The Hunting of the New Herschel Conduits

[MathAndCode]

A catalyst usage guide I had made for myself at one point:
Included are some ways for common catalysts to go wrong, and replacement catalysts that don't.

Yes; that seems useful.
Also, it is possible to guess which catalyst was used by the perturbations in the active region? I'm trying to reverse-engineer a catalytic reaction in order to make the dot and possibly the domino in 5bo$2bo$b4o$o$b2o$2b2o$4bo2bo$5bobo.

[Entity Valkyrie 2]

While I was searching for conduits accepting C*, I found an oscillator partial.

Code: Select all

x = 11, y = 8, rule = B3/S23
9b2o$9bo$6b2obo$8bo$4bo$2o2bo$2o2bo2bo$5b2o!

Can it be completed to make what would be one of the smallest p5 oscillators known?

Silver's p5:

Code: Select all

x = 11, y = 15, rule = B3/S23
9b2o$9bo$7bobo$7b2o$b2ob2o$o4bo$o5bo$b2o2b2o3$5bo2bo$4bo2b2o$4b2o2b2o$
8bobo$9b2o!

[yujh]

While I was searching for conduits accepting C*, I found an oscillator partial.

Code: Select all

x = 11, y = 8, rule = B3/S23
9b2o$9bo$6b2obo$8bo$4bo$2o2bo$2o2bo2bo$5b2o!

Can it be completed to make what would be one of the smallest p5 oscillators known?

Silver's p5:

Code: Select all

x = 11, y = 15, rule = B3/S23
9b2o$9bo$7bobo$7b2o$b2ob2o$o4bo$o5bo$b2o2b2o3$5bo2bo$4bo2b2o$4b2o2b2o$
8bobo$9b2o!

Nice but posted elsewhere?
Edit:oh and why here in the hunting of new Herschel conduits?

[Entity Valkyrie 2]

I don't think I saw it anywhere else in the patterns forum.

[yujh]

I don't think I saw it anywhere else in the patterns forum.

In sandbox. What about question 2?

Code: Select all

x = 11, y = 8, rule = B3/S23
9b2o$9bo$6b2obo$8bo$4bo$2o2bo$2o2bo2bo$5b2o!                                                                                                                                                                #C [[ THEME 6 LOOP 5 ]]

Interestingly, that can be supported by part of a Silver's p5:

Code: Select all

x = 15, y = 11, rule = B3/S23
8b2o$8b2o2$9bo$2b2o4b2o$bo2bo2bobo$bo2bo2bo$5bo2bo2b2o$2bobo6bo$2o10b
3o$2o12bo!

[MathAndCode]

An example of a snake usage would be this here. I was sadly unable to fully stabilize it, and the snake gets destroyed soon after its catalysis of the reaction.

Code: Select all

x = 17, y = 28, rule = LifeHistory
5.A$5.3A$8.A$7.2A7$12.D$A.2A7.3D$2A.A9.D12$2.2A$2.2A5.C$8.3C4.2A$10.C
4.2A!

Here is another R–to–R conduit based off of that.

Code: Select all

x = 15, y = 22, rule = LifeHistory
13.A$11.3A$10.A$10.2A8$3.2D$4.2D$4.D5$2A$2A5.C$6.3C4.2A$8.C4.2A!

(I'm not sure whether or not I had already created this particular R–to–R conduit.)

Also, I think that this might be progress on trying to make a backwards Herschel.

Code: Select all

x = 12, y = 13, rule = LifeHistory
5.2A$5.2A4$.2D6.3A$3D6.A$.2D6.A$3.2D5.A$5.2D$7.D$9.2A$9.2A!
#C [[ T 18 PAUSE 1 ]]

[dvgrn]

An example of a snake usage would be this here. I was sadly unable to fully stabilize it, and the snake gets destroyed soon after its catalysis of the reaction.

Here is another R–to–R conduit based off of that...
Also, I think that this might be progress on trying to make a backwards Herschel...

I don't want to discourage new discoveries in this area, but on the other hand this thread already has forty-four pages even when people mostly only post actual working conduits. These last two seem an awful long way from working. Any time a catalyst doesn't survive say fifty ticks after its work is done, it's highly speculative to call it a catalyst at all.

So maybe a good place for posts like this would be a new "Thread For Your HIghly Speculative Herschel Conduits", to try to keep the clutter down a bit on this collection thread. The problem is that here are just so many thousands of speculative reactions that look about as good as these ones. I have to discard a few dozen of them every time I do a catgl search; if I'm not kind of picky about only investigating the (very rare) really promising results, I usually waste a whole lot of time and don't end up with anything useful.

[pcallahan]

Not too surprisingly, the first stable X-to-Ls also tended to be composite monstrosities. I seem to remember one by Dieter Leithner, but can't lay hands on it at the moment. Here's an early one assembled by Paul Callahan, 17 May 1997:

Interesting. I have absolutely no recollection of this. Maybe I was too busy planning my move back from Zürich. I wonder where the actual Herschel to LWSS reaction came from.

[MathAndCode]

The problem is that here are just so many thousands of speculative reactions that look about as good as these ones. I have to discard a few dozen of them every time I do a catgl search; if I'm not kind of picky about only investigating the (very rare) really promising results, I usually waste a whole lot of time and don't end up with anything useful.

That's a good point. Also, thank you for the advice. (I had been trying to complete a partial ∏-to-blockade predecessor conduit until I read your post.)

[Sphenocorona]

(I had been trying to complete a partial ∏-to-blockade predecessor conduit until I read your post.)

To be fair, a sufficiently edgy LOM (lumps of muck, produces blockade) output, or a new suitable conduit with a LOM input, could potentially qualify it as a valid new branch point for the Elementary Conduits Collection. Currently there's only a single composite conduit with a LOM intermediate, which uses a much more distant and accessable LOM output than any other known way to get a LOM from a conduit besides direct synthesis. There's other known potential LOM conduits, but so far none of them are compatible with known ways of either getting a signal into them or getting a signal back out.

But yeah, an "Incomplete Conduits" thread or such might be good if we want to keep both this and the Elementary Conduits thread clean (...why do we have two different threads for basically the same thing anyway?)

[MathAndCode]

To be fair, a sufficiently edgy LOM (lumps of muck, produces blockade) output, or a new suitable conduit with a LOM input, could potentially qualify it as a valid new branch point for the Elementary Conduits Collection. Currently there's only a single composite conduit with a LOM intermediate, which uses a much more distant and accessable LOM output than any other known way to get a LOM from a conduit besides direct synthesis. There's other known potential LOM conduits, but so far none of them are compatible with known ways of either getting a signal into them or getting a signal back out.

Unfortunately, I had deleted my progress, so now let's hope that I remember enough of what I did.



Edit: It's not nearly as edgy as I remembered (probably because that's not the LOM that I was trying to extract).

Code: Select all

x = 15, y = 5, rule = B3/S23
13b2o$13b2o$3o$obo$obo!

[wwei23]

Hold on. Do I see a block on boat there?

[MathAndCode]

Hold on. Do I see a block on boat there?

One of my other catalyst attempts returned a beehive on fishhook. Would you like that?

[wwei23]

One of my other catalyst attempts returned a beehive on fishhook. Would you like that?

Sounds interesting enough to me.

[MathAndCode]

Sounds interesting enough to me.

I'm trying to find it again, but I'm having difficulty remembering what I did the first time. Here's an integral sign in order to make up for it.

Code: Select all

x = 13, y = 3, rule = B3/S23
3o8b2o$obo8b2o$obo!

[bubblegum]

Sounds interesting enough to me. :P

I'm trying to find it again, but I'm having difficulty remembering what I did the first time. Here's an integral sign in order to make up for it.

Code: Select all

x = 13, y = 3, rule = B3/S23
3o8b2o$obo8b2o$obo!

I'm trying to help find it too. Not having much success but here's a transparent block:

Code: Select all

x = 15, y = 6, rule = B3/S23
13b2o$13b2o2$3o$obo$obo!

[wwei23]

Here's an integral sign in order to make up for it.

It's okay! Seems like it'll be way harder to take that B-heptomino and resynthesize that block though.

[yujh]

Sounds interesting enough to me.

I'm trying to find it again, but I'm having difficulty remembering what I did the first time. Here's an integral sign in order to make up for it.

Code: Select all

x = 13, y = 3, rule = B3/S23
3o8b2o$obo8b2o$obo!

Oops!
I found the thing was discovered accidentally!

Code: Select all

 https://www.conwaylife.com/forums/viewtopic.php?f=2&t=279&hilit=spark+glider+discoveries#p1344

(in the first viewer, where the reaction happens later)

[dvgrn]

I found the thing was discovered accidentally!

Code: Select all

 https://www.conwaylife.com/forums/viewtopic.php?f=2&t=279&hilit=spark+glider+discoveries#p1344

(in the first viewer, where the reaction happens later)

A modified pi explosion is a very old and well-known way to make an integral sign. There are three examples in Mark Niemiec's database, for example, which are probably a good bit older than Extrementhusiast's accidental discovery.

In the future, can we please not allow discussions like this to spiral out of control on the Herschel conduit collection thread?

[MathAndCode]

I'm trying to help find it too. Not having much success but here's a transparent block:

Code: Select all

x = 15, y = 6, rule = B3/S23
13b2o$13b2o2$3o$obo$obo!

I should have said this earlier, but the beehive on fishhook wasn't from searching ∏-accepting conduits but instead from something like this.

Code: Select all

x = 12, y = 17, rule = B3/S23
11bo$2o7b3o$bo6bo$bobo4b2o$2b2o8$6b2o$6bobo$5bo2bo$5bobo$6bo!

A modified pi explosion is a very old and well-known way to make an integral sign.

I figured that, but wwei23 seems to be amazed by even mildly uncommon objects. (I'm still mildly impressed whenever I see a long boat.)

[EvinZL]

Can a catalyst replace the yellow block?

Code: Select all

x = 35, y = 42, rule = LifeHistory
23.B$22.3D$22.BDB$21.2B3D$21.5B$21.6B$21.6B$21.5B$20.6B$21.6B$20.7B$20.
6B$20.6B$20.6B$19.8B$20.8B$19.9B$19.9B$19.10B$9.A9.5B2A3B$9.3A7.5B2A4B
3.2A$12.A6.11B3.A$11.2A3.B2.4BD7BA.A$11.8B2.B3D4B2.2A$13.8BD2B2D2B$13.
13B$12.14B$10.17B$8.19B$8.2BC15B$7.3BCBC4B.8B$8.2B3C4B2.B4.2E$7.5BC4B
7.2E$6.10B$5.4B$4.4B$3.4B$2.4B$.4B$4B$3B$.B!

A fishhook almost works, but a backwards spark destroys it.

[dvgrn]

Can a catalyst replace the yellow block?...
A fishhook almost works, but a backwards spark destroys it.

I hope Bellman will be able to turn something up -- that would make a nice Lx104. I was worried that this might turn out to be an L112 variant or something, but an "Lx104" search doesn't turn up any matches, except for your "Extended Life Extensions" post today!

At worst, the block can be replaced "easily" by a glider-to-block converter. But that's just a way of turning a nice compact almost-conduit into a huge slow useless technically-working one:

Code: Select all

x = 72, y = 85, rule = LifeHistory
42.B$41.3D$41.BDB$40.2B3D$40.5B$40.6B$40.6B$40.5B$39.6B$35.2A3.6B$34.
A.A2.7B$34.A4.6B$21.A11.2A4.6B$19.3A17.6B$18.A19.8B$18.2A19.8B$16.4B
18.9B$15.3B20.9B$B14.4B19.10B$2B12.5B9.A9.5B2A3B$3B11.6B8.3A7.5B2A4B
3.2A$10B2.8B11.A6.11B3.A$.5BC14B9.2A3.B2.4BD7BA.A$2.2B3C13B10.8B2.B3D
4B2.2A$.3BCBC4B.7B13.8BD2B2D2B$2.2BC15B12.13B$2.18B11.14B$4.17B8.17B$
6.15B.2B2.20B$7.22BD15B$7.8B.13BDBD4B.8B$5.8B3.2B2A9B3D4B2.B4.2AB$5.
2A3.B5.2B2A11BD4B6.B2A2B$6.A10.18B7.6B6.2A$3.3A4.A7.B.3B.4B9.2A.7BAB
5.A$3.A6.3A10.4B9.A.A7BABA2.BA.A$13.A8.4B10.A3.6B2A2B.B2A$12.2A7.4B
10.2A3.12B$12.5B3.4B16.12B$14.3B2.4B17.12B$4.2A7.9B16.2AB.9B$4.A8.8B
16.A.AB.10B$.2A.A.B3.10B17.A4.6B.4B$.A2.3AB.2B2A7B16.2A5.5B2.4B$2.2A
2.BA3B2A7B24.3B4.4B$4.4A12B24.2B6.4B$4.A.2B3.7B.B2A21.4B6.4B10.2A$5.
3AB2.7B.BA.A21.B2AB6.4B9.A$8.A4.4B5.A22.2A8.4B10.A$3.5A5.4B5.2A25.2A
5.4B5.5A$3.A10.4B32.A5.4B4.A$5.A9.4B31.A.AB.7B2.B3A$4.2A10.4B31.2AB.
7B3.2B.A$17.4B32.12B4A$18.4B31.7B2A3BAB2.2A$19.4B30.7B2A2B.B3A2.A$20.
4B29.10B3.B.A.2A$21.4B27.8B8.A$22.4B25.9B7.2A$23.4B23.4B2.3B$24.4B21.
4B3.5B$25.4B19.4B7.2A$26.4B10.A6.4B8.A$27.4B7.3A5.4B10.3A$28.4B5.A7.
4B13.A$29.4B4.2A5.4B$30.9B4.4B$31.6B5.4B$31.8B2.4B$29.15B$29.14B$29.
13B$27.2AB.10B$26.A.AB3.B2A3B$26.A6.B2A3B$25.2A6.4B$34.3B$35.2B.BA$
34.B2ABA.A$33.BABABA.A$31.A2.A.A.A.A.2A$31.4A.2A2.A2.A$35.A4.2A$33.A.
A$33.2A!

[MathAndCode]

Can a catalyst replace the yellow block?

Here's an I→R conduit with the same problem.

Code: Select all

x = 18, y = 19, rule = LifeHistory
6.D$5.3D$5.D2$2C$2C3$8.C$7.2C7.2A$5.3C8.2A$5.C6$7.2A$7.2A!

(There's also a wing at generation 36, but I chose to focus on isolating the R.)
In case it's useful, here is what I originally found.

Code: Select all

x = 15, y = 22, rule = B3/S23
2o$2o15$13b2o$10b2ob2o$10b2o2$12b2o$12b2o!

Getting a century conduit seemed much less likely, although now that I think back to my attempts at a C→I converter, it seems more plausible to me than it did ten minutes ago.



Edit: I tried to make a clean C→C without success. Also, here's the partial I→W.

Code: Select all

x = 13, y = 26, rule = LifeHistory
A$3A$3.A$2.2A8$2E$2E3$8.C$7.2C2.2D$5.3C2.D.D$5.C3.D2.D$10.2D5$7.2A$7.2A!

[EvinZL]

Using AforAwesome(?)'s observation that Fx119 contains a R to B+H, I made this B254

Code: Select all

x = 94, y = 74, rule = LifeHistory
48.4B$47.4B$33.A12.4B$24.A7.A.A10.4B$24.3A4.2BAB9.4B$27.A4.2B9.4B$26.
2A5.4B5.4B$14.B.7B3.3B3.6B3.4B$8.B.13B5.3B2.5B2.4B$3.21BD10BD7B$.D21B
D10B2D6B$B3D18B2D11B2D4B$.DBD15B2A2B2D16B5.2A$3.D15B2A3BD16B5.A$5.2B
2.2B3.27B2.BA.A$15.10B2.B3.B.9B.B2A$18.6B9.11B$20.3B10.11B5.A$22.B8.
2AB.9B3.3A$22.2A6.A.AB2.7B3.A$23.A6.A5.8B2.2A$20.3A6.2A4.13B$20.A14.
11B$34.13B$35.12B4.2A$35.12B4.A$34.13B.BA.A$33.9B3D2B.B2A$33.10BD5B$
34.7B3D4B$33.16B$31.18B$29.19B$29.2BC17B$28.3BCBC4B.11B$29.2B3C4B2.3B
D7B$28.5BC4B2.2B3D2B.4B$27.10B4.D2B2DB2.4B9.3B$26.4B10.7B3.4B7.5B$25.
4B12.6B4.4B6.6B$24.4B13.7B4.4B5.2BA4B$23.4B14.7B5.4B4.BABA4B$22.4B14.
9B5.4B2.2BABA6B$21.4B15.2BD7B5.4B2.2BA8B$20.4B15.2B3D7B5.4B2.10B$19.
4B17.3BD2B2.4B5.4B.10B$18.4B18.7B2.4B5.15B$17.4B17.11B.4B5.14B$16.4B
16.B.2B2A13B5.13B$15.4B16.2A3B2A9B.4B5.12B$17.B17.2A14B2.4B5.11B$36.
14B4.4B5.12B$34.5B2.B2.2B.3B5.4B3.14B2.B$34.2A4.3B5.2A6.4B.19B$35.A3.
B2AB5.A8.24B$32.3A5.2A7.3A6.14B2A7B$32.A18.A4.16B2A8B$56.26B$54.2AB.
24B$53.A.AB4.21B$53.A7.22B$52.2A9.20B$64.19B2A$64.14B4.A2BA2.2A$65.
13B4.ABAB3.A2.A$63.2AB.B.9B5.A4B.A.A.A$62.A.AB4.8B7.B2A.A2.A$62.A10.
3B9.BAB.A$61.2A20.A4.A$83.5A2$85.A$84.A.A$85.A!

Don't ask about the repeat time

[bubblegum]

Using AforAwesome(?)'s observation that Fx119 contains a R to B+H, I made this B254

Code: Select all

x = 94, y = 74, rule = LifeHistory
48.4B$47.4B$33.A12.4B$24.A7.A.A10.4B$24.3A4.2BAB9.4B$27.A4.2B9.4B$26.
2A5.4B5.4B$14.B.7B3.3B3.6B3.4B$8.B.13B5.3B2.5B2.4B$3.21BD10BD7B$.D21B
D10B2D6B$B3D18B2D11B2D4B$.DBD15B2A2B2D16B5.2A$3.D15B2A3BD16B5.A$5.2B
2.2B3.27B2.BA.A$15.10B2.B3.B.9B.B2A$18.6B9.11B$20.3B10.11B5.A$22.B8.
2AB.9B3.3A$22.2A6.A.AB2.7B3.A$23.A6.A5.8B2.2A$20.3A6.2A4.13B$20.A14.
11B$34.13B$35.12B4.2A$35.12B4.A$34.13B.BA.A$33.9B3D2B.B2A$33.10BD5B$
34.7B3D4B$33.16B$31.18B$29.19B$29.2BC17B$28.3BCBC4B.11B$29.2B3C4B2.3B
D7B$28.5BC4B2.2B3D2B.4B$27.10B4.D2B2DB2.4B9.3B$26.4B10.7B3.4B7.5B$25.
4B12.6B4.4B6.6B$24.4B13.7B4.4B5.2BA4B$23.4B14.7B5.4B4.BABA4B$22.4B14.
9B5.4B2.2BABA6B$21.4B15.2BD7B5.4B2.2BA8B$20.4B15.2B3D7B5.4B2.10B$19.
4B17.3BD2B2.4B5.4B.10B$18.4B18.7B2.4B5.15B$17.4B17.11B.4B5.14B$16.4B
16.B.2B2A13B5.13B$15.4B16.2A3B2A9B.4B5.12B$17.B17.2A14B2.4B5.11B$36.
14B4.4B5.12B$34.5B2.B2.2B.3B5.4B3.14B2.B$34.2A4.3B5.2A6.4B.19B$35.A3.
B2AB5.A8.24B$32.3A5.2A7.3A6.14B2A7B$32.A18.A4.16B2A8B$56.26B$54.2AB.
24B$53.A.AB4.21B$53.A7.22B$52.2A9.20B$64.19B2A$64.14B4.A2BA2.2A$65.
13B4.ABAB3.A2.A$63.2AB.B.9B5.A4B.A.A.A$62.A.AB4.8B7.B2A.A2.A$62.A10.
3B9.BAB.A$61.2A20.A4.A$83.5A2$85.A$84.A.A$85.A!

Don't ask about the repeat time :?

307.