math-and-meta

[frosty owl]

In an ideal world...

[oblique hollow]

we've been into it before: satis is in fact not deterministic

[wicked tinsel]

so what likely happen, they are written in gaming way, to advance simulation by fraction of second, and this breaks when that fraction of second is too long

and probably for 780 belts, they are so fast that even properly timed simulation is too slow for them

at least that what i got it from empiric experiments

[oblique hollow]

its the transition segments

[frosty owl]

Which, considering how there's never no drop of FPS or whatever at all, makes it so that there is never no throughput loss, even with 60+ "stable" FPS

[oblique hollow]

the belt to belt connections

pipes dont have pipe-to-pipe throughput issues funny enough, but they have other issues as we know

or maybe they do, but because they are in 1000x units, its less noticable

[wind spade]

it's all belt to belt connections, not just mk5

[wicked tinsel]

yes, but the issue is more prominent the faster you get

[frosty owl]

I got a few posts on it on Reddit, if you'd like more context
Or you can just check the related thread in this channel

[oblique hollow]

then again factorio has the curve issue, so its not like stuff is perfect
i think you posted that, didnt you Ven?

[frosty owl]

Just a link to someone mentioning it

[wind spade]

yeah, but that means it isn't related to mk5 belts being to fast, but b2b connections being weird

[oblique hollow]

is that recent even?

[frosty owl]

The bug?
I think they fixed it, by recent convos on Discord relating to that

You should ask @barren elm for more

[oblique hollow]

you cant really find anything on goggle regarding that so cant verify xd

[wind spade]

it's possibly just the bug they had a few years back (and fixed)

[wintry aurora]

Which bug are we talking about here?

[oblique hollow]

Factorio belt and Satis belt issues

[barren elm]

Factorio curve compression bug I think? Pretty sure that's fixed

[vapid gorge]

wait what? fixed b2b?

[wind spade]

no, we were talking about factorio

[vapid gorge]

Right, got my hopes up there

[median heath]

If FICSIT wanted you to have hope, they would have issued it to you.

[vapid gorge]

it comes in patch form

[median heath]

FICSIT giveth (Sprint Lining)
and FICSIT taketh (Blueprints..)

[uncut sigil]

Is Hope as bad as Bliss or can we keep civilisation this time?

[median heath]

Bliss is allowed.

[fierce ruin]

I am curious, FPS or compute aside what is the fundamentally limiting factor in terms of an upper bound in Satisfactory?

[barren elm]

I mean the answer is fps so I don't know why you'd exclude that

[fierce ruin]

I am excluding it because I want to know what factor otherwise would define the upper bound, since that gives a better understanding of the game on a mechanical level.

[barren elm]

Having a hard time translating the question into something answerable

Are you asking what in-game resource you're likely to run out of first?

[versed violet]

TIME

[fierce ruin]

That would be one potential upper bound yes. It equally could be throughput for example.

[barren elm]

Most end-game products have enough alt recipes to allow you to use a wide variety of resources, so as you run out of one resource, you can just use another

Contrary to what you often see around here, there's rarely a "best" alt recipe

Steel is a prime example, if you were to use the whole planet's worth of resource, you'd end up using every single steel recipe at some point (excluding the default)

[oblique hollow]

Most limits are due to engine and the fact that you can't multithread stuff that timely and orderly depends on stuff before it, which is most of the factory logic

The engine limit is the amount of "things" the dev can put in the game before the Engine's Garbage Collector freaks out and crashes the game. Can't increase that limit without unforseen instability issues

The unreal objects make up all of the stuff in game, to the most degree

Building more stuff increases the number of objects that the GC tracks

[fierce ruin]

I am aware of Uobject limits to be clear.

[oblique hollow]

right, then the other limit is simply CPU, due to the Factory Logic

[fierce ruin]

Again my interest lies not in technical limitations. My interest lies in where the mathematical boundaries lie mechanically

[oblique hollow]

what would that be? got an example question?

[fierce ruin]

Basically assuming infinite compute, what would be the ultimate bottleneck

What would stop the factory from growing first.

[oblique hollow]

assuming you never crash and the game always performs well: resource extraction limits

[fierce ruin]

Its almost safe to assume its not space, so it is either throughput at some point, or its a given resource

[oblique hollow]

for the higher tier parts, its usually bauxite / quartz / nitrogen

so that limits your high-tech parts

on the lower tiers, its oil and copper i think

and even further down, coal

[arctic willow]

theoretically it could be space; the power can be maximised by underclocking everything

run everything at 1% to minimise power requirement and you might be able to run out of room before resource

[oblique hollow]

that would also cause you to die of old age

before you get even one heavy modular frame produced

so if you maximize power efficiency, the mechanical limit is your life expectancy

[fierce ruin]

That is an interesting point. Basically what to optimize or maximize for

[oblique hollow]

definitely not power, you got nuclear

that is finite, but enough to yield a few TW i think

if you use plutonium that is

with that however, you have a finite space and time for your save

[fierce ruin]

I think that is almost a given in any endgame base no?

[oblique hollow]

before everything is just containers of waste

[fierce ruin]

To use plutonium I mean

[oblique hollow]

eh, not needed

most cant even use all of uranium power fully

only if you REALLY build big

which isnt the norm actually

with U7 it MIGHT due to blueprints, buuut i doubt that

[fierce ruin]

I mean yeah only fraction of the playerbase actively pushes a given game to its limits.

My best example in that regard is that the civ community was adamant that the vast majority plays and wins on diety regularly

until Firaxis gave them telemetry that showed that over 80 percent never won above settler (the easiest difficulty)

Meaning to say those involved in communities always tend to be more competent in a given game simply due to their longer and or more intense engagement with a given community

[oblique hollow]

in that case, most of the large save owners / large builders here run into the limits of bauxite i think

but alts complicate all this

you can shift the balance around heavily with those

[fierce ruin]

I love the whole alternative recipe idea

it gave the game a lot more depth

[oblique hollow]

definitely

a simple example i can give: solid steel

it uses iron ingots and coal to make steel
not that interesting at first

UNTIL you look into recipes that have great iron ingot efficiency

namely iron alloy and pure iron

with those, coal is your limit on how much steel you can make

theres only enough coal to cover like 30% of ALL iron ingots made

hence the other steel recipes are needed to maximize steel

[fierce ruin]

To be clear I am still at best middling in Satisfactory. Just recently came back to it because I needed a pallete cleanser after doing a Krastorio 2 and Space exploration mod run in Factorio which nearly broke me.

[oblique hollow]

theres not enough coal or oil or sulfur to use all the iron on the map to make steel

[fierce ruin]

But I love to push games to their limit thus my curiosity

[oblique hollow]

im currently dabbling in modding, having taken a slight inspiration from factorio regarding oil processing, else ive tried to give everything my own twist

in vanilla, im at tier 8

but i have yet to automate most elevator parts

but yeah, due tu us NOT being in an ideal world, your PC resources are a major limiting factor

with enough size, your save might not even boot

else, if you somehow get it to load every time, the limit really are the resources

theres enough space for factories to use up all the high tier resources

[fierce ruin]

I have a middling machine. Ryzen 5 3600, 16 gigs of ddr4 ram and a RTX 3060

[versed violet]

are flower petals still a thing? I noticed the flowers in grasslands are not harvestable anymore

[oblique hollow]

yes they are

still used to make color cartridges (used to apply decorative patterns)

[versed violet]

so they only drop from trees now?

and there is 4 types of alien body poarts, not just organs and carapaces?

is there any reason to keep the alien remains around after unlocking mam research?

[sonic wasp]

Yes

They are going add ability to sink them for points separate from the other sink points

[median heath]

the ones in the GF are bugged iirc?
If you want to farm them go to the NF.

[median heath]

The individual remains? No.
Everything can just be processed into Protein. Which you can either keep around for Inhalers or as Boaty said process into Alien Data Capsules and sink for Coupons.

[versed violet]

good reminder. guess I'll keep a box of each.

SCIM does not have the tape locations yet??

[sonic wasp]

kibitz has a vid

[smoky quiver]

Has anyone figured out the optimal Uranium plant to Plutonium plant to maximize the power output?

Currently, my math says just under 115 uranium plants and 305 plutonium plants.

Uses all the uranium, and outputs a ridiculous amount of power.

[upbeat tide]

I built mine w 50.4 uranium rods and 22.4 plut rods

Plut being sunk atm with 252 active reactors

But if/when I burn the plut can get a max of 476 reactors

[wind spade]

what Verios said is the highest possible 🙂

[arctic willow]

can you not sacrifice some uranium for more plutonium?

(i don't know, i'm genuinely curious)

[upbeat tide]

You dont need to as the base non fissle recipe doesnt need uranium ore

Only side effect is slightly less non-fissle and you lock out making nuke nobelisks unless you pilfer from the belt/factories

[smoky quiver]

Fertile uranium is more efficient with Uwaste though, at the cost of using some raw Uranium. my math create 30.5 Prods out of 1146 Uwaste

[near zenith]

49.8 urods, 12.45 plu will get you one nuke nobelisk per minute fyi

[arctic willow]

according to wiki you lose more power in uranium than you gain in plutonium by using the alt

[upbeat tide]

Yea fertile really isnt that good

[smoky quiver]

I disagree, with 115 and 305, you end up with 543,760k MJ and with 252 and 224 you end up with only 525,000k MJ

[upbeat tide]

Uum each reactor makes 2500 MJ

252 reactors = 630,000MJ
224 reactors = 560,000 MJ
Totaling 1,190,000 MJ

Your setup is close but 140,000 MJ lower in max

[smoky quiver]

I don't know where my math got all screwy, your math checks out.

[glacial hemlock]

1 reactor makes 2.5 GW, or 2500MW, or 2500MJ every second

[fierce ruin]

2500MW != 2500MJ

1 MWh = 3600MJ

[median heath]

What if Sprint-Lining cost you like 5-10MW to use? 🙂

[glacial hemlock]

Could be, that's negligible cost

[median heath]

People would still manage to crash their grids with it.

[glacial hemlock]

Ouch...

[next bramble]

is there a calculator that would more help me when it comes to min/maxing my production from ores?
i have 4 pure iron nodes really close to one another but currently am using 2
so I want a more automatic form of calculating how to use those 240 iron ores by choosing my end recipes

[sand epoch]

Calculator.. or production line planning?

[next bramble]

uh production line planning, but without the planning, I just put in my max ores i can provide and it calculates and plans everything out for me

that is also nicely rounded

so all machines end up at 100% efficiency

and i wouldn't have for example a machine working like 15% of the time while the other one working 100% of the time

daniel's planner on github works close enough, but i still gotta adjust my end products so it works as close as possible to be rounded up

[fierce cypress]

have you tried

https://satisfactorytools.com/production

[eager igloo]

ok so i need some math help to double-check some numbers

i have 5 blenders in a line using 104 water each per minute and they produce 86.667 per minute. so i have a total of 433.335 run-off per minute. if I feed that back in to the loop after a valve choked down to 166.7 that should give me 600 correct? im not using the full 600 in the pipe though. will that cause a back up?

[fierce cypress]

which part of the system has the valve? the fresh water or the byproduct water

[eager igloo]

fresh water

i have the run-off going through a pump to keep it one-way

[fierce cypress]

it could work but i'd probably rather use a VIP just in case, or even better segregate the fresh and byproduct water systems entirely

[eager igloo]

I've never had success with vip

I've tried several times and they never worked for me

I've been forced to use wet concrete to get rid of it in the past and I want to do it right this time

[fierce cypress]

the most stable solution and the one i use is to segregate the systems

[eager igloo]

I'll do that if I have no choice but are my numbers right?

[fierce cypress]

your numbers are right yes

what clockspeed are the blenders at?

[eager igloo]

173.333 I'm processing 260 bauxite per minute so I can feed 3 blenders on 1 mk.5 belt

And I have 36 of them total

[fierce cypress]

instant scrap right?

[eager igloo]

Yeah

[next bramble]

does not give me rounded max production

or at least i couldn't find such option

[fierce cypress]

wdym by rounded max prod?

[vapid gorge]

I'm guessing not repeating decimals?

that link does do 1900 flat though

[next bramble]

sry not that, but rounded production cycle

[fierce cypress]

i see

[vapid gorge]

whats wrong with 1.5 copper sheet per min?

[fierce cypress]

but thats just on you to find numbers to work really i suppose

[next bramble]

i don't wanna do 1 and a half constructors, i wanna aim either for one, or for two and if i use one, then take into consideration a power slug

[fierce cypress]

i dont think theres a calc that does that for you

[fierce cypress]

you can just build 2 and underclock

[vapid gorge]

do two and clock one to 50%?
Or do two and clock both at 75%?

no slugs needed

[next bramble]

i'd still have to unlock clocking

[fierce cypress]

and thats a problem why?

[next bramble]

or is it now auto unlocked?

[fierce cypress]

nope, still in the mam

[vapid gorge]

ok just have two and have the production cycle stall sometimes

or do double of both and have 12 for the wire and 3 for the sheets xD there are many options

[next bramble]

yeah i'd rather have my brain turned off and just follow a already made plan

thinking makes my head hurt

[vapid gorge]

so build 2 constructors from that 'ready made plan' and they'll only each be on 75% of the time and you won't need to under clock shrug

[fierce cypress]

@eager igloo for a segregated system do something like this:
all you need is 86.666 recurring fresh water in to feed that rightmost blender, and the other 5 are fed off of the waste water, then for the alu output you get your 5 orange lines with 520/min each - you just combine the two weird clockspeed refs together

or alternatively the other way that people do with instant is they feed the 50 waste water/min back into the sulfuric acid one which also needs 50 water/min so its a 1:1 ratio there

[median heath]

Cries in imprecision.

[fierce cypress]

their numbers not mine

[glacial hemlock]

For me, 0.0001 doesn't matter because it can take thousands of minutes before taking effect. And even if it does, the machine maybe only interrupted by few seconds

[fierce cypress]

imprecision is imprecision and sev no likey

[eager igloo]

Will this still let me use full mk.5 belts for the bauxite? Or am I gonna have to redo my belt work for these to work

Also how is it imprecise if it uses the full amount and runs at 100%?

[fluid forge]

Can you try to building it twice?

So then you have 3x

And otherwise just build 2, one will just work 50% or both 75% of the time

[oblique hollow]

just take that sheet and double it

6 becomes 12, 1.5 becomes 3

[median heath]

Random thought about the blueprinter:

Everything he showed involved the Hoverpack, so will be interesting to see how it operates for people like me who build from the ground level.

[barren elm]

Presumably the same but more annoying

Kinda funny, my first thought on that vid was that I'm surprised he wasn't using some kind of fly command instead

[median heath]

If it can snap it would be the same.
If it can't my current building method is faster 🤷‍♂️

[barren elm]

It can snap

It showed foundation to foundation snapping, alignment snapping and "it's on a foundation snap to the nearest 10cm" snapping

[uncut sigil]

A lookout tower is almost as good as a hoverpack if you're building on a small area. I think it's going to be a little earlier. I'd say phase three would fit in terms of scale, but I can see it being with steel somewhere if they want it to be a major convenience for pulling people in.
Especially considering they mentioned wanting to fit coal generators into the box. Why bother if it's going to be way past fuel until you get to use blueprints?

[barren elm]

You get to use blueprints from minute 0

Blueprints are independent of your save file so you can just import whatever you want

It's only the "blueprint maker" that needs building and unlocking

[median heath]

VIP junction.

[uncut sigil]

Ok, chicken and egg thing aside, the game needs to introduce blueprints at some point, and it's not going to do that by saying: btw, here's a thing you can download off random people on the internet 😛

[upbeat tide]

I kinda wish the bp maker was 40x40x40. Would make fitting in refineries and the inputs a bit easier

[median heath]

Plz... no.

[barren elm]

Yeah Snutt could barely fit the manifold in

Like I can see their goal of not wanting to create something that lets you just vomit out factories

But it's so small that even individual factory chunks look cramped

[median heath]

Good.

[barren elm]

Let's hope the dimension checking is sloppy enough to allow larger prints

[median heath]

I am against that hope 😁

[barren elm]

I'm just happy it'll be able to fit my "sushi but better" layout in

[ruby field]

What happens of you push too much plastic in there?

[median heath]

I don't see how that is "better", but it's a clean layout.

[median heath]

Standard rule of sushi manifolds is overflow gets pushed elsewhere.
Something to be added after the blueprint.

Like if I end up using the BP machine at all I won't put belts in it. Because I sometimes change my mind about direction and I'd rather not delete all of them.

[barren elm]

Nothing, it's just a manifold with all the belts horrifically clipped into one superbelt

[sand epoch]

Wonder if bp will let you preset ss and ps...

[median heath]

sp?

[magic island]

I think everything inside the blueprint will preserve settings. machines will definitely preserve recipe, clockspeed, and shards

that's why snutt was able to make distinct rod/plate manifolds, not just a generic 8-constructor manifold. logically, smart splitter settings should carry over too

[sand epoch]

Yea, but they already support copy/paste.. splitters don't..

[median heath]

Splitters do.

They added that in U6.

[sand epoch]

Must have missed that memo

[barren elm]

It'd be kinda funny if you could make a bp with an alt recipe that you don't have

[median heath]

The only thing I want to try is using beams to make a lot of random wire crossings, then delete the beams and save it so you can copy/paste a floating ball of yarn.

[median heath]

It was packed in with the UI update to them.

[oblique hollow]

great news: you can fit lookout towers into the designer in case you ever need 4 x 4 lookout towers xd

[winter grotto]

thats rad

[median heath]

https://tenor.com/view/totally-rad-gif-3527228

[barren elm]

The more I build and test this rail network

The more I realize path signals have no reason to exist

Even in junctions with crossed rails they're more harm than good

[median heath]

They work decently when chained. But you have to build with that in mind. Otherwise yes, using just blocks works for the vast majority of cases.

[fierce cypress]

the beltwork should stay the same yea, that setup will use 1300 baux/min which should be the same

[fierce cypress]

epic reply fail

[eager igloo]

?

oh

[fierce cypress]

this message should have replied but it didnt

[eager igloo]

no worries

[fierce cypress]

all ive done is split the 5th refinery into two, the last two refineries on the right hand side will consume the same and produce the same amounts as one of the other ones clocked at 173.3333%

[eager igloo]

ok. also i posted a pic of my pipework in the screenshots if you wanna give it a lookover.

i can easily adapt it over to yours though too

the white pipes are run-off water

[fierce cypress]

the system looks fine, but i find valves to be somewhat unreliable, a VIP would probably be better suited but it doesn't look like you have much space down there, i'd say if you're confident just turn it on and in a worst case scenario where it breaks then you have other alternatives

[eager igloo]

I've never had success with VIP. but yeah i designed it to easily send the run-off to a concrete sink if all else fails

if this doesn't work im gonna change to give your design a try first though

[median heath]

@deft lichen has anyone done testing on the new nobelisks?
AoE radius, effect duration, etc.

[deft lichen]

nothing at all, we don't even have damage values

[median heath]

Base Nobelisk: 8m Radius
Pulse Nobelisk: 10m Radius
Cluster Nobelisk: 8m Radius, shoots 5 additional nobs up to 8m from original blast, each with their own 8m radius.
Gas Nobelisk: 15m Radius, 30s duration.

Testing NukeNobs atm.

NukeNob: 25m Blast Radius, 30s total Radiation duration, 50m Radiation Radius that rapidly decays towards origin.

As for damage numbers, 🤷‍♂️

Hope that helps @deft lichen

[median heath]

Clarification: Wiki says 7.5 for nobelisk radius, which could very well be it. Trying to distinguish between 7.5 and 8 is not an exact science with in-game testing.

7.5 would also fit the 7.5/10/15 radius pattern a bit better.
But until one of the devs says "this is the number we coded it to" idk 😬

[upbeat tide]

@oblique hollow why VIP distruster?

something change with the method?

[median heath]

They still work, he just distrusts them because they shouldn't work if I recall his reasoning.

[upbeat tide]

aah ok

[supple belfry]

I've been wondering about this myself. Regretably built out a bit of an elaborate series of junctions early in my rail network, and I'm too lazy to fix it. Using block signals, locomotives seem to slow down at every intersection. Will path signals fix much of this? It's not terribly high-traffic, but there is the possibility of collision there....

[fierce cypress]

blocks can be used in place of paths but not visa versa, path signals can increase throughput by allowing multiple trains through an intersection at the same time but trains will still slow down before a path signal

[fading coyote]

Wow I wonder if im doing way too much math for being so early in the game

gonna have to start writing programs by the time i unlocked the next tier

[fierce cypress]

https://www.satisfactorytools.com/production

[fading coyote]

hmm

[fierce cypress]

or do it your own way, both are valid

[fading coyote]

oh wowers

Im trying to calculate ratios to split things into

At least it looks like my math is right so far

rod needs to be split 5:9 for rotors, etc

[vapid gorge]

Have you heard about our lord and savior ‘overflow manifold’?

[fading coyote]

Nope

Just barely put down the space elevator guy

[vapid gorge]

https://m.youtube.com/watch?v=HwguwhizdlE

[vapid gorge]

Manifolds are the simplest and most basic way to feed multiple machines in a line

[fading coyote]

hmm

I mean hey conveyor splitter/merger can get you any fractional split

[vapid gorge]

You can ‘load balance’ and split exact number sure but manifolds are just as efficient as load balancing and much easier to set up

[fading coyote]

What are the definition of those two phrases

I'll show what im imaging rn if ya dont mind

[vapid gorge]

Load balancing is when you split belts in such a way that exact numbers get fed into a machine. What you were talking about

[fading coyote]

Oh yea

[vapid gorge]

An overflow manifold is just making sure the belt has the same number of parts that the machines connected to it have. The first machine will fill up first but eventually it’ll self balance

They are generally just called manifolds to make it simple

Watch the vid I linked. Much less math

[fading coyote]

So given this, if you want to make exclusively one part, you can configure A/B/C 'load balancers' in some fashion

Now im not sure if every split is a fraction that can perfectly divide

Or if its even worth it to design it in that way

[vapid gorge]

You can manifold all that and it’ll still work

[fading coyote]

hm

[vapid gorge]

Ok so for example how many ingots do your rods and plates need?

[fading coyote]

i guess its more of a factory design decision if you want to just pre-decide how much of everything you want

Because that way you wouldnt have any overbuilt components

[vapid gorge]

You can do that with manifolds too 😉

I have extremely complex factories with precise numbers all manifolded

[fading coyote]

I was thinking, you could design it which could be reconfigured in any way to fully utilize the input to make any one component. But that would mean, say, completely disabling the rod-> factory segment

whether or not thats wasteful i guess depends on the person

[vapid gorge]

As in you could hit a switch and do that?

[fading coyote]

I have no idea if the game has switches that could enable/disable that sort of stuff

But at the very least you could destroy/lay down conveyors as your configuration

[vapid gorge]

I mean you could turn off a section of a factory with a power switch?

[fading coyote]

Like if sometimes you need 50/50, and sometimes 100/0. You can just have a splitter, and delete/place the conveyor

Is there a circuitry system you can manage that kind of thing?

[vapid gorge]

No. You could just turn off a factory and items will back up there and flow elsewhere

[fading coyote]

Not so much disable a section but rather change distribution of a sections output

it could potentially end up fully disabling a section

[vapid gorge]

It wouldn’t really be manageable for more than one tier of production because the NEXT tiers will need those parts still

If you turn off one section you essentially turn off every section that relies on that part

[fading coyote]

Yeah so theres a few problems i forsee, one is that the number of configurations will continue to increase as new end-products come

I dont want to turn off a section completely tho

[vapid gorge]

There’s no way around that really. It’s automation and if one step stops working everything down the line stops

[fading coyote]

Like in this diagram, I want <A> to be a 'balancer' which can do 1:0, 0:1, and 3:4

so whenever theres 1:0 yeah it disables an entire section

But if you needed something like disconnected, 1:0, 0:1, 5:9, and 20:57

two of those configs wouldnt shut down a section (althought i dont know how the hell you would split 20:57)

[vapid gorge]

Ok, the easier method is putting the machines you want ‘off’ on a separate power grid and turn it off so they don’t use parts

[fading coyote]

i see

Does that mean each section needs its own power supplies?

and not a centralized grid

[vapid gorge]

It would have the same effect at the push of a button.

I’m not saying you CANT do what you want. Just not seeing the point

[vapid gorge]

Later you can break off grids with switches

[fading coyote]

ah ok

[vapid gorge]

Do you even have coal power yet?

[fading coyote]

nope lol

[vapid gorge]

Ok honestly just slap things together until you set up a coal power station

Until you have coal power the game is just a long tutorial

[fading coyote]

How would disabling portions end up creating efficient splits?

[vapid gorge]

I just commented because you said ‘should I be doing so much math this early??’

And the answer is no, you’re just doing it to yourself

[fading coyote]

Unless you just dont need that component at all

I mean yeah I know i dont NEED to do the math

I was more asking 1. Is the math required going to change later (probably yes, since switches will exist) and 2. are there already tools that solve exactly what im trying to

[vapid gorge]

It wouldn’t be efficient.

If you set up miner to feed A and B production and turn off B your iron will back up

[fading coyote]

Right

[vapid gorge]

Math doesn’t change unless you want it to

You would have to create your own tools if it is even possible. I’m guessing no but I might be misunderstanding what you want

Like I said: all these parts depend on the parts before them. Turning off step 4a will just turn off every factory that needs whatever 4a is all the way down the line.

Can you do it? Sure. Just don’t see why

[fading coyote]

well like you said, you gain new abilities with coal power like electrical switches. So in some sense I might be wasting my time if this work will need to be re-done with more efficient tools

My idea is basically to build the factory in such a way that it can generate any single item at maximum capacity.

And you need to re-configure how items are split to do that

[vapid gorge]

Because the way the game is you’ll probably be completely rebuilding factories as you learn

[fading coyote]

oh yeah im not against rebuilding factories thats not a problem

[vapid gorge]

I mean you do you. I just find a new node to build things from

[fading coyote]

but if im gonna do 2 hours of math and thinking. But in 30 minutes I'll unlock a tech that makes that math not useful. Then there isnt much of a point in building advanced early-game structures

[vapid gorge]

And building ‘any item at maximum capacity’ will be VERY hard at more than 1 step of production.

I imagine you’d actually need calculus

[vapid gorge]

And no- all my stuff is slap dash until I unlock everything

[fading coyote]

Well, the image above (admittedly only 3 splits) doesnt seem impossible

the more splits you have the harder it is tho

[vapid gorge]

It’ll get worse the more parts you need.

My plan for heavy modular frames starts out with 75,000 parts per minute getting turned into 300 parts per minute

[fading coyote]

Yeah

[vapid gorge]

You need A LOT of basic parts. And yeah the tech will constantly change what you need and you’ll need more of everything.

Personally I find 2 normal nodes dedicated to each basic part can make good progression if you keep upgrading miners and belts

That’s a throwing things together to grind out tech levels

[fading coyote]

I think its just more of some graph analysis to work out what splitters you need for each component

dont think (?) youd need any calculus or anything

[vapid gorge]

The other thing about manifolds is they are very easy to upgrade as you get better tech if you leave yourself space

You can’t ever upgrade load balancers

[fading coyote]

im gonna finish watching this vid about manifold to understand what exactly it is

[vapid gorge]

Some parts need like 20 steps. If you want to mix and match that for max output of anything? Calculus

[fading coyote]

I dont think so

I worked it out by hands for the basic components. For rotors. 100% of the iron needs to go to rods. 5/9ths of the rods go into screws, 100% of that goes into rotors, and the other 4/9ths of rods go into the rotor as well

and for reinforced plates, it's 3/4ths of the iron into plates, 1/4th into rods. Then 100% of the rods into screws, and then all that into the reinforced plates

Just continue the path, calculate every different split % you need, and figure out a way to implement that in a non-stupid way

[vapid gorge]

Yeah but what if to make max computers you need to turn off 1/34th of your screws so more wire is made?

[fading coyote]

You dont need those extra screws tho, thats the logic of this design at least. In reality, I imagine you eventually could make use of every component you manufacture

Now. Because you need to account for the cases where 100% of the iron goes into rods. Sometimes you don't use 100% of the machinery
that's a downside.

[vapid gorge]

Most people I think do make use of every component

Design top down is often easier

[fading coyote]

Top-down meaning what exactly

Like " i need a lot of iron ", so you make machinery to smelt 100% of your iron mining
then "i need rods" so you do that, "plates" etc, and just continue to stack everything up?

So actually another question about the game overall, not math related at all. Is the modularity. 1. Is there efficient long distance item travel you could make use of to simplify and segregate component factories? and 2. Is there any sort of automatic-construction stuff where you could make designs and copy/paste them effectively?

[fading coyote]

This seems more bottom-up when I think

[vapid gorge]

They are introducing new blueprints but for very small sections.

Load balancers tend to need to be individually calculated and designed though

[fading coyote]

It would be cool to be able to just design a like "50x50x50 500 rotors/minute module" that you can just lay out

[vapid gorge]

Also when I say production gets complicated #math-and-meta message

[fading coyote]

Oh boy

So like that stuff makes me think you NEED modularity of some sort

[vapid gorge]

Ehhhh I in fact do the opposite of modularity

All designed top down and separated

[fading coyote]

I play openTTD and it would be very cool to have something that can efficiently supply multiple factories with, say, "electrode- aluminum scrap" (that diagram above has 3 outgoing destinations of that component)

Instead of making the alu scrap factory. And bolting on the 3 outgoing factories onto the main one into some gigantic cube

[vapid gorge]

You often don’t want to use the same recipe all the time though depending on where you’re building

[upbeat tide]

I will say im building to a similar goals, and already use about 40% of all my aluminum ingot supply 😭

[vapid gorge]

Like honestly. Unlock coal, work through all the tiers once, then figure out what you want to do. You’re coming at this without actually knowing what tools and materials you have to work with

[fading coyote]

Like lets say i make a factory that does 500 screws/minute, and two factories which consume 250screws/min each. A train just looping around in a timely way keeping input buffers full would be very cool

[long crow]

What do I need to OC fuel generators at to burn 100 turbo fuel/min across 11 fuel generators? I'm not sure I understand how the oc scales exponentially to get it so each fuel generator burns exactly 9.0909... turbo fuel/min

[upbeat tide]

Its best to not really train low end goods,

Well thats the opinion of the anti-train guy...me

Tldr long time ago they pissed me off and went down a belt only road

[fading coyote]

I have only heard that trains exist, I dont know how they function at all

i play a train transportation game sometimes and im just thinking if it works like that it'd be very nice

[upbeat tide]

They function well now but im on a 3k+ hour save. Beyond the point of easily adding them in 🙂

[fading coyote]

but yeah, low end components dont really need modularity

[upbeat tide]

Yea the addition of train signaling fixed alot of the early issues

[fading coyote]

Of course you would integrate every pure iron-only components into one factory, or at least directly adjacent boxes

Modularity awsome in theory. I dont have enough experience to comment on doing it in-game

[vapid gorge]

Honestly don’t. It’s not clean maths and they are changing how oc and generators work in the update

[vapid gorge]

If you’re doing a big world project in one spot your computer will die

[long crow]

I thought in the most recent update they said they were pushing back oc changes to the next update after this one

[fading coyote]

yea idk anything about how well this game runs at scale

[vapid gorge]

U7 which is soon. Last I saw anyway

[vapid gorge]

Massive world projects like what I linked will chug your computer even with multiple large hubs spread out.

Won’t handle it all in a spot

[fading coyote]

Does the game only simulate things in a certain radius of ur guy?

[vapid gorge]

Ehhhh to a certain degree. It doesn’t keep all the objects loaded past a certain distance but it still has to do a ton of maths

[fading coyote]

Yeah that was my next question, if things still generate items

[vapid gorge]

But every object in the game total puts load on it. The more obj directly near you the worse

[vapid gorge]

As in machines making things?

[fading coyote]

yeah

If you have a miner attached to a storage box. does it stop producing item at some distance or anything?

[vapid gorge]

Yes ofc, otherwise the whole automation collapses

[fading coyote]

Well the whole thing could just freeze

like yknow minecraft just doesnt simulate chunks after a distance, maybe this game similar

I feel like there would be nice tricks to closely approximate the net effect of a factory, never tried it myself, that sounds like annoying math

[vapid gorge]

Well if you need the computers being flown in at the other side of the map for supercomputers it’d never work if the distant factory stopped working if you moved away from it

[fading coyote]

Yea thats what i wanted to know

[vapid gorge]

Items and factories just exist as math if they aren’t rendered

[fading coyote]

gotcha

[vapid gorge]

Though again- play through the tiers. Then see what you want to do. I changed my mind at least twice

[fading coyote]

calculating an efficient balancer %'s is definitely just some graph traversal the more i think about it

Yeah im just gonna upgrade my factory and keep playing lmao

[vapid gorge]

I’m on my 5th map

[frosty owl]

Yo, asking since there's activity: did you notice the custom-colour bug getting fixed?

[vapid gorge]

That part yes- designing for multiple stage flexibility in output for any given part that changes based on other variables? Calculus

[vapid gorge]

I didn’t know it existed

[fading coyote]

Well

any factory that can handle, say, 500 input/minute. Could also efficiently handle 250 input/minute, no?

So you just over-build each sub-factory to handle the maximum throughput

[vapid gorge]

Except everything behind it will back up

[frosty owl]

The one that makes it so that all my lifts (custom- coloured) load as black, remember

[fading coyote]

No

If a sub-factory is backed up, then the balancer is splitting incorrectly

[vapid gorge]

Sorry, misunderstood, yes

[fading coyote]

And other factories in the chain should be picking up the slack

[frosty owl]

I'd say it could process 250 at 50% efficiency. Obviously, either half the machines aren't working, or they're all starving half the time

[vapid gorge]

Nope! I had that happen to all my color swatches XD

[fading coyote]

Yeah, it has to be over-built, and there will be idling machinery. However, you sacrifice that idle machinery for the sake of optimizing the end product

[frosty owl]

Elaborate, please

[vapid gorge]

Actually @frosty owl is someone you should talk to about this XD

[fading coyote]

hai

[frosty owl]

Sure.
I was just answering the point about efficency, btw ^^

[fading coyote]

Yeah, i think its just a decision

[frosty owl]

... What?
I've got very little time atm, if my assistance is needed, please be quick about it or wait for later 😅

[fading coyote]

no no we're just discussing

[vapid gorge]

Rugg wants to set up a factory that is load balanced in a way that could have parts turned off to nearly maximize any item

Ven does sushi load balancing.

[vapid gorge]

Ah fair XD

[fading coyote]

Yeah pretty much right on, so some kind of load balancer that can adjust it's distribution. So that you can have any component maxed out for its original inputs

[fierce cypress]

absolute mad man

[fading coyote]

gotta do some thinkin

[vapid gorge]

I’m pretty sure you could do that mostly with manifolds tbh

[frosty owl]

I did make an "easily configurable load balancer" recently...

[fading coyote]

im not even close to experienced enough with the game to know effective designs

But I know for a fact, obviously, you can brute force any fractional balance

[frosty owl]

Might not be what you're looking for but (apologies) I have no time to discuss it atm.
See my recently posted links to find it if interested

[fading coyote]

yea no problem

[vapid gorge]

Ven also does needless weird things that are cool XD

Try to find his post with 1 belt feeding 4 parts into a manufacturer

[fading coyote]

Huh

not sure how that would work unless you can ensure the order of the belt's input

(also if splitters split in a deterministic way)

Ahhh, ok. I'm learning

[vapid gorge]

As long as the ratio stays stable. But it’s a lot of work

[fading coyote]

Manifold distributes each 'child' 1/2, 1/4, 1/8th. Until the first one gets backed up from it being overflowed, and that continues for a while. Seems like that would cause a pain point on an extremely long first child, it would take forever for it to begin backing up

[vapid gorge]

And it’s one belt AND one input

You can do 1 belt 4 inputs pretty easily . 1 and 1 is hard

[fading coyote]

Well hmm, nevermind I think

This idea seems easily solvable with how manifolds work, yeah.

[vapid gorge]

The big hurdle will be ‘maximum efficiency’ though

[fading coyote]

If you want rotors, you just kill these edges. And once any buffers in the blocked edge fill up, everything would start flooding thru the available pathways

So like you said

Just have a switch to disable certain sub-factories. I get it now

[vapid gorge]

Oh and the other thing- you’ll have to build a full factory for every node because you’ll be limited by belt speeds.

So every iron node will need a screw factory and plate factory and a rod factory

[fading coyote]

mmm would u?

each thing just needs enough bandwidth for worst case, so just extra belt tracks

[vapid gorge]

If you do that I’m not sure you can turn things off with a switch since you can’t connect EVERY belt

[fading coyote]

You could just turn off the first machine in the sub-section though right?

[vapid gorge]

Which? What if every node needs to be making iron plate?

[fading coyote]

Ummmm

[vapid gorge]

How maximised do you want to go?

[fading coyote]

Wouldn't this self-optimize to make rotors?

[vapid gorge]

Like, you can also make wire from iron.

Do you want to have an optional iron wire factory?

[fading coyote]

(with the crossed out edges being turned off)

[vapid gorge]

Maybe in one step processes

[fading coyote]

Im just thinking with these components because it's what im actually working with

Im really early in the game we know

let me do a little thinking ill show u smth

[vapid gorge]

I'm sure you could do it with the set up you showed before.

I'm talking about the big factory idea yo uwere sharing

But yeah go do things and set up a big coal power plant

continuous power changes the whole game

[fading coyote]

So yeah this is not 100% optimized because each path could have more specific bandwidths. But you could just take the maximum production of each component, and that is the absolute worst-case bandwidth of anything that could possibly utilize the component

So, using mk1 belts for the sake of simplicity
obviously Rotors do not need a 720/60 = 12 belt wide screw input. But it is the worst-possible-case

You could work backwards given an ore rate and figure out how wide rotor needs to be of course. but this seems quite simple and sort of an obvious way to build a factory

[vapid gorge]

Work through the tiers and keep me updated 🙂

[fading coyote]

Will do lol

[vapid gorge]

Working through the tiers what I did:

Made 1 pure node of each basic part - connected it to multiple 2nd tier part reguardless of how many items needed/used there were, parts of hte factory shut itself down when containers filled uip dirrecting other parts to sections

[fading coyote]

Mind you, if there become better transportation methods. I will always prioritize modularization to squeezing out efficiecny

[vapid gorge]

Well you can't get more than 100% efficiency. And honestly it's not hard to do.

[fading coyote]

If scheduled trains exist, i dont even know. I will be switching to small factories with large input buffers that get resupplied regularly

[vapid gorge]

you don't get to 'schedule' them, they just do their routes

[fading coyote]

What do you get to tell them?

At least which stations they go to?

[vapid gorge]

yup point a to b to c if you want

[fading coyote]

So what would be sufficient is something like
"Go to the screw factory, and load up 500 items worth"
"Go to the rotor factory, and unload 250"
"Go to the reinforced plate factory, and unload 250"
"Loop this path"

[vapid gorge]

nope

[fading coyote]

sadge

How does it decide?

[vapid gorge]

just go to screw factory
go to rotor
go to plate

it doesn't decide. You set an auto pilot and it does it's thing

[fading coyote]

it transports items tho right?

[vapid gorge]

it'll just follow the pattern you tell it to and grab whatever it can take

[fading coyote]

ah.

Could still be doable, but gets more scuffed as the limitations build

[vapid gorge]

Come back to me in a month as we can talk about it XD

[fading coyote]

will do lmao i keep talking too much

[vapid gorge]

shoo 😛

[fierce ruin]

im horrible at math can someone tell me or help me if i got 2 miners mk 1 they spit out 1 copper ore every 30 per minute how many smelters should i then have total to evenly have it

[fierce cypress]

well 2x mk1 miner on an impure node = 60 ore/min
and a smelter can process 30 ore/min
so we divide the 60 ore produced by the 30 each smelter can handle to work out that we need 2 smelters to process 60 ore/min

[fierce ruin]

ok i get that but why would i have extra accumulating in background like i have enough that the machine is so stocked if i did just 2 smelters

[fierce cypress]

is it backing up?

[fierce ruin]

sometime like in my mind i think if it backs up it needs more smelters but math is like law right ?

[fierce cypress]

yes, math is like a law, but only if your numbers are correct

[fierce ruin]

i just read what the machines said and you was right about 60 ore produced by the 30 each smelter can handle work out that we need 2 smelters thing

like i tried turning it off then on again make sure numbers were correct and they are but still smelters have alot ore extra i feel i should add extra smelters but then when i do then they insufficient and run out its just weird

[fierce cypress]

since supply = demand, stuff that was already backed up in it won't get consumed but it won't back up further either

you can test it but removing all the ore from inside of the smelter

and watch, it shouldn't back up

[fierce ruin]

ok will do thanks 🙂

so if we got a constructor next if its does 30 per minute then we only need 1 constructor correct?

[fierce cypress]

if it consumes 30 and you give it 30 then it will work fine, yes

[fierce ruin]

thanks again

[fierce ruin]

so im on concrete now if i got a miner mk1 does 60 per minute and the constructors can only do 15 per minute i would need 4 constructors correct ?

[fierce cypress]

yup 👍

but im pretty sure you're misreading here - they do 45/min and output 15/min

assuming you're using the default recipe

[fierce cypress]

the constructor i mean

[fierce ruin]

oh yes constructor is 15 per minute

basically dont you divide the 60 divided by 15 and you get 4 ? maybe my math is wrong one second ill use caculator

[fierce cypress]

60/15=4, your math is right yes

but i think you're reading the machine wrong

its 45 limestone > 15 concrete

[fierce ruin]

still not following ya :/

[fierce cypress]

can you send a pic of the constructor ui?

[fierce cypress]

the red is what it needs, the blue is what it makes

it needs 45 limestone/min - not 15

it produces 15 concrete/min

[fierce ruin]

sorry dude ill ask my dad explain it later. im just not following you learning issues. i havent been in highschool since 2012 so plus got bad memory to math learning

thanks for trying tho 🙂

[fierce ruin]

backed up

[fierce cypress]

because the output is blocked

the ingots have nowhere to go

[fierce ruin]

no i got constructor that take them

[fierce cypress]

well then its either blocked as well or its going too slow or not working for some other reason

[fierce ruin]

ok thanks

[oblique hollow]

the easiest thing to do is compare how many items per minute the next machine actually consumes

does it actually use up 30 ingots/minute?

you can use notes or, if feasible, short term memory

[loud trellis]

What else should go in here? Right now it's just doing 225/min Concrete.

[fierce cypress]

whatever other production you need, or leave the space for future production

[wind spade]

Or don't build limited spaces before you build a factory

[loud trellis]

I was thinking about putting Encased Industrial Beams in here since it's mainly concrete stuff, the current setup uses up all limestone I have around the area.

But now that I think about it the limestone will get increased, so, future production it is.

[upbeat tide]

👆
This is why I dont pre build factory structures. Many times have found a build to need more space, less space, or something in between. Having a pre self limitation on floor space makes life much more annoying than it needs to be

[wind spade]

why the middle finger tho xD

[barren elm]

I think it's meant as a pointing up/"this"

[upbeat tide]

Yea oops wrong one its early

[barren elm]

Either that or they feel very strongly about preplanning

[upbeat tide]

Fixed

[wind spade]

yeah got confusing messages from your post xD

[upbeat tide]

Yea just saw a up pointy emoji went with it 😄

[mystic tusk]

The wiki mentioned two very efficient production lines.
One for HMF based on the Heavy Encased Frame alternate recipe with a bunch of other alt recipes and one which was labeled as "diluted fuel cycle" with recycled plastic/rubber, diluted fuel, diluted packaged fuel and heavy oil residue.
Can someone please show me a layout of these cycles? the web calculator isn't really helpful because I haven't understood how these cycles work.

[wind spade]

all lines are 100% efficient

we can't really show you any layouts if we don't know exactly which recipes are used 🤔

[mystic tusk]

For HMF:
Bolted Frame
Cast Screw
Encased Industrial Pipes
Heavy Encased Frame
Iron Wire
Pure Iron Ingots
Solid Steel Ingot
Stitched Iron Plates
Wet Concrete
mentioned here: https://satisfactory.fandom.com/wiki/Heavy_Modular_Frame#Tips

The second one I have found a schematic for already

This one

@wind spade does this help?

[wind spade]

the screenshot you shared is already your setup

[mystic tusk]

No, my current setup is the "default" one:
oil into fuel plus resin
resin and water into plastic/rubber

The screenshot above is from the wiki and shows a more efficient setup

[wind spade]

no, what I mean by "your setup" is "the layout you wanted to see"

you asked for a layout, the website provides it 🙂

[mystic tusk]

oh, yeah, it's exactly that

sadly the wiki doesn't provide a layout for the HMF setup

[wind spade]

you can make your own in the calculator

[mystic tusk]

I'm trying rn

will definitely be my most complex production line

[mystic tusk]

Done

[wind spade]

I thought you wanted to use the same calculator as wiki did?

[mystic tusk]

I don't know where to find that one, so I used the one I knew

[wind spade]

https://u6.satisfactorytools.com/

[mystic tusk]

nice, thanks

[glacial hemlock]

You have got a very powerful tools. from now on you probably can't design a complex production line without it

[fierce ruin]

It just takes the step out to make a calculator for it in an excel sheet

[frosty owl]

Scoffs in brain power

The only time I took factory notes was for a maximized Turbomotors megafactory

[wind spade]

you want to make... calculator for calculator?

or do I not understand what you mean? 😄

[fierce ruin]

My point was more that while its a neat tool what makes it a good tool is the frontend it provides. The math itself is not that complicated.

In regards to the comment above that once you use it you can not do without it

[frosty owl]

I see both points. As a counter to your point, I've seen some (which is already too many imo) become almost dependant of having everything nicely spread out and shown as in SFTools in order to imagine things. Like, they stop being able to just visualize production layouts

[fierce ruin]

I somewhat agree with you on that one. It largely depends how you learn really

I am not a visual learner myself so I do better with rows and tables of data

[wind spade]

tell that to the 150-200 rows matrix that is needed to calculate the results 🙂

linear optimisation is hardly easy math

[frosty owl]

Gets easier if you know how to ignore the bad paths from the get-go like my incredibly advanced quantum computer brain does :P
Jokes aside, I think their point was that it's not hard to calculate the production needed once one is familiar with the recipes (the codex improvement helps tons with planning too, now!)

[fierce ruin]

Yeah, if you are used to modded factorio which has stuff like recursive recipes the ratios in Satisfactory are rather clean in comparision.

[median heath]

Quantum Sushi 👀

[frosty owl]

Will have to wait for 1.0 for that, I think

Well, sort of "quantum" sushi... It will have quantum parts on it

[fierce ruin]

Wait will 1.0 have one more tier in it?

[frosty owl]

Recycled plubber joins the chat

[fierce ruin]

huh?

[oblique hollow]

recycled plastic + rubber

its a dual recursive loop

[frosty owl]

Dunno, I'm just assuming the quantum-items (anticipated via holiday events) will appear in 1.0 and not sooner. Just my assumption

[fierce ruin]

I just recently returned from a longer hiatus so I might be out of the loop

[frosty owl]

Appropriate wording

[median heath]

Most likely not.

[oblique hollow]

see the loop?

[frosty owl]

(Recycled plastic-rubber makes for loops)

[median heath]

➿

[fierce ruin]

Now take that and imagine that it only works n percent of the time, and if it does not work it produces some sort of waste that you have to deal with, and you got modded Factorio 🙂

[oblique hollow]

yeah im aware of the probability stuff

this and packaging loops are the only recursions we got

[fierce ruin]

I quite like the cleaner math in Satisfactory though

[oblique hollow]

hmmm... technically water recycling also counts as loops

but doesnt have to be a cycle, so theres that

afaik the only vanilla loop in factorio is the kovarex process

[fierce ruin]

Yup

[barren elm]

Almost certainly, there's a few quantum items that as of yet can't be made

[fierce ruin]

I would just love one more tier to cap it off that is brutally complex and requires a lot of material as a long term goal.

[median heath]

Most likely not going to happen.

[fierce ruin]

I know, sadly.

[barren elm]

You could argue that the current tier 7 stuff is "brutally complex" tbh

[median heath]

I'm more excited to see how they expand current tiers.

And how SAM Ore is going to "make existing production easier"

[barren elm]

Like throw a plutonium fuel cell into the calculator and revel for a second at the absolute spiderweb it vomits at you

[fierce ruin]

I so far have not dabbled in mods but probably will once 1.0 is out and stable

[median heath]

Bleh

[barren elm]

I can't play without Just Fly anymore tbh

The hoverpack gave me a taste of flight, and then it was just a case of "why am I even dealing with this weird power distance mechanic"

[frosty owl]

Meh. Can fit it all in like... A flat 30x30 (kinda guessing, but not a massive are imo)

[median heath]

Jetpack > Hoverpack

[barren elm]

30x30 is bigger than the size you need to finish the whole game in

(assuming you just make a factory capable of finishing the final milestone)

[frosty owl]

Is it?
I'm just spiteyeballing based on a portion of the size my nuclear factory is, but I can't quite recall it's size exactly

[barren elm]

But yeah, 10 is a nice number, there's weird quantum stuff they keep givng us for christmas, and teleportation feels like an easy carrot on a stick

There's also as of yet nothing like factorio's personal logistics system which is another easy carrot on a stick reward

Feels like it'd fit nicely into a teleportation themed tier, you know, teleporting items into your pocket dimension

[frosty owl]

You're right, that's quite huge (I took a look at a screenshot)
I can make a better guess: the entirety of nuclear processing (excluding the power generator) could probably fit in a range between 10x10 and 15x15
Big uncertainty because I'm unsure how the particle accelerator would fit in

[fierce ruin]

The 5x5 challenge exists

[oblique hollow]

the last 2 x 5000 parts + 2 x 1000 are already, relatively, quite a lot.
maybe the devs increase that number, but that will likely be the final delivery

[fierce ruin]

I heard the opposite really that some numbers will be reduced since they were only set as high due to it being temporary as a sort of endgame.

[oblique hollow]

we dunno honestly what they will do with it

[frosty owl]

7x7 is a thing too!
#screenshots message

[fierce ruin]

According to what I read that was an official statement from CSS on the topic but I do not know how valid that claim was/is

I like my factories pretty though. Monuments of brutalist architecture. If they are not evoking bladerunner vibes I am doing it wrong 😉

[frosty owl]

Maybe as a sub-goal...?
If they add more costly items to the Awesome Shop, getting them all could require sinking those "brutally complex and a lot of materials" (or waiting forever)

[barren elm]

From a game design point of view, satisfactory is already kinda long, I can't see them stretching it too much

[frosty owl]

Rewording my point: I partly agree with McGalleon and think that the current end-game is already the max a general user can take without having too much people feel like it's too much work.
Thus my thought that having a higher ceiling might be left for "sub-goals" or achievemt-like things such as the AWESOME SHOP

[fierce ruin]

Might be the old man in me talking but I liked games way back when that a sizeable chunk of people playing it could not beat.

To give a more recent years example of that something like Spaechem

which was brutally hard, but so gosh darn rewarding to beat just a single level.

Generally all Zachatronics games really.

To be clear I get why devs do not do that anymore, and agree with the sentiment of it.

[frosty owl]

I disagree only on the basis of the vibe of this game. (Imo ofc) it's supposed to be chill and relaxing. Can get challenging, sure, but shouldn't become a chore. Having many not finish the game because "too hard/too long" isn't in line with that

[barren elm]

Give space exploration mod in factorio a spin if you liked spacechem and enjoy pointlessly long and complicated unlocks

[fierce ruin]

I did beat Space exploration and Krastrorio 2

[barren elm]

Well fair enough lol

[fierce ruin]

Only thing left is Pyanodon which even I dread

Because that is in a league of its own

[barren elm]

I dream of a space exploration style mod for satisfactory, tho the added complexity of a 3d world probably makes it unrealistic

[fierce ruin]

I started a Satisfactory run as a pallet cleanser basically 🙂

[oblique hollow]

a bit too much right now for any currently sized small mod team

in factorio, the 2D aspect makes things easier because you just need sprites for everything

here its literally a whole other dimension on top of that

[barren elm]

Yeah I don't even know what the state of satisfactory's mod tools are

I just presume it's the usual affair of "none"

[oblique hollow]

what tools

[frosty owl]

Some nerd somewhere: "Hold my beer"

[upbeat tide]

is there a formula for drone battery usage? Tried searching the wiki but may hazve missed it

[median heath]

4 for takeoff and 1 per km iirc.

[upbeat tide]

cool, was just trying to put a ballpark estimate on max support for a 400 a min battery supply

[still brook]

I'm looking at setting up a bunch of trains to slurp up ores
In order to ensure that I get a 780 belt on the outputs side is it as simple as having a ISC buffer on either side and make sure the train loop time is low enough?

[median heath]

Why are you limiting your throughput per car to 780?

[frosty owl]

As long as the source ISC can completely empty out (after a couple trips), the system is good

[still brook]

To make it consistent throughout the system, I don't want to have to deal with odd numbers like 1553 ore/min
How else might I tackle it?

[median heath]

Just seems like you're using more cars than you need to imo.
Manifolds are consistent as long as you supply them correctly.

If the way you've chosen to build requires limiting cars and using more of them, then stick with that.

It's moreso something to consider during the planning phase.

[still brook]

Yeah that's fair, right now im just doing the math waiting for blueprints
It'll be annoying to have 1 car per miner, it means there's going to be a lot of trains on the network, but even with manifolds I cant have the miners stall mining while the trains load

[glacial hemlock]

it can be treated as 2 full belts

[still brook]

How would you achieve a consistent load of 2 full belts with the train load/unload stalling mechanic?

[frosty owl]

With unfulfilled wishes

[glacial hemlock]

by limiting a train car to only 1 full belt. Then 2 cars = 2 full belts

you can also do it with 2 car: 3 belts, or 3 cars:5 belts, or whatever ratio you can manage

[still brook]

I feel like it'll be easier to just make each car 780 rather than needing to balance them out later on

[upbeat tide]

As a hard rule for my game I just dont max mk5 belts. Have had pretty bad issues in the past with 780 throughput going weird

Like my max for any mk5 belt is 750/min.

[wind spade]

one segment is fine, then you split 🤷‍♂️

[frosty owl]

Or place a merger/buffer

[upbeat tide]

I mean for production and yea do split

[frosty owl]

The throughput issues are not specific to production-related or storage-related beltwork 🤔
It's the same for all belts

[upbeat tide]

Yea ik, sorry explaining bad

[frosty owl]

Eh, whatever 😆
As long as no wrong info is conveyed~

[still brook]

Yeah i've just been reading that, sounds more problematic in larger factories
maybe the go would be to limit it to 600 then it's even easier to do the math against normal and impure nodes

That should work fine for iron, but for copper and caterium im a lot closer to the limit

[upbeat tide]

Yea 600 based production lines is my primary goal. Some stuff just doesnt fit nicley into 600.

Silica for example. The cheap silica alt can take in 600 quartz and limestone but makes 750 silica.

[frosty owl]

600 is definetly a good number of to work with.
Just throwing it out there for clarity's sake though: using max throughput from a belt is not impossible, "just" needing extra care if the belt is longer than one belt segment

[still brook]

Oh right, I'll keep that in mind to always split it up for things like quick wire

[frosty owl]

If interested, you can find some numbers about the loss you can expect depending on your segment-count in the "b2b" thread in this channel

[versed violet]

Satis wiki keeps opening in dark layout. I switched to light layout several times, but it gets reset to dark. Is that normal or is my browser borked?
-> satisfactory.fandom.com

[median heath]

It's been doing the same for me. 😦

[barren elm]

treat your train station docks just as you would any other manifold

throw in every input via splitter manifold, and every output via merger manifold, then you only have to worry about remembering the total inputs/outputs

Sure it's technically inefficient and could be done without a manifold, but if you wanna remove the need to remember weird numbers...

[versed violet]

the game update respawned all the crab hatchers, including ones inside our factory floors? Any way to handle them?

[median heath]

Delete, kill, rebuild 🤷‍♂️

[versed violet]

nobelisks don't work through walls

[median heath]

oof

[versed violet]

It looks our steel foundry has an bug infestation! [roleplays any staple alien hunting game]

[wind spade]

finally dark by default 🙂

[median heath]

How do I change it back?

[wind spade]

no idea. I have dark theme everywhere and never bothered with light theme

[upbeat tide]

Solution. Build a ton of nuke nobelisks, drop everywhere, go hide far away, detonate 50 at once
... watch PC cry

[median heath]

@versed violet just delete everything, kill them all, then use your modded "undo button" to build it all back in 1 move 😉

[barren elm]

Official undo button when

[median heath]

Never 🙂

[wind spade]

undo button is almost impossible

[versed violet]

does nuke nobelisks work through walls? 🤔

[versed violet]

No undo button... yet. There is one for foliage if you blast too much, but not for things you built.

Is this supposed to happen on bidirectional rail, if both sides have block signals on entry?

[median heath]

So... | There is not a mod for that

😉

[versed violet]

Yet 😉

[wind spade]

I don't think there ever will be

how do you implement undo button in a running simulation?

[versed violet]

you would have to keep track of all building construction/deconstruction. Would be quite heavy memory wise & touching a lot of game internals, but theoretically, possible

[wind spade]

not really

[median heath]

128 GB ram required. 😉

64 to run the game, 64 to run the undo button.

[wind spade]

what if because of the simulation the undo isn't possible? what kind of things do you want to undo anyway?

[versed violet]

I'd assume building/dismantling things. this is a building game after all.

[wind spade]

what if you dismantle something and then other player builds something in that place and you press undo?

[versed violet]

you assume the mod would be multiplayer compatible

[wind spade]

I don't assume it's a mod, I assume it's a functionality that needs to work

what if you dismantle something and trash the items from it and then want to undo it?

[versed violet]

error: not enough materials

[wind spade]

what if you dismantle a wire that crashes your network and then want to undo it?

[versed violet]

it restores the wire. you still have to reset fuse manually

if that happened to be one important wire in mess of dozen of wires, undo would be nice

[wind spade]

there's just so many cases that need to be handled because it's a simulation, it doesn't have states to go back to

pretty much impossible to do properly

[versed violet]

you don't have to store states, this isn't mspaint

just store the actions and replay them backward/forward.

kinda like text editors do

[wind spade]

which isn't always possible because it's a simulation 🤷‍♂️

text editors have states, they are not a simulation

[versed violet]

text editors store undo/redo as operations, not as state of whole document. at least some of them do, implementation may vary

[wind spade]

that doesn't change anything about it having or not having a state

point is that you can always do the "operation" to get back to the same state as you were before. Which isn't possible with a simulation

because the state may have already been changed between the action and it's undo

[versed violet]

you don't need to store state, if you can apply a set of operations. and whether operation is possible to execute, thats a game logic thing.

[wind spade]

I never said anything about storing a state though

[versed violet]

then there is no problem. just store a list of operations and it should work.

[wind spade]

it's not about storing, it's about possibility to make the functionality

if a state can change on it's own between action and undo, then making proper undo button is impossible

to follow up to your text editor example: if you wrote abcd and then you wanted to delete bc, but in between state changes to abXcd, what happens now?

[versed violet]

false premise. You assume world state and buildings state to be the same thing. I picture the undo button affecting solely the buildables. whatever world does in the meantime is irrelevant - only the buildings being added destroyed. You could compare it to writing abc, changing font color and taking back the bc. It doesn't conflict.

[wind spade]

buildings are part of the game state tho

[versed violet]

as far as I'm aware, creating buildings and dismantling them does not crash the game, so such operations are allowed.

even more, dismantling buildings while they are processing parts is also allowed and works.

we also have functionality to mass diamntle and mass construct (soon to be in vanilla), so basic building pieces are available, we only need a way to track player actions.

[vapid gorge]

it sounds like the g ame would have to remember every obj you build which would very quickly increase the number of obj in the game

[versed violet]

it would have to remember kind and location of objects, and sequence of them - basically what it stores in savegame, just uncompressed. Probably limited to some sane number of steps - 10, 100? or couple last minutes.

[wicked tinsel]

i also support having undo functionality, it would really improve a lot of things

especially since actions are not symmetric in this game, particularly due to belts not auto attaching

sure, it might be tricky to implement this, but ultimately, it would be worth it, imho

it doesnt need to store everything, it can be depth limited, a lot of undo functionality is and its perfectly acceptable

also tbh, i have no idea what greeny was trying to said, state in this game can only be influenced by player and nothing else

the only exception is railway, but its a minor one

and even that is very simple to work around with, "cant undo railway construction, train in the way"

[versed violet]

a programmatic undo would handle belts as well, they are a buildable.

[wicked tinsel]

you can dismantle everything and you can build everything at any moment, its impossible to get into a state where last action cant be undone

but undoing last action manually is often epic pain

[versed violet]

if you deleted your last grid-aligned support - YES

[wicked tinsel]

even such mundane actions like accidentally deleting a manufacturer

how much pain it is to rebuild it is inconvincible

nevermind dismantling stuff like train station by accident

gl rebuilding that

(or you zooped some foundations bellow map and now its impossible to delete them :D)

happened at least few times

[versed violet]

that really could use an undo button

[fierce ruin]

could be interesting

[wicked tinsel]

https://questions.satisfactorygame.com/post/5e5bfa29c672159e0cdf89b5 there is a ticket for it

[fierce ruin]

specially when you deconstruct smth by accident when mass deconstructing 😭

[wicked tinsel]

from my experience, you know very fast that you fcked up, even having "undo up to last 5 actions" would be a great thing

it doesnt need to undo all the way to beginning of game

with blueprints its going to be even worse as collisions are no longer a thing

so you can perfectly misclick and mash your blueprint right into core of your factory 😄

gl fixing that

[versed violet]

nobody said bluerints will ignore colissions. that would be wrong

[wicked tinsel]

why wouldnt they ignore collisions?

soft clearance is a thing, it probably will be for blueprints too

[versed violet]

you can't mash two machines together.

[wicked tinsel]

well no

but you can perfectly mash bunch of splitters right into your manufacturer

[versed violet]

that could happen

[wicked tinsel]

and how are you going to fix this 😄

[versed violet]

but you will have plenty of time to observe in horror as they slowly buld up

[wicked tinsel]

yep 😄

cool main storage sorter you have here, it would be a shame if you pasted a balancer on top of it

[raw flame]

i need some help, i need to split a line of 290 into a line a 155 and a line of 135. how do i do this?

[median heath]

Single splitter.

[raw flame]

oh

thats it?

[median heath]

290/2 = 145
145 > 135

135 side eventually fills and from then on can take only 135.
After which you have a perfect 135/155 split.

[raw flame]

so just put one splitter and it will work?

[median heath]

Yes.

If you want to skip the spool time just pre-fill all the machines on the 135 side so it backs up immediately.

[raw flame]

ok

[raw flame]

@median heath I have a line of 155 and I need to split it into a line of 95 and a line of 60. Is it just 1 splitter?

[signal nimbus]

I'll confirm it. Yes.

[median heath]

Same concept, so yes.

[signal nimbus]

One of the lines out of there needs to be a Mk 1.

[median heath]

False.

[signal nimbus]

I'm listening, but that's the simplest way to split out 60 from any line.

[median heath]

You can do a 95/60 split with all mk5s if you want.

[raw flame]

so one splitter?

[median heath]

Yes.

[raw flame]

i have mk4s

[median heath]

Works just fine.

[raw flame]

ok

[signal nimbus]

I'm sure you can... how is it simpler?

[median heath]

No one argued simplicity.
I said your claim of "needs to be mk1" was false.

[signal nimbus]

I'm still not hearing how it works.

Manifold/overflow, yeah, doesn't matter.

[median heath]

155/2 = 77.5
77.5 > 60

60 side eventually fills and from then on can take only 60.
After which you have a perfect 95/60 split.

[signal nimbus]

So Manifold/overflow.

[raw flame]

sorry to bother you again but i have a line of 380 and i need to split it into a 200 and a 180, is it one splitter again?

[signal nimbus]

Look, I agree that it's a correct answer. When someone asks "how do I split X from Y" and I answer "like this", don't... don't put it down.

[raw flame]

sooooooo i shouldnt trust that answer?

[signal nimbus]

It's... correct.

It's not elegant and it's not load-balancing.

[raw flame]

@median heath

[median heath]

Yes.

[raw flame]

ok

[signal nimbus]

But by that standard, the correct answer will always be "one splitter".

[median heath]

You don't ever need to load balance because resources are infinite.

If you choose to spend more time and take more space making balancers to get exact splits, that's a perfectly fine choice you can make.

Manifolding is simpler, achieves the exact same end result, and takes far less space.

[signal nimbus]

You don't need to preach it, I use manifolds almost 100% of the time.

But when someone asks that question, they're generally looking for load balancing, and there are reasons for that.

[median heath]

So?

Most people across the past 4 years of asking that question are thinking balancer because they haven't had manifolds explained and it shown how they are equivalent in results.

[signal nimbus]

I'm aware.

[median heath]

So for most people telling them 1 splitter is both an answer and showing them a simpler way to do things.

If someone needs a balancer for a specific reason, they state that after being told "1 splitter".

These are opposing statements to me.

[signal nimbus]

They... really aren't.

[median heath]

Either you're aware that my way of answering is helpful, or you thinking that people want only balancer-related answers.

[signal nimbus]

My intent is to answer the question asked, not say "you're thinking wrong".

I'm aware you're technically correct and I hate it when someone puts down the way I'm trying to do it.

[median heath]

Yeah I'm out.
We are of opposing minds on this and aren't going to agree, neither will we change how we answer people.

[cinder silo]

I reserve balancers for radioactive items alone, manifolds are just plain simpler.

[paper girder]

The original question was splitting 380 into a line of 200 and a line of 180, so a single splitter wouldn't be a helpful answer as a splitter splits evenly right? Which would make it 2 lines of 190. Not trying to be antagonistic even though it's plainly coming off that way, just pure drunk curiosity

[median heath]

Which would make it 2 lines of 190.

Yes, which I start with. And I explain how that evens out once the side lower than 190 fills.

[frosty owl]

I don't get how no fill-up time and no stacks of items in machines can be summed up with "no difference"
As if turning on factories at once (takes forever manifolding) or dealing with radiation isn't a thing at all? O.o

[median heath]

Radiation scenarios are rare and usually clarified.
I also denote the "end result" is the same, accounting for spool time. Which I also explain how to skip.

[cinder silo]

You can split 380 to 180 and 200 by using 1 splitter, but you also need a merger, the 180 is created using a mk1 belt (60) and a mk2 belt (120) , those two merge on to a mk3 belt to take the 180, the third and final splitter output will be 200, again on a mk3 belt, choked outputs as described are quite messy.

[wind spade]

Or you just need a single splitter and let it overflow

[cinder silo]

I know, I was merely answering the question without bias.

[tidal wren]

this was a lot of communication for what amount to :

choked balancing is the answer you seek
manifold is the answer we feel you need

[oblique hollow]

the best way to balance outputs is to play with machine clockspeed and NOT merging all products onto one belt directly

in a way, the machines are your ratio splitters

[barren elm]

That's a lesson screws are meant to teach you

It hurts seeing people make 50 belts worth of screws before shipping them off to other factories

[wind spade]

or ditching screws altogether

"screws suck"
well they suck because you make them centralised

[frosty owl]

Assuming one has a stable enough input being provided and using a smart splitter... This can be true ^^

In other words: a normal splitter may not work as intended and discontinuous input could lead to issues even when using a smart splitter

[fierce ruin]

stiched reinforced iron plates to the rescue 🙂

[wind spade]

well there's more usages for screws than RIPs 🙂

[fierce ruin]

I know, my point was more that when I can get rid of screws anywhere, I will take the opportunity generally

[wind spade]

f.e. rotors are great with screws

[frosty owl]

I don't get what you're saying about nuclear.
Regardless, I believe that having the machines run with a full inventory or an almost empty one is not a difference that should be just ignored or brought up whenever one feels like it: saying they're the "exact same" does that imo

[barren elm]

Nuclear punishes you for holding above the bare minimum of each radioactive resource

[wind spade]

"punishes" only if you are near the factory

[barren elm]

True, you absolutely don't need to care about said punishment, just that it exists

[cinder silo]

Not much of a punishment even if you manifold radioactive sources, that's what the rad suit is for, some of us mitigate the hazards, others are fine if the world burns, this is a sandbox and we do our own thing 😛

[barren elm]

Right, but it exists

I'm not saying that you must care about it

I actually really like the nuclear mechanic and hope there's other rifs on this "manifolds perhaps not the best" concept in future

[frosty owl]

I'm inclined to agree, but I reckon that depends on how the Devs want the logistics to feel "complicated"
Having a "basic" and relatively easy solution to all problems could be somethinf they don't want to stray from (having manifold not work in some situations Vs how they currently just cause slow warm-up & radiation in some situations)

[wind spade]

direct input bestest

[barren elm]

I do wonder whether having a gigantic direct input factory, even if some machines end up with really small overclock values, lags less than a traditional factory

[frosty owl]

Funfact: direct input makes for some of the most reliable single-input sushis

[wicked tinsel]

manifolds would work fine if their implementation wasnt so braindead

[frosty owl]

Oh, the "ideal factory for performance/logistics" argument. Love that

[frosty owl]

You really have a way with words...

[wicked tinsel]

there is absolutely no reason for machines to automatically load up to stack size

and this is the core problem with manifolds

[barren elm]

Every factory game works that way

[wicked tinsel]

well, factorio does not, and this game clones factorio 😄

[wind spade]

it... doesn't?

[cinder silo]

Avoiding the startup issues with manifolds is straight forward, either do a staged startup where the whole thing is primed by its own action then is filled before the next stage, or just add the stuff manually.

[oblique hollow]

its similar, yet the ideas branch

its more like parallel universe factorio

[wicked tinsel]

core concepts are similar

and manifold is one of them

its just badly implemented here

[oblique hollow]

yeah but core concepts are things that kinda define genres

[frosty owl]

Without it, any long-range manifold could need having a loop for the items to travel around to keep up with the deliveries.
Items come in, pass in front of all machines (giving 1/2 to each) and just overflow out of the system (unless it's a closed one ofc), leaving the system empty until the next delivery

"long range" as in: fed by long-range transport

[cinder silo]

Funny you mention loops, I considered that for some of the sushi and the initial uranium intake to cut radiation 🙂

[wicked tinsel]

no, it would change nothing for long range transport

[barren elm]

Factorio adds in double the resource requirement of the assembler which can be way more than a single stack size

[oblique hollow]

i think for most recipes, 3 times the demand per cycle should be an ok stack size. and if 3 x is larger than the items normal stack size, it only does double

[frosty owl]

To short a message to comment on 🤷‍♂️
I can only repeat that I don't think so (see my prior message)

[oblique hollow]

our recipes are all set so they never go over half a particular item's stack size

so reduction is ok

[wicked tinsel]

most of them dont even go to 1/10th of stack

[oblique hollow]

its not off the table yet for future updates

[wicked tinsel]

it should load like up to two cycles of resources, and then everything is fine

[oblique hollow]

i think theres a QA site ticket for that... right? there better be

[wicked tinsel]

there is

[oblique hollow]

got a link?

[frosty owl]

Loops are usually a bad idea for sushi 🤔
Could you walk me through your thought process there? I often hear of people considering sushi loops but don't quite get where this (imo bad) thinking stems from

[wicked tinsel]

but people like to pretend that the issue is with the manifold concept itself, where in reality, its just an issue with auto loading

[oblique hollow]

probably because they simply gave machines a normal inventory

with no extra checks

eh, they'll likely redo that once they do a pass over all recipes

thats the perfect time for that

[wicked tinsel]

https://questions.satisfactorygame.com/post/60518819aa0ba107e32511e3

[oblique hollow]

thanks

[cinder silo]

I said I considered it not went with it, I realised early on that it would unbalance things and cause a clog, Just some parts of my facility flow through with excess inputs and deals with them down the line, I was hoping to eliminate that, the uranium plant has its uranium ore on its own feed and I wondered if looping that would work, even that didn't go as planned simply because of the blender buffer holding on to a shit ton of ore, the idea while it sounded cool was abandoned 😦

[wicked tinsel]

manifolding stuff like encased beams is epic pain

[oblique hollow]

and balancers arent really that meta

...at least thats what it feels like

[cinder silo]

Manifolding encased beams is dead easy, what are you on about?

[oblique hollow]

it takes time to fill

[wicked tinsel]

well, you need to preload 500 concrete per machine

[oblique hollow]

its not difficult to manifold

but long winded to get it running

[wicked tinsel]

make 20 machines and you need to produce 10000 concrete before it can start working

[snow dove]

time consuming≠difficult

[wicked tinsel]

painful != difficult

[oblique hollow]

its the same issue of fuel gen setups or fluid stuff in general.

a lot of time wasted

[cinder silo]

YAY!, someone who actually uses the proper does not equals sign instead of just an equals.

[oblique hollow]

waiting for it all to fill

50 m3 is a lot

[snow dove]

just letting it run in the background while you grab a snack is painful?

[oblique hollow]

a few minutes is funny

20 turbofuel gens take 30 minutes until you can detect problems

[wind spade]

!= ftw

[arctic willow]

concrete isn't the issue with manifolding eibs so much as the steel component

[oblique hollow]

thats 30 minutes of time wasted until you can do troubleshooting

[cinder silo]

When I see that I just consider it as equals 🤷‍♂️ , ≠ or bust.

[frosty owl]

You have a factory receiving a bunch of items at once every 10 minutes. Factory A processes most of such items, factory B uses most overflow and the remaining goes to storage
Machines need 1 item per production cycle (stack of 100) so they take up to 2 whenever possible (your logic).
What happens is: items come in, deposit 2 items per machine and everything else reaches factory B (more than one wants). Same happens for factory B and a humongous amount of items end up reaching the storage (just 2 items per machine stay behind) while the machines will starve until the next delivery comes

Please correct what I got wrong

[wicked tinsel]

a single 300 rock mine makes 100 concrete a minute, it will take hour and 40 minutes for it to fill 20 machines

[oblique hollow]

non-progs will not understand

[wicked tinsel]

its not few minutes

[wind spade]

well that may be just my coding "career" then 😄

[oblique hollow]

thats a logistical issue for you to solve

put a storage container somewhere

[wicked tinsel]

normally, you would use buffer that will average out the janky input

like .. train station's internal buffer

[oblique hollow]

if you have discontinuous production, you use buffers

thats factory 101

[wicked tinsel]

sure, there is an unrelated problem of not being able to limit buffer's internal size

[oblique hollow]

but it doesnt need to be the machine buffer

[wicked tinsel]

but its not a core problem of machines themselves

[oblique hollow]

would also give fluid buffers more reason to be used

[wicked tinsel]

^

[cinder silo]

Probably, I wasn't good at coding back in college, I learned the hard way just how much of a pain it can be when bug hunting in even the most basic program, 1 typo in a data field had me rewriting the whole damned thing causing numerous extra bugs with each change until the whole thing was unrecognisable, then I saw the typo, epic fail on my part that told me it is one career I couldn't do.

[oblique hollow]

there plain currently is no demand for buffer spaces because fluid recipes by default already store like 20 times the demand

except for alumina

or blender recipes

[wicked tinsel]

the big problem with current approach is, some items take forever to construct

and then manifolding those take days to fill

[frosty owl]

I feel like you're not attempting to have me understand...
What's the station' buffer gonna do about the situation I described?
It'll just empty out as fast as the belts allow, putting 1/2 items inside each machine and having all the rest of the items reach the end of the line. The machines will get more items only if they manage to finish a production cycle before the item-stream ends

[wicked tinsel]

like uranium reprocessing, you need to fill it with Pressure Conversion Cubes, 50 for each machine

[frosty owl]

How is that an issue when one can easily load-balance?
High-tier items are especially easy to balance, btw. Most have quite handy numbers/min

[cinder silo]

The only manifold that I encountered that would have taken days is my reactor hall, filling 100 reactors with fuel, everything else has been much quicker than that.

I tore it all out after three hours and built the 1-100 balancer, the entire thing was producing power within 5 minutes and no death zone.

[oblique hollow]

from experience it can still take hours

waiting hours until a problem shows up isnt realy fun 😢

[cinder silo]

Three hours in my case had 18 reactors of the 100 running, and a lot of radiation.

[wicked tinsel]

say your factory A needs 100 items per 10 minutes, it means it needs 10 items per minute
you have a buffer building that holds 100 items, it gets consumed over 10 minutes, then you refill it with your 700items/min belt in few seconds once items are available

you just moved the buffering from machines to dedicated building

[frosty owl]

That applies to machines as well 🤷‍♂️
If an input belt has discontinuous input, the internal buffer of a machine can (and will) help with that

[oblique hollow]

eh, they delay any issue

but you dont want delays

you want responsiveness

[wicked tinsel]

the problem is that you dont have options and this should be a game about solving issues

[frosty owl]

Please address the example I provided accurately. You're assuming that the items can back up after factory A, that is not the case for the example I gave