bo12b2ob2o28bo5bo$6b2o4b2o44bo9bo$6bo6bo5b2o3b2o4b2o$8bo
2bo7bo2bobo6bo30b3o$21b2o3bo2bo28bo9bo$60bo5bo$59b2o5b2o3$63bo3$8bo2bo
3b2o$6bo6bobo2bo7bo2bo$6b2o4b2o3b2o5bo6bo17bo2bo3b2o$24b2o4b2o15bo6bob
o2bo$6b2ob2o36b2o4b2o3b2o$3bo2bobobo7bo8b2ob2o$6b2ob2o16bobobo2bo12b2o
b2o$27b2ob2o12bo2bobobo7bo$6b2o4b2o33b2ob2o11b2o$6bo6bo5b2o3b2o4b2o31b
o$8bo2bo7bo2bobo6bo15b2o4b2o10bo$21b2o3bo2bo17bo6bo9b2o$49bo2bo$56b2o
5b2o$57bo5bo$55bo9bo$59b3o$59bobo$55bo4bo4bo$57bobobobo$56b2ob3ob2o3$
60bo!

[dani]

:c , I guess the p19 gun will be our lowest period until someone (winks and nods) does something with it.
In other news, here's a 3-dot side-push reaction with R's:

Code: Select all

x = 40, y = 7, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
b2o$2bo4b3o23bo2bo2bo$3o6bo5b2o$8b2o6bo4b3o$14b3o6bo5b2o$22b2o6bo$28b
3o!

It allows 2 off-R-phase dots to travel up with the central on-R-phase dot
Toroidalet's synthesis inspired a partial pull reaction:

Code: Select all

x = 15, y = 7, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
11bo$9b5o$b2o6bo3bo$ob2obo2b2obob2o$b2o6bo3bo$9b5o$11bo!

I hope it can be completed. Here's another cool thing:

Code: Select all

x = 14, y = 7, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
11bo$9b5o$b2o6bo3bo$ob2obo2b2obobo$b2o6bo3bo$9b5o$11bo!

[2718281828]

:c , I guess the p19 gun will be our lowest period until someone (winks and nods) does something with it.

Well, we can construct G (pseudo) guns for all periods >=15, and with the 3G to R collision with a repeat time of 16, we would directly have R guns for all periods>=16. Of course, just pseudo guns, but guns. I also think that a p14 gun is possible, we just need a suitable p14 sparking oscillator as I described somewhere above.

EDIT1:
Maybe this will help us for a p17 gun:

Code: Select all

x = 16, y = 11, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
5bo2bo$3bo6bo4bo$3b2o4b2o2b3o$13bo$6bo3bo$o5bo3bo$6bo3bo3bo2$3b2o4b2o$
3bo6bo$5bo2bo!

Code: Select all

x = 17, y = 12, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
5bo2bo$3bo6bo4bo$3b2o4b2o2b3o$13bo$6bo3bo$o5bo3bo$6bo3bo2$3b2o4b2o$3bo
6bo$5bo2bo4b3o$13bobo!

[2718281828]

here's a really hard but possible synthesis for xq4_1f9c1263 (the hard part is the interaction between the snowflake predecessor and the glider):

Code: Select all

x = 8, y = 18, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
2o$b2o$bo3bobo$o4b3o2$bo4bo2$6bo5$6bo4$5b3o$5bobo!

It works out (5G+3R for the messy version, and 3 additional G for clean up):

Code: Select all

x = 88, y = 42, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
2$18bo45bo$17bobo43bobo$17b3o43b3o$18bo45bo2$46b2o$47bo$4b3o39b2o2b3o$
6bo45bo$5b2o44b2o2$10bobo43bobo$10b3o43b3o$24b2o44b2o$24bo45bo$10b2o
12b3o29b2o12b3o$11bo45bo$10b2o36b2o6b2o$49bo$48b2o$85bo$28b2o44b2o$15b
3o10bo32b3o10bo10bo$15bobo10b3o30bobo10b3o4$85b2o$85bo$87bo$15b3o43b3o
$15bobo43bobo5$57b3o$57bobo!

The only problem is that I just got it to work with a front R, so the repeat time in a gun would be large using this recipe, depending on how fast we can insert the R.
I tried it also with:

Code: Select all

x = 30, y = 27, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
19bo$18bobo$18b3o$9b3o7bo$11bo$10b2o3$14bo$16b2o$17bo$15b3o3$20b3o$20b
obo2$25b2o$25bo$25b3o3$16b3o$16bobo!

but it did not work out to create the dot behind the R fast enough.

EDIT1:
It also works for the fast version:

Code: Select all

x = 28, y = 33, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
15bo$14bobo$14b3o$15bo4$b3o$3bo$2b2o3$2o$bo6b2o$2o7bo$7b3o6$25b2o$16b
3o6bo$16bobo6b3o7$12b3o$12bobo!

EDIT2:
The fastest clean up I found so far has a repeat time of 46, so a p46 xq4_1f9c1263-gun seems to be realistic:

Code: Select all

x = 97, y = 92, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
51bo$50bobo$50b3o$44bobo4bo$44b3o9$47bo$46bobo$46b3o$47bo7$51bo$50bobo
$50b3o$44bobo4bo$44b3o9$47bo$46bobo$46b3o$47bo4$11b2o20b3o$12bo22bo$
10b3o21b2o3$9b2o21b2o$10bo5b3o14bo6b2o$9b2o7bo13b2o7bo$b2o14b2o5b2o13b
3o$ob2o19bob2o$b2o4b2o15b2o4b2o$6bob2o19bob2o$7b2o21b2o$92bo$57b2o21b
3o$48b3o6bo22bo11bo$48bobo6b3o20b2o4$92b2o$92bo$94bo$44b3o$44bobo14$
48b3o$48bobo7$44b3o$44bobo!

But getting the required p46 salvo's (esp. the 3G+2R back salvo) is likely a hard job.

And something different which I found accidentally, a flotilla synthesis:

Code: Select all

x = 14, y = 11, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
10bo$2b3o5bobo$4bo5b3o$3b2o3$b2o$2bo6b2o$b2o7bo$8b3o!

EDIT3:

Code: Select all

x = 74, y = 80, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
31bobo$31b3o5bo$38bobo$38b3o$39bo7$36bo$35bobo$35b3o$36bo4$31bobo$31b
3o5bo$38bobo$38b3o$39bo7$36bo$35bobo$35b3o$36bo4$4b3o15b3o$6bo17bo$5b
2o16b2o3$3b2o16b2o$4bo6b2o9bo6b2o$3b2o7bo8b2o7bo$10b3o15b3o$b2o16b2o$
2bo17bo$3o15b3o2$52b2o16b2o$46b2o4bo11b2o4bo$37b3o6bo5b3o9bo5b3o$37bob
o6b3o15b3o7$33b3o$33bobo9$37b3o$37bobo7$33b3o$33bobo!

A modified clean up, 6G+5R (the back 3R+G is likely most difficult to achieve, but likely easier than the thing above). Minimal repeat time is 36. But am sceptical if we will manage to construct a p36 xq4_1f9c1263-gun.

[2718281828]

A minimal sqrt-gun with 24 cells with the restriction that only one bounding box component is growing, so it is really suitable for a gun:

Code: Select all

x = 17, y = 48, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
9bo12$7bobo2$14bobo$7bobo5$3bo2$o$13bobo$o$7bobo$13bobo4$2bo3$4bobo8$
13bo3$9bo3$11bobo!

Without the bounding box restriction, I have a 17 cell solution:

Code: Select all

x = 32, y = 32, rule = B2ci3ai4c8/S02ae3eijkq4ciz5ar6i7e8
29bo11$o7bo2bo17bobo$22bobo$8bo2bo$29bobo9$29bobo5$25bo3$27bobo!

Also an update for the quadratic growth, 55 cells:

Code: Select all

x = 32, y = 54, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
$15bobo3$19bo5$14bobo6$5bo$18bo$2bo$11bo5bo$2bo$17bo2bo$6bo2bo$30bo$9b
o20b2o$29b2o$8bo$14bobo$29b2o$5bo24bo$29b2o$2bo$11bo$2bo20bo$12bobo11b
obo$16bo$16bo$12bo2b3o10bo$16bo7$9bobo4bobo3$9bobo2$21bo3$17bobo!

EDIT1: the minimal sqrt-gun. (two growing sides of the box).

[BlinkerSpawn]

A minimal sqrt-gun with 24 cells with the restriction that only one bounding box component is growing:

Code: Select all

x = 17, y = 48, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
9bo12$7bobo2$14bobo$7bobo5$3bo2$o$13bobo$o$7bobo$13bobo4$2bo3$4bobo8$
13bo3$9bo3$11bobo!

This one breaks due to some timing constraint I couldn't work out, but this 25-cell one works:

Code: Select all

x = 17, y = 48, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
9bo12$7bobo2$14bobo$7bobo5$3bo2$o$13b3o$o12bobo$7bobo5$2bo3$4bobo5$13b
o2$13bo4$9bo3$11bobo!

[2718281828]

A minimal sqrt-gun with 24 cells with the restriction that only one bounding box component is growing:

Code: Select all

x = 17, y = 48, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
9bo12$7bobo2$14bobo$7bobo5$3bo2$o$13bobo$o$7bobo$13bobo4$2bo3$4bobo8$
13bo3$9bo3$11bobo!

This one breaks due to some timing constraint I couldn't work out, but this 25-cell one works:

Code: Select all

x = 17, y = 48, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
9bo12$7bobo2$14bobo$7bobo5$3bo2$o$13b3o$o12bobo$7bobo5$2bo3$4bobo5$13b
o2$13bo4$9bo3$11bobo!

Thanks to BlinkerSpawn and danny for correcting me. Yes, the 23 cell version I published does not work. Therefore I made a file 24-cell version but published the wrong one (but I mentioned 24 cell ).
Here we go with 24 cells:

Code: Select all

x = 17, y = 48, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
9bo12$7bobo2$14bobo$7bobo5$3bo2$o$13bobo$o$7bobo$13bobo4$2bo3$4bobo6$
12bo$15bo4$9bo3$11bobo!

[2718281828]

I'd wonder if there is a 5-cell pattern which last more than 100 generations

116M:

Code: Select all

RLE

A methuselah collection, with a surprisingly long lasting:
#2: 5M
#3: 21M
#4: 71M
#5: 120M

Code: Select all

x = 33, y = 64, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
o2bo17bobo16$o2bo2$21bob2o$o18$o2bo$28bo2$o2bo17bob2o17$o2bo$28bo2$o2b
o17bob2o7bo3$o!

For some reason all the pattern are nested in each other.

[2718281828]

Does anybody have the longest c/4 partial? That first methuselah I shared looks like a potential frontend

As you were asking within the methuselah post.

The longest partial using ntzfind with width 11 appears (as almost always in this rule) at odd-symmetric symmetry.
The ntzfind-length is only 43 (11 real), so c/4 is not very promising.

Code: Select all

x = 21, y = 11, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
10bo3$6bo2b3o2bo$9bobo2$bo3b3o2bo2b3o3bo$7bob3obo$4b3o2b3o2b3o$3obobo
2bobo2bobob3o$o5b9o5bo!

2c/8 at width 10 is not really better:

Code: Select all

x = 13, y = 12, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
$6bo3$2bo2b3o2bo$5bobo2$b3o2bo2b3o$3bob3obo$3o2b3o2b3o$obo2bobo2bobo$
2b9o!

I think promising speeds to find ships are c/7 and maybe 9c/21 as we have the fuses.

In general all 1/Kc and 2/Kc for large K (roughly >= 10) seem to be promising and we know, that 2c/19 works.
Here my most promising partials (2c/13 at w7, c/11 at w8, c/16 at w5, c/17 at w5):

Code: Select all

x = 76, y = 35, rule = B2ci3ai4c8/S02ae3eijkq4iz5ar6i7e
9bo22bo19bo16bo$31bobo17bobo13b5o$30b5o15b5o12bo3bo$7bobobo19bobo17bob
o12b2obob2o$8b3o21bo19bo14bo3bo$9bo57b5o$69bo$6b2obob2o16bo5bo13bo5bo
11bo3bo$7bo3bo20bo19bo16bo$7bo3bo15bo3b3o3bo30bobo$8b3o19b5o33b3o$30bo
bobo17bo$7b5o15bo9bo14bo$7b2ob2o18bobobo17bo15b3o$31bobo17b3o14bobo$6b
o5bo18b3o18bo14bo3bo$9bo21b3o$32bo36bo$31b3o35bo$52bo14b5o$51b3o13bo3b
o$6b3ob3o38bobo15bo$6bob3obo15b2ob3ob2o14b3o$6b3ob3o15b9o$31b3o18bo16b
o$31b3o$5bo7bo34bo3bo3bo9bo5bo$4b2o3bo3b2o36bobo14b3o$5bo2b3o2bo36bo3b
o14bo$8b3o38b2o3b2o$4b3obobob3o$6bo5bo$3bobo7bobo$3bo2bo5bo2bo$8b3o!