Page 1 of 1

1,437-generation methuselah

Posted: September 20th, 2013, 11:28 am
by Rhombic

Code: Select all

x = 7, y = 4, rule = B3/S23
2b3o$4bo$3o$2bo2b2o!
This 10-cell methuselah stabilises in 1,437 generations. After trying to tweak some initial cells for it to grow larger and more diffuse so it could build itself around still lives for a longer period of time, for it to stay longer. This is my first attempt at a medium-lasting methuselah so don't be too hard on me :P Hopefully it hadn't been discovered yet.

Re: 1,437-generation methuselah

Posted: September 20th, 2013, 2:13 pm
by towerator
Hmm... 1437 isn't that much for 10 cells (rabbits have a 10x longer lifespan for 9 cells), but it may be useful in some ways.

Re: 1,437-generation methuselah

Posted: September 27th, 2013, 5:40 pm
by TWLife
On the second generation it's 9 cells.... and it would last for 1436 gens.

Re: 1,437-generation methuselah

Posted: September 27th, 2013, 7:23 pm
by Tropylium
Rhombic wrote:This is my first attempt at a medium-lasting methuselah so don't be too hard on me :P Hopefully it hadn't been discovered yet.
TBH I don't think anyone is really keeping track of this kind of stuff.

Also, the most efficient way to make mid-range (1 to 10 kilogens) methuselahs appears to be to collide smaller methuselahs (R, B, C, pi, acorn, etc.) with small amounts of junk. These are much less likely to fizzle out than compact patterns with a similar cell count. I ran a full analysis of pi + block and pi+ blinker(which can be rewinded to 8 cells) a while ago which yielded 137 patterns that ran at least 2000 generations.

I've no idea how many compact up to 8-cell patterns that match the same there are (though this might be a completely enumerable set of patterns?) — but less than 5 have been found altogether, IIRC.