Spirit

[wet parcel]

So right now, I typically use one of two rail blueprint books I've made depending on what I'm doing.

[hasty sage]

Thanks very much for taking the time

Just one thing please, I don't want to be exposed to specific designs, I consider them spoilery

But I am interested in the top-level discussion and concepts

[wet parcel]

Yeah, I figured! I wasn't going to share any images or blueprints.

One is based on a ~||3||x||3|| chunks for ||1||x||2|| trains. There's a straight section, a diagonal section, T intersections, diagonal offshoots, etc. Largely for "Point and shoot" rails for early game stuff. "There's this resource, let's go as fast as we can from point a to point b and drop down a station (also in the blueprint book). It's not really expected to support super high throughput. Just for getting mid-game resources and production setup outside of the primary base to assist to late game stuff. So while it does support diagonals, there's no intention for it to be laid out in a grid structure. Just plop stuff down where it fits, it's all intended to be ripped up later.

[hasty sage]

Yea that's what I wanted to do first

But I aimed for 2x3 train size

and aimed to make it 2-way 1-rail

thinking that minimizes resources

since it is for early/mid-game

[wet parcel]

The other is based on a ~<stupid-big>x<stupid-big> chunk layout for ||4||x||8|| trains. Each blueprint is an entire square, grid aligned. Very strict in it's layout, there's really no flex ability there. No diagonals. There's also a loading bay blueprint that you add after the fact where you need it.

[hasty sage]

yea I wanted to add loading bay bps too

but I don't think I was gonna make this grid aligned

just in general would make them

and connect the 2-way rail to it manually

at the outpost or in base

[wet parcel]

Ohh, I forgot to mention. The first set I talked about is 2 one way rails, the second is 4 one way rails.

[hasty sage]

isn't 1-way 2-rail more resources to connect 2 points?

compared to 2-way 1-rail with a few 1-way 2-rail bypasses?

at some scale the bypasses might make the 2-way 1-rail system start approaching the rail count of the 1-way 2-rail system, but the 2-way 1-rail is not meant to scale anyway

so if I am not mistaken it should have lower rail count at small scales no?

[wet parcel]

So the reason most people chunk align their rail blueprints is because they're using 2 rails. It's really easy to move a single line of rail over a couple tiles. It can get really messy when you're running two parallel tracks.

[hasty sage]

I see, so it's not necessarily for the QoL of tiling

that's an added benefit

but not the main purpose

I was making this tile for the QoL aspect

[wet parcel]

Yeah, I think because you're using two way rails, you get that QoL for "free. "

[hasty sage]

mm I see that makes sense

so are these the concepts ppl had in mind when they kept criticizing my approach?

and advising me to do 1-way 2-rail instead

though ofc I think likely I also didn't properly communicate my reasoning that led me to arrive at my approach

I was even linked this https://xyproblem.info/

xD

though to be fair to myself, I also knew very little

I didn't even know about the distinctions between 2-way and 1-way and such

I just happened to make 2-way the first time I played the tutorial

[wet parcel]

Right. And since you're ultimately going to hit the throughput cap of a single two way rail failry fast unless you invest in bypasses everywhere, it's easier to just start with two one ways.

[hasty sage]

and decided it's good enough at low scale

[wet parcel]

Yes, a two way rail should have smaller rail counts at small scales.

[hasty sage]

and so then after finishing the game spaghetti-style, and starting the next step of my journey of wanting to make bps for my next playthrough, I thought about making a 2-way rail bp book

also when I finished the game, I only had three 2-way rails and 3 trains, two of the lines overlapping slightly with no bypass, so the trains waited at their outposts, but the outposts weren't far from my base or each other, so I thought it's fine

and then started hitting some snags in the design process and asked questions

and was met with criticism at it being 2-way

and started learning broader concepts from the things people were saying

was I supposed to somehow better communicate my intentions? xD was I committing https://xyproblem.info/ ?

[wet parcel]

I think critizing may be the wrong word. I think intent is very hard to convey in a forum. I think it's more of a "please don't shoot yourself in the foot, a lot of us have tried the same thing and it didn't work out, we don't want you to be disappointed."
Have you ever heard the story of Lego collectors? The new guy comes in "hey, look, I came up with a new sorting method! I'm starting out and sorting everything by color!" and everyone else that started there went, "Uhh, we've all been there, you don't want do to do that, you want to sort by shape. Learn from our mistakes"

[wet parcel]

Not knowing what you don't know is also a very difficult thing.

[hasty sage]

mm I see

though also different ppl had different types of reactions

[wet parcel]

And yeah, and those scales, I totally agree 2 way rails work absolutely fine. It's just how far do you plan on taking that 2 way rail?

[hasty sage]

and some felt to me like their reactions were kinda like "I hate that" etc.

After what I learnt I was gonna finish this 2-way design and then make a 1-way 2-rail design

and try to compare their throughputs at various scales

by generating a map at my settings

and duplicating the nearby ore patch locations

and see where the 2-way starts to struggle in throughput

cos I have no idea how the 2 systems compare in that regard as they try to scale

[hasty sage]

my goal atm is to accomplish the vanilla sub8 hour achievo

though I have been told I likely don't need much designing to do so

but I am kinda enjoying replaying the game now in a way where I design a little in advance

every time I am close to the next science milestone, I pause the run and go design the things I need for the next science milestone

in that run I have just finished building a car and scouting nearby and am ready to get oil and an iron outpost and go for blue science

so I paused the run and wrote down a list of things I need to design

one of them is a simple rail system and its stations

including loading/unloading

[wet parcel]

So let's take a simple example. You have infinite trains at your disposal and they load and unload instantly. You want to run a rail between your main base and an ore patch that's 2k tiles away. With two way rail, your fastest throughput is the total amount of time it takes a train to travel 4k tiles as you can only have one train moving at any given time.
With two way rails, your throughput is as fast as you can load and unload them as you can have trains going both ways.

[hasty sage]

a station design for just items, another for just fluids, and another for both in case I have both types of wagons on one train

and in each type, two separate designs for load and offload

[wet parcel]

Sub 8 can be done with zero rails without that much more effort. You really only need to hit one extra iron and copper patch, and those are typically close enough to belt in without any problems.

Actually, it's probably more effort to run rails.

[hasty sage]

yea I have been told so after I communicated my goals

I didn't realize this

but also I have been purposely avoiding learning the "correct" or "optimal" ways of doing things

because I value discovering and figuring things out on my own

I thought that even for the nearest mining outposts I should use trains

and from my experience in the first run I knew I will need at least one iron mine if not more

to be fair I didn't calculate the cost difference between running 2-3 yellow saturated belts from a nearby mine, relative to running a train from it

as a specific example, here is where I want to expand in my current run:

it felt to me I should use trains

oh and also, I thought I must use trains anyway for oil

my plan was and is to not look at existing designs or speedrun strats until I finish 8h and am satisfied with my bp collection

at which point I would look only at what the top speedruns do, but still not at any existing designs (beyond what I am exposed to when looking at the speedruns)

and then likely I would consider the "launch a rocket fast" portion of my vanilla journey complete

[wet parcel]

Is this a rail world?

[hasty sage]

and I haven't yet been interested in megabases so was planning to next look at one of the big content mods, I had my eyes on SE though I have been also recommended Krastorio, Pysomething and Seablock

[hasty sage]

no I haven't changed any settings

I just scouted by car and those are the patches I found first

I haven't scouted in the other 3 cardinal directions and wasn't planning to once I found these

[wet parcel]

You may want to look n/e/w from your base. That's an unusually long distance for the next-closest patch. You got extremely unlucky there 😂

[hasty sage]

are you saying I should have some meta knowledge and know there should be closer patches?

on the settings I chose (which is default)

[wet parcel]

Now that I think about it... Yes. It's just one of those pieces of knowledge you take for granted after playing for so long.

Did you look at the map preview? That typically will show you where your next patches are as well.

[hasty sage]

no I avoid spoilers of that kind

and I play anything it serves me

so far I died twice, both times on a desert start xD

and I won once on a forest start

and this is my 4th run

[wet parcel]

That's why you look at the map preview 😂

[hasty sage]

but I don't wanna

[winged thicket]

it's worth noting that speedrunners play the game very differently from everyone else

[hasty sage]

I wanna crash land wherever on the planet

and do my best

[hasty sage]

yea I understand, I would've looked at their runs mainly to just marvel at what you can accomplish when you optimze for going fast

not necessarily to adopt what they do in a general sense xD

but possibly learn some things too

but I don't wanna be "served" those potential lessons (if there are any in looking at the top speedruns) without first discovering on my own

[wet parcel]

You mean it's not normal to build 10M in 4 hours?
/s

[hasty sage]

and after that I still wouldn't want to look at bp designs because then I wanna play modded content and I don't want looking at existing designs to affect that journey of discovery either

though ofc vanilla bps likely don't spoiler much in modded content

but even then, I prolly won't look at designs anyway

cos SA comes eventually

which will likely force me to scale my designs bigger

so I want to be able to discover that without being affected by existing designs

is this too extreme of a mindset?

[wet parcel]

Nope! You do you. Play the game how you enjoy it.

I would recommend going into the new map menu, opening the preview, and hitting "new seed" a few times just to get a sense of what's "normal."

[hasty sage]

hmm

I have to think about how I feel about that

because beside all the other stuff, I think I do like a little bit of roleplay mixed in too

[wet parcel]

But I feel like we've gotten off of a tangent compared to where we started. I totally understand the "I need to run rail for my next patch" based upon that screenshot you shared.

[hasty sage]

immersing into the story of the game of crash landing onto an unknown planet

in which case I shouldn't know what the usual topography of the planet tends to be

ofc over time I will start learning it as I play more (or rather, learning the traits of the default generation settings)

but I can prolong not knowing as well by not looking for it

[hasty sage]

haha yea we have

but thanks for the discussions

[wet parcel]

To play devil's advocate, I think if you were traveling interstellerly, I think you'd have all of the planets failry well mapped out in your vacintiy. If you knew you were going down, you can make sure they'd be figuring out what the immediate area looks like.

[hasty sage]

true, but the story didn't tell me about that

and I didn't make headcanon

BUT I see your point

in that I can easily make that headcanon

and allow myself to look

and pretend that I looked at the planet's traits as I was going down

among all the other stuff like existing bp designs, I suppose this aspect is the smallest of possible spoiler types xD

[wet parcel]

Depending on how far you want to take it, you can even consider the "random seed" button to be your "Push the ship in a different direction" button.

[hasty sage]

especially considering the fact that the map generator has tons of settings that would change things

[hasty sage]

true, I like that haha

so far I have been kinda extreme/purist in that I play whatever it serves me

even if it's an arid desert

so maybe I should scout in a spiral in the future

for some reason I didn't think about that

even though I saw the radars scouting in a spiral

xD

though forests and water and nests make it a bit harder

an I just found myself where I am in that ss

mainly due to this pass

idk looking at the ss it feels that I didn't go THAT far

but I certainly wasn't thinking that I am looking for something withing belting distance, either

[wet parcel]

It's going to blow your mind when you find your first belt-only megabase.

[hasty sage]

I would prolly consider something like this as belting distance

for early-game

without having done any calculations

on how much belting costs relative to train lines

at the same throughput

(which should be fully utilizing the mine)

(at least ~60 eminers)

so 2 yellow belts is my benchmark

for what I would expect my mining outpost to output

idk if that's accurate

but it's what I went with for my benchmark after looking at my iron mine in my first playthrough

but in this run I found a 10M mine

I haven't looked at how much eminers that can house

oh shit

I didn't see that it's under a forest

lmao

[wet parcel]

I think trains are always cheaper (in materials) compared to belts as soon as you add a second belt. But the big things about belts is how convenient they are.

[hasty sage]

hmm

some of these eminers are hanging off the edge but, looks like this will house 100+ easily

I guess the benchmark is prolly around 60 eminers (2 yellow belts) per ~5M ore patch or so?

this one is 10M

[hasty sage]

gotcha

even within that approximate square I drew?

a train with 2 yellow belts of throughput is cheaper than 2 belts?

actually I haven't looked at train throughput in terms of belts

hmm

idk what the numbers are

per wagon

assuming you match it to the total eminers' output

like if you want the train to match the eminers' throughput

idk what it is per wagon

assuming stacksize 2 fast inserters

[wet parcel]

Theoretical train throughput can be as fast as the slowest of your loading or unloading time.

[hasty sage]

although maybe I need stackinserters?

I haven't looked at that either

so far I have never used stack inserters

[hasty sage]

right makes sense

I haven't looked at stack inserters at all

and haven't done the maths on fast inserters per wagon either

[wet parcel]

Let me do the math, one minute.

[hasty sage]

is it 5 inserters per wagon side?

I'll try to calc it too

now

sorry nvm you are calcing the belt vs train

I am calcing the load/unload throughput

the belt vs train calc is dependent on distance no?

well the belts are linear across distance

but the train isn't unless we ignore the stations?

I guess we ignore the stations

and add them as a static cost at the end

so it would have three components:

belt cost / distance

rail cost / distance

station cost

right?

[wet parcel]

So assuming you get your fuel for free (solar/nuclear) and don't care about stone or coal (what else are you going to use it for, landfill?) a single piece of rail takes 2.8iron ore (1.9 moduled) to craft. Two tiles of belts takes 3 iron (2.1 moduled) to craft.

[hasty sage]

you mean producing them per minute?

[wet parcel]

Ohh, I meant just for transporting materials.

[hasty sage]

so you aren't simply calcing their cost / distance and comparing that?

[wet parcel]

Correct, I thought you were looking for that here.

[hasty sage]

hmmm

not sure if I understand correctly

in my mind I am assuming already that a train can definitely support 2 yellow belts of throughput across any reasonable distance

possibly even at 0 distance

actually yea it has to even at 0 distance

because the load/unload MUST be able to support the throughput

though it will have a larger delay than the belts

by 0 distance I mean full load then followed by unload

so since I assume it can support the throughput

what I want to know is how much it costs to cover x distance using belts vs using a train

for 2 yellow belts of throughput

[hasty sage]

to figure that out I would need this calculation I mentioned right?

[wet parcel]

Yeah, that's what my numbers were for. Multiply the belt cost by 2 as one train tile spans two tiles and you want two belts.

[hasty sage]

what's the iron ore a minute thing?

if it's a cost to build the infrastructure why does it have time

[wet parcel]

Ohh, "a minute" was not supposed to he there.

[hasty sage]

so it's this yea?

2 rails cost 0.5+5 = 5.5

1 rail = 2.75

2 belts cost 3

1 rail is size 2, 2 belts are size 2

is that correct?

[wet parcel]

Yep.

[hasty sage]

OK so rails are 2.75 per a unit of distance (defining a unit of distance as 2 for simplicity of comparison)

and belts are 3

as mentioned in your message

gotcha

then I would need to have my loading station that supports 2 yellow belts of throughput

since I haven't done the inserter maths idk how many wagons or inserters that is

also wait

wagon count is dependent on delivery distance no?

since if I have too few wagons, the ore will accummulate in chests (or the eminers will stall production if no storage)

or is that directly taken into account once I do the inserter maths?

I guess it is

each wagon will have a maximum throughput

so once I calculate it for 1 wagon, assuming some form of inserter config, I will know its maximum throughput

and I need it to support 2 belts

or wait no

just doing that is not enough

if the train travels too long

that's part of the throughput too

no?

so if it travels too long it needs more wagons

so distance affects wagon count no?

[wet parcel]

I think the throughput of a single wagon is like ~4 blue belts each.

[hasty sage]

at loading/unloading yea?

[wet parcel]

Yeah.

[hasty sage]

but how about how long the train takes on the path

the belts traversing the path maintain their throughput throughout

[wet parcel]

If you have many trains, it doesn't matter how long it takes.

[hasty sage]

ohhh I was assuming 1 train

in your version what is the assumption?

that we have enough trains to never fill storage at loading?

or rather

that we have enough trains to match the throughput

of the mining outpost

which in this scenario is 2 yellow belts

but how do I calculate how many trains that is

given distance

and given wagon count

[wet parcel]

My assumption is that you're overthinking it. Less time theory crafting, more time building! 😝
But yes, I would assume you have enough trains that storage doesn't backup.

[hasty sage]

it has to be: wagon storage / (loading time + unloading time + travel time) = 30 items per second right?

but wait this assumes one wagon in the whole system no?

[hasty sage]

haha yea I understand

but I am now interested in learning how to calculate this

even if it is overkill for my scenario

which seems to be a common theme

of the rabbitholes I find myself in

[wet parcel]

If we assume one wagon, one engine, and eight fast inserters, that's 588 iron (and 51 copper) as a fixed cost for the train.
So y=2.75x+588 for trains, y=6x for belts, trains are cheaper in materials after just about 180 tiles.
And that's for two belts of yellow belts.

[hasty sage]

wagon_capacity: 2000 ore

fast inserter_stacksize: 2
stack inserter_stacksize: 4

num_inserters (per wagon): 6

fast/stack inserter rotation_speed: 864 (degrees)

------------------------------------------------------------------

## Equations

num_swings: wagon_capacity / (inserter_stacksize * num_inserters)

swings_per_sec: rotation_speed / 360

load_time: num_swings / swings_per_sec

------------------------------------------------------------------

## Fast inserter calculation:

num_swings = 2000 / (2 * 6) = 500/3

swings_per_sec = 864/360 = 2.4

load_time = (500/3) / 2.4 = 500/7.2 = ~69.44s

when measured in practice: ~72s

------------------------------------------------------------------

## Stack inserter calculation: 

num_swings = 2000 / (4 * 6) = 250/3

swings_per_sec = 864/360 = 2.4

load_time = (250/3) / 2.4 = 250/7.2 = ~34.72s

when measured in practice: ~36s

I used this setup to measure them in practice:

[hasty sage]

why 8 sorry?

is the calculation not target_throughput / inserter_throughput?

2 yellow belts = 30/s

fast inserter at stacksize 2 = ~ 4.62?

or hmm wait

the faster we offload the better

to minimize load_time

which should minimize wagon count needed across distance

so we should do the max number of inserters per wagon no?

which is 12?

and offload into a chest

and from the chests, offload at a rate of 30/s

or am I mistaken in something?

wagon_capacity: 2000 ore

fast inserter_stacksize: 2
stack inserter_stacksize: 4

num_inserters (per wagon): 12

fast/stack inserter rotation_speed: 864 (degrees)

------------------------------------------------------------------

## Equations

num_swings: wagon_capacity / (inserter_stacksize * num_inserters)

swings_per_sec: rotation_speed / 360

load_time: num_swings / swings_per_sec

------------------------------------------------------------------

## Fast inserter calculation: 

num_swings = 2000 / (2 * 12) = 250/3

swings_per_sec = 864/360 = 2.4

load_time = (250/3) / 2.4 = 250/7.2 = ~34.72s

when measured in practice: ~36s

------------------------------------------------------------------

## Stack inserter calculation: 

num_swings = 2000 / (4 * 12) = 125/3

swings_per_sec = 864/360 = 2.4

load_time = (125/3) / 2.4 = 125/7.2 = ~17.36s

when measured in practice: ~18s

here is the 12 inserter version

so now I minimized how much time it spends at a station given this inserter config (per wagon)

I also need to estimate travel time over distance

and then I will have all the components of load_time + unload_time + travel_time

hmm wait

now travel_time is the only variable

(and wagon count)

so I can just assume each wagon_count

and see what the max travel time is

to maintain 30/s throughput

[wet parcel]

For two yellow belts, 4 per wagon. 2 to load to a chest, 2 more to load to the train. I forgot to account for the chests.
To be honest, if you're doing the cost / benifit analysis down to the tile level, you're doing it wrong. Either you've got a rail network setup that you're utalizing, or you're belting stuff for the ease of use.
In the long term, the transport costs are negligible.

[hasty sage]

and I can use that to figure out how many trains I need for that wagon_count

right?

[hasty sage]

yea makes sense

I am interested in knowing how to calculate this though

because it is relevant in general

like to calculate how many trains with a given wagon_count and loading/unloading inserter_config do I need to maintain x throughput

in a typical scenario I assume what would arise is I have a mine, the output rate of which determines the throughput

I have a given inserter config which deteremines the load/unload time

and I have a given path which determines travel time (assuming no disruptions)

so now the only variable is train and wagon count

I can assume a certain wagon count

and calculate trains needed from this

[wet parcel]

Don't calculate it. Throw some trains out there. Not enough, add more trains.

[hasty sage]

and then maybe add +1 or +2 to account for disruptions

[hasty sage]

hmm OK xD

but there is one further step too, do you think it is irrelevant/unnecessary too?

which is that a particular railway setup has max train count

which limits max throughput

should I not calculate what that is either?

I think I should no?

so I know how much throughput I can support approx for that setup

to know whether I need a better setup or not

if I am planning for higher throughput

[wet parcel]

Your railway setup is pretty much always going to be bottlenecked by your intersection design, but even then, on a grid layout, you've got a ton of parallel intersections. Unless you're doing something very silly, it'll be hard to hit that limit.

[hasty sage]

hmm

so this means that I can more or less assume that the bottleneck is unload speed?

and therefore wagon count at unload station?

I mean as a shorthand for quick calculation

I just need to ensure my offload speed supports the throughput coming in

and that's all?

beyond that I just throw trains into the system

until all outposts are operating at max rate

without their storage getting full

(assuming I only offload at 1 central location)

btw sorry we went off on tangents and it dragged on 😛

but ty for everything, I learnt some things

[wet parcel]

No. Bottleneck is ultimately intersections. You can load/unload at the same rate. It's getting trains around other trains that's the largest concern.

[hasty sage]

Hmm

idk if there is something I am missing

why are there intersections in a 2-rail system?

isn't it one big loop?

oh I guess it can have branches

and should have branches

[wet parcel]

Yeah.

I think someone did the math. Even a basic roundabout with 1-1 trains can support 1.1k spm with all trains going through the same intersection, so even then it's not typically a concern.

So tl;dr just start building.

[hasty sage]

oh

and 1.1k spm is tons of ore throughput

xD

but wait

in that scenario how do I know how fast I must offload?

so the trains don't wait in queue

in other words, how many wagons and/or offload stations I require

like how do I know in advance "OK now I will need a 2nd iron offload station"

or "OK now I should add 1 extra wagon to each train (to increase offload speed by 1 wagon's worth of inserters)"

though I guess in a scaled system I will immediately put the trains at the max target wagon count

so the second part is not relevant

but the number of offload stations is

at some point I may need to increase number of offload stations no?

if I have too much throughput coming in

or wait

OK wait I think the shorthand I mentioned earlier handles this no?

because if trains are waiting in queue for offload

then it's kinda as if the whole train system is one big chest buffer

so logically it's almost like:

mine output --> mine storage buffer --> railway storage buffer (one giant chest split across the wagons and distance) --> offload

so if there is large queueing happening at offload it likely means that my mine output (the throughput coming in) is larger than my offload throughput (because the railway buffer storage is filling up, so I am adding to it faster than I am removing from it)

so I can simply calculate my offload throughput per offload station and compare it to mining output

and that tells me when I need more offload stations

right?

in that scenario adding more trains to the system won't fix the problem

it will just slightly increase the railway storage buffer

and after that fills, the outposts' storage buffer will fill back up again

the only way to fix it is to offload faster (have enough offload throughput to support the mining output)

so that shorthand works right?

comparing mining output to offload speed

to know how many offload stations I need

I think the key lesson is I should try to moderate my rabbithole delving more

and not bite off more than necessary to chew the given bit

or wait that comparison doesn't work xD

but anyway xD

[wet parcel]

Keep your number of loading and offload belts the same. Then it's just making sure you have enough trains to support that transport delay.

[hasty sage]

yea makes sense

but loading expands over time

right so I just expand offload accordingly

I mean sure that works as an even simpler offhand

but that is wasteful too

because offload rate per station is likely higher

[wet parcel]

And depending on throughput, you can always have one producer feed multiple consumers, or vise versa.

[hasty sage]

oh you said belts sorry

not stations

nvm my bad

OK so that's essentially the same as my shorthand right?

because belts define number of stations needed

[wet parcel]

Indirectly, yes. You probably don't want to try planning for more than ||four|| ||blue|| belts at wagon.

That's X color if you don't want the spoiler.

[hasty sage]

yea thanks I opened it

and now I am going down that rabbithole xD

and I can't help myself

do I assume max capacity bonus stack inserters?

if I am doing blue belt offload

hmm half a blue belt is 22.5

and one inserter can only offload onto half a belt

ofc I can quickly merge belts

but yea max cap stack inserter is faster than half a blue belt

so technically I can max out 3 blue belts on each side of the wagon with them

which is 6 blue belts

but if I do that, I am offloading at a rate of 6*45 = 270/s

[wet parcel]

That's chest to chest numbers, not chest to belt.

[hasty sage]

which empties the wagon in 7.4s

[wet parcel]

But yes, max bonus.

[hasty sage]

oh right

OK so technically I could have 2 inserters for the outer chests

going onto belts

but only one for the inner chests

ah and with multiple wagons, all chests are inner chests except the very first and very last one

WAIT

you can offload diagonally

does that increase inserter count?

it should I think?

OK I am now leaving the rabbithole

this rabbithole in particular is extremely bad