Partial B conduit

[fluffykitty]

Partial R39 for B heptominos, it needs a catalyst to replace the snake and a way to replace the block:

Code: Select all

x = 16, y = 11, rule = B3/S23
6b2o5bo$2bo3b2o5b2o$b3o11bo$2obo9b2o4$7b2o$8bo$7bo$7b2o!

The environment of the problem:

Code: Select all

...AAA
...AAA
..*%..
...*..
..*...
..**..

The % cell dies, and an ordinary eater is broken. If you remove the snake, the active junk catches up to the new B and kills it. (Also, the b+block was discovered accidentally.)

[codeholic]

There is no chance to make a conduit from it, even if you find a proper catalyst instead of expendable snake, because the block is restored in the wrong place.

[fluffykitty]

Well here's a reaction where the tub turns into a boat.

Code: Select all

x = 16, y = 10, rule = Life
6.2A5.A$2.A3.2A5.2A$.3A11.A$2A.A9.2A4$8.A$7.A.A$8.A!

The block still dies, though.

[Kazyan]

An eater nearby the tub can prevent it from becoming a boat.

Code: Select all

x = 16, y = 12, rule = B3/S23
6b2o$2bo3b2o$b3o$2obo4$8bo$7bobo2b2o$8bo3bobo$14bo$14b2o!

[fluffykitty]

That's the first step in making the catalyst. Next step is restoring the block in the right place.

There is no chance to make a conduit from it, even if you find a proper catalyst instead of expendable snake, because the block is restored in the wrong place.

Well expandable snake is obsolete as of 6:58 pm yesterday! I wonder if some Bellman catalyst might do the trick. (I myself do not really know how to do searches, but I might be able to come up with an input file.)

[codeholic]

It's unlikely. Catalysts can only tweak a reaction on its boundaries, while this block is an integral part of this reaction.

[fluffykitty]

Yeah, I guess it is. But there's still the top. For example:

Code: Select all

x = 19, y = 21, rule = Life
13.A$11.3A$10.A$10.2A5$7.2A$3.A3.2A$2.3A$.2A.A4$9.A$8.A.A2.2A$9.A3.A.
A$15.A$15.2A!

Also, evolution of conduit:

Code: Select all

x = 79, y = 19, rule = Life
3$39.2A5.A$35.A3.2A5.2A19.2A5.A$10.2A5.A16.3A11.A14.A3.2A5.2A$6.A3.2A
5.2A14.2A.A9.2A14.3A11.A$5.3A11.A41.2A.A9.2A$4.2A.A9.2A2$40.2A$41.A
27.A$40.A27.A.A2.2A$40.2A27.A3.A.A$75.A$75.2A!

[Kazyan]

Another partial conduit here. Pay attention to generation 61 and for a few afterwards. (But this might already be known; I dunno.)

Code: Select all

x = 13, y = 20, rule = LifeHistory
.2A$2.A$2.A.A$3.2A2.2A$7.2A6$6.2C$6.2C$2A$2A4$10.A$9.3A$9.A.2A!

[fluffykitty]

Another partial conduit here. Pay attention to generation 61 and for a few afterwards. (But this might already be known; I dunno.)

Code: Select all

x = 13, y = 20, rule = LifeHistory
.2A$2.A$2.A.A$3.2A2.2A$7.2A6$6.2C$6.2C$2A$2A4$10.A$9.3A$9.A.2A!

This is mostly a thread for the R39 B reaction. I could start a new "Thread For your Partial Patterns" for your new reaction and more.

[fluffykitty]

Here's a fun pattern based on the snake-R39:

Code: Select all

x = 13, y = 14, rule = MultiLife
2$7.2A$3.A3.2A$2.3A$.2A.A4$8.2A$8.A$9.3A$11.A!

So, the eater is turned into a glider, the glider hits the back of the B and is shifted, and then is turned into a TL+block when it collides with the Herschel.

[fluffykitty]

I wonder if the output B block could be moved/rebuilt in the original position using the output herschel. (Coincidentally, it is a 59-gen B->H, same as conduit 1.)

[codeholic]

It certainly can, but the recovery time might be quite large.

[fluffykitty]

It certainly can, but the recovery time might be quite large.

Well recovery time doesn't really matter for me. I would like to have it anyway, even if there is a huge recovery time.

[dvgrn]

It certainly can, but the recovery time might be quite large.

Well recovery time doesn't really matter for me. I would like to have it anyway, even if there is a huge recovery time.

Unfortunately there's an even bigger problem than recovery time. It doesn't look to me as if there are any known conduits that can produce a B-heptomino in that location -- check out the H-to-B, R-to-B, B-to-B, etc. lists in Calcyman's collection.

It would be fairly easy to move the block if you really wanted to -- it could be done with a diagonal Herschel track connected to a series of edge-shooting reflectors, for example, producing a slow salvo. The right slow salvo could move the block to anywhere you wanted it.

But such a repair circuit would come out ridiculously large in bounding box as well as recovery time, and without a workable B-heptomino source the conduit could never be used anyway. Might as well start over with a different reaction.

Just for example, the base reaction for the 31c/240 caterpillar could be seen as an imperfect conduit quite similar to this one -- it burns up a block and drops a couple of other blocks in the wrong places. Here again a salvo could easily be designed that uses one of those wrong blocks as a target, and restores the correct block. Unfortunately the resulting "conduit" would be so large, slow, and unwieldy as to be unusable.

[fluffykitty]

Oh, right. I forgot to mention it doesn't have to be in the same location. Someone should search for a stable catalyst for it. (Of course, ideally the B would still be in the same location and not have an outrageous recovery time. Also it would lead to a lot of new conduits if it worked.)

[dvgrn]

Oh, right. I forgot to mention it doesn't have to be in the same location. Someone should search for a stable catalyst for it.

Have I given a clear enough description of the problem with putting an input B-heptomino in the right location? When hunting for new conduits, it's really important to start with an input location that can be reached by at least one known conduit. Otherwise you have to find not only a way to complete your conduit, but also some totally new unknown way to usefully feed a B-heptomino (or whatever) into it.

Unless this B+block reaction can be shown to be compatible with existing conduits, there doesn't seem to be much point in setting up a speculative search for a stable catalyst. (?) Bellman-type searches are not too likely to succeed for any particular problem of this type, so this kind of search is usually done only for fairly well-defined problems where a solution would be really useful.