New methuselahs

[fsrm]

This is a topic for your new methuselahs.

[fsrm]

4218-gen:

Code: Select all

x = 31, y = 25, rule = B3/S23
15b2o$16bo$16bo18$o$o28bo$o27bo$28bo$28b3o!

[Kiran]

Not notable

[drc]

Not notable

Would above 40k be notable? (I don't have one, just asking) [Not for it's own thread] {batteries not included}

[muzik]

How was that 40k made anyway? I want to see if I can top it off with a few more generations at most

[gameoflifeboy]

How was that 40k made anyway? I want to see if I can top it off with a few more generations at most

It was made from a 15-cell methuselah found by Richard Wobus, presumably found by the Online Life-Like CA Soup Search:

Code: Select all

3o20$3bo$3bo$3bo5$20b3o$9b3o10bo$22bo$21bo!

This methuselah is pretty much the best example of a typical methuselah: it has only 15 cells and already lasts over 32000 generations. However, knightlife discovered that a blinker could be added to one side that wouldn't get touched for about 30000 generations, and when touched, would increase the lifespan by about 8000 generations.

Despite being the longest lived notable methuselah, 40514M has too large a bounding box to be considered as good a methuselah as the 15-cell one, or Lidka.

[muzik]

How was that 40k made anyway? I want to see if I can top it off with a few more generations at most

Despite being the longest lived notable methuselah, 40514M has too large a bounding box to be considered as good a methuselah as the 15-cell one, or Lidka.

Someone should probably invent some sort of measurement, involving the amount of generations for stabilisation compared to the bounding box.

[dvgrn]

Someone should probably invent some sort of measurement, involving the amount of generations for stabilisation compared to the bounding box.

Oh, please don't! Check out the last three columns in the table in Kiran's link --

Not notable

People have been inventing methuselah metrics for forty-six years now, including Gustavo last year who added "escapeability" to an already very overcrowded field.

As a rough approximation, no two Life enthusiasts will be interested in exactly the same methuselah metric, and all metrics about equally arbitrary (or seem that way to somebody).

The best methuselah according to one metric will be different from the best according to another metric... there are so many different "best" methuselahs that none of them end up being really notable.

Basically, in the absence of a large prize for The Methuselah With The Highest Value Of Some Particular Metric, my experience has been that methuselah metrics are pretty much useless, except for causing arbitrary disagreements...!

[fsrm]

12677-gen:

Code: Select all

x = 128, y = 170, rule = B3/S23
125b2o$127bo$127bo$127bo44$35bo$35b2o107$bo$2bo$2bo$3o9$16bo$16bo$16bo
!

[dvgrn]

12677-gen...]

@fsrm, did you follow Kiran's Not notable link?

The pattern you posted starts with 17 ON cells in a 128x170 bounding box.

With 11 ON cells in a 6x4 bounding box, you can get a methuselah that lasts about 5000 ticks longer than your 17-cell one.

That means that it's possible to write a script that could randomly generate millions (at least) of different methuselahs that start with less than 17 cells and last longer than 12,677 generations.

But it's not clear that any of those millions of methuselahs would be worth drawing any special attention to, really. What would you say is interesting about the one you posted?

[Kiran]

I was about to post this (copied from one of your posts), you beat me to it:
This was found in 1997:

Code: Select all

#C 14321M -- Nick Gotts, 3 November 1997
x = 34, y = 6, rule = B3/S23
31b3o$30bo$30bo3$3o!