ConwayLife.com - A community for Conway's Game of Life and related cellular automata
Home  •  LifeWiki  •  Forums  •  Download Golly

A common structure between some catalysts?

For discussion of specific patterns or specific families of patterns, both newly-discovered and well-known.

A common structure between some catalysts?

Postby Scorbie » January 17th, 2015, 10:42 am

Just noticed that catalysts of a variety of reactions share a certain common structure.
Maybe this structure is very robust in catalysis of protrusions, and I think there might be other uses for this catalyst structure.
The one on the top: needless to say.
1. Two of my attempts to make a 2c/3 double signal turner.
2. Calcyman's recent post from here
3&4. Calcyman's eaters from a long time ago. http://www.conwaylife.com/forums/viewtopic.php?f=2&t=315&p=1919&hilit=signal#p1683
(And maybe Calcyman had known about this common structure)
x = 259, y = 161, rule = B3/S23
120bo$120b3o$123bo$122b2o3$125bo$125bobo$125b2o2$132b2o$125b2o5bobo$
125b2o7bo$134b2o2$121bo$120bobob2o$120bobobobo$119b2obobobo2bo$120bo2b
2ob4o$120bo4bo$121b3obo2b2o$123b2o3b2o39$85bobobo$23bo122bobobo$89bo$
23bo126bo78bo3bo$85bobobo$23bo122bobobo78bo3bo$85bo$23bo126bo78bobobo$
85bobobo$23bo122bobobo82bo2$233bo6$252b2o$252bo$249b2obo$244b2o2bobobo
b2o$244bo3bobobo2bo$39bo2bo202b3obobobo2b2o$37b6o204bobobob2obobo$36bo
6b2o50b3o149bobobo4bobo$36b4obobo45bo5b3o150b2ob4o2bo$41bobo44bobo5bo
43bo34bo79b2o$34b4obobob2o43bobo2bo5bo33bo4b3o27bo4b3o77b2o$34bo4bobob
o2bo42b2obobo3bobo33bo2bo31bo2bo44b2o35bo$7b2o20bob2obo2bobobo3b2o45bo
bo3bobo24bobo4b3obobo6b2o20b3obobo6b2o30b2o3b3o34bobo$7bo21b2obobobobo
2bo42b2o3b3obob2ob2o25b2obo8bo7bo26bo7bo36b2o36b2o$8bo5bo2bo16bobo4b2o
41b2o2bo2bobobobo30b3o15bo34bo30b3o35b2o$7b2o5b6o13bo2bo51b3o2bobobo
27b2o4bo13b2o33b2o26b2o2b2o36bobo$20bo13b2o12b2o34b4o5bobobob2o24bob5o
3b2o6bo26b2o6bo29b3ob2o38bo$2o10b5obobo22b2obo2bo34bo3b5o2bobob2o26bo
7b2o2b2obob2obo13b2o7b2o2b2obob2obo27bob2o38b2o$o2bob2o5bo5bob2o20bobo
b3o38bo7bobo28bo3b2o7b2obobob2o14bo2b2o7b2obobob2o27bo3bo27b2o$b3obobo
5b2obobobo14b2o2bo2bobo42b2o2b4o2bo27bo3bobo10bo19bobobo10bo32bo3bo27b
2o$5bobo2bo5bobobo14b2o2b4ob7o28b2o9bo6b2o26b2o3bo3b2o3b2obo18b2o2bo3b
2o3b2obo32bob2o38b2o$3b3ob4o4bo2bob2o21bo6bo28bo2bob2o5b2o40bobo4bobob
o14bo2bo5bobo4bobobo28b3ob2o38bo$2bo3bo9b2o17b5ob2o2b2ob2o30b3obobo5bo
40bobobo2bo2b2o14b2o7bobobo2bo2b2o28b2o2b2o36bobo$3b2obob2obo23bo4b2o
4bobo35bobo2bo2bobo37b2obob4o26b2obob4o36b3o35b2o$4bob2obob3o24bo4bo2b
obo31b2ob2ob4o3b2o35bo5bo28bo5bo45b2o36b2o$2bo11bo23b6o3bo32bo3bo45b5o
bo2b2o24b5obo2b2o36b2o3b3o34bobo$2b2o9b2o66b2obo3b2o45bo3b2o29bo3b2o
41b2o35bo$40b2o41bo4b2o40b2obo31b2obo84b2o$40b2o41bobo44b2ob2o30b2ob2o
85b2o$84b2o162b2ob4o2bo$247bobobo4bobo$247bobobob2obobo$245b3obobobo2b
2o$244bo3bobobo2bo$244b2o2bobobob2o$249b2obo$252bo$128bo18bo18bo18bo
66b2o$129bob2ob2o12bob2ob2o12bob2ob2o12bob2ob2o$127b3o2bob2o10b3o2bob
2o10b3o2bob2o10b3o2bob2o$132bo18bo18bo18bo$121b2o7bobo7b2o7bobo16bobo
16bobo$120bo2bo6b2o7bo2bo6b2o9b2o6b2o9b2o6b2o$119bobo2bo13bobo2bo16bob
o16bobo$118bo2b2obo12bo2b2obo18bo18bo$119b2o3b3o3b2o6b2o3b3o3b2o11b3o
3b2o11b3o3b2o$121b3o3bo2b2o8b3o3bo2b2o9b2o3bo2b2o9b2o3bo2b2o$121bo2b4o
12bo2b4o12bo2b4o12bo2b4o$122b2o17b2o16b3o16b3o$119b3o2b4o10b3o2b4o15b
4o15b4o$119bo2b2o3bo3b2o5bo2b2o3bo10b5o3bo3b2o5b5o3bo$120b2o2b2obobo2b
o6b2o2b2obobob2o5bo4b2obobo2bo5bo4b2obobob2o$122bo2bobob3o9bo2bobob2ob
o8bo2bobob3o9bo2bobob2obo$122b2o2bo14b2o4bo12b2o2bo14b2o4bo$127b3o13b
3obo17b3o13b3obo$129bo13bo2bo20bo13bo2bo$144b2o36b2o4$132b2o17b2o17b2o
17b2o$132bobo16bobo16bobo16bobo$128bo5bo12bo5bo12bo5bo12bo5bo$129bo4bo
b2o10bo4bob2o10bo4bob2o10bo4bob2o$127b3ob2obobo9b3ob2obobo9b3ob2obobo
9b3ob2obobo$132bobobo14bobobo14bobobo14bobobo$132bo2bo15bo2bo15bo2bo
15bo2bo$121b2o7bobo7b2o7bobo16bobo16bobo$120bo2bo6b2o7bo2bo6b2o9b2o6b
2o9b2o6b2o$119bobo2bo13bobo2bo16bobo16bobo$118bo2b2obo12bo2b2obo18bo
18bo$119b2o3b3o3b2o6b2o3b3o3b2o11b3o3b2o11b3o3b2o$121b3o3bo2b2o8b3o3bo
2b2o9b2o3bo2b2o9b2o3bo2b2o$121bo2b4o12bo2b4o12bo2b4o12bo2b4o$122b2o17b
2o16b3o16b3o$119b3o2b4o10b3o2b4o15b4o15b4o$119bo2b2o3bo3b2o5bo2b2o3bo
10b5o3bo3b2o5b5o3bo$120b2o2b2obobo2bo6b2o2b2obobob2o5bo4b2obobo2bo5bo
4b2obobob2o$122bo2bobob3o9bo2bobob2obo8bo2bobob3o9bo2bobob2obo$122b2o
2bo14b2o4bo12b2o2bo14b2o4bo$127b3o13b3obo17b3o13b3obo$129bo13bo2bo20bo
13bo2bo$144b2o36b2o!


EDIT: In a completely unrelated note, pentadecathlon.com is back!
EDIT2: Adding general catalyst to Catgl/ptbsearch sounds like a good idea.
Last edited by Scorbie on January 17th, 2015, 12:17 pm, edited 2 times in total.
Best wishes to you, Scorbie
User avatar
Scorbie
 
Posts: 1381
Joined: December 7th, 2013, 1:05 am

Re: A common structure between some catalysts?

Postby simsim314 » January 17th, 2015, 11:47 am

I was thinking about something similar.

I was considering to use bellman to make a list of catalysts with say max-active = 10 and repair-interval = 10. For some list of interactions (say glider, pi, r-pentomino, hersch, honey farm predecessor etc.).

Then use this list in something like catalyst or ptbsearch - this could yield much more results than the usual searches, and give a bunch of new conduits/stable reflectors/G->H etc.
User avatar
simsim314
 
Posts: 1702
Joined: February 10th, 2014, 1:27 pm

Re: A common structure between some catalysts?

Postby Sokwe » January 17th, 2015, 8:41 pm

simsim314 wrote:I was considering to use bellman to make a list of catalysts with say max-active = 10 and repair-interval = 10.... Then use this list in something like catalyst or ptbsearch

Back when Mike Playle was first posting his new catalysts, I tried adding several of them to ptbsearch. I didn't find very many useful reactions with the searches I did, but I didn't run very many searches with them (I mostly searched for reactions with honeyfarms in an attempt to find honeyfarm hasslers). I certainly think it's worthwhile. I posted some ramblings on this idea here. I also included these alternate catalysts in a file that I included with ptbsearch here. Here are the catalysts that I found somewhat useful:
#Eater
...*!.***!*!zz!
..*z!.*.z!.*!**!
zz!*!.***!...*!
z*!z.*!..*!..**!
*!***!...*!..zz!
**!.*!.*.z!..*z!
..zz!...*!***!*!
..**!..*!z.*!z*!

#Dual Eater
...zz!...*.*!.*.*.***!*.*.....*!**.....*.*!........**!
.....zz..!....*.*..!..***.*.*!.*.....*.*!*.*.....**!**!
...**!..*.*!...*!z**!z!.**!..*!..*.*!...*.*!....**!
....**!...*.*!..*.*!..*!.**!z!z**!...*!..*.*!...**!
**!*.*.....**!.*.....*.*!..***.*.*!....*.*!.....zz!
........**!**.....*.*!*.*.....*!.*.*.***!...*.*!...zz!
**!*.*!.*.*!...*!...**!.....z!...**z!..*!.*.*!.**!
.**!.*.*!..*!...**z!.....z!...**!...*!.*.*!*.*!**!

#Block
zz!**!
**!zz!
z*!z*!
*z!*z!

#Block 2
*z!zz!
zz!z*!

#Block 3
*z!zz!
zz!z*!
z*!zz!
zz!*z!

#Transparent Block
zz!zz!

#Trans Beehive
.zz!z..z!.zz!
.z!z.z!z.z!.z!

#Tub
.z!*.*!.*!
.*!*.z!.*!
.*!*.*!.z!
.*!z.*!.*!

#Tub 2
.*!*.z!.z!
.z!z.*!.*!

#Transparent Tub
.z!z.z!.z!

#Eater 2
zz.*!zz.***!......*!**.***!.*.*!.*.*!..*!
..*!.*.*!.*.*!**.***!......*!zz.***!zz.*!
...*.zz!.***.zz!*!.***.**!...*.*!...*.*!....*!
....*!...*.*!...*.*!.***.**!*!.***.zz!...*.zz!

#Boat
.**!z.*!.z!
.z!z.*!.**!
.z!*.z!**!
**!*.z!.z!

#Boat 2
.z*!z.z!.z!
.z!z.z!.z*!
.z!z.z!*z!
*z!z.z!.z!

#Transparent Loaf
.zz!z..z!.z.z!..z!
..z!.z.z!z..z!.zz!
.z!z.z!z..z!.zz!
.zz!z..z!z.z!.z!

#Eater 3
....**.*!....*.**!.!.....*****!**..*....*!*..*..*!.*.*.*z!**.*.....z!...*....z.z!...**..z..z!........zz!
...*.**!...**.*!.!.*****!.*....*..**!....*..*..*!....z*.*.*!.z.....*.**!z.z....*!z..z..**!.zz!
.zz!z..z..**!z.z....*!.z.....*.**!....z*.*.*!....*..*..*!.*....*..**!.*****!.!...**.*!...*.**!
........zz!...**..z..z!...*....z.z!**.*.....z!.*.*.*z!*..*..*!**..*....*!.....*****!.!....*.**!....**.*!

#Hook with Tail
**!.*!.*.**!..*.*!
...*!.***!*!.*!**!
*.*!**.*!...*!...**!
..**!..*!...*!***!*!

#Long Hook with Tail
...*!.***!*!.*!..*!.**!
.**!..*!.*!*!.***!...*!
....**!**..*!*.*.*!...*!
**!.*..**!.*.*.*!..*!
*!***!...*!..*!.*!.**!
.**!.*!..*!...*!***!*!
...*!*.*.*!**..*!....**!
..*!.*.*.*!.*..**!**!

#Ship
zz!z.*!.**!
.zz!*.z!**!
**!*.z!.zz!
.**!z.*!zz!

#Ship 2
zz!z.z!.z*!
.zz!z.z!*z!
*z!z.z!.zz!
.z*!z.z!zz!

#Ship and Eater
......*!....***!...*!...z*!!*z!*.z!.zz!
.zz!*.z!*z!!...z*!...*!....***!......*!
*!***!...*!..*z!!.....z*!....z.*!....zz!
....zz!....z.*!.....z*!!..*z!...*!***!*!
**!.*!.*.*!..*z!......zz!.....z.z!.....**!
.....**!.....z.z!......zz!..*z!.*.*!.*!**!
......**!......*!....*.*!....z*!zz!z.z!.**!
.**!z.z!zz!....z*!....*.*!......*!......**!

#Bookend
......*!....***!...*!...****!......*!.***!*..z!zz!
z*!z.*!..*!.z*.**!....*.*!....*.*!...**.**!
zz!*..z!.***!......*!...****!...*!....***!......*!
...**.**!....*.*!....*.*!.z*.**!..*!z.*!z*!
*!***!...*!****!*!...***!...z..*!.....zz!
......*z!.....*.z!.....*!..**.*z!.*.*!.*.*!**.**!
.....zz!...z..*!...***!*!****!...*!***!*!
**.**!.*.*!.*.*!..**.*z!.....*!.....*.z!......*z!

#Tub-with-tail eater (block)
**!.*!.*.*!..*.z!...z...z*!.......z*!
.....*!...***!..*!.z.*!..z!!!zz!**!
*z!*z...z!....z.*!.....*.*!.......*!.......**!
....**!....zz!!!...z!..*.z!...*!***!*!
.......**!.......*!.....*.*!....z.*!*z...z!*z!
*!***!...*!..*.z!...z!!!....zz!....**!
.......z*!...z...z*!..*.z!.*.*!.*!**
**!zz!!!..z!.z.*!..*!...***!.....*!

#Gourmet catalyst
....**!...*.*!..*.*!..*!.**!!.**!z.*!.*!..***!....*!
.*!.***!....*!...*.z!...**!!...**!...*!.*.*!*.*!**!
**!*.*!.*.*!...*!...**!!...**!...*.z!....*!.***!.*!
....*!..***!.*!z.*!.**!!.**!..*!..*.*!...*.*!....**!
...z!..*.*.*!.*.**.***!.*.......*!**......*.*!.........**!
**!*.*......**!.*.......*!..***.**.*!....*.*.*!.......z!
.......z!....*.*.*!..***.**.*!.*.......*!*.*......**!**!
.........**!**......*.*!.*.......*!.*.**.***!..*.*.*!...z!

#MikeP 1
......z!..**.z.z!*..*.**!**..*!.....*.*!......**!
..**!..*!....*!...**!..*!.*.*z!*..*.z!**..z!
**!*.*!...*..**!.**.*..*!z.z.**!.z!
.z..**!z.*..*!.z*.*!...*!.**!.*!...*!..**!
.z!z.z.**!.**.*..*!...*..**!*.*!**!
..**!...*!.*!.**!...*!.z*.*!z.*..*!.z..**!
......**!.....*.*!**..*!*..*.**!..**.z.z!......z!
**..z!*..*.z!.*.*z!..*!...**!....*!..*!..**!

#MikeP 2
...**!....*!..*!..**..**!.....*.*!****.*!*..z.zz!
...**!...*!.....*!**..**!*.*!..*.****!.zz.z..*!
..**!z..*!z**!.....**!z*.*..*!.*.**!.*!**!
...**!...*..z!....**z!**!*..*.*z!..**.*!.....*!.....**!
*..z.zz!****.*!.....*.*!..**..**!..*!....*!...**!
.zz.z..*!..*.****!*.*!**..**!.....*!...*!...**!
**!.*!.*.**!z*.*..*!.....**!z**!z..*!..**!
.....**!.....*!..**.*!*..*.*z!**!....**z!...*..z!...**!

#Dead spark coil
zz.zz!z...z!.***!!.***!z...z!zz.zz!
zz...zz!z.*.*.z!..*.*!z.*.*.z!zz...zz!

#Reflector hook
...zz!..z.z!..*!...*!***!*!
*!***!...*!..*!..z.z!...zz!
zz!z.z!..*!.*!..***!....*!
....*!..***!.*!..*!z.z!zz!
....zz!..*..z!.*.*z!.*!**!
**!.*!.*.*z!..*..z!....zz!
zz!z..*!.z*.*!....*!....**!
....**!....*!.z*.*!z..*!zz!

#Beehive tail thing
...z*.*z!....*.*!....*.*!.zz.*.*.zz!z..*...*..z!.**.....**!
.**.....**!z..*...*..z!.zz.*.*.zz!....*.*!....*.*!...z*.*z!
.z!*.z!*.z!.*...z!..****!!..****!.*...z!*.z!*.z!.z!
....z!...z.*!...z.*!z...*!****!!****!z...*!...z.*!...z.*!....z!

#Long hook
...zz!z...z!****!!****!z...z!...zz!
zz!z...z!.****!!.****!z...z!zz!
zz...zz!z.*.*.z!..*.*!..*.*!.z*.*z!
.z*.*z!..*.*!..*.*!z.*.*.z!zz...zz!

To use these catalysts in ptbsearch, you need to create a new file which includes only the patterns (lines with '.', '*', 'z', and '!'). For example, here is "spartan.list" which contains the spartan catalysts:
zz!zz!
.z!z.z!.z!
.zz!z..z!.zz!
.z!z.z!z.z!.z!
.zz!z..z!.z.z!..z!
..z!.z.z!z..z!.zz!
.z!z.z!z..z!.zz!
.zz!z..z!z.z!.z!
*z!zz!
zz!z*!
z*!zz!
zz!*z!
.*!*.z!.z!
.z!z.*!.*!
.z*!z.z!.z!
.z!z.z!.z*!
.z!z.z!*z!
*z!z.z!.z!
zz!z.z!.z*!
.zz!z.z!*z!
*z!z.z!.zz!
.z*!z.z!zz!
...*!.***!*!zz!
..*z!.*.z!.*!**!
zz!*!.***!...*!
z*!z.*!..*!..**!
*!***!...*!..zz!
**!.*!.*.z!..*z!
..zz!...*!***!*!
..**!..*!z.*!z*!
......*!....***!...*!...z*!!*z!*.z!.zz!
.zz!*.z!*z!!...z*!...*!....***!......*!
*!***!...*!..*z!!.....z*!....z.*!....zz!
....zz!....z.*!.....z*!!..*z!...*!***!*!
**!.*!.*.*!..*z!......zz!.....z.z!.....**!
.....**!.....z.z!......zz!..*z!.*.*!.*!**!
......**!......*!....*.*!....z*!zz!z.z!.**!
.**!z.z!zz!....z*!....*.*!......*!......**!


Another thing to try is to use ptbsearch to find an "almost working" pattern and then try to complete it with Bellman. For example, one could try to replace the p6 in Extrementhusiast's fast Herschel conduit with a still life:
x = 27, y = 34, rule = B3/S23
15b2o$14bo2bo$17bo$17bo$11bo3bobo$10bo3bobo$10bo4bo$11b4o7$22bo$2o20bo
$bo20b3o$bobo20bo$2b2o3$20bo3b2o$15b2o2bobo3bo$15bo3b2o3bo$16b3o4bo$
18bob4obo$19b2o3bobo$23bo2bo$24b2o2$3b3o7b2o$4bo8bobo$4b3o8bo$15b2o!
-Matthias Merzenich
Sokwe
Moderator
 
Posts: 1480
Joined: July 9th, 2009, 2:44 pm

Re: A common structure between some catalysts?

Postby dvgrn » January 17th, 2015, 10:34 pm

In the spirit of other recent high-volume data collection and filtering -- apgsearch, fusefinder, etc. -- it occurs to me that it might be interesting to measure the catalyzing abilities of a fairly wide range of objects. Presumably the ones that score the highest would be the best bets for adding to ptbsearch or catalyst.

I'm not sure exactly how to conduct a randomized test; it doesn't seem like a good idea to start with completely random soups, since catalysts are supposed to interact with the leading edge of an active reaction, and those are very far from random. So maybe just catalogue all the leading edges of a lot of long-lived methuselahs...?

-- Then place the object to be tested at every possible location near each leading edge, and see if it happens to survive the initial contact. If it's still there after 25 ticks or so, that probably counts as success. For the average object the success rate will be very low, but still generally above zero (because of transparent reactions if nothing else).

Good candidates for catalysts would then be the objects with the highest success rates. I suspect there would be a few interesting surprises in the results, along with the obvious winners like blocks and eaters.
User avatar
dvgrn
Moderator
 
Posts: 5831
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: A common structure between some catalysts?

Postby simsim314 » January 18th, 2015, 2:19 am

dvgrn wrote:it doesn't seem like a good idea to start with completely random soups, since catalysts are supposed to interact with the leading edge of an active reaction, and those are very far from random.


So let the random soup to run for ~50 generations and then place potential catalysts. If you want "really" stay out of randomness take larger number, taking 300 will keep you very far from randomness.

Another approach would be to run bellman in some local area of such patterns, and see what kind of catalysts are emerging, and conduct statistics on them.
User avatar
simsim314
 
Posts: 1702
Joined: February 10th, 2014, 1:27 pm

Re: A common structure between some catalysts?

Postby Extrementhusiast » January 18th, 2015, 5:03 pm

dvgrn wrote:I'm not sure exactly how to conduct a randomized test; it doesn't seem like a good idea to start with completely random soups, since catalysts are supposed to interact with the leading edge of an active reaction, and those are very far from random. So maybe just catalogue all the leading edges of a lot of long-lived methuselahs...?

What about via that one variant of apgsearch that collides gliders with a field of random still lifes? If testing like this were to happen, then one could also find components, not just eaters of some sort.

While your approach is good, I was actually thinking of more of a depth-first search, by starting with a Herschel (or some other common methuselah), and then randomly adding catalysts until something works.
I Like My Heisenburps! (and others)
User avatar
Extrementhusiast
 
Posts: 1794
Joined: June 16th, 2009, 11:24 pm
Location: USA

Re: A common structure between some catalysts?

Postby dvgrn » January 18th, 2015, 5:26 pm

Extrementhusiast wrote:What about via that one variant of apgsearch that collides gliders with a field of random still lifes? If testing like this were to happen, then one could also find components, not just eaters of some sort.

The approach I outlined would certainly find a lot of eater reactions, but it would also find all kinds of successful modifications of active reactions -- including the occasional transparent reaction, where the catalyst reappears and a glider or Herschel or whatever shows up on the other side of the catalyst.

Extrementhusiast wrote:While your approach is good, I was actually thinking of more of a depth-first search, by starting with a Herschel (or some other common methuselah), and then randomly adding catalysts until something works.

Clearly that's ultimately the right direction to go. What I was speculating about was a way of pre-testing catalysts for viability: how do you know what the most likely catalysts are to add? There are a lot of possible still lifes that might occasionally do something useful. If we can get relative survival ratios for common still lifes and constellations in a large randomized test, that might give a sense of which catalysts are really worth trying. E.g., might there be a transparent tub or pond reaction out there somewhere?

Once we a larger but not-too-large set of catalysts to try in ptbsearch or catalyst, maybe we can figure out how to distribute a new Herschel conduit search across a larger number of computers, and just see how much new stuff turns up.
User avatar
dvgrn
Moderator
 
Posts: 5831
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: A common structure between some catalysts?

Postby Kazyan » January 18th, 2015, 5:52 pm

The first approach has the problem of cross-contamination. For example, maybe a ship gets destroyed by a LoM, then a Herschel swoops in and deposits another ship in the same spot. You can't really call that a transparent reaction. That will be a significant confounding variable.
Tanner Jacobi
User avatar
Kazyan
 
Posts: 860
Joined: February 6th, 2014, 11:02 pm

Re: A common structure between some catalysts?

Postby Extrementhusiast » January 18th, 2015, 6:33 pm

dvgrn wrote:Clearly that's ultimately the right direction to go. What I was speculating about was a way of pre-testing catalysts for viability: how do you know what the most likely catalysts are to add? There are a lot of possible still lifes that might occasionally do something useful. If we can get relative survival ratios for common still lifes and constellations in a large randomized test, that might give a sense of which catalysts are really worth trying. E.g., might there be a transparent tub or pond reaction out there somewhere?


Well, let's start with the Spartan SLs, because even just those have a lot of possibilities between them, including this one (which I found manually):
x = 9, y = 17, rule = B3/S23
bo$obo$obo$bo11$5b3o$6bo$6b3o!


And here's an example of Kazyan's point:
x = 23, y = 16, rule = B3/S23
20b2o$19bo2bo$20b2o11$3o$bo$b3o!
I Like My Heisenburps! (and others)
User avatar
Extrementhusiast
 
Posts: 1794
Joined: June 16th, 2009, 11:24 pm
Location: USA

Re: A common structure between some catalysts?

Postby dvgrn » January 18th, 2015, 10:53 pm

Kazyan wrote:The first approach has the problem of cross-contamination. For example, maybe a ship gets destroyed by a LoM, then a Herschel swoops in and deposits another ship in the same spot. You can't really call that a transparent reaction. That will be a significant confounding variable.

Hmm, not sure about that. The block+snake combination in many Herschel conduits could be described as "a block gets destroyed by an R-pentomino, and then a B-heptomino swoops in and deposits another block in the same spot." That block still counts as transparent... Not very much time passes between the destruction and the replacement, so maybe that's the key: when you're testing catalysts for success, they can only disappear for (e.g.) N<25 ticks at most, and then they have to reappear and stay in place for at least M>15 ticks (let's say).

There's no particular reason to run a reaction to completion and require that the catalyst should appear in the final product, because we'll probably want to place other catalysts to modify any leftover junk anyway. Running to completion just takes extra computation for no benefit in this context.

There are all kinds of subtleties that are very hard to measure, such as the accessibility of the sparks around a potential transparent catalyst. Many otherwise hopeful Herschel conduits are ruined because some final decaying reaction gets stuck in the middle of an earlier part of the reaction envelope -- there's no way to add further catalysts to get rid of some piece of annoying junk. There's probably a way to algorithmically rate the "hopefulness" of a reaction along these lines -- record the number of ticks since the envelope increased in size, maybe, and do something with that? -- but it's certainly not a trivial problem.
User avatar
dvgrn
Moderator
 
Posts: 5831
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: A common structure between some catalysts?

Postby simsim314 » January 19th, 2015, 3:01 am

I think that if we just add another 20 SLs with nice potential - we could find a lot more conduits coming from similar searches we conducted previously.

Here is some example of a find to show-case the potential of "new" SL that seems to have nice potential as catalyst:

x = 16, y = 21, rule = B3/S23
10b2o$9bo2bo$10bobo$9b2o2b2o$11b2o2bo$9b2o2b2o$3o7bobo$bo7bo2bo$b3o6b
2o9$2b2o$2bo$3b3o$5bo!
User avatar
simsim314
 
Posts: 1702
Joined: February 10th, 2014, 1:27 pm

Re: A common structure between some catalysts?

Postby Sokwe » January 19th, 2015, 3:38 am

simsim314 wrote:Here is some example of a find to show-case the potential of "new" SL that seems to have nice potential as catalyst

I think this one has a lot of potential because it can also perform a "common" reaction that I have been using another still life combination for:
x = 55, y = 16, rule = B3/S23
8bo39bo$6b3o37b3o$5bo39bo$5b2o38b2o2$2b3o37b3o$bo3bo35bo3bo8bo$o5bo5b
2o3b2o21bo5bo5b3o$o5bo4bo2bobo2bo20bo5bo4bo$o5bo5bobobobo21bo5bo5bo$bo
3bo5b2o2bo2b2o21bo3bo5b2o$2b3o8bobobo24b3o$11b2o2bo2b2o31b2o$12bobobob
o33bo$11bo2bobo2bo32bobo$12b2o3b2o34b2o!

Obviously, the new still life is still a little too obtrusive in this case, but in most cases this probably won't matter. For ptbsearch, it might be fastest to give the oscillator more symmetry and only include two orientations:
x = 22, y = 9, rule = B3/S23
b2o3b2o6bobobobo$o2bobo2bo4bob2ob2obo$bobobobo5bo3bo3bo$2o2bo2b2o5b2o
3b2o$2bobobo9b3o$2o2bo2b2o5b2o3b2o$bobobobo5bo3bo3bo$o2bobo2bo4bob2ob
2obo$b2o3b2o6bobobobo!

It's not really any more obtrusive in this form.
-Matthias Merzenich
Sokwe
Moderator
 
Posts: 1480
Joined: July 9th, 2009, 2:44 pm


Return to Patterns

Who is online

Users browsing this forum: No registered users and 6 guests

cron