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
Kazyan
Posts: 887
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Post by Kazyan » February 20th, 2015, 3:10 pm

dvgrn wrote:Hmm. Adding a block here extracts a glider, and protects the catalyst from immediate destruction, and gets a Herschel out to the north. But it's probably still a bit too messy to be a very likely candidate for a complete cleanup:
Hmm. Maybe run the south area's late evolution through catgl just in case; that's where the troublesome debris shows up.
Tanner Jacobi

User avatar
Extrementhusiast
Posts: 1801
Joined: June 16th, 2009, 11:24 pm
Location: USA

Re: The Hunting of the New Herschel Conduits

Post by Extrementhusiast » February 20th, 2015, 8:11 pm

A possibly known H-to-glider:

Code: Select all

x = 25, y = 31, rule = B3/S23
10bo$10b3o$13bo$12b2o$23b2o$23bo$21bobo$21b2o4$2o$bo$bobo17b2obo$2b2o
17bob2o7$2bo$2bobo$2b3o$4bo2$14b2o$14b2o2b2o$18bobo$20bo$20b2o!
An R-to-glider:

Code: Select all

x = 19, y = 18, rule = B3/S23
4b2o$4b2o5$17b2o$17bo$15bobo$15b2o2$2o$b2o$bo3$7b2o$7b2o!
I Like My Heisenburps! (and others)

User avatar
dvgrn
Moderator
Posts: 5978
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » February 21st, 2015, 2:09 am

Extrementhusiast wrote:A possibly known H-to-glider...
That looks like a new Spartan NW15 (EDIT: using NW newlanes here now). NW8 and NW17 are the nearest ones in the old collection. I haven't gone back and hunted through the last several years of forum postings to collect all the H-to-G's yet, though -- it's possible Guam or someone found a NW15. Someone should start a new thread to get these things organized...!
Extrementhusiast wrote:An R-to-glider...

Hm, I think that one's new too. It combines with one of the known H-to-R to make a passable H-to-G7; we already have one of those (and not much use for it that I know of) but it looks quite different. New one on the left, old one on the right, both with a F117 conduit leading in:

Code: Select all

#C [[ VIEWONLY HEIGHT 240 ZOOM 3.5 THUMBNAIL ]]
x = 178, y = 59, rule = LifeHistory
61.A$61.3A$64.A$63.2A$63.6B$65.5B$63.7B$62.8B$63.6B$62.7B.2B$58.3B2.
8B2A60.4B3.4B$56.15B2A61.4B3.4B$56.13B.2B63.4B3.4B$55.2A13B66.4B3.4B$
55.2AB.10B.2B65.4B3.4B$56.B2.15B64.4B3.4B$59.8BD6B65.4B3.4B5.2A$60.6B
3D5B66.4B3.4B3.B2AB$62.4BD7B67.4B3.4B2.3B$62.11B69.4B3.4B2.B.B$62.9B
72.4B3.8B$60.11B73.4B3.7B$60.10B75.4B2.7B$59.12B75.4B.7B$58.14B75.12B
4.2A$57.18B73.11B3.A.A$56.18B2A71.13B2.BA$55.19B2A70.17B$54.21B71.17B
$53.18B67.2A6.18B.3B$30.2A20.18B50.2A17.A6.23B$31.A19.19B51.A17.A.AB
4.23B$18.A11.A19.19B39.A11.A19.2AB.27B$18.3A9.2A17.23B36.3A9.2A20.30B
$21.A9.B16.24B39.A9.B20.31B$8.2A10.2A9.3B13.4B2.20B25.2A10.2A9.3B19.
32B$9.A10.5B5.6B10.4B2.22B25.A10.5B5.6B16.34B$9.A.AB9.4B3.10B6.4B.24B
25.A.AB9.4B3.10B11.36B$10.2AB.3B4.6B2.11B3.30B27.2AB.3B4.6B2.11B8.39B
$12.7B.13B2A15BD13B.10B27.7B.13B2A15BD35B2A$12.21B2A15BDBD4B.7B2.2B4.
2A27.21B2A15BDBD4B.28B2A$13.37B3D4B2.3B.B4.B4.A29.37B3D4B.29B$12.40BD
4B4.2B2A3.2A4.3A25.40BD4B.8B3.15B$10.46B4.BA2.A4.A6.A23.46B.9B4.3B.9B
$8.18B.B3.13B.4B12.3A5.A.A26.18B.B3.13B.4B6.10B.2B3.B3.7B$8.2BC13B7.
10B3.4B22.2A26.2BC13B7.10B3.4B5.14B2A$7.3BCBC4B.6B12.4B3.4B14.3A33.3B
CBC4B.6B19.4B5.15B2A$8.2B3C4B2.B.5B9.4B3.4B15.A2.A33.2B3C4B2.B.5B17.
2A6.17B$7.5BC4B7.2A8.4B3.4B18.2A32.5BC4B7.2A18.A7.13B$6.10B8.A8.4B3.
4B52.10B8.A16.3A11.10B$5.4B16.3A4.4B3.4B52.4B16.3A13.A13.11B$4.4B19.A
3.4B3.4B52.4B19.A27.10B$3.B2AB23.4B3.4B52.B2AB49.8B$2.3BA23.4B3.4B52.
3BA53.3B$.B3A23.4B3.4B52.B3A55.4B$2BAB23.4B3.4B52.2BAB58.2A$26.4B3.4B
115.A$25.4B3.4B117.3A$24.4B3.4B120.A!

Sphenocorona
Posts: 481
Joined: April 9th, 2013, 11:03 pm

Re: The Hunting of the New Herschel Conduits

Post by Sphenocorona » February 21st, 2015, 6:01 am

Is this R-to-Pi known?

Code: Select all

#C [[ THUMBNAIL AUTOSTART LOOP 40 GPS 10 ]]
x = 16, y = 17, rule = LifeHistory
3.2A$3.2A2$3.4A3.2A$3.A2.AB.B2AB$4.3B2.B.2B$2.B2.7B$.12B$13B3D$B2C12B
D$2B2C9B3D$2BC11B$5B.4B$10B$8B$6B$.4B!

chris_c
Posts: 921
Joined: June 28th, 2014, 7:15 am

Re: The Hunting of the New Herschel Conduits

Post by chris_c » February 21st, 2015, 11:14 am

dvgrn wrote: Unfortunately this new variant isn't quite Spartan either, or it would be quite handy in self-constructing circuitry.
Here is a 6 SL seed for the eater on eater:

Code: Select all

x = 23, y = 18, rule = B3/S23
11b2o$11b2o5$2o$2o4bo4b2o$5bobo2bo2bo$5b2o4b2o$21b2o$21b2o$8b2o$8bobo$
9bo$b3o$3bo$2bo!
It was found with the help of Catagolue http://catagolue.appspot.com/object?apg ... rule=b3s23

User avatar
dvgrn
Moderator
Posts: 5978
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » February 21st, 2015, 12:25 pm

Sphenocorona wrote:Is this R-to-Pi known?
No, but it doesn't quite look like a usable conduit -- there are only a dozen ticks to do anything with the pi before the block catalyst is irretrievably threatened. Might get an R-to-G out of it somehow, but I don't think the pi output can be attached to any known components.

Without the extra block, the table performs the standard R-to-B conversion. I seem to recall the table catalysis has been seen before, though it's not in Calcyman's conduit collection -- I've added it here:

Code: Select all

#C [[ VIEWONLY THUMBNAIL X 0 Y 0 ZOOM 4.62 HEIGHT 240 ]]
x = 135, y = 43, rule = LifeHistory
41.5D29.5D$41.D4.D28.D4.D$41.D5.D15.D11.D5.D$41.D5.D16.D10.D5.D$41.D
5.D17.D9.D5.D$41.D4.D19.D8.D4.D$41.5D9.13D7.5D$41.D2.D21.D8.D4.D$41.D
3.D19.D9.D5.D$41.D3.D18.D10.D5.D$41.D4.D16.D11.D5.D$41.D4.D28.D4.D$
41.D5.D27.5D7$112.B$111.2B$110.3B$109.4B$7.D29.D29.D29.D10.4B15.D$6.
3D27.3D27.3D27.3D8.4B15.3D$5.2D2BD25.2D2BD25.2D2BD4.2A19.2D2BD6.4B15.
2DB.D$4.6B24.6B24.6B4.A19.6B5.4B15.4B.B$4.6B24.6B24.6B.BA.A19.6B4.4B
16.6B$3.7B23.7B23.7B.B2A19.7B3.4B16.7B$4.8B22.8B22.8B22.12B18.8B$4.8B
22.8B22.7B23.11B19.8B$4.9B18.AB.9B21.7B23.10B18.B.9B$3.10B17.A.A10B
20.7B4.B18.10B.B16.2A10B$4.10B16.2A.7B.3B20.6B.B.2BA18.8B.2BA15.A.11B
$3.12B20.10B18.11BA.A16.11BA.A11.2A.A.B.10B$3.5BD6B20.3BD6B18.4B.D4B.
BA17.5BD4B.BA12.2A.2A2.3BD6B$3.4B3D5B19.3B3D5B19.3B3D3B20.4B3D3B21.3B
3D5B$3.4BD7B20.2BD7B20.2BD5B22.2BD4B23.2BD7B$2.12B22.8B22.6B24.6B24.
8B$.2A2B96.B2A$A2BAB96.BA.A$ABAB100.A$.A2B100.2A!

User avatar
dvgrn
Moderator
Posts: 5978
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » February 21st, 2015, 1:51 pm

chris_c wrote:
dvgrn wrote: Unfortunately this new variant isn't quite Spartan either, or it would be quite handy in self-constructing circuitry.
Here is a 6 SL seed for the eater on eater...
That's a pretty nice one! For this purpose the rightmost block can usually be omitted, since it only suppresses a block that could be won't get in the way of the catalyst.

It's also convenient that this is the same still life needed for the old F116 conduit. In Spartan constructible circuitry so far, only the F116 variant followed by F166 or Lx200 has been allowable. This is unfortunate, because circuits are generally useful in inverse proportion to their size and timing. That means that in your average unconstrained Hersrch search from (0,0,0) to big (X,Y,T), F116s are part of the best solutions a lot -- probably more often than anything besides Fx77.

That said, even a 6sL seed is fairly expensive... especially since you have to account for construction direction, and three quarters of the time it will be necessary to build an extra boat-or-something to bounce a final glider in to trigger the seed. Also the eater-on-eater will usually have to be constructed first, which causes all kinds of headaches for some orientations. For the H-to-G6, the nearby eater will have to be built after the eater-on-eater seed is triggered, and the same is (just barely) true for the nearby block in the F116.

We're pretty close to having workable scripts for "compiling" any regular Spartan pattern into either a one-arm or a two-arm slow-salvo construction recipe. None of the pieces are much bigger or more difficult to build than any other piece, so we can always work incrementally from the most distant still lifes back toward the shotgun(s).

Throwing in eater-on-eaters or spark coils or tub-with-tail eaters or eater2s or any other larger-but-incredibly-useful objects into constructible circuitry means that, well, somebody has to do a lot more coding.

And it's not really clear what the payoff will be. Circuitry with these catalysts in it will be smaller and more efficient (lower repeat time) than the pure-Spartan version of the same circuit. But so far it's not really clear that it will be enough smaller. It's possible that the extra cost of the Spartan++ objects will make the final construction recipe longer and slower overall.

(I don't really believe that -- it's probably significantly more efficient for (most) self-constructing circuity to use some Spartan++ objects... maybe even a Snark here or there, for all I know. But it will probably take an order of magnitude longer to get an automated compiler working with such objects. Same with compiling Spartan++ circuitry recipes by hand, which is how I've been doing all the Spartan Geminoid self-constructing patterns so far.)

To make a given object really painless to use in self-constructing circuitry, it will be necessary to collect a large set of construction recipes for it, similar to what was done for the fishhook eater. For each orientation we'll need a recipe that builds the object at the back edge of its reaction envelope, and one on the left edge, and one on the right edge. Once we've collected all of those recipes for some non-Spartan object, it becomes just as good as any Spartan object (except for the extra cost in gliders.)

But even with Guam's work on tub-with-tail eaters and spark coils (links above) we're not anywhere near there yet.

User avatar
Kazyan
Posts: 887
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Post by Kazyan » February 21st, 2015, 5:50 pm

Why have I not seen these converters before?

Code: Select all

x = 18, y = 33, rule = B3/S23
16b2o$16bo$14bobo$5b2o7b2o$5bobo$6b2o$2b2o$bobo$bo$2o$5b2o4bo$5b2o3b3o
$9b2obo11$5b2o$5bobo$6b2o$2b2o$bobo$bo$2o$5b2o4bo$5b2o3b3o$9b2obo!
Tanner Jacobi

User avatar
dvgrn
Moderator
Posts: 5978
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » February 21st, 2015, 9:50 pm

Kazyan wrote:Why have I not seen these converters before?
Wow! Apparently, because nobody has ever done the search you just did for ways to use the ship+eater catalyst.

It's not terribly easy to connect the inputs and outputs in the B-to-R, but it can certainly be done. Here's a very nice L200, which I assume I would know about if it had already been invented. We have an L198 and an Lx200, but this is something new and different:

Code: Select all

x = 49, y = 61, rule = LifeHistory
34.2A$33.A2.A$34.2A4$33.8B$33.9B$33.9B$33.9B$34.4B3DB$35.4BD2B$35.2B
3D2B$36.6B$36.7B$36.8B$37.8B$37.9B$36.6B.3B$36.7B.2B$37.6B2.B$37.6B$
37.6B$36.8B$35.8B$35.9B$35.9B$34.10B$34.3B2A5B$33.4B2A5B$33.11B$28.2A
.A7BD4B$28.A.2A2.4B3D2B$35.2B2D2BD4.2A$36.6B4.A$36.6B.BA.A$35.7B.B2A$
30.A5.8B$30.3A3.7B$7.2A24.A2.7B$8.A13.2A6.3A2.7B4.B$8.A.AB9.B2A2B4.A
5.6B.B.2BA$9.2AB.3B6.4B.3B5.11BA.A$11.7B2.11B4.5BD4B.BA$11.22B.B.3B3D
3B$12.27BD5B$11.25B2D7B$9.28B2D6B$7.29B2D6B$7.2BC26BD7B$6.3BCBC4B.24B
.3B$7.2B3C4B2.16B2.B.B3.B2A$6.5BC4B2.7B.4B6.B2AB2.BA.A$5.10B4.6B2.2B
8.2A4.B2A$4.4B11.6B16.4B$4.3B13.3B18.2A$2.4B12.7B17.A$2.2A14.2A.B.2A
14.3A$3.A15.A3.A15.A$3A13.3A5.3A$A15.A9.A!
#C [[ VIEWONLY THUMBNAIL ]]
Very close to Spartan, too, though the hat and the ship+eater might be a little tricky in some orientations. There's never been any particular reason to slow-construct a hat before (or a ship, either, unless I'm forgetting something).

The H-to-G is new also, of course -- it's a little harder to hook up than the H-to-R, actually, since the output glider means that the above hat trick doesn't work. But there are options:

Code: Select all

#C new H-to-G, path 56
x = 57, y = 38, rule = LifeHistory
40.B$40.2B$7.2A31.3B$8.A14.2A15.4B$8.A.AB10.B2AB15.4B$9.2AB.2B7.4B6.
4B6.4B$11.5B6.4B5.6B6.4B9.A$5.2A4.7B4.5B4.8B5.4B6.3A$5.A5.7B3.7B.12B
4.4B4.A$6.3A2.7B.17B.4B.B2.5B.B.2A$8.A24BD10BD10B$10.23B2D10BD7B$11.
21B2D11B2D6B$11.33B2D6B$2.2A3.2A.B.19B.12BD7B$2.A.A.A.A23B.B2.13B.3B$
4.A.A.B.21B5.2B2.2B2.B.B3.B2A$3.2A.2A22B7.AB5.B2AB2.BA.A$7.B.21B6.A.A
6.2A4.B2A$9.19B8.2A11.4B$10.13B2.2B22.2A$11.11B28.A$8.14B25.3A$7.15B
25.A$7.14B$7.12B$7.B3C4B.2B$7.2BC4B$7.2B3C2B$7.6B$6.7B$5.4B.B$4.4B$4.
3B$2.2AB$.A.AB$.A$2A!
#C [[ VIEWONLY THUMBNAIL ]]
Maybe the nicest thing about this converter is that it hooks up nicely to a Herschel receiver -- a near-Spartan circuit again, this time made slightly awkward by a spark coil instead of a hat.

So we at last have a way to turn a tandem glider 90 degrees, without mirror-reflecting it, that's reasonably compact and has the best possible compression rate of 117 ticks. The L200 itself recovers in as little as 59 ticks if you suppress the output Herschel's glider; that's very nearly as good as any Herschel conduit can manage.

Time for some congratulatory fireworks, I think!

Code: Select all

#C L200 + Herschel transceiver emu loop
#C [[ AUTOSTART LOOP 5616 ZOOM 3.6 THEME 2 ANGLE 45 ]]
x = 168, y = 168, rule = B3/S23
89b2o$88bobo$76b2o3b2o5bo$76bobobobo4b2o$78bobo$76bobobobo$76b2o3b2o7$
81b3o$82bo$82b3o7b2o$92bobo$94bo$94b2o13$83b2o$83b2o22b2o$107b2o$90bob
2o$90b2obo$74b2o33b2o$75bo33b2o$75bobo44b2o$76b2o44bo$120bobo$120b2o$
42b2o80b2o$42b2o67b2o11b2o$38bo35bo35b2o3bo$38b3o32bobo33bo6bo$41bo32b
o20b2o11bo3bo3b7o$40b2o11bo40b2obo11bo2bobob4o$52bobo7b2o30bo2bo12b6ob
obo3bo$52bobo7b2o25b3ob2ob2o17b2o4b2o3b2o$53bo35bo2b2o2bo19bo8bo2bo$
89bo5bo29bobo$90b2o2bo24bo6bo$92b2o24b2o$77b2o33bo5bo$50b2o24bobo4b2o
7bo19bo$50b2o24b2o5b2o6bobo18bo$36b2o54b2o$36b2o41b2o$32b2o45bo48b3o$
32b2o46b3o45bo$82bo46bo$118b2o$118b2o5$133b2o$132b2o$134bo2$18bo$16b3o
103bo$15bo18b2o20b2o63bobo7bo$15b2o17bo20bobo64bo6b3o$35bo20bo54b2o15b
o32b2ob2o$2o32b2o75bobo14b2o31bo3bo$o111b2o48b3o$b3o104b2o$3bo103bobo
52b3o$107bo46bo6bo3bo$106b2o44b3o6b2ob2o$15bo15b2o22b2o54b2o22b2o15bob
o$13bobo15b2o22b2o54b2o22b2o15bo$2b2ob2o6b3o44b2o$2bo3bo6bo46bo$3b3o
52bobo103bo$58b2o104b3o$3b3o48b2o60b3o48bo$2bo3bo31b2o14bobo58bo2bo13b
2o32b2o$2b2ob2o32bo15b2o54bo3bo2bo13bo$36b3o6bo64bobobo2bo15bo17b2o$
36bo7bobo63b2o2bo2b2o13b2o18bo$45bo69bo2b3o28b3o$116bo3b2o27bo$117b2o
2bo$33bo84b3o$34b2o$33b2o5$48b2o$48b2o$38bo46bo$39bo45b3o46b2o$37b3o
48bo32bo12b2o$87b2o31bobo7b2o$74b2o43bo3bo6b2o$55bo18bobo6b2o5b2o27bo
3b2o$55bo19bo7b2o4bobo19b3o5b3o2bo$49bo5bo33b2o28bo$48b2o24b2o43b2o$
41bo6bo24bo2b2o40b2o3bo$40bobo29bo5bo38b2ob3o$39bo2bo8bo19bo2b2o2bo40b
3o$40b2o3b2o4b2o17b2ob2ob3o25b2o7b2o5b2o$44bo3bobob6o12bo2bo30b2o8b2o
3b3o$48b4obobo2bo11bob2o47bo4b2o$45b7o3bo3bo11b2o20bo24bo2bo4bo$51bo6b
o33bobo23bo2bo5b3o$52bo3b2o35bo25bo9bo$42b2o11b2o67b2o$42b2o72b2o6b2o$
46b2o67bo2bo$45bobo68bobo$45bo44b2o15b2o8bo$44b2o44bobo13bobo$57b2o33b
o15bo$57b2o33b2o$74bob2o20bo$74b2obo20b2o$59b2o36bobo$59b2o22b2o$83b2o
13$72b2o$73bo$73bobo$74b2o7b3o$85bo$84b3o7$85b2o3b2o$85bobobobo$87bobo
$79b2o4bobobobo$79bo5b2o3b2o$77bobo$77b2o!

User avatar
dvgrn
Moderator
Posts: 5978
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » February 22nd, 2015, 12:37 am

I threw together one more sample pattern using the L200... well, technically when it starts with a Herschel transceiver it's an L432+4N conduit.

With a F117+R64 added, the repeat time rises to 156 ticks -- still not too shabby. You end up with decent adjustable p1282+8N loops, which can build any gun you want down to p156 by adding more signals in the loop, with the right adjustment:

Code: Select all

#C p1282+8N loop, with attached Snark-assisted Herschel output
#C [[ THUMBNAIL AUTOSTART THEME 3 ZOOM 3.1 ANGLE 270 STOP 2564 ]]
x = 144, y = 200, rule = B3/S23
5b2o$5bobo$7bo5b2o3b2o$7b2o4bobobobo$15bobo$13bobobobo$13b2o3b2o2$74bo
20bo$72b3o11bo7bobo$71bo14b3o6bo$71b2o16bo15b2o$88b2o14bobo$48bo55b2o$
34bo13b3o57b2o$2b2o28b3o16bo56bobo$bobo27bo18b2o58bo$bo29b2o77b2o$2o
79b2o22b2o$81b2o22b2o5$41b2o$41b2o41b2o$64b2o19bo20bo$64bobo17bo20bobo
$66bo17b2o20b2o$4b2o48b2o10b2o$4b2o48bo$10b2o32b2o9b3o$10b2o33bo11bo$
42b3o$42bo$8b2o$8b2o5b2o$15b2o3$50b2o$50b2o$55bo26b2o$53b3o26b2o$52bo
33b2o$40bo11b2o32b2o$30b2o7bobo58b2o$30b2o7bobo58b2o$40bo4$103bo$42b2o
58bobo7b2o$42b2o58bobo7b2o$56b2o32b2o11bo$56b2o33bo$60b2o26b3o$60b2o
26bo$92b2o$92b2o3$127b2o$127b2o5b2o$134b2o$101bo$99b3o$86bo11bo33b2o$
86b3o9b2o20b2o$89bo30b2o8bo2bo4b2o$76b2o10b2o35b2o3bo7b2o$36b2o20b2o
17bo47bob4ob2o$36bobo20bo17bobo46b2o4b2o$37bo20bo19b2o47bo4b2o$58b2o
41b2o25bo2bo$101b2o26b3o3$114bo$114bobo$37b2o22b2o15bo35b2o$37b2o22b2o
15bobo61b2o$32b2o44b3o61bo$33bo46bo11b2o46bobo$33bobo56bo47b2o$34b2o
57b3o$38b2o55bo$37bobo14b2o$37b2o15bo$48bo6b3o$47bobo7bo$48bo2$124b2o
3b2o$124bobobobo$126bobo$124bobobobo4b2o$124b2o3b2o5bo$136bobo$137b2o
14$46bo$44b2o$45b2o6$23bo$23b3o$26bo$25b2o3$14b2ob2o$14b2obo$17bo$17b
3o4b2o$15b2o3bo3b2o$14bo2b4o$14b2obo15b2o$15bo2b3o12bobo$15bo5bo13bo$
16b5o14b2o$18bo9$55b2o$55b2o$60bo$58b3o$57bo$57b2o$35b2o$35b2o6$47b2o$
47b2o$61b2o$61b2o$65b2o$65b2o14$41b2o20b2o$41bobo20bo$42bo20bo$63b2o2$
94bo$93bobo$93bobo$94bo$42b2o22b2o$42b2o22b2o$37b2o$38bo$38bobo$39b2o$
43b2o$42bobo14b2o$42b2o15bo$53bo6b3o$52bobo7bo$53bo!
There's nothing new about the ability to build Herschel transceiver loops down to p156 or even p117 -- just that some interesting new geometries are available now.

When a new conduit shows up, I always hope that it will allow a smaller or faster Snark-assisted G-to-H, but there's no hope of that here since there aren't any extra glider outputs. Still lots of other things to try, though. I definitely haven't even looked at the rest of Calcyman's component collection (which is starting to need some serious updating) to see what else these converters might be attached to -- I know there are at least a few more workable options out there, but I think it's somebody else's turn...!

Sokwe
Moderator
Posts: 1488
Joined: July 9th, 2009, 2:44 pm

Re: The Hunting of the New Herschel Conduits

Post by Sokwe » February 22nd, 2015, 12:56 am

Kazyan wrote:Why have I not seen these converters before?
It might be partly because there is generally less focus on B->X converters (that block is hard to get rid of). Congratulations on the great discovery!
-Matthias Merzenich

User avatar
Kazyan
Posts: 887
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Post by Kazyan » February 22nd, 2015, 1:26 am

Yay, I did a thing! And it wasn't a search--it was just manual catalyst-placing.

The ship+eater and double-eater catalyst are broadly applicable, but obscure and apparently less-explored. I just started using them and soon, discovery.
Tanner Jacobi

chris_c
Posts: 921
Joined: June 28th, 2014, 7:15 am

Re: The Hunting of the New Herschel Conduits

Post by chris_c » February 22nd, 2015, 2:06 am

dvgrn wrote:I know there are at least a few more workable options out there, but I think it's somebody else's turn...!
Ok so here is an Fx245:

Code: Select all

x = 63, y = 57, rule = LifeHistory
29.A$28.A.A7.A$28.BA2B4.3A$18.2A9.2B4.A$17.BA.AB2.6B5.2A$17.2B2A10B3.
3B3.7B.B$15.2AB.11B2.3B5.13B.B$14.A.AB4.38B$14.A6.40BD$13.2A4.B.38B3D
B$18.2A22B2A15BDBD$18.2A.21B2A15BD$19.30B3.2B2.2B$21.6B.2B.B3.B2.10B$
21.6B12.6B$20.7B13.3B$21.8B11.B$21.8B10.2A$18.AB.9B9.A$17.A.A10B10.A$
17.2A.7B.3B8.2A$22.10B$22.10B$21.11B$22.10B$21.B.8B$20.9B$18.11B$18.
10B$17.12B$16.14B$16.17B$17.15B2A$17.15B2A$16.17B$13.16B$12.16B$12.
16B$11.16B$10.20B$10.20B$11.20B$10.22B$8.24B$6.25B$6.2BA13B.10B$5.3BA
BA4B.7B2.2B4.2A$6.2B3A4B2.3B.B4.B4.A$5.5BA4B4.2B2A3.2A4.3A$4.10B4.BA
2.A4.A6.A$3.4B12.3A5.A.A$2.4B22.2A$.4B14.3A$4B15.A2.A$3B18.2A$2B$B!
(EDIT: you must have known about this one already given your almost identical H->G, but never mind)

User avatar
dvgrn
Moderator
Posts: 5978
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » February 22nd, 2015, 1:05 pm

For the sake of completeness, it's also possible to string together an L560+4N, or L588+4N for the transceiver terminated by R64:

Code: Select all

#C L560+4N (left) and R328 (right)
#C Repeat time 271 (or a few overclocked smaller periods)
#C thanks to the slow B-to-H at the end
#C [[ VIEWONLY THUMBNAIL ZOOM 2.8 ]]
x = 214, y = 85, rule = LifeHistory
99.A$99.3A68.2A$82.2A18.A68.A14.2A$83.A17.2A4.B63.A.AB10.B2AB8.2A$82.
A18.8B63.2AB.2B7.4B7.B2A2B$82.2A19.8B5.2A56.5B6.4B8.4B$83.B18.9B6.A
50.2A4.7B4.5B5.6B5.2A$83.3B15.12B.3A51.A5.7B3.7B.8B.2B2.B2AB$82.6B12.
14BA54.3A2.7B.22B2.2B$78.B2.10B8.15B57.A35B$75.17B3.2B.15B60.36B$75.
10B2A15BC11B60.32B2A2B$75.10B2A15BCBC10B59.32B2A3B$75.27B3C4B.B2AB2A
49.2A3.2A.B.35B$75.29BC4B2.A3.A49.A.A.A.A38B$75.33B4.3A52.A.A.B.30B.
7B$75.3B5.13B.4B65.2A.2A23B3.B2.4B.6B$77.3B3.7B.4B.4B12.3A55.B.21B3.
3B2.9B$78.2A10.4B.4B12.A3.A56.20B3.B2AB3.8B$78.A10.4B.4B13.2A.2A57.
13B.4B5.2A5.8B$79.3A6.4B.4B77.11B.4B14.7B$81.A5.4B.4B75.18B16.6B$86.
4B.4B75.18B17.2B3D2B$77.2A6.4B.4B76.17B18.2BD4B$76.B2AB4.4B.4B77.16B
19.B3D4B$41.A35.3B3.4B.4B78.8B.6B20.9B$39.3A34.B.B3.4B.4B79.7B3.4B21.
9B$38.A36.10B.4B80.7B2.4B22.9B$38.2A33.11B.4B81.6B2.4B24.8B$36.4B11.
2A19.11B.4B81.7B.4B$35.2AB4.2A6.A.A17.12B.4B81.12B$35.A.AB2.B2AB5.BA
17.12B.4B81.12B$36.2AB3.B.B2.B3.2B15.9B2A6B81.12B33.2A$36.3B.13B2.B
10.10BA2BA4B.B79.7BD4B33.A2.A$37.20B8.12B2A9B78.5B3D2B35.2A$37.21B6.
6B2A17B77.B.2B2D2BD4.2A$36.22B4.8B2A16B75.2A4.6B4.A$36.22B2.29B75.A4.
6B.BA.A$36.53B72.3A4.7B.B2A$36.9B3.4B.35B72.A8.8B$33.AB.10B2.40B73.5A
3.7B$32.A.A11B4.31B2.B.B54.2A24.A2.7B$33.A2B.B.6B7.28B2.3B56.A13.2A6.
3A2.7B4.B$34.B4.7B7.27B2.B2AB55.A.AB9.B2A2B4.A5.6B.B.2BA$38.7B10.24B
4.2A57.2AB.3B6.4B.3B5.11BA.A$38.7B13.17B.B.B2A63.7B2.11B4.5BD4B.BA$
37.8B15.9B.6B2.BA.A62.22B.B.3B3D3B$35.2AB.7B16.6B5.2B6.A63.27BD5B$34.
A.AB.6B5.2A10.8B11.2A61.25B2D7B$34.A4.6B3.B.A12.5B2A72.28B2D6B$33.2A
4.D2B2D2B.BA.A14.3B2A70.29B2D6B$.2A35.2B3D2B.2B2A16.2B72.2BC26BD7B$A
2.A33.4BD2B.4B16.4B70.3BCBC4B.24B.3B$.2A33.12B17.B2AB71.2B3C4B2.16B2.
B.B3.B2A$35.12B19.2A71.5BC4B2.7B.4B6.B2AB2.BA.A$34.12B92.10B4.6B2.2B
8.2A4.B2A$33.4B.7B92.4B11.6B16.4B$8B24.4B2.6B93.3B13.3B18.2A$9B22.4B
2.7B91.4B12.7B17.A$9B21.4B3.7B91.2A14.2A.B.2A14.3A$9B20.6B.8B92.A15.A
3.A15.A$.4B3DB19.16B89.3A13.3A5.3A$2.4BD2B18.17B89.A15.A9.A$2.2B3D2B
17.18B$3.6B16.18B$3.7B14.4B.11B$3.8B5.2A5.4B.13B$4.8B3.B2AB3.20B$4.9B
2.3B3.21B.B$3.6B.4B2.B3.23B2A.2A$3.7B.30B.B.A.A$4.38BA.A.A.A$4.35B.B.
2A3.2A$3.3B2A32B$4.2B2A32B$5.36B$7.35BA$6.2B2.22B.7B2.3A$5.B2AB2.2B.
8B.7B3.7B5.A$6.2A5.6B5.5B4.7B4.2A$13.4B8.4B6.5B$13.2B2AB7.4B7.2B.B2A$
15.2A8.B2AB10.BA.A$26.2A14.A$42.2A!
Strangest-looking left turn I ever saw, and again they're both really slow to recover. These are definitely not on the most usable end of the conduit spectrum. Are there other viable combinations that I'm not seeing?

Also, I was wrong in a previous post about the R-to-pi converter having the pi too close to be useful. A variant of that R-to-pi is already in Calcyman's collection: it's a B-to-pi, using a block in that same position relative to the pi. The table catalyst now seems vaguely familiar, too, so it's possible someone (Guam?) may have posted it before.

Anyway, as an apology, I've included the table+block R-to-pi in a probably-never-before-seen composite Fx conduit (and G1 tandem-glider emitter). It's probably a good idea to forget about this one quickly -- best not to even collect the abysmal statistics on it:

Code: Select all

#C [[ THUMBNAIL AUTOSTART STOP 999 ]]
x = 127, y = 107, rule = LifeHistory
43.2A3.2A$43.2A2.A.3A$47.A4.A$43.4A.2A2.A$43.A2.A.A.A.2A$46.A.A.A.A$
47.2A.A.A$51.A2$37.2A$38.A7.2A$38.A.A5.2A17.A$39.2A22.3A$62.A$62.2A3$
70.2A.2A$71.A.2A$49.2A20.A$49.A13.2A4.3A$30.A19.3A10.2A3.A3.2A$28.5A
14.3A2.A15.4A2.A$27.A5.A13.A2.A.A.2A15.A.2A$27.A2.3A12.A.A2.A2.A.A12.
3A2.A$26.2A.A15.2A4.A.A13.A5.A$26.A2.4A19.2A14.5A13.2A3.2A$27.2A3.A3.
2A32.A13.3A.A2.2A$29.3A4.2A45.A4.A$29.A53.A2.2A.4A$26.2A.A52.2A.A.A.A
2.A$26.2A.2A52.A.A.A.A$83.A.A.2A$84.A$37.2A$38.A58.2A$35.3A50.2A7.A$
35.A52.2A5.A.A$95.2A7$85.2A$86.A$83.3A$83.A2$19.A$19.3A6.2A$22.A5.2A$
21.2A2$30.D7.D$22.2A7.D6.D.D50.A$22.2A7.2D5.3D.2A45.3A$30.2D8.D.2A22.
D21.A$30.D35.2D8.D11.2A$67.2D5.3D$22.2A43.D6.D$22.2A42.D7.D3$116.D$
22.2A90.3D$22.A74.2A15.D$19.2A.A42.2A30.2A15.D$19.2A.2A5.D35.2A7.2A
49.A$28.3D42.A.A48.A.A$28.D44.A50.A.A$51.2A19.2A10.2A39.A$52.A32.A$
49.3A30.3A9.2A$49.A32.A11.A$61.2A32.A$61.A32.2A$62.3A$36.2A26.A$36.2A
12$12.C$12.C.C20.2A$12.3C20.A$14.C10.2A3.2A4.3A$23.A2.A4.A6.A$23.3A5.
A.A$32.2A$23.3A$3.A.2A16.A2.A$.3A.2A18.2A$A$.3A.2A$3.A.A$3.A.A$4.A!
I think that the above mess, possibly along with some equally hideous variants involving the Lx200 dependent conduit, is the only known way to use that table+block R-to-pi. (But I've been wrong before.) It's necessary to suppress the output Herschel's glider. Otherwise the glider will kill off the table, making the conduit useless.

There are other dependent conduits as well, but besides the Lx198 which doesn't quite fit, I think all the rest are periodic, and I'm refusing to look at those -- it might encourage them.

I remember Guam did some really clever things with that Fx119-like pi-to-Herschel component, though, that I haven't really gotten used to yet -- so I might easily be missing something good.

chris_c
Posts: 921
Joined: June 28th, 2014, 7:15 am

Re: The Hunting of the New Herschel Conduits

Post by chris_c » February 22nd, 2015, 1:09 pm

dvgrn wrote:For each orientation we'll need a recipe that builds the object at the back edge of its reaction envelope, and one on the left edge, and one on the right edge. Once we've collected all of those recipes for some non-Spartan object, it becomes just as good as any Spartan object (except for the extra cost in gliders.)

But even with Guam's work on tub-with-tail eaters and spark coils (links above) we're not anywhere near there yet.
Here is a dead spark coil near to the back of the reaction envelope in ten gliders from a traffic light. I reckon it should be usuable for all 4 orientations of the Herschel transmitter without massive difficulty.

Code: Select all

x = 918, y = 891, rule = B3/S23
913bo$913bo$913bo2$909b3o3b3o2$913bo$913bo$913bo29$885bo$885b2o$884bob
o79$790b3o$792bo$791bo98$692b3o$694bo$693bo85$613bo$613b2o$612bobo99$
518b3o$520bo$519bo98$421b3o$423bo$422bo80$301b2o$300bobo$302bo98$196b
2o$195bobo$197bo98$102b2o$101bobo$103bo98$b2o$obo$2bo!

User avatar
dvgrn
Moderator
Posts: 5978
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » February 22nd, 2015, 1:44 pm

chris_c wrote:Here is a dead spark coil near to the back of the reaction envelope in ten gliders from a traffic light. I reckon it should be usuable for all 4 orientations of the Herschel transmitter without massive difficulty.
Yes, indeed -- a little messy, maybe but quite workable.

I was worried about the orientation where the associated eater in the tandem-glider converter is behind the spark coil -- see the last pattern in this posting for all four construction cases (how convenient!). Luckily there's just exactly enough room for the eater to be constructed first. Even the extra beehive behind the spark coil can be cleaned up, though it's out of the way so it could just be left there for decoration.

That takes care of single-arm constructions of H->2G. I'm not sure about Guam's two-arm construction recipes, but there's generally a lot more flexibility there -- and of course we can always just use one-arm recipes instead! I wonder how often it will turn out to be more efficient for the two arms to wander off and do single-arm slow-salvo constructions in two places at once?

Once the current excitement dies down, I'm thinking it's time to finish stringing together a workable design for a two-arm quadratic-growth Geminoid replicator, quite possibly including things like these spark coils.

(On the other hand, I'd like to finish the Demonoid spaceship before getting too deep into the replicator project. Too many interesting things to do around here...!)

Sphenocorona
Posts: 481
Joined: April 9th, 2013, 11:03 pm

Re: The Hunting of the New Herschel Conduits

Post by Sphenocorona » February 22nd, 2015, 5:08 pm

Any chance of restoring the yellow block here (preferably without wiring up a glider to it and keeping the advantages of the FNG suppression?)

Code: Select all

x = 27, y = 17, rule = LifeHistory
15.2C$3.2A11.C$4.A11.C.C$3.A13.2C$3.2A$4.B17.2E$4.3B15.2E$3.6B$2.10B$
2.11B3.2B2.2B$2BD3B2A15BC$2B2D2B2A15BCBC$.2B2D18B3CB$2.BD21BC$2.D21B$
4.13B.B$4.7B.B!

User avatar
Kazyan
Posts: 887
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Post by Kazyan » February 22nd, 2015, 7:20 pm

Sphenocorona wrote:Any chance of restoring the yellow block here (preferably without wiring up a glider to it and keeping the advantages of the FNG suppression?)

Code: Select all

x = 27, y = 17, rule = LifeHistory
15.2C$3.2A11.C$4.A11.C.C$3.A13.2C$3.2A$4.B17.2E$4.3B15.2E$3.6B$2.10B$
2.11B3.2B2.2B$2BD3B2A15BC$2B2D2B2A15BCBC$.2B2D18B3CB$2.BD21BC$2.D21B$
4.13B.B$4.7B.B!
Yes, actually! He're something that works but kills the Herschel, and something that almost works but leaves activity.

Code: Select all

x = 26, y = 15, rule = LifeHistory
2A19.2A$A8.A12.A$.3A3.3A12.A.2A$3.A2.A16.A.A$6.2A2$2C18.2C$2C18.2C4$.
C19.C$.C.C17.C.C$.3C17.3C$3.C19.C!
Looks like a job for Bellman to get the best of both worlds.
Tanner Jacobi

Sphenocorona
Posts: 481
Joined: April 9th, 2013, 11:03 pm

Re: The Hunting of the New Herschel Conduits

Post by Sphenocorona » February 22nd, 2015, 9:15 pm

Here's another transparent reaction with a different mechanism (though this is probably not as useful):

Code: Select all

x = 10, y = 16, rule = LifeHistory
2.2C$.C.C$.2C3$2C$2C4$.C$.C.C$.3C$3.C!
The results from manual search with various types of catalysts seems somewhat promising, though I can't say if Bellman will be able to find something that is remotely like what I'm investigating:

Code: Select all

x = 13, y = 20, rule = LifeHistory
3.A2.A2.2A$3.4A2.2A2$.10A$A.A8.A$A3.2A.2A2.A$.3A.A.C.A.2A$3.A.C3.A$4.
2C2.CA3$.2C$.2C4$2.C$2.C.C$2.3C$4.C!
EDIT: This catalyst doesn't really accomplish anything before its destruction, but it continues reacting for a surprising amount of time before finally exploding for a catalyst found by hand:

Code: Select all

#C [[ THUMBNAIL GPS 10 STOP 42 ]]
x = 14, y = 20, rule = LifeHistory
.2E4.E$E2.E3.3E$3E.E5.E$4.7E$2.2E7.2E$.E3.2E.2E3.E$.E.2E.E.E.3C$2.E.C
.E.C.C$5.2C.C.C$9.C2$2.2C$2.2C4$3.C$3.C.C$3.3C$5.C!

Sokwe
Moderator
Posts: 1488
Joined: July 9th, 2009, 2:44 pm

Re: The Hunting of the New Herschel Conduits

Post by Sokwe » February 22nd, 2015, 11:08 pm

dvgrn wrote:The L200 itself recovers in as little as 59 ticks if you suppress the output Herschel's glider.
Conveniently, it allows for a much smaller p59 emu:

Code: Select all

x = 190, y = 90, rule = B3/S23
36bo35b2o12bo35b2o12bo$23bo9bo2b3o5b2o27bo12b3o5b2o27bo12b3o5b2o24bo$
23b3o5b3o5bo4bo28bobo13bo4bo28bobo13bo4bo16bo7bobo$10bo15bo3bo7bo3bobo
29b2o12bo3bobo29b2o12bo3bobo16b3o6bo$8b3o14b2o3b2o5bo3bobo43bo3bobo43b
o3bobo20bo15b2o$7bo29b2o3bo44b2o3bo44b2o3bo20b2o14bobo$7b2o57bo49bo62b
2o$64b3o47b3o66b2o$4b2o5b2o24bo2bo22bo33b2o14bo45bo15bo7bobo$4bobo4b2o
23bo3bo9b2o11b2o31bobo14b2o42b2ob2o12b3o8bo$5b2o19b3o7bo4bo8b2o25b3o
15b2obo58bo15b2o2bo7b2o$22bo13bo4bo35bo2bo15b2o41b2o16bo3bo11bo3bob3o$
21b3o13bo3bo17b3o15bob2o16bo40b3o17b3o12bobo4b2o$20b2ob2o13bobo17b2obo
14bo57b2o4bo38bo$19b3o2b2o31b2obo13bo2bo25b2o29bo2b2o35b2ob2o$20b2o2b
2o32bobo13bobobo3bo19bo32b2o39bo$23b3o33b2o4bo9b2obo4bo31bo20bo30bo$6b
3o14b2o38bobo11bo4bo30bobo50bobo$2bo5bo55b2o47b3o9b2o32b2o5bobo$bobo4b
2obo27b2o47b2o23bo11b2o11b2o20bo4b2ob2o$2bo6b3o7bo6b2o11bobo46bo12b3o
34bo20bo$10bo6b3o6b2o13bo47b3o47b3o17b2o$16bo24b2o48bo11bo37bo$17b3o
42b2o3bo35b2o7b2o3bo69b2o$19bo42bo3bobo34b2o7bo3bobo56bo11bo$4b2o43b2o
12bo3bobo29b2o12bo3bobo54b2o9bobo$3bobo42bobo13bo4bo28bobo13bo4bo48b2o
3b4o8b2o$3bo44bo12b3o5b2o27bo12b3o5b2o47b2o2b2o3b3o$2b2o43b2o12bo35b2o
12bo61bo5bo$8b2o8b2obo157bo$7bo2bo7bob2o151bo4bo6b2o$7bo2bo161b2o2bo8b
obo$7bo164b2obo11bo$7b2o3bo160b2o12b2o$8bo2b2obo$9b3o2bo$188b2o$184b2o
2bo$184bobobo$169b2o16bo$170bo$8b3o156b3o15bo$8bo2b2o154bo16bobo$9b3o
173bo$bo8bo175b3o$b3o175bo8bo$4bo173b3o$3bobo16bo154b2o2bo$4bo15b3o
156b3o$19bo$2bo16b2o$bobobo$bo2b2o$2o$175bo2b3o$175bob2o2bo$b2o12b2o
160bo3b2o$2bo11bob2o164bo$2bobo8bo2b2o161bo2bo$3b2o6bo4bo151b2obo7bo2b
o$10bo157bob2o8b2o$10bo5bo61bo12b2o35bo12b2o43b2o$10b3o3b2o2b2o47b2o5b
3o12bo27b2o5b3o12bo44bo$3b2o8b4o3b2o48bo4bo13bobo28bo4bo13bobo42bobo$
2bobo9b2o54bobo3bo12b2o29bobo3bo12b2o43b2o$2bo11bo56bobo3bo7b2o34bobo
3bo42bo$b2o69bo3b2o7b2o35bo3b2o42b3o$48bo37bo11bo48b2o24bo$29b2o17b3o
47b3o47bo13b2o6b3o6bo$30bo20bo34b3o12bo46bobo11b2o6bo7b3o6bo$20b2ob2o
4bo20b2o11b2o11bo23b2o47b2o27bob2o4bobo$21bobo5b2o32b2o9b3o47b2o55bo5b
o$21bobo50bobo30bo4bo11bobo38b2o14b3o$22bo30bo20bo31bo4bob2o9bo4b2o33b
3o$13bo39b2o32bo19bo3bobobo13bobo32b2o2b2o$11b2ob2o35b2o2bo29b2o25bo2b
o13bob2o31b2o2b3o$10bo38bo4b2o57bo14bob2o17bobo13b2ob2o$8b2o4bobo12b3o
17b3o40bo16b2obo15b3o17bo3bo13b3o$8b3obo3bo11bo3bo16b2o41b2o15bo2bo35b
o4bo13bo$3b2o7bo2b2o15bo58bob2o15b3o25b2o8bo4bo7b3o19b2o$4bo8b3o12b2ob
2o42b2o14bobo31b2o11b2o9bo3bo23b2o4bobo$4bobo7bo15bo45bo14b2o33bo22bo
2bo24b2o5b2o$5b2o66b3o47b3o$9b2o62bo49bo57b2o$8bobo14b2o20bo3b2o44bo3b
2o44bo3b2o29bo$8b2o15bo20bobo3bo43bobo3bo43bobo3bo5b2o3b2o14b3o$19bo6b
3o16bobo3bo12b2o29bobo3bo12b2o29bobo3bo7bo3bo15bo$18bobo7bo16bo4bo13bo
bo28bo4bo13bobo28bo4bo5b3o5b3o$19bo24b2o5b3o12bo27b2o5b3o12bo27b2o5b3o
2bo9bo$53bo12b2o35bo12b2o35bo!
dvgrn wrote:I was wrong in a previous post about the R-to-pi converter having the pi too close to be useful.... I think that the above mess, possibly along with some equally hideous variants involving the Lx200 dependent conduit, is the only known way to use that table+block R-to-pi.
There is also this B->G which gives a "relatively nice" R->Pi->B->G:

Code: Select all

x = 40, y = 45, rule = LifeHistory
3.2A16.2A$3.2A16.A$19.A.A$3.4A3.2A3.2A2.2A$3.A2.A3.2A3.2A4$13.3D4.2A$
.2C12.D4.2A$2.2C9.3D$2.C10.2D2.D$14.3D$15.D3$9.2A$9.2A6$2A$.A$.A.A$2.
2A$38.2D$37.D.D$39.D4$3.2A$2.A.A$2.A$.2A2$12.2A$12.A.A$13.A$28.2A$28.
A$29.3A$31.A!
That B+block reaction has the potential to be quite useful, as it can remove the B's natural block in some very tight spaces, but I just haven't been able to get anything better than a glider out of it (at least in any sort of nice way).
dvgrn wrote:The table catalyst now seems vaguely familiar, too, so it's possible someone (Guam?) may have posted it before.
I specifically remember Guam posting the table variant of the R->B.
-Matthias Merzenich

Sphenocorona
Posts: 481
Joined: April 9th, 2013, 11:03 pm

Re: The Hunting of the New Herschel Conduits

Post by Sphenocorona » February 23rd, 2015, 12:49 am

I've never seen an Eater 2 catalyze something like it does in this new, 6th R->B variant:

Code: Select all

#C [[ AUTOSTART THUMBNAIL GPS 7 LOOP 28 ]]
x = 16, y = 19, rule = LifeHistory
4.A$3.A.A$3.A.A6.2A$.3A.2A6.A$A6.B2.3A$.3AB2A3BA$3.A.2A2B$4.B.2B2.B$
2.B2.8B$.12BD$13B2D$B2C11B2D$2B2C10BD$2BC10BD$5B.4B$10B$8B$6B$.4B!
Maybe there's some other situation where this sort of thing is handy?

User avatar
Kazyan
Posts: 887
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Post by Kazyan » February 24th, 2015, 12:58 am

I'm starting to really like the ship-and-eater catalyst.

Code: Select all

#C Pi-to-Pi
#C [[ THUMBNAIL AUTOSTART ZOOM 12 HEIGHT 480 LOOP 40 ]]
x = 37, y = 10, rule = B3/S23
5b2o23b2o$6bo23bo$6bobo19bobo$2o5b2o19b2o5b2o$bo33bo$bobo13b3o13bobo$
2b2o12bo3bo12b2o$6b2o8b2ob2o8b2o$5bobo21bobo$5b2o23b2o!
Cant' be strung together, unfortunately, but I'll ask Bellman if it has any better ideas.
Tanner Jacobi

User avatar
dvgrn
Moderator
Posts: 5978
Joined: May 17th, 2009, 11:00 pm
Location: Madison, WI
Contact:

Re: The Hunting of the New Herschel Conduits

Post by dvgrn » February 24th, 2015, 1:24 am

Kazyan wrote:I'm starting to really like the ship-and-eater catalyst.
#C Pi-to-Pi
...Can't be strung together, unfortunately, but I'll ask Bellman if it has any better ideas.
Yup -- as usual, you can get a pi in, but it may still be tricky to get a signal out:

Code: Select all

x = 37, y = 46, rule = LifeHistory
17.3D$17.D.D$17.D.D$5.2A23.2A$6.A23.A$6.A.A19.A.A$2A5.2A19.2A5.2A$.A
33.A$.A.A29.A.A$2.2A13.3D13.2A$6.2A9.D.D9.2A$5.A.A9.D.D9.A.A$5.2A23.
2A3$32.A$30.3A$29.A$29.2A3$7.2A19.2A$7.2A19.2A10$24.3C$15.2A8.C$16.A
6.3C$13.3A$13.A6$33.2A$33.A.A$35.A$35.2A!
See also the bottom of this old weblog posting for a periodic (but chainable) version of the same 6-cell pi-to-pi reaction.

This may be a case for ptbsearch, actually: I'm not sure if anyone has done a thorough search for likely transparent block/beehive/whatever locations to draw a pi explosion away from a conduit like this, before the backwash destroys the nearest catalysts. Ptbsearch hasn't been applied to this since the late 1990's, anyway, I think. And it will also be interesting to see what Bellman comes up with, of course!

User avatar
Kazyan
Posts: 887
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Post by Kazyan » February 24th, 2015, 1:32 am

Yeah, I noticed that none of calcyman's elementary conduits allowed the Pi to escape, so we just need to get the catalysts further downstream so that something will fit. Chainability isn't even necessary if getting a 6-cell spacer like this would be useful somehow.
Tanner Jacobi

User avatar
Kazyan
Posts: 887
Joined: February 6th, 2014, 11:02 pm

Re: The Hunting of the New Herschel Conduits

Post by Kazyan » February 25th, 2015, 2:21 am

Bellman blew up trying to do the previous (as in, it started enumerating all still lifes that could be welded to an eater2).

Anyway, here's a transparent block reaction found through sheer dumb luck. A quick 2-catalyst run from generations 90-130 using the catalyst searcher doesn't yield anything, so I guess it's up to manual searches or Bellman catalysts to get that working (and hopefully, delete that annoying blinker with a glider or something).

Code: Select all

x = 16, y = 24, rule = LifeHistory
10.2A$10.A$8.A.A$4.2A2.2A$4.2A6$14.2A$14.A$12.A.A$12.2A4$2C$2C3$5.A$
4.3A$3.2A.A!
Tanner Jacobi

Post Reply