Thread for small diehards

For discussion of specific patterns or specific families of patterns, both newly-discovered and well-known.
Post Reply
Pyry Virtanen
Posts: 32
Joined: July 22nd, 2022, 2:09 pm

Thread for small diehards

Post by Pyry Virtanen » June 26th, 2023, 1:13 pm

There has been some talk about making a list of the best known small diehards, but no-one has actually posted such a list, so I'm doing it here.
dvgrn wrote:
August 17th, 2017, 8:04 am
I don't think there's a List of Record-Breaking Diehards on the wiki yet, but there's no harm in putting together a trial version and seeing if it gets contributions or adjustments from other people.
These are the longest-lasting diehards for each population size from 1 to 17 I know about. There has to be some sort of bounding box restriction to avoid degenerate glider+blonk constructions and similar, so each pattern on this list fits into a 2Nx2N bounding box, where N is the pattern's population. Many of these can probably be improved.

1 cell --> 1 tick, trivial
2 cells --> 1 tick, trivial
3 cells --> 2 ticks, e.g. banana spark
4 cells --> 5 ticks, the four 4-cell parents of the P-pentomino
5 cells --> 8 ticks, if my math is correct, there are seven different 5-cell patterns that vanish in 8 ticks, all parents of the same heptomino (Edit: Actually I later found that there are at least 15 of such patterns, not just seven). Here is one of them:

Code: Select all

x = 5, y = 2, rule = B3/S23
2o$2b3o!
6 cells --> 46 ticks, the eight 6-cell parents of the Z-hexomino. Here is one of them:

Code: Select all

x = 6, y = 3, rule = B3/S23
2bo$2o2b2o$3bo!
7 cells --> 130 ticks, the standard Die Hard:

Code: Select all

x = 7, y = 3, rule = B3/S23
6bo$2o2bobo$bo4bo!
8 cells --> 140 ticks, I wish I had something better:

Code: Select all

x = 7, y = 5, rule = B3/S23
bo$2o$2bobobo$5bo$5bo!
9 cells --> 229 ticks, found by edwin:

Code: Select all

x = 7, y = 4, rule = B3/S23
3bobo$bo3b2o$2bobo$2o!
10 cells --> 270 ticks:

Code: Select all

x = 5, y = 7, rule = B3/S23
2bo$bo$bo$b3o$o3bo$4bo$4bo!
11 cells --> 419 ticks:

Code: Select all

x = 8, y = 13, rule = B3/S23
3o$2bo$2bo5$6bo$6bo$6bo$6bo$6bo$7bo!
12 cells --> 495 ticks:

Code: Select all

x = 7, y = 10, rule = B3/S23
6bo$4bobo$bo4bo$2o4$2bo$b3o$bo!
13 cells --> 738 ticks:

Code: Select all

x = 22, y = 7, rule = B3/S23
16b3o$2bo18bo$b2o18bo$21bo$19bo2$3o!
14 cells --> 779 ticks, related to the previous one:

Code: Select all

x = 18, y = 9, rule = B3/S23
13bo$14bo$bo12bobo$2o13bobo$16b2o2$o$o$o!
15 cells --> 1071 ticks:

Code: Select all

x = 19, y = 8, rule = B3/S23
8bo$8b2o$17bo$o15bobo$obo14bo$3bo$2bo2b2o$5bo!
16 cells --> 1100 ticks:

Code: Select all

x = 22, y = 17, rule = B3/S23
16bo$16b2o5$10bo$10b3o4$2bo$2bo16b3o$2bo2$b2o$o!
17 cells --> 1237 ticks:

Code: Select all

x = 23, y = 27, rule = B3/S23
9bo$9bo$9bo16$2o20bo$obo18bo$bo19bo$20bo$21b2o4$9b3o!

Pyry Virtanen
Posts: 32
Joined: July 22nd, 2022, 2:09 pm

Re: Thread for small diehards

Post by Pyry Virtanen » August 9th, 2023, 7:50 am

Here are seven diehards with small bounding boxes that I have discovered. I have already posted most of them in other threads, but I decided to put them here too so that they can all be in the same place. The third one is new. These are the longest-lasting diehards for their bounding boxes I know about, but better ones very possibly exist.

5x6 bounding box, 270 ticks:

Code: Select all

x = 6, y = 5, rule = B3/S23
b2o$bo$o2bo$bo2b2o$2bob2o!
6x7 bounding box, 571 ticks:

Code: Select all

x = 7, y = 6, rule = B3/S23
obo2b2o$3o$ob2o2bo$2b3obo$2b2obo$5bo!
7x8 bounding box, 1002 ticks:

Code: Select all

x = 8, y = 7, rule = B3/S23
6bo$o5bo$4ob2o$bo4bo$2b2obobo$2o4bo$3ob3o!
8x9 bounding box, 1069 ticks:

Code: Select all

x = 9, y = 8, rule = B3/S23
o2bobob2o$ob2o$o2bobob2o$5bo2bo$6bo$2b2o2bo$2b4ob2o$4b3o!
8x11 bounding box, 1119 ticks:

Code: Select all

x = 11, y = 8, rule = B3/S23
3o$5b2o$7bo$o5bo$bob4o3bo$2bobo5bo$b2o2bo3b2o$2bo!
3x14 bounding box, 997 ticks:

Code: Select all

x = 14, y = 3, rule = B3/S23
3o6bob2o$2b2o2bo4bobo$2bob2o2b3o!
4x17 bounding box, 1067 ticks:

Code: Select all

x = 17, y = 4, rule = B3/S23
bo4b6o4bo$bo7bo3b4o$3o5bo5bobo$bob4o3b2o2b2o!
Also, years ago simeks posted these very interesting diehards, which are apparently the best possible ones for their bounding boxes. I would love to see more of these for other bounding boxes.
simeks wrote:
August 9th, 2017, 5:13 pm
The longest running die-hard in a 7-by-3 bounding box is this (214 gens):

Code: Select all

6bo$3o2b2o$3obobo!
The longest running die-hard in a bounding box area of 21 cells or less is this (234 gens):

Code: Select all

bo2bo$5o$4bo$5o!
If you allow a slightly larger bounding box there is this 499 gens die-hard in an 8-by-4 box:

Code: Select all

2bo2b3o$o3b2o$o2bo2bo$3o4bo!

Pyry Virtanen
Posts: 32
Joined: July 22nd, 2022, 2:09 pm

Re: Thread for small diehards

Post by Pyry Virtanen » September 5th, 2023, 2:19 pm

I was able to improve the record 17-cell diehard by 10 ticks by using a more distant century predecessor. I also replaced the boat with a predecessor to decrease the pattern's bounding box area. Its lifespan is now 1247 ticks.

Code: Select all

x = 25, y = 27, rule = B3/S23
8bo$8bo$8bo15$o$o20bo$obo18bo$o20bo$23b2o$24bo4$8b3o!

Pyry Virtanen
Posts: 32
Joined: July 22nd, 2022, 2:09 pm

Re: Thread for small diehards

Post by Pyry Virtanen » September 29th, 2023, 6:11 pm

A new 8-cell record diehard, it lasts 176 ticks:

Code: Select all

x = 8, y = 4, rule = B3/S23
5bo$2bo2b2o$obo4bo$2bo!
It is based on a soup. It would be very doable to find the longest-lasting diehards with population counts 10 and below with a brute force search, but I don't have resources to do that.

User avatar
Entity Valkyrie 2
Posts: 1758
Joined: February 26th, 2019, 7:13 pm
Contact:

Re: Thread for small diehards

Post by Entity Valkyrie 2 » September 29th, 2023, 10:31 pm

Pyry Virtanen wrote:
September 29th, 2023, 6:11 pm
A new 8-cell record diehard, it lasts 176 ticks:

Code: Select all

x = 8, y = 4, rule = B3/S23
5bo$2bo2b2o$obo4bo$2bo!
It is based on a soup. It would be very doable to find the longest-lasting diehards with population counts 10 and below with a brute force search, but I don't have resources to do that.
I guess you won’t be counting things like a glider hitting an arbitrarily distant blinker?
Bx222 IS MY WORST ENEMY.

Please click here for my own pages.

My recent rules:
StateInvestigator 3.0
B3-kq4ej5i6ckn7e/S2-i34q6a7
B3-kq4ej5y6c/S2-i34q5e
Move the Box

Pyry Virtanen
Posts: 32
Joined: July 22nd, 2022, 2:09 pm

Re: Thread for small diehards

Post by Pyry Virtanen » September 30th, 2023, 6:06 am

Entity Valkyrie 2 wrote:
September 29th, 2023, 10:31 pm
I guess you won’t be counting things like a glider hitting an arbitrarily distant blinker?
I set a bounding box restriction in the first post of this thread to avoid things like that. An N-cell diehard must fit into a 2Nx2N bounding box. I think this is generous enough but it rules out the glider + blonk nonsense and things like that.

User avatar
PHPBB12345
Posts: 1096
Joined: August 5th, 2015, 11:55 pm
Contact:

Re: Thread for small diehards

Post by PHPBB12345 » September 30th, 2023, 6:28 am

Pyry Virtanen wrote:
September 30th, 2023, 6:06 am
I set a bounding box restriction in the first post of this thread to avoid things like that. An N-cell diehard must fit into a 2Nx2N bounding box. I think this is generous enough but it rules out the glider + blonk nonsense and things like that.
Is Engineered diehards counts?

Pyry Virtanen
Posts: 32
Joined: July 22nd, 2022, 2:09 pm

Re: Thread for small diehards

Post by Pyry Virtanen » September 30th, 2023, 7:48 am

PHPBB12345 wrote:
September 30th, 2023, 6:28 am
Is Engineered diehards counts?
It's true that pretty much every engineered diehard satisfies my bounding box criterion, but they tend to be relatively large. I try to focus on smaller diehards, let's say patterns with 30 cells or less, or alternatively patterns with bounding box areas 100 or less and no population limit. With a restriction like that it's probably not possible to make good engineered diehards. My main method is looking through Catagolue soups and then if I find something good I try to improve it manually.

Pyry Virtanen
Posts: 32
Joined: July 22nd, 2022, 2:09 pm

Re: Thread for small diehards

Post by Pyry Virtanen » October 1st, 2023, 8:22 pm

Here is a 7x11 diehard that lasts 1119 ticks. It is an improved version of an older 8x11 form.

Code: Select all

x = 11, y = 7, rule = B3/S23
3o$5b2o$7bo$6bo$2ob4o3bo$2bobo5bo$3obo2bob2o!

galoomba
Posts: 111
Joined: February 28th, 2023, 10:19 am

Re: Thread for small diehards

Post by galoomba » October 5th, 2023, 7:43 pm

Started a wiki page, in User namespace for now.

Pyry Virtanen
Posts: 32
Joined: July 22nd, 2022, 2:09 pm

Re: Thread for small diehards

Post by Pyry Virtanen » October 13th, 2023, 9:21 am

galoomba wrote:
October 5th, 2023, 7:43 pm
Started a wiki page, in User namespace for now.
Here are two more diehards that I think should be on that list:

9x10 that lasts 1118 ticks. It's an easy variation of the 7x11 diehard:

Code: Select all

x = 10, y = 9, rule = B3/S23
o$o$o4bo$4bobo$3bob2o$4obo3bo$5bo3bo$2o6b2o$bo!
9x12 that lasts 1143 ticks:

Code: Select all

x = 12, y = 9, rule = B3/S23
5bobo$2bo3b4o$2obo5bobo$2bo3bo3b2o$obob2o2bo2bo$b3o4bo$2o4b2o2bo$3b3o$
5bo!

Pyry Virtanen
Posts: 32
Joined: July 22nd, 2022, 2:09 pm

Re: Thread for small diehards

Post by Pyry Virtanen » November 8th, 2023, 9:45 am

A new record 12-cell diehard, it lasts 515 ticks:

Code: Select all

x = 11, y = 8, rule = B3/S23
8b3o$10bo$3bo$5bo$4bo$4bo$3bo$3o!
Edit: Nevermind, here is an even better 12-cell diehard. It lasts 541 ticks:

Code: Select all

x = 20, y = 4, rule = B3/S23
19bo$bo9bo6bo$obo8b2o5bo$bo16b2o!

User avatar
HerscheltheHerschel
Posts: 589
Joined: September 4th, 2023, 5:23 am

Re: Thread for small diehards

Post by HerscheltheHerschel » November 29th, 2023, 3:26 am

16 cells, 140 generations. This is the best I can do, and was found while trying to search for a normal methuselah.

Code: Select all

x = 10, y = 12, rule = B3/S23
8bo$6b4o$7bo2$bo$2o$2bo$6b3o2$5bo$5bo$5bo!
Another one (population 20, 168 generations, too lazy to calculate bounding box):

Code: Select all

x = 10, y = 20, rule = B3/S23
o2bo2b4o$o4b3obo$ob2o3b3o16$2bo$2b2o!
Population 15, 43 generations:

Code: Select all

x = 10, y = 4, rule = B3/S23
8b2o$2bo3b2obo$4o2bo$bo4b3o!
Predecessor with a population of 12 and lasts for 46 generations:

Code: Select all

x = 10, y = 4, rule = B3/S23
9bo$2bo5b2o$4o3b2o$bo6bo!
Another predecessor with population 12 (again) and 59 generations:

Code: Select all

x = 14, y = 5, rule = B3/S23
10bo$2bo7bo$4o6bo$bo10b2o$13bo!
superstrings, fuses, waves, wicks, and agars are cool
30P5H2V0 IS A BAD, UNMEMORIZABLE NAME
moved to new account hth

User avatar
HerscheltheHerschel
Posts: 589
Joined: September 4th, 2023, 5:23 am

Re: Thread for small diehards

Post by HerscheltheHerschel » December 11th, 2023, 11:39 am

(Sorry for necro/multipost)
205 generations, 7x5 bounding box, 13 cells:

Code: Select all

x = 6, y = 6, rule = B3/S23
3b3o$2bo2bo$3b2o$b2o$obo$2o!
superstrings, fuses, waves, wicks, and agars are cool
30P5H2V0 IS A BAD, UNMEMORIZABLE NAME
moved to new account hth

User avatar
HerscheltheHerschel
Posts: 589
Joined: September 4th, 2023, 5:23 am

Re: Thread for small diehards

Post by HerscheltheHerschel » December 30th, 2023, 12:46 pm

Why isn't anybody posting here?
superstrings, fuses, waves, wicks, and agars are cool
30P5H2V0 IS A BAD, UNMEMORIZABLE NAME
moved to new account hth

Post Reply