math-and-meta

[plucky tusk]

blender turbo fuel is pretty fun ngl

[prisma kraken]

yeah, its a pretty great recipe

this is how i set it up for a build which allows me to split it into 4 independent units

[plucky tusk]

yeah im just gonna mirror that and then surround with gens ig

10 blenders max oc is what im makin

like 1125 turbofuel

[prisma kraken]

how that actually ends up looking

i've already done a rocket fuel conversion of it, so its 3x what the tf would yield 😄

[plucky tusk]

this what im at rn

[prisma kraken]

yeah, i clock everything up, so probably using less machines

[plucky tusk]

prolly gonna make pet coke off to the side and belt it in

then its just miner-belt-splitter for the sulfur

[prisma kraken]

i just group 3 hor refineries output together and push that exact amount to a dilluted fuel blender, a coke refinery along side and then the 4 tf blenders

[plucky tusk]

yeah

u could totally do that

[prisma kraken]

if you make the exact amt of coke needed for the blenders right nearby it saves some belting headaches

[plucky tusk]

belting doesnt give me headaches

its the math that does

[prisma kraken]

4 input blender recipes aren't a cakewalk to deal with, so any shortcuts I totally try to use

[plucky tusk]

yeah well i usually optimize toward logistics so its the other way around for me math hard belt easy

like the left side is done

[prisma kraken]

all the beltwork for one of my modules

[plucky tusk]

belt floor?

[prisma kraken]

yeah, just a 4m subfloor

[plucky tusk]

thats crazy when you can just put it all on the machines lol

[prisma kraken]

ehh, i'm bringing the sulfur in by train

i didn't want to run belts a long distance across potential building space

[plucky tusk]

yeah i thought of something crazy lol

5 freight platforms each side 1-1 feeding the blenders lmao

[prisma kraken]

i mean you can, but it makes more sense to use 125% or 250% clocking to turn a 5 into a 4

[plucky tusk]

yeah ur right u can totally do that

[prisma kraken]

that's what i'm doing

[plucky tusk]

i only need like 56.25 pm for each blender so i just thought it would be funny to have 10 freight platforms

[prisma kraken]

i use a 4 frieght standard size for my trains. anything that gets loaded into a train is evenly distributed btw 4 cars

for lower volume stuff, it means trains can make longer round trips and then for high volume stuff, i'm splitting it all up where it is produced rather than where consumed, and i'm finding that's just a nicer way of doing things

gets rid of a bunch of crazy balancer building

[plucky tusk]

sadly cant fit 2 staions on top so ill just split 5 ig

idk man i think the train looks stupid, also cant split 5 in 2 with trans

[regal pier]

I am confused. If I have mk.2 pipes, they move 600 units per min. And Fuel Gens use 7.5 turbofuel per min. I should be able to fill 80 Gens, right?

[vapid gorge]

if you set it up right.
pipes aren't belts

[regal pier]

the last 10 or so are on and off again and I thought I was matching them

[vapid gorge]

over head images of the set up would help

[regal pier]

I have a big loop that goes around the gens

[vapid gorge]

as high up as you can

[crimson moat]

Manifolds cause backflow which cuts into the effective flow rate of the pipes, so if you want to manifold from a 100% flowrate pipe you have to split it before.

turbofuel gens are worst case scenario for that issue

[vapid gorge]

?
systems with the lowest input per min tend to be more stable. Pre 1.0 you could run a 600 line w/o loop of TF

[regal pier]

using photo mode

one sec I'll grab a screen from the Calculator

[vapid gorge]

can you sketch how the pipes are laid out under the walkways? it's all covered

or save, delete most of hte walk ways and show? can reload after

and maybe a few shots from the side

[regal pier]

Does this help for pipes?

I can send more screens as well, one sec again

[vapid gorge]

yeah sides

[regal pier]

outside looking in

center splits

[vapid gorge]

how does that look at the start of the manifold?

[regal pier]

inside of outside

its a single pipe that splits left and right to make the loop

[vapid gorge]

ok so only the right one is currently hooked up?

[regal pier]

yeah. I noticed it wasnt running right when I started working on the next batch

[vapid gorge]

ok did you pre flood the system?

[regal pier]

I went around and double checked the connections between blueprints and pipes, no missing connectors

I did not, But I could turn it off to flood the pipes. The fluid buffers are full

[vapid gorge]

what fluid buffers? the internal ones?

[regal pier]

I have 2 fluid buffers on each pipeline. They are right after the blenders

i think thats 4800 units?

[vapid gorge]

delete them
flood the system by down clocking a few fuel gens like 50%
once everything is flooded, clock them back up and see the behaviour

[regal pier]

I will try that

Why no buffers?

[vapid gorge]

buffers wil generally just mess up a system.

basically the only place to use them is to buffer fluid trains

[regal pier]

makes some sense, I will try it

[vapid gorge]

essentially they don't give you any benefit and can cause issues

you can have them in a system if you know what they are doing and it can run. But it won't be running Because of the buffers, they'll be running despite them

[plucky tusk]

I was gonna put buffers on mine but the big ones didnt fit and small ones were too small

[vapid gorge]

small ones were also too big

unless you mean for trains

[plucky tusk]

No off the blenders

[vapid gorge]

yeah. don't use them. just makes your life easier

[plucky tusk]

🤓

It was more for looks tbh

Since the refineries hang off the back of the blenders

[vapid gorge]

clip the pipes through them then. I use them for aesthetics myself

[plucky tusk]

Well like i said neither worked out

Also im not just running a pipe through of course that wouldnt work

Set it up the same way u do buffers for fluid trains

[vapid gorge]

if it's not for trains it's not going to give any benefit to a fluid system

at best, they do nothing.

[plucky tusk]

i kniw but all the pipes would look cool

[vapid gorge]

so just clip the pipes through the buffers. It looks the same and you don't get the mess of having a buffer in your system

[plucky tusk]

I dont want fake stuff tho im fine with waiting for the fill. everything will be connected and running before i even build the generators, so it will give me tine to fix stuff anyway

[regal pier]

Looks like everything is working. I took out the buffers and sent the fluid vertical for a "water tower" effect. Gens all seem to be staying on now

[vapid gorge]

well, water towers are unlikely to hurt your system but it's certainly not doing anything. Especially not doing anything a pump at the start of the manifold wouldn't do

[latent anchor]

In some setups that you want to have automatically adjust, yes. In the dark matter crystals example you could be taking as much as possible from the factory (necessitating that it switches to maximal dark matter efficiency) but given that both recipes are supplied with DMR and both are merged into the output without priority, occasionally a crystal will flow from the inefficient recipe and use extra DMR for no additional crystals

[vapid gorge]

i've never had a darm matter process that a smart splitter, at worst, wasn't a solution

[latent anchor]

Even if you were to use overflow systems to fuel the DMR-efficient recipe before the diamond-efficient one, as soon as DMR became abundant and diamonds not so much you'd face the same problem

Btw you can certainly get 99.999...% as I said but not 100

[frosty owl]

Note: 100% priority mergers have been invented in pre-1.0 vanilla 😉
Ah, the beauty of brainpower

[vapid gorge]

also never had an issue with a manifold, in any set up, not being 100%

is this going to be teh aeryn 'manifolds aren't effcient' argument again?

[latent anchor]

Nope

Just saying that there are mathematical (if not practical) advantages to priority mergers

[frosty owl]

They do be a lot more compact than their pre-1.0 coubterparts ^^

[latent anchor]

Yep! I managed to come up with a design centred around an industrial storage. It needed 47 stacks of concrete

[vapid gorge]

I've never managed to design a system, involving those items, that wasn't 100% effcient

[frosty owl]

A design using PMs (Priorury Mergers) or trying to make a PM?

[latent anchor]

Making a PM in pre-1.0

[latent anchor]

Mine took time to settle but settled perfectly

[frosty owl]

Ah. Was that a 99% model?
The 100% model needed 2 ISCs (iirc)

[latent anchor]

Nope, mine only needed 1 additional ordinary storage container that was almost redundant

[frosty owl]

Very small contraptions anyway, needing entire STORAGE CONTAINERS to run 😆

[vapid gorge]

I mean, I can make those systems, w/o p mergers, be 100%. I don't understand why you think they need them

[latent anchor]

Full storage containers, too

[frosty owl]

Eh, tbf, that's easy enough using low-stack items... Dunno why Concrete (500-stack) was your first choice in that regard :P

[latent anchor]

You make it practically 100%, whether it's 99.9% or 99.9999999%, while I wish for 100% (for some reason idk). If there's some other way to circumvent the problem of selecting between a number of recipes according to demand always having a case where a non-desirable recipe gets used sometimes, let me know

[vapid gorge]

No. My manifolds are 100%.

like, draw a diagram of the case you're talkign about, but my dark matter manifolds run at 100%

[latent anchor]

I'll try

Gotta get discord working on my laptop again rq

[vapid gorge]

cause it's entirely possible to build those systems, w/o pmergs, and have them run at 100%

in fact, I would say it's trivial

[frosty owl]

I'd argue that if it were trivial, you would've managed to explain your disagreement by now

[vapid gorge]

It's not clear what the hell they're talking about. You can make the dark matter waste production run perfectly efficiently

[latent anchor]

Let me show you

[vapid gorge]

they've talked about random item production that they claim can't be run at 100% efficiency w/o p mergers.

[latent anchor]

Gonna MS paint for a moment

[vapid gorge]

while, I, have made those items, perfectly efficiently, w/o them

can you set up systems in wildly dumb ways that might require a p merg to work properly? sure

[latent anchor]

I'm not debating that you have them almost indistinguishably close to perfect

[vapid gorge]

ok bugger off with your claims the yaren't perfect

I'm utterly sick of you bs

[latent anchor]

Sorry, I'm a pedantic mathematician! I'll stop after I draw this diagram. Didn't know I ticked you off, now I can see I've been a prick tho

[vapid gorge]

youve never explained why any partricular manifold wouldn't be at 100% I no longer care about anythign you ever have to say

[latent anchor]

I need time!

[frosty owl]

Reminder of #rules about politeness and such~

[vapid gorge]

yeah, random, unsubstantiated nonsense gets to be called bs. And I get to tell people I no longer care about what they have to say

[latent anchor]

That's fine by me, I'll make my point to others even if you're not interested

[thorn bane]

do you?
<@&387163995947270144>

[latent anchor]

Still drawing my stuff

Please don't punish this dude

[frosty owl]

Eh, worth making that one point clear imo (the ping's active already anyway)

[edgy leaf]

how does your dark matter crystal production look like COD? cause to me it seemed like a priority merge of sorts was necessary if you want to use trap with dark matter crystallization while recycling the dark matter

not interested in the argument, just curious

[frosty owl]

I am saddened that there is no "belt" emoji for me to add as a reaction to this...

[wind spade]

Clock machines and do direct input?

[frosty owl]

Can't do that because No InSErTeRs!

[vapid gorge]

any of the dark matter residue over flows were dealt with similarly to keeping alumina waste water split? Like, I put so little thought into it and got it working it has not made a lasting impression.

[edgy leaf]

wouldn't that potentially get clogged if the belt fills up past the merger?

[wind spade]

there is no merger if it's direct input?

or I'm not sure what do you mean

[edgy leaf]

ah, with two machines? yea that'd work I think

[latent anchor]

In this setup, things are 100% efficient when you need more crystals and the 5 -> 2 recipe is used, but if the output backs up to where you want to only use the 10 -> 1 recipe, the non-prio merger means sometimes you use 5 -> 2 and waste time crystals.

Second setup in a moment

[frosty owl]

You can press P to take screenshots...

[latent anchor]

On the other hand, if you wire it up to prioritize the 10 -> 1, it will get used even when you want to use only 5 -> 2

[wind spade]

you can press printscreen to take screenshots 😛

[latent anchor]

Discord discriminates against us Australians and our bad internet 😔 I cannot open it on my PC

I use phone-cam because it needs to end up on this device one way or another...

[frosty owl]

Don't mind my mumbling (unless you find it funny)~

[latent anchor]

I find most things funny

[unique cypress]

Wouldn't this require a priority splitter for gasses?

[latent anchor]

Trains!

They're not needed at all when you use a prio-merge setup

[unique cypress]

What is this belt/pipe for?

DMR?

[frosty owl]

@latent anchor why are the priorities for input opposite of the outputs? It also looks like you prioritized the output of 10:1 over 5:2 in the first picture and the opposite in the other, which would lead to the opposite result of what you described

[latent anchor]

In this case it would be, yes.

[thorn bane]

if you know how much crystals you need then you never need the 10->1
as in math out how much youre making and you wont need the prio merger

[unique cypress]

How do you plan on making a priority splitter for DMR?

[frosty owl]

Let's calm down and not pile on poor Bilbo too much 😆

[latent anchor]

So that the system auto-balances. The idea of the first one is that the 5 -> 2 runs with priority but if it backs up and stalls it switches to the diamond-efficient 10 -> 1 recipe. If you give a recipe high input and output priority, the other one sees no use

[frosty owl]

Gases tend to fill the first machine connected to a pipe... I think one can leverage that to make some pretty good priority

[prisma kraken]

i might ask why you even need the priority merger. it seems like it ought to self-level

[frosty owl]

The output priority may suffice imo too

[latent anchor]

You can't always know. If you make a dark matter crystal setup for future factories and need all the diamonds/time crystals you get (like in a megabase) you can't know how many crystals you need at a moment in time

[latent anchor]

Trains!

[frosty owl]

I think I understood the first part correctly, but didn't you (in the first picture) give the 5:2 system high input priority but low output priority, countering your explanation?

Might just be a typo, but I want to clarify and make sure ^^

[thorn bane]

well you can if you know your plans and dont build for the future
as in treat it all as 1 factory and use those numbers
but ye i agree that this way is easier

[unique cypress]

Oh I get it now... This is so dumb lol 🤦‍♂️

[thorn bane]

ye think so as well
it would just fill and stall
i guess you save the 200 dark crystals xD

[latent anchor]

Assume the system is empty. DMR flows to the 5 -> 2 and as the 10 -> 1 is not running, the items are free to flow through. The 5 -> 2 keeps running. Now assume the system is full but has maximal demand for crystals. The 10 -> 1 setup can't keep up with demand, the 5 -> 2 boots up and starves the 10 -> 1

[latent anchor]

It's cool, too. I love seeing a factory adapt. Life finds a way

[frosty owl]

So one of the conditions is that if one works, the other can't (Edit) get input, right? ^^

[thorn bane]

i think the point of people like cobalt is that you can make it without this if you just math it out

[latent anchor]

Yep! If the 5 -> 2 is needed the 10 -> 1 will stop, and if the 10 -> 1 will do the 5 -> 2 backs up and clogs

[frosty owl]

That part was a bit confusing as it's not obvious (imo) ^^

[latent anchor]

Exactly. I was happy to walk away and allow each their own as a factory that works works, but as a mathematician I feel the need to be a jerk online inform people

[thorn bane]

wait if youre already using trains isnt it super simple to make a perfect priority merger with trains?

[frosty owl]

I reckon it's more compact using the actual priority merger or even the ISC-based pre-1.1 contraptions...

[latent anchor]

It is, yes. You set a train to pick up from the high priority station, then the second-highest priorty, etc. and then unload it where you need

[thorn bane]

oh flash backs to where i did alu that way by shipping water
it was horrible xD

[latent anchor]

"Simple" here isn't exactly "compact"

[latent anchor]

Your first mistake in the factory

Your second was to persist with the idea, was it not?

[prisma kraken]

trainifolds are like the secret weapon of the game

[thorn bane]

4 years ago fuck im old
#math-and-meta message
well i like restarting a lot so not really xD

[unique cypress]

I'm doing my 1.1 aluminum water with the new priority merger. Yes, I'll be packaging it to do that

[thorn bane]

could already do that in 1.0 since that doesnt need a perfect priority
but ye fuck packagers it makes it like 3x as big

[frosty owl]

Do you like to restart or do you dislike to finish?

[unique cypress]

I know, but it never occurred to me before

[thorn bane]

i finish fast so i restart a lot

[latent anchor]

My first priority merger design involved packaging/unpackaging alumina solution... the way things circle back

[latent anchor]

Insert crude, typical follow-up statement

[thorn bane]

"i think im actually going insane considering this btw"
#math-and-meta message
no way, the foreshadowing

[fierce ruin]

Is this bad? Does going over 20 headlift reduce the flow rate or lead to any other such issues?

And if so, does this seemingly false "headflift is exceeding recommended level" issue mean anything?

[deft lichen]

you mustn't exceed the maximum, which is 12 m for buildings, 22 m for mk1 pumps and 55 m for mk2 pumps

the warning is just that, a warning

[dusky dust]

You can start seeing weird flow issues if you go too far over. In that example you seem to be right on the limit, so you're probably fine. But in general I like staying a meter or two below the limit

[deft lichen]

!wikisearch head+lift

[brisk shore]

https://satisfactory.wiki.gg/index.php?search=head+lift

[deft lichen]

hm, image doesn't embed anymore

[dusky dust]

I still need to check over that image; I know at least the water extractor is misleading on there

[fierce ruin]

So going over 20 leads to issues?

[deft lichen]

no, over 22 does

[fierce ruin]

I can't seem to go higher than 22 no matter what. The pumps just aren't pumping higher than that

But, whatever I guess

[deft lichen]

yes, that is the maximum

you need to place another pump after those 22

test https://satisfactory.wiki.gg/wiki/Head_lift

hurrah, embed

[dusky dust]

I still wouldn't recommend pushing the pump placement right up to the limits. Flow can get weird if you do (like it'll seem to be fine for awhile but then start failing, etc)

[vapid gorge]

yeah don't push that extra bit. Thats leniency. Even the pumps themselves say 20m

[fierce ruin]

So if I can get water up here, it should be able to flow all the way to the power plants, right?

This is the highest point in the pipeline, and I'm pretty sure the collective rises in elevation do not go higher than this

[vapid gorge]

sure, but you should move the coal to the water. Much easier

or just burn coal that's already next to water.

I'm pretty sure there are more coal nodes near water than not

[fierce ruin]

But the aesthetic of having a giant power plant in the middle of a desert, and a pipeline leading to it

...It has taken me about twelve hours to get this to almost vaguely work

[vapid gorge]

eh, if you want aesthetics? sure.

and yeah long pipes have that effect especially if you're not experienced

[wind spade]

rip index.php

[restive sparrow]

On more than one occasion I've attached a pump to the head lift bubble and ended up with a system that was getting water as high as I needed but throughout was limited. Once I moved the pumps down to be less "on the line" throughput capped out nicely.

[unique cypress]

Iirc the game suggests placing the pumps exactly at the 22/55m mark (the blue rings), and it usually does work fine, but going above means you're relying on the fluid to slosh higher than it should (and that does not work reliably).

I'd still recommend staying within recommend lift

[restive sparrow]

I'm recommending going lower than the recommended because I've been burned by using it. At least twice if not more.

[patent blaze]

I always went overkill and placed mk2 every 30 meters or closer if i was in doubt. It doesnt hurt to be safe

[fierce ruin]

Is there a graph for showing how to split 45 into 20 and 25?

[unique cypress]

Yes

As long as the 20 side consumes 20, it'll work

Alternatively, if only the 25 side consumes 25, you can use a smart splitter instead

[vapid gorge]

it'll just self balance

[plain rivet]

eventually. you might want to prefill the machine(s) on the 20 side if you can so it backs up faster. Otherwise the 25 side could stutter for potentially hours depending on how big the item stack size is and how many machines are on the 20 side.

[vapid gorge]

it won't

even wire doesn't take that long

and that's almost an even split of probably not wire

[plain rivet]

22.5 would be going to each side. that's 2.5 extra to the right. wire would take over 3 hours to stabilize at only one machine

[radiant sinew]

You could always load balance it

[unique cypress]

A single splitter is a balancer. 1:2.

[restive sparrow]

If you wanted to be an absolute lunatic about it (I say because the machines should balance themselves so there shouldn't be any reason to do this) you could get a 20/25 split right on the nose with 4 splitters (breaking the original into 9 pieces by way of 1:3 2 layers deep) and 4 mergers (bring together 4 of the 9 for the 20 with the first 2 mergers and bring together 5 of the 9 for the 25 using the other 2 mergers).

Again, this is purely for the sake of mathing it out and I believe actually doing this would be lunacy.

[unique cypress]

You only need 2 splitters and 2 mergers for a 4:5 ratio splitter. Split into 3, then split one of those into 3 again. Take the second of the first thirds and one of the second thirds (a ninth) and you've got 4/9. Take the remaining belts (a third and 2 ninths) and you've got 5/9.
9x -> 4x, 5x

[restive sparrow]

Good thinking. Bringing the entire thing down to 1/9ths isn't necessary, and actually quite wasteful.

[fierce ruin]

Thanks

I just let it auto-balance a while back lol

[white bloom]

that works of course, but this with just 2 extra machines is as simple a balancer as it gets

[patent blaze]

i never understood these complex items/min splitters, just clock the machine, and overflow the rest into the other belt

[white bloom]

it's 2 extra. complex is a stretch

[patent blaze]

yeah true, but it could be only 1.

[unique cypress]

This is not a balancer lol

[white bloom]

call it what you want, it achieves a specific desired split without external backfill

mapping [45] to [25,20] without requiring downstream buffers to fill up first

[white bloom]

it's faster to build 2 extra than to wait for a buffer of a full item stack to fill up before it settles and works properly

[patent blaze]

true but once it gets filled it'll work indefinietly if the rest of the factory is set up properly

[white bloom]

but it consumes more factory footprint

[patent blaze]

im just lazy i guess

[white bloom]

or mindful of your footprint! don't be too hard on yourself 😉

[patent blaze]

i do tend to cram all the machines into one box so yeah i'd struggle to fit stuff like that

i underestimated the pressure conversion cube production in my nuclear setup greatly, its like a cobweb.

eh anyway gtg work

[unique cypress]

You could also do this so you don't have to wait for the mk1 to saturate

[white bloom]

since the mk1 just needs to be a very short segment it will saturate almost instantly, but yes, that works too with just 1 more machine. So many options

[unique cypress]

The items need to loop around a few times so idk it might take a few minutes

Still, I'd always just use a single splitter

[plucky tusk]

Same lol

[unique cypress]

That's exactly what I posted?

[hoary wigeon]

simple

[unique cypress]

Ah, yes, splitting into 3 only to merge all 3 unchanged. Efficient

[grizzled grove]

Yours looks smelly

[hoary wigeon]

if it works it works

[grizzled grove]

Also… https://icemoonmagic.github.io/Satisfactory-Splitter-Calculator/

[wind spade]

what's so bad about single splitter

[plain rivet]

some splitters are happy on their own. quit pressuring them, grandma

[white bloom]

I like KYO's visual layout more tbh. And as he said, topologically it's exactly the same.

[grizzled grove]

The numbers mason

[plucky tusk]

Yall still talkin about splitting 45 into 20 and 25? Its just a splitter dude like what

[grizzled grove]

It’s not

[white bloom]

well we gotta talk about something in a talk channel don't we

[grizzled grove]

45/2 is not 25 and 20

[unique cypress]

It is if one belt backs up

that's literally how manifolds work

[white bloom]

them: BuT YoU cAn WaIt FoR bAcKfIlL

[grizzled grove]

If manifolded

[white bloom]

(they're right of course but it's boring)

[grizzled grove]

If manifolded you don’t need balancer

[unique cypress]

1 splitter is a balancer. This isn't

[grizzled grove]

If sushi you’ll need it ig

Or if you over an overflow belt

[plucky tusk]

Honestly if you get stuck with weird splits its a planning issue coulda avoided it entirely

[white bloom]

but there is no problem with sorting out a weird split. Using otherwise for one's purposes inferior recipes just so the numbers work out more nicely is... well that does have opportunity cost

[unique cypress]

A splitter can do any split. There are no "weird splits"

[grizzled grove]

Splitters split

[white bloom]

... with backfill, yeah. If you don't want to bother implementing a specific split, which many people call a balancer, but KYO does not, then hook them up arbitrarily and make sure overflow doesn't escape the system and eventually it works

[unique cypress]

If this is a balancer, then manifolds are balancers. I'm happy with that, but I recon most people would disagree

[white bloom]

KYO just to avoid confusion in terminology what do you call a splitter-balancer network that achieves a specific input to output mapping without relying on outputs' backfill? Cuz we need a catchy term for that

[grizzled grove]

Tbh I only find use in balancers from resource nodes themselves

[white bloom]

I agree balancer doesn't feel entirely right when you think about it too much but that's just what kinda stuck with the community because it's the "manifolds vs balancers" dichotomy

[unique cypress]

I call these ratio splitters

[unique cypress]

This wouldn't be one because it uses a lower tier belt that's meant to be saturated, so it's a rate or precision splitter to me. I called them both

[white bloom]

ok. well it isn't just splitting though, sometimes there are more inputs then outputs. are these ratio mergers then? even though the difference doesn't matter conceptually

[grizzled grove]

[white bloom]

and if number of inputs = number of outputs you call them ratio/rate balancers then to make the confusion perfect? 😄

i like the ratio vs rate nuance for without/with belt saturation, gonna adopt that

maybe

[grizzled grove]

You also can rate limit with priority splitters

[unique cypress]

Any balancer has to meet these requirements:

1. every input connected to every output
2. equal input priority
3. equal output priority

Any good one also needs full belt throughput between inputs and outputs and groups of inputs and outputs

[white bloom]

yeah that's not community standard in Satisfactory

but I agree it would make sense

it's just a too rare use case

[unique cypress]

I've never seen anyone do ratios with anything else than 1 input and multiple outputs

[white bloom]

so the more frequent use case kind of took over the term within the Satisfactory community

[unique cypress]

It's in factorio ¯_(ツ)_/¯

[grizzled grove]

Have ether of you ever needed like a 4 belt to 3 belt balancer…

(Off topic question)

[unique cypress]

And Satisfactory balancers work indentically

So I hold them to the same rules and standards

[unique cypress]

Needed? Probably not. Used? Several

[white bloom]

you can build equivalent networks as factorio balancers, but they're not satisfactory balancers by what most people think of here

I don't have any stakes in what gets called what I just like consistency, this risks some confusion so it's just a warning

[grizzled grove]

1 very rarely do I have 4 belts or X that can all be on 1 belt. 2 you can just fully saturate overflow the belts, unless you want to balance to specific input in that case a 4 to 3 balancer isn’t the necessary end goal

[white bloom]

these do happen frequently, for sure

[amber umbra]

Ah man, balancer discussion.

[white bloom]

I mean it is kind of related. An example of a more-than-1 to many

the most elegant one I know splits one of them into 3 and merges one of each with one of the other inputs

4 machines, no feedback loops with bottlenecks, no belt saturation

pretty sure there is nothing smaller than that

[unique cypress]

That assumes all 4 inputs are equal and all 3 outputs are used equally

[white bloom]

yes

that's a default assumption

if he had uneven belt load he would have specified that

[unique cypress]

???

[grizzled grove]

Balance :3

[unique cypress]

I use balancers specifically because they don't care about item rates on the belts

Anything that has requirements is useless to me

[white bloom]

So yes, the use case for balancers in the narrow sense (your / the factorio sense) is when you have unknown / variable belt loads or supplies and want to ensure the resources get adaptively allocated. You pay for that by building redundancy into the system

I find adaptive networks and production lines intriguing, but whenever I actually play the game I just never get to a point where I think "here it would actually make sense to spend twice as much time to build everything robust / adaptively". The world is not procedurally generated, you can plan your whole production line statically

[thorn bane]

this comes up in 4/min rotors
here is how i build that

[unique cypress]

To me it's less about adaptability, and more about universality. Whenever I have to handle multiple full belts of one item in one factory, I can just paste a balancer from a blueprint and not care about how many items are produced/needed on each individual belt. It just works.

[thorn bane]

can build this easily by dragging a line between the 4 constructers and then attaching the splitters/mergers on to the line

[thorn bane]

can also just paste a belt compressor and its works the same while being smaller

[white bloom]

hmm. At the end of the day, you still need to ensure that the total production matches total consumption for each item, otherwise you've built some expensive production buildings that don't get to fully run because they're starved for inputs or stalled for outputs. And if you do have matching totals, there is no need for balancers in the strict sense anymore, any network topology (barring unintended internal belt bottlenecks) will work. Or as the Satisfactory kids like to call it: "just manifold it"

[thorn bane]

do me its the expandability
while im going through the tiers, i change production/consumption (even recipes) a lot so it pays to have those distribute evenly even when i change a small thing
and adding this #math-and-meta message REALLY doesnt take a long time

[unique cypress]

Without doing some insane spaghetti, they're 8*max(n,m) m long

[thorn bane]

except when you feed them from storage
like build up 500 modular frames
use them to instantly have HMFs as soon as you unluck them

[white bloom]

expandability is the huge advantage of manifolds in the stricter sense, with translationally symmetric layout

[unique cypress]

(barring unintended internal belt bottlenecks)

That's exactly what balancers avoid

I cannot manifold 12 belts, because 12 full belts don't fit on one belt

[white bloom]

you don't need to build a complete bipartite graph every time to avoid bottlenecks though, typically that's going to be overkill by a huge factor

[thorn bane]

ye
ofcourse you manifold
the question is how you feed those manifolds

[unique cypress]

I manifold. But only up to one belt. Because more items don't fit on one belt. Then I put a balancer between the manifolds

[thorn bane]

ye i manifold and use belt compressors so the manifolds are always full

here is what i mean
https://docs.google.com/document/d/1_rXrJ8CWyjNExcayArby3HjRpDUXV5sd8QvFrdNZxBc/edit?pli=1&tab=t.0#heading=h.i4mh5jaa6ir1
(i should really update this to priority mergers, it makes the BC 1 splitter and 1 merger instead of 2)

[white bloom]

ok, suppose for simplicity I have N incoming belts and N outgoing belts with equal load (imagine each of these gathers or feeds a manifold for example). There is no point in building a complete bipartite graph / strict balancer compared to just matching the belts one-to-one, unless the belt loads are unknown/variable.

the compressors are a decent way to achieve this in that it doesn't build as much belts as a complete / strict balancer, but it's also relatively hassle-free compared to custom fitting a connective network and being wary of local bottlenecks everywhere

[unique cypress]

How do you intend to match 4x420 -> 150, 300, 300, 450, 480 one to one?

[thorn bane]

you dont build 420 in the first place

[oblique hollow]

there comes a point where you must ask "what the heck am i doing this even for"
Its either out of boredom or because you wouldnt know a better way to do it

4 x 420 isnt a game-given goal or limitation.
You were the one who put that limit up yourself

[thorn bane]

ye imo its just way simpler through the game to have (somewhat) unknown amounts on your belts
like image i make 270/steel and use that to make 270/min worth of pipes and 270/min worth of beams
ofcourse i dont have the resources for that, id need 540
but it works as long as i dont consume 100% of both
now the ratio of pipes to beams might change depending on my factory
i might switch to encased industrial pipes
i might use steel screws
i might use molded pipe
but all of that just works if i just split it, so i dont have to worry about doing the math everytime i unlock a recipe or get an alt

[thorn bane]

well the reason to do 420 is "probably" that its easier to build that way
and if a balancer (which ever you choose) is just a bluleprint click away, then it becomes overall easier to just build 420 and balance

the math alone to figure out which lane needs how much takes longer than pasting a blueprint

[unique cypress]

Exactly

I don't look at the item counts at all and once I balanced 4x360 to 4x360

because I just saw 4 belts in 4 belts out and pasted a blueprint

[summer flare]

Yeah, I got out the popcorn to see where it went.

[white bloom]

Hmm, you have a point. If the build costs aren't prohibitive, just slapping a strict balancer in there from a blueprint - even though it may have quite some redundant edges - is even more convenient than using belt compressors or custom fitting a sufficient network.

[thorn bane]

you paste a belt compressor and it does exactly the same

my issue with blalancers is that they are big so i would need extra space to paste it on to my bus

[unique cypress]

They aren't suitable for depart when empty/full trains

[white bloom]

only when the number of endpoints (not their load) to be connected gets large will this become infeasible, because the complete bipartite graph m,n has something of order O(m * n) edges, when just to accomodate throughput you only need O(max(m, n))

yet, we typically aren't connecting that large number of belts that asymptotic scaling would really matter

so for small numbers, complete bipartite connections (your strict balancers) are probably the most convenient indeed

[thorn bane]

me balancing my 7 lanes for quickwire xd

[unique cypress]

7? Rookie numbers lol. I have 20

[thorn bane]

i mean even for 2-2 its 2 instead of 4 units so twice as big

[prisma kraken]

sometimes i just need to shake my head at things

[restive sparrow]

I'm over here just making what I need like an IDIOT

[unique cypress]

And this isn't even the biggest balaner I've built lol

[thorn bane]

cause you havent tried it
once you go adaptive you never go back

[unique cypress]

Though it is the biggest I've actually used (so far)

[white bloom]

yes, it's technically overkill at any scale, the point I credit him is that you could just slap out that blueprint for two extra belts and don't even have to worry about compressing the right one to the left or vice versa or whatever, it just works without a second thought if you're willing to pay the cost in build materials + wasted space

[prisma kraken]

i have tried it and found that direct belting eliminates those headaches

[thorn bane]

its not a headache its a problem
we have found the solution
now the problem is trivial (select blueprint, paste)

[prisma kraken]

most problems you invent have trivial solutions (i.e. don't do it)

[oblique hollow]

most problems are trivially invented

[thorn bane]

i mean if you have a super gadget does it all blueprint
why use the simple solution when you can use a sledgehammer to crack a nut

[thorn bane]

except the "dont do it" part involves a lot of math, planning and foresight

[white bloom]

it remains a problem of scale though. If you were to connect 1000 inputs to 1000 outputs that's 1 million belts with a strict balancer even if you could connect 1000 belts to one and the same splitter/merger. While accounting just for the un-evenness at hand you'll custom fit something that's on the order of not much more than 1000 belts to achieve that connection. Maybe 2000, depends on the loads, point is something on that scale.
Of course you'd never need that, just to illustrate the point. It inevitably becomes unfeasible at some size to go with connectivity in a higher dimension than required throughput

you've convinced me I'm kinda digging this idea for small handy numbers now though, the universality is enticing

cool stuff. Thank you both for entertaining this discussion with me. I gotta go, see you around!

[thorn bane]

except belt compressors (god i hate that word) are O(n) instead of O(n^2)
thats why everyone uses them in factorio

[white bloom]

I know. The O(n^2) was referring to the strict balancer (complete bipartite graph)

[unique cypress]

These are the stats for a 100:100 balancer

[white bloom]

using belt compressors as needed is roughly the same as custom fitting the connection, you have to watch and follow along keeping track of the flow rates and therefore end up with something O(n)

[unique cypress]

(Belt count is slightly inflated due to length limits)

[white bloom]

that should be more than 10,000 by my understanding of what you explained a balancer to be by your definition. Maybe I misunderstood you

anyways I'll respond later, gotta go

[wind zinc]

end of the line
i never thought i'd run out of space here lol

just a few machines

[unique cypress]

FYI, this summary is inaccurate

it just adds the building counts together

[wind zinc]

+this

[unique cypress]

completely ingnoring that 1.5 constructors for 1 thing and 2.5 for another is not 4 constructors total but 5

[wind zinc]

eh, close enough

[wind zinc]

those are for betteries / drones

[unique cypress]

Just another reason why that calculator is garbage

[wind zinc]

what does FYI mean ?

[unique cypress]

"for your information"

[wind zinc]

Ahh

Had no idea

[oblique hollow]

it just tells you the effective "percentage summary" needed

you can clock machines however you want and that then changes the Actual total building count

1.5 for 1 thing and 2.5 for another adds up to 400% effective constructor work force

how many machines you actually build for this is up to you

[wind zinc]

I didn't OC much

[oblique hollow]

not just OC; but also UC

[wind zinc]

But i did later go in and change some recipes

[unique cypress]

How is this information useful when it's added for multiple different items

[oblique hollow]

its not useful to summarize it at all imo.

[wind zinc]

Did UC quite a bit

[oblique hollow]

but it is useful to give you a rough estimate for how many machines it likely needs

if you start messing around with clocking, it starts deviating anyway

So that number is never ever 100% useful or accurate

just a guideline essentially.

[unique cypress]

I never OC/UC anything other than miners

[oblique hollow]

well, thats a you thing and i dont care about that

[unique cypress]

So when it's rounded up for each recipe individually and then added, it's 100% accurate

[oblique hollow]

neither might others care about it.
for them, this might be more useful

[wind zinc]

Like, i change the wire recipe to caterium wire (Plz don't judge me), i had a ton of caterium i wasn't gonna use
And that allowed me to save some space

[unique cypress]

Or even if you UC uniformly, but don't go so low that you'd need extra machines vs 100%, it's also 100% accurate if it's rounded before adding

Adding without rounding makes it inaccurate and useless for everyone

[oblique hollow]

Eh, who knows, maybe the next big update to tools that is supposed to change a lot of things changes this too

[unique cypress]

Tools does it correctly

[oblique hollow]

Not like i personally ever use machine summaries anyway

[unique cypress]

It's only the SCIM planner that fucks it up (as usual)

[oblique hollow]

i honestly didnt realize this was SCIM lol

[wind zinc]

What IS with you guys and constantly sh*ting on SCIM just cuz you see it ??????

[unique cypress]

Because it's gubage

[wind zinc]

BRUH

[oblique hollow]

meh, i dont care much bout scim.
useful as long as it gets stuff right.
once it doesnt, its time to switch

On a completely unrelated note:
new transmutation table available

[wind spade]

SCIM doesn't really get things "wrong", but given it can't do like half of the mid-lategame builds...

[unique cypress]

I'm reasonably sure the total power is slightly incorrect lol

[wind zinc]

It Does what I want it to do 🤷‍♂️ and i didn't even know that there were others until a few days ago TBH lol

[unique cypress]

I only bothered to check one example but couldn't get the result it did

[unique cypress]

That's only until you want to be oil efficient

It literally can't calculate the most efficient recipe chain for rubber/plastic

and the second best takes like double the oil lol

[wind zinc]

I know it cant do recycled rubber/plastic it doesn't bother me TBH
And i don't mind doing a bit of math from time to time

But yeah, that's a fair point

[wind zinc]

Nice
Im kinda mad that you cant convert nitrogen to anything

It would be quite useful

IDK, like turn it into uranium or quartz

[unique cypress]

Unfortunately the converter doesn't have a fluid input

[wind zinc]

Well, useful is a strong word
More like, fun.

[wind zinc]

Really ? I though it did lol

[unique cypress]

Only fluid output

[wind zinc]

do you guys think it'd be worth making a dissolved silica / leached quartz crystal factory ?

[unique cypress]

If you can use everything you produce, both crystal and silica, yes

If you're gonna end up sinking a significant portion of either, no

[wind zinc]

Aight

Im gonna run through the numbers tomorrow

[thorn bane]

it reduces amount of buildings so yes

[wind zinc]

I just found these recipes really interesting when i first found them
So i thought i'd ask

and how much quartz crystals/silica does T9 take ?

[dusky dust]

Depends what kind of recipes you end up using

[wind zinc]

Fair enough

[wind zinc]

It would actually let use the foundry recipe for aluminum

Which would be nice cuz that recipe has a better rate

Would i need it though ? 🤔

[unique cypress]

Calculators generally do not use default alu ingot unless forced

It's "expensive"

I wouldn't set up silica specifically for it, unless I really needed that little bit extra bauxite efficiency

But if you need crystals and have distilled silica to spare, it's certainly cheaper

[thorn bane]

you trade 60 quartz for 28 bauxite
kinda meh imo

[wind zinc]

I thought it'd be nice later on for trigons and nuclear
Cuz i think I've used up most of the maps bauxite
(I don't wanna use cat/iron for the ficsite stuff)

[thorn bane]

wdym you used up the bauxite

[unique cypress]

Yeah what?

You either have to really waste it or build a factory so large it lags most PCs

[thorn bane]

i didnt use any
my uses were
insulated crystal osc.
silicon HSC
silicon CB

[wind zinc]

I used up most not all

[thorn bane]

for what?

[wind zinc]

There still a bunch in the swamp

[wind zinc]

This #math-and-meta message
Im still building more stuff that'll use bauxite, im not done yet

[opaque quartz]

I’ve never done the quartz purification chain but I’ve been intrigued by it

[wind zinc]

Well, actually the bauxite refinement place is somewhere else
But that's where the Alu goes

[unique cypress]

Bro that is 10% of the map's bauxite

[wind zinc]

"im not done yet"

And im making like 3500 Alu ingots
Theres no way that's only 10%

[unique cypress]

https://www.satisfactorytools.com/1.0/production?share=eBD98Uw1V4C5FMQhFNiN

[wind zinc]

Whats that for ?

[thorn bane]

arent those the items youre making?

[unique cypress]

The items you listed, but in tools

[wind zinc]

A : why ?
B : im not done using the aluminum yet

[unique cypress]

Well you said that this is using most of the map's bauxite

But even if you're making 3.5k alu ingots, that's still 30%

[wind zinc]

I've collected every node in the red forest like WAT

and some in the titan forest

Well 1200bauxite from the titan forest

4 nodes IIRC

[wind zinc]

Oh wait
Right, cuz that's 30% assuming you have MK6 belts and MK 3 miners

Which i do not in fact have

[thorn bane]

ah

[wind zinc]

Well i technically have MK 3 miners now but i didn't have those when i set the Alu refinement up

[plucky tusk]

I think i just did 300 bauxite to get started on p4 but im def gonna need more

Its like 270 ingots split between my 2 factories

[thorn bane]

p5 kinda inhales alu as well
i was up to 4k at the end

[wind zinc]

And i am using lot's of recipes that use aluminum and aren't included here with tools cuz they're "inefficient"

They've saved lots of space and headaches for me

[thorn bane]

wait which ones?
arent the resource efficient one the space efficient ones?

[wind zinc]

Cuz i don't have to make anywhere as much iron and steel as i otherwise would have

[unique cypress]

Using aluminum instead of iron sounds like the opposite of space efficient, easy or convenient

[thorn bane]

alclad casing uses an assembler, but its also 2x the output
heat fused frame makes 3/min instead of 1.5/min and uses less alu
heat exchanger makes 10/min instead of 7.5 and uses less alu
actually not the case for RCUs, radio connection uses less than radio control system but its 3.75 vs 4.5

[harsh schooner]

Cheese

[wind zinc]

Aluminum rods and beams.
Plus the tools site also uses the regular screw recipe,steel rods,a different recipe for RCU,iron pipe, encased pipes,caterium CB,iron wire lol,caterium computer and etc
Point is its worlds apart from what i built

[thorn bane]

ah ye right alu rods and beams are a thing, i kinda forgot about them

[wind zinc]

I don't think i had heat fused frame before i started building it

[unique cypress]

Except you're saving resources, so you need less of everything upstream

[thorn bane]

well it depends, sometimes you need other stuff

[wind zinc]

I even if you need 110 more constructors ?

[wind zinc]

More often than not the more rescorce efficient recipes take up more space than *some of the less efficient ones in my experience (repost but this time without autocorrect being annoying)

(That came before KYO's comment )

[unique cypress]

If I disable all alts on that plan, the building count literally doubles

Tools optimizes for resource efficiency

[thorn bane]

pure recipes absolutely
advanced part recipes? i feel like its the opposite

[unique cypress]

So I'd say the more resource efficient recipes are on average more space efficient as well

[wind zinc]

It depends

[wind zinc]

🤦‍♂️

Of course !

What did you expect ???

[unique cypress]

Well, you seem to have expected the opposite

After all the more resource efficient recipes take up more space, right?

So if I made it less resource efficient, it should be more space efficient, no?

[wind zinc]

OMFG okay im gonna dip cuz its 11:40pm and i rather not couch my braincells by someone purposefully misunderstanding what i said

[thorn bane]

the thing is the further down the production chain you go the more resource efficiency matters as it has a ripling effect of needing more of the previous items as well
so as a trend the more complex the more important item efficiency is over output numbers

[thorn bane]

is it possible to have a solid train unload directly into machine instead of an ISC and let the machines be the buffer?
how would one go about that exactly?

wouldnt it work to just split the machines in 2 and feed half with 1 belt coming from the station and the other half with the second?
the main issue is that you cant just merge both belts since its throughput is sometimes high sometimes 0 but just using 2 belts should work right?

[unique cypress]

If I'm not wrong, the minimum buffer you need is belt speed * 27/60 items

Maybe double that?

[thorn bane]

well the machines should have that

hm actually if you do it as a manifold then the machines would already be full so you dont have the buffer
i guess if you balance it it works

[unique cypress]

If machines are full, it's still a buffer. They'll empty during the pause and refill after it ends

[thorn bane]

oh true
i guess you also have some amount of belt buffer

[unique cypress]

I think to make sure it works consistently, you only need to make sure no machine runs out of input if it doesn't receive anything for 27 seconds

Though idk, some belts shenanigans might make that delay longer?

Sounds like something that could happen

Yeah, if you manifold, the first machine will get refilled after 27 seconds but the latter ones will take longer

But they'll lose supply pretty much at the same time, with only a little bit on the belts

So the latter ones will take longer to refill than 27 seconds

They'll get some items after 27, but not enough to start filling, only slow the drain

[regal pier]

Can I ask about your username tag line? I just set up 120 fuel gens using mk2 pipes

[thorn bane]

mk2 pipes in long manifolds dont do 600/min, more like ~550
this is caused by junction splitting unevenly like 600->570/30 creating backflow that reduces the overall throughput
splitting evenly like 300/300 is fine
you can fix this by splitting your pipe into 2x 300/min pipes at the manifold, for example by building a loops that feed 300/min from the start, 300/min from the end or by connecting it in the middle so it splits 2 ways

[regal pier]

That makes sense and seems like an easy adjustment to existing systems

[thorn bane]

if your setup is not 1 long line but multiple lines that your first split into from 1 pipe then its fine

[regal pier]

Yeah mine feeds from one pipe into a 2 way split then makes a loop all the way around. I also sent you my save file. I've been cleaning up the platforms since I sent it though

[unique cypress]

Incorrect. I have several pipelines happily doing exactly 600/min

[thorn bane]

are they in long manifolds

[unique cypress]

Depends what you consider long

I have tested a 20 refinery long one once and it worked fine

I don't remember if I have one that long in actual use, but I have done tests and it worked

[thorn bane]

did you build it as a loop?

[unique cypress]

No

Just a straight pipe with junctions

[thorn bane]

hm interesting, do you have a screenshot?
600->570/30 (600/20 = 30) should cause problems
how did you feed it?
from the start or from the middle?
also from above, ground level or from below?

always happy to look at more test cases, but that setup (20 refineries) didnt work when i tested it

[unique cypress]

I don't have screenshots (anymore? Idk if I took any). Can try re-doing it in like 16 hours and @ you if you want

[thorn bane]

sure

[unique cypress]

The most recent test I did was with a line of plastic or rubber refineries, fed from a single, fully overclocked oil extractor connected to the end of the manifold

The extractor was below the refineries

But I don't remember where the manifold was

Either on the same level as the refinery inputs or above

Definitely not below, I never do that

The pipes were not pre-filled

I also remember having 50 fuel gens running off a single 600/min pipe (pre 1.0's consumption change from 12/min to 20/min).

[regal pier]

I find that keeping the pipes feeding downward works well in long lines as long as I can feed the machines at a rate like n+10%

[fierce ruin]

What happened to beacons? It's been about a year and a half since I last played

I can't find the beacons

[quick gorge]

They were removed in 1.0 among some other things that became redundant

[fierce ruin]

Redundant?

Is there another way to mark things now?

[quick gorge]

Left Alt + Middle Mouse Button

[vapid gorge]

poke the map

[quick gorge]

I legit had to google it because I never mark anything on my map (I really should...(I won't))

[thorn bane]

now thats a name i havent heared in a long time
PTSD of CRYSTAL BEACON IS BETTER THAN DEFAULT BEACON

[thorn bane]

i mark crash pods and highlight them
nice to find where it is

[quick gorge]

I just... remember where they all are because I know the map like the back of my hand
(I am quite aware I have no life thank you)

[thorn bane]

aint no way you have all 100 memorized
like if im close then i know the area but all????´
someone needs to make the "name all states" but for crash sites

[quick gorge]

Most of the crash sites's requirements are often close in proximity so I don't bother remember their requirements.
Saying that that black powder one with the arch above it always gets me tbh

[thorn bane]

haha ye

actually remembered to put some sulfur and coal into dimensional storage this time
quite proud of myself

[quick gorge]

I have several laps of the map for spheres, sloops and drives in my head, the starter lap to get the basic shite.
Once I get nobes/blades (now I can get those from the cave and get on the first lap)
and when I get hazmat to clean up the rest

[quick gorge]

Actually peak gameplay

[quick gorge]

If only they had names I would... No not BP_DropPod33,BP_DropPod_C_8

This is maths and meta after all, I do someshit around this

It is a tad harder to do it from the map view since I'm use to pathing it in first person but ECH close ebloodynough

[thorn trail]

more like maps and meta

[quick gorge]

I will not draw up my first real life, I would need record myself doing it in game and then apply it to the map and I have shit to do

Like dig through the game files to attempt to find spooky lore words to put into a thing I'm messing with

[thorn trail]

do you do this first lap before or after getting blade runners? Just curious because I always start in the rocky desert and the first leg of my journey is immediately to the doggo cave to grab the blade runners there then looping around to grab a 'first lap' of starting stuff from sites

[thorn bane]

i should start doing doggo cave but its so far away from northern forest...

[quick gorge]

I unlock foundations first, than ramp up into the Doggo cave and pick up the runners/basher

[thorn trail]

you could always start in the rocky desert, grab the boots then quickly head for the forest picking up stuff along the way

[quick gorge]

Like if I'm playing a server I start the lap but head toward the cave to pick them up, if they're in my way might as well pick them up but you could count those as false starts I guess

I could autistically list everything I would depending on starting on location... but do we really want that?

[thorn bane]

have you seen the conversations in this place?

[quick gorge]

(We in this case being the royal we, aka: me)

Whenever SF+ is stable™️ for a dedicated server in 2029 I have a group of people who want to do a SF+ run.

Needless to say, I am the scout

Much love McGalleon

[vapid gorge]

@nova vortex

line of rail, zooped along the way, then build pylons at intervals to keep it tidy

[nova vortex]

I dont have those things yet…

[vapid gorge]

this is only foundations and rail

plus some pillar feet, but that's just for looks

[nova vortex]

oh i just build high up and went straight

[vapid gorge]

that doesn't negate teh fact that this is a way to keep your rails tidy and properly spaced with little effort

[nova vortex]

it is tidy

[vapid gorge]

you sounded like making 2 lines was an extreme effort somehow? so I was giving you a quick way to do things

[nova vortex]

oh no mb i didnt mean to make it sound like that im sorry.

[whole bane]

Sorry I am brand new kind of just starting but would I need to make a 3 to 5 balancer for this, or would that not be necessary because it's only 45 units/minute?

[prisma kraken]

use a manifold instead, when machines in the line are oversupplied, they only consume from the line what is needed, and eventually the line overflows dto the rest of the machines. All that being said, the rotor problem in the early game has some other clever solutions

one small hint to start you down the correct path on that is that an intro rotor build is easier to do at 2x the size

[whole bane]

I will have to google the rotor thing but manifold makes sense, and thats because the lines will backup because the 5 constructors will be off by 5 units? IF I did have 50 units a minute would you suggest a balancer?

[prisma kraken]

yeah, eventually everything ends up going where it needs to get to

another thing that is probably the easiest solution to your problem is hunt down some hard drives to get this recipe

[whole bane]

oh will do, forgot about that when watching an intro video!

[prisma kraken]

i'm trying not to just solve the problem for you 🙂

[whole bane]

I appreciate it.

[prisma kraken]

another way to approach it all is to realize that a 3-way splitter can take 15/min and split it into 5+(5+5)

[crimson moat]

Another thing that you can do is just split each rod machine into two for screws and accept that they're not running all of the time (or with a little tech, underclock them)

1:1.5 ratio is awkward, rounding to 1:2 is less so

[unique cypress]

Balancers aren't particularly useful if the total amount of items can fit on one belt

They only shine when you've got several belts

And they're even more useful if the belts each carry a different amount of items

Also, obligatory SCIM Production Planner slander: it's garbage

Well, it's not like it's unusable, but it's by far the worst calculator I've ever had the "pleasure" of using, so if you're new, I suggest you find something else before you get used to SCIM

I recommend
Satisfactory Tools:
satisfactorytools.com/1.0/production

FactorioLab:
factoriolab.github.io/sfy

Satisfactory Logistics: satisfactory-logistics.xyz

Satisfactory Modeler
(available to download on Steam)

Try all of them and find one you like the most

[thorn bane]

yo whats up with satisfactory-logistics.xyz Minimize Area?

[unique cypress]

That its typical behaviour in that mode

Though I have to say, using five different recipes for one item is new to me

Generally, if you're trying to minimise area, whole numbers of buildings are more space efficient than fractional ones

[thorn bane]

why the fuck is this happening
and does this mean that mode is just useless?

[thorn bane]

but its worse than default steel ingot
and if i increase the amount of steel then it only changes the amount of default steel

also why are there 5 iron ingot recipes?????
this has to be a bug right

[unique cypress]

The only solution I have is to look at which recipe it wants to use the most and disable everything except that one

[unique cypress]

Idk how they wrote their optimization function. I imagine it's not that easy when anything between ɛ and 1 adds exactly 1 building's worth of area

This doesn't sound linear

[thorn bane]

how long does a manifold need to fill after 27s of downtime?
lets say im shipping 1650/min steel ingots to turn into 1100 steel pipes so 55 machines

it would be (27.08/60 * 1650)/(2400-1650) = ~1 minute right?
so we need to make sure that the roundtrip time is longer that 1min which checks out since it takes 1min20 to fill a train of steel ingots
and we need to make sure that the machines dont run out which is fine since steel ingots stack to 100 and the recipe consumes 30/min (empty after 3min20)
soooo you dont need a buffer for this?

[fallow siren]

no need buffer, just hand fed to warm the manifold line

[unique cypress]

Sounds about right, I think

Manifold fill time is normally more complicated because the consumption changes based on the number of machines currently with items, but here all should have items in them and the consumption should be constant 1650/min

At first they're losing 1650/min for 27 seconds, then gaining belt speed - 1650 until full

If your max belt speed is 2400 then yeah the whole process takes about 1.5 minutes

And emptying a full wagon at full belt speed takes 1.33 minutes so you won't run out of items for the reload even if you only run one wagon

[wind zinc]

That's the " make it as annoying as it possibly can be" mode

[wind zinc]

i guess distilled silica would use a lot less space than the other options

[vapid gorge]

that's interesting

[fierce ruin]

Perhaps not math related, but I blew this guy up. Is he dead for good, or is he gonna get back up and start spewin' his fart cloud again?

[grizzled grove]

More quartz per quartz

[fierce ruin]

Ohp, the corpse is gone now.

Guess he's not fartin' no more

[wind zinc]

another one will spawn

[dusky dust]

No, those shouldn't respawn unless the game bugs out (which has been known to happen, of course)

Once they've been bombed, they're supposed to be gone

[wind zinc]

oh

[quick gorge]

They are not meant to respawn

However: They do not care for the rules and do so anyway

[fierce ruin]

Do the fart pillars respawn?

[opaque quartz]

No

[wind zinc]

those really should not

[quick gorge]

I've tad them respawn after updates.. as I said; they do not care for rules

[wind zinc]

TBF there was also an update in which a bunch of somesloops and mercer spheres respanwed

so, anything is possible

[quick gorge]

They do not care for rules :3

[cerulean stratus]

Ok I just wanna show you guys something
a 7 segment display, using belt logic

[quick gorge]

What black magic is this?

[restive sparrow]

Lol I love it

[quick gorge]

Can you make a calculator?
Can you be the true math within https://discord.com/channels/370472939054956546/558721941410807812

[cerulean stratus]

actually yes

[quick gorge]

My god

[restive sparrow]

You could improve this by controlling with power switches by running your materials by sinks you're switching on and off

(smart splitters overflow)

[gloomy bridge]

If people can make a computer in minecraft, it can be done in satisfactory. It just takes an electrical/math engineer to do it.

[restive sparrow]

If you wanted to be a complete lunatic you could extend the logic to have a single switch create a number for example.

[cerulean stratus]

actually we don't use sinks that much. Hopefully we'll get modding for 1.1 soon, because that's where i'm using the priority merger

[restive sparrow]

Possible, but would be very difficult in satisfactory cause there isn't a lot you can use for logical switching

[cerulean stratus]

yeah it's pretty difficult in satisfactory

NOT gates take a ton of space, because each wants a storage container

[vapid gorge]

I can't recall if you can actually make logic gates in SF so it may not be possible

[cerulean stratus]

https://www.reddit.com/r/SatisfactoryGame/comments/1judnyr/belt_logic_now_with_priority_mergers/

yes, you can do logic gates

[gloomy bridge]

Def can do logic gates.

I use logic gates in my current builds. Smart Splitters and Programmable Splitters make the best logic gates for simple things.

[cerulean stratus]

The priority merger is a game changer for Not gates

[gloomy bridge]

Yes, which makes things even more spicy in the programming realm.

[upbeat summit]

uhh help, explored and scanned a bit too many hard drives, can't select the alt recipe

[vapid gorge]

change the UI size

[upbeat summit]

where is it?

[vapid gorge]

settings

[upbeat summit]

where?

[vapid gorge]

probalby scaling. Scale can refer to size

[upbeat summit]

ohh got it, i'm dumb

ty

also is there any list which shows which recipes are the best (kind of like a tier list?

[vapid gorge]

all tier lists are poorly made

[unique cypress]

There is. I suggest you don't use it

[vapid gorge]

each recipe has it's own situations where they can benefit you is the short of it.

some recipes tend to be more general and thus more occasions to be used - these are usually lower tier recipes

[upbeat summit]

then how should i select alt recepies?

[vapid gorge]

are either of the recipes part of a factory you'll be building soon?
if yes, pick it

if no, rescan or pick at random

you're never locked out of recipes. You can always get all of them

[unique cypress]

Use a calculator (or do the math yourself) and see which ones work best for you

[faint locust]

100% true! the only thing to consider is what you have nearby and how much power does it cost.

[upbeat summit]

alr

[vapid gorge]

eh, power costs are pretty trivial. You can easily build more power.

it's really how they suit your style, what you want to do, and where

[faint locust]

true!

I wasn't too thrilled with water geyser

500mw for something I could do in 200mw for the same output

[vapid gorge]

but if you want recipes that will give you more opportunities to use them? lower tier items. Wire, screws, plates, Alloy ingots. Things like that. They are versatile because they apply the affect at the start.

[faint locust]

yeah, having a way to build basic components without a long belt or trains, helps alot.

iron wire, does suck, but it is wire.

[vapid gorge]

why? it basically costs the same amount of iron as it does copper. And iron is more common

once I get iron wire I rarely make it from something else

[faint locust]

I get that, I never had the shortage of wire from copper to supplement a section to force myself into iron wire. I did it a few times, but copper is just easier from the foundry

[vapid gorge]

it's not even about saving copper. It means you can start making wire and cable things in spots w/o any copper

[faint locust]

true

[vapid gorge]

but sometimes you do need the copper for other things, like sheets. Helps spread the burden around

[faint locust]

you mean the million billion uses for sheets

[cerulean stratus]

eh, wherever there's wire there's copper nearby

[amber umbra]

When in doubt, avoid saying “xyz alt is bad” here unless you want to bait replies.

[vapid gorge]

basically xD
I hate steamed sheets but damn is it useful

[unique cypress]

There's 2.5 times more iron than copper so it's easier to find

[faint locust]

I don't think every alt is bad, OTHER THAN a few of the recipes.

I'll be honest, I don't want to make them, that overcomplicates something that was simple

[upbeat summit]

yeah selected many of the alt recepies, i was wondering since i'm at phase 3 and obv don't have things like rocket fuel, and aluminium unlocked, what would happen is i try to get all the hard disks, will the game give me alt recepies which i haven't unlocked yet?

[opaque quartz]

One of the things to keep in mind about alt recipes is that some recipes have good synergy with other alt recipes when considered together. This is one of the main reasons why tier lists are bad

[vapid gorge]

there are some recipes that are very niche.

quick wire stator.
mainly used for saving on steel.
which isn't commonly a limiter

[vapid gorge]

there's more hard drives than recipes, so yes

[faint locust]

man, quick wire is amazing and also the hardest thing to get right.

since its not equal in slugging it

[upbeat summit]

damm that's interesting

[opaque quartz]

You can only unlock recipes for milestones you have unlocked. It won’t let you scan more drives if there are no more available at your current phase

[upbeat summit]

alr

[vapid gorge]

slugging?

[faint locust]

power slug

[opaque quartz]

The wiki has the milestone prerequisites listed for each recipe too if you need to see it

[unique cypress]

And they're wrong sometimes lol

[vapid gorge]

ah, not sure why slugging it is hard though?

[faint locust]

its 2 to 3 in ratio.

[opaque quartz]

Which ones are wrong? Submit edits to the wiki if so

[faint locust]

hard 2 to 3, you can balance it I suppose

I wish I had that alt that made more bars and check it out.

[unique cypress]

The ones that have 2 requirements listed as one or the other. I checked the code for 2 of them and they're actually AND, not OR

Though I still have to check in-game whether I'm understanding the code right

[faint locust]

wait, no. there was no alt that could just beat the basic if you considered power usage

[opaque quartz]

Interesting. Got an example?

[vapid gorge]

the top recipe?

[faint locust]

I was thinking the ingot itself

[faint locust]

yeah, top is 2 from smelter and 3 in the constructor as max.

[vapid gorge]

yeah I've never had issues using power slugs on machines for these. Not sure what you mean

[faint locust]

I was thinking there was an easier filler for the ingot, but nope.

sulfuric acid is probably the way to go, but that is just asking alot. haha

[vapid gorge]

most sulfur recipes are a bit awkward imo. personally can't be bothered with them. Tempered is nice though. much less piping and fewer refineries

[unique cypress]

silicon circuit board was the one I checked recently. On the wiki, it's listed as either oil processing or quartz (the research), but the code makes it sound like it needs both

I checked it because someone complained that having quartz didn't make it available to be rolled from a drive

[vapid gorge]

anyway sleep time 🙂 gn all!

[faint locust]

silicon board? that one is super duper worth it

[faint locust]

later man! have a good night

oh, well to roll into the rng, you have to have one of the tiers unlocked, BUT that doesn't mean you will see it early

I saw sloppy at the end of phase 4.

[wind zinc]

i finally finished the assembly director systems

6/min though thats with supercomputers and adaptive control units slooped
and then after that i also slooped these
so im actually getting 12/min 😎

[viral sparrow]

just a warning that this will be able to make 48 of the next part (without sloops)

[wind zinc]

wait, what

that sounds insane

[viral sparrow]

yeah the next one is 1:4

[wind zinc]

perhaps i overbuilt : )

but that'll be very fun then

[ruby shoal]

Is this kind of backup an issue? 4 foundaries are working beneath this floor to produce 180 steel ingots/minutes (45x4). It's hard to see in a picture, but these three constructors are all accepting 60, so it balances with what it's being fed (60x3=180).

I understand that the problem is probably the belts and conveyors, which are mk3 and transporting 270 items per minute. This was done to prevent the machine from starving, but it ended up overflowing the entrances. Is this an issue?

This kind of problem is occurring everywhere, but from what I understand, it really shouldn't matter what kind of belts I use, because the machines will always be fed and never idle, which means they're at maximum efficiency, right?

[oblique hollow]

if the constructors use 180 total and you make 180 total its no issue usually

[opaque quartz]

This constructor has a yellow status light which means it’s not running 100%. Share a screenshot of its console

[ruby shoal]

It's green right now

[opaque quartz]

Was the output backed up earlier? Do you have a high enough tier belt to handle carrying away all the output from the machines?

[sullen barn]

What planner do u guys recommend? I dont really like the one from calculator cuz u cant entangle the productions properly especially while using alternative recipes

[cerulean stratus]

satisfactory modeler

[cerulean stratus]

It's one of the best rn

[sullen barn]

That a site or an app?

[cerulean stratus]

app, on steam

[sullen barn]

Ty

[dusky dust]

https://www.satisfactorytools.com/1.0/production is, IMO, the best solver out there. As with any solver I would recommend keeping graphs separate -- don't entangle your production in general and you'll get much nicer, cleaner graphs. :)

Modeller's not really a solver per se, and its graphs are absolute dogs#$@ at conveying information to other users (there's no labels on anything, so you're stuck interpreting bunches of tiny little icons instead), though folks who use it to plan factories seem to like it

[unique cypress]

Yeah, I don't like modeler either. Way too manual. I use Tools for all my calculation and planning needs

[dusky dust]

It's really two different use cases entirely; I can see the appeal of planning stuff out with Modeller, even if I don't use it myself

Solvers + planners occupy different problem spaces

[sullen barn]

How? I cant properly entagle alternative recipes with this one or can i?

[sullen barn]

I only want something where i can somewhat visualize my production line with input, machines required and output. I did that on paper before but its annoying to have it like that for bigger projects

[dusky dust]

Well, as I say, my favorite tool is the sftools site listed above. Give it a try; there are plenty others if you don't like it. You can choose recipes + available resources + available machines and all sorts of stuff on the tabs there

(I actually mostly use my own homegrown tools when actually building, but I'll often use sftools as a spot-check and to more easily try out different recipe chains, since my own tools are pretty clunky for that)

[sullen barn]

Can i link multiple productions? It seems to use the quickest route to ge a certain item which isnt the most efficient

[dusky dust]

You can specify multiple outputs, if that's what you mean

But no matter what tool you use, you're gonna end up with unweildy graphs if you add too much in

As I say, I recommend keeping production lines separate no matter what tool you use. Satisfactory production chains are inevitably rather complex, eventually

[plucky tusk]

🧠

[sullen barn]

I just want something so i can manually select the input, recipe and sth to put that product into another factory with a different recipe. Like making recycled rubber from recycled plastic for example

[dusky dust]

Sounds like something that sftools (and probably other tools) could do

[plucky tusk]

You could prolly do it in game tbh

[dusky dust]

Though keep in mind that sftools is a solver, which generally tries to solve for resource efficiency (using some hardcoded resource weights based on the total availability of things on the map)

[sullen barn]

Im unable to use it that way rn tho. I dont understand how

[plucky tusk]

But i might be misunderstanding smth

[dusky dust]

So if you have more than one recipe for an item chosen in the recipe list, it may not choose the exact one you want. If you want to use a specific recipe, make sure that's the only recipe for that resource that's enabled

[sullen barn]

I dont want a solver i just want a planner

[sullen barn]

That’s what my hope is lmao

[dusky dust]

Heh, well I led with "sftools is a solver," so we could've stopped there if that was the requirement. :)

Anyway, I'm not personally familiar with planners. I assume Modeller is probably the most popular at the moment

[sullen barn]

Well i dunno what a solver is but it couldnt really help me

[plucky tusk]

Nah dude i usually do it all in my head so idk what planner would be good

My best advice would be to start from one end of the production and work ur way down the line tor figuring out what u need

Normally when i plan in game ill just place machines down and run all the numbers from there

Its not super complicated math but its a lot of calculations

[dusky dust]

A solver is basically telling it "I want to make 20/min RIP + 10/min Modular Frames" and then it spits out this: https://www.satisfactorytools.com/1.0/production?share=JIQPVbLZjPFH9VVRUJ7o (using vanilla recipes on that one; you'll also have to clickthrough to the actual tab once it loads)

(I know I said I recommend against mixing production chains, but early-game ones aren't bad and since you wanted to do it anyway, I figured I'd throw it in)

Use the "recipes" tab to tweak what recipes it'll consider, and/or the "inputs" tab to specify what resources are available

ie: it solves the factory chain for you (again, in sftools' case it's generally solving for resource efficiency, using some hardcoded resource weights)

But since you can tweak all the recipes/inputs available it's generally pretty easy to get it to do what you want, if you have some specific recipe/resource you want to prefer

But yeah, if you want a planner instead of a solver, sftools isn't for you. :)

[sullen barn]

Why cant i use a raw resource as input

Makes me kinda angry

Well doesnt matter

[dusky dust]

(I suppose I do have one opinion about planners: IMO the best planner is just an alternate Advanced Game Mode save where you've got fly mode and everything unlocked, etc. That'll generally always give you the best knowledge of how to fit things together)

[dusky dust]

Not sure what you mean? The available raw resource counts are right there in the "Items/Input" tab.

[sullen barn]

I saw that before but i understood the conditions of it choosing those resources

[wind spade]

how would it be different from the existing raw resource limits?

[sullen barn]

Thats not what i meant. Why problem was it not choosing a raw resource but a processed one as input. Doesnt matter

[wind spade]

if you provide more info (e.g. a share link), we can help you

[dusky dust]

If you put in a specific input (like you're producing 30/min Modular Frames that you want to make use of), sftools will indeed prefer those Modular Frames versus making new ones, if some part of the production chain requires Modular Frames

If you don't want those MFs used in that graph, you'd have to not add them as an input. :)

[fierce ruin]

A new belt tier was added to fully utilize tier 3 overclocked drills on pure nodes. When are we gonna get a belt to utilize this?

[quick gorge]

no

[unique cypress]

Doesn't the engine cap out at like 2k/min?

[quick gorge]

Well the mk6 belt still doesn't seem to be stable when transferring shite according to #satisfactory-experimental

[plucky tusk]

Yeah would be cool

Then i could fully max out my liquid biofuel

Or at least the solid biofuel.prolly got enough liquid biofuel burners

[unique cypress]

I haven't had issues myself so idk

[quick gorge]

I love consistency, I wish we had some 💀

[plucky tusk]

Inconsistency is natural

[quick gorge]

This is a factory, where is any of that natural. We've been put on this planet to do very unnatural things

[sullen barn]

Thanks greeny. Everything is working just fine

[woven swan]

i have something extremely stupid planned :D using the 1200 m^3 per minute oil nodes + diluted fuel + ionized fuel to create the worlds most terrifying rocket fuel power plant

ill let y'all know how much this insanity makes

supposedly makes 4800m^3 rocket fuel per minute :D

realizing this is incredibly stupid now because oh my god that would take so many fuel generators

[prisma kraken]

it seems in 1.1 that the mk6 belt can slightly exceed 1200/min allowing it to make up for hiccups

[quick gorge]

Interesting, I haven't personally tested those things... I'm riding this drive to build pretties as much as I can

I will put industrial storages at the connection point of all the belts, after the balancer and just before going on the belts thag will eventually lead to the processing part

[prisma kraken]

yeah, my observations are pretty unscientific, just the belt item counter seems at times to bounce up to 1201/1202 on a full line

[prisma kraken]

i did a 2000/min rocket fuel using default rf + blended tf and i found it to be a big but not awful project. clocking the gennies to 250, it ended up at 192 of them

if you were then going to convert the rf to ion fuel, that pushes the gen count up to something like 266.6666

[plucky tusk]

There was a guy in the other chat makin 9444 rocket fuel

[vapid gorge]

yeah, it's... weird honestly. spamming 2000 generators is just tedium.

[plucky tusk]

Yeah

I used to always max stuff out but u always get burnout doin that. Havin more fun on my new save doin the bare minimum bc it gives me more options for factories

[sullen barn]

@wind spade why does ur calculator enjoy packaging stuff so much?

[wind spade]

what do you mean?

share the production line you're talking about please 🙂

(blue share button in top right corner)

[sullen barn]

https://www.satisfactorytools.com/1.0/production?share=ZM3y2a9V389lWGmJUXHj

Wanted to send a photo lol

[wind spade]

this way it's easier for me, as I can look at everything 🙂

[sullen barn]

Isnt the packaging here just a waste of energy?

[wind spade]

no? it's using the diluted packaged fuel loop

that's normal

the diluted packaged fuel recipe makes packaged fuel, so it needs to unpackage it

and it requires packaged water, so it needs to package the water

[sullen barn]

Oh wiki says its 5 residue fuel + 10 water in a blender

[unique cypress]

Don't use maximise mode

[wind spade]

that's Diluted fuel, not Diluted Packaged Fuel
those are two different recipes

[unique cypress]

Or, if you do, copy the amount produced and switch back to items/min mode

[sullen barn]

But why does it make the packaged fuel?

Nvm

[wind spade]

because the recipe makes it?

[plucky tusk]

Yeah

This why i plan in game

[sullen barn]

Why does it use that recipe? Why not the liquid one? Than all that packaging wouldnt be required

[unique cypress]

Because you're using maximise mode

[wind spade]

both recipes make the same amount of fuel, so for the tool it doesn't matter which one you use

[sullen barn]

Yea but one requires tons of energy and the other one doesnt

[wind spade]

blenders use more energy than refineries, the total difference is very small

and Tools don't care about power

[sullen barn]

kk

[wind spade]

(and you shouldn't much either 😛 )

but yeah, the hint about not using maximise also applies

[sullen barn]

Ye

[unique cypress]

Diluted fuel uses only 10% less energy than diluted packaged fuel

It's mostly for convenience

[deft lichen]

diluted fuel is more power efficient than diluted packaged fuel though

you can't conclude that just by comparing the power draw of a blender with a refinery 😛

there is no reason to use DPF once DF is available, DF is superior to DFP but unlocked later

[wind spade]

unless you don't want to do blenders (e.g. lack of materials)

and I personally don't like DF, it feels it should be slightly less oil efficient than DPF

[unique cypress]

It's unlocked 2 tiers later, I feel like it's allowed to be better than its predecessor

[wind spade]

given that majority of the recipes and stuff in the game is designed to never be objectively better, I'd disagree

[deft lichen]

we talked about that earlier

[unique cypress]

Yeah, but you never really get anything that's a direct upgrade: same inputs, same outputs

[deft lichen]

it's apparent DF is what they intended, but couldn't implement because blenders didn't exist back then
yet removing DPF now would have a pretty significant effect on how oil is dealt with before tier 7

[wind spade]

I agree with the fact that it's how they intended it to be, but they could've still balanced the two options in line with rest of the game (more complex options yield more goods)

[plucky tusk]

Yeah i think its fine the way it is

[unique cypress]

I think it's already balanced by the fact that it's unlocked later

[wind spade]

unlocked later doesn't mean it needs to be better in all parameters

[plucky tusk]

Since when

[wind spade]

since the whole game is balanced around having options, not having to redo your whole production every tier

[oblique hollow]

Blender benefits from the fact that you get good output via overclocking and also Slooping (if you wanna be that funny)

[wind spade]

yeah and my point is that I'd like the DPF to be better in some way 🙂

e.g. do 300 oil -> 850 fuel instead of 300->800

[plucky tusk]

Doesnt need to be

[wind spade]

would be nice, to fit in line with pretty much everything else in the game

[unique cypress]

I think they have bigger balancing issues than that lol

[wind spade]

not disputing that either

[unique cypress]

Wth is that layout bro

[crimson moat]

recycler loop off google for someone 😄

[unique cypress]

But why is it in a circle

[crimson moat]

A lot of stuff goes in a circle in P3-P5

This particular recipe is:

1 fuel + 1 plastic = 2 rubber

1 fuel + 1 rubber = 2 plastic

If you send half of the rubber/plastic back into the other side of the recipe, what you get is this:

1 fuel = 1 plastic
1 fuel = 1 rubber

because you cancel out the rubber/plastic input requirement

[unique cypress]

I'm not talking about those recipes. They're fine

[crimson moat]

oh the position of the nodes? Complain to the guy who put it on reddit 2 years ago 😄

[unique cypress]

I see lol

Also, in case you didn't know, the recycling loop can make all 3 items in almost any ratio. You don't need to make equal amounts of all 3

[slow crater]

Does it make more sense to have factories constantly producing and sinking excess or have them fill up buffer containers and thus shut off once the system is full? Basically is power hard enough in the long run to keep up with that I should ditch the big number go brrrr of the sink?

[fallow siren]

if youre saving power, sure

[plain rivet]

I try to sink end products so the lines stay busy. Power gen is easy enough to scale to keep pace with demand even if everything is running

[fallow siren]

i just sunk all overflow to sink, to keep factory running and free points

[plain rivet]

Plus relying on idle machines to stay under power limits seems like a bad idea. You grab too much stuff at once from storage/depot… all those machines kick on and your power fails

[wind spade]

generally there's no reason to not run factories at full speed and sink excess

since power is produced (ideally) at a fixed rate, there's no reason to save power (except maybe pre-coal)

[vapid gorge]

as I'm going up the tiers I'll let most of the factories fill and shut down. even a few sections being fed into sinks gets you everything you need

[slow crater]

This is a good point. I’ll just keep it as is and try to keep scaling up power

[restive sparrow]

I think there is at least one very good reason to keep things running and sinking: interconnected dependencies. If you're making plastic and rubber in a recycling loop you can't just let you plastic stop producing just cause you're not using plastic, because that will kill your rubber production (which you may be using) and visa versa.

Edit: Similar applies if you're making power from fuel and you're also making plastic or rubber from those same refineries. If you're not using the plastic or rubber you'd need to sink them if you don't want your fuel production (and power) to die.

Mostly just let stuff die off if you're not using, but it's important to think about dependencies once you get into the refinery age and beyond.

[vapid gorge]

sure but in my example those dependences get full and also get on standby xD

and if you split your recycled plubber into their own systems they don't mess about

[restive sparrow]

I'm having a hard time imagining how a split plubber system could be stable even if you stop using one or the other, but even if you manage it, I also don't see the reason to think about it that hard when, as Greeny pointed out, power production is mostly static so if you've got it budgeted in why not just let it roll.

[oblique hollow]

overflow

[unique cypress]

if you design your recycling loop properly, there are no issues like that

[restive sparrow]

"properly"

[unique cypress]

Frankly, I'm not sure how would you need to build it to have this issue

[restive sparrow]

thing feed into thing feed into thing. Stop taking thing...then thing stop too. I dunno what to say. Seems like a pretty likely scenario.

Go make some rubber with your fuel refineries that you're making power with. Now stop using the rubber. How's that power doing?

[vapid gorge]

not even over flow

start off with your final desired product of each.
keep halving each system until you know how much residual rubber you start with

[unique cypress]

That's a completely different situation? You're using a byproduct from one item to make another.

[vapid gorge]

so in this example 800 of each, the side numbers are fuel being injected

[unique cypress]

This is not the case in a recycling loop

If you take some plastic while rubber is backed up, it'll just take some of that rubber?

[restive sparrow]

I did mention both scenarios.

[restive sparrow]

Yeah that makes sense. In the case of the recycler loop perhaps I wasn't thinking it through completely.

[unique cypress]

And I only mentioned the recycling loop, because yes, the other setup absolutely will have this issue. I don't see how the recycling loop would

[vapid gorge]

does the diagram make sense orbain?

[restive sparrow]

Yes.

[vapid gorge]

cool. personally I find it a much simpler design and layout problem keeping them split. My first one was... not split. But worked. But was definitely more work

[restive sparrow]

Yeah mine is certainly not split, but perhaps my next one will be.

[vapid gorge]

oh and I think this was an example of each of hte blue sections for the rubber on the left

at least one possible layout for the first 3 steps of it

sleep time though!

[frosty owl]

When power production was not static, I found it fun to set up my Plubber running on overflow Fuel from my Turbofuel production: the closer I was to my max power, the less Plubber I made 😆 (not all that useful, but fun to run)

It could have been useful if I dismantled my Turbofuel plant when switching to Nuclear

[quick gorge]

@cinder silo has the sun always been in the void?
... wait... that's.

Oh no

[tall hazel]

Does anyone know the equation used to calculate the time it would take several different machines in a manifold setup to reach full efficiency?
As an FYI, I'm aware of the online calculator that exists but I want to setup my own calculation using excel.

[opaque quartz]

It would be a function of:
stack size
belt speed
Input amount/rate
number of machines
Recipe consumption item rate

[unique cypress]

also consumption rate per machine and input rate

[opaque quartz]

Yes good catch

[unique cypress]

I don't think there's a single equation

or if there is, it'll be very complicated

[opaque quartz]

It would be a geometric series, I think