The Hunting of the Elementary Conduits

[MathAndCode]

Unidentified to H:

Code: Select all

x = 13, y = 48, rule = B3/S23
11b2o$11bo$9b3o16$9b2o$9b2o2$3bo$b3o$o$2o$5bo$4bobo$4b2o9$11b2o$4b2o5b
2o$5bo$5bobo$6b2o3$9b3o$10bobo$12bo$9bob2o$9b2o!

How'd you decide to work on that particular region?

[Entity Valkyrie 2]

I chose an active region that doesn't have a name, but this happened when a pi was hit by a unwanted spark.

Code: Select all

x = 7, y = 7, rule = B3/S23
bo2$2b3o$bo3bo$o5bo$o5bo$3ob3o!
#C [[ STOP 4 ]]

[wwei23]

I chose an active region that doesn't have a name, but this happened when a pi was hit by a unwanted spark.

How'd you find that transparent block?

[MathAndCode]

I chose an active region that doesn't have a name, but this happened when a pi was hit by a unwanted spark.

Ah; thank you for explaining.
Also, remember that I made this thread for conduits like that.

[wwei23]

I was able to turn a pi into that active region, but I don't think it's possible to get a pi into this:

Code: Select all

x = 25, y = 45, rule = B3/S23
6bo$6b3o$9bo$8b2o5$10b3o$3b2o5bobo$3bo2bo3bobo$4b3o2$2obob2o$ob2obo$5b
o$5b2o6$6bo$6b3o$9bo$8b2o5$5b2o3b3o$5b2o3bobo$10bobo!

[MathAndCode]

I was able to turn a pi into that active region, but I don't think it's possible to get a pi into this:

Code: Select all

x = 25, y = 45, rule = B3/S23
6bo$6b3o$9bo$8b2o5$10b3o$3b2o5bobo$3bo2bo3bobo$4b3o2$2obob2o$ob2obo$5b
o$5b2o6$6bo$6b3o$9bo$8b2o5$5b2o3b3o$5b2o3bobo$10bobo!

The bottom conduit looks feasible.

[calcyman]

I was able to turn a pi into that active region, but I don't think it's possible to get a pi into this:

Code: Select all

x = 25, y = 45, rule = B3/S23
6bo$6b3o$9bo$8b2o5$10b3o$3b2o5bobo$3bo2bo3bobo$4b3o2$2obob2o$ob2obo$5b
o$5b2o6$6bo$6b3o$9bo$8b2o5$5b2o3b3o$5b2o3bobo$10bobo!

The bottom conduit looks feasible.

Yes, here's a 107-tick H-to-unidentified:

Code: Select all

x = 34, y = 43, rule = LifeHistory
11.A$11.3A$14.A$13.2A$13.5B$15.4B$13.5B3D$11.B.6BDBD$10.2A9BD$10.2A6B
DB2D$11.7B2D$11.8B$14.6B$13.7B$2.A10.8B$2.3A8.9B$5.A7.9B$4.2A6.10B$4.
5B2.13B$6.18B.2B$5.2A19B2A$5.2A17B.B2A$6.B.17B.B$8.16B$9.14B$10.8B2.
4B$11.7B3.4B$8.11B3.4B$7.12B4.4B$7.12B5.4B$7.11B7.4B$7.B3C4B.4B6.4B$
7.2BC4B4.2A7.3B$7.2B3C2B4.A$7.6B6.3A8.2A$6.7B8.A8.A$5.4B.B20.3A$4.4B
25.A$4.3B$2.2AB$.A.AB$.A$2A!

It's sadly not compatible with EV2's unidentified-to-H, though.

[MathAndCode]

Yes, here's a 107-tick H-to-unidentified:

There might be another way to get something out of that undefined region. Is that the ∏-making conduit with the best output clearance?



Edit: I did not find any other conduits accepting that undefined region, but I did find this:

Code: Select all

x = 11, y = 14, rule = B3/S23
6bo$4b3o$3bo$3b2o5$3o$obo$obo4b2o$7bobo$9bo$9b2o!

The region at generation 22 that resulted from the loaf-fishhook interaction is the grandchild of a Herschel. (I have found over ten of these.) I doubt that that particular reaction will end up being useful, but Bellman could probably find a catalyst for converting loaves that are dropped near the edge of an active region in certain ways into Herschels.



Another edit: Since I seem to be finding a lot of Herschels with alternative predecessors, some coming from alternative directions, I may as well start finding conduits to accommodate this.

Code: Select all

x = 16, y = 11, rule = LifeHistory
.A12.2A$A.A11.A$.2A9.A.A$7.A4.2A$4.2D.3A$5.2DA.A$5.D3.A2$.D$2D$.2D!

Obviously, the fishhook needs to be replaced. The boat can be replaced with at least fourteen other sacrificial objects and probably a stable catalyst.

[Hunting]

Random attempt at catalyzing this unnamed object: (With the help of Kittyst)

Code: Select all

x = 23, y = 25, rule = B3/S23
9b2o$9bobo$11bo$2b2o6b3o4b2o$2b2o13b2o4$16b2o$16b2o$2o$2o7$16b2obo$16b
2ob3o$22bo$16b2ob3o$17bobo$17bobo$18bo!

Code: Select all

x = 20, y = 12, rule = B3/S23
9b2o$9bobo$11bo$2b2o6b3o4b2o$2b2o13b2o5$18b2o$2o16b2o$2o!

[dvgrn]

Does anybody mind if I split this thread, and put the recent piles of speculative / partial conduits into a new thread? I definitely don't want to discourage this kind of research, since eventually it will probably produce some amazing new mechanisms.

But on the other hand, the Elementary Conduits thread used to get a new post mostly just when someone found a new actual elementary conduit ... and it's still already thirteen pages long, going up fast. It's getting difficult to find the actual working conduits in among all the speculation.

If nobody objects to splitting the thread, what should the new thread be called? "Conduit Research Thread", let's say? My idea would be that a conduit should be cross-posted to this thread as soon as there are multiple distinct ways to start from active object X, convert it to some other known active object already used in the ECC, and then convert back again to object X -- in such a way that arbitrarily long chains can be created.

... I'm trying not to be Herschel-centric here. Last time I mentioned making Herschel conduits as a test for elementary-conduit usefulness, it was really just an attempt to say the same thing I'm saying here. Generally a useful elementary conduit will be able to connect to H-to-X and X-to-H conduits to make a valid Herschel conduit. With luck it can also connect B-to-X and X-to-B, etc., etc., but if it can't do the H-to-X and X-to-H trick then most likely it's pretty awkward to use, when it comes right down to it..

[MathAndCode]

Does anybody mind if I split this thread, and put the recent piles of speculative / partial conduits into a new thread? I definitely don't want to discourage this kind of research, since eventually it will probably produce some amazing new mechanisms.

But on the other hand, the Elementary Conduits thread used to get a new post mostly just when someone found a new actual elementary conduit ... and it's still already thirteen pages long, going up fast. It's getting difficult to find the actual working conduits in among all the speculation.

If nobody objects to splitting the thread, what should the new thread be called? "Conduit Research Thread", let's say?

That's a good name. I would probably call it, "Thread for your promising partial conduits," but I guess that the title doesn't have to explain its entire purpose.

My idea would be that a conduit should be cross-posted to this thread as soon as there are multiple distinct ways to start from active object X, convert it to some other known active object already used in the ECC, and then convert back again to object X -- in such a way that arbitrarily long chains can be created.

Upon reading this, I realize that I may have been confused initially. If you were talking about conduits that used unnamed active regions instead of partial conduits, then there is already a thread for that.

I'm trying not to be Herschel-centric here. Last time I mentioned making Herschel conduits as a test for elementary-conduit usefulness, it was really just an attempt to say the same thing I'm saying here. Generally a useful elementary conduit will be able to connect to H-to-X and X-to-H conduits to make a valid Herschel conduit. With luck it can also connect B-to-X and X-to-B, etc., etc., but if it can't do the H-to-X and X-to-H trick then most likely it's pretty awkward to use, when it comes right down to it..

From what I gather, one of the requirements for a conduit to be listed in the Elementary Conduits Collection is that it can be preceded and followed by other conduits in the Elementary Conduits Collection, thus ensuring that it is connectable with the other conduits. These conduits can intake or output Herschels, but I don't believe that they should have to in order for the conduit to be counted as valid, i.e. connecting the new conduit with other conduits should also work. After all, the Elementary Conduits Collection includes many H→X and X→H conduits, and inclusion in the Elementary Conduits Collection implies connectability (typically), so anyone who wants a Herschel conduit can connect more conduits in order to get a composite H→H conduit.

[wwei23]

so anyone who wants a Herschel conduit can connect more conduits in order to get a composite H→H conduit.

Which is why I suggested letting people put together the Lego pieces however they needed them put together, instead of prepackaging them in Herschel conduits.

[MathAndCode]

Which is why I suggested letting people put together the Lego pieces however they needed them put together, instead of prepackaging them in Herschel conduits.

Also, having search programs store every elementary conduit with any input and output is much more feasible than having search programs store every conduit with a Herschel input, a Herschel output, and no Herschel intermediate.

[dvgrn]

From what I gather, one of the requirements for a conduit to be listed in the Elementary Conduits Collection is that it can be preceded and followed by other conduits in the Elementary Conduits Collection, thus ensuring that it is connectable with the other conduits.

Yup, that's pretty much the key criterion. On the other hand, it's easy to end up with situations where a new candidate only connects to one specific conduit A as an input, and another specific conduit B as an output -- and then the geometry ends up being no good, such that you can't both get an input into conduit A and an output out of Conduit B. Which makes it pretty questionable whether the candidate conduit should really find its way into the ECC after all, since there's still no known way to use it.

I guess maybe the real criterion that I'm trying to get at is something like "conduit chains can be created that include N copies of the candidate conduit, for any chosen number N".

Which is why I suggested letting people put together the Lego pieces however they needed them put together, instead of prepackaging them in Herschel conduits. :P

Also, having search programs store every elementary conduit with any input and output is much more feasible than having search programs store every conduit with a Herschel input, a Herschel output, and no Herschel intermediate.

Nobody is arguing in favor of anything even vaguely like that, as far as I can tell. The idea is strictly this: if it isn't possible for a new candidate elementary conduit to be part of a standard composite Herschel conduit, then the odds are pretty low that that candidate conduit will ever turn out to be useful for anything. The doesn't seem much point in collecting conduits that are not likely to ever be useful for anything -- they'll mostly just clutter up the collection.

There's been talk sometimes about making a separate collection of "hopeful conduits" that currently don't measure up to ECC criteria, but might turn out to be useful if someone finds better inputs or outputs for them. I guess I'd just rather have a separate thread to hold most of these "hopeful conduits", so that this main Elementary Conduits thread can just collect conduits after they've actually been shown to be connectable.

[MathAndCode]

From what I gather, one of the requirements for a conduit to be listed in the Elementary Conduits Collection is that it can be preceded and followed by other conduits in the Elementary Conduits Collection, thus ensuring that it is connectable with the other conduits.

Yup, that's pretty much the key criterion. On the other hand, it's easy to end up with situations where a new candidate only connects to one specific conduit A as an input, and another specific conduit B as an output -- and then the geometry ends up being no good, such that you can't both get an input into conduit A and an output out of Conduit B. Which makes it pretty questionable whether the candidate conduit should really find its way into the ECC after all, since there's still no known way to use it.

I guess maybe the real criterion that I'm trying to get at is something like "conduit chains can be created that include N copies of the candidate conduit, for any chosen number N".

Nobody is arguing in favor of anything even vaguely like that, as far as I can tell. The idea is strictly this: if it isn't possible for a new candidate elementary conduit to be part of a standard composite Herschel conduit, then the odds are pretty low that that candidate conduit will ever turn out to be useful for anything. The doesn't seem much point in collecting conduits that are not likely to ever be useful for anything -- they'll mostly just clutter up the collection.

You have a point about the proceeding and succeeding conduits interfering with each other, although that doesn't happen very often. A good method is probably to make sure that the conduit can be connected with a beginning and ending conduit that fit some clearance requirement, e.g. we probably don't have to worry about anything that can be fed into conduit 1 having enough output clearance. Starting or ending with gliders should also be okay. I think that this is better than simply making sure that everything begins and ends with a Herschel. For example, I once had a conduit that made a Herschel's grandchild and needed to be followed by a dependent conduit, but L122 didn't fit.

[Kazyan]

Probably no dice on the dot-spark-to-X idea from the Uselss Discoveries thread (a Bellman search up to max-active 12 found noting; see addendum below), but I did find a more specific small spark that can quickly be turned into a Herschel:

Code: Select all

x = 26, y = 17, rule = LifeHistory
4.A$4.3A$7.A5.2C$6.2A4.B.B$6.5B3C$8.8B.B$.2A4.16B$.A.A2.5B2A11BD$3.A
2.5B2A9B3DB$3.3A16BDBD$.2A2.BA.14BD$A2.4A2B.2B3.2B2.2B$2A.A.4B$3.A.2A
$3.A.A$2.2A.A.A$6.2A!

The placement of the eater is inconvenient. I'd just say to keep an eye out for conduits (or FNG terminations) that happen to throw off a spark like this.

Bellman input file for the 'no dice' result:

Code: Select all

#S max-active 12
#S repair-interval 20
#S stable-interval 4
#S max-live 500
#S last-encounter 11
#S first-encounter 5
#P 0 0
.
.
..?????????????????????????..
..?????????????????????????..
..?????????????????????????..
..?????????????????????????..
..?????????????????????????..
..?????????????????????????..
..?????????????????????????..
..?????????????????????????..
..?????????????????????????..
..?????????????????????????..
..?????????????????????????..
..?????????????????????????..
..?????????????????????????..
..????????????**???????????..
.
.
.
.
.............@
............@.@
............@.@@
.............@
.........@@
........@.@
........@
.......@@
.
.

[dvgrn]

The placement of the eater is inconvenient. I'd just say to keep an eye out for conduits (or FNG terminations) that happen to throw off a spark like this.

I like the idea of hunting for conduits like this that can produce a viable output signal from an edge-shot spark of some kind. Seems like what we could really use is a way of automatically generating a big pile of high-clearance H-to-spark conversions -- or let's just say "X-to-spark", so that no one has to comment about Herschels being too specific.

Really a big collection of X-to-anything-edgy might be good to have. Just brute-force applying each reaction to various hopeful conduits might allow for a few lucky finds.

Mostly Random Tangent -- Don't Expect Relevance
In ancient times (2004) there were several known ways of doing 2G->H conversions, including one that could be used as a merge circuit. Here's an early p39 gun showing the reaction:

Code: Select all

#C p39 gun based on overclocking a Herschel track with p156s
#C December 18, 2014
x = 236, y = 131, rule = LifeHistory
177.2A$175.BABA2B$175.BA4B$174.AB2A4B$172.7BA4B$166.2A3.2BA3BA2BA4B3.
2B$165.A2BAB.3B4ABA2B2AB.5B$166.2A13BA2BA5B$163.B2.16B2A6B2.A$162.29B
ABA$162.5B2.8B2A8B2.2BABA$163.4B3.3B2ABA3BA5B3.3BA$163.5B3.2B2A5BA4B
3.5B$164.5B2.9BA4B2.5B$162.9B.4BABA5B.6BA2B$161.11B.3B3A4B.6BABA2B$
161.BA5B2A3B3.4B3.7BABA2B$160.A2BA4B2A3B4.2B4.8BA4B$96.A61.5BA9B10.2B
3A5BA4B$94.3A11.A49.3A2BA4BA2B2AB8.10BABA3B$93.A14.3A46.A2BAB2A3BA4BA
2B6.2B2A3BA8B2A$93.2A16.A14.A30.2A8BA3B2A2B6.2BA4BA3B2ABA2BA$91.4B15.
2A12.3A31.3BABA10B8.B2A2BA4BA2B3A$70.A19.3B5.B.7B3.3B.2B7.A34.4BA5B3A
2B10.9BA5B$70.3A16.4B.11BAB5.5B6.2A35.4BA8B4.2B4.3B2A4BA2BA$63.A9.A8.
15B4A4BA12B2.5B36.2BABA7B3.4B3.3B2A5BAB$63.3A6.2A7.15B2AB2A3BABA16B
30.A7.2BABA6B.4B3A3B.11B$66.A5.5B.B2.14B2A3BA3BA17B2A27.3A8.2BA6B.5BA
BA4B.9B$65.2A7.6B.15B2A6B3A15B2A26.A14.4B2.4BA9B2.5B$61.2A2.4B4.28B2A
2B2A14B.B27.2A12.4B3.4BA5B2A2B3.5B$62.A4.4BA19B2.2B3.B3A4B2A13B27.4B
11.4B3.5BA3BAB2A3B3.4B$62.A.AB.2BA18B11.2ABA3B2AB2.9B27.3B12.4B3.8B2A
8B2.5B$63.2AB.B2A3BA12B16.BA4B3.8B25.6B11.4B4.25B$65.6BA3B2A8B19.3B4.
7B26.6B10.4B3.2AB.B2A16B2.B$65.6BABA2BA8B21.B3.8BABA23.5B2A8.4B3.A.AB
.A2BA13B2A24.2A$64.7BA4BA2.8B19.2A2.7BA2BAB21.A4B2A2B6.4B4.A4.B2A2BAB
4A3B.BA2BA21.2B2A2B$61.AB.12B5.B3.2A20.A2.6B2A3BA20.B2A3B.BA2B4.4B4.
2A4.4BA2BA3BA2B3.2A22.6B$60.A.A9BA4BA7.A18.3A4.10BA2.2A2.2A.2A8.BABA
3B2.4B2.4B12.4BA7B27.8B$35.2A24.A2B.8B3.BA8.3A15.A4.4B.8B2.A.A2.A.A8.
4B2.B4.8B15.4B2ABA27.3B2A7B$36.A25.B.10B2.A11.A20.2A4.B3A2BA3.B.2A3.A
6.4B9.2BCBCB3.A13.4BAB22.2A3.2A2B2A8B3.2A$34.A28.10B4.3A30.A4.5B2A2.
2B2.A.2A6.4B11.B2CB3.A.A6.A5.2BABAB21.A2BAB.2A12B.BA2BA$18.A13.4A26.
11B6.A27.3A6.B3A2B.2BAB.A.A6.4B11.3BC2B.2BAB4.3A7.2A24.2A20B2A$18.3A
10.A4.A24.12B34.A8.B3A4BA.A.A.A5.4B11.8B.2B4.A5.2A26.B2.24B2.B$21.A8.
A.5A23.ABAB3.7B42.8B.2A.2A5.4B11.4B2.4B.3B2.2A4.A26.32B$7.2A11.2A8.A.
3B2.2A20.B2AB4.6B44.5B12.4B4.B.9B3.12B.BA.A26.5B2.18B2.5B$6.A.A11.4B
5.2A.ABABA2.A18.3BA5.6B44.6B10.4B.14B4.10B3.B2A28.4B3.7B2A3BA3B3.4B$
6.A15.BAB2.2BA.BABA.A.2A17.4B6.7B42.6B4.5BA2BA35B30.5B3.3B2A6B2AB3.5B
$2A2.2A.4A10.2ABA.A3B2A.A2.A19.4B7.5B2AB2.2A.A35.7B2.5BA3B2A34B31.5B
2.3B2A5BAB2A2.5B$A2.A.B.A2.A7.2A.5B.2BA.B.A.2A18.4B6.3BAB4A3BA.2A36.
6B2.5BA3BA34B30.2A7B.2BA6BABA.A8B$2.2A.A3B2.2A5.A7BA2.2A.A.A19.4B7.B
2A2B2A4B41.7B.5BA3BA12B2C20B5.2A22.B2A8B.2B2A4B2A.3A8B$3.A.3B3A2.A5.
2A5B4.A.A.A18.4B8.3B3A5B40.15BA2BA12B2C20B5.A23.12B3.4B3.A2B2A7B$3.A.
2A2B2.2A.A4.2B2A2BA4.A2.A18.4B9.9BA41.17BAB2.32B2.BA.A22.9B3AB4.2B4.
3A10B$4.A3.ABA3.A5.2B3AB5.2A18.4B10.6BABAB40.17B6.4B.10B2.B3.B.9B.B2A
21.3B2A6B2ABA10.15B$5.3A.B.3A5.8B23.4B11.6B2AB37.B3.15B7.4B5.6B9.11B
23.3B2A9BAB8.16B$7.ABABA6.4B2.4B21.4B12.2B3A2B2A36.2AB.15B7.4B8.3B10.
11B22.2A9BA6B6.6BA9B2A$8.3A6.4B4.4B19.4B14.B2A3B2A36.2A18B5.2C2B11.B
8.2AB.9B22.2A9BA6B6.6BA9B2A$8.BAB5.4B6.4B17.4B14.2BA5B38.B.3B2A12B4.C
BCB12.2A6.A.AB2.7B24.16B8.BA9B2A3B$9.B.B3.4B8.4B15.4B9.2A5.B2ABAB42.
2B2A11B4.3BC14.A6.A5.8B23.15B10.AB2A6B2A3B$10.8B10.4B13.4B11.A5.2B3AB
43.2B2.10B3.4B14.A6.2A4.8B26.10B3A4.2B4.B3A9B$8.BA7B12.4B11.4B12.A.AB
2.3BA2B42.2B3.6B.B4.4B15.2A11.7B28.7B2A2BA3.4B3.12B$7.BABA5B14.4B9.4B
14.2AB2.7B34.2A4.B2AB2.4B7.4B28.11B25.8B3A.2A4B2A2B.8B2AB$7.2B2A5B15.
4B7.4B17.3B.6B32.2B2A2B3.2A3.2B2AB5.4B30.11B25.8BA.ABA6BA2B.7B2A$5.2A
B.8B15.4B5.BA2B18.9B33.6B10.2A5.4B31.11B29.3B2.2ABA5B2A3B2.5B$4.A.AB
3.7B15.4B3.2A2B20.8B32.8B15.4B32.11B2.2A2.2A.2A9.A6.4B3.B2A6B2A3B3.5B
$4.A5.2A2B.4B15.4B.2B2A22.8B29.7B2A3B12.4B31.2AB2.8B2.A.A2.A.A8.3A6.
2A4.3BA3B2A7B3.4B$3.2A5.2A4.4B15.7B24.7B23.2B3.8B2A2B2A4.2A4.4B31.A.A
B3.3B3AB3.B.2A3.A6.A10.A3.18B2.5B$17.4B15.5B26.6B7.2A13.5B.12B2A4.A4.
4B32.A6.2BA3BA2.2B2.A.2A7.2A6.3A3.26B$18.4B14.5B26.7B6.A15.19B.BA.A3.
4B32.2A7.6B.2BAB.A.A6.4B6.A4.23B2.B$19.4B12.7B25.7B3.BA.A12.A2.19B.B
2A3.4B42.A7BA.A.A.A5.3B12.4B.17B2A$20.B3A10.4B.4B24.8B2.B2A12.ABA22B
4.4B43.B5A2B.2A.2A5.4B11.4B2.12B2A.BA2BA$21.A3B8.4B3.3BA23.11B14.ABA
2B2.18B3.4B45.3BAB12.4B4.B.9B3.8B2A2B2A3.2A$22.A3B6.4B5.ABAB22.11B15.
A3B3.2B2A6B2A4B3.4B46.6B10.4B.14B5.7B2A3B$23.4B4.4B7.2A2B21.11B4.A10.
5B3.BABA4B3A3B3.4B46.6B4.26B7.8B$24.4B2.4B9.4B21.11B2.A.A10.5B2.BA6BA
2BA2B2.4B47.7B2.28B7.6B$25.8B11.4B23.7B3.A.A8.9B.7B3A2B.7B2A45.6B2.
28B7.2B2A2B$19.2A5.6B13.4B21.8B2.2A.3A5.7B3AB.ABA4B2AB.8B2AB44.7B.22B
2A4B9.2A$18.A2.A4.7B13.4B19.8B4.B4.A4.7BA3BA3.4B3.2BA9B43.3B2A26B2A4B
$17.A.A.A4.7B14.4B19.7B2.ABAB3A4.9BA3B4.2B4.2AB2A8B42.3B2A14B2.16B$
17.A.A.2A2.7B2A14.4B17.6B3A.2BA.A4.14BA10.2AB2A5B2A3B39.BA15B6.4B.10B
$15.2A.A.B.A2B.2B2AB.2A15.4B16.6BA6B6.16B8.2B3A6B2A3B35.B3.A2BA11B7.
4B5.6B$15.A2.A.2A4B.ABAB19.4B15.11B7.2A16B6.16B2A33.2AB.5BA9B7.4B8.3B
$12.2A.A.ABAB.A2B2.BAB21.4B12.2AB.3BABA3B.B5.2A16B6.16B2A33.2A18B5.4B
11.B$12.A2.ABABA.2A5.4B20.4B10.A.AB2.7BA4B4.3B2A6B3A2B8.16B35.B.3BA
13B4.4B12.2A$13.2A2.3B.A8.2A21.4B9.A5.A2B3A7B3.3B2A5B2AB2A10.A14B38.
2B2A3BA7B4.4B14.A$15.5A.A8.A23.4B7.2A5.14B4.8B2AB2A4.2B4.3BA9B41.2B2.
ABA7B3.4B14.A$15.A4.A10.3A21.4B13.B2A11B5.9BA2B3.4B3.A3BA7B41.2B3.6B.
B4.3AB15.2A$16.4A13.A22.4B11.3BA11B5.B2A8B.B2A4BABA.B3A7B40.B2AB2.4B
7.3BA$17.A39.4B9.14B.B2A4.2A7B.2B3A7B.9B42.2A3.2B2AB5.3BA$15.A42.4B7.
4B2.5B5.BA.A5.5B2.2BA2BA6BAB2.5B51.2A5.4B$15.2A42.3BA5.4B3.6B7.A4.5B
3.3B3A4BABAB3.5B40.2A14.4B$60.3BA3.4B4.6B7.2A3.4B3.4B2A6B2A2B3.3BA38.
2B2A2B11.4B$61.3AB.4B4.6B12.5B2.18B2.2BABA37.6B10.4B$62.7B.B2.3BA3B
12.29BABA36.8B8.4B$63.12BABA2B13.B2.24B2.A35.7B4AB5.4B$63.12BABA2B16.
2A22B32.2B3.8B5AB3.4B$63.13BA3B15.A2BAB.2A12B.5B30.5B.12B2A2.4B$61.4B
.12B.B2A14.2A3.2A2B2A8B3.2B32.19B.4B$61.2A4.10B2.BA.A19.3B2A7B35.A2.
23B4.A$62.A5.7B7.A21.8B36.ABA23B3.3A$59.3A7.2BA3B7.2A21.6B37.ABA2B2.
5B3ABA8B3.A$59.A10.2B2A31.2B2A2B38.A3B3.3BAB2ABA7B4.2A$71.2A2B32.2A
40.5B3.2BA2BABA6B3.4B$72.4B74.5B2.2B2ABA8B2.3B$73.4B71.9B.2B3AB3A3B.
7BAB$74.4B69.11B.2B5A3B.8B2AB$75.4B68.7B3A2B3.4B3.10B2A$76.4B66.7BA2B
2AB4.2B4.10B2AB$77.4B63.9B2A2B2A10.BA8B2A3B$78.4B62.9B6AB8.B2A8B2A3B$
79.3BA60.2A13BA2B6.2B2AB3A8B2A$80.ABAB59.2A8B3AB2A2B6.2BA13B2A$81.2A
2B59.3B2A8B2AB8.B6A9B$82.4B58.3B2A8BAB10.2A2B2A9B$83.4B59.B2A10B4.2B
4.B2A2BA7B$84.4B59.2A10B3.4B3.2B3A7B$85.4B58.B2A8B.3B5A2B.11B$86.4B
58.BA7B.3B3AB3A2B.9B$87.4B59.5B2.8BAB2A2B2.5B$88.4B57.5B3.6BABA2BA2B
3.5B$89.BABA56.4B3.7BAB2ABA3B3.3BA$90.B2AB54.5B2.8BAB3A5B2.2BABA$91.A
3B53.29BABA$92.4B53.B2.24B2.A$93.4B55.2A22B$94.4B53.A2BAB.2A12B.5B$
95.4B53.2A3.B5A8B3.2B$96.4B58.B4A7B$97.4B59.8B$98.3BA59.6B$99.3BA58.
2B2A2B$100.3AB59.2A!

Maybe there's some spark yet to be found that can manage this kind of merging trick, all from one side, hitting a BFx59H from the side with no conduits and ending up with a B in the right place -- while also allowing other signals to connect to the BFx59H?

[Extrementhusiast]

Mostly Random Tangent -- Don't Expect Relevance
In ancient times (2004) there were several known ways of doing 2G->H conversions, including one that could be used as a merge circuit. Here's an early p39 gun showing the reaction:

Code: Select all

RLE

Maybe there's some spark yet to be found that can manage this kind of merging trick, all from one side, hitting a BFx59H from the side with no conduits and ending up with a B in the right place -- while also allowing other signals to connect to the BFx59H?

I had found this R+glider collision a while back that is perhaps an even more general inserter:

Code: Select all

x = 33, y = 25, rule = LifeHistory
15.2A14.2F$15.2A14.2F10$10.C18.3D$10.2C19.D$9.2C19.3D2$.2A$A.A$2.A4$
10.2D$10.D$11.3D$13.D!

Unfortunately, the only way I found to insert the R was from the ancient G2/5/6-to-R, and then the extra glider must also be edgeshot. (The indicated eater merely suppresses a large spark that may otherwise interfere with future conduits.)

[MathAndCode]

I had found this R+glider collision a while back that is perhaps an even more general inserter:

Code: Select all

x = 33, y = 25, rule = LifeHistory
15.2A14.2F$15.2A14.2F10$10.C18.3D$10.2C19.D$9.2C19.3D2$.2A$A.A$2.A4$
10.2D$10.D$11.3D$13.D!

If we can catalyze that spark to form a boat (or probably another object) in the right place at the right time, then that gives us an R+G→∏.

Code: Select all

x = 33, y = 18, rule = LifeHistory
15.2A$15.2A10$10.C18.3D$10.2C19.D$9.2C19.3D2$.2A$A.A$2.A!
#C [[ PASTEMODE COPY PASTET 56 PASTE 2C$C.C$.C! 28 17 ]]

[MathAndCode]

Here's a C→∏.

Code: Select all

x = 10, y = 7, rule = B3/S23
2b2o$3o$bo$6b2o$6bobo$8bo$8b2o!

[MathAndCode]

Figuring out how to remove that extra block would yield a new C→G.

Code: Select all

x = 23, y = 13, rule = B3/S23
13b2o$12bobo$13bo3$bo$2bo$3bo$2bo$2o2$21b2o$21b2o!

[wwei23]

Figuring out how to remove that extra block would yield a new C→G.

A blocker works no matter how the century is timed.

Code: Select all

x = 33, y = 229, rule = B3/S23
4b2o$2obo2bob2o$2o2bo4bo$5bo$6bobo2$23b2o$22bobo$23bo3$11bo$12bo$13bo$
12bo$10b2o2$31b2o$31b2o12$4b2o$4o4b2o$3ob2o2b2o$5bo3$23b2o$22bobo$23bo
3$11bo$12bo$13bo$12bo$10b2o2$31b2o$31b2o12$b4o$o7b2o$o3b2o2b2o$bo2b2o
3$23b2o$22bobo$23bo3$11bo$12bo$13bo$12bo$10b2o2$31b2o$31b2o11$2b2o$b3o
$obo2bo2b2o$2o2b2o2b2o$4b2o3$23b2o$22bobo$23bo3$11bo$12bo$13bo$12bo$
10b2o2$31b2o$31b2o11$bobo$4bo$o4bo2b2o$2obo2bob2o$4b2o3$23b2o$22bobo$
23bo3$11bo$12bo$13bo$12bo$10b2o2$31b2o$31b2o12$4bo$2o2b2ob3o$2o4b4o$4b
2o3$23b2o$22bobo$23bo3$11bo$12bo$13bo$12bo$10b2o2$31b2o$31b2o12$4b2o2b
o$2o2b2o3bo$2o7bo$5b4o3$23b2o$22bobo$23bo3$11bo$12bo$13bo$12bo$10b2o2$
31b2o$31b2o12$4b2o$2o2b2o2b2o$2o2bo2bobo$6b3o$6b2o2$23b2o$22bobo$23bo
3$11bo$12bo$13bo$12bo$10b2o2$31b2o$31b2o!

[dvgrn]

Figuring out how to remove that extra block would yield a new C→G.

Code: Select all

x = 23, y = 13, rule = B3/S23
13b2o$12bobo$13bo3$bo$2bo$3bo$2bo$2o2$21b2o$21b2o!

This is an idea based on the known BSE22T31 conduits, right?

Code: Select all

x = 94, y = 96, rule = LifeHistory
27.4D3.3D2.5D2.3D3.3D2.5D2.3D4.D$27.D3.D.D3.D.D5.D3.D.D3.D3.D3.D3.D2.
2D$27.D3.D.D5.D9.D5.D3.D7.D3.D$27.4D3.3D2.3D6.D5.D4.D5.2D4.D$27.D3.D
5.D.D7.D5.D5.D7.D3.D$27.D3.D.D3.D.D6.D5.D6.D3.D3.D3.D$27.4D3.3D2.5D.
5D.5D3.D4.3D3.3D10$54.5B$53.8B$52.10B$52.11B$35.2A6.B2.2B3.12B.B$34.B
2AB4.22B2A$34.3B4.23B2A$35.B5.19B.4B$33.4B3.21B.3B$33.21B.6B$33.4BD
15B3.4B$33.5BD14B4.4B$32.6B2D13B5.4B$31.6B2D14B6.4B$32.5BD15B7.4B$32.
22B.BA4.4B$36.B4.B2A11BA.A4.4B$42.2A11B.2A5.4B$46.8B10.4B$47.6B12.4B$
48.3B15.3B$67.2B$68.B19$4D2.D3.D.D3.D9.D4.3D2.5D2.3D9.3D2.5D2.3D3.3D
2.5D2.3D4.D$D3.D.2D2.D.D3.D8.2D3.D3.D3.D3.D3.D7.D3.D.D5.D3.D.D3.D3.D
3.D3.D2.2D$D3.D.D.D.D.D3.D9.D7.D3.D3.D3.D7.D5.D9.D5.D3.D7.D3.D$4D2.D
2.2D.D.D.D.5D3.D6.D4.D4.4D8.3D2.3D6.D5.D4.D5.2D4.D$D3.D.D3.D.D.D.D9.D
5.D5.D7.D11.D.D7.D5.D5.D7.D3.D$D3.D.D3.D.2D.2D9.D4.D6.D7.D7.D3.D.D6.D
5.D6.D3.D3.D3.D$4D2.D3.D.D3.D8.3D2.5D3.D4.3D2.5D2.3D2.5D.5D.5D3.D4.3D
3.3D10$54.5B$53.8B$52.10B$52.11B$43.B2.2B3.12B.B$42.22B2A$35.2A4.23B
2A$34.B2AB3.19B.4B$24.B9.4B2.21B.3B$24.2B9.19B.6B$24.3B7.3BD15B3.4B$
24.4B.9BD14B4.4B$25.13B2D13B5.4B$26.11B2D14B6.4B$27.10BD15B7.4B$28.
26B.BA4.4B$29.6B.B4.B2A11BA.A4.4B$30.3B9.2A11B.2A5.4B$46.8B10.4B$47.
6B12.4B$48.3B15.3B$67.2B$68.B!

Just mention the prior art you're starting from, when you can. It tends to make things a lot less confusing.

[MathAndCode]

A blocker works no matter how the century is timed.

A unix also works with any timing.

Code: Select all

x = 32, y = 23, rule = TripleB3S23
2.2G$A.FG2$2.D$AGBG$.FC.G2.2G$3.CBD.FG$3.FG$4.A2.A2$22.2G$21.G.G$22.G3$10.G$11.G$12.G$11.G$9.2G2$30.2G$30.2G!

Code: Select all

x = 32, y = 23, rule = TripleB3S23
2.2G$.FAEB$DFEA$DFA$.G.G3.B$2.G.G.AEG$3.G.AEAG$4.G3F$5.2D2$22.2G$21.G.G$22.G3$10.G$11.G$12.G$11.G$9.2G2$30.2G$30.2G!

This is an idea based on the known BSE22T31 conduits, right?

Actually, what happened is that while searching for conduits accepting the century, I (independently) noticed that the century has an internal glider that later gets destroyed.

Code: Select all

x = 18, y = 23, rule = Immigration
2A$2A7.2A$8.A$11.A$9.2A2$2A$2A6.A$7.A.2A$7.A4.A$8.A4.A$2A8.B$2A7.2B2.A2.A$10.2B2.A2.A$16.A3$9.2A2.2A$9.2A2.2A$13.A$13.A2.A$13.A.A$13.2A!

I then looked for ways to liberate that glider. I found that a block gets rid of the debris that the glider crashes into and deletes the blinker. I then tried to clean up the three remaining blocks, and I was able to use a boat to suppress two of them. I checked all of the century→glider conduits in the elementary conduits collection, and none liberated that internal glider, but I don't know how I would check for conduits that made that century by consuming the extra block.
Speaking of transparent block reactions, for which common objects have ways been searched to place a block (and possibly a beehive) in the path of those objects so that the object is consumed and then regenerated?

[dvgrn]

Speaking of transparent block reactions, for which common objects have ways been searched to place a block (and possibly a beehive) in the path of those objects so that the object is consumed and then regenerated?

It's safe to assume that R's, B's, Herschels, pi-heptominoes and gliders have had fairly extensive transparent-object searches done. But a lot of that happened in a big push in the mid-1990s, mostly by Paul Callahan. So there is definitely search space that was not reachable in practice back then, that would be reachable now -- even with the same exact ptbsearch code.

For example, not many searches have been done that involve two simultaneous transparent objects. In the 1990s that would have taken too much time and memory, but with the right limits set it might turn up something interesting now. Kazyan might have done some searching along these lines -- maybe he can say more about the boundaries of recent searches.