The Hunting of the New Herschel Conduits

For discussion of specific patterns or specific families of patterns, both newly-discovered and well-known.
User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » May 30th, 2015, 1:01 am

chris_c wrote:Meanwhile, here is a lower bound for the size of a G->boat for the recently discussed highway robber...
Sometimes it's easier to just build a working sample of the whole circuit, just to get a better sense of what is going to need to be optimized:

Code: Select all

#C badly designed highway robber --
#C current recovery time is 1797 ticks,
#C   but it will be easy to improve that.
x = 236, y = 259, rule = LifeHistory
95.A$93.2A$94.2A2$92.A$90.2A$91.2A$88.A$86.2A$87.2A2$85.E$83.2E$84.2E
39$55.A$53.3A$52.A9.2A3.2A$52.2A8.2A3.2A3$74.2A$74.A$72.A.A$72.2A3$
37.D$36.D.D$36.D.D$37.D$69.2A$37.D31.A$36.D.D31.3A$37.2D33.A41.A11.2A
$113.A.A10.2A$105.A3.2A2.A.A$104.A.A2.A2.2A.2A$104.A.A3.A.A$105.A.4A
2.A.2A$107.A3.A.A.2A$56.2A48.A3.A.A$56.2A47.A3.A.A$64.2A39.2A3.A$64.A
26.A$65.3A9.A13.3A$67.A7.3A16.A$74.A18.2A23.2A$74.2A42.2A$133.2A$132.
A2.A$133.2A.A$136.A$36.2A15.2C11.C69.2A$36.2A15.C.C10.3C52.2A$53.C12.
C.C15.2A36.A$35.C7.A.2A21.C15.2A33.3A$34.3C6.2A.A60.2A10.A$34.C.2C69.
A.A$109.A$25.2A20.2A48.2A10.2A$26.A20.2A48.A$26.A.A24.2A32.2A9.3A$27.
2A24.2A33.A11.A$87.A$87.2A$41.2A8.2A$41.A.A7.2A5.2A$43.A14.2A$43.2A4$
18.2A$17.A2.A2.2A$18.2A2.A.A$20.2A16.2A$20.A17.A$17.2A.A2.A15.3A$17.A
.2A.A.A16.A$21.A.A$18.2A2.A$16.3A.2A$15.A$16.3A.2A$18.A.2A2$28.2A$28.
2A7.2A$37.A$35.A.A$35.2A4$15.2A$15.2A5$31.A$30.A.A$30.A.A20.2A$31.A
21.A$32.3A.2A17.A$34.A.A14.5A$36.A13.A$36.A.A12.3A$37.2A15.A$51.4A$
46.2A3.A3.2A$46.2A4.3A2.A$54.A.2A$54.A$53.2A5.2A11.A$60.2A10.A.A$72.A
.A2.2A3.A$45.2A24.2A.2A2.A2.A.A$45.A29.A.A3.A.A12.2A$46.3A22.2A.A2.4A
.A13.A$48.A22.2A.A.A3.A13.A.A$75.A.A3.A12.2A$76.A.A3.A$77.A3.2A$102.
2A$103.A$103.A.A$68.2A34.2A$68.2A$53.2A$52.A2.A$51.A.2A70.2A$51.A73.
2A$50.2A$45.2A18.2A82.A$46.A18.A83.3A$3.2A39.A21.3A24.2A57.A$4.A39.5A
14.3A2.A11.2A11.2A56.A.A$2.A46.A13.A2.A14.A69.A.A$2.5A14.2A23.3A12.A.
A2.2A10.3A71.A$7.A13.A23.A15.2A15.A$4.3A12.A.A23.4A$3.A15.2A22.2A3.A
3.2A42.2A$3.4A35.A2.3A4.2A41.A.A133.2A$.2A3.A3.2A30.2A.A49.A71.2A62.A
$A2.3A4.2A33.A31.A16.2A71.2A64.A$2A.A41.2A29.A.A7.A11.2A15.2A96.2A14.
5A$3.A72.A.A6.A.A10.2A15.2A16.2A79.A13.A$3.2A68.2A.A.A.2A3.A.A44.A.A
79.A.A12.3A$53.2A18.A2.A.A.A.A.2A.2A37.2A4.A14.2A66.2A15.A$54.A16.A.A
2.A.A3.A.A39.A2.A2.2A.4A9.A.A80.4A$11.2A38.3A17.2A4.A.4A2.A.2A35.2A.A
.A.A.A2.A9.A77.2A3.A3.2A$12.A38.A27.A3.A.A.2A38.A.A.A.A11.2A7.2A68.2A
4.3A2.A$9.3A66.A3.A.A42.A.A.2A21.2A76.A.2A$9.A67.A3.A.A44.A103.A$62.A
14.2A3.A79.2A.A65.2A$61.A.A77.2A19.2A.3A$62.A69.2A7.A26.A$132.2A5.A.A
20.2A.3A55.2A$90.2A47.2A20.A2.2A57.A$90.2A68.A.A61.3A$105.2A35.A16.A.
A.2A.A59.A$104.A2.A34.3A15.A2.A.2A$61.2A42.2A.A36.A17.A$61.2A45.A35.
2A16.2A$108.2A49.A.A2.2A$93.2A34.2A28.2A2.A2.A$94.A13.4A18.A33.2A$91.
3A14.A3.A14.3A$79.2A10.A19.2A14.A$79.A.A14.2A$81.A13.A.A$67.2A12.2A6.
2A4.A$67.2A18.A2.A2.2A.4A$87.2A.A.A.A.A2.A$90.A.A.A.A39.2A19.2A$90.A.
A.2A25.2A13.2A19.2A$91.A22.2A5.A.A$114.2A7.A$104.2A17.2A33.2A$95.2A7.
A53.A$95.2A5.A.A5.A48.3A$102.2A5.A.A.2A46.A$109.A.A.A.A$106.2A.A.A.A.
A2.A$106.A2.A2.2A.4A$108.2A4.A$114.A.A$115.2A$92.2A$93.A$70.A19.3A$
70.3A17.A80.A$73.A95.3A$72.2A94.A$168.2A2$146.A$117.A27.A.A$117.3A24.
A.A$120.A21.3A2.3A3.2A$82.2A35.2A20.A3.2A3.A2.2A$75.2A5.A.A57.3A2.2A.
A7.2A$75.2A7.A59.A.A2.A7.A.A5.2A$84.2A59.A.A9.A7.2A$146.A9.2A$71.A$
70.A.A.2A94.A$70.A.A.A.A52.2A35.2A.A.A$67.2A.A.A.A.A2.A42.2A5.A.A33.A
.A.A.A$67.A2.A2.2A.4A42.2A7.A30.A2.A.A.A.A.2A$69.2A4.A55.2A29.4A.2A2.
A2.A$75.A.A88.A4.2A$76.2A40.A45.A.A$117.A.A.2A41.2A$117.A.A.A.A$114.
2A.A.A.A.A2.A$114.A2.A2.2A.4A$116.2A4.A$122.A.A$123.2A!
Probably a different edge-shooter for the return glider will work out better -- maybe the new SE39T12, or the old NE30T3?

I don't really have a good sense yet of how much the recovery time can be reduced. It will be easy to find a custom Herschel track that wanders off further to the southwest, out of the way -- that should save several hundred ticks anyway.

Then there might be ways to use staged-recovery mechanisms to make things easier. E.g., use something like Brice Due's demultiplexer, or the more recent H-to-boat, to bounce the highway-robber output glider out of the path of the return glider:

Code: Select all

x = 40, y = 35, rule = LifeHistory
A.A$.2A$.A17$38.2A$38.A$36.A.A$36.2A6$17.2A$16.A.A18.A$16.A18.A.A$15.
2A4.2A12.3A$20.A.A12.A$20.A$19.2A!
Then you no longer need an edge shooter to build the return glider... though this particular demultiplexer sends out a glider on a bad lane, so some kind of transparent conduit would still be needed. Some other bait still life would probably be better, as long as the conduit generating it was transparent to the return glider path. Anyway, then there's all kinds of time to route an extra signal around to rebuild the bait still life.

Synchronizing the B-heptomino and the return glider is clearly what takes most of the space and time, so as soon as a good edgy H-to-boat converter shows up, the circuitry will get a lot smaller.

User avatar
Scorbie
Posts: 1692
Joined: December 7th, 2013, 1:05 am

Re: The Hunting of the New Herschel Conduits

Post by Scorbie » May 30th, 2015, 2:00 am

How about generating the bait with glider and xwss collisions? Is this possible for making a perfect highway robber?

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » May 30th, 2015, 9:12 am

Scorbie wrote:How about generating the bait with glider and xwss collisions? Is this possible for making a perfect highway robber?
I thought about that, while figuring out the replacement for the beehive inserter in chris_c's original version of the highway robber. it would certainly minimize the amount of activity near the highway, so it might work, but I don't know how exactly.

Unfortunately synchronizing LWSSes and gliders is just as time- and space-consuming as synchronizing Herschels and gliders. The current version is only a few ticks away from optimal in any case.

Another interesting idea would be to find a bait object that leaves a block or something behind, and rebuild the bait with a slow salvo. More signals, but no synchronization.

Here's a slight improvement -- 1624 ticks recovery, and a smaller bounding box. Obviously the next thing to tackle is the edge shooter in the middle:

Code: Select all

x = 207, y = 240, rule = LifeHistory
116.A$114.2A$115.2A2$113.A$111.2A$112.2A$109.A$107.2A$108.2A2$106.E$
104.2E$105.2E39$76.A$74.3A$73.A9.2A3.2A$73.2A8.2A3.2A3$95.2A$95.A$93.
A.A$93.2A3$58.D$57.D.D$57.D.D$58.D$90.2A$58.D31.A$57.D.D31.3A$58.2D
33.A7$77.2A$77.2A$85.2A$85.A$86.3A9.A14.2A$88.A7.3A14.A24.2A$95.A18.
3A6.2A13.A$95.2A19.A6.2A11.A.A$136.2A13.2A$151.2A3$57.2A15.2C11.C69.
2A$57.2A15.C.C10.3C67.2A$74.C12.C.C15.2A46.2A$56.C7.A.2A21.C15.2A46.
2A$55.3C6.2A.A$55.C.2C2$46.2A20.2A88.2A$47.A20.2A88.2A$47.A.A24.2A32.
2A$48.2A24.2A33.A$108.A13.2A3.2A$108.2A13.A3.A$62.2A8.2A46.3A5.3A3.2A
$62.A.A7.2A5.2A39.A9.A2.A2.A2.2A$64.A14.2A53.2A2.A.A$64.2A70.2A$136.A
$19.2A11.A100.2A.A2.A$19.2A10.A.A99.A.2A.A.A10.2A$31.A.A2.2A3.A95.A.A
11.A$30.2A.2A2.A2.A.A91.2A2.A13.3A$34.A.A3.A.A89.3A.2A16.A$30.2A.A2.
4A.A17.2A70.A$30.2A.A.A3.A19.A72.3A.2A$34.A.A3.A19.3A71.A.2A$35.A.A3.
A20.A$36.A3.2A102.2A$144.2A7.2A$153.A$151.A.A$27.2A122.2A$27.2A$12.2A
$11.A2.A$10.A.2A117.2A$10.A120.2A$9.2A$24.2A9.2A21.2A$3.2A19.A9.A.A
21.2A$4.A20.3A6.A17.2A$2.A24.A5.2A17.2A93.A$2.5A14.2A123.A.A$7.A13.A
40.2A11.A70.A.A$4.3A12.A.A32.2A6.2A10.A.A70.A$3.A15.2A26.2A5.2A18.A.A
2.2A3.A63.3A$3.4A16.A6.2A15.2A24.2A.2A2.A2.A.A64.A$.2A3.A3.2A10.A.A5.
2A45.A.A3.A.A12.2A$A2.3A4.2A11.A49.2A.A2.4A.A13.A$2A.A69.2A.A.A3.A13.
A.A$3.A73.A.A3.A12.2A$3.2A73.A.A3.A$39.2A38.A3.2A$39.2A63.2A$11.2A92.
A$12.A92.A.A$9.3A5.2A51.2A34.2A$9.A8.A51.2A$15.3A37.2A$15.A38.A2.A89.
A$53.A.2A70.2A18.3A$13.5A35.A73.2A21.A$12.A5.A2.2A29.2A95.A.A$13.5A2.
A.A44.2A80.A.A$18.2A16.2A29.A82.A39.A$15.2A.A17.A31.3A24.2A91.3A$15.A
2.A2.A15.3A30.A11.2A11.2A90.A$16.3A.A.A16.A43.A103.2A$19.A.A58.3A$16.
2A2.A20.2A37.A84.2A$14.3A.2A22.A122.2A6.2A.2A$13.A26.A57.2A74.A.A2.A$
14.3A.2A20.5A14.2A36.A.A31.2A41.A.A.2A$16.A.2A25.A13.A37.A32.A.A42.2A
13.2A$42.3A12.A.A36.2A26.2A4.A14.2A31.2A10.2A$26.2A13.A15.2A58.2A3.A
2.A2.2A.4A9.A.A28.3A2.A$26.2A7.2A4.4A72.2A3.2A.A.A.A.A2.A9.A29.A3.2A$
35.A3.2A3.A3.2A75.A.A.A.A11.2A7.2A21.A2.A$33.A.A2.A2.3A4.2A75.A.A.2A
21.2A22.A.A$33.2A3.2A.A84.A50.A21.A$41.A118.2A.A33.3A$41.2A96.2A19.2A
.3A30.A$130.2A7.A26.A29.2A$13.2A115.2A5.A.A20.2A.3A$13.2A34.2A86.2A
20.A2.2A$50.A107.A.A13.2A19.2A$47.3A90.A16.A.A.2A.A9.2A19.2A$47.A92.
3A15.A2.A.2A$143.A17.A$29.A112.2A16.2A$28.A.A126.A.A2.2A$28.A.A96.2A
28.2A2.A2.A$29.A98.A33.2A13.2A$30.3A92.3A50.A$32.A92.A49.3A24.2A$175.
A21.2A2.A2.A$191.3D3.A.A2.2A$182.2A8.D7.2A$183.A6.3D8.A$180.3A15.A2.A
.2A$134.2A19.2A23.A16.A.A.2A.A$134.2A19.2A41.A.A$199.A2.2A$200.2A.3A$
156.2A48.A$156.A43.2A.3A$157.3A40.2A.A$159.A$192.2A$183.2A7.2A$184.A$
184.A.A$185.2A4$205.2A$205.2A4$77.A$77.3A64.A29.A15.A$80.A62.A.A26.3A
14.A.A$79.2A61.A.A26.A17.A.A$140.3A2.3A3.2A18.2A17.A$139.A3.2A3.A2.2A
34.3A$111.A28.3A2.2A.A38.A$111.3A28.A.A2.A$114.A28.A.A$113.2A29.A$89.
2A$82.2A5.A.A69.2A$82.2A7.A68.A.A5.2A$91.2A67.A7.2A$159.2A$78.A$77.A.
A.2A40.2A48.A$77.A.A.A.A32.2A5.A.A43.2A.A.A$74.2A.A.A.A.A2.A29.2A7.A
42.A.A.A.A$74.A2.A2.2A.4A38.2A38.A2.A.A.A.A.2A$76.2A4.A82.4A.2A2.A2.A
$82.A.A27.A56.A4.2A$83.2A26.A.A.2A50.A.A$111.A.A.A.A49.2A$108.2A.A.A.
A.A2.A$108.A2.A2.2A.4A$110.2A4.A$116.A.A$117.2A!

User avatar
praosylen
Posts: 2443
Joined: September 13th, 2014, 5:36 pm
Location: Pembina University, Home of the Gliders
Contact:

Re: The Hunting of the New Herschel Conduits

Post by praosylen » May 30th, 2015, 11:54 am

Tiny (yet probably useless) H-to-R:

Code: Select all

x = 14, y = 16, rule = LifeHistory
9.2A$9.A$.A.2A.2A.A$.2A.A.A.A$6.BA$5.4B$5.8B$5.9B$4.10B$4.10B$2.2AB2.
3B3CB$.A.AB2.4BC2B$.A5.2B3C2B$2A6.6B$8.5B$10.B!
EDIT: H-to-B with the eater in a different place. I think it's new:

Code: Select all

x = 14, y = 16, rule = LifeHistory
8.2A$8.A$A.2A.2A.A$2A.A.A.A$5.BA$4.4B$4.8B$4.9B$4.9B$4.9B$2.7B3CB$2.
2A2.4BC2D$3.A2.2B3CB2D$3A4.4B2D$A6.5B$9.B!
former username: A for Awesome
praosylen#5847 (Discord)

The only decision I made was made
of flowers, to jump universes to one of springtime in
a land of former winter, where no invisible walls stood,
or could stand for more than a few hours at most...

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: The Hunting of the New Herschel Conduits

Post by simsim314 » May 30th, 2015, 12:56 pm

H->G9:

Code: Select all

x = 161, y = 61, rule = LifeHistory
12$28.C$28.3C97.2C$31.C95.C2.C$30.2C96.C.C$129.C6$18.C30.2C67.C30.2C$
18.3C27.C2.C66.3C27.C2.C$21.C26.C.C70.C26.C.C$20.2C27.C70.2C27.C11$
21.2C98.2C$20.C.C97.C.C$20.2C98.2C$35.3C97.3C$23.2C11.C86.2C11.C$23.C
10.3C86.C10.3C$24.3C97.3C$26.C99.C!
EDIT Without the eater on top, it's pretty nice H->Block factory:

Code: Select all

x = 52, y = 36, rule = LifeHistory
6$7.C30.2C$7.3C27.C2.C$10.C26.C.C$9.2C27.C11$10.2C$9.C.C$9.2C$24.3C$
12.2C11.C$12.C10.3C$13.3C$15.C!
EDIT Another H to first natural B:

Code: Select all

x = 43, y = 36, rule = LifeHistory
7$34.C$27.2C3.3C$26.C.C2.C$11.2C6.4B3.BC3.2C$12.C5.7B2.6B$12.C.CB.10B
.3B$13.2CB.15B$15.17B.2B$15.19B2C$14.18B.B2C$14.2BC15B2.B$13.3BCBC4B.
7B$14.2B3C4B2.B2DBDB$13.5BC5B.2B3D2B$12.13B.2BD3B$11.4B7.9B$10.4B9.8B
$9.4B11.8B$9.2C14.7B$10.C15.6B$7.3C16.7B$7.C18.7B!
EDIT2 Looks like something I saw, but can't find this H->G:

Code: Select all

x = 50, y = 43, rule = LifeHistory
22.4B$23.4B$24.4B$25.4B$26.4B8.A$19.2A6.4B5.3A$20.A7.4B3.A$20.A.AB4.
4B3.2A$21.2AB.2B.5B.3B$23.12B$23.14B$23.14B$24.14B$25.14B$15.2A8.13B$
16.A9.11B$16.A.AB4.14B$17.2A20B$19.20B$19.19B$20.20B$19.22B$17.23B$
15.24B$15.2BC22B$14.3BCBC4B.10B2.4B$15.2B3C4B2.7B5.4B$14.5BC4B2.8B5.
4B$13.10B2.11B4.4B$12.4B9.11B5.4B$12.3B11.11B5.4B$10.4B11.B.10B6.4B$
10.2A12.2B.9B8.4B$11.A9.2B.12B9.4B$8.3A9.2A13B11.4B$8.A11.2A13B12.3B$
21.2B.13B11.2B$25.6B4.2A12.B$35.A$36.3A$38.A!
EDIT3 This variation found using bellman:

Code: Select all

x = 49, y = 60, rule = LifeHistory
10.4B$11.4B$12.4B$13.4B$14.4B$15.4B$16.4B$17.4B$18.4B8.A$11.2A6.4B5.
3A$12.A7.4B3.A$12.A.AB4.4B3.2A$13.2AB.2B.5B.3B$15.12B$15.14B$15.14B$
16.14B$17.14B$7.2A8.13B$8.A9.11B$8.A.AB4.14B$9.2A20B$11.20B$11.19B$
12.20B$11.22B$9.23B$7.24B$7.2BC22B$6.3BCBC4B.10B2.4B$7.2B3C4B2.7B5.4B
$6.5BC4B2.8B5.4B$5.10B2.12B3.4B$4.4B9.12B4.4B$4.3B11.12B4.4B$2.4B11.B
.10B6.4B$2.2A12.13B2.2A3.4B$3.A9.2B.13B3.A4.4B$3A9.2A17B.A.A3.4B$A11.
2A15B2A.2A.A3.4B$13.2B.12B2.A4.A4.4B$17.7B3.BA.A4.2A4.4B$19.4B4.B2A6.
A5.4B$18.4B3.A8.A8.4B$17.4B4.5A4.2A8.4B$16.4B9.A15.4B$15.4B8.A18.3B$
14.4B9.2A18.2B$13.4B31.B$12.4B$11.4B$10.4B$9.4B$8.4B$7.4B$6.4B$5.4B$
4.4B$3.4B$2.4B!
EDIT4 Another (totally unrelated) bellman assisted H->G9 (with H into 4 gliders):

Code: Select all

x = 37, y = 52, rule = LifeHistory
9$30.2C$29.C.C$29.C$28.2C$27.C$28.2C$29.C$28.C$5.2C22.3C$6.C24.C$6.C.
C$7.2C5$28.C$27.C.C$7.C20.C$7.C.C$7.3C$9.C8$16.2C$17.C$14.3C$14.C!

User avatar
Scorbie
Posts: 1692
Joined: December 7th, 2013, 1:05 am

Re: The Hunting of the New Herschel Conduits

Post by Scorbie » May 31st, 2015, 2:02 am

simsim314 wrote:H->G9:
Technically, that looks like H->BG combined with BG->G(Cool to see that transparent loaf, by the way.) But I don't think the BG could be converted into something else.

Code: Select all

x = 72, y = 28, rule = LifeHistory
A30.2A21.2A$3A7.2B4.2B12.A2BA19.A2BA$3.A5.4B2.4B10.BABAB19.BABA2B$2.
2A.B.13B3.B.6BA2B19.2BA4B$2.30B23.6B$4.26B27.4B$2.15B3D10B27.4B.B$2.
15BDBD10B26.8B$2.14BD2BD10B25.9B$2.14BDBD11B24.12B$3.14BD12B24.13B$4.
14B2.10B24.14B$4.13B5.8B24.15B$3.3B.13B4.6B24.16B$3.2AB2.13B4.4B25.
16B$2.A.AB3.12B33.16B$2.2AB4.12B34.7B2C5B$3.4B3.2B.4B3CB35.5BCBC4B$5.
2A7.4BC2B34.5B2C4B$5.A8.2B3C2B33.12B$6.3A6.6B34.11B$8.A6.7B34.10B$17.
B.4B34.10B$20.4B34.10B$58.6B.4B$60.3B3.4B$61.2B4.4B$68.4B!
By the way, I think I saw a BG->G with a single block somewhere... I just can't find it.

Sokwe
Moderator
Posts: 2644
Joined: July 9th, 2009, 2:44 pm

Re: The Hunting of the New Herschel Conduits

Post by Sokwe » May 31st, 2015, 4:20 am

I have updated the conduit collection:
Conduit_Collection-30May2015.rle.gz
(68.27 KiB) Downloaded 591 times
conduit_compiler-30May2015.zip
(111.52 KiB) Downloaded 540 times
The update includes everything I suggested in this post, most things posted in this topic after that post (except for H->G converters), and a few miscellaneous patterns that for some reason didn't make it into the original collection. I did not add the following recent patterns, since I don't think they are likely to ever work with anything else:

Code: Select all

x = 73, y = 13, rule = LifeHistory
59.A$59.3A7.2A$3.2A57.A6.A$3.2A5.2A49.2A7.A$10.A28.2A28.2A$8.A.A27.A.
A$8.2A24.2A3.A29.2A$33.A2.A31.A.A$34.2A33.A$C29.C29.C9.3A$C.C27.C.C
27.C.C9.A$3C27.3C27.3C$2.C29.C29.C!
I also did not add patterns of this sort, since I didn't know how to classify them:

Code: Select all

x = 109, y = 47, rule = LifeHistory
105.2A$76.2A27.2A$66.2A8.2A$66.2A5$57.2A$47.2A8.2A$47.2A48.A$83.2A11.
A.A$83.2A12.2A8.D$108.D$106.3D$38.2A$28.2A8.2A$28.2A$64.2A13.2A$64.2A
13.A$77.A.A$77.2A$19.2A$9.2A8.2A52.2A$9.2A61.A.A$45.2A13.2A11.A$45.2A
13.A$2A56.A.A$2A56.2A2$54.2A$53.A.A$26.2A13.2A11.A$2.C23.2A13.A$3.C
35.A.A$3.2C34.2A$2.2C$2.C32.2A$34.A.A$7.2A13.2A11.A$7.2A13.A$20.A.A$
20.2A2$16.2A$15.A.A$16.A!
I also made some minor corrections and did some reordering of the patterns that were already in the collection. The following patterns in the collection are actually composite and should perhaps be removed:

Code: Select all

b_b_04.rle
b_b_05.rle
b_b_06.rle
b_pi_02.rle
b_pi_03.rle
h_b_17.rle
h_h_18.rle
pi_h_02.rle
pi_h_03.rle
I made no attempt to update or correct the H->G section.

We still have the possibility of using white cells for catalysts to denote some property of certain conduits (like we do with yellow). Possible ideas are if such conduits are actually just variants of earlier conduits, or alternatively, if such conduits don't connect to any other known conduits.

Thanks again to Chris for making the collection so easy to update.
-Matthias Merzenich

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » May 31st, 2015, 5:43 am

simsim314 wrote:H->G9...
EDIT Without the eater on top, it's pretty nice H->Block factory...
Nice clearance. The block-to-pi-to-Herschel conduit from the U.C. ladder doesn't quite fit here, but there are lots of other ways that this might come in handy. The tub-with-tail trick below is nice (though I don't know about really useful) because the trigger glider is harmlessly absorbed unless the block is present:

Code: Select all

x = 94, y = 135, rule = LifeHistory
82.A$82.A.A$82.2A29$28.2A$28.A$13.2A11.A.A$13.A.A10.2A$14.2A$10.2A$9.
A.A$9.A$8.2A$53.A$52.A.A$52.A.A$51.2A.2A$50.A2.A2.A$34.2D2.3D8.A.A.A.
A.A$34.2D4.D9.A.A.A.A$10.C27.3D11.A.A$10.C.C39.A2.A$10.3C40.2A$12.C2$
52.2A$52.2A2$3.2A6.2A$4.A5.A.A$.3A6.A$.A7.2A3$27.2A$26.A2.A61.A$27.A.
A61.A.A$28.A62.2A28$61.A$61.A.A$61.2A7$27.2A$27.A$12.2A11.A.A$12.A.A
10.2A$13.2A$9.2A$8.A.A$8.A$7.2A6$33.2D$33.2D$9.C$9.C.C$9.3C30.2A$11.C
26.A3.2A$37.A.A$36.A.A$36.A$35.2A$2.2A6.2A$3.A5.A.A$3A6.A$A7.2A3$26.
2A$25.A2.A$26.A.A$27.A!
simsim314 wrote:EDIT2 Looks like something I saw, but can't find this H->G...
EDIT3 This variation found using bellman...
It's superficially similar to a couple of the old known H-to-Gs, #24 and #25, and the NW31T120 glider is the old standard SNG. But the SE22T74 and the SW22T218 are both new as far as I know.
simsim314 wrote:EDIT4 Another (totally unrelated) bellman assisted H->G9 (with H into 4 gliders)...
This one, on the other hand, is all known gliders, according to the identification script -- it is in fact a variant HtoG#24, with various other H-to-Gs bolted on. The Bellman catalyst allows the SNG to escape cleanly, where otherwise it has to be reflected to clean up a block.

I'm working on a May 31 update to the H-to-G stamp collection; with any luck it will be posted by the end of the day.

Sokwe
Moderator
Posts: 2644
Joined: July 9th, 2009, 2:44 pm

Re: The Hunting of the New Herschel Conduits

Post by Sokwe » May 31st, 2015, 6:03 am

dvgrn wrote:
simsim314 wrote:H->G9...
EDIT Without the eater on top, it's pretty nice H->Block factory...
Nice clearance. The block-to-pi-to-Herschel conduit from the U.C. ladder doesn't quite fit here, but there are lots of other ways that this might come in handy. The tub-with-tail trick below is nice (though I don't know about really useful) because the trigger glider is harmlessly absorbed unless the block is present.
On a somewhat related note, the H->BG which is part of the H->G7 can function as a block keeper, and should work in a manner similar to what you described above:

Code: Select all

x = 117, y = 35, rule = LifeHistory
24.2A68.2A$24.2A68.2A7$32.2A68.2A$31.A.A67.A.A$32.A69.A3$7.2A68.2A$8.
A69.A$8.A.A67.A.A$9.2A68.2A3$32.2D68.2D$33.D69.D$31.2D68.2D$31.D69.D$
9.C35.2D32.C35.2C$9.C.C33.2D32.C.C33.2C$9.3C67.3C$11.C69.C3$26.2A68.
2A$26.2A68.2A$2.2A68.2A$3.A69.A$3A67.3A$A69.A!
-Matthias Merzenich

chris_c
Posts: 966
Joined: June 28th, 2014, 7:15 am

Re: The Hunting of the New Herschel Conduits

Post by chris_c » May 31st, 2015, 11:01 am

Sokwe wrote: I also made some minor corrections and did some reordering of the patterns that were already in the collection. The following patterns in the collection are actually composite and should perhaps be removed:

Code: Select all

b_b_04.rle
b_b_05.rle
b_pi_02.rle
b_pi_03.rle
I think I added the last 2 of these. The point is that the second half of the conduit is not quite a copy-and-paste from elsewhere in the collection. The trick is to get rid of the first block of the Herschel without using the standard eater. Perhaps it would be better to delete those 4 conduits and illustrate the trick just once with something like this:

Code: Select all

#C Possibly add as b_h_02b.rle ?
x = 42, y = 41, rule = LifeHistory
17.2A$16.B2A2B$17.4B$6.A11.4B3.3B$6.3A8.4B4.5B$9.A6.14B$8.2AB5.14B$8.
5B3.16B$10.5B.16B$10.22B$10.23B$9.24B4.3B$9.24B.7B$9.23B.B2A5B$8.25BA
2BA5B$6.4BC23B2A6B$5.6BC29B$5.6B2C17B.6B.2B$6.4B2C16B2.8B.B$7.3BC16B
4.6B.B2A$7.20B8.B3.A.A$7.21B12.2A$8.19B4.B$9.20B.B2A$9.8B3D11B2A$8.9B
D14B$7.9B3D5B2D6B$6.18B2D4B$5.22B.B$4.10B.12B$3.8B5.11B$2.8B9.7B$.8B
9.8B$8B9.8B$7B11.7B$6B13.5B$5B14.5B$4B13.2CB.2B$3B13.C.CB$2B14.C$B14.
2C!
Sokwe wrote:Thanks again to Chris for making the collection so easy to update.
No problem. Thanks for this latest update.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » May 31st, 2015, 12:00 pm

Sokwe wrote:On a somewhat related note, the H->BG which is part of the H->G7 can function as a block keeper, and should work in a manner similar to what you described above...
Thanks, yes, that one has more clearance, and because of the block-keeper function it can be SET multiple times by a Herschel before receiving the TEST glider signal. And if the block is not there, the glider passes cleanly through the circuit, and that signal can be caught and used elsewhere:

Code: Select all

x = 122, y = 97, rule = LifeHistory
119.A$119.A.A$119.2A18$99.A$99.A.A$99.2A25$16.A$16.3A5.2A$19.A4.2A$
18.2A30.A$13.A29.A6.3A$13.3A27.3A7.A$16.A29.A5.2A$15.2A28.2A$41.2A$
32.2A8.A$31.A.A8.A.A$32.A10.2A3$7.2A$8.A$8.A.A$9.2A5$88.2A$88.A$9.C
35.2D39.A.A$9.C.C33.2D39.2A$9.3C$11.C3$26.2A$26.2A32.2A$2.2A56.2A$3.A
$3A9.2A93.A$A11.2A34.2A56.A.A$49.A56.A.A$46.3A58.A$46.A49.D$50.2A44.D
.D$49.A.A4.2A27.2A9.3D$49.A6.2A14.2A11.2A11.D$48.2A23.A$70.3A$70.A2$
89.2A$90.A$87.3A$87.A!
All in all, that's about as clean a Spartan stop-and-go conduit as one could ask for. Won't set any records for recovery time, of course, but it's usually easy enough to design around that.

It's impressive that all six block-annihilating SW glider lanes are transparent to the block-keeper. I can't think of any use offhand for a resettable width-6 eater, but there might be a case where we want to suppress a leading glider and then let through a following glider on an adjacent lane. In that case, one of these is a lot cheaper than a highway robber.

User avatar
simsim314
Posts: 1823
Joined: February 10th, 2014, 1:27 pm

Re: The Hunting of the New Herschel Conduits

Post by simsim314 » May 31st, 2015, 2:49 pm

Here is nice and edgy H->boat that works with dvgrn construction above (not sure what is exactly the purpose of it):

Code: Select all

x = 104, y = 116, rule = B3/S23
13$87bobo$87b2o$88bo2$84bobo$84b2o$85bo$80bobo$80b2o$81bo2$77bobo$77b
2o$78bo36$45bo$44bobo$44bobo$45bo4$19bo$19b3o$22bo$21b2o5$30bo$28b3o$
28bobo$28bo7$28b2o$27bobo$27bo$26b2o5$39b2o$39b2o3$36b2o$35bobo$35bo$
34b2o!
EDIT Never saw this H eater:

Code: Select all

x = 9, y = 17, rule = LifeHistory
.2C.2C$C.C.C.C$.C3.C12$6.3C$7.C$5.3C!

User avatar
Kazyan
Posts: 1247
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Post by Kazyan » May 31st, 2015, 3:31 pm

If someone can figure out how to delete the honeybit pond, this would be a G11->R with potential as a G13->Pi.

EDIT: Figured it out--we just leave it there, basically!

Code: Select all

x = 36, y = 48, rule = LifeHistory
21.A$19.3A$10.A7.A$8.3A7.2A$7.A$7.2A2$4.2A10.2A7.2A$4.A.A8.A2.A6.A$5.
2A4.2C2.A2.A4.A.A$2A9.2C3.2A5.2A$.A$.A.A$2.2A3$27.2A$27.2A5$22.C$21.
2C$21.C.C21$34.C$33.2C$33.C.C!
Tanner Jacobi
Coldlander, a novel, available in paperback and as an ebook. Now on Amazon.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » May 31st, 2015, 5:11 pm

Kazyan wrote:If someone can figure out how to delete the honeybit pond, this would be a G11->R with potential as a G13->Pi.

EDIT: Figured it out--we just leave it there, basically!
Here's hoping that that turns out to be really useful for something -- but even if it doesn't, that's a really nice little Spartan circuit. Does anyone remember seeing a G11 tandem-glider output anywhere -- preferably an "in passing" output, so as not to have to use up a Herschel? I've gotten stuck sorting through the H-to-G collection, so haven't gotten very far on the H-to-G(n) stamp collection yet.

Here are Spartan G-to-3G and G-to-2G reflectors based on the above reaction. With any luck they can be made smaller somehow. Are there any new short fast parity-changing F conduits, or a workable R-to-H or B-to-H that produces the other color of Herschel?

Code: Select all

x = 238, y = 162, rule = B3/S23
60bo119bo$58b3o117b3o$49bo7bo111bo7bo$47b3o7b2o108b3o7b2o$46bo119bo$
46b2o118b2o2$43b2o10b2o7b2o97b2o10b2o7b2o$43bobo8bo2bo6bo98bobo8bo2bo
6bo$44b2o4b2o2bo2bo4bobo99b2o4b2o2bo2bo4bobo$39b2o9b2o3b2o5b2o95b2o9b
2o3b2o5b2o$40bo119bo$40bobo117bobo$41b2o118b2o3$45b2o19b2o97b2o19b2o$
44bo2bo18b2o96bo2bo18b2o$45b2o118b2o3$37b2o118b2o$37bobo117bobo$38bo
119bo7$33bob2o116bob2o$33b2obo116b2obo2$42b2o118b2o$42b2o118b2o15$32b
2o191b2o$33bo182bo8b2o$33bobo180b3o$34b2o62bo8b2o110bo$98b3o6b2o109b2o
$101bo$100b2o13b2o73b2o$115bo74bo$113bobo72bobo28bo$113b2o69b2o2b2o28b
obo$150b2o32b2o32bo2bo10b2o$106b2o42b2o67b2o11b2o$106b2o3$35b2o75bo42b
2o$35b2o14b2o58bobo41b2o$51bo60b2o37b2o$52b3o96b2o68b2o$54bo166b2o$
103b2o$103b2o52b2o31b2o$157b2o13b2o16bobo$171bobo18bo$48bo122bo20b2o$
46b3o121b2o4b2o$45bo129bobo$45b2o128bo$174b2o7b2o$183b2o4$27b2o$28bo$
28bobo$29b2o3$46b2o67b2o118b2o$29b2o15b2o67bobo117bobo$28bobo84bo119bo
$28bo$27b2o$173b2o$174bo$174bobo$46b2o127b2o$47bo$47bobo$48b2o4$29bob
2o$29b2obo2$38b2o151b2o$38b2o151b2o2$173b2o$172bobo$172bo$171b2o$176b
2o$177bo$174b3o$174bo5$28bo$28b3o$31bo$14bo15b2o$12b3o$11bo$11b2o3$23b
2o$23b2o4$43b2o$2o41b2o$obo$bo$44b2o$44bo$31b2o12b3o$32bo14bo$29b3o$
29bo9$4bo$3bobo$4b2o7b2o$13b2o2$4b2o$5bo$2b3o$2bo!
#C [[ AUTOSTART ZOOM 2 STEP 8 LOOP 1600 ]]

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » June 1st, 2015, 8:18 am

Sokwe wrote:I have updated the conduit collection...
The update includes everything I suggested in this post, most things posted in this topic after that post (except for H->G converters), and a few miscellaneous patterns that for some reason didn't make it into the original collection...
Thanks again to Chris for making the collection so easy to update.
Thanks to both of you, and everyone else who has been contributing! Transparent ships and loaves and boats, and lots of new objects for input and output and new ways to connect them together.

A nice trickle of new Spartan circuitry has been showing up, too. One of the big surprises that I just noticed this morning is the Spartan G2 in Sokwe's latest additions. It's the same lanes as a previous G2, but with better relative timing:

Code: Select all

x = 193, y = 152, rule = LifeHistory
99.3D2.D3.D2.3D2.5D2.3D3.3D3.3D15.3D2.D3.D2.3D2.5D2.3D3.3D3.3D$98.D3.
D.D3.D.D3.D3.D3.D3.D.D3.D.D3.D13.D3.D.D3.D.D3.D3.D3.D3.D.D3.D.D3.D$
98.D5.D3.D.D2.2D3.D7.D5.D.D2.2D13.D5.D3.D.D2.2D3.D7.D5.D.D2.2D$99.3D
2.D.D.D.D.D.D3.D5.2D5.D2.D.D.D14.3D2.D.D.D.D.D.D3.D6.D4.2D2.D.D.D$
102.D.D.D.D.2D2.D3.D7.D3.D3.2D2.D17.D.D.D.D.2D2.D3.D5.D7.D.2D2.D$98.D
3.D.2D.2D.D3.D3.D3.D3.D2.D4.D3.D13.D3.D.2D.2D.D3.D3.D4.D4.D3.D.D3.D$
99.3D2.D3.D2.3D4.D4.3D2.5D2.3D15.3D2.D3.D2.3D4.D3.5D2.3D3.3D4$99.3D2.
D3.D8.3D2.5D2.3D4.D16.3D2.D3.D8.3D2.5D2.3D4.D$98.D3.D.D3.D7.D3.D3.D3.
D3.D2.2D15.D3.D.D3.D7.D3.D3.D3.D3.D2.2D$98.D5.D3.D11.D3.D7.D3.D15.D5.
D3.D11.D3.D7.D3.D$99.3D2.D.D.D.5D4.D4.D6.D4.D16.3D2.D.D.D.5D4.D4.D6.D
4.D$102.D.D.D.D9.D5.D5.D5.D19.D.D.D.D9.D5.D5.D5.D$98.D3.D.2D.2D8.D6.D
4.D6.D15.D3.D.2D.2D8.D6.D4.D6.D$99.3D2.D3.D7.5D3.D3.5D2.3D15.3D2.D3.D
7.5D3.D3.5D2.3D6$120.2A$120.A3.2A39.A$108.2A12.A2.A39.3A$109.A11.4A
43.A$109.A.A55.2A$110.2A9.2A$121.A.A$106.2A14.2A$106.A.A67.2A$108.A
67.2A$108.2A$157.2A$158.A$123.2A33.A.A$123.A.A33.2A$103.2A20.A$94.2A
7.2A20.2A53.2A$95.A83.A.A$95.A.A82.A$96.2A$157.2A$89.2A66.2A$90.A$90.
A.A$91.2A5$126.2A$126.A$91.C35.3A21.2A$91.C.C35.A22.A$91.3C29.A28.A.A
$93.C28.A.A28.2A16.2A$121.A.A2.A44.2A$103.2A16.A.A.2A$103.2A2.2A9.2A.
A.A$107.A.A9.A.A.A$109.A7.A.A.2A48.2A$109.2A6.2A52.2A$153.C$153.C.C$
153.3C$155.C11$164.2A$164.2A2$22.E$20.3E$19.E$19.2E81.2A$101.A.A15.2A
$101.A8.2A6.A2.A$98.2A.2A7.2A7.2A$99.A$99.A$26.2E66.A.2A.2A$26.2E66.
2A.A$97.A$97.2A8$131.2A$132.A$111.2A19.A.A$10.2E99.2A20.A.A$10.2E122.
A3.2A$18.2E118.2A$9.3D6.E$10.D8.3E$8.3D10.E$126.2A11.D$126.2A9.3D$
137.D.D$137.D$114.2A32.E$115.A31.E.E$112.3A32.E.E$112.A35.E$116.2A$
10.2A103.A.A4.2A$10.2A103.A6.2A$114.2A2$.2A20.A$.2A19.A.A$22.A.A10.A$
23.A10.A.A$34.A.A$35.A$2.2A$3.A$3A3.2A$A4.A.A$5.A$4.2A11.2A$17.2A15.
2A$34.2A7$36.2A$34.A2.A$34.3A$31.A$31.4A$28.A5.A$28.4A$31.A$30.A$30.
2A!
The above ambidextrous G2 version of the new 135-degree receiver should probably be added to the conduit collection, since the G5 and G6 versions are there.

Might there still be a Spartan way to get another signal out of that H-to-G2? I notice there's a good R output, but so far I have to add a Snark to clean up the extra block.

Composite Lx256 with one-Snark cleanup:

Code: Select all

x = 52, y = 94, rule = LifeHistory
42.B2A$40.2BA2BAB$40.3B2A3B$38.12B$37.14B$38.13B$38.14B$37.15B$37.14B
$37.13B$37.B3D4B.3B$37.2BD4B$37.2B3D2B$37.6B$36.7B$35.8B$34.8B$33.9B$
32.4B.6B$31.4B.7B$30.4B2.6B$29.4B3.6B$28.4B4.6B$27.4B4.8B$12.A13.4B6.
8B$12.3A10.4B6.9B$15.A8.4B7.9B$14.2A7.4B8.10B$14.5B3.4B9.5B2A3B$16.3B
2.4B10.5B2A4B$6.2A7.9B11.11B$6.A8.8B12.4BD7BA.2A$3.2A.A.B3.10B15.B3D
4B2.2A.A$3.A2.3AB.2B2A7B15.D2B2D2B$4.2A2.BA3B2A7B15.6B$6.4A12B15.6B$
6.A.2B3.7B.B2A13.7B$7.3AB2.7B.BA.A7.B2.8B$10.A4.4B5.A6.3B.8B$5.5A5.4B
5.2A5.12B.BA$5.A10.4B10.14BA.A$7.A9.4B8.7B.7B.2A$6.2A10.4B7.13B$19.4B
6.9BD3B$20.4B5.8B3D3B$21.4B4.10BD2B$17.2A3.4B2.13B$18.A4.13B$18.A.AB
2.11B$19.2AB2.12B$21.16B$21.18B.2A$21.6B2D10BA.A$22.5B2D11BA$18.B.20B
$17.2AB.19B$17.2A19B$18.B.17B$20.16B$20.16B$20.15B$20.14B$20.13B$20.
12B$19.12B$19.11B$11.2A6.10B$12.A7.9B$12.A.AB4.9B.2B$13.2AB.14B2A$15.
16B2A$15.14B.2B$16.13B$15.14B.2B$13.18B2A$11.18B.B2A$11.2BC15B2.B$10.
3BCBC4B.8B$11.2B3C4B2.7B$10.5BC4B2.7B$9.10B3.7B$8.6B9.7B$7.6B10.7B$6.
6B11.6B$5.6B13.5B$4.6B13.5B$3.6B14.5B$2.6B16.3B$.6B16.4B$6B16.4B$5B
18.B2AB$4B20.2A$3B$2B!

User avatar
praosylen
Posts: 2443
Joined: September 13th, 2014, 5:36 pm
Location: Pembina University, Home of the Gliders
Contact:

Re: The Hunting of the New Herschel Conduits

Post by praosylen » June 1st, 2015, 9:53 am

Why is this simple pi-turning conduit not in the collection:

Code: Select all

x = 10, y = 10, rule = LifeHistory
2.2A$3.A$2.A$2.2A2$2.2A$2.A.A2.3C$3.A3.C.C$3A4.C.C$A!
?
I know it's known, as it appears in an oscillator on LifeWiki.
former username: A for Awesome
praosylen#5847 (Discord)

The only decision I made was made
of flowers, to jump universes to one of springtime in
a land of former winter, where no invisible walls stood,
or could stand for more than a few hours at most...

User avatar
praosylen
Posts: 2443
Joined: September 13th, 2014, 5:36 pm
Location: Pembina University, Home of the Gliders
Contact:

Re: The Hunting of the New Herschel Conduits

Post by praosylen » June 2nd, 2015, 10:07 am

Another BG-to-G that can't connect with anything:

Code: Select all

x = 23, y = 32, rule = LifeHistory
12.2C$13.C$11.C$11.5C$15.C$9.4CB$9.C2.CB$10.6B$3.C6.6B$3.3C3.8B$6.C2.
8B$5.2C3.7B3.B$5.12B2.4B$7.15B$3.18B$3.18B$.13B2C5B$.12BCBC4B$2.10B2C
4B$.17B$.15B$.15B$.15B$.15B$2.14B$5.13B$5.4B7.2C$3.5B8.C$3.2CB11.3C$
4.C14.C$.3C$.C!
former username: A for Awesome
praosylen#5847 (Discord)

The only decision I made was made
of flowers, to jump universes to one of springtime in
a land of former winter, where no invisible walls stood,
or could stand for more than a few hours at most...

User avatar
Extrementhusiast
Posts: 1966
Joined: June 16th, 2009, 11:24 pm
Location: USA

Re: The Hunting of the New Herschel Conduits

Post by Extrementhusiast » June 3rd, 2015, 5:20 pm

A nice seed for the two-eater combo (which actually works from three directions):

Code: Select all

x = 26, y = 17, rule = B3/S23
2b3o3$24b2o$24b2o$3o$2bo$bo2b3o6$11b2o3b2o$11b2o2bo2bo$15bobo$16bo!
A couple of other predecessors for it, which could be made into seeds:

Code: Select all

x = 56, y = 22, rule = B3/S23
4bo$5bo$5bo2$6b2o$5bo2bo$6b3o34b2o$7bo35b2o2$2o$2o$54b2o$54b2o$43b2o$
42bo2bob3o$43b2o2bo$47b2obo$50bo3$11b2o$11b2o!
The rightmost block is just there to suppress a pi.
I Like My Heisenburps! (and others)

simeks
Posts: 401
Joined: March 11th, 2015, 12:03 pm
Location: Sweden

Re: The Hunting of the New Herschel Conduits

Post by simeks » June 4th, 2015, 2:57 pm

Kazyan wrote here:
I recall seeing a catalyst like this somewhere on the forum before--Simkin, I think?
I happened to stuble upon this post by simsim314, so here is a link to it!

chris_c
Posts: 966
Joined: June 28th, 2014, 7:15 am

Re: The Hunting of the New Herschel Conduits

Post by chris_c » June 7th, 2015, 7:15 am

I did some investigation of glider pairs interacting with a small number of still lifes. I'm not sure if this thread is quite the place for such things but here are the results anyway:

Code: Select all

#C 90 degree reflector, 180 degree refelctor, duplicator, MWSS and two Herschels
x = 287, y = 39, rule = B3/S23
2o$bo$bobo56bo$2b2o55bobo$60bo65b2o51bo$126bo50b3o$124bobo49bo$124b2o
45b2o3b2o$59b2o110b2o$59b2o219bo$278b3o$4bo43b2o227bo$3bobo42b2o227b2o
$3bobo10b2o203b2o$4bo11bobo202b2o$18bo$18b2o154b3o3b2o89b2o$63b3o49b2o
57bo5bobo88b2o$63bo51b2o58bo4bo$12b3o49bo149b2o$12bo200bobo$13bo60bo
138bo$73b2o105bo31b2o$73bobo51b2o50b2o102b2o$127bobo49bobo101bobo$127b
o147b3o7bo$112b2o7b3o151bo9b2o$112b2o7bo154bo$122bo2$233b3o36b3o$19bo
213bo38bo$18b2o214bo38bo$18bobo3$244b2o$244bobo$244bo!

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » June 7th, 2015, 10:25 am

chris_c wrote:I did some investigation of glider pairs interacting with a small number of still lifes. I'm not sure if this thread is quite the place for such things but here are the results anyway...
Definitely some interesting stuff here. Maybe those synchronized salvos will all need a much-too-big engineered gun to produce them, but if a simple converter ever showed up that produced exactly one of these salvos, we'd be in business.

How extensive was the search that produced these results?

Does it seem vaguely possible that a similar search would turn up anything interesting, using only known easy-to-produce glider salvos as input? Probably the current options in the H-to-G(n) collection are sparse enough that the search space will be too small to find anything in (?).

The 3G-to-MWSS reaction made me think about G(n)-to-MWSS and G(n)-to-LWSS:

Code: Select all

x = 199, y = 109, rule = LifeHistory
111.C$C.C109.C$.2C107.3C$.C59$3.C109.C$4.C109.C$2.3C107.3C7$182.D$71.
D.D109.D$72.2D107.3D$72.D3$64.A109.A$64.3A107.3A$67.A109.A$66.2A108.
2A2$47.A109.A$47.3A107.3A$50.A109.A$37.A11.2A11.2A11.D71.A11.2A11.2A
11.D$35.3A24.2A9.3D69.3A24.2A9.3D$34.A38.D.D68.A38.D.D$34.2A37.D70.2A
37.D$19.2A108.2A$20.A66.2A41.A66.2A$20.A.2A63.2A41.A.2A63.2A$21.A2.A
23.D82.A2.A23.D$22.2A24.D.D81.2A24.D.D$37.2A9.3D96.2A9.3D$37.2A11.D
22.2A72.2A11.D22.2A$72.A.A107.A.A$72.A109.A$71.2A108.2A$46.A3.2A104.A
3.2A$45.A.A3.A103.A.A3.A$44.A.A3.A12.2A89.A.A3.A12.2A$40.2A.A.A3.A13.
A.A84.2A.A.A3.A13.A.A$40.2A.A2.4A.A13.A84.2A.A2.4A.A13.A$44.A.A3.A.A
12.2A87.A.A3.A.A12.2A$40.2A.2A2.A2.A.A97.2A.2A2.A2.A.A$41.A.A2.2A3.A
99.A.A2.2A3.A$29.2A10.A.A95.2A10.A.A$29.2A11.A96.2A11.A!
I haven't worked out what the smallest Fx conduit is, that can replace the Fx77 shown here, to make an adjustable G10+2N or G11+2N circuit produce exactly the right synchronized salvo to make an LWSS or MWSS. And of course the same trick could be done with pure Spartan circuitry, at the usual cost in increased size and slower recovery time.

I'm still most interested in this line of research, in terms of finding 2G salvos that are cheaper to build than these 3G salvos, that can be aimed at an invariant turner to produce 90-degree gliders for an armless universal constructor. Of this new collection, only the beehive+2eater invariant turner seems suitable for this purpose, unless we add another eater to the 2block+eater to suppress the second glider output.

I suppose the beehive+2eater will be significantly cheaper to build at a distance (with either slow salvos or LWSS+G collisions) than a loaf2GtoG. So maybe it's worth looking into the cheapest Spartan way to produce that synchronized G7 tandem glider.

simeks
Posts: 401
Joined: March 11th, 2015, 12:03 pm
Location: Sweden

Re: The Hunting of the New Herschel Conduits

Post by simeks » June 7th, 2015, 6:05 pm

A new R->G:

Code: Select all

x = 22, y = 21, rule = LifeHistory
9.2C$8.B2CB$8.3B$6.7B6.2C$5.9B5.C$6.8B2.BC.C$6.9B.B2C$6.2BE8B$6.2B2E
7B$6.B2E7B$6.9B$7.10B$7.7B.2C.C$6.7B2.2CB3C$5.4B7.B4.C$4.4B7.2C.3C$3.
4B9.C.C$2.4B10.C.C$.D3B12.C$D3B$3D!
It fits with the most basic of the H->R:s to create a new H->G7. Also shown a new H->G11 preceeded by conduit 1 to show something can connect to it:

Code: Select all

x = 195, y = 102, rule = LifeHistory
179.C$178.C.C9.C$178.C.C7.3C$176.3C.2C5.C$175.C4.B6.2C$176.3CB2CB2.4B
$178.C.2CB.3B$182.6B$181.9B$180.11B$180.11B.B$180.12B2C$180.5BD6B2C$
179.5B3D6B$178.6BD6B$177.5B.8B$53.C122.14B$52.C.C9.C110.14B$52.C.C7.
3C109.14B$50.3C.2C5.C111.16B$49.C4.B6.2C109.18B$50.3CB2CB2.4B108.4B.
17B$52.C.2CB.3B109.4B3.15B2A$48.A7.6B107.4B4.15B2A$48.3A4.9B104.4B4.
17B$39.A11.A2.11B102.4B2.16B$39.3A8.2A2.11B.B79.2A18.4B2.16B$42.A7.
16B2C78.A.A16.4B3.16B$41.2A3.B5.7BD6B2C80.A15.4B3.16B$41.8B2.7B3D6B
81.2A13.4B3.20B$43.15BD6B84.B12.4B4.20B$43.22B84.3B9.4B6.20B$42.19B2.
B84.6B6.4B6.22B$40.21B86.10B2.4B5.24B$38.24B85.16B2.26B$38.2BE14B4.4B
82.2BD3B2A15BE13B.10B$37.3BEBE11B6.4B81.2B2D2B2A15BEBE4B.7B2.2B4.2A$
38.2B3E7B.5B5.4B81.2B2D18B3E4B2.3B.B4.B4.A$37.5BE5B6.2A6.4B81.BD21BE
4B4.2B2A3.2A4.3A$36.11B7.A8.4B80.D26B4.BA2.A4.A6.A$35.4B3.4B9.3A6.4B
81.13B.4B12.3A5.A.A$34.4B3.4B12.A7.4B80.7B.B4.4B22.2A$33.4B3.4B22.4B
80.4B7.4B14.3A$32.4B3.4B24.4B78.4B7.4B15.A2.A$31.4B3.4B26.4B76.4B7.4B
18.2A$30.4B3.4B28.4B74.BD2B7.4B$29.4B3.4B30.4B72.2D2B7.4B$28.4B3.DBDB
32.4B72.2D7.4B$27.4B4.2DB34.4B79.4B$26.4B6.D36.4B77.4B$25.4B45.4B75.
4B$24.4B47.4B73.4B$23.4B49.4B71.4B$22.4B51.4B69.4B$21.4B53.4B67.4B$
20.4B55.4B65.4B$19.4B57.4B63.4B$18.4B59.4B61.4B$17.4B61.4B59.4B$16.4B
63.4B57.4B$15.4B65.4B55.4B$14.4B67.4B53.4B$13.4B69.BDBD51.4B$12.4B71.
B2D50.4B$11.4B73.D50.4B$10.4B124.4B$9.4B124.4B$8.4B124.4B$7.4B124.4B$
6.4B124.4B$5.4B124.4B$4.4B124.4B$3.4B124.4B$2.4B124.4B$.D3B124.4B$D3B
124.4B$3D124.4B$126.4B$125.4B$124.4B$123.4B$122.4B$121.4B$120.4B$119.
4B$118.4B$117.4B$116.4B$115.4B$114.4B$113.4B$112.4B$111.4B$110.4B$
109.4B$108.4B$107.4B$106.4B$105.4B$104.D3B$103.D3B$103.3D!

chris_c
Posts: 966
Joined: June 28th, 2014, 7:15 am

Re: The Hunting of the New Herschel Conduits

Post by chris_c » June 8th, 2015, 6:30 am

dvgrn wrote: How extensive was the search that produced these results?
The gliders had to be within 8 lanes of each other and 40 ticks. The bait catalyst was a block or a beehive and two non-transparent catalysts were allowed from block, eater, boat and tub.
dvgrn wrote: Does it seem vaguely possible that a similar search would turn up anything interesting, using only known easy-to-produce glider salvos as input? Probably the current options in the H-to-G(n) collection are sparse enough that the search space will be too small to find anything in (?).
Vaguely possible I suppose. But in the realm of Herschel conduits we are already pretty well served for glider-pair->G and glider-pair->H so it would seem difficult to beat the technology that already exists.

User avatar
Kazyan
Posts: 1247
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Post by Kazyan » June 8th, 2015, 12:53 pm

This produces an explosion so comically large it's practically a methuselah--looks like a good candidate for bait restoration attempts for that Spartan G-to-X dream.

Code: Select all

x = 35, y = 32, rule = B3/S23
2bo$o$4bo$2bo$6bo$4bo$8bo$6bo$10bo$8bo$12bo$10bo$14bo$12bo$16bo$14bo$
18bo$16bobo$17b2o4$28bo4b2o$27bobo3b2o$22b2o3bobo$22b2o4bo3$31b2o$31bo
$32b3o$34bo!
Tanner Jacobi
Coldlander, a novel, available in paperback and as an ebook. Now on Amazon.

User avatar
dvgrn
Moderator
Posts: 10610
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » June 8th, 2015, 2:58 pm

Kazyan wrote:This produces an explosion so comically large it's practically a methuselah--looks like a good candidate for bait restoration attempts for that Spartan G-to-X dream.
That does look nice. There's room for a dozen catalysts around the edges of that reaction, even before the explosion gets unmanageably large.

I've set up an old-style catgl search -- three catalysts, reacting between T=70 and T=120 -- with a target pattern so that I'll know with a simple text search whether the bait block ever shows up again. Might take a few days to finish running that.

Of course that won't even scratch the surface of the tip of the iceberg for this search space. For example, if you add a boat catalyst to extract a B output signal from the obvious R-pentomino, early in the reaction, that will radically change the later evolution of the pattern... it might be worth running a separate catgl search starting with that added boat:

Code: Select all

x = 43, y = 49, rule = LifeHistory
2.A$A$4.A$2.A$6.A$4.A$8.A$6.A$10.A$8.A$12.A$10.A$14.A$12.A$16.A$14.A$
18.A$16.A$20.A$18.A$22.A$20.A$24.A$22.A$26.A$24.A.A$25.2A4$36.A4.2A$
35.A.A3.2A$30.2A3.A.A$30.2A4.A2$13.2A$13.A.A23.2A$14.A24.A$40.3A$42.A
4$18.2D2.D$19.3D$17.2A.D$18.A$15.3A$15.A!
However, I kind of suspect that with the active reaction moved that far north, it will be much less likely for a block to get thrown back to exactly the right spot. More likely that southward R will have to be used up to re-produce the block, and an output signal will have to come from somewhere else. Further bulletins as events warrant...!

Post Reply