Catalysts

[Sphenocorona]

This topic is for talking about catalysts in general, not necessarily for herschel tracks. First thing I have is an interesting reaction between a tub/snake on long dock and many types of evolving patterns like B-heptaminos, R-pentaminos, and others:

Code: Select all

x = 43, y = 14, rule = B3/S23
40bo$38bobo$19b2obo15b3o$20b3o$21bo2$2o3b2o10b2o3b2o$o2bo2bo10bo2bo2bo
12b2o3b2o$b5o12b5o13bo2bo2bo$37b5o$3bo16bo$2bobo14bobo17bo$3bo16bo17bo
bo$39bo!

The long dock repairs itself but unfortunately the evolving pattern always comes back and damages it. Is there any way to prevent that?

Another thing I wanted to suggest is the idea of 'proteins', built out of things like hats, buns, aircraft carriers, an whatnot strung together, sort of like the Snark. Just like proteins in real life, the shape of the combined building blocks is what would make one work. Some building blocks are more reactive or volatile than others, and these are usually mostly covered. By looking for catalysts from this perspective we might find some more interesting things.

[Tropylium]

The reaction probably works only on a small subset of reactions. Much of the time, "elbows" develop on either side of the perturbed object which then re-react to destroy the longdock.

Two examples that nearly stop a B, where a right shoulder does not develop:

Code: Select all

x = 20, y = 16, rule = B3/S23
3bo12bo$2bobo10bobo$3bo12bo2$b5o8b5o$o2bo2bo6bo2bo2bo$2o3b2o6b2o3b2o7$
2bo12bo$b3o10b3o$bob2o8b2obo!

The left shoulder seems unavoidable, but can sometimes die off cleanly, as in here:

Code: Select all

x = 12, y = 17, rule = B3/S23
$4bo$3bobo$4bo2$2b5o$bo2bo2bo$b2o3b2o6$3bo$2b3o$b2obo!

---

The 2nd of the large non-spartan catalyst list Sokwe mentioned here seems to actually work in a fairly similar fashion.

And continuing that discussion:

this works just as well with the eater-hive-hook complex replaced with a hook with tail

I think this particular catalyst might be a good candidate to add to ptbsearch. Here is another (useless) example of this reaction that I found a while ago:

Code: Select all

x = 41, y = 16, rule = B3/S23
8bo25bo$6b3o23b3o$5bo25bo$5b2o24b2o2$2b3o8b2o13b3o$bo3bo6bobo12bo3bo8b
o$o5bo5bo13bo5bo5b3o$o5bo4b2o13bo5bo4bo$o5bo19bo5bo5bo$bo3bo5b2o14bo3b
o5b2o$2b3o7bo15b3o$11bo25b2o$12b3o23bo$14bo23bobo$39b2o!

Also this three-point arrangement can actully grab "trimino horns" both "forward" and "backward". Here's a couple century eaters for the sake of example.

Code: Select all

x = 42, y = 35, rule = B3/S23
$4b2o17bo$5bo16bobo$4bo17bobo11b2o$4b2o10b2o3b2ob2o10bo$16bo17bobo$4b
2o8bobo4b2ob2o8b2o$4bo9b2o6bobo$5bo2b3o11bo2bo2b3o$2b3o3bo2bo11b2o3bo
2bo$2bo7b2o18b2o12$9b2o22b2o$9b2o22bo$2bo15bo12bobo$2b3o13b3o10b2o$5bo
15bo$4bo3b3o9bo3b3o$4b2o2bo2bo8b2o2bo2bo$10b2o14b2o$4b2o14b2o$4bo15bo$
2bobo13bobo$2b2o14b2o!

The upper two run "upward", the lower two "downward".

One last catalyst that I think has potential is the following:

Code: Select all

x = 17, y = 16, rule = B3/S23
2$6bo2$7b3o$8bo2$8b2o2bo$7bob4o$5b3o$4bo3b3o$5b3o2bo$7bo!

I have not yet tried running ptbsearch with it, but it is essentially a solution to this problem which I have encountered fairly often:

Code: Select all

x = 4, y = 7, rule = B3/S23
o2$b3o$2bo2$2b2o$2b2o!

I'm not sure if I follow what you're trying to do there, but doesn't the rock functionality of a hook with tail cover this already?

Code: Select all

x = 5, y = 9, rule = B3/S23
bo2$2b3o$3bo2$2bobo$bob2o$bo$2o!

[Sokwe]

doesn't the rock functionality of a hook with tail cover this already?

It does in that particular case, but not others. Obviously, that was a bad example. Here is the sort of reaction I was intending to show:

Code: Select all

x = 9, y = 12, rule = B3/S23
2bo$3bo$3b2obo$4bobo$5bo2$4b2o2bo$3bob4o$b3o$o3b3o$b3o2bo$3bo!

[Tropylium]

doesn't the rock functionality of a hook with tail cover this already?

It does in that particular case, but not others. Obviously, that was a bad example. Here is the sort of reaction I was intending to show:

Code: Select all

x = 9, y = 12, rule = B3/S23
2bo$3bo$3b2obo$4bobo$5bo2$4b2o2bo$3bob4o$b3o$o3b3o$b3o2bo$3bo!

Still could use tweaking then, I suppose; here you can just shift the block to the other side of the nose.

Code: Select all

x = 20, y = 8, rule = B3/S23
o14bo$bo14bo$b2obo11b2obo$2bobo12bobo$3bo14bo2$3b2o12b2o$3b2o12b2o!

I know I've seen that block reaction before, but it hasn't struck me as particularly frequently being unworkaroundable.

---

A more general problem statement though:

Code: Select all

x = 36, y = 38, rule = B3/S23
b5o5b3obo5b2ob2o5bobobo$2b3o7b3o7b3o7b3o$3bo9bo9bo9bo2$3b2o8b2o8b2o7b
2o$3b2o8b2o8b2o7b2o7$bo9bo20b2o$o9bo13b2o4bo3bo$bo9bo19bob3o$b2o2bo5b
2obo6b2o2b2o4bob2o$2b3o7b3o7b3o7b3o$3bo9bo9bo9bo2$3b2o8b2o9b2o6b2o$3b
2o8b2o2b2o6bo6b2o$17bobo4bo$19bo4b2o$19b2o6$2bo9bo$bo3bo5bo2bo$2b3o7b
3o$3bo9bo2$2bobo7b2o$2b2obo6b2o$5bo$5b2o!

Given a growth engine in a t-tetromino stage, depending on what the 3rd row looks like,
1) several forms have no working way to be stopped at all;
2) some forms can be sometimes stopped, depending on what goes on further away in the pattern;
3) two forms can be fairly consistently stopped.

A catalyst that could do something to group 1 should allow perturbing reactions & seeds at several new points.

---

And on another angle. By now most small still lifes have been found to work as catalysts in some fashion. I don't recall seeing aircraft carriers, long snakes, ponds, or mangoes used though? Yet given the wide variety of reactions that a particular object can work with, there would probably be ways to use these too. (The last two, perhaps, may be already covered by loaves though.)

[Sokwe]

Still could use tweaking then, I suppose; here you can just shift the block to the other side of the nose.

Here is a reaction with a distinctly different outcome in this case:

Code: Select all

x = 32, y = 12, rule = B3/S23
o12bo12bo$o12bo12bo$2obo9b2obo9b2obo$bobo10bobo10bobo$2bo12bo12bo2$15b
2o10b2o2bo$15b2o9bob4o$24b3o$23bo3b3o$24b3o2bo$26bo!

I don't recall seeing aircraft carriers, long snakes, ponds, or mangoes used though?

The aircraft carrier and long snake work pretty much anywhere a snake does. Since all of these can be replaced by an eater, it would not help much to add them to ptbsearch. I'm not sure if the pond or mango would have any use, as most things that react with them tend to destroy them. The only stable technology I know of that uses a pond is Brice Due's HWSS eater:

Code: Select all

x = 24, y = 8, rule = B3/S23
21b2o$b6o13bo2bo$o5bo13bo2bo$6bo14b2o$o4bo$2b2o$20b2o$20b2o!

[Tropylium]

Still could use tweaking then, I suppose; here you can just shift the block to the other side of the nose.

Here is a reaction with a distinctly different outcome in this case:

Code: Select all

x = 32, y = 12, rule = B3/S23
o12bo12bo$o12bo12bo$2obo9b2obo9b2obo$bobo10bobo10bobo$2bo12bo12bo2$15b
2o10b2o2bo$15b2o9bob4o$24b3o$23bo3b3o$24b3o2bo$26bo!

Ah I see, that puts out a different spark.

The aircraft carrier and long snake work pretty much anywhere a snake does.

Sure. They have different side profiles though, I'm thinking that there probably exist mechanisms specific to them.

For example a long snake can do this (a fairly similar catalysis to one of the TwT eater variants):

Code: Select all

x = 7, y = 6, rule = B3/S23
$3b4o$4bo$2o$obobo$3b2o!

Now this particular reaction a long hook with tail can do better, but anyway it shows a mechanism separate from a plain snake.

[codeholic]

I found an interesting rectifier-based catalyst with Bellman. Can anyone do something with the final r-pentamino? Probably not, because it's in the middle of the reactor zone

Code: Select all

x = 27, y = 31, rule = B3/S23
bobo7bo$2b2o6bobo$2bo7bobo$11bo12$20b2o$20b2o2$2b2o$bobo$bo$2o4$11b2o
12bo$10bobo11bobo$10bo8b2o4bo$9b2o8bo$20bo$19b2o!