The Hunting of the New Herschel Conduits

[Kazyan]

I've been looking into the century repeatedly. It's not very cooperative.

Here's an unusual R-to-G (though I get the feeling it's a duplicate of something):

Code: Select all

x = 14, y = 14, rule = B3/S23
9bo$7b3o$6bo$6b2o5$b2o$o2bo$o2bo$b2o9b2o$11b2o$12bo!

[Scorbie]

Here's an unusual R-to-G

Well, congrats for finding the second useful pattern that uses a transparent pond!
(After the honeybit HWSS eater)
Edit: This one:

Code: Select all

x = 21, y = 9, rule = B3/S23
17b2o$b2o12bo4bo$o2bo10bo$o2bo10bo5bo$b2o11b6o3$2b2o$2b2o!

[Kazyan]

Well, congrats for finding the second useful pattern that uses a transparent pond! (After the honeybit HWSS eater)

The only other use for a pond is converting it into a ship, isn't it? It's nigh-useless for a spartan still life; even ships are more handy.

[simsim314]

I've been looking into the century repeatedly. It's not very cooperative.

I would suggest to try the new ptbsearch on Century, while CatForce is not searching in depth. CatForce approach is not working so well for exploding interactions, and C tends to explode. While ptbsearch is depth oriented, so it works on non exploding areas first. I would also try to search with more catalysts and last gen to be pretty early.

Here's an unusual R-to-G (though I get the feeling it's a duplicate of something):

In dvgrn's latest collection there is no R->G at all. And be assured that transparent pond is never duplicate, as it's certainly unique interaction never seen before. I'm adding pond to every CatForce search and it never works at all, I mean I never get anything with a pond out of pretty large searches, i.e. transparent pond is very rare.

Congrats on the finding!

[Kazyan]

Well then! Thanks.

Here's a Spartan version of the R-to-R in the collection, which otherwise requires that long-hook SL sometimes used with a tub-with-tail. Also, one of the eaters just suppresses a B, which might be extractable as another signal; I've left that eater out here,

Code: Select all

x = 18, y = 16, rule = LifeHistory
16.2D$15.2D$16.D$2.D$.2D$2D11.2A$.D10.2A$2.D10.A3$4.2A$5.A$2.3A$2.A$
12.2A$12.2A!

Another R-to-G.

Code: Select all

x = 15, y = 17, rule = B3/S23
2o$bo$bobo$2b2o3$13b2o$12b2o$5b2o6bo$5b2o3$3b2o$2bobo$2bo$b2o9b2o$12b
2o!

This might be a non-Spartan R-to-R, but there's that errant block. There's a small chance Bellman could clean it up by interacting with it from the cell I've placed a dot in...

Code: Select all

x = 18, y = 15, rule = B3/S23
o$3o$3bo$2b2o$17bo4$5b2o$5b2o9b2o$15b2o$16bo2$3b2o$3b2o!

[dvgrn]

Here's an unusual R-to-G (though I get the feeling it's a duplicate of something):

In dvgrn's latest collection there is no R->G at all. And be assured that transparent pond is never duplicate, as it's certainly unique interaction never seen before.

Here it is making a new pi-to-G17:

Code: Select all

x = 40, y = 27, rule = LifeHistory
5.A$5.3A22.A$8.A21.3A$7.2A24.A$7.7B18.2A$7.8B$6.10B3.B$5.17B$6.17B$7.
16B14.2A$6.2B3C12B13.A2.A$6.4BC12B2.3B8.A2.A$4.4B3C15B2DB8.2A$4.23B2D
$3.24BD3B$3.29B$3.30B$3.30B$5.28B$5.17B3.7B$5.17B3.5B$4.4B.7B11.4B$3.
4B3.5B13.B2AB$2.4B5.3B15.A.A$.4B6.3B13.A.A.3A$4B8.B14.2A5.A$3B30.2A!

Not Spartan -- the eater-tie-eater we could handle with a slow-salvo construction in an emergency, but the only known pi source that works here requires another minor eater weld. Wouldn't you know, G17 is just one cell too narrow to work with Snarks or semi-Snarks, or to allow an extra passthrough glider (to be reflected later to support some future conduit) for a syringe.

I went through all the R sources and didn't find anything else that could connect to this -- except one of the B-to-Rs, BLx19R, and then none of the X-to-Bs seemed to connect to that. What did I miss?

Anyway, it's nice to fill the R-to-G spot in the table, especially with such a beautifully simple new reaction.

Speaking of beautifully simple, the new H-to-G4 enables a couple of new infinite families of Spartan SE H-to-Gs, on new output lanes that were previously unreachable:

Code: Select all

x = 127, y = 60, rule = LifeHistory
52.A69.A$50.3A67.3A$33.A15.A53.A15.A$33.3A13.2A4.2A46.3A13.2A4.2A$36.
A9.5B4.A50.A9.5B4.A$35.2A3.B4.4B3.BA.A49.2A3.B4.4B3.BA.A$35.8B.7B.B2A
50.8B.7B.B2A$37.16B54.16B$37.16B54.16B$36.17B53.17B$34.18B52.18B$32.
19B51.19B$32.2BA15B52.2BA15B$31.3BABA13B51.3BABA13B$32.2B3A13B52.2B3A
13B$31.5BA12B52.5BA12B$30.10B3.6B51.10B3.6B$29.8B6.7B49.8B6.7B$28.8B
9.5B48.8B9.5B$27.8B11.4B47.8B11.4B$26.8B11.B2A2B46.8B11.B2A2B$25.8B
13.2A47.8B13.2A$24.8B62.8B$23.8B62.8B$22.8B62.8B$21.8B62.8B$20.8B62.
8B$19.8B62.8B$18.8B62.8B$11.A5.8B62.8B$9.3A4.8B62.8B$8.A6.8B62.8B$8.
2A4.8B62.8B$13.8B62.8B$6.2B4.8B56.2A4.8B$5.2C2B2.8B56.B2AB2.8B$4.B2C
2B.8B57.4B.8B$4.13B59.2B.8B$3.13B60.10B$.3B2C9B59.11B$.3B2C10B55.13B$
2A14B.2B47.B3.13B$2A16B2A45.2AB.9BC4B$.15B.B2A45.2A10BCBC4B$3.13B2.B
47.B.9BC2BC4B$3.14B52.9B2C6B$5.12B53.11B2.4B$5.2B2A9B53.8B5.4B$6.B2A
5B.4B52.10B4.4B$6.9B.4B52.3B4.2A5.4B$6.9B2.4B50.4B4.A7.4B$4.2A.8B3.4B
48.4B6.3A5.4B$4.A2.6B6.4B48.B2AB7.A6.4B$2.A.A3.5B7.4B48.2A16.4B$2.2A
3.6B8.4B66.4B$8.4B10.3B67.3B$9.2B$10.2B$9.B2AB$10.2A!
#C [[ AUTOSTART PAUSE 2 ZOOM 3.5 LOOP 400 ]]

Even more impressive is the new family of ambidextrous Spartan Herschel transceivers (!!) Here's F262+4N:

Code: Select all

x = 60, y = 46, rule = LifeHistory
55.A$53.3A$36.A15.A$36.3A13.2A4.2A$39.A9.5B4.A$38.2A3.B4.4B3.BA.A$38.
8B.7B.B2A$40.16B$40.16B$39.17B$24.2B11.18B$24.3B3.2B2.20B$24.13BC15B$
24.13BCBC13B$24.13B3C13B$24.15BC12B$24.19B3.6B$24.7B.8B6.7B$24.B.B4.
8B9.5B$30.8B11.4B$29.8B11.B2A2B$28.8B13.2A$27.8B$6.B2AB16.8B$7.BA7.2A
7.8B$8.ABA4.B2AB5.8B3.2A$9.2A.B3.3B4.8B5.A$10.5B2.B4.8B5.A$11.9B.8B6.
2A$8.3B.16B8.B$8.19B9.3B$.B3.21B9.6B$2AB.22B.B3.B2.10B$2A43B3.2B2.2B$
.B.31BD3B2A15BD$4.7B2A21B2D2B2A15BDBD$5.B.4B2A9B2D11B2D18B3DB$9.14B2D
10BD21BD$8.A3.11BD10BD21B$6.3A5.12B2.3B5.13B.B$5.A8.13B3.3B3.7B.B$5.
2A9.5B.4B5.2A$16.B2A2B4.2B4.A$17.2A5.BA2B4.3A$24.A.A7.A$25.A!
#C [[ AUTOSTART PAUSE 2 ZOOM 7 STOP 262 ]]

The extra glider output should come in very handy. There's even a G-to-H variant that recovers in 542 ticks, and the G-to-G recovers in 546 ticks and has only twenty still lifes, just like the Silver reflector. Clockwise from upper left: Silver reflector, new G-to-H, new G-to-G:

Code: Select all

x = 172, y = 170, rule = LifeHistory
10.2A$10.2A$31.A$29.3A11.A$28.A14.3A$28.2A16.A14.A$45.2A12.3A$58.A$
58.2A3$57.2A$38.2A17.2A$38.2A2$2A$2A3$41.2A$21.2A19.A$21.A.A15.3A$23.
A15.A$23.2A19.2A$45.A$42.3A$42.A6$35.2A$36.A$36.A.A15.2A$37.2A15.2A2$
167.A$165.3A$148.A15.A$129.2A17.3A13.2A4.2A$130.A20.A18.A$117.A11.A
20.2A16.A.A$74.A42.3A9.2A37.2A$75.A44.A$73.3A31.2A10.2A$82.A9.A15.A$
55.2A25.3A5.3A15.A.A$55.A.A27.A3.A19.2A$57.A26.2A3.2A41.2A$57.2A73.2A
7$98.2A61.2A$98.2A23.2A36.2A$45.2A39.2A35.A$45.2A38.A2.A35.3A$33.2A
45.2A4.2A38.A$32.A.A44.A.A$32.A46.A$31.2A45.2A$88.2A$56.2A30.A$35.2A
19.A32.3A$36.A17.A.A34.A$36.A.A15.2A$37.2A4.A$42.A.A$42.A.A$43.A10.2A
$54.A.A58.2A$56.A58.2A20.A$56.2A16.A60.3A$41.2A31.3A57.A$42.A34.A56.
2A$39.3A34.2A$39.A51.2A72.A$91.A72.A.A$89.A.A7.2A63.A.A$78.A10.2A8.2A
64.A$77.A.A57.2A$77.A.A30.2A25.2A$72.2A4.A31.2A$71.A.A15.2A$71.A17.A.
A15.A$70.2A19.A13.3A$91.2A11.A$104.2A24.2A$116.2A12.A16.2A$116.2A6.A
6.3A14.A$123.A.A7.A11.3A$124.A20.A$80.2A$80.2A6$47.2A$47.2A8.A$55.3A
35.A$54.A36.3A$54.2A34.A$90.2A3$54.A$53.A.A$40.2A10.A2.A$40.2A11.2A$
95.2A$95.A$93.A.A$93.2A3$51.2A26.2A$51.2A25.A.A$78.A$77.2A3$95.2A$95.
A.A$97.A$88.2A7.2A$88.2A15$98.2A$98.A$96.A.A$96.2A10$80.2A$80.2A2$98.
2A$98.A.A$100.A$100.2A$95.2A$95.A$96.3A$98.A!
#C [[ AUTOSTART ZOOM 2.5 STEP 8 ]]

[simsim314]

Well I was thinking about using G4 as tandem signal on its own. With the latest discovery, it has recovery time of 118:

Code: Select all

x = 137, y = 151, rule = B3/S23
21$109bo$108bo$108b3o14$89bobo$89b2o$90bo12$79bo$79bobo$79b2o$118bo$
52b2o62b3o$53bo7b2o36bo15bo$53bobo5b2o20bo15b3o13b2o4b2o$54b2o25b3o18b
o18bo$80bo20b2o16bobo$80b2o37b2o4$45b2o$45b2o$61bo21b2o$56b2ob2o22b2o$
56b2o2b2o2$53bo$51b3o$50bo$50b2o24b2o$62b2o12bo35b2o$62b2o6bo6b3o32b2o
$69bobo7bo$70bo37$98bo$32b2o62b3o$33bo7b2o36bo15bo$33bobo5b2o20bo15b3o
13b2o4b2o$34b2o25b3o18bo18bo$60bo20b2o16bobo$60b2o37b2o4$25b2o$25b2o$
63b2o$36b2o25b2o$36b2o2$33bo$31b3o$30bo$30b2o24b2o$42b2o12bo35b2o$42b
2o6bo6b3o32b2o$49bobo7bo$50bo!

Unfortunately it doesn't combines well with Chris H->G4 becuase of recovery 325. And if we want to use G4 signal at this usual recovery, we need all circuits to work around 120.

Which brings me to the question: what is the fastest recovery H splitter?

EDIT Thinking of it again, maybe we can have a design that can work with only one signal reflector, and signal->glider. Ideally we could have non destructive glider edge shooter and 10hd arm, but we can also use something like opposite glider collision mechanism. The signal will travel into two distant locations, shoot a glider in opposite directions and come back into smaller data-stream.

EDIT Looking into the old Serizawa thread. This trick could be a base for 0-hd up to 3hd arm construction:

Code: Select all

x = 69, y = 43, rule = B3/S23
bo$2bo$3o6$14b2o$15bo$15bobo$16b2o3$32b2o$21b3o8bo$21bo8bobo$20b3o7b2o
5$44bo$34b2o6b3o$34bo6bo$32bobo6b2o$32b2o3$19b2o$18bobo$18bo25b2o$17b
2o25b2o$67b2o$67b2o2$21b2o$20bobo$20bo4b2o$19b2o5bo$23b3o5bo$23bo6bobo
$31bo!
#C [[ THUMBNAIL ]]

The only thing missing is arm recipes for those. But I guess there is no real limitation about it, only search that should be conducted.

[dvgrn]

Which brings me to the question: what is the fastest recovery H splitter?

The fastest Spartan H splitter, you mean? Otherwise the syringe's recovery time of 78 ticks is the obvious lower limit.

I keep trying to build Spartan structures that can split a signal quickly, but they all have some fatal flaw that drops the recovery time down to around 250-300 ticks. Seems as if the best should be better than this now, but the fastest splitter circuit I can think of is still the one based on ancient technology that Guam pointed out three years ago -- 204 ticks. There are a couple of very closely spaced eaters at the beginning, and out of superstition I've always avoided using that variant L156b in self-constructing circuitry -- but I don't think that it presents any insurmountable construction difficulties, it's just a little trickier and more expensive.

It does seem to me that we might have dug up enough new Herschel conduits recently that throw a spare glider forward, that we might be able to string together a version of this with a sub-200 repeat rate. The old and new variants of the receivers can be made to work in the 110-130-tick range... so is there a good way to string together new fast-recovery conduits to emit 90 degree intersecting gliders within 130 ticks of each other?

The most promising composite mechanisms still seem to work out to just above 200 ticks (even if they were finished):

Code: Select all

x = 190, y = 108, rule = LifeHistory
AB$B2A$2A2B$.4B$2.4B$3.4B$4.4B$5.4B$6.4B$7.4B$8.4B$9.4B$10.4B$11.4B$
12.4B$13.4B$14.4B$15.4B$16.4B$17.4B$18.4B$19.4B$20.4B$21.4B$22.4B$23.
4B$24.4B$25.4B$26.4B$27.4B$28.4B$29.4B81.BA$30.4B79.ABAB$31.4B79.2A2B
$32.4B79.4B$33.4B79.4B$34.4B79.4B$35.4B79.4B$36.4B79.4B$37.4B79.4B$
38.4B79.4B$39.4B79.4B$40.4B79.4B$41.4B79.4B$42.4B79.4B$43.4B79.4B$44.
4B79.4B$45.4B79.4B$46.4B79.4B$47.4B79.4B$48.4B79.4B39.3B$49.4B79.4B
38.5B$50.4B79.4B36.6B$51.4B79.4B10.A24.8B$52.3BA4.A9.A64.4B7.3A24.8B$
53.ABAB3.3A5.3A65.4B5.A21.B4.9B$54.2A2B5.A3.A69.4B4.2A19.3B3.11B$55.
4B3.2A.B.2A61.2A6.9B17.6B2.13B$56.4B2.7B62.A7.6B13.4B2.7B.13B$57.4B3.
BDB23.4B4.A32.A.2A5.6B3.B2.2B2.28B$58.4B.BDBDB21.5B3.A.A32.A2.A4.45B$
59.5BDBDB19.7B3.A.A33.2AB3.46B$60.5BD3B6.2B3.2B2.10B4.A35.14B2A13B2A
21B$61.26BD6B41.13B2A13B2A22BD$62.25BDBD4B42.51B2D$62.14B2A9B3D4B42.
17B.B5.28B2D$60.16B2A11BD4B43.15B10.B2.22B2D$60.4B2A27B44.15B12.24B$
59.4BA2BA7B.2B.3B.4B52.13B12.24B$58.2AB2.B2A2B.B.4B6.4B55.13B9.25B$
57.A.AB2.5B4.4B4.2B.B55.8B4.2AB7.4B.20B$57.A6.2B7.4B2.4B56.6B5.B2A2B
5.4B2.18B$56.2A6.4B6.8B57.5B6.5B4.4B4.18B.B$66.2A7.6B58.B.B7.7B2.4B6.
18B2A$66.A9.4B60.3B4.3B2A9B7.18B2A$67.3A5.6B58.B2AB4.3B2A10B6.3B3.13B
$69.A4.8B58.2A4.2A14B.2B4.B2.3B.11B$72.10B64.2A16B2A6.2A2.9B$71.2B2D
7B65.15B.B2A7.A3.7B$71.2BDBD6B67.13B2.B5.3A5.4B$71.3BD6B68.14B7.A8.5B
$66.3B2.6B74.11B20.2A$65.12B74.2B2A7B20.A$63.14B75.B2A8B20.3A$63.17B
72.12B21.A$63.18B71.13B$63.18B69.2A.8B.4B$61.A19B69.A2.6B4.4B$60.ABAB
.14B.B2A65.A.A3.5B5.2B2A$60.A2BA2.10B4.BA.A64.2A3.6B6.BABA$58.3B2A3.
8B9.A70.4B8.AB$58.3B5.7B10.2A70.2B$57.4B4.11B80.2B$56.3AB5.12B78.B2AB
$56.2BA6.12B79.2A$57.A8.11B$64.4B.4B3AB2.2A$64.2A4.4BA2B2.2A$65.A4.2B
3A2B$62.3A6.6B$62.A8.7B$73.B.4B$76.4B$77.3B$79.B2A$79.BA.A$82.A$82.2A!

What other tricks could be tried here?

[praosylen]

Conduit that may or may not be useful:

Code: Select all

x = 32, y = 22, rule = LifeHistory
12.2C6.2C4.2D$10.2B2CB4.B2CB2.2B2D$.2C7.4B6.2B2.2BDB$2.C7.6B2.3B2.4B$
.C8.16B$.2C5.B.15B$3.CB.20B$.3C22B$C2.B.10B3C8BD$2C3.9BC3BC6B3D$4.10B
2C2BC5B2D2BD$4.25B$2.2CB.23B$.C.CB2.21B$.C6.9B.6B.3B2.2C$2C7.7B3.3B5.
BC2.C$10.5B12.B3C$11.4B10.C$25.4C$28.C$27.C$27.2C!

[chris_c]

A dependent H->B that can be used to make an F256 Herschel conduit and a rather slow H->G9:

Code: Select all

x = 96, y = 38, rule = B3/S23
28b2o56bo$9b2o18bo35b2o19b3o$10bo17bo37bo22bo$10bobo15b2o36bobo19b2o$
11b2o2b2o50b2o2b2o$15b2o54b2o11b2o5b2o$84b2o4bobo$90b2o2$31b2o$31b2o4$
92b2o$92bo$93b3o$95bo$b2o54b2o$3ob2o50b3ob2o$b2ob3ob2o47b2ob3ob2o$3ob
2o2b2o46b3ob2o2b2o$2o54b2o6$6b2o$5bobo21b2o54b2o$5bo23bobo53bobo$4b2o
25bo55bo$31b2o54b2o4$22b2o54b2o$22b2o54b2o!

[simeks]

A H-to-G6. It seems simple but I can't find it anywhere so maybe it's new?

If preceded by F171 it forms a different H-to-G6.

By replacing an eater with a block it becomes a 98 tick H-to-B. The B is in a difficult position, but a known B-to-2G luckily needs a block in exactly that position, and together it becomes a H->G0 with an extra glider to spare:

(Edited typo: B-to-2G not B-to-G2)

Code: Select all

x = 221, y = 181, rule = LifeHistory
42.D$41.2DB$41.DBDB$42.4B$43.4B$44.4B$45.4B$46.4B$47.4B$48.4B$49.4B$
50.4B$51.4B$52.4B$53.4B$54.4B$55.4B$56.4B$57.4B$58.4B$59.4B$60.4B$61.
4B$62.4B$63.4B$64.4B$65.4B$66.4B$67.4B$68.4B$69.4B$70.4B$71.4B$72.4B$
73.4B$74.4B$75.4B$76.4B$77.4B$78.4B$79.4B$80.4B$81.4B$82.4B$83.4B$84.
4B$85.4B96.2C$86.4B94.B2C2B6.2D$87.4B94.4B5.DBD$88.4B91.6B4.3BD$89.4B
89.6B4.4B$90.4B40.C11.C34.6B.2B.4B$91.4B39.3C7.3C34.13B$92.4B25.C15.C
5.C36.13B$93.4B24.3C12.2C5.2C35.12B4.2C$94.4B26.C11.9B35.11B4.C.C$95.
4B7.A16.2C3.2B.12B37.10B5.BC$96.4B6.3A14.4B.16B36.9B7.2B$97.4B8.A15.
21B35.9B5.4B$98.4B6.2A14.24B34.18B$99.4B5.6B10.26B34.16B$100.4B6.5B9.
27B34.16B$101.4B4.7B7.27B35.16B$102.4B2.8B6.26B37.16B$103.14B2.28B37.
16B$104.22BD13B3.6B35.15B$100.2A3.21BDBD4B.6B7.2C.C34.13B$101.A4.20B
3D4B2.B.5B5.2CB3C32.13B$41.C11.C47.A.AB.22BD4B7.2C6.B4.C30.12B2C3.A
11.A$41.3C7.3C48.2A28B8.C6.2C.3C32.11B2C3.3A7.3A$28.C15.C5.C52.17B.4B
16.3C4.C.C35.11B7.A5.A$28.3C12.2C5.2C53.19B19.C4.C.C37.8B7.2A5.2A$31.
C11.9B44.A10.16B26.C39.3BD3B7.9B$30.2C3.2B.12B46.3A8.15B67.2B3D3B.12B
$30.4B.16B48.A7.14B66.3BD2B2D16B$32.21B45.2A3.B3.13B66.2C25B$31.24B
43.8B.12B67.2C27B$31.26B43.18B69.2B.27B$31.27B42.17B73.28B$30.27B42.
17B75.26B$29.26B42.18B76.24B$28.26B41.19B77.23B$27.6BE13B3.6B39.2BE
15B78.2BE13B3.6B$26.7BEBE4B.6B7.2C.C36.3BEBE14B76.3BEBE4B.6B7.2A.A$
25.D3B2.2B3E4B2.B.5B5.2CB3C35.2B3E5B.7B78.2B3E4B2.B.5B5.2AB3A$24.D3B
2.5BE4B7.2C6.B4.C33.5BE4B4.5B77.5BE4B7.2A6.B4.A$24.3D2.10B8.C6.2C.3C
33.10B3.5B78.10B8.A6.2A.3A$28.4B16.3C4.C.C34.4B2.4B4.2A80.4B16.3A4.A.
A$27.4B19.C4.C.C33.4B2.4B6.A79.4B19.A4.A.A$26.4B26.C33.4B2.4B4.3A79.
4B26.A$25.4B60.4B2.4B5.A80.4B$24.4B60.4B2.4B86.4B$23.4B60.4B2.4B86.4B
$22.4B60.4B2.4B86.4B$21.4B60.4B2.4B86.4B$20.4B60.4B2.4B86.4B$19.4B60.
4B2.DBDB86.4B$18.4B60.4B3.2DB86.4B$17.4B60.4B5.D86.4B$16.4B60.4B92.4B
$15.4B60.4B92.4B$14.4B60.4B92.4B$13.4B60.4B92.D3B$12.4B60.4B92.D3B$
11.4B60.4B92.B3D$10.4B60.4B92.4B$9.4B60.4B92.4B$8.D3B60.4B92.4B$7.D3B
60.4B92.4B$7.3D60.4B92.4B$69.4B92.4B$68.4B92.4B$67.4B92.4B$66.4B92.4B
$65.4B92.4B$64.4B92.4B$63.4B92.4B$62.4B92.4B$61.4B92.4B$60.4B92.4B$
59.4B92.4B$58.4B92.4B$57.4B92.4B$56.4B92.4B$55.4B92.4B$54.4B92.4B$53.
4B92.4B$52.4B92.4B$51.4B92.4B$50.4B92.4B$49.4B92.4B$48.4B92.4B$47.4B
92.4B$46.4B92.D3B$45.4B92.D3B$44.4B93.3D$43.4B$42.4B$41.4B$40.4B$39.
4B$38.4B$37.4B$36.4B$35.4B$34.4B$33.4B$32.4B$31.4B$30.4B$29.4B$28.4B$
27.4B$26.4B$25.4B$24.4B$23.4B$22.4B$21.4B$20.4B$19.4B$18.4B$17.4B$16.
4B$15.4B$14.4B$13.4B$12.4B$11.4B$10.4B$9.4B$8.4B$7.4B$6.4B$5.4B$4.4B$
3.4B$2.4B$.D3B$D3B$3D!

[Freywa]

and together it becomes a H->G0 with an extra glider to spare:

So a H-to-G0; add a divide-by-2/semi-snark (G0-to-G) and a syringe and you get yet another Herschel transceiver, though not one of the more useful ones.

Anyway how many H-to-G0s have we really seen? Is this the first one?

[Sokwe]

A H-to-G6.

The eater 2 can be replaced with two blocks and an eater, which saves one column on the right:

Code: Select all

x = 52, y = 42, rule = LifeHistory
34.A11.A$34.3A7.3A$21.A15.A5.A6.A$21.3A12.2A5.2A3.3A$24.A11.9B2.A$23.
2A3.2B.12B3.B2A$23.4B.16B.4B$25.23B$24.24B$24.24B.2B$24.26B2A$23.25B.
B2A$22.26B2.B$21.21B.3B$20.6BE13B5.2B$19.7BEBE4B.6B4.B2AB$18.D3B2.2B
3E4B2.B.5B3.2A$17.D3B2.5BE4B7.2A$17.3D2.10B8.A$21.4B16.3A$20.4B19.A$
19.4B$18.4B$17.4B$16.4B$15.4B$14.4B$13.4B$12.4B$11.4B$10.4B$9.4B$8.4B
$7.4B$6.4B$5.4B$4.4B$3.4B$2.4B$.D3B$D3B$3D!

By replacing an eater with a block it becomes a 98 tick H-to-B. The B is in a difficult position, but a known B-to-2G luckily needs a block in exactly that position, and together it becomes a H->G0 with an extra glider to spare.

Nice observation! I didn't think that particular B->2G would ever be useful since it is nearly impossible to place a B-heptomino that close to the block. There is also this alternative B->2G that uses the same starting reaction:

Code: Select all

x = 81, y = 93, rule = LifeHistory
39.A11.2A$39.3A7.BABAB$42.A3.2B.B2AB$41.2A.6B.B$41.13B$43.12B$43.12B$
39.2D.13B$38.2D14B4.2A$39.BD13B3.A.A$40.12B5.BA$41.11B6.2B$42.10B5.4B
$43.19B$44.18B$45.18B$45.18B$46.17B$46.16B$46.15B$44.16B$44.16B$42.2A
B.12B2A3.A11.A$41.A.AB2.11B2A3.3A7.3A$41.A6.11B7.A5.A6.A$40.2A8.8B7.
2A5.2A3.3A$51.3BD3B7.9B2.A$51.2B3D3B.12B3.B2A$49.3BD2B2D16B.4B$48.2A
27B$48.2A27B$49.28B.2B$50.29B2A$53.24B.B2A$53.24B2.B$53.18B.3B$53.2BE
13B5.2B$52.3BEBE4B.6B4.B2AB$53.2B3E4B2.B.5B3.2A$52.5BE4B7.2A$51.10B8.
A$50.4B16.3A$49.4B19.A$48.4B$47.4B$46.4B$45.4B$44.4B$43.4B$42.4B$41.
4B$40.4B$39.4B$38.4B$37.4B$36.4B$35.4B$34.4B$33.4B$32.D3B$31.D3B$30.B
3D$29.4B$28.4B$27.4B$26.4B$25.4B$24.4B$23.4B$22.4B$21.4B$20.4B$19.4B$
18.4B$17.4B$16.4B$15.4B$14.4B$13.4B$12.4B$11.4B$10.4B$9.4B$8.4B$7.4B$
6.4B$5.4B$4.4B$3.4B$2.4B$.D3B$D3B$3D!

[simsim314]

Period multiplier 2H->G

Code: Select all

x = 61, y = 33, rule = LifeHistory
$20.C$20.3C$23.C$22.2C31.C$10.2C43.3C$11.C46.C$11.C.C43.2C$12.2C31.2C
$46.C$46.C.C$47.2C4$12.C$12.C.C$12.3C$14.C32.C$47.C.C8.2C$47.3C8.2C$
49.C2$5.2C$6.C$3.3C$3.C17.2C17.2C$22.C18.C$19.3C16.3C$19.C18.C17.2C$
57.C$54.3C$54.C!

[dvgrn]

Period multiplier 2H->G...

It might make sense to put period multipliers in a separate thread -- and luckily there's one already started. I posted all the multipliers that I knew about in the response to the first post. This 2H-to-G is the third one down in the second column.

A H-to-G6. It seems simple but I can't find it anywhere so maybe it's new?

Definitely new, and very useful I think. I was just hoping recently that a Spartan H-to-G0 would turn up. The only thing better would be an actual Spartan Herschel conduit that puts out a G0 pair in passing.

This component creates another series of ways to build Spartan receivers using a G0 plus a surplus parallel glider, with a trivial choice of two timings for the continuation just by moving the semi-Snark block. Obviously this will only be useful when we don't care about fast recovery times:

Code: Select all

x = 226, y = 121, rule = B3/S23
184bo11b2o$184b3o8bobo$187bo7b2o$186b2o2$94bo$92b3o$91bo$83b2o6b2o110b
2o$83bobo116bobo$84bo118bo2$89b2o$89b2o5b2o$96b2o3$79b2o$79b2o2$94b2o$
94bobo$96bo90b2o14b2o3bo11bo$96b2o88bobo14b2o3b3o7b3o$186bo24bo5bo6bo$
185b2o23b2o5b2o3b3o$221bo$183bo37b2o$181b3o$168bo11bo12b2o$168b3o9b2o
11b2o$105b2o64bo$105b2o51b2o10b2o52b2o$143bo15bo64b2o$141b3o15bobo$
140bo19b2o$140b2o41b2o15bo$131b2o50b2o15bobo16b2o$131bobo66b3o16b2o$
132bo69bo11b2o$105b2o107bo$104bo2bo107b3o$105b2o36b2o72bo$112b2o29b2o$
112b2o$174b2o$174bo$175b3o$177bo3$99b2o$99b2o$103b2o$102bobo$102bo$
101b2o$114b2o$108b2o4b2o$108b2o2$116b2o$110b2o4b2o$110b2o$123b2o$123bo
13bo$121bobo12bo$121b2o13b3o$125b2o$125b2o3$48bo$48b3o$51bo$50b2o$112b
2o$81b2o29b2o$8bo72b2o36b2o$8b3o107bo2bo$11bo107b2o$10b2o81bo$5b2o85bo
bo$5b2o34b2o50b2o$41b2o41b2o$64b2o19bo$64bobo15b3o$2o64bo15bo$2o52b2o
10b2o51b2o$54bo64b2o$31b2o11b2o9b3o$31b2o12bo11bo$42b3o$3b2o37bo$4bo$b
3o3b2o5b2o23b2o$bo6bo5bo24bo$5b3o7b3o3b2o14bobo88b2o$5bo11bo3b2o14b2o
90bo$129bobo$130b2o2$145b2o$145b2o2$136b2o$128b2o6b2o$128b2o3$22bo118b
o$21bobo116bobo$21b2o110b2o6b2o$134bo$131b3o$131bo4$32b2o$32b2o!
#C [[ AUTOSTART STEP 8 ZOOM 2 HEIGHT 320 THEME 7 ]]

In general I'd say that the new Spartan H-to-G4 seems to be a more efficient way of doing this kind of single-reflection trick, though. H-to-G0s are a lot more useful in a non-Spartan context when we can also use Snarks, to reflect either the doubled stream before the semi-Snark, or the single stream after it.

Anyway how many H-to-G0s have we really seen? Is this the first one?

Look in the new elementary-conduits collection. There's a known H-to-G0, but it needs an eater2 -- unless someone can see how to replace that one, too...!

[simsim314]

Pi->2H although not sure if unknown/useful (recovery 98):

Code: Select all

x = 69, y = 41, rule = LifeHistory
24.C$24.3C$27.C$26.2C$26.B16.2C$24.3B15.C2BCB$22.6B15.2C2B$19.10B2.B
3.B3.B4.3B$5.6B2.2B3.24B.5B$.B2.3BD15B2C23B$2CB.BDBD15B2C24B$2C3B3D
41B$.2B.BD43B$4.45B$12.B.13B5.3B.13B$18.B.7B5.2B.13B$35.15B$36.15B$
35.16B$35.17B$35.17B$36.17B$38.B3C12B$39.2BC12B$39.3C11B$37.14B$36.
11B$36.12B.2B.3B.B$37.28B$37.13B2C11BDB.2B$38.12B2C9B3D3B2C$40.21BDBD
B.B2C$38.23BD3B2.B$38.2C.11B2.2B2.6B$39.C2.9B$36.3C3.9B$36.C4.8B$40.
2CB.4B$39.C.CB.2B2CB$39.C6.2C$38.2C!

[simsim314]

Looks interesting to me, but I couldn't find any way to fix/improve it:

Code: Select all

x = 35, y = 43, rule = B3/S23
28bo$27bo$27b3o4$9b2o$2o6bo2bo$2o7b2o16$b2o$b2o6$16b2o$16b2o15b2o$33b
2o2$4b2o$5bo$2b3o$2bo$6b2o$5bobo4b2o$5bo6b2o$4b2o!

[Kazyan]

Well, Bellman can get rid of the difficult block, which I know is not the answer you're looking for. But that still makes this a non-Spartan G2/5/6 receiver that outputs at a 45 degree angle instead of 135 degrees, which is interesting in its own right. Nice find!

Code: Select all

x = 53, y = 50, rule = B3/S23
51bo$50bo$50b3o10$8b2o$7bobo15b2o$7bo8b2o6bo2bo$4b2ob2o7b2o7b2o$5bo$5b
o$ob2ob2o$2obo$3bo$3b2o$35bo$34bo$34b3o$27b2o$26bo2bo$27b2o4$17b2o$17b
2o6$32b2o$32b2o15b2o$49b2o2$20b2o$21bo$18b3o$18bo$22b2o$21bobo4b2o$21b
o6b2o$20b2o!

[simeks]

Inspired by Chris's recent gunmaking with different ways to shoot down the block of the natural blocklaying B->H, it occured to me that this new H->G6 can also be made into a B->G:

Code: Select all

x = 64, y = 54, rule = LifeHistory
46.C11.C$46.3C7.3C$33.C15.C5.C6.C$33.3C12.2C5.2C3.3C$36.C11.9B2.C$35.
2C3.2B.12B3.B2C$28.C6.4B.16B.4B$28.3C6.23B$31.C4.24B$30.2C4.24B.2B$
30.32B2C$30.30B.B2C$29.31B2.B$28.26B.3B$28.2BE21B5.2B$28.3BE13B.6B4.B
2CB$27.4B2E12B2.B.5B3.2C$27.3B2E13B7.2C$27.3BE13B8.C$27.15B11.3C$26.
12B17.C$26.13B$26.12B$27.10B2.B$28.9B.B2C$27.12B2C$26.12B.B$25.4B4.4B
$24.4B5.3B$23.4B$22.4B$21.4B$20.4B$19.4B$18.4B$17.4B$16.4B$15.4B$14.
4B$13.4B$12.4B$11.4B$10.4B$9.4B$8.4B$7.4B$6.4B$5.4B$4.4B$3.4B$2.4B$.D
3B$D3B$3D!

Edit: Connects to this version of the 58 tick H->B in F117:

Code: Select all

x = 55, y = 26, rule = LifeHistory
10.C19.2A$10.3C17.2A$13.C$2C10.2C$.C$.C.C42.A$2.2C40.3A$43.A$43.2A$
48.2A$48.2A3$2.E$2.E.E48.2A$2.3E48.2A$4.E10.2C4.2A$15.2C5.A$11.2C6.3A
$12.C6.A$9.3C14.2A22.2A$9.C17.A22.A$24.3A12.2A5.2A3.3A$24.A15.A5.A6.A
$37.3A7.3A$37.A11.A!

[chris_c]

Inspired by Chris's recent gunmaking with different ways to shoot down the block of the natural blocklaying B->H, it occured to me that this new H->G6 can also be made into a B->G.

Ah, nice. I just pushed some more updates to the gun collection a few minutes ago but my solution was much bulkier than yours. On the plus side my method did allow this fun looking p151:

Code: Select all

x = 98, y = 55, rule = B3/S23
73bo11b2o$72bobo10b2o$64bo3b2o2bobo$63bobo2bo2b2ob2o$63bobo3bobo$64bob
4o2bob2o$32bo33bo3bobob2o$30b3o32bo3bobo$29bo11bo18bo3bo3bobo$29b2o8b
3o16b3o3b2o3bo$38bo18bo$38b2o17b2o$17b2o34b2o$9bo7b2o34b2o22b2o$9b3o
65b2o3b2o$12bo69bobo7b2o$11b2o69bo8bo2bo$92b2obo$46b2o47bo$3b2o40bo2bo
46b2o$3bo40b2o2bo31b2o$2obo21b2o18b2o2bo31bo$o2b3o4b2o14bo19b4o28b3o7b
o$b2o3bo3b2o11b3o40b2o10bo8bobo$3b4o16bo42bobo18bo2bo$3bo15b2o47bo20bo
$4b3o12bobo34b2o10b2o$7bo13bo34bo36b2o$2b5o14b2o34b3o13b2o17bo2bo$2bo
56bo14bo17bobo$4bo69bobo16bo$3b2o59bo10b2o12b3o$18bo11bo31b3o24b2o$19b
2o7b3o30bo$18b2o7bo33b2o$27b2o17b2o48b2o$47bo48b2o$47bob2o$48bo2bo$49b
2o$53b2o2b2o$52b3o2b3o$17b2o34b2obo3bo$16bobo5b2o30b5obo$16bo7b2o30bo
5bo$15b2o39bo2b2obo10bo3b2o$57bo4bo9bobo3bo$29bo31b2o8bobo3bo$25b2obob
o26b3obo5b2obobo3bo$24bobobobo26bo9b2obo2b4obo5b2o$21bo2bobobobob2o24b
2o11bobo3bobo5bo$21b4ob2o2bo2bo33b2ob2o2bo2bobo2b3o$25bo4b2o36bobo2b2o
3bo3bo13bo$23bobo30b2o10bobo26bo$23b2o31b2o11bo25b3o!

I can get your B->G working from p453 and above. If there is a way to get it working at p445 then drop it over on the other thread I guess.

[dvgrn]

Pi->2H although not sure if unknown/useful (recovery 98)...

Paul Callahan came up with this exact same pi-to-2H circuit on 29 Oct 1997 -- the paragraph started "Here's something frustrating"...!

He used it as the basis for the standard Fx176 Herschel conduit, but was unable to successfully extract the second Herschel. I've tried to get some use out of it several times since then, but it seems as if that second Herschel output always has to be sacrificed.

One vaguely successful derivative circuit was the Spartan ladder used in the original prototype universal constructor (Golly's Patterns/Life/Signal-Circuitry/constructor-memory-tape.rle). I guess this is a Spartan version of simeks' recent invention involving freezing and thawing a Herschel. There are several other Spartan "frozen Herschel" tricks involving boats and beehives and other block pairs, too. Here's an extract of the old ladder circuit:

Code: Select all

x = 248, y = 244, rule = LifeHistory
245.A$245.A.A$245.2A69$149.2A$150.A$150.A.A$151.2A10$158.2D$158.2D18$
152.2A$152.2A2$159.A$159.3A$162.A$161.2A2$142.2A$142.A.A$144.A$144.2A
4$163.2A$163.A$161.A.A$144.2A15.2A$144.2A3$161.2A$161.A.A$163.A$163.
2A3$89.A$89.3A$92.A$91.2A52.2A$87.2A57.A$88.A54.3A$88.A.A52.A$82.2A5.
2A$83.A$83.A.A$84.2A7$96.2D$96.2D36.2A$134.A$132.A.A18.2A$132.2A10.2A
7.2A$145.A$145.A.A$140.2A4.2A$141.A20.2A$141.A.A18.A$106.2A34.2A16.A.
A$106.2A52.2A3$94.2A$95.A$92.3A$92.A$96.2A$95.A.A4.2A27.2A$95.A6.2A
14.2A11.2A$94.2A23.A$85.2A29.3A35.2A$85.2A29.A37.2A2.2A$158.A.A$92.A
42.2A23.A$92.3A41.A23.2A$95.A37.3A$94.2A37.A2$75.2A$75.A.A$77.A$77.2A
4$96.2A$96.A$94.A.A$77.2A15.2A$77.2A3$94.2A$94.A.A$96.A$96.2A3$22.A$
22.3A$25.A19.A$24.2A19.A.A30.2A$20.2A23.2A32.A$21.A54.3A$21.A.A52.A$
22.2A2$16.A$A15.3A$3A16.A$3.A14.2A$2.2A5$11.C55.2A$9.3C55.A$9.C.C53.A
.A18.2A$9.C55.2A10.2A7.2A$78.A$78.A.A$73.2A4.2A$74.A20.2A$74.A.A18.A$
39.2A34.2A16.A.A$9.2A28.2A52.2A$8.A.A$8.A10.2A$7.2A11.A6.2A$17.3A8.A$
17.A7.3A$25.A$29.2A$28.A.A4.2A27.2A$28.A6.2A14.2A11.2A$27.2A23.A$49.
3A35.2A$49.A37.2A2.2A$91.A.A$68.2A23.A$69.A23.2A$66.3A$66.A!

There's also the option of catching a boat-bit and turning it into a block and then a Herschel with an unsynchronized G2 glider pair:

Code: Select all

x = 69, y = 47, rule = LifeHistory
10.A18.A$3.A6.3A16.A.A$3.3A7.A15.2A$6.A5.2A$5.2A$.2A$2.A$2.A.A$3.2A3$
20.A$20.A.A$20.2A3$9.A$9.A.A$9.2A$2A46.2A$.A46.A$.A.A42.A.A$2.2A42.2A
6$20.2A$20.2A2$67.A$8.2A56.A.A$9.A56.A.A$6.3A58.A$6.A49.D$10.2A44.D.D
$9.A.A4.2A27.2A9.3D$9.A6.2A14.2A11.2A11.D$8.2A23.A$30.3A$30.A2$49.2A$
50.A$47.3A$47.A!

Notice that if the first glider is not present to create the boat-bit, the other two gliders can pass through the circuit and be caught by a Herschel receiver (or another boat-bit-to-pi circuit) farther down the line. We have H-to-G2s that are almost Spartan now -- just need a pair of bookends, and those are relatively slow-constructible now as I recall.

[Extrementhusiast]

In other news, an O-to-G:

Code: Select all

x = 11, y = 15, rule = LifeHistory
9.2A$9.A$7.A.A$7.2A3$.C$3C$C.C$2.2C3$8.A$7.A.A$8.A!

[praosylen]

Possibly known R-to-Pi:

Code: Select all

x = 18, y = 16, rule = LifeHistory
7.3D$7.DBD$7.DBDB$3.2B.5B$2.2C7B$2.2C.6B$3.10B$5.8B$5.7B$4.7B.B$.CB2.
6B.2B$C.C11B.2B$.C7BC6B2C$2.B.4B3C5B2C$6.2BC5B.2B$7.6B!

Is there any collection of X-to-Y converters that I could use to check whether my discoveries are known so that I don't have to waste posts on old converters that I've rediscovered?
Edit: it seems to be a Spartan variation of this one that I found in an old post in this topic:

Code: Select all

x = 17, y = 16, rule = LifeHistory
8.3D$8.DBD$8.DBDB$4.2B.5B$3.2A7B$3.2A.6B$4.10B$6.8B$4.B.9B$3.2A10B$3.
A.7B.3B$2A.A.B.10B$2A.2A2.3BC6B$6.3B3C5B$7.2BC7B$8.8B!

[dvgrn]

Is there any collection of X-to-Y converters that I could use to check whether my discoveries are known so that I don't have to waste posts on old converters that I've rediscovered?

The new elementary conduits stamp collection is aiming to be a reasonably complete table of all known converters.

Edit: it seems to be a Spartan variation of this one that I found in an old post in this topic...

Eventually I'm hoping to put a label on each converter conduit. Seems to me that this will make it a lot easier to notice if a new discovery is a variant of a known converter or not.

This particular reaction would boil down to "RF29Pi" -- input is an R pentomino, the reaction moves forward 29 steps and turns into a pi heptomino. With any luck we won't very often see cases where two elementary conduits put their outputs in different locations but take the same amount of time, and thus get the same identifier string. [For those very rare cases we can maybe add location information: RF29Pi(12,1).]

Pi heptominos are symmetrical, so there's no such thing as a RFx(n)Pi conduit (where "x" means "flipped") -- and there's no such thing as a left or right turn starting from a pi, queen bee, honeyfarm, traffic light, etc. From symmetric starting points there are only turns, "T" instead of "L" or "R", which can have standard or flipped output if the resulting object is aysmmetrical -- e.g., PiTx9B.

Canonical orientations for each of the standard input and output object types are shown in the elementary conduits stamp collection, and there's a viewer image of the common ones in the same posting.

[dvgrn]

If I could make a suggestion for this thread: it would be very useful if posts of candidate additions to the elementary-conduits collection could also each include a second pattern showing a sample way to connect up the inputs and outputs to other known conduits.

When I look back to the old posting in question, it appears that this latest R-to-pi shares the same limitation as the table-and-block version. Currently RF29Pi variants can only be used in one extraordinarily unlikely-looking still-not-quite-Spartan H-to-G converter: EDIT: and the other H-to-G that Sokwe mentions a few posts down.

Code: Select all

x = 61, y = 78, rule = LifeHistory
35.4B$36.4B$37.4B$38.4B$39.4B$40.4B$41.4B$42.4B$43.4B$44.4B$45.4B$46.
6B$45.8B$45.9B$44.11B$43.11B5.2A$42.12B5.A$42.11B3.BA.A$41.14B.B2A$
40.17B$38.19B$37.20B$32.B2.21B$31.3B.22B$16.A14.26B$16.3A5.2A5.28B$
19.A3.B2AB3.30B$18.2A4.3B3.29B$18.5B2.B2.30B$20.7BD30B$19.2A7BD23B.4B
$19.2A3B3DB2D21B2.3B$20.B.2BDB3D22B4.AB$22.2BDB2D23B3.A.A$20.2B.28B3.
2A$19.2A30B$19.2A.6B2.2B2A18B$20.10B2.2A2B4.2B.8B$22.8B4.B9.6B$22.9B
12.7B$21.11B11.6B$18.AB2.10B11.8B$17.A.A14B7.5B3.2A$18.A7BD6B2A6.2AB
5.A$19.B.4B3D5B2A7.A7.3A$16.B5.3BD5B.2B5.3A10.A$22.B.6B9.A$21.9B$19.
11B$19.10B$18.12B$17.14B$17.17B$18.15B2A$18.15B2A$17.17B$14.16B$13.
16B$13.16B$12.16B$11.20B$11.20B$12.20B$11.22B$9.24B$7.25B$7.2BC13B.
10B$6.3BCBC4B.7B2.2B4.2A$7.2B3C4B2.3B.B4.B4.A$6.5BC4B4.2B2A3.2A4.3A$
5.10B4.BA2.A4.A6.A$4.4B12.3A5.A.A$3.4B22.2A$2.4B14.3A$.B2AB15.A2.A$.
2BA18.2A$3A$A!

This is not to say that other connecting conduits might not be found in the future. It's just good to know whether a new conduit is likely to have lots of immediate applications or not.