mniemiec wrote:gmc_nxtman wrote:P3 oscillator? ... Looks similar to trice tongs.

To see whether things like this are even remotely viable, it's necessarily to put on one's Sherlock Holmes hat to examine the rotor, and see what kind of bushing it would require. The top part survives, while the bottom loaf part doesn't, so let's assume the top is unchanged, and we are free to change the loaf to anything else we want. Generation 2 looks like this:

Code: Select all

```
..OO
.O..O
O..Opr
O.OzAx
.OqBCDE
..syFh
....G
```

For this to work, it is necessary that

1) cells A and B die

2) cell C-G remain alive

3) no births form at x, y, z, r, s

4) births form at p and q

This breaks down to three possibilities:

1) A and B are alive, as in your example. This means they must be overwhelmed. r, s, x, and y must be dead, otherwise no births could form at p or q, so D and F must also be alive (which kills C, making this non-viable)

2) A and B are dead, so one of r and x must be alive and underpopulated, and similar with s and y. This means C must be dead, to prevent birth at z. This would require D and F to be alive, which would cause births at A and B (also non-viable).

3) A is alive and B is dead. To kill A, either x or r must be alive, which would prevent the birth at p (making this non-viable)

4) B is alive and A is dead (mirror-image of 3 above)

Since there is no arrangement of any of the lettered cells that will return your rotor back to its original form from generation 2 to generation 3, this cannot possibly be made to work.

gmc_nxtman wrote:very similar:

Try using similar deductive techniques to determine what it would take to fix the two leftmost cells on the bottom row.

Very very good point, thanks for pointing that out!

About the methuselah guidelines thing, although I have posted some methuselahs in the past, I quickly learned that methuselahs are not notable. I do find this aspect of life very interesting, especially "natural" reactions that lengthen a methuselah's lifespan, (should I be capitalizing methuselah?) so here's what I propose.

Methuselahs are not notable if:

• They have a bounding box of over 20x20, unless they're a non-trivial record holder

• They involve a trivial collision between a glider and some object

• The cell count to area of bounding box ratio is poor (~10%)

• They have a lifespan of less than 10,000 ticks

• They have less than 10 cells and a lifespan of less than 7,500 ticks

• They have no unique or interesting objects in their census

*and* a lifespan of less 7,500 ticks

Any methuselahs which defy one of the above bullets, in my opinion, is notable.

Note that methuselahs with interesting objects in their census are still not notable unless the lifespan is > 7500 ticks.

Also, long hooks eating boats:

Code: Select all

```
x = 11, y = 9, rule = B3/S23
5bo$bob2obo$obo3bo$2o4bob2o$5b2obo$2o4bobobo$obo3bo2b2o$bob2obo$5bo!
```

And a pretty weird component:

Code: Select all

```
x = 15, y = 12, rule = B3/S23
6bobo$2o4bobo4b2o$2o4b3o4b2o5$3bobo3bobo$3b2obobob2o$6bobo$6bobo$7bo!
```

This seems like it could be part of a larger eater, but for now it's just double:

Code: Select all

```
x = 8, y = 17, rule = B3/S23
bo$2bo$3o2$6b2o$2bo3b2o$bobo$b2o2$b2o$bobo$2bo3b2o$6b2o2$3o$2bo$bo!
```

If it's turned asymmetrical, the only problem is a slight bit of induction at the bottom, turning the boat (bob$obo$2ob$!) into a pre-bun (bob$obo$2ob$bob$!).

**EDIT** Aha, so there is! I did see the similarities between this and the TWIT eater variant.

Code: Select all

```
x = 8, y = 8, rule = B3/S23
bo3b2o$obo3bo$2o3bo$4bo$5obo$o4bobo$2bo2bobo$b2o3bo!
```