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

The Hunting of the New Herschel Conduits

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

Re: The Hunting of the New Herschel Conduits

Postby Kazyan » June 12th, 2015, 4:59 pm

Extrementhusiast wrote:One-bit spark plus beehive restores the beehive:
x = 16, y = 34, rule = LifeHistory
2D$.D$.D.D10.2A$2.2D10.2A11$10.C$9.C.C$9.C.C$7.A2.C15$13.2D$13.2D!

Optional eater shown to get rid of extra block (if needed), optional block shown as transparent catalyst. (With that, since the reaction seems to go off to the northwest, wouldn't it be hilarious if we could get a gun from this?)


In addition to the chance of a hilarious gun, there are tons of possibilities for perturbing the junk produced by the reaction and then crashing stuff into it to get an X-to-G. I'm shooting for a G-to-G right now.

EDIT: Accidental conduit discovery, though probably will never hook up to an input:

x = 47, y = 32, rule = B3/S23
36b2o$36bo$22b2o10bobo$22b2o10b2o12$27bo$26b3o$28bo3$45b2o$45bo$43bobo
$43b2o2$2o$o$b3o$3bo38b2o$42bobo$9b2o33bo$9b2o33b2o!
Tanner Jacobi
User avatar
Kazyan
 
Posts: 807
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Postby dvgrn » June 12th, 2015, 7:40 pm

Extrementhusiast wrote:One-bit spark plus beehive restores the beehive...
Optional eater shown to get rid of extra block (if needed), optional block shown as transparent catalyst.

Guam's version from a few years ago used the same two blocks but a different eater to the north:

x = 68, y = 50, rule = LifeHistory
$6.3B14.A$6.4B11.3A$7.4B9.A$8.4B8.2A$9.4B2.2B.4B$10.4B.5B$10.11B$7.
14B.2B$6.19B$5.20B$6.20B$6.21B$6.21B$6.22B$5.24B$5.24B14.2A$5.23B3.B
11.A$4.32B4.BA.A$3.19BA16B.B2A$4.37B$4.19B2A16B$4.18BA2BA15B$4.19B2A
14B$3.28B5.B.B$4.2B2A7B2.14B4.3B$5.B2A8B5.8B6.B2AB$5.12B4.6B9.2A$6.6B
2.4B4.4B$8.4B3.4B$16.4B$17.4B$18.4B$19.4B$20.4B$21.4B$22.3B!

Kazyan wrote:In addition to the chance of a hilarious gun, there are tons of possibilities for perturbing the junk produced by the reaction and then crashing stuff into it to get an X-to-G...

Hey, that's a really good idea -- why didn't I think of that when I first saw Guam's bit-plus-beehive? You mean catalyzing the post-beehive active reaction to produce things like this block, right?

x = 18, y = 34, rule = LifeHistory
2.2A$3.A$3.A.A10.2A$4.2A10.2A9$3.2B$3.B2A$.3B2A2B4.C$9B2.C.C$9B2.C.C$
9B3.C$.2A6B$ABA5B$.BA3B$2.4B$3.2B10$15.2D$15.2D!

How many different common G+object collisions are there that produce a clean one-bit spark with enough clearance? There must be quite a few, and each one is a separate search. Guess I'll just keep my fingers crossed for now...!

EDIT: Random sample, much less likely to work out than the block, for the most part:

x = 261, y = 34, rule = LifeHistory
5.2A58.2A58.2A58.2A58.2A$6.A59.A59.A59.A59.A$6.A.A10.2A45.A.A10.2A45.
A.A10.2A45.A.A10.2A45.A.A10.2A$7.2A10.2A46.2A10.2A46.2A10.2A46.2A10.
2A46.2A10.2A7$64.A$65.A$63.3A2$15.C59.C59.C59.C59.C$14.C.C57.C.C57.C.
C57.C.C57.C.C$14.C.C57.C.C57.C.C57.C.C57.C.C$10.2AD2.C56.D2.C53.2A.D
2.C56.D2.C53.A2.D2.C$9.A2.A115.A.A117.A.A$10.2A116.2A60.2A57.2A$190.A
.A$68.2A121.A$.2A65.A.A58.2A$A.A66.A58.A.A$2.A127.A53.2A57.3A$183.A.A
59.A$185.A58.A6$18.2D58.2D58.2D58.2D58.2D$18.2D58.2D58.2D58.2D58.2D!
User avatar
dvgrn
Moderator
 
Posts: 5546
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: The Hunting of the New Herschel Conduits

Postby A for awesome » June 13th, 2015, 4:49 pm

Does anyone know of A. a catalyst or B. a G-to-block that can complete this conduit:
x = 22, y = 48, rule = LifeHistory
10.B$9.3D$9.BDB$8.2B3D$8.5B$8.6B$8.6B$8.5B$7.6B$8.6B$7.7B$7.6B$7.6B$
7.6B$6.8B$7.8B$6.9B$6.9B$6.10B$6.5B2C3B$6.5B2C4B$6.11B$6.12BC$8.8B2.
3C$8.7B6.C$8.6B6.2C$8.6B$8.7B$6.8B$6.8B$5.9B.BC$5.10BC.C$4.11B.2C$3.
10B$3.10B$3.11B$3.10B$3.9B$2.10B$.12B$2C11B$2CB.9B$5.4B3CB$6.4BC2B$6.
2B3C2B$7.6B$7.7B$9.B.4B!

?
x₁=ηx
V ⃰_η=c²√(Λη)
K=(Λu²)/2
Pₐ=1−1/(∫^∞_t₀(p(t)ˡ⁽ᵗ⁾)dt)

$$x_1=\eta x$$
$$V^*_\eta=c^2\sqrt{\Lambda\eta}$$
$$K=\frac{\Lambda u^2}2$$
$$P_a=1-\frac1{\int^\infty_{t_0}p(t)^{l(t)}dt}$$

http://conwaylife.com/wiki/A_for_all

Aidan F. Pierce
User avatar
A for awesome
 
Posts: 1778
Joined: September 13th, 2014, 5:36 pm
Location: 0x-1

Re: The Hunting of the New Herschel Conduits

Postby Scorbie » June 14th, 2015, 3:48 am

Almost H-to-R with an annoying block:
x = 48, y = 31, rule = B3/S23
4bo$3bobo$3b2o2$28bob2o$27bob2obo$27bo4bo$24b2ob6ob2o$24bobo6bobo$26bo
b4obo$25b2obo2bob2o4$2b2o$bobo$bo44b2o$2o44b2o9$35b3o$26b2o8bo$27bo6b
3o$24b3o$24bo!


EDIT: and an H-to-QB:
x = 46, y = 30, rule = B3/S23
26bob2o$25bob2obo$25bo4bo$22b2ob6ob2o$22bobo6bobo$24bob4obo$23b2obo2bo
b2o4$2b2o$bobo$bo42b2o$2o42b2o6$11b2o$11b2o2$33b3o$24b2o8bo$25bo6b3o$
22b3o$22bo2$39b2o$39b2o!


EDIT: a brief 1-catalyst search gives these potential candidates:
x = 262, y = 97, rule = B3/S23
166b2o3b2o$166bobo2b2o$168bo$21bob2o42bob2o93b4ob4o$20bob2obo40bob2obo
62bob2o26bo2bobo2bo69bob2o$20bo4bo40bo4bo61bob2obo102bob2obo$17b2ob6ob
2o34b2ob6ob2o58bo4bo88b2o12bo4bo$17bobo6bobo34bobo6bobo55b2ob6ob2o43bo
b2o38b2o9b2ob6ob2o$19bob4obo38bob4obo57bobo6bobo42bob2obo48bobo6bobo$
18b2obo2bob2o36b2obo2bob2o58bob4obo44bo4bo50bob4obo$131b2obo2bob2o40b
2ob6ob2o46b2obo2bob2o$2o179bobo6bobo$2o181bob4obo$106b2o74b2obo2bob2o$
107bo$39b2o44b2o17b3o$39b2o44b2o17bo47b2o106b2o$152b2o106b2o2$203b2o$
203b2o$56bobo$55bob2o$55bo$54b2ob2o$28b3o26bo16b3o$19b2o8bo24b2obo17bo
65b3o105b3o$20bo6b3o24bobo16b3o56b2o8bo97b2o8bo$17b3o113bo6b3o98bo6b3o
$17bo112b3o59b3o43b3o$130bo52b2o8bo44bo$34b2o44b2o102bo6b3o$34b2o44b2o
65b2o32b3o71b2o$147b2o32bo73b2o2$198b2o$198b2o26$196bob2o$195bob2obo$
4b2o98b2o89bo4bo$4bo99bo87b2ob6ob2o$2bobo97bobo87bobo6bobo$2b2o6bob2o
88b2o4bob2o82bob4obo$9bob2obo92bob2obo80b2obo2bob2o$9bo4bo92bo4bo$6b2o
b6ob2o86b2ob6ob2o$6bobo6bobo86bobo6bobo$8bob4obo90bob4obo$7b2obo2bob2o
88b2obo2bob2o$214b2o$214b2o2$175b2o$175bo2bo$28b2o96b2o48b3o$28b2o96b
2o$172b2obob2o$172bob2obo$177bo$177b2o24b3o$194b2o8bo$195bo6b3o$192b3o
$192bo$17b3o95b3o$8b2o8bo87b2o8bo92b2o$9bo6b3o88bo6b3o92b2o$6b3o95b3o$
6bo97bo2$23b2o96b2o$23b2o96b2o!
Best wishes to you, Scorbie
User avatar
Scorbie
 
Posts: 1375
Joined: December 7th, 2013, 1:05 am

Re: The Hunting of the New Herschel Conduits

Postby dvgrn » June 14th, 2015, 9:47 am

Scorbie wrote:Almost H-to-R with an annoying block...
EDIT: and an H-to-QB...

This looks like partly known territory, unfortunately. The elementary-conduits collection has only one H-to-QB in it, and it produces an identical output. Is there anything the non-Spartan catalyst can do that two blocks can't?

x = 90, y = 46, rule = LifeHistory
20.A49.A$20.3A47.3A$23.A49.A$22.2A48.2A5$35.D49.D$35.D.D47.D.D$36.D.D
47.D.D$13.2A21.D2.D23.2A21.D2.D$13.2A21.D.D24.2A21.D.D$35.D.D47.D.D$
35.D49.D8$7.2A20.2A26.2A$8.A18.A2.A27.A$8.A.A16.3A28.A.A$9.2A19.3A26.
2A16.2A$27.2A.A2.A43.2A$28.A.A.A$28.A.A.2A$27.2A.A2.A$30.3A44.2A$27.
3A47.2A$9.C17.A2.A28.C$9.C.C17.2A28.C.C$9.3C47.3C$11.C49.C5$2.2A48.2A
$3.A49.A$3A47.3A$A49.A$20.2A48.2A$20.2A48.2A!

I'm hoping to convince more people to post new conduits with a standard name to go with them -- HF145QB, in this case. Then eventually it will make sense to try a quick text search. "HF145QB" does actually turn up the relevant prior art, but I think it's going to be a good while before that's true in general.

I think the blocked H-to-R may actually be new. I don't see any mention of an F258 (i.e., HF258H) elsewhere in this thread:

F258 (not Spartan, unfortunately, in spite of the potential):
x = 157, y = 65, rule = LifeHistory
44.2A78.2A$20.A24.A54.A24.A$20.3A21.A55.3A21.A$23.A20.2A57.A20.2A$22.
2A.2B3.2B3.2A8.B56.2A.2B3.2B3.2A8.B$22.11B2.A.A7.3B19.4B4.A26.11B2.A.
A7.3B19.4B4.A$24.9B3.AB6.6B16.5B3.A.A27.9B3.AB6.6B16.5B3.A.A$22.12B.
2B3.B2.10B11.7B3.A.A25.12B.2B3.B2.10B11.7B3.A.A$21.13B.19B3.2B2.10B4.
A25.13B.19B3.2B2.10B4.A$21.22BD3B2A15BD6B30.22BD3B2A15BD6B$20.23B2D2B
2A15BDBD4B29.23B2D2B2A15BDBD4B$20.B2.8B2D11B2D18B3D4B29.B2.8B2D11B2D
18B3D4B$18.3B2.9B2D10BD21BD4B27.3B2.9B2D10BD21BD4B$17.15BD10BD26B27.
15BD10BD26B$16.19B.4B5.13B.4B33.19B.4B5.13B.4B$16.24B5.7B.B4.4B34.24B
5.7B.B5.3B$16.22B19.4B35.22B19.4B$16.20B20.4B36.20B21.2A$16.19B20.4B
37.19B23.A$16.17B21.4B38.17B22.3A$16.18B19.4B39.18B21.A$16.10B2.2B.4B
6.A10.4B40.10B2.2B.4B$16.10B6.4B5.3A7.4B41.10B6.4B$15.10B8.4B7.A5.4B
41.10B8.4B$15.10B9.4B5.2A4.4B42.10B9.3B$7.2A6.10B10.4B4.9B35.2A6.10B
10.2B$8.A7.9B11.4B5.6B37.A7.9B11.B$8.A.AB4.9B.2B9.4B2.8B37.A.AB4.9B.
2B$9.2AB.14B2A9.15B36.2AB.14B2A8.4B$11.16B2A10.14B38.16B2A9.4B$11.14B
.2B12.13B38.14B.2B11.4B$12.13B16.10B.B2A37.13B15.4B$11.14B.2B15.3B2AB
3.BA.A35.14B.2B13.4B$9.18B2A14.3B2AB6.A33.18B2A13.4B8.A$7.18B.B2A16.
4B6.2A30.18B.B2A7.2A5.4B5.5A$7.2BC15B2.B17.3B39.2BC15B2.B9.A5.4B4.A5.
A$6.3BCBC4B.8B17.AB.2B39.3BCBC4B.8B12.A.AB.7B2.B3A2.A$7.2B3C4B2.7B16.
A.AB2AB39.2B3C4B2.7B13.2AB.7B3.2B.A.2A$6.5BC4B2.7B16.A.ABABAB37.5BC4B
2.7B15.12B4A2.A$5.10B3.7B15.2A.A.A.A2.A34.10B3.7B15.7B2A3BAB2.2A$4.4B
11.7B15.A2.2A.4A33.4B8.11BA12.7B2A2B.B3A$4.3B12.7B15.A4.A37.3B10.9BA.
A11.10B3.B.A$2.4B13.6B17.3A.A2.2A31.4B12.9BA4.A6.8B8.A.2A$2.2A16.5B
19.2A3.2A31.2A15.6B.B3.3A5.9B7.2A.2A$3.A14.6B59.A14.6B5.A7.4B2.3B$3A
15.6B56.3A15.7B4.2A5.4B3.5B$A17.4B58.A16.2A12B4.4B7.2A$18.2B2AB73.A.A
10B5.4B8.A$20.2A74.A3.11B2.4B10.3A$95.2A3.16B13.A$101.14B$101.13B$99.
2AB.10B$98.A.AB3.B2A5B$98.A6.B2A6B$97.2A6.4B2.4B$106.3B3.4B$107.2B.BA
$106.B2ABA.A$105.BABABA.A$103.A2.A.A.A.2A$103.4A.2A2.A$107.A4.A$103.
2A2.A.3A$103.2A3.2A!

I might be missing a tricky combination somewhere, though -- that block is accessible from three of the four possible diagonals, and even the fourth one isn't impossible if the input H comes from one of those strange new first-natural-block-eating conduits, I suppose.
User avatar
dvgrn
Moderator
 
Posts: 5546
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: The Hunting of the New Herschel Conduits

Postby A for awesome » June 14th, 2015, 1:11 pm

A HL226B:
x = 52, y = 42, rule = LifeHistory
16.C$10.2C3.C.C$9.C.C3.BC2B$9.BC5.2B$10.2B2.B.3B$8.11B$7.D12B$6.2D13B
$5.2D14B$6.D14B$7.D14B$11.11B$10.12B$9.13B$9.13B3.B$8.15B.B2C$6.B.17B
2C$5.18B.2B$3.20B$3.19B8.2C3.2C$2.21B6.B2CB.B2CB$2.21B6.4B.3B$2.23B5.
2B3.B6.2B$4.42B$2.47B.B$2.2C2.42B.B2C$3.C3.43B2C$3C4.41B.2B$C8.36B.2B
$10.32B$11.28B$12.26B$13.B2.3B6.3B2.11B$30.12B$30.12B$31.11B$29.4B.4B
3CB$29.2C4.4BC2B$30.C4.2B3C2B$27.3C6.6B$27.C8.7B$38.B.4B!

Edit: Fixed the name; I put R instead of L.
Last edited by A for awesome on June 15th, 2015, 9:50 am, edited 1 time in total.
x₁=ηx
V ⃰_η=c²√(Λη)
K=(Λu²)/2
Pₐ=1−1/(∫^∞_t₀(p(t)ˡ⁽ᵗ⁾)dt)

$$x_1=\eta x$$
$$V^*_\eta=c^2\sqrt{\Lambda\eta}$$
$$K=\frac{\Lambda u^2}2$$
$$P_a=1-\frac1{\int^\infty_{t_0}p(t)^{l(t)}dt}$$

http://conwaylife.com/wiki/A_for_all

Aidan F. Pierce
User avatar
A for awesome
 
Posts: 1778
Joined: September 13th, 2014, 5:36 pm
Location: 0x-1

Re: The Hunting of the New Herschel Conduits

Postby Kazyan » June 14th, 2015, 2:06 pm

A for awesome wrote:Does anyone know of A. a catalyst or B. a G-to-block that can complete this conduit:
x = 22, y = 48, rule = LifeHistory
10.B$9.3D$9.BDB$8.2B3D$8.5B$8.6B$8.6B$8.5B$7.6B$8.6B$7.7B$7.6B$7.6B$
7.6B$6.8B$7.8B$6.9B$6.9B$6.10B$6.5B2C3B$6.5B2C4B$6.11B$6.12BC$8.8B2.
3C$8.7B6.C$8.6B6.2C$8.6B$8.7B$6.8B$6.8B$5.9B.BC$5.10BC.C$4.11B.2C$3.
10B$3.10B$3.11B$3.10B$3.9B$2.10B$.12B$2C11B$2CB.9B$5.4B3CB$6.4BC2B$6.
2B3C2B$7.6B$7.7B$9.B.4B!

?


The Spartan block factory would work...but the color of the FNG is wrong, if reflected via Snark. So is the first glider of a dependent conduit and the glider of the large B-to-H non-spartan conduit. Bah humbug. Adding more conduits to get a different-colored glider will do it, but then the repeat time starts to get sketchy.
Tanner Jacobi
User avatar
Kazyan
 
Posts: 807
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Postby dvgrn » June 14th, 2015, 5:10 pm

A for awesome wrote:A HR226B...

Very nice! That's another conduit for the Spartan collection -- Lx285:

Code: Select all
x = 51, y = 78, rule = LifeHistory
43.2A$42.A2.A$43.2A4$38.8B$37.9B$37.9B$37.9B$37.B3D4B$37.2BD4B$37.2B
3D2B$37.6B$36.7B$35.8B$34.8B$34.8B$32.2AB2.6B$31.A.AB.7B$31.A4.6B$30.
2A4.6B$36.6B$35.8B$36.8B$35.9B$19.2A14.9B$20.A14.10B$20.A.AB.3B7.5B2A
3B$21.2AB.5B5.5B2A4B3.2A$23.7B5.11B3.A$23.8B4.4BD7BA.A$22.10B5.B3D4B
2.2A$20.11B6.D2B2D2B$20.13B4.6B$19.16B2.6B.BA$18.18B.7BA.A$17.27B.2A$
16.27B$15.28B$16.28B$16.27B2.BA$15.30BA.A$15.31BA$15.29B.B$16.26B$16.
25B$16.22B$16.12B.4B$16.10B5.B$16.10B4.3B$15.10B5.B2AB$15.10B6.2A$7.
2A6.10B$8.A7.9B$8.A.AB4.9B.2B$9.2AB.14B2A$11.16B2A$11.14B.2B$12.13B$
11.14B.2B$9.18B2A$7.18B.B2A$7.2BC15B2.B$6.3BCBC4B.8B$7.2B3C4B2.7B$6.
5BC4B2.7B$5.10B3.7B$4.4B11.7B$4.3B12.7B$2.4B13.6B$2.2A16.5B$3.A15.5B$
3A16.5B$A20.B.B$20.3B$20.B2AB$21.2A!
#C [[ VIEWONLY THUMBNAIL ZOOM 6 HEIGHT 500 ]]

(It's a left turn in my book, not a right turn -- you're starting with a simple rotation of the standard Herschel orientation, not a reflection of it.)
User avatar
dvgrn
Moderator
 
Posts: 5546
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: The Hunting of the New Herschel Conduits

Postby dvgrn » June 14th, 2015, 5:38 pm

Kazyan wrote:
A for awesome wrote:Does anyone know of A. a catalyst or B. a G-to-block that can complete this conduit...

The Spartan block factory would work...but the color of the FNG is wrong, if reflected via Snark. So is the first glider of a dependent conduit and the glider of the large B-to-H non-spartan conduit. Bah humbug. Adding more conduits to get a different-colored glider will do it, but then the repeat time starts to get sketchy.

It looks like it might turn out to be useful somewhere somehow -- nice and simple, even if we don't quite know right now where we need a block-eating conduit.

Maybe we can keep track of it as a composite Fx293 for now. It works okay with a prefix F171 and a couple of Snarks -- the repeat time is no sketchier than usual for an F171 (227 ticks).

x = 97, y = 69, rule = LifeHistory
22.2A3.2A$22.2A2.A.3A$26.A4.A$22.4A.2A2.A$22.A2.A.A.A.2A$25.A.A.A.A$
26.2A.A.A$30.A2$16.2A$17.A7.2A$17.A.A5.2A$18.2A4$44.2A$44.2A2$28.2A$
28.A$29.3A21.2A$31.A21.A$51.A.A$51.2A2$4.A$2.5A14.2A31.2A$.A5.A13.A
15.2A4.2A9.A$.A2.3A12.A.A14.A.A3.A.A7.A.A$2A.A15.2A15.A6.A8.2A$A2.4A
16.A11.2A$.2A3.A3.2A11.3A20.2A$3.3A4.2A14.A19.2A$3.A21.2A$2A.A$2A.2A
3$11.2A$12.A30.D$9.3A5.2A24.D.D40.D$9.A8.A24.3D38.3D$18.A.A24.D21.2C
15.D.D$19.2A46.2C15.D$95.C$94.C.C$13.A42.C37.C.C$13.3A39.C.C37.C$16.A
38.2C$15.2A47.2C$64.C$65.C$64.2C4$14.C$14.C.C$14.3C$16.C2$23.2A$24.A$
21.3A$7.2A12.A$8.A$5.3A$5.A!
User avatar
dvgrn
Moderator
 
Posts: 5546
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: The Hunting of the New Herschel Conduits

Postby simsim314 » June 14th, 2015, 7:53 pm

Not so useful, and still pretty edgy block factory based on Kazyan's finding:

x = 29, y = 28, rule = B3/S23
21bo$19b3o$10bo7bo$8b3o7b2o$7bo$7b2o2$4b2o10b2o7b2o$4bobo8bo2bo6bo$5b
2o4b2o2bo2bo4bobo$2o9b2o3b2o5b2o$bo$bobo$2b2o3$27b2o$27b2o6$21b3o$7b2o
12bo$6bobo13bo$6bo$5b2o!
User avatar
simsim314
 
Posts: 1676
Joined: February 10th, 2014, 1:27 pm

Re: The Hunting of the New Herschel Conduits

Postby Scorbie » June 15th, 2015, 4:57 am

This H-B seems new:
x = 30, y = 33, rule = B3/S23
7b2o$7bobo$5bobob3o10bob2o$4bobo5bo9b2obo$2o3bo5b2o$2o18b5o$14b2o4bo4b
o2b2o$13bobo7bo2bo2bo$13b2o8b2obobo$20bo5bob2o$19bobo4bo$19bo2bo2b2o$
20b2o12$11b3o6b2o$12bo7b2o$10b3o5$16b2o$16b2o!


And there's an H-boat, although I'm not sure how useful this would be:
x = 32, y = 14, rule = B3/S23
8b2o$3b2o2bo2bo17b2o$3bo4bobo17bobo$2obo5bo20bo$bobob2o23b2o$o2bo2bo$
2o2bo4bo$5b5o2$4bob2o12b3o$4b2obo3b2o8bo$12bo6b3o$9b3o$9bo!
Best wishes to you, Scorbie
User avatar
Scorbie
 
Posts: 1375
Joined: December 7th, 2013, 1:05 am

Re: The Hunting of the New Herschel Conduits

Postby A for awesome » June 15th, 2015, 10:24 am

A for awesome wrote:A HL226B:
rle

A related HL227Pi that looks hard to place:
x = 51, y = 48, rule = LifeHistory
16.C$15.C.C$9.2C4.BC2B$8.B2CB4.2B$8.2B.B2.B.3B$9.10B$8.12B$5.3D13B$5.
D15B$5.3D13B$7.15B$11.11B$10.12B$9.13B$9.13B3.B$8.15B.B2C$6.B.17B2C$
5.18B.2B$3.20B$3.19B8.2C3.2C$2.21B6.B2CB.B2CB$2.21B6.4B.3B$2.23B5.2B
3.B6.2B$4.42B$2.46B$2.2C2.42B.B$3.C3.42B2C$3C4.42B2C$C8.36B.4B$10.32B
4.4B$11.28B$12.26B$13.B2.3B6.3B2.11B$30.12B$30.12B$31.11B$29.4B.4B3CB
$29.2C4.4BC2B$30.C4.2B3C2B$27.3C6.6B$27.C8.7B$38.B.4B$41.4B$42.3B$44.
B2C$44.BC.C$47.C$47.2C!

Edit: And a slightly different HF182E:
x = 51, y = 36, rule = LifeHistory
13.6B$11.8B$11.11B$9.13B$9.11B2D$8.13B2D$.B3.14B3D$2CB.18B8.2C3.2C$2C
21B6.B2CB.B2CB$.22B6.4B.3B$.24B5.2B3.B6.2B$3.B2.40B$6.42B$7.41B.B$9.
40B2C$10.39B2C$11.34B.4B$11.31B4.4B$9.5B.24B$9.2C4.B.21B$10.C5.2B7.3B
2.11B$7.3C5.B2CB11.12B$7.C8.2C12.12B$31.11B$29.4B.4B3CB$29.2C4.4BC2B$
30.C4.2B3C2B$27.3C6.6B$27.C8.7B$38.B.4B$41.4B$42.3B$44.B2C$44.BC.C$
47.C$47.2C!

Edit 2: An eater 5 can function as an ERx42B:
x = 15, y = 22, rule = LifeHistory
8.D$7.3D$6.2DBDB$6.5B$6.5B$6.6B$5.7B$5.6B$3.10B$.9BC3B$9B3C2B$.8BCBC
3B$3.8BC2B$4.9B$5.8B$2.B3.5B$.2CB.3B$.2C3BC$2.B.BC.C$5.BC.C$8.C$8.2C!

Unfortunately, it is incompatible with my HF182E.
Edit 3: If anyone ever wants an edgeshooting non-Spartan hive inserter...:
x = 48, y = 39, rule = LifeHistory
19.C$17.3C$16.C$15.C.C$14.BC.C$13.3BC$11.4B$9.6B$5.2B.7B$2.5B.8B$.2D
14B.B8.2C3.2C$D2BD16B6.B2CB.B2CB$.2D17B6.4B.3B$2.20B5.2B3.B6.2B$4.39B
$5.40B$6.39B.B$7.39B2C$7.39B2C$7.35B.4B$7.32B4.4B$7.29B$7.4B.B.21B$8.
3B2.2B7.3B2.11B$12.B2CB11.12B$13.2C12.12B$28.11B$26.4B.4B3CB$26.2C4.
4BC2B$27.C4.2B3C2B$24.3C6.6B$24.C8.7B$35.B.4B$38.4B$39.3B$41.B2C$41.B
C.C$44.C$44.2C!

Edit 4: A HR(x?)204BG:
x = 47, y = 36, rule = LifeHistory
11.4B$11.5B$10.7B$9.9B$8.5B2D6B$8.4BDBD8B$8.3BD2BD9B.2B$7.5B2D12B2C$
8.18B2C$2C5.17B.2B$.C4.17B$.C.CB2.15B$2.2CB.16B$4.13B8.2C3.2C$4.14B6.
B2CB.B2CB$4.14B6.4B.3B$4.16B5.2B3.B6.2B$4.37B$2.2C40B.B$.C.C39B.B2C$.
C4.39B2C$2C5.B.34B.2B$9.31B.2B$10.27B$10.24B$10.B.21B$11.2B7.3B2.11B$
10.B2CB11.12B$11.2C12.12B$26.11B$24.4B.4B3CB$24.2C4.4BC2B$25.C4.2B3C
2B$22.3C6.6B$22.C8.7B$33.B.4B!
Last edited by A for awesome on June 15th, 2015, 12:33 pm, edited 3 times in total.
x₁=ηx
V ⃰_η=c²√(Λη)
K=(Λu²)/2
Pₐ=1−1/(∫^∞_t₀(p(t)ˡ⁽ᵗ⁾)dt)

$$x_1=\eta x$$
$$V^*_\eta=c^2\sqrt{\Lambda\eta}$$
$$K=\frac{\Lambda u^2}2$$
$$P_a=1-\frac1{\int^\infty_{t_0}p(t)^{l(t)}dt}$$

http://conwaylife.com/wiki/A_for_all

Aidan F. Pierce
User avatar
A for awesome
 
Posts: 1778
Joined: September 13th, 2014, 5:36 pm
Location: 0x-1

Re: The Hunting of the New Herschel Conduits

Postby dvgrn » June 15th, 2015, 10:47 am

Scorbie wrote:This H-B seems new...

With that early glider output so far forward, it seems hard to imagine getting any use out of the B output -- I looked through all the B inputs, and it seems as if there are always going to be conduits in the way of that glider. The resulting Herschel could be used, possibly, but then there's an extra block to dispose of. Do you have a workable connection in mind here?

Scorbie wrote:And there's an H-boat, although I'm not sure how useful this would be...

This one has been around for a while -- it's been used in multi-stage stable reflectors, for example, because the boat can be used to quickly reflect the first output glider from a Silver reflector to delete the extra beehive. Then you just have to get a Herschel to the H-to-boat converter somehow... Syringes seem a lot simpler nowadays, given that the converter isn't Spartan (it can be made so at the cost of an extra glider, but that isn't really an improvement!). It's always nice to have another resettable glider reflector in the toolbox, though.
User avatar
dvgrn
Moderator
 
Posts: 5546
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: The Hunting of the New Herschel Conduits

Postby dvgrn » June 15th, 2015, 11:05 am

A for awesome wrote:A related HL227Pi that looks hard to place...
Edit: And a slightly different HF182E...
Edit 2: An eater 5 can function as an ERx42B... Unfortunately, it is incompatible with my HF182E.

I'm pretty skeptical about adding unlikely-looking conduits like these to the current elementary conduits collection, even in the hope that someday something might be found that connects to it. Would have to inspect and reject them too often in the meantime, and the odds seem fairly good that no reasonable connection is possible. If someone wants to start a separate Stamp Collection of Misfit Conduits, though, I'd be most grateful...!

Also, where does the "E" come from exactly? Are there other conduits with this output that I've missed, and is there a working E-to-X that this conduit does connect to? I don't recognize the shape, and it seems to explode southwards fairly quickly and takes out one of the catalysts in the H182E.

Without a standard orientation and direction defined for the "E" object, it doesn't really make sense to include F-for-forward in the conduit name. This "HF182E" defines a standard orientation and direction, I suppose, but it seems as if the primary direction of travel for the "E" is really south, not east (if you rotate the conduit 90 degrees right so that the Herschel is in standard orientation). That would make this conduit some kind of right turn (?)
User avatar
dvgrn
Moderator
 
Posts: 5546
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: The Hunting of the New Herschel Conduits

Postby A for awesome » June 15th, 2015, 11:12 am

dvgrn wrote:Also, where does the "E" come from exactly? Are there other conduits with this output that I've missed, and is there a working E-to-X that this conduit does connect to? I don't recognize the shape, and it seems to explode southwards fairly quickly and takes out one of the catalysts in the H182E.

Without a standard orientation and direction defined for the "E" object, it doesn't really make sense to include F-for-forward in the conduit name. This "HF182E" defines a standard orientation and direction, I suppose, but it seems as if the primary direction of travel for the "E" is really south, not east (if you rotate the Herschel 90 degrees right into standard orientation). That would make this conduit some kind of right turn (?)

The "E" comes from here.
I suppose it might be a right turn, but I was taking the standard orientation as the one where it matches the Herschel standard orientation most closely:
x = 11, y = 4, rule = LifeHistory
C7.C$C.C5.C.C$3C5.3C$2.C6.C!

Also, it does start moving east in that orientation; it only goes south after the LOM disappears. I think it could go either way. Your call.

Edit: I've tried out all Spartan catalyses of the E from the H(F?)182E I know of by hand and there is no way to get it to do anything useful. If there is a conduit that connects, it would probably have to be non-Spartan or transparent. As I don't have any search programs that can find catalysts, this conduit will probably not connect to anything in the near future.
x₁=ηx
V ⃰_η=c²√(Λη)
K=(Λu²)/2
Pₐ=1−1/(∫^∞_t₀(p(t)ˡ⁽ᵗ⁾)dt)

$$x_1=\eta x$$
$$V^*_\eta=c^2\sqrt{\Lambda\eta}$$
$$K=\frac{\Lambda u^2}2$$
$$P_a=1-\frac1{\int^\infty_{t_0}p(t)^{l(t)}dt}$$

http://conwaylife.com/wiki/A_for_all

Aidan F. Pierce
User avatar
A for awesome
 
Posts: 1778
Joined: September 13th, 2014, 5:36 pm
Location: 0x-1

Re: The Hunting of the New Herschel Conduits

Postby dvgrn » June 15th, 2015, 12:55 pm

A for awesome wrote:The "E" comes from here.
I suppose it might be a right turn, but I was taking the standard orientation as the one where it matches the Herschel standard orientation most closely... Also, it does start moving east in that orientation; it only goes south after the LOM disappears.

Yeah, I saw that, but it's a fairly short-lived spark. Would probably need a transparent catalyst to pull the reaction in that direction, as you say. A CatForce search would definitely be a good idea here.

Meanwhile, the other converter inputs and outputs tend to have a fairly clear direction -- either the direction the reaction is usually traveling away from (to reach an output location) or the direction the object is about to travel toward (starting from an input). Objects that just wobble around more or less in one place are much harder to get hold of with a series of catalysts, even if they explode over time.

That makes the E-heptomino potentially pretty interesting, because it does travel quite well. Not very cleanly, but R's and pi's are about equally messy, and the tub-with-tail is a good cleanup. You've shown that the E can be produced "naturally" from a catalyzed reaction, so it's just a question of whether it appears near the edge of a reaction envelope, in a workable orientation, often enough to be useful.

-- I wonder, how long does it take before one of these shows up in a frame-by-frame analysis of large random soups, at T>=100 let's say? I'm wondering if the frequency of natural E objects might be comparable to QBs at least.

Actually I'm surprised that E-heptominos haven't been mentioned before (that I can remember, anyway). They're not in simsim314's list of potential input/output objects, for example... and yet I suspect they would start showing up all over the place if people started looking for them. There's a really long low-population bottleneck before the E shows up -- population counts

10, 8, 10, 12, 12, 11, 12, 10, 10, 8,10, 7, 7, 9, 10 ...

with the second "7" being the E. Any time an active population drops below a dozen cells or so, the resulting pattern acts as an attractor of sorts: populations fluctuate up and down, and in the "up" direction there are untold billions of possibilities... but when the population drops into single digits, there are relatively very few arrangements, so the odds are much improved that the reaction will drop into a known "channel" and produce a clean pi or B or E.

[For similar reasons, a much higher percentage of small-pop-count active objects are symmetrical, which accounts for some of the prevalence of honeyfarms and traffic lights and pi-heptominos and so on. Unfortunately the E-heptomino won't get a boost from that effect.]

It occurs to me that it's vaguely possible that a known conduit already includes an E stage, most likely 42 ticks before some B output or 62 ticks before an H output -- unrecognized just because nobody knew to look for it. If something like that can be found, that could give new working input and output options.
User avatar
dvgrn
Moderator
 
Posts: 5546
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: The Hunting of the New Herschel Conduits

Postby Kazyan » June 15th, 2015, 1:33 pm

I missed this C-to-G in previous searches because I'd messed up the symmetries and reflections in the BTS variant I'm using, but have since corrected that.

x = 13, y = 10, rule = B3/S23
5b2o$5bo$ob2obo4b2o$2obob2o3b2o2$4b3o$3bo2bo$3b2o5b3o$9bo2bo$9b2o!
Tanner Jacobi
User avatar
Kazyan
 
Posts: 807
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Postby Extrementhusiast » June 15th, 2015, 1:56 pm

Doing a trawl through the first one hundred C1 soups for xs14_354cgs26 (that two-eater combo used in F116 and other conduits), I found this reaction that adds the second eater onto the first:
x = 19, y = 36, rule = B3/S23
11bo$10b3o$9bo$3b2o4b2o$3b2o5$15b2o$14bobo$7b2o5b2o$7b2o11$12bo$2o9bob
o$o10bobo$b3o8bo$3bo$7b2o7b2o$6bo2bo5bo2bo$7b2o7b2o2$12bo$11bobo$11bob
o$12bo!

However, it gets even better, as the only relevant part at gen 56 is this:
x = 19, y = 14, rule = B3/S23
12bo$10bo2bo$2o8b2obo$o10bobo$b3o8bo$3bo$7b2o7b2o$6bo2bo5bo2bo$7b2o7b
2o2$12bo$11bobo$11bobo$12bo!

The predecessor to that need not be exact, as the interaction only lasts three generations.
I Like My Heisenburps! (and others)
User avatar
Extrementhusiast
 
Posts: 1745
Joined: June 16th, 2009, 11:24 pm
Location: USA

Re: The Hunting of the New Herschel Conduits

Postby dvgrn » June 15th, 2015, 2:36 pm

A for awesome wrote:Edit 4: A HR(x?)204BG...

To cut down on the confusion if possible (though it may be inevitable) I just updated the first post of this thread with an illustration of the canonical orientations of most of the input/output objects we've been dealing with lately. Also threw in a link to the message with the current latest version of the elementary-conduits collection -- please remind me if the link gets out of date!

If you rotate the HR202BG so that the Herschel is in canonical form, then the BG grandchild is a 90-degree right rotation of the BG's canonical form (plus two ticks). No mirror reflection, therefore no "x":

Code: Select all
x = 42, y = 47, rule = LifeHistory
20.A11.A$20.3A7.3A$23.A5.A$22.2A5.2A$22.9B$22.7B$21.9B.B$20.13B.B$21.
17B$19.20B$14.B.24B$13.2AB.25B$13.2A19B2D6B$14.B.19BDBD4B$16.20B2D4B$
16.25B$16.24B$16.12B.10B$16.10B3.9B$16.10B3.9B$15.10B4.9B$15.10B4.8B$
7.2A6.10B6.6B$8.A7.9B7.4B$8.A.AB4.9B.2B5.2B$9.2AB.14B2A3.4B$11.16B2A
3.B2AB$11.14B.2B5.2A$12.13B$11.14B.2B$9.18B2A$7.18B.B2A$7.2BC15B2.B$
6.3BCBC4B.8B$7.2B3C4B2.7B$6.5BC4B2.7B$5.10B3.7B$4.4B11.7B$4.3B12.7B$
2.4B13.6B$2.2A16.5B$3.A15.5B$3A16.5B$A20.B.B$20.3B$20.B2AB$21.2A!
#C [[ VIEWONLY THUMBNAIL ZOOM 9 ]]

The form shown above never actually appears in the HR202BG reaction -- but if it did, it would appear where it's shown, and so T=202 is the right number of ticks to measure to.

I really have no idea if I picked a good canonical version of the BG. Maybe it should be changed to the "R-loaf" form two ticks later. But I think even then there are some BG variants that don't go through that stage. So for now, given that the elementary-conduits stamp collection uses the above form, it's probably worth trying to make that work.

... Here again it seems pretty worrisome that the output is so far inside the reaction envelope, maybe especially for this orientation. It doesn't seem to leave very much clean area for catalysts, before the final HR202BG block catalyst gets destroyed.
User avatar
dvgrn
Moderator
 
Posts: 5546
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: The Hunting of the New Herschel Conduits

Postby Kazyan » June 15th, 2015, 6:02 pm

A for awesome wrote:Edit 3: If anyone ever wants an edgeshooting non-Spartan hive inserter...:
x = 48, y = 39, rule = LifeHistory
19.C$17.3C$16.C$15.C.C$14.BC.C$13.3BC$11.4B$9.6B$5.2B.7B$2.5B.8B$.2D
14B.B8.2C3.2C$D2BD16B6.B2CB.B2CB$.2D17B6.4B.3B$2.20B5.2B3.B6.2B$4.39B
$5.40B$6.39B.B$7.39B2C$7.39B2C$7.35B.4B$7.32B4.4B$7.29B$7.4B.B.21B$8.
3B2.2B7.3B2.11B$12.B2CB11.12B$13.2C12.12B$28.11B$26.4B.4B3CB$26.2C4.
4BC2B$27.C4.2B3C2B$24.3C6.6B$24.C8.7B$35.B.4B$38.4B$39.3B$41.B2C$41.B
C.C$44.C$44.2C!


Simkin found a way to get a "nice glider" out of where that suppressed TL would appear. Since this is non-Spartan anyway, redirecting it with snarks shows that it happens to be the right color to make it an H-to-R with a repeat time of 219--which is at least better than F171.

x = 64, y = 67, rule = LifeHistory
44.2C$44.C$41.BC.C$41.B2C$39.3B$38.4B$35.B.4B$24.C8.7B$24.3C6.6B$27.C
4.2B3C2B$26.2C4.4BC2B$26.4B.4B3CB$28.11B$13.2C12.12B$12.B2CB11.12B$8.
3B2.2B7.3B2.11B10.C$7.4B.B.21B11.3C$7.29B9.C8.A$7.32B6.2C5.3A$7.40B4.
A$7.38B6.2A$7.39B2.5B$6.40B2.3B$5.47B7.2A$4.39B.8B8.A$2.20B5.2B3.B6.
3B3.10B3.B.A.2A$.19B6.4B.3B5.BC2B2.7B2A2B.B3A2.A$D19B6.B2CB.B2CB4.C.C
3.7B2A3BAB2.2A$3D14B.B8.2C3.2C6.C4.12B4A$.D5B.8B27.2AB.7B3.2B.A$3.4B.
7B27.A.AB.7B2.B3A$2.4B3.6B27.A5.4B4.A$3.4B4.4B26.2A5.4B5.5A$4.4B5.3BC
30.4B10.A$5.4B5.BC.C28.4B9.A$6.4B5.C.C27.4B10.2A$7.4B5.C27.4B$8.4B5.
3C23.4B$9.4B6.C22.4B$10.4B27.4B$11.4B25.4B$12.4B23.4B$13.4B21.4B$14.
4B19.4B$15.4B6.A10.4B$16.4B5.3A7.4B$17.4B7.A5.4B$18.4B5.2A4.4B$19.4B
4.9B$20.4B5.6B$21.4B2.8B$22.15B$23.14B$24.13B$25.10B.B2A$27.3B2AB3.BA
.A$27.3B2AB6.A$29.4B6.2A$29.3B$26.AB.2B$25.A.AB2AB$25.A.ABABAB$22.2A.
A.A.A.A2.A$22.A2.A2.2A.4A$24.2A4.A$30.A.A$31.2A!


Speaking of Aiden Pierce's ideas, I asked gencols if any convenient signals could come out of colliding an existing one with a block, like that H-to-R that was solved with F171. These are the results that looked "easy", though admittedly I was begin generous with the Herschel:

Code: Select all
#C [[ VIEWONLY ]]
x = 250, y = 160, rule = LifeHistory
10.240F$10.F28.2F28.2F28.2F28.2F28.2F28.2F28.2F28.F$10.F28.2F28.2F28.
2F28.2F28.2F28.2F28.2F14.2A12.F$10.F28.2F28.2F28.2F28.2F28.2F28.2F15.
2A11.2F14.2A12.F$10.F28.2F16.2A10.2F28.2F28.2F28.2F28.2F15.2A11.2F28.
F$10.F28.2F16.2A10.2F16.2A10.2F28.2F18.2A8.2F28.2F28.2F28.F$10.F28.2F
28.2F16.2A10.2F28.2F18.2A8.2F28.2F28.2F28.F$10.F28.2F28.2F28.2F17.2A
9.2F28.2F28.2F28.2F28.F$10.F28.2F28.2F28.2F17.2A9.2F28.2F20.2A6.2F28.
2F28.F$10.F28.2F28.2F28.2F28.2F28.2F20.2A6.2F28.2F28.F$2D4.2D2.F9.A
18.2F9.A18.2F9.A18.2F9.A18.2F9.A18.2F9.A18.2F28.2F9.A18.F$2D4.2D2.F9.
A.A16.2F9.A.A16.2F9.A.A16.2F9.A.A16.2F9.A.A16.2F9.A.A16.2F28.2F9.A.A
16.F$2D4.2D2.F9.3A16.2F9.3A16.2F9.3A16.2F9.3A16.2F9.3A16.2F9.3A16.2F
6.A21.2F9.3A16.F$2D4.2D2.F11.A16.2F11.A16.2F11.A16.2F11.A16.2F11.A16.
2F11.A16.2F6.A.A19.2F11.A16.F$8D2.F28.2F28.2F28.2F28.2F28.2F28.2F6.3A
19.2F28.F$8D2.F14.2A12.2F28.2F28.2F28.2F28.2F28.2F8.A19.2F28.F$2D4.2D
2.F14.2A12.2F28.2F28.2F10.2A16.2F28.2F28.2F28.2F28.F$2D4.2D2.F28.2F
28.2F28.2F10.2A16.2F28.2F28.2F28.2F28.F$2D4.2D2.F28.2F28.2F28.2F28.2F
28.2F28.2F28.2F28.F$2D4.2D2.240F$10.120F$10.F28.2F28.2F28.2F28.F$10.F
28.2F11.2A15.2F28.2F28.F$10.F28.2F11.2A15.2F28.2F28.F$10.F28.2F28.2F
28.2F28.F$10.F28.2F28.2F28.2F28.F$10.F28.2F28.2F28.2F7.2A19.F$10.F28.
2F28.2F13.2A13.2F7.2A19.F$10.F13.2A13.2F28.2F13.2A13.2F28.F$10.F13.2A
13.2F28.2F28.2F28.F$7D3.F28.2F28.2F28.2F28.F$8D2.F28.2F28.2F28.2F28.F
$2D4.2D2.F28.2F28.2F28.2F28.F$2D4.2D2.F28.2F28.2F28.2F28.F$7D3.F10.A
17.2F10.A17.2F10.A17.2F10.A17.F$7D3.F9.3A16.2F9.3A16.2F9.3A16.2F9.3A
16.F$2D4.2D2.F9.A.2A15.2F9.A.2A15.2F9.A.2A15.2F9.A.2A15.F$2D4.2D2.F
28.2F28.2F28.2F28.F$8D2.F28.2F28.2F28.2F28.F$7D3.F28.2F28.2F28.2F28.F
$10.120F$10.F28.F$10.F28.F$10.F28.F$10.F28.F$10.F28.F$10.F10.2D4.2D
10.F$10.F10.2D4.2D10.F$10.F11.2D2.2D11.F$10.F12.4D12.F$7D3.F13.2D13.F
$8D2.F13.2D13.F$2D4.2D2.F12.4D12.F$2D4.2D2.F11.2D2.2D11.F$7D3.F10.2D
4.2D10.F$7D3.F10.2D4.2D10.F$2D4.2D2.F28.F$2D4.2D2.F28.F$2D4.2D2.F28.F
$2D4.2D2.F28.F$10.90F$10.F28.2F28.2F28.F$10.F28.2F28.2F28.F$10.F28.2F
28.2F28.F$10.F28.2F11.2A15.2F28.F$10.F28.2F11.2A15.2F28.F$10.F28.2F
28.2F28.F$10.F28.2F28.2F28.F$10.F28.2F28.2F28.F$10.F28.2F28.2F28.F$8D
2.F28.2F28.2F28.F$8D2.F28.2F28.2F28.F$2D4.2D2.F14.2A12.2F28.2F28.F$2D
4.2D2.F14.2A12.2F28.2F28.F$2D4.2D2.F9.3A16.2F9.3A16.2F9.3A8.2A6.F$2D
4.2D2.F9.A.A16.2F9.A.A16.2F9.A.A8.2A6.F$2D4.2D2.F9.A.A16.2F9.A.A16.2F
9.A.A16.F$2D4.2D2.F28.2F28.2F28.F$2D4.2D2.F28.2F28.2F28.F$2D4.2D2.F
28.2F28.2F28.F$10.120F$10.F28.2F28.2F28.2F28.F$10.F28.2F28.2F28.2F28.
F$10.F28.2F28.2F28.2F28.F$10.F28.2F28.2F28.2F28.F$10.F11.2A15.2F28.2F
28.2F28.F$10.F11.2A15.2F28.2F28.2F28.F$10.F28.2F28.2F28.2F28.F$10.F
28.2F28.2F28.2F28.F$10.F28.2F28.2F28.2F13.2A13.F$.6D3.F28.2F28.2F28.
2F13.2A13.F$8D2.F28.2F5.2A21.2F8.2A18.2F28.F$2D4.2D2.F28.2F5.2A21.2F
8.2A18.2F28.F$2D8.F28.2F28.2F28.2F28.F$2D8.F9.3A16.2F9.3A16.2F19.3A6.
2F9.3A16.F$2D8.F9.A2.A15.2F9.A2.A15.2F19.A2.A5.2F9.A2.A15.F$2D8.F11.
2A15.2F11.2A15.2F21.2A5.2F11.2A15.F$2D4.2D2.F28.2F28.2F28.2F28.F$8D2.
F28.2F28.2F28.2F28.F$.6D3.F28.2F28.2F28.2F28.F$10.150F$10.F28.2F28.2F
28.2F28.2F28.F$10.F28.2F28.2F28.2F16.2A10.2F28.F$10.F28.2F28.2F28.2F
16.2A10.2F28.F$10.F28.2F28.2F28.2F28.2F28.F$10.F28.2F28.2F28.2F28.2F
28.F$10.F28.2F28.2F28.2F28.2F28.F$10.F28.2F28.2F10.2A16.2F28.2F28.F$
10.F28.2F28.2F10.2A16.2F28.2F7.2A19.F$10.F28.2F28.2F28.2F28.2F7.2A19.
F$2D.2D.2D2.F10.2A16.2F5.2A21.2F28.2F28.2F28.F$2D.2D.2D2.F10.2A16.2F
5.2A21.2F28.2F28.2F28.F$2D.2D.2D2.F28.2F28.2F28.2F28.2F28.F$2D.2D.2D
2.F28.2F28.2F28.2F28.2F28.F$2D.2D.2D2.F28.2F28.2F28.2F28.2F28.F$2D.2D
.2D2.F28.2F28.2F28.2F28.2F28.F$2D.2D.2D2.F9.2A17.2F9.2A17.2F9.2A17.2F
9.2A17.2F9.2A17.F$2D.2D.2D2.F9.A.A16.2F9.A.A16.2F9.A.A16.2F9.A.A16.2F
9.A.A16.F$8D2.F11.2A15.2F11.2A15.2F11.2A15.2F11.2A15.2F11.2A15.F$.6D
3.F28.2F28.2F28.2F28.2F28.F$10.150F$10.F28.F$10.F28.F$10.F28.F$10.F
28.F$10.F28.F$10.F28.F$10.F16.2A10.F$10.F16.2A10.F$10.F28.F$.6D3.F28.
F$8D2.F12.A15.F$2D4.2D2.F11.A.A14.F$2D4.2D2.F10.A3.A13.F$2D4.2D2.F11.
3A14.F$2D4.2D2.F9.2A3.2A12.F$2D2.4D2.F28.F$2D2.3D3.F28.F$8D2.F28.F$.
4D.2D2.F28.F$10.210F$10.F28.2F28.2F28.2F28.2F28.2F28.2F28.F$10.F28.2F
28.2F28.2F28.2F28.2F28.2F28.F$10.F28.2F28.2F28.2F28.2F28.2F28.2F28.F$
10.F28.2F28.2F28.2F28.2F12.2A14.2F28.2F28.F$10.F9.2A17.2F28.2F15.2A
11.2F28.2F12.2A14.2F28.2F28.F$10.F9.2A17.2F28.2F15.2A11.2F10.2A16.2F
28.2F28.2F28.F$10.F28.2F28.2F28.2F10.2A16.2F28.2F15.2A11.2F17.2A9.F$
10.F28.2F28.2F28.2F28.2F28.2F15.2A11.2F17.2A9.F$10.F9.A18.2F9.A18.2F
9.A18.2F28.2F9.A18.2F9.A18.2F9.A18.F$8D2.F9.A.A16.2F9.A.A16.2F9.A.A
16.2F9.A18.2F9.A.A16.2F9.A.A16.2F9.A.A16.F$8D2.F9.3A16.2F9.3A16.2F9.
3A16.2F9.A.A16.2F9.3A16.2F9.3A16.2F9.3A16.F$2D8.F10.A17.2F10.A17.2F
10.A17.2F9.3A16.2F10.A17.2F10.A17.2F10.A17.F$2D8.F28.2F14.2A12.2F28.
2F10.A17.2F28.2F28.2F28.F$7D3.F28.2F14.2A12.2F28.2F28.2F28.2F28.2F28.
F$7D3.F28.2F28.2F28.2F28.2F28.2F28.2F28.F$2D8.F28.2F28.2F28.2F28.2F
28.2F28.2F28.F$2D8.F28.2F28.2F28.2F28.2F28.2F28.2F28.F$8D2.F28.2F28.
2F28.2F28.2F28.2F28.2F28.F$8D2.210F!


(W is 'wing', which is the R-loaf phase of a BG. Writing out BG and messing up the single-letter column annoyed me.)

For these, the obvious procedure is to clean up all the junk, and use a glider form somewhere else in circuit to repair the block. It may be worth collecting H-to-blocks...I'll do that if no one else wants to. Snarking an FNG back into the spartan block factory will only work less than half the time, since it's color-dependent and will not always fit; putting an FNG into a syringe hooked up to an H-to-block may be more reliable.
Tanner Jacobi
User avatar
Kazyan
 
Posts: 807
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Postby dvgrn » June 15th, 2015, 7:27 pm

Kazyan wrote:Simkin found a way to get a "nice glider" out of where that suppressed TL would appear. Since this is non-Spartan anyway, redirecting it with snarks shows that it happens to be the right color to make it an H-to-R with a repeat time of 219--which is at least better than F171.

So this is another nice example of a stop-and-restart adjustable Herschel conduit -- fairly compact, at least for low N. Here it is as an HFx(307+8N)R, HFx(335+8N)B, or F394+8N -- i.e., HF(394+8N)H:

x = 67, y = 75, rule = LifeHistory
43.A$41.3A$40.A$40.2A$31.2A8.B$31.A.A7.3B$32.AB6.6B$31.2B3.B2.10B$28.
B2.19B3.2B2.2B$27.12BD3B2A15BD$26.13B2D2B2A15BDBD$26.B2D11B2D18B3DB$
26.2B2D10BD21BD$25.3BD10BD21B$25.11B5.13B.B$24.12B5.7B.B$23.12B$22.
14B$16.15B2.4B$15.16B3.4B$15.17B3.4B$15.17B4.4B$17.16B4.4B$14.B.17B5.
4B$13.2CB.17B5.4B$13.2C20B5.4B$14.B.14B3.B2C5.4B$16.13B4.C2.C5.4B$16.
12B6.2C.C5.4B$16.13B8.C6.4B$16.12B9.2C6.4B$16.10B20.4B$16.10B21.4B$
15.10B23.4B10.2A$15.10B24.4B9.A$7.2C6.10B25.4B10.A$8.C7.9B19.2A5.4B5.
5A$8.C.CB4.9B.2B17.A5.4B4.A$9.2CB.14B2C16.A.AB.7B2.B3A$11.16B2C17.2AB
.7B3.2B.A$11.14B.2B20.12B4A$12.13B23.7B2A3BAB2.2A$11.14B.2B20.7B2A2B.
B3A2.A$9.18B2C19.10B3.B.A.2A$7.18B.B2C18.8B8.A$7.2BC15B2.B18.9B7.2A$
6.3BCBC4B.8B20.4B2.3B$7.2B3C4B2.7B19.4B3.5B$6.5BC4B2.7B18.4B7.2A$5.
10B3.7B17.4B8.A$4.4B11.8BC13.4B10.3A$4.3B12.7BC.C11.4B13.A$2.4B13.8BC
4.A6.4B$2.2C15.6B.B3.3A5.4B$3.C15.5B5.A7.4B$3C16.6B4.2A5.4B$C16.2CB.
10B4.4B$16.C.CB3.6B5.4B$16.C6.8B2.4B$15.2C4.15B$21.14B$21.13B$19.2AB.
10B$18.A.AB3.B2A3B$18.A6.B2A3B$17.2A6.4B$26.3B$27.2B.BA$26.B2ABA.A$
25.BABABA.A$23.A2.A.A.A.A.2A$23.4A.2A2.A2.A$27.A4.2A$25.A.A$25.2A!

Kazyan wrote:Speaking of Aiden Pierce's ideas, I asked gencols if any convenient signals could come out of colliding an existing one with a block, like that H-to-R that was solved with F171. These are the results that looked "easy"...

If we go very far along these lines, it might be time to start a whole new incomplete-conduits collection -- say, all conduits that either leave an extra piece of junk behind in some accessible location, or lose a small still life from some accessible location.

Just recording the time it takes to get from R to B, H to QB, or whatever won't really be good enough any more, I think: for things like this to be useful, they'll have to have unique (X, Y, T) offsets, to allow very short circuit connections to be made that would otherwise be impossible. Bigger Herschel circuits nearly always have a few spare gliders, so it might work out fairly often to liberate a glider early and bounce it around to hit an incomplete conduit somewhere downstream... as long as there aren't more than one or two incomplete conduits.

Maybe eventually something like Hersrch can be adapted to offer candidate connections using a database of incomplete conduits. Until then, though, things like this are somewhat of a solution in search of a problem, aren't they?

There are certainly lots of specific cases where a new incomplete conduit might just-barely-possibly be a big help -- "I'd really like to be able to connect this H at (0,0,0) to this other H at (X, Y, T)".

But it's going to take a really big new database to change the odds of success very much -- and then, of course, half the time the only available repair glider will obstinately be the wrong color, or the second Snark won't _quite_ fit in the available space, or what have you.

...Herschel plumbing is a big pain, really -- you're always wrestling with combinatorial dragons, and victory usually consists of quietly giving up on the idea of finding the absolute smallest possible solution...!
User avatar
dvgrn
Moderator
 
Posts: 5546
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: The Hunting of the New Herschel Conduits

Postby simsim314 » June 16th, 2015, 12:02 am

I really don't see any good reason for super complex Herschel search. I would rather simplify and use more of syringe.

We have two major applications for Herschels, self constructing and not self constructing.

For the not self constructing we already saw in the gun collection that all we need is H->X, X->G. We don't need to make monsters, only one-two short conduits together with nice glider generation options is more than enough.

For the self constructing conduits, we could use modified syringe. It has the cost of silver, but it works much faster. Another option there is to use some tricks with periodic constructions (like p120/p240 gun). So for the self constructing stuff we have bunch of different ideas that work for them. I would still prefer spartan low cost G->H, but I'm saying that we have enough options already to eliminate the need for some crazy Herschel plumbing.

In general I would say crazy Herschel plumbing is much less needed today, for most constructions we don't need more than maximum three conduits. The days when Gemini used some crazy splitters and heavy conduits to make G->G Salvo are history.

We still need to use some sort of Hersrch, mainly because we need all the timing to work, but as a rule of thumb I think the guys who built the whole gun collection didn't use Hersrch even once, just to show how less needed this is today. The modern Hersrch should also work for all possibilities, and consider gliders. For example we have G4->R and R->G, so we can skip Hesrchels all together many times, they already not the default conduit, although they are the most versatile.
User avatar
simsim314
 
Posts: 1676
Joined: February 10th, 2014, 1:27 pm

Re: The Hunting of the New Herschel Conduits

Postby Kazyan » June 16th, 2015, 5:06 am

Good points. I'm getting ahead of myself with this Herschel stuff.

This thing here is periodic using a sparker (because Bellman couldn't do anything to the stray block at max-active 9 and repair-interval 12), but I'm posting it anyway because it's fast and direct enough to maybe be useful for guns.

x = 29, y = 26, rule = LifeHistory
2.2A$2B2AB$2B$2BA$BABA3.B4.BA$.AB.A2.2A2.BABA$3.2BA.2A2.BABA$3.2A3B4.
BA2B$4.4B2.5B$8.9B$8.14B$8.2B2D17B$9.B2D14BD2B$7.19B2DB$5.22B2D$5.2BC
18B2DB$4.3BCBC4B.8B.5B$5.2B3C4B2.7B$4.5BC4B2.5B$3.10B3.4B$2.4B10.2B2A
B$.4B13.A.BA$4B12.A.A.2A$3B11.3A.A$2B11.A4.A$B12.2A3.2A!


EDIT: I wonder, would it be possible to JLS something that makes both of these sparks?

x = 85, y = 40, rule = LifeHistory
68.A$26.2A3.2A26.A7.A.A$27.A4.A26.3A4.2BAB$27.A.3A30.A4.2B9.2A$24.2A.
A.A31.2A5.6B2.BA.AB$7.2A15.AB.A21.B.7B3.3B3.10B2A2B$6.B2AB6.2A7.B2A2B
13.B.13B5.3B2.11B.B2A$7.3B5.B2AB7.4B3.26BD10BD5B4.BA.A$8.B7.3B6.5B2.
4BA21BD10B2D6B6.A$6.5B3.2B.B5.7B2.4B3A18B2D11B2D5B.B4.2A$4.4BA10BD2B.
8B.4BABA15B2A2B2D13BD4B2A$3.4BABA8BD19BA15B2A3BD12B3D2B.2A$3.4BABA7B
2D23B2.2B3.22BD2B2D3B$4.4BA9B2D19B11.10B2.B3.B.2B.6B$6.13BD14B2DB16.
6B12.6B$8.26B2D2B17.3B12.7B$8.13B3.14B19.B13.6B$9.12B8.8B15.BA3.2A12.
5B$8.13B10.5B.C13.BABA3.A9.8B$9.12B4.2A3.2BAB17.BABA3.A.A5.10B$10.10B
5.A.A3.ABAB17.BA2B3.2A4.12B$10.8B9.A3.ABAB13.C.5B10.13B$10.5B12.2A3.A
B14.9B8.12B$9.6B13.B19.14B3.13B$9.7B12.3B17.2B2D26B$9.6B12.6B16.B2D
14BD13B$9.6B.2B.B3.B2.10B11.19B2D9BA4B$7.3B2D2BD22B3.2B2.23B2D7BABA4B
$6.2A.2B3D12BD3B2A15BD19BD8BABA4B$6.2A4BD13B2D2B2A15BDBD4B.8B.2BD10BA
4B$.2A4.B.5B2D11B2D18B3D4B2.7B5.B.2B3.5B$2.A6.6B2D10BD21BD4B2.5B6.3B
7.B$2.A.AB4.5BD10BD26B3.4B7.B2AB5.3B$3.2AB.11B2.3B5.13B.B13.2B2AB7.2A
6.B2AB$5.2B2A10B3.3B3.7B.B21.A.BA15.2A$5.BA.AB2.6B5.2A31.A.A.2A$6.2A
9.2B4.A30.3A.A$16.BA2B4.3A26.A4.A$16.A.A7.A26.2A3.2A$17.A!
Tanner Jacobi
User avatar
Kazyan
 
Posts: 807
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Postby dvgrn » June 16th, 2015, 5:59 pm

simsim314 wrote:I really don't see any good reason for super complex Herschel search. I would rather simplify and use more of syringe.
...
In general I would say crazy Herschel plumbing is much less needed today, for most constructions we don't need more than maximum three conduits. The days when Gemini used some crazy splitters and heavy conduits to make G->G Salvo are history.

True enough. It would be interesting to see how much smaller some of those crazy old for-fun constructions would be nowadays -- like the glider-to-Cordership converter (I've been wanting to rebuild that one properly for half a decade now -- Calcyman never took any kind of unhealthy interest in bounding-box optimization.) Or the 2c/3-wire pattern produced by heisenburp.py -- nowadays I'd put a 90-degree elbow in the wire, and of course there are better highway robbers available too.

It's funny, but the Gemini project used very-far-from-optimal Herschel technology, even for 2010 -- the Spartan construction arms were copied out cell for cell from the old prototype universal constructor. The prototype was just an alpha first draft that Paul Chapman put together in a couple of late nights, just to get a working demonstration of the idea. It was never intended for actual use; even in 2004 it would have been possible to boil the same functionality down to half as many Spartan still lifes -- but it would have taken several more late nights, and nobody ever bothered to do it.

simsim314 wrote:We still need to use some sort of Hersrch, mainly because we need all the timing to work, but as a rule of thumb I think the guys who built the whole gun collection didn't use Hersrch even once, just to show how less needed this is today.

Here's the exception that proves the rule... the second exception doesn't really count.

I think that Hersrch would get used quite a bit more if it had a Golly GUI interface, where you'd select an input object and then an output object, specify the travel time, and then let NewHersrch display candidates one at a time -- optionally only ones that don't overlap with existing obstacles in the pattern.

simsim314 wrote:The modern Hersrch should also work for all possibilities, and consider gliders. For example we have G4->R and R->G, so we can skip Hesrchels all together many times, they already not the default conduit, although they are the most versatile.

Writing that new Hersrch is going to be an interesting problem. Hersrch's author started a discussion about a Java-based Hersrch, back in 2012, but the project didn't get very far. Might be a good thread to bump now, though, especially if anyone has ideas for a new algorithm that can handle things like gliders and tandem gliders -- the current Hersrch would be swamped pretty quickly if someone tried to add all the known adjustable conduits to it.
User avatar
dvgrn
Moderator
 
Posts: 5546
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI

Re: The Hunting of the New Herschel Conduits

Postby A for awesome » June 17th, 2015, 10:22 am

Ia there any way to use the R's exhaust to restore this boat:
#C Almost-H-to-R+G0
x = 24, y = 23, rule = LifeHistory
14.2D$13.BDBD$12.BD2BDB$11.3B2D2B$9.5BD3B$8.10B$9.8B$10.7B$8.8B$6.13B
$2.2C.2B2C11B$2.C.2BCBC11B$3.C.2BC12B$3C3.10B3CB$C8.2B2.4BC2B$13.2B3C
2B$14.6B$14.7B$16.B.4B$19.4B$20.4B$21.3B$22.2B!
?
x₁=ηx
V ⃰_η=c²√(Λη)
K=(Λu²)/2
Pₐ=1−1/(∫^∞_t₀(p(t)ˡ⁽ᵗ⁾)dt)

$$x_1=\eta x$$
$$V^*_\eta=c^2\sqrt{\Lambda\eta}$$
$$K=\frac{\Lambda u^2}2$$
$$P_a=1-\frac1{\int^\infty_{t_0}p(t)^{l(t)}dt}$$

http://conwaylife.com/wiki/A_for_all

Aidan F. Pierce
User avatar
A for awesome
 
Posts: 1778
Joined: September 13th, 2014, 5:36 pm
Location: 0x-1

PreviousNext

Return to Patterns

Who is online

Users browsing this forum: No registered users and 4 guests