The Hunting of the Elementary Conduits

[MathAndCode]

I prefer looking for conduits that produce it first, and then building a receiving converter after the first one.

Do you mean for the purpose of ensuring that it is practical or for the purpose of ensuring that we know which way the object would come in?

[bubblegum]

I prefer looking for conduits that produce it first, and then building a receiving converter after the first one.

Do you mean for the purpose of ensuring that it is practical or for the purpose of ensuring that we know which way the object would come in?

Practicality.

[MathAndCode]

Practicality.

Ah; thank you for clarifying.
Of course, it's possible that it's popped up several times, and we just don't know because no one knew to look for it. I've mentioned this before, and I'm considering doing a soup search and tracking which active regions are the most common in order to get around this catch-22.

[Extrementhusiast]

Practicality.

Ah; thank you for clarifying.
Of course, it's possible that it's popped up several times, and we just don't know because no one knew to look for it. I've mentioned this before, and I'm considering doing a soup search and tracking which active regions are the most common in order to get around this catch-22.

This has been done before.

[Hunting]

Practicality.

Ah; thank you for clarifying.
Of course, it's possible that it's popped up several times, and we just don't know because no one knew to look for it. I've mentioned this before, and I'm considering doing a soup search and tracking which active regions are the most common in order to get around this catch-22.

This has been done before.

This has been done two times.

[MathAndCode]

This has been done before.

This has been done two times.

Thank you for informing me. This certainly is very interesting. For example, the I-sequence is indeed over twice as common as the dove sequence (as well as us having more conduits for making it and more conduits for turning it into something besides a glider), yet the latter but not the former is included in the conduits collection. As another example, contrary to dvgrn's assumptions, block-tie-blinker is fairly common (over one and a half times as common as the dove sequence). I would happily find conduits that accept some of the more active and common active regions that haven't been investigated much yet if someone can tell me which direction they typically come from or give me examples of them being made by conduits.

[bubblegum]

For example, the I-sequence is indeed over twice as common as the dove sequence (as well as us having more conduits for making it and more conduits for turning it into something besides a glider), yet the latter but not the former is included in the conduits collection.

For this, we need to analyse two things.
1) The alternate heptominoes (E,I) never received as much attention as the dove until you came along with your belief that you can come along and overhaul everything we've been doing for the past two decades. (It's actually possible, so go ahead.)
2) The dove spreads out, leaving room for catalysts to modify it easily. The I really hates its blinker and tub and drops them as soon as possible, meaning conduits have to throw it back into its debris.

[MathAndCode]

1) The alternate heptominoes (E,I) never received as much attention as the dove until you came along with your belief that you can come along and overhaul everything we've been doing for the past two decades. (It's actually possible, so go ahead.)

I am not trying to overhaul the entire catalyst system. For example, I have been using the standard conduit nomenclature, and I demonstrate the connectability of the conduits that I find by connecting them to other conduits that have already been found. I merely think that some active regions have been underinvestigated and do not like Herschel-centrism. Also, part of the reason why I tend to investigate active regions that have not already been extensively investigated is so that the conduits that I find will have an increased chance of being new.

2) The dove spreads out, leaving room for catalysts to modify it easily. The I really hates its blinker and tub and drops them as soon as possible, meaning conduits have to throw it back into its debris.

You can say that the dove should be more useful as an active region than the I-sequence, but the fact of the matter is that we know of three stable conduits that make an I-sequence but only one stable conduit that makes a dove sequence, and while we have three stable conduits that accept each, we have two stable conduits that turn an I sequence into something besides a glider and no stable conduits that turn a dove sequence into something besides a glider. It seems that so far, the I sequence has proven more useful conduit-wise than the dove sequence.



Edit: I found a reaction that suggests a partial ∏→E.

Code: Select all

x = 9, y = 21, rule = LifeHistory
.A$2.2A$.2CD2.2A$.D4.A.A$.3D3.2A13$.D$3D$D.D$2.D!

The region near the E might also be useful for a loafer gun that doesn't completely rely on a glider synthesis.



Another edit: While trying to complete the ∏→E, I found something even more interesting.

Code: Select all

x = 12, y = 14, rule = LifeHistory
A$3A$3.A$2.2A6$10.2A$10.2A$5.3C$5.C$5.3C!

[bubblegum]

2) The dove spreads out, leaving room for catalysts to modify it easily. The I really hates its blinker and tub and drops them as soon as possible, meaning conduits have to throw it back into its debris.

You can say that the dove should be more useful as an active region than the I-sequence, but the fact of the matter is that we know of three stable conduits that make an I-sequence but only one stable conduit that makes a dove sequence, and while we have three stable conduits that accept each, we have two stable conduits that turn an I sequence into something besides a glider and no stable conduits that turn a dove sequence into something besides a glider. It seems that so far, the I sequence has proven more useful conduit-wise than the dove sequence.

No, it's because there are more people who care about the I (like 2) than the dove.

[MathAndCode]

No, it's because there are more people who care about the I (like 2) than the dove.

I also investigated the dove for stable conduits that accept it, and I did not find any.

[wwei23]

Another edit: While trying to complete the ∏→E, I found something even more interesting.

Code: Select all

x = 23, y = 14, rule = LifeHistory
11.A$11.3A$14.A$13.2A6$2A19.2A$A.A18.2A$.2A13.3C$16.C$16.3C!

I also investigated the dove for stable conduits that accept it, and I did not find any.

Does a beehive basically on top of where the dove is supposed to be considered an output?
At the same time, I don't remember being able to get anything starting from an I either.

[MathAndCode]

Code: Select all

x = 23, y = 14, rule = LifeHistory
11.A$11.3A$14.A$13.2A6$2A19.2A$A.A18.2A$.2A13.3C$16.C$16.3C!

That's impressive how you were able to find one object that gets rid of everything else. However, unless you know of a way to regenerate the block, ship, and fishhook quickly enough so that the next ∏ can come with repeat time at most 300, we're going to have to find another way to make this work. Replacing the fishhook or adding another catalyst might be feasible, but I highly doubt that the block can be replaced with a catalyst that can perform both catalyses without getting destroyed, especially a catalyst small enough to allow the ∏ to get where it needs to be, considering that it already looks difficult.
Of course, I realized that getting the ∏ in there in the first place would be difficult earlier, but I posted the reaction anyway since we don't have any good ways to make LWSSes for conduits. However, letting the block be transparent, which we now realize is necessary, makes getting the ∏ in much easier. I see three main possibilities: there might be a way to create an active region that edgeshoots a ∏ with trailing junk that settles into a block (or a beehive) in the right place, and I suspect that there's some collection of near-miss conduits that make active objects with a residual object or two behind them, but this particular block and ∏ combination might even be in an E-making conduit; we could try to make the ∏ and the block separately with synchronized signals, although I'm concerned about the crowding if we try that; or we could just use the method that I found to make the ∏ and block in the first place, namely shooting a glider at a ship (or boat). Of course, the fishhook would get in the way of the glider, but I don't think that there's a way to prevent it from getting destroyed even if we manage to insert the ∏ via another mechanism, so we should replace it with a transparent block anyway.

Code: Select all

x = 21, y = 34, rule = DoubleB3S23
6.2C$6.2C$11.C$10.2C$10.C.C$11.C12$2.C$C.C$.2C12$18.2C$18.C.C$19.CB!

The honey farm predecessor colliding with a block is an example of a block insertion reaction, but it doesn't have to be the one used. The block will have to be edgeshot, but only by two diagonals, which I'm sure is feasible. The block insertion and glider will have to be synchronized, but we have a window of about forty generations, which is probably enough. I'm not very concerned about the ship/boat. If nothing else, the Herschel edgeshoots a ship even without any catalysts. (We would need to clean up the blocks somehow, but there's already a widely-known method for suppressing the first block.) The block factory used in this H→∏ has enough clearance for this situation, so I'll start trying to synchronize it with a glider soon unless someone gives me a more suitable block factory before then.



Edit: Unfortunately, this doesn't work.

Code: Select all

x = 59, y = 104, rule = LifeHistory
16.2E.E$16.E.2E2$17.5E$12.2E2.E4.E$12.E2.E.BEB$13.E.E.2EB$12.2E.E.4BE
$15.E3.BEBE$15.2E2.E2BE$20.2E3B$22.7B$22.9B$22.10B$22.11B$18.14B$18.
14B$17.16B$17.17B$17.11B2E4B$18.10B2E9B$18.24B$18.25B$19.24B$20.24B$
19.26B$20.26B$18.29B$18.2E.21B2E4B$19.E3.18BE2BE3B$16.3E5.9B.8B2E4B$
16.E7.8B3.4B2.6B$26.5B5.4B2.B$26.6B5.4B$26.7B4.4B$27.2B.4B3.5B2.B$25.
4B2.4B.6B.B2E$25.2E5.12B2EB$26.E6.9B.3B$23.3E8.8B.2B$23.E10.8B.4B.2E.
E$24.2E8.14BE.2E$25.E9.13B$25.E.EB4.15B$26.2EB.17B$28.20B$28.19B$29.
17B.2B$25.2B.20B2E$24.22B.B2E$23.22B3.B$23.3BC17B$23.3BCBC16B$22.4B3C
17B$9.A12.6BC17B6.2E$9.3A10.25B5.E$12.A9.26B.BE.E$11.2A12.23B.B2E$19.
2B4.26B$27.23B$3.2A20.25B$3.A21.2E.B.18B5.A$2A.A22.E4.17B4.A.A$A2.3A
4.2A11.3E5.16B6.A$.2A3.A3.2A11.E7.17B$3.4A23.20B7.2D$3.A15.2A9.21B6.
2D$4.3A12.A.A6.2EB.20BD$7.A13.A5.E.EB3.18BDBD$2.5A14.2A4.E5.19B3DB$2.
A23.2E4.22BD$4.A28.20B$3.2A27.16B$32.15B$33.12B$34.7B.3B$34.6B$32.2EB
.3B$31.E.EB3.2B$31.E5.B2EB$30.2E6.2E2$36.4A$35.A3.A$35.2A7$25.2A$24.A
.A5.2A$24.A7.2A$23.2A2$37.A$33.2A.A.A$32.A.A.A.A$29.A2.A.A.A.A.2A$29.
4A.2A2.A2.A$33.A4.2A$31.A.A$31.2A!

I shall keep looking.

[bubblegum]

No, it's because there are more people who care about the I (like 2) than the dove.

I also investigated the dove for stable conduits that accept it, and I did not find any.

Here's some proof-of-concept work:

Code: Select all

x = 24, y = 20, rule = B3/S23
b2o$o2bo18b2o$bo2bo17b2o$2b3o10$12b2o$12bobo$14bo$14b2o2$2o$2o!

We want something other than dove-to-inaccessible-B-plus-lots-of-junk, though.

(This looks more promising than that pi-to-LWSS though.)

[MathAndCode]

Here's some proof-of-concept work:

Code: Select all

x = 24, y = 20, rule = B3/S23
b2o$o2bo18b2o$bo2bo17b2o$2b3o10$12b2o$12bobo$14bo$14b2o2$2o$2o!

We want something other than dove-to-inaccessible-B-plus-lots-of-junk, though.

(This looks more promising than that pi-to-LWSS though.)

I have also found many promising interactions with the I sequence that have not yielded actual conduits. I get your point, and if more dove conduits are been found in the future, this will change, but right now, doves are not very useful to conduits. They do show promise, but they are currently not useful in conduit making.
By the way, blonk-tie seems promising. So far today, I have tried one catalyst placement, and I got a clean ∏.

Code: Select all

x = 8, y = 11, rule = B3/S23
5b2o$5b2o$7bo$7bo$7bo3$2b2o$bobo$bo$2o!

Also, it produces a B without any catalysts, so maybe that could be made into a B-producing or H-producing conduit.

[Hunting]

Here's some proof-of-concept work:

Code: Select all

x = 24, y = 20, rule = B3/S23
b2o$o2bo18b2o$bo2bo17b2o$2b3o10$12b2o$12bobo$14bo$14b2o2$2o$2o!

We want something other than dove-to-inaccessible-B-plus-lots-of-junk, though.

(This looks more promising than that pi-to-LWSS though.)

I have also found many promising interactions with the I sequence that have not yielded actual conduits. I get your point, and if more dove conduits are been found in the future, this will change, but right now, doves are not very useful to conduits. They do show promise, but they are currently not useful in conduit making.
By the way, blonk-tie seems promising. So far today, I have tried one catalyst placement, and I got a clean ∏.

Code: Select all

x = 8, y = 11, rule = B3/S23
5b2o$5b2o$7bo$7bo$7bo3$2b2o$bobo$bo$2o!

Also, it produces a B without any catalysts, so maybe that could be made into a B-producing or H-producing conduit.

Code: Select all

x = 13, y = 17, rule = B3/S23
2o$bo$bobo$2b2o2$12bo$12bo$12bo$10b2o$2o8b2o$o2bob2o$2b2ob2o2$2b2ob2o$
bobobo$obo4bo$2o4b2o!

[MathAndCode]

Code: Select all

x = 13, y = 17, rule = B3/S23
2o$bo$bobo$2b2o2$12bo$12bo$12bo$10b2o$2o8b2o$o2bob2o$2b2ob2o2$2b2ob2o$
bobobo$obo4bo$2o4b2o!

Active regions, especially those that already have known catalysts accepting them, are typically more useful than stationary objects, especially stationary objects that are not edgeshoot.

Also, here are two Bs that I was unable to extract from the dove.

Code: Select all

x = 27, y = 33, rule = LifeHistory
10.2A$9.A2.A$10.A2.A11.2A$11.3A11.2A6$2D.D$.3D$2.D19$17.2D.D$18.3D$19.D!

Edit: I have seen several close calls with my work on the LWSS-making conduit. Here is the latest.

Code: Select all

x = 74, y = 84, rule = LifeHistory
12.A$11.A.A$12.2A11$24.B$17.A5.3D$16.A.A4.BDB$17.A4.3D2B$22.5B$21.6B$
21.6B$22.5B$22.6B$21.6B$21.7B$22.6B$22.6B$4.A17.2B2AB$3.A.A15.3B2A2B$
3.A.A15.7B$.3A.2A15.5B$A4.B6.3B6.B.4B10.2A$.3AB2A4.4B5.6B10.A.A$3.A.
2A2.19B8.A$5.3B.19B6.3A$7.23B2.BAB2.2A$6.25B.B4A2.A$6.23B2D2B.B2.A.A$
6.22B2D3B2AB2A.2A$7.22BD2B.AB5.A$8.B2.20B3.2A.2A$11.19B5.A.A$11.8B3.
6B7.A.A$11.16B9.A32.2A$11.14B44.A$11.12B48.A$11.12B28.2A14.5A$12.2B.
4B3DB29.A13.A$16.4BD2B29.A.A12.3A$16.2B3D2B30.2A15.A$17.6B44.4A$17.7B
38.2A3.A3.2A$19.B.4B37.2A4.3A2.A$22.4B44.A.2A$23.3B44.A$24.2B2A41.2A$
25.BA.A$28.A9.A$28.3A5.3A22.2A$31.A3.A25.A$30.2A3.2A25.3A$64.A2$43.C$
43.3C$18.2A23.C.C$18.2A25.C2$9.2A.A7.D$9.A.2A6.3D$18.2D.D4$15.2A$14.A
.A$14.A$13.2A26.2A$41.A$35.2A.A3.3A$35.2A.4A2.A$41.A$35.2A.3A$34.A2.
2A$34.2A!

[Scorbie]

Just three small interrelated historical context for the Herschel-centric theme (Sorry for being offtopic, just this post)
I don't mind you disagree with minor facts BTW (feel free to correct me), I'm just suggesting the big picture.

1. People have been looking through Herschels since a long time ago.
The first "conduit"s was B/R/H based. The field of "Stable signal circuitry" was probably made because of the H conduits and the first signal reflector(1996, I think). For quite a long time "Stable signal circuitry" just meant "Herschel circuitry" because these were the only ways to make reasonable-sized stable circuits.
'
By the way, This was where I used to hang out before finding the ConwayLife Forum :)
On the "Selected Topics" you can see some fun articles in the past era of Life.

2. It was supported by a circuit assembling program called Hersrch. (Used for making guns in the pre-Syringe era)
Actually I'd rather say the elementary conduits aren't supported enough, I guess. I think there is a chance for shrinking some of the guns with an automated run by an automated tool, but nobody has done that feat yet. (And I guess I have some responsibility over that too)

3. Herschel is easily found in random ash, which led people to explore things. (Which I guess led to 1 and 2 and so on.)
Yes I think there are selection bias, yes, I feel this looks like QWERTY, yes, this seems like a positive feedback loop, but that was how it went I guess, and it worked well enough for a gazillion of circuitry you see now.

(And 4. dvgrn maintains the conduit collections, but that's (hopefully) ongoing history, so I'll make that as a separate point.)

By the way, I think the Herschel-centralism (and the conduits in general) is fading at this point with the advent of the Syringe and the Snark.
If you find a G->I->2G or G->E->2G then I'm pretty sure people would race to find I and E conduits without any encouragement from anyone.
People aren't biased because of other reasons, they're just using the best tool for the job for what it is now.
(My guess is that the conduits are going to be used more sparingly if there pops out a more compact and fast G-to-2G)

[Hunting]

advent of the Syringe and the Snark.

Missing the +1 button and the heart button from lichess. But really, didn't syringe encourage Herschel-centralism?

[bubblegum]

advent of the Syringe and the Snark.

Missing the +1 button and the heart button from lichess. But really, didn't syringe encourage Herschel-centralism?

Yes, because it allowed glider-to-Herschels easily (the missing core part of Herschel-centrism), and no, because it solidified the single glider as a viable signal medium with its own interconversions.

[MathAndCode]

1. People have been looking through Herschels since a long time ago.
The first "conduit"s was B/R/H based. The field of "Stable signal circuitry" was probably made because of the H conduits and the first signal reflector(1996, I think). For quite a long time "Stable signal circuitry" just meant "Herschel circuitry" because these were the only ways to make reasonable-sized stable circuits.

I don't have a problem with the fact that people were looking for Herschel conduits. I have used Herschel conduits myself, e.g. here. My problem is when users practice Herschel-centrism to the detriment of non-Herschel conduits, including chaining them into composite Herschel conduits and refusing to acknowledge elementary conduits until they have been chained as such.

2. It was supported by a circuit assembling program called Hersrch. (Used for making guns in the pre-Syringe era)
Actually I'd rather say the elementary conduits aren't supported enough, I guess. I think there is a chance for shrinking some of the guns with an automated run by an automated tool, but nobody has done that feat yet. (And I guess I have some responsibility over that too)

Yes, Herschel conduits are useful for making guns, but other conduits can be used as well. As you mentioned, it's possible that running a similar search program will result in more efficient solutions at some periods.

3. Herschel is easily found in random ash, which led people to explore things. (Which I guess led to 1 and 2 and so on.)

The Herschel sequence does not occur very often besides being spawned by a B-sequence, which in turn comes from an R-sequence more often than it doesn't, but there aren't nearly as many R→R conduits as H→H conduits. Also, the ∏-sequence is about as common as the R-sequence, but again, there aren't nearly ∏→∏ conduits as H→H conduits.

[bubblegum]

I don't have a problem with the fact that people were looking for Herschel conduits. I have used Herschel conduits myself, e.g. here. My problem is when users practice Herschel-centrism to the detriment of non-Herschel conduits, including chaining them into composite Herschel conduits and refusing to acknowledge elementary conduits until they have been chained as such.

I agree that the second is a (albeit unsolvable due to practicality reasons) problem, but the first sounds like you're just prejudiced against Herschels because they're the standard.

Yes, Herschel conduits are useful for making guns, but other conduits can be used as well. As you mentioned, it's possible that running a similar search program will result in more efficient solutions at some periods.

Yes, but we need to know what reactions are useful, and Herschels have shown themselves reliably usable (just look at a handful of the variable guns).

The Herschel sequence does not occur very often besides being spawned by a B-sequence, which in turn comes from an R-sequence more often than it doesn't, but there aren't nearly as many R→R conduits as H→H conduits. Also, the ∏-sequence is about as common as the R-sequence, but again, there aren't nearly ∏→∏ conduits as H→H conduits.

The Herschel instead of B tradition came from that one annoying block that has to be removed; a good way to remove it is (often the only sensible one that fits) Conduit 1 but that creates a Herschel <s>which is unacceptable</s>. The R of course also drops a bit of annoying debris which can be removed easily but that's an R->B. (The only R->R I can remember translates it backwards.) The pi (the symbol you're using by the way is the continued product symbol which is not only distinct from any of the actual pi symbols Π,π,ϖ, but is pi in the way that C is Γ, i.e. it's not pi) has an unhelpful reaction envelope which either forces the pi's catalysts to be very far out or squishes them into a space you can barely fit a pi into.

[Kazyan]

There's one major, characteristic trait of Herschels that has allowed Herschel-centrism, which I don't think has been mentioned yet: they move. They do it in a convenient way, too. A Herschel that comes out of a conduit will transports its 'explosion' in a distinct path and direction, and this path vacates the place where it first appeared, without flailing about in such a way that it is nigh-guaranteed to knock over catalysts from the conduit it came from. (Looking at you, century and Pi, but I'm also giving hard side-eye at R.) That is why Herschels are so great. For any new conduit, if the input is an H, it is almost guaranteed to be able to connect to something as long as you don't put a catalyst directly behind the Herschel. If its output in a Herschel, it has probably (but not always) formed by launching itself 20 cells away from an active region via B-heptomino. It has also probably formed in a way such that it's pointing away from your new conduit, rather than an R or C, which like to do the opposite of that.

This is part of the motivation behind the (unnecessary) packaging of elementary conduits into Herschel conduits. Once you prove that you can get back to a Herschel, you're home free to connect that to whatever you want.

The "no connections" graveyard is built on centuries, Pi heptominos, B heptminos, and Rs. It is not built on Herschels. The only signals that do better in this respect are gliders.

[MathAndCode]

I agree that the second is a (albeit unsolvable due to practicality reasons) problem, but the first sounds like you're just prejudiced against Herschels because they're the standard.

I just used Herschel conduits again. (They are composite conduits, but they were from the original universal list of Herschel conduits.)

Code: Select all

x = 72, y = 92, rule = LifeHistory
2A$A.A$.2A12$6.A5.3D$5.A.A5.D$6.A4.3D16$11.2A$11.2A2$2.2A.A7.D$2.A.2A
6.3D$11.2D.D2$22.2A$22.A$20.A.A$20.2A3$6.2A$5.A.A$5.A$4.2A3$67.2A$67.
A$69.A$18.3D28.2A14.5A$9.2A8.D30.A13.A$10.A6.3D30.A.A12.3A$7.3A41.2A
15.A$7.A57.4A$60.2A3.A3.2A$60.2A4.3A2.A$68.A.2A$25.2A41.A$25.A.A39.2A
$27.A9.A$27.3A5.3A$30.A3.A24.2A$29.2A3.2A23.A$60.3A$62.A$42.C$42.3C$
17.2A23.C.C$17.2A25.C2$8.2A.A7.D$8.A.2A6.3D$17.2D.D4$14.2A$13.A.A$13.
A$12.2A26.2A$40.A$34.2A.A3.3A$34.2A.4A2.A$40.A$34.2A.3A$33.A2.2A$33.2A!

[bubblegum]

I just used Herschel conduits again. (They are composite conduits, but they were from the original universal list of Herschel conduits.)

Code: Select all

x = 72, y = 92, rule = LifeHistory
2A$A.A$.2A12$6.A5.3D$5.A.A5.D$6.A4.3D16$11.2A$11.2A2$2.2A.A7.D$2.A.2A
6.3D$11.2D.D2$22.2A$22.A$20.A.A$20.2A3$6.2A$5.A.A$5.A$4.2A3$67.2A$67.
A$69.A$18.3D28.2A14.5A$9.2A8.D30.A13.A$10.A6.3D30.A.A12.3A$7.3A41.2A
15.A$7.A57.4A$60.2A3.A3.2A$60.2A4.3A2.A$68.A.2A$25.2A41.A$25.A.A39.2A
$27.A9.A$27.3A5.3A$30.A3.A24.2A$29.2A3.2A23.A$60.3A$62.A$42.C$42.3C$
17.2A23.C.C$17.2A25.C2$8.2A.A7.D$8.A.2A6.3D$17.2D.D4$14.2A$13.A.A$13.
A$12.2A26.2A$40.A$34.2A.A3.3A$34.2A.4A2.A$40.A$34.2A.3A$33.A2.2A$33.2A!

That sacrificial ship is painful.

[MathAndCode]

That sacrificial ship is painful.

I'm hoping to figure out a way to restore it (or a boat in the same position), but I choose to focus synchronizing the glider and the block first.



Edit: I managed to get rid of the junk except for a beehive near the ship's original position.

Code: Select all

x = 72, y = 112, rule = LifeHistory
13.2A5.2A$13.2A5.2A17$26.2A$26.2A$2A$A.A$.2A12$6.A5.3D$5.A.A5.D$6.A4.
3D16$11.2A$11.2A2$2.2A.A7.D$2.A.2A6.3D$11.2D.D2$22.2A$22.A$20.A.A$20.
2A3$6.2A$5.A.A$5.A$4.2A3$67.2A$67.A$69.A$18.3D28.2A14.5A$9.2A8.D30.A13.
A$10.A6.3D30.A.A12.3A$7.3A41.2A15.A$7.A57.4A$60.2A3.A3.2A$60.2A4.3A2.
A$68.A.2A$25.2A41.A$25.A.A39.2A$27.A9.A$27.3A5.3A$30.A3.A24.2A$29.2A3.
2A23.A$60.3A$62.A$42.C$42.3C$17.2A23.C.C$17.2A25.C2$8.2A.A7.D$8.A.2A6.
3D$17.2D.D4$14.2A$13.A.A$13.A$12.2A26.2A$40.A$34.2A.A3.3A$34.2A.4A2.A
$40.A$34.2A.3A$33.A2.2A$33.2A!

Is there any way to make the reaction that forms that beehive make a ship or boat instead?