1x1 24-12 Stacker No Collisions

[lofty crater]

This stacker has no collisions between launchers and 2 high buildings or belt elevators. Belt routing can probably be done cleaner, this is more of a proof of concept.

Current newest version: #1276111404989939823 message

[zenith geyser]

Blueprint Infos :

Version : 1095 / 0.0.8, Blueprint type : Platform, Blueprint cost : 2,838, Platform unit cost : 2
Building count : 454, Building size : 16x15x3, Building tiles : 538
Platform count : 1, Platform size : 1x1, Platform tiles : 1
Icons : Platforms, building.StackerDefaultVariant, <empty>, Number3

[lofty crater]

So if anyone manages to decrease the building count. I'd be happy to hear about it.

[dark crater]

this thing is seriously impressive

congratz

[lofty crater]

Thanks, it took me way too much time lol.

[dark crater]

i didnt believe it was actually homogenous inputs at first but it is

[lofty crater]

I fixed a small unnecessary asymmetry.

[zenith geyser]

Blueprint Infos :

Version : 1095 / 0.0.8, Blueprint type : Platform, Blueprint cost : 2,822, Platform unit cost : 2
Building count : 452, Building size : 16x15x3, Building tiles : 536
Platform count : 1, Platform size : 1x1, Platform tiles : 1
Icons : Platforms, building.StackerDefaultVariant, <empty>, Number3

[dark crater]

im trying to play spot the difference and i cant notice any

ah

found it

[copper mulch]

#1276021379694858362 message I have a version of this that's 14 buildings less

(without labels & trash for direction indication that is, which is what is represented by what I linked)

[dark crater]

this one has launchers through lifts

[copper mulch]

is that getting removed?

[dark crater]

which is probably okay, but an unknown

what is or is not considered collision has yet to be determined

i hope that launchers through lifts will remain

[copper mulch]

yeah

[dark crater]

but there is a world where it doesnt work

[lofty crater]

Avoiding lift collisions was the hardest part of this. I also hope launchers through lifts will remain but personally, it would make no sense to me if launchers collide with stackers but not lifts.

[copper mulch]

how long did you end up spending on it

it looks like you took a totally different approach than mine or Sikerow's one and did a 3 3 cluster layout on the bottom and then a 3 3 cluster rotated on the layer above it

[lofty crater]

Like 3-4 hours. I had most of it done in 1.5 hours but had an issue I couldnt solve. I went to bed, woke up, and solved it in like 15 minutes.

[dark crater]

the only "issue" that remains with these designs is that they thoroughly jumble which shapes go to which output

its not really a problem in a practical sense but many people like to keep belts segregated

[lofty crater]

Yes but I figured it doesnt matter.

[copper mulch]

the one I have matches notches for stacking at least

[lofty crater]

I have a 8-8 2 high design that keeps the input belts together. As in opposite belts go to the same 4 bent stacker array.

[copper mulch]

so the northmost left notch will act as the bottom for the northmost right notch

[dark crater]

this is much much more important for usability yeah

this one has homogenous notches aswell

the earliest 12 belt stackers had mixed notches

a notch is a full set of 12 belts

not the individual catchers

[copper mulch]

oh

[lofty crater]

with this you mean a single belt completely goes to a single 4 stacker array? In my design this is not always the case yeah.

[dark crater]

1 belt is processed by these 4 stackers

but each pair of stackers here outputs to a different belt

so the inputs from 1 belt are scattered amongst the outputs

[lofty crater]

I see. For me these kinds of solutions are most fun to find.

[dark crater]

yeah

[fossil crater]

I'd say in order of flexibility the most important is having the notches segregated, with having the floors of each notch being segregated as a nice to have, and each individual belt on each floor being segregated being even nicer

[copper mulch]

Yeah for this one the northmost right 3rd level is shared as the top for northmost left 3rd level and northmost left 2nd level

[dark crater]

floor by floor is nice because it guaranteed that 1 train-unloader of crates will perfectly load up another crate in the next step

without separating floors, this can turn out to not be the case

[lofty crater]

To me the most important thing is using as little buildings as possible. This design was not aiming to accomplish that. I just wanted something that works.

[copper mulch]

I wonder if it's possible to match them up like my design is doing while also avoiding launchers through elevators

[lofty crater]

It's probably possible but it requires quite a few extra buildings.

[dark crater]

honestly i think floor-by-floor is more important than belt-by-belt
because if you have a belt-by-belt design but the corrosponding belts are not on the same floor

then you lose the train crate guarantee even still

[copper mulch]

floor by floor?

[dark crater]

yeah so like, all floor 1 inputs are floor 1 outputs

[lofty crater]

I havent played with trains yet lol. So I have no idea what this problem entails.

[dark crater]

lemme see if i can find my pics

[copper mulch]

That's what the rewired version of mine is doing

I might take inspiration from the way this blueprint does stacker clusters

[lofty crater]

What is a stacker cluster?

[dark crater]

i loaded exactly 1 crate of input per floor into my 12 belt stacker, but because of simulation lag or whatever other inconsistencies, the floors did not perfectly distribute the outputs
ive tested in a sandbox/empty world and it does work, so this was due to some form of lag or something

but because my design isnt floor by floor

i lost 1:1 train crates

not a problem if you are making continuous throughput regardless

[copper mulch]

I was experimenting with a lot of designs earlier (I actually ended up making the final one pretty early in but I wasted a lot of time on designs that didn't work 😭)
I consider a cluster to basically just be the "cluster" of the 4 bent stackers

[lofty crater]

Do you think this will be fixed?

[dark crater]

no

i hope it will, but the fix is not so simple

[lofty crater]

Ok thanks.

[copper mulch]

The one I have has 4 clusters with 2 clusters on top of it in the same orientation and then that's mirrored on the back side

the oldest 12-12 has 3 clusters with 3 clusters on top

[fossil crater]

which version is that? I cant find a blueprint on the thread you linked that preserves the floors?

[copper mulch]

which seems to be very bad since it creates a lot of spaghetti

[copper mulch]

wdym by preserve floors? like for output?

I might have misunderstood

[fossil crater]

as in floor 1 on both inputs = floor 1 of output, etc

[copper mulch]

Ah

yeah mine doesn't do that

I was thinking of something else

[copper mulch]

This one is doing 3 clusters on the bottom and then an interesting approach of having 3 clusters on top but they are a different direction

[lofty crater]

I actually dont know if mine does that.

[fossil crater]

it doesnt. none of the 1x1 designs so far have

[copper mulch]

I wonder if it would be optimal to do 4 bottom 2 on top rotated

or if that is at least just another valid layout for the stackers

[fossil crater]

it'll certainly be an achievement for whoever can design the first 1x1 24-12 stacker with independent floors or even independent belts

[copper mulch]

I also have a list of axioms for 1x1 12-12 stackers

to help design them

[lofty crater]

Independent belts will probably not happen by modifying my design but I might be able to seperate floors, since I did not think about that at all while designing.

[dark crater]

honestly i'd be interested in a belt-by-belt 1x2

i have no faith in that being possible on a 1x1

[copper mulch]

• All of the 3rd floor bottom belts will have to be moved down
• All of the 1st level top belts will have to be moved up
• There will have to be at least 2 belt transfers for the higher and lower floors of the bottom and top belts respectively
• 4 clusters will have to be moved to the 3rd level
• At least 2 clusters will have to move down to the 1st level

usually the 4 clusters moved up to the 3rd level are in the back since they are launched across everything i find

[copper mulch]

what exactly is meant by independent floors?

[dark crater]

the inputs from floor X are all output on floor X

[copper mulch]

ah

[fossil crater]

it means that floor X of input a + floor X input b = floor X output

[dark crater]

i mean technically

you could say "the inputs of floor 1 are stacked with the inputs of floor 3 and all output on floor 2"

and that would still be train-loader safe if youre using homogenous notches

but usually what people mean is that all the floors inputs are not mixed with each other

[lofty crater]

nvm this wont happen lol. I have too many belts that are not segregated.

[fossil crater]

yeah it's not a simple task

[lofty crater]

I can feel it is possible though.

lol

[fossil crater]

ive been tempted to write a constraint programming automated blueprint finder for compact 1x1 designs lol

it's such a huge problem space that it's impossible to brute force solve it, so it'd need a fairly smart solver

[lofty crater]

That is an interesting idea.

[copper mulch]

It's pretty tricky since you have to deal with 3 floors and a lot of stuff

[lofty crater]

It is probably harder to make such a program than to find the design by accident.

[copper mulch]

I came up with a graph system for representing the structures but I think it would be really difficult to actually make a program out of it

[dark crater]

btw you can straighten these out a bit

it makes 0 difference but it looks nicer

[stuck terrace]

I can't believe it

[lofty crater]

To achieve floor segregation with my design I think I would have to drop the no launchers through lifts constraint.

[fossil crater]

probably yeah. there's around 600 tile locations in a 1x1, with around 50 different things you'd want to be able to place in each tile for a stacker blueprint (different rotations of each belt, lift, etc), so that's at least 50^600 (about 10^1000) possibilities to narrow down

[lofty crater]

But I will let someone else figure that out.

[lofty crater]

You can solve subproblems though, with a computer, and then put those solutions together as a human. For example you might need a stacker cluster to fit in a certain area with certain constraints, but it is impossible. Then you find a ton of different configurations with a computer, and see if there is one that can help you.

[fossil crater]

true

[lofty crater]

But then there is a chance you get stuck in a local optimum.

[copper mulch]

here's what 8 bottom 4 rotated top could look like

no idea if it has any more potential than other layouts

wait oops I forgot 2 lmao

[lofty crater]

I only see 2 at the top.

[copper mulch]

probably not very useful

idk I think there's a lot of different ways to arrange the clusters

I wonder if any of them are beneficial towards making something that doesn't jumble the inputs

I feel like Shapez 2 is too limited right now to really go wild with designs like piston doors in MC

[west barn]

Order of floors doesn't necessarily need to be preserved. When loading a train, all that matters is that each floor is homogeneous.

Stackers have a lot more criterias they can follow or break, so I believe the next challenge is to create a system that segregates the content of each floor first, without worrying about final order.

A more elegant design is having a clean floor order. A clean belt order will be incredibly tedious and not worth the little benefit is provides.

And then, all of this with adjacent inputs...

[dark crater]

[lofty crater]

Adjacent inputs 😟

[copper mulch]

Also something asymmetrical might be ideal

The issue is it’s a lot of effort to try and set up belts for a given cluster design and that design might not support avoiding launchers through stuff for example

I’m pretty sure on the 4-4-2-2 cluster design I did it’s impossible to route it in such a way that launchers though elevators are avoided

But on the other hand on that design it is possible to route it so belt stacking lines up (not super useful though) which idk is possible with this 3-3-3-3 cluster setup

[copper mulch]

It’d be nice to know what launcher collision is going to be to actually design with it in mind

[floral crater]

Who are these people with these “issues”? Are there people who send a different shape on each layer rather than belts full of a same shape in the middle/end game?

[dark crater]

its just a design restriction that they like to adhere to

making your own requirements is as much a part of playing the game as building is

it also does have positive consequences about throughput, like the train safety as mentioned later

[copper mulch]

It feels like every blueprint currently looms in the shadow of launcher collision update

since launchers are the only way at the moment to create more complicated flow on a single level\

ngl I wish they just didn't add collision at all

it just sounds tedious and is going to make really cool designs not possible anymore

[dark crater]

thats a silly one to make a setting

[copper mulch]

yeah

[dark crater]

why not make building processing ratios a setting at that point

let everything be 1/belt

:P

[lofty crater]

12 building stacker platform

[copper mulch]

I know there was collision in the demo and they removed it

[dark crater]

yeah it was too buggy as implemented there

so it was removed temporarily

while they work on adding other features and optimizing other things

[copper mulch]

it's just weird since they put the game out and if they re-add it it will break so many peoples factories

and designs, blueprints, etc.

it's really weird to not have an extremely integral mechanic like that not in the game right now

[west barn]

It's not silly if it implies breaking a backbone of your factory, especially with how restrictive they can be.

It's a weird risk they took to release the game, albeit in EA, with no collisions knowing the impact it might have

[dark crater]

honestly i think a sufficient warning via devlog and releasing it on a major patch number will ease all that out

[obsidian igloo]

to me that sounds like the person that was saying that since the game released in EA it should already have multithreading implemented

[dark crater]

players who are gone wont be bothered, players who are here have time to prepare

[copper mulch]

I mean I get making the game more challenging but I feel like if they add launcher collision they should make up for it by adding more ways of routing

It feels like they're just discouraging making cool designs

[lofty crater]

lift splitters would compact designs by an insane amount.

[copper mulch]

yeah seriously

there's a bunch of restrictions with conveyors right now that are also a bit annoying :p

[obsidian igloo]

I'm pretty sure the reason that was rejected is because it would make designs too easy

[dark crater]

the only belt mechanic i still really want is 2:2s

they come up just often enough with valid cases, and its not interesting when you have to avoid them

[copper mulch]

idk at the very least I think they should consider just not adding launcher collision since it makes designs way more flexible

while still being tricky

although part of that right now is also that launchers are extremely rigid they aren't variable length

[lofty crater]

I like that

Designs shouldnt become too easy to make. That takes away the fun.

[copper mulch]

I think if they add launcher collision though and don't add any more routing tools then this design will probably be the best stacker design possible since I don't know if it will be physically possible to make one that has good routing which is kind of sad

[copper mulch]

That's kind of true

[lofty crater]

Imo belt 2:2s and lift, building collisions add to the game.

[copper mulch]

what are belt 2:2s

[lofty crater]

2 in 2 out intersection

[copper mulch]

oh

[dark crater]

i have a list of message links of people wanting 2:2s

heres a random one

#1274307010170585169 message

here

[copper mulch]

yeah imo that's the problem

[copper mulch]

I think stacker setups are the ultimate challenge tbh everything else is a lot easier since you can just make it flat

I hope they add more two level buildings and routing stuff

[lofty crater]

I think the swapper is quite challenging to work with too

And the painter, although a nice layout has been found already.

[copper mulch]

The swapper is fun but stackers are a bit better since swappers layouts can still be mostly flat

Painter was already conquered 😦

[dark crater]

crystallizer

[copper mulch]

Haven’t got it yet

[lucid pagoda]

er

[serene trail]

a

[lofty crater]

Made some small tweaks that slightly reduce the building count. It also looks better now imo.

[zenith geyser]

Blueprint Infos :

Version : 1095 / 0.0.8, Blueprint type : Platform, Blueprint cost : 2,708, Platform unit cost : 2
Building count : 438, Building size : 16x15x3, Building tiles : 517
Platform count : 1, Platform size : 1x1, Platform tiles : 1
Icons : Platforms, <empty>, <empty>, {RuRuRuRu}

[lofty crater]

1x1 12,12-12 Stacker No Collisions

[dark crater]

usually we just call it 24-12

[lofty crater]

1x1 24-12 Stacker No Collisions

Got it thanks.

[vestal crow]

Anyone know how hard it would be to make this so the output is across fron the input, instead of between them?

[dark crater]

it hasnt been done to a degree that i would call satisfactory yet

[lofty crater]

You mean adjacent inputs?

[vestal crow]

Yeah! Sorry I couldn't think of the right word

[lofty crater]

Im quite sure not without collisions but I dont know if it has been done at all.

[lofty crater]

I did it for one of the variations https://discord.com/channels/1000343719314198548/1277273526537162783

[vestal crow]

Amazing!!