math-and-meta

[leaden depot]

yes, I have buffers on the station, and its all ficsmas stuff that stacks to 500. It is not a throughput issue.

[prisma kraken]

transfer rates for stations are inexact... there's always a little variance due to the fact that their moving quantized stacks, and if you have multiple trains on the same circuit the times in the station will always have averages that swing one way and then the other with each train's delivery

if you're seeing something like 718/720, that's 'good enough'

[vapid gorge]

yeah don't trust the efficiency meters or transfer rates as absolute

[thick mortar]

Is there a possible way to split 5x 60 output into 3x 100 with only 120 belts?

[vapid gorge]

clock the machines to make 6x 50 is the simplest

or clock the things that consume the product to use 2x 120 belts and 1x 60 belt

managing specific outputs at the belt stage is usually much more of a pain in the ass than managing it from the machine ends

[thick mortar]

Clocking isn't an option right now, it would disrupt the rest of the design

Too much*

[vapid gorge]

youve built the next stage w/o building the middle belts?

rip

[thick mortar]

No I just spent like an hour and a half doing maths and designing and don't want to restart

[vapid gorge]

and you can always over clock a machine and keep the numbers you have

[vapid gorge]

for future arithmetic https://www.satisfactorytools.com/production

[thick mortar]

Yeah, we've been using that for ages, just never really liked it much.

[drowsy thorn]

you can get to 100 by back splitting a 60 line then back splitting it again and just moving 10 foreward so split down to 30. then on the other side split to 15, then split again to 5 5 5 and push the last 5 back. that gives you 40. Add that to 60. Profit.

[vapid gorge]

that will probably take much more room than changing a bit of clocking but could work

[drowsy thorn]

that WILL take a lot more room than changing clocking.

it's.. 4 splitters? and you'd need to do it three times.

you didn't ask if it was a good diea. you asked if it was possible 😉

[vapid gorge]

yeah I don't know the exact machines and needs but just at a glance you could probably alter the clocking groups and not change the layout of the factory at all

[drowsy thorn]

or do a 5 to 3 load balancer. wind up with three lines coming out. one side would back up if it's only consuming 100. it's effectively an even split.

also takes a fair amount of space.

[vapid gorge]

I'm not 100% sure they could do a 5:3 LB with those belts? but I'm not going to dedicate brain space to it

at least not w/o just doing a bunch of tiny splits

[drowsy thorn]

well heh. you could do Merger - Splitter - Merger
Splitter - Merger - Splitter
it's not a real 5:3 merge. but it'd kinda work if the consumption was limited to 100.

[vapid gorge]

ugh balancers just give me a headache

[drowsy thorn]

yeah. but they can be so.. engaging.

[vapid gorge]

but ... its so much work for no real benefit xD the aesthetic of it is so minute too

[drowsy thorn]

eh. I dunno. I use them to do alternate production lines. when one backs up the other line kicks off.

But they're ugly as sin.

it's more fun to do sushi belts with ratios

[vapid gorge]

shouldn't a LB never back up?

[drowsy thorn]

when the output buffer filles it always backs up.

[vapid gorge]

oh if you don't sink the excess - manifolds will do that too though

[drowsy thorn]

yes they will 🙂 I mean is it a manifold or is it balanced, because thats just a way to push the excess over to another side. I wouldn't count on it to balance 😉

[main dirge]

Does using lots of balancers (or splitters/mergers in general) impact performance ie compared to machines? Idk much about how the game is optimised

[vapid gorge]

I would guess if it ends up using more objects to accomplish the same job it would be more but probably not hugely significant compared to overall game play

like the difference in obj amount could easily be overshadowed by people decorating a bit more than others

[main dirge]

Fairs, lights are problematic right? Was considering them but might remove if so

[vapid gorge]

I've heard signs specifically have had an issue lately - no idea how much they fixed that up though

[main dirge]

Putting a few huge factories near each other so trying to be cautious before I get to the next one

[vapid gorge]

the 'huge factories' bit will be more the issue here 😄

[main dirge]

I'd be screwed without my notes everywhere for throughput 💀

[vapid gorge]

oh yeah, notes on what your belts are doing are super useful

[main dirge]

The desert mesa - swamp - titan forest intersect just has way too many resources nearby (un)fortunately

Keeping it mostly depowered for now so should be able to build

[vapid gorge]

those sound pretty spread out tbh

[main dirge]

Nah the point between the three

[vapid gorge]

ah

[main dirge]

Approx bc I've just turned off computer

[vapid gorge]

yeah doesn't look too bad

[main dirge]

Get issues around the refinery but way oversized that one

Would power up steel to test but I've put off nuclear for a few too many hundred hours xd

[wind spade]

out of interest, what don't you like on it?

[heady vine]

i'm glad to share fact nobody needs:
its nice to plan dilluted fuel in patches of 10 generators
cause it needs exactly 120 water / min
(and in my extremely unlucky/blessed case, it fits on mk2/mk4 belt)

[snow dove]

i grouped it up into groups of 6 blenders, which then output 300 fuel, which is perfect for 25 generators

[quaint ridge]

I don't use it either, I find my spreadsheet more ergonomic. The visualization is not easy to read, it's not fast and easy to tweak the numbers and check the result to find which solution suits you best.
I created a spreadsheet that shows me all the (chosen) recipes (I chose to use the alternate recipes that require to build the less buildings possible), then I just input the amount wanted for a given item and all the sheet update, telling me the amount of building needed for each item. I find tables way easier to work with than graphs, I know where the information is from the beginning without having to look around and follow arrows from one node to another

(this being said, good job for creating that tool ^^)

[wind spade]

there's building/item lists in SFTools as well 🤔

[true junco]

I like my overclocked BP that turns 75 oil into 200 fuel.

[quaint ridge]

this list gives you the total amount for a given building, I'd prefer "you need 10 constructors for that item, 4 constructors for that other item, ... "

[wind spade]

you can click each building to expand a list for each item, see the screenshot

[quaint ridge]

Oh ok that's how you use it

[wind spade]

(the same for items and power sections)

[quaint ridge]

(but I'm still used to my spreadsheet :p the information is more compacted and easy to work with (for me ))

[wind spade]

yeah that's completely fair. I'm just looking for feedback on what could be improved, so I often ask people what they do not like. I'm completely fine with people not liking/using my tool, I can't make it so that everyone will like it, but if people don't like it, I may try to fix the reason why they don't like it 🙂

[primal flicker]

Note to self:
Never try to build fuel in manifold groups again. This shit just won't saturate and give a nice flat power graph.

[fierce ruin]

load balancing 4 life

[primal flicker]

it worked on paper
not so much, in practice

[fierce ruin]

dude thats how I feel even with load balancing. I get everything right up until the miners so now I have a little system of mine where I create "transformer boxes" that are splitter systems for my miners. kinda hides the ugliness of having 50 trillion splitters/mergers.

it would be nice to have "smart splitters" that split specific numbers instead of items.

tbh I would work work for coffee stain studios for free just to shoot ideas on work on stuff every once in a while 👀

[vapid gorge]

can't load balance fluid

[fierce ruin]

that is not true

[vapid gorge]

fluids are bidirectional, literally impossible

you can do splits over and over but it doesn't load balance anything

[mystic moon]

It is actually impossible

[fierce ruin]

yes it does lol. if you do a 2 split to 2 split it will evenly split 4 fluids you just have to oversaturate your system and have pumps. In my experience it will load balance.

just place the pumps close before they start to fail

[vapid gorge]

you can get it to stabalise - but pipes aren't belts, fluids are bi directional

and putting pumps just sets where any back flow might stop

the way you handle pipes is pre flooding it to manage backflow

[fierce ruin]

it might not be perfect but its the closest thing to load balancing. In an engine you can gain pretty perfect efficiency by distributing the fluid equally and in an even distrubution. The same mechanics work in pipes for liquids. I promise you it might not be perfect but if you saturate the system 3:1ps instead of 2:1ps then it will load balance perfectly. I have a coal plant in a world currently that the water is perfectly loadbalanced to power all of my coal plant efficiently. If I wanted to run it 3:1ps I would but I'm wonky and like my factories to be completely saturated. fluids are quite bidirectional you just have to have to have enough pressure for it to distrubute properly. Atleast this is applicable to the real world. In terms of satisfactory I oversaturate my system to make it more aesthetic. But I promise you that it is bidirectional lol.

tridirectional idk you have to have valves atleast in the real world. In my experience in Satisfactory however, I don't even need valves..

[vapid gorge]

but that's the thing - it's not delivered as a load balance. It's imaginary load balance. You even admit you try to put pumps on it as you know it can flow back.

The closest you can do is the appearance of fluid being evenly and continuously delivered. Which you're free to do, but you get the same effect with a fluid manifold

and if you can get a fluid system working like that? fine sure, shoot your shot. But it's not load balanced

[fierce ruin]

if you want to add valves to each manifold then sure it is the same but if you add pumps to reestablish the pressure that was initially in the system then it is a load balanced system.. It's kinda like an exhaust manifold how a 4:1 is going to cause a lot of pressure in one area where as in a 4:2:1 it relieves the pressure into seperate blocks. Idk if satisfactory has the same physics that apply to real life but in terms of an actual fluid delivery system a load balanced pipe would be a much better solution than a straight manifold lol

and it's only imaginary because we haven't reached a point in physics in which we can distribute fluids with 100% efficiency

if we do I'm fucked I'm not going to take fluid dynamics again XD

btw I am talking about the horizontal axis. when it comes to vertical im sure you know but you have to do a whole other range of calculations.. However, I've done an experiment on the distrubution of water through an intersecting pipe with complex shapes to change the flow of fluids and my final result was a pretty evenly distributed amount of water(L). I don't know what the effect would be on other molecules I'm just speaking off experience 🤷‍♂️

[delicate chasm]

I was scrolling down, skimming, and had to do a double take because I thought this was just an America meme.

[vapid gorge]

nah, the valves still just add an extra back flow point. But having a stable system that consumes the right amount of fluid pm does not a load balance make. Belts can evenly and continuously distribute items. You can only pretend that happens with fluids in this game

[delicate chasm]

Reading over the conversation above, I feel it might be helpful to point out that our liquid simulation is incomplete and has a few known quirks (both of these may change in future updates). But for now:

• There is no pressure. The system works on equalization by measuring the fill level of the segment along its length and calculating where in the pipe the water should be, if less than full.
• Head lift is our analog to pressure. It's a real concept and works as you would expect. If you aren't familiar, a simple explanation is: liquid in SF does not have any energy to flow upward against gravity. Every machine adds 10m of head lift, and pumps add 20 and 50 for mk1 and 2 respectively. This only represents the ability of the liquid to continue to flow forward through a vertical rise. It will still prefer to fall to the lowest point in a pipe segment that is not full.
• Machines along a horizontal input manifold will "chug" as we often call it, as much liquid as they can the moment they finish a cycle. This reduces the fill level of the pipe segment immediately connected to them. That makes it thirstier than the downstream pipe segment in the manifold, so the flow reverses (to equalize the fill level, remember?) because it's not pressurized.
• Pumps and valves prevent backwards flow but they don't prevent the pipe segment from trying to equalize on both sides of the valve or pump, which creates some odd interactions together with machine chugging. Recommended to not use.
• Buffers are like pipe segments but they have exaggerated qualities when it comes to fill level equalizing and flow rate. The reason that a horizontal manifold that isn't prefilled won't work can be put on display by simply building a fluid buffer in the middle of it and turning it on.

[fierce ruin]

I dont know how to explain this in any other way then they do. I promise. if you have enough pressure in your system and the pulse is the same throughout the entire system it is load balanced. Again this probably isn't applicable to Satisfactory; However, in real life is the frequency of fluid remains the same throughout the piping it will remain load balanced. The efficiency can be gained by adding valves; However, again we're trying to discover ways to load balance fluids perfectly without valves. In terms of satisfactory. When you keep the same pulse throughout the entire system and it can distribute properly amongst the output then it becomes the most efficient way possible because you remove the necessity for valves which subtracts an entire pay group from having to be implemented. My current design is most efficient because I dont have to use as much resources + development to implement my system.

[vapid gorge]

the main issue is that you can't control flow at every point, you can't prevent fluid from moving back at every point - you can, on average have the amount a machine needs move through a sections - this does not mean it moving it consistently nor that it might not move backwards at points

[fierce ruin]

I'm intoxicated so please bear my weird wording but I promise you its possible XD and more efficient

[vapid gorge]

BTW - were you a member of this server ages ago and quit?

[vapid gorge]

you can't get more than 100% efficient so not sure what you're on about

[delicate chasm]

All good. Point of order, isn't a load balancer defined as a directional (loading ->) input though? We all know it's bidirectional, so it's probably just a semantical difference here.

[vapid gorge]

it'd be consistently moving an even amount along the branchs, not delivering the amount needed on average

it sounds like a small distinction but it immediately stops being the thing w/o it

[delicate chasm]

Yeah fluids won't do that. The whole network is connected and there will be flow direction reversals and uneven flow rates at all times.

But OVERALL...

As you say, that's just a manifold.

[vapid gorge]

exactly.

the very long handle and argument feel very familiar to me. I think they used to be on the server

my main issue is I don't want them telling people you can load balance fluids - they can do whatever else they like

[vapid gorge]

look I see you typing there - the only way you can load balance fluids is if you actually make pipes mono directional - explain how you do that and you win

and thats mono directional at every point , not jsut at random pumps or valves

[median heath]

Also you've already stated there is no pressure to which their reply was "if there is enough pressure".

[fierce ruin]

you're arguing with me over the same theory.. You can load balance fluids 3:1 instead of 2:1 because of "the game mechanics" or "our current theory in fluid dynamics". However, it is load balanced to the most efficient system. If you create a manifold system then you deal with expessive pressure at one point rather than distributing it throughout the piping. Thus, creating an inefficient system because if the developers decided out of a whim to implement new fluid dynamics theory system then you would be creating a point that burst and your manifold system wouldn't be effective. I state my point that youre creating an inefficient system.

[median heath]

It's been proven beyond a shadow of any doubt multiple times that you cannot load balance fluids.
So your premise is incorrect.

[delicate chasm]

The lines are getting blurry between which parts are about Satisfactory and which parts are about real life. Kind of hard to understand now.

It might be good to come back to this another time. :)

[delicate chasm]

They are intoxicated and typing a bit slowly, and let us know that earlier. It's alright.

But that's also why it would be a good idea to pick up the topic a different time.

[vapid gorge]

Yeah, just gonna assume it's drunk talk. Night all

[median heath]

Sheds light on earlier conversation too.

Sleep well.

[fierce ruin]

Dude if you show me your transcript that proves youve taken atleast principles of physics II then i will listen to what you say but please stop arguing with me when I know how fluid move throughout a pipe. Fluids in Satisfactory move like a frequency in real life. I have experimented with this applying what I have learned into the actual game. It's not insane but it's the game. The pipe might be fluctuating between 15m^3ps an 0m^3ps but it is also producing and supplying that amount of liquid. The only reason why it's slight inefficient is because it differentiated between decimal numbers. In this instance you have to round to numbers greater than just 2:1. I'm pretty intoxicated but I did my own experiment in satisfactory just to figure this out. I'm unsure about vertical pipes because of the vertical physics going down but keeping it at that same frequency that it started off with will keep it in relative distribution. 😦 . I will provide a screenshot of my plant tomorrow but if I were to start a new node of the factory it would run at a 3:1 ratio because of the water. I really couldn't tell you the physics behind it and I'm sure like avelon said the physics are completely different but I've done it I promise. The physics are actually quite similar to real life besides the measurements which is why I like it. it is difficult because you cant necessarily have pipes that are uniformly distrubuted without valves but youre reducing resource cost and complexity when you distribute the liquids with pumps. interiors of pipes can have certain shapes to direct flow differently and for some reason in my experience satisfactory distributes it quite equally. So I dont bother with valves and the chance of innefficiency because I wanted to gain a little power.. Irl I would add valves anyway to ensure that it's recieving the right flow but in satisfactory not only does it look more satisfing for it to be distributed equally like conveyers but it's the same if not slightly

less efficient. Have a good night 🙂

[median heath]

Book that begins with "show me your credentials"
And pinging just me when there were other people in the conversation telling you the same thing...

👍

[fierce ruin]

🤦‍♂️ if youre that intrigued by discord that you reply to notifications than maybe hop off and find a job that pays you a salary. Myb if I was replying to you but in no way does that require you to respond with nothing that adds. Again I said I'm slightly intoxicated. Again you did not read and instead commented on how I pinged you. I have not made any remark specifically toward you and yet you make it a point to feel as if I'm attacking you. If it's that big of an issue than maybe consider using your medical insurance to pay for therapy. I am not targeting you but making a point and yet I can differenciate me making a joke between you making an obvious opinion to keep things the way you like it. The only thing I've seen you make a point on is how you feel like other people are targeting you because of your opinion. Not consider that your opinion is fascist towards others point of view and also just making the most obsurd comments to justify you being weird and unreceptive. I dont know what your "role" is but I've been playing the game since they published it and I would only like to add some suggestions. You dont have to read what I said because I know you have a stigma about long text but I'm only contributing to the game as a fan. I wouldn't defend myself so much if it weren't for people like you who only respond to me with a simple comment. I hope that you wake up and consider that my input into the game is only subjective and I would like it to be more optimized like other influencers. If you think that LetsGameItOut content is hurtful than I can only imagine what your views are about the world.

[median heath]

if youre that intrigued by discord that you reply to notifications

Wait so the point of being able to receive messages from people is to.... not reply to them?

Been doing that wrong since email and text were invented...

If it's that big of an issue than maybe consider using your medical insurance to pay for therapy.

This is a fun comment 😄

[lavish quiver]

Slapstick

[fierce ruin]

dont be so fast to dismiss my idea. "load balancing" or whatever pipes looks really good lol

[frosty owl]

Link?

[main dirge]

All Americans know is oil, steel, nuclear fuel, eat hot chip, and lie 😔

[oblique hollow]

i think this is all just semantics

[oblique hollow]

Idk how much of this is drunk talk but in case you care about this:

Yeah, Satis pipes pretty much are simulated waves (so akin to frequncies).

The fill volume of a pipe determines how much head lift there will be (despite whats stated above, that IS in fact pressure).
0 to 1.5 m head lift proportionally affects flow rate, 0 to 300 m3/min.
Above that, it just sustains flow up to 300 and is mostly used for vertical traversal now.

If you work with them and with valves, they are quite similar to electronic circuits.

Which is where the whole load balance spiel frankly doesnt matter much. Equivalent resistances and all. What we do know is that for mk 1 pipes, it doesn't matter at all how you build them because functionally their slow speed prevents any malicious complications like water hammer or the like.

Mk 2 pipes are a little more complicated, but thats due to flaws in the coding itself.

[frosty owl]

Note: many simple balancing designs and even some loop-back ones need as many or less belts/splitters than single-sided manifolds

I was quite surprised by it when I ran some numbers

[vapid gorge]

that's where the 'if it ends up using more' comes in 🙂 I do happen to use more single sided manifolds these days

[frosty owl]

The things we do for aestetics~

[oblique hollow]

technically any load balanced pipe system is, topologically speaking, also a manifold

its just the injection point that is different

so its both in the end

[vapid gorge]

how would it be both?

[oblique hollow]

If you dont change the manifold but change the entry points, the behavoir switches

[vapid gorge]

talk to me like I'm 5

[oblique hollow]

Junctions split flow.
Manifold have unequal split
Load balance have equal split

what makes unequal split into equal split?

If you have 5 marbles, how do you unevenly split them in 2 groups vs how do you evenly split them in 2 groups?

How do you unevenly split 6 marbles into 3 groups / how do you split 6 marbles evenly into 3 groups?

[vapid gorge]

I may be too sleepy for this to see the point you're trying to drive, sorry xD

[oblique hollow]

Just answer the question about the five marbles

I give you 5 marbles. now split them into 2 uneven groups

How many marbles per group?

[vapid gorge]

any number =/= 2.5?

[oblique hollow]

basically yea

1 / 4, 2 / 3

[vapid gorge]

but fluid doesn't work in whole units

[oblique hollow]

or for sake of completion, 0 / 5

i know, this isnt about flow

its about junction divisions

where you put the input pipe in a manifold is basically the same as moving the divider when you divide your marbles into groups

[vapid gorge]

ok ok sure I think I get what you're trying to say - but even assuming pipes were monodirectional there'd only be one spot where you could inject a system to have it LB right?

[oblique hollow]

assuming you only have one input......

the fun part is having 2 inputs can be seen as a load balancer feeding into a load balancer

depending on how much each one inputs

[vapid gorge]

I think this might the sort of thing I'd need to sit next to you with some sketching going - very visual and don't have internal imagery

[oblique hollow]

Junction behaviour, when not met by any backflow, is inherently like conveyor splitters: equal splitting

[vapid gorge]

sure

[oblique hollow]

Yeah its very visual

[vapid gorge]

hmm wait, assuming junction was mono direction would it still evenly split the fluid if there was a difference in volume in the pipes ahead?

like if one path was 0% and the other was 20% full

[oblique hollow]

it would already be unequal in that case as the 20% full one will try to flow to the 0% full one

it will try to do an absolute pressure equalization, but thats a bit complex

and usually takes a few turns

like, the full input pipe would probably just get split wash through both

[vapid gorge]

that's what I though

[oblique hollow]

but a test would be more determining

even with equal splits, theres supposedly a tolerance of like 20% concerning how much more one side can chug on the junction

this is ultra obscure info back from the Update 3 launch

[vapid gorge]

yeah fair. I suppose you could, in theory, set up like a simple 1:2 Pipe system, make sure everything is empty and maybe try to claim it's load balanced.

[oblique hollow]

once one side becomes unbalanced for some reason, theres a pretty gray area when it comes to " how will this mess up the flow"

Thats where knowing how a backflow wave acts would be useful to know

[vapid gorge]

yeah any disturbance and it goes out the window

[oblique hollow]

the question is for how long

because a constantly unsteady state is unlikely

unless theres periodic disturbances

[vapid gorge]

if the machine used very little fluid pm it could be very stable

[ashen stirrup]

So... the pinned pipeline guide gives several examples of when valves should be used. I'm pretty sure I've seen here something to the effect of "don't valve, ever". Does that mean the guide is wrong?

[wind spade]

well it's more that the setups that use valves are kinda pointless as well

the "don't use valves ever" basically applies to "unless you really know what you're doing" (but simple pipes = best pipes, so I'd keep valveless)

[ashen stirrup]

Gotcha, I thought it was to do with a bug with valves

[wind spade]

it kinda has

valves can be only set to 255 different values, so if your desired amount isn't in incremenets of 2.35294 (600/255), then the valve would be set to different amount than you want.

[vapid gorge]

It's also got to do with if you have a stabalised system working it'll work with or w/o the valve

[wind spade]

and since pipes work best when they are full, valves that split partial amount are kinda pointless given pipe is full everywhere

[oblique hollow]

If you use valves, you kinda have to set it to a value thats marginally bigger then whatever flow you want out of it

and even then, the manual lists the many issues you face when you use valves in an improper way

[vapid gorge]

I use valves for aesthetics? that's a good use for them 😄

[wind spade]

or just leave it at full, but then you can just do without valve anyway

[oblique hollow]

like how you need to apply a limit to every output

[vapid gorge]

just clip the pipes through

[ashen stirrup]

Ok, I just wasn't sure if the guide's guidance on the flaws was incomplete

[oblique hollow]

Meta constantly moves forward and the manual only ever gets changed when its "time"

[vapid gorge]

I don't really remember the last time anything signifcant changed for hte pipe manual though?

[oblique hollow]

So unlikely it is ever completely out of sync with current knowledge, but a few points can get a bit "outdated"

[vapid gorge]

like the last big pipe thing I remember was them fixing the fluid loss bug, but that didn't effect set ups afaik

[ashen stirrup]

I'm mainly just sitting here waiting for my train and it's itching at me that 1 to 3 of my coal gens occasionally cut out intermittently, even though I'm giving them more coal and more water than they need.

[vapid gorge]

is your train delivering water?

[ashen stirrup]

Oh no, an irl train

It's delivering people

[vapid gorge]

oh right 😄 , well there's a ton of tiny issues that can be happening with your coal set up

once you know pipes it's actually really simple - but it's also very easy to make tiny errors

feel free to share a pic of the set up, pref from a bit over head, and usually someone is on that can help diagnose an issue

I'm going to bed shortly, if others also wander off I'm pretty sure literally every coal issue has been discussed in #1038092680493801533 a million times and you'd be able to find a variety of posts

[ashen stirrup]

Thanks! It's a 3-tier plant with six groups of 4 plants each, so screenshots will be hard—I might try at some point though. In the meantime, that's a great idea. Thanks for the reading material, and have a good night!

[vapid gorge]

have you seen the simple coal gen set up diagrams? as long as you keep it simple and make sure you have enough head lift and not trying to cram more than the pipe can move at any point you should be good

[ashen stirrup]

If I don't have enough head lift, no water would get there, right?

[vapid gorge]

ish? it might be that you're just at the edge of it

Also if you're using floor holes they are a bit buggy and sometimes kill head lift in a pipe

one of the first trouble shooting things you need to do is click on your coal gens and open up their panels to see if they are missing coal or water

[ashen stirrup]

Yeah, I do have floor holes. I thought that was also an all-or-nothing thing though

It's frustrating as well, since the fluctuation will stop for several minutes at a time and then happen for 20 seconds or so.

I'll have to go through and check it again

[vapid gorge]

it should be yes

and it's probably just a small error 🙂

[ashen stirrup]

Is there a downside to Encased Industrial Pipe I'm not seeing?

[wind spade]

slower production per machine

[ashen stirrup]

Yeah, it requires more assemblers but fewer constructors

If that's it, I'll gladly take that

[wind spade]

also uses different material, which may be a downside if you already have tons of beams (but then again, why would you make something you don't need yet 🤷‍♂️ )

[spare zenith]

Hi Pioneers, I was wondering if someone among you would be able to help me to retrieve the mathematical proof behind the prime numbers splitting patterns (I don't know very well english mathematics' language, sry for approximation in term of lexic). I did establish it before, but I can't remember how...
So let's take an example : You want to split a stream in 5 parts. You split in 6 and reinject 1/6 of the stream in the entry of the splitting, so mathematically, this is :

Let (x,d) any numbers from (R,N) as x>0,d>1 let also u(n) R(N) n>=0 we have :

u(n)=(x+u(n-1))/d+1
u(n-1)=((u(0)+u(1)/d+1)+u(2)/d+1)...+u(n-2))/d+1
And we want to prove that when n comes to +inf, u(n) comes to x/d.
lim(u(n))n->+inf=x/d
But I don't know how to deal with it. Could someone help ?
I may wrote wrong things just below, feel free to correct me

[ashen stirrup]

I'm assuming you've read https://satisfactory.wiki.gg/wiki/Tutorial:Prime_splitter_arrays#Proof_of_existence ?

[wind spade]

https://satisfactory.wiki.gg/wiki/Tutorial:Prime_splitter_arrays

[spare zenith]

Thank you all !

[wind spade]

damn too slow

[spare zenith]

Mmh, that is not very rigourous as it's said

If someone has a better proof, based on the POV I took, I would be glad to discuss

I can vocal too

(would be a good update for the wiki too tho)

[frosty owl]

Maybe @stone delta woukd like to take the challenge?

[ashen stirrup]

Calculator... what? Why is it making all this extra concrete? 😆

[wind spade]

send link 🙂

[ashen stirrup]

https://www.satisfactorytools.com/production?share=VolHjp63qdFLk9GrNmf6

[wind spade]

change maximise to items/min 🙂

https://www.satisfactorytools.com/production?share=T2VrjQ0Dd8l1CrqL4fJP

[delicate chasm]

I don't want to misinform anyone. Where in our fluid system is the pressure, how can I observe it at work?

[ashen stirrup]

Okay, so that fixes it. Why would maximizing iron pipe also maximize limestone to a concrete byproduct though?

[wind spade]

maximise doesn't optimise for raw resources

it finds a solution, not most resource efficient solution

[ashen stirrup]

It is indeed a solution

[wind spade]

yeah, and it leads to max production of pipes 🤷‍♂️

[ashen stirrup]

Yeah, I just don't get the "byproduct" thing. It seems to have seen that I have 540 Limestone available and used all of that for concrete, then used only what of that it needed for the encased pipe assemblers... though none of that has any bearing on pipe production

[wind spade]

again, it stops when it finds first maximised solution, it doesn't optimise it further

[true junco]

Tools does some odd things as a result. Ive had entirely independant side productions show up before.

[wind spade]

that's mostly just rounding error (-0x constructor)

because computers and decimals aren't friends

[ashen stirrup]

I've gotta say though, this is massively useful. And I love playing with the flowchart to get something nice and compact.

[true junco]

Ive been using steel screws so much I forgot how many machines you need for other paths. 1000% of caste screw constructors vs 192.3077% of Steel Screw Constructors.

[primal flicker]

Compare to steel rod + default screw

[main dirge]

Big manifold user but gotta say I'm a big fan of how satisfying big belt managers look when the math doesn't work out into multiples of 780/600

[main dirge]

Here's the area I'm working in, trying to figure out where to put nuclear fuel production (if it needs moving away from 15k steel or splitting into two factories ie)

[delicate chasm]

That lake all the way to the left-center of your image is a good spot for UFR. Proper coal spot with immediate access to normal sulfur, a pure iron, and limestone - plus adjacent access to copper, and Titan quartz.

[main dirge]

Concerned about building in the forest itself though as there's not that much space, especially as the building goes up

Expecting a floor space of about 24 by 113 foundations

If I've done my math right

I'd guess 7-8 floors

27 water extractors needed but with power shards (if I have any left after nuclear ;-;) and nearby pools that's not too much an issue

[delicate chasm]

Yeah, that side of Titan is also where the land bridges are.

You can probably fit your footprint into that spot by running the length east-west. ENTERING Titan Forest with your build could be an interesting opportunity... Your final output area doesn't need to be tall for example. Or your consolidated inputs area, if bringing from east.

[main dirge]

Definitely want the general area though for the quartz, sulfur, and caterium

[delicate chasm]

27 water extractors should fit in that lake but I'd double check that and the foundations footprint of course.

Looking forward to seeing your build :)

[main dirge]

Filled in a patch of 24 by 56 - ...I may need to build more vertically xd

Or mb put it on top of the reactors actually

That build's already taller than I'd like though

plus lag

[main dirge]

Here's the steel at least

[true junco]

I dont care for that combo at all honestly. Conserving steel is just silly when its possibly to make way over 100k steel ingots per minute. Lol

[main dirge]

Is that not over a bunch of trees though?

[delicate chasm]

_>

[main dirge]

thinking one of these rn

[delicate chasm]

Sorry, had to. 😁 But yeah, and if you go west from there you might actually want to turn 90 degrees when you hit the geyser area and go into the canyon, or south and up the hill onto the relatively flat open area, with the water fracking site and etc.

But ah yes, why not use the desert? It's right there anyway.

[main dirge]

Start running into my old steel factory further West - unfortunately miscalculated some stuff so I need an extra belt of steel over the 15.6k from the factory so don't really want to tear it down

Plus further I go from that location the worse the resources are

[primal flicker]

LOL true.
I need to get back to my "smallest footprint, no matter what" approach.

[main dirge]

These resouce nodes give me enough ore whilst staying within my v low tolerance limits

Can only waste ~200/m of each across the map and these give ~60 above what I need so I can just use these without worrying about needing to link into other production later and splitting it well

[true junco]

Im not quite doing that. But reducing foot print drastically at the expense of the most common materials on the map is definitely something i think about a lot.

[frosty owl]

The FPS run out faster than the nodes

[main dirge]

Thinking about it more now though I may want to split it in half - oil refinery is unplayable near it bc of noise issues when on and it's got 750 buildings, this needs approx 850

[primal flicker]

My FPS seems fine.
Right up until the game freezes.

[main dirge]

Maybe iron (wire), caterium (ingots), and quartz processing at 1 and the rest at 2/3

[oblique hollow]

Head Lift, range 0 to 1.5 m

Head Lift's internal name is Pressure

and at roughly this range, its job is generating flow rate

above that, its job is to push fluid upwards

[primal flicker]

Update: it's the pipe sounds bug. Looks like I get to play with no audio for a while.

[prisma kraken]

@primal flicker 900 rubber is the recycling loop, i didn't draw that up for quickness, but should be factored in

[primal flicker]

Basically equivalent footprint/power. Interesting.

[prisma kraken]

the recycling loop just still wins, but there's probably a case to be made for the coated & insulated thing in something like a speedrun

[primal flicker]

Simplicity is its own benefit, sometimes.

[oblique hollow]

caterium cable

[delicate chasm]

I see. In that case, respectfully, I don't think it's inaccurate to say head lift isn't pressurization because it might be referred to as that internally, but it's not coded to simulate pressure; it's coded to simulate the results of pressure.

Do you agree that's an important enough distinction to avoid calling the system "pressurized"?

[oblique hollow]

Oh it definitely is pressurized because theres a volume overfill portion of pipes

about 40% above the normal volume

[delicate chasm]

That's just buffering though is it not?

[oblique hollow]

and that range, in code, also contributes to pressure generation

its not just buffering as that range is designed and defined as the pressurization range

[primal flicker]

Given that pressure can be stated as heat-feet of a standard fluid, I don't see any need for the distinction between the two.

[oblique hollow]

hence why outside of code they give us the neat head lift name

but yes, in code pressure is normalized as pressure columns

in meters

[primal flicker]

Atmospheric pressure is commonly stated as inches of Hg which is just barometric head.

[oblique hollow]

it is the difference in pressure columns between pipes that generates flow rates

which is just straight up pressure

just an incredibly small range for pressure that stops affecting flow above a certain point

[delicate chasm]

🤔 And what about splits? The splits will have a full pipe behind them and we're missing the mechanic where the fill rate of the pipe stays full while liquid passes through the segment to the end point.

i.e. the fill rate lowers when a machine chugs. Pressurized systems would remain full.

[oblique hollow]

above 1.5 m, pipes accelerate to max flow in nearly every case, but the pipe gauges are coded to average flow

[oblique hollow]

they will have a full pipe... once full. Not guaranteed to be full the instant you start filling them

[primal flicker]

Does that mean that very short segments of pipe can restrict fill/flow rates?

[oblique hollow]

theres a reason very short pipes have a minimum volume of 5 m3

the overfill portion of that is only 2 m3

and theres a reason (well, one of them) why mk 2 is so unstable

[delicate chasm]

Doesn't really address my point though, that the simulation involves removing liquid from the pipe which is then replaced in 2 steps rather than it being an ACTUALLY continuous flow.

[oblique hollow]

it has the SAME volume and pressure ranges, but twice the speed

so it can empty the overfill portion twice as fast

[oblique hollow]

Then i dont get your point

[delicate chasm]

A filled pipe is a happy pipe and works as we expect a pressurized system to.

Until you fill it, that's not true. How do we reconcile that? Until now I've just been saying we have a simulation (incomplete) that is trying to look like pressure, but under the hood is fundamentally not working the same way pressure does IRL.

[oblique hollow]

I think the problem is a semantic nitpicking of what pressurization means and thus defining what pressure is

[delicate chasm]

Yeah, not really nitpicking but investigating to make sure I am not ignorant of an entire mechanic in the simulation.

I THINK we're on the same page though?

[oblique hollow]

To us IRL, pressure has no limits. More pressure = fluid move faster

In satis, pressure has a limit for affecting flow

above a certain point, pressure only sustains the maximum flow, it doesnt increase it

And, fundamentally, none of us have experience IRL with large 1.5m diameter pipes i would wager

Also, IRL, fluid energy has a few components, like static and dynamic head

Satis only knows static head

Flowing fluid does not generate more pressure than non-moving fluid.
There is no dynamic component

[primal flicker]

Just be glad SF doesn't simulate headloss across the length and turns of a pipe. 😂

[oblique hollow]

Wells there is some headloss

[delicate chasm]

Funny you should say that... 😁

[oblique hollow]

its like.... 10%?
That think keeps 2 parallel pipes from self inducing each other

[main dirge]

🤔

[oblique hollow]

it used to be an issue in Update 4, 5 and 6 and 7

But in 8, one of my old test cases is now solved

[primal flicker]

What was the test case?

[oblique hollow]

apparently, at the cost of the shared head lift mechanic

[delicate chasm]

Was it your horizontal to vertical one started working but the same thing diagonal (Auto/Auto2D) doesn't work?

[primal flicker]

i.e. breaking water towers (which were broken to begin with)

[oblique hollow]

#1174954053650894980 message

[oblique hollow]

its the diagonal pipe thing

[delicate chasm]

Unknown channel link.

[mystic moon]

Click it

Q & H threads show up that way at first

[delicate chasm]

Ohhh, TIL. Thanks, I thought it was a role-restricted channel.

[oblique hollow]

Its not a porn site i promise Lmao

[delicate chasm]

You don't know which numbers I like or how much I like them.

[oblique hollow]

The relevant QA post is the second one

[primal flicker]

Is there pi? I'm a fan of pi...

[oblique hollow]

the major bug report

[oblique hollow]

in the volume calc for pipes

pipes are defined by radius

and their volume is literally just radius x radius x pi x length

[delicate chasm]

No way, you serious? 🤣
I just assumed they vector the entire length of the pipe and check the z-level difference between each point and the center.

But that explains the Auto2D discrepencies.

[oblique hollow]

nah they just take the spline length and multiply it with the base area

[oblique hollow]

that what you talk about is for head lift math

they use both end points and calculate a mid point

they dont actually use the actual spline mid point

they derive it just from the end points

probably to keep pipe math uniform between the build modes

[delicate chasm]

This is what I feared, and what I suspected was the reason diagonal slopes are wonky compared to horizontal or vertical segments.

[median heath]

What does the H stand for?

[wind spade]

help?

I mean it's in the name, #1038092680493801533

[median heath]

Oh.

I'm thinking Quality and Hassurance.

😂😂

I'm dumb.

[wind spade]

hi dumb

[primal flicker]

I thought it was Quodlibets and Heuristics.

[median heath]

Nice!

[main dirge]

Building so wide the pop-in distance of pipe supports is giving me issues bc I can't tell where I've built 🥲

[vapid gorge]

how so?

[main dirge]

keep thinking I've left rows of supports like on the ceiling here unfished

[vapid gorge]

oh as in you can't link them in one go, gotcha 😄

[main dirge]

getting distracted from what I'm doing and going to check and immediately realising I'm dumb 🥲

[main dirge]

gotta say the fog really adds to the pipe jungle though

[prisma kraken]

that's one sexy extractor farm

[main dirge]

I think it's funny that max uranium power is approximately pi times more water extractors than for 100% pure copper production

[plain pawn]

Pics of water making me sad

[vapid gorge]

nuclear is one of the reasons I will always use the Big Water Extractor mod

[main dirge]

like c'mon it's just perfect

[plain pawn]

its 6 water to 16 coal so that should work. Is the the fact that I am putting them all essnetially in 2 pipes the reason its not working?

[vapid gorge]

ok so the whole layout is making me cry a bit, but I don't see anything that specifically will break it - you seem to have the water coming in at multiple points

[main dirge]

now it is time to regret never fully automating supercomputers

[vapid gorge]

I'd turn off a couple generators and let the whole system flood until you check all the gens. then turn them back on

[plain pawn]

so let each gen be full then turn back on?

ok i try

[vapid gorge]

pre flooding a system is pretty important and a good way to trouble shoot.

if everything is flooded and works for a bit, and then stops working you know there's an issue

[plain pawn]

it makes my cry too btw I wanted organized but pipe placment was frusrating.

next time

[vapid gorge]

practice makes perfect 🙂

[toxic current]

guys how do i get the area of a circle

[vapid gorge]

pi r squared?

or was it 2pi r squared?

[toxic current]

no there circles HA HA HA

sorry my math teacher really like that one

[vapid gorge]

... I'm a math teacher and I don't get it sorry xD

[toxic current]

they are cirlces, pies are cirles, saying pi r sqaured is like saying pies are a sqaure... i guess

[vapid gorge]

oh thats truly awful, I love it

[toxic current]

yep lol

[vapid gorge]

I'll add that one to the list of how to torture students

[main dirge]

2 pi r is the perimeter :)

[median heath]

Area is pi r squared, yes.

pi * d is circumference.

[vapid gorge]

always something I double check 😄

[median heath]

Which is fun because circumference is literally how pi is defined.

Pi was discovered as the constant for circumference / diameter

[vapid gorge]

what do you expect from circular logic

[median heath]

Math people never heard of "define the word without using the word" 😛

[plain pawn]

ok so the pipes and the coal gens are slowly running out of water

but also not all the extractors are running

hmm i pipes on non running extractor and now they run and coal gens are staying full

[vapid gorge]

can you take a series of pics from where extractors are connected all the way up? might have to delete a few foudnations to get a clear view

[primal flicker]

Phase 3 construction mall plan:
I worked backwards from the high tier stuff to make as many machines as possible have nice numbers, and to keep resource inputs near nice numbered limits too. Priority given to lowest building numbers in most cases, unless that conflicted with convenient resource limits (like default silica requiring more than one Mk2 normal quartz node). The ratio of solid/coke steel is purely a function of limiting 960 iron.
https://www.satisfactorytools.com/production?share=7izjA5u7PvmxnKpO5fJX

[prisma kraken]

NF start?

also, taking a look at your numbers there, it looks like you're doing a lot of endgame-style builds in phase 3, you might want to scale some of those back a little and focus more on something like making a lot of MF's or motors rather than trying to max things out at 480 and needing a rebuild in phase 4

[wind spade]

Why would they need a rebuild?

[prisma kraken]

to scale up

[wind spade]

Just build new

No reason to rebuild old

[prisma kraken]

there's some builds in phase 3 that you can do with mk4 tech, and some that its better to wait for the faster speed stuff

i'm just saying that things like tapping a couple of pure quartz nodes for 480/min probably will end up needing revisited in phase 4

[wind spade]

Split after miner 🙂

[prisma kraken]

of course, but definitely worth planning that into a phase 3 build so that you can run the belts for what that mess turns into

cheap silica always ends up being sort of a really messy thing

in any event, the gist of what i was saying is that looking beyond phase 3 in terms of planning is probably a good idea

i build as little as i can during the phase and focus more on getting all the research goodies

i am kind of thinking though that perhaps getting a big motor build done during the phase isn't a horrible idea

i kind of jammed myself up in my current playthrough by not doing one that stuck around

[primal flicker]

NF start, just finished optimizing my fuel power on northern lake. (+some plastic, rubber, and a bit of packaged fuel).
Looking at tapping one of the Titan Flat quartz nodes, which is why that's soft capped at 300, and the impure oil in the canyon.
This is to start accumulating "(almost) all of the things", to support T7/8 construction, and to get some practice with construction and logistics involving a large number of processes. Once complete, I'll draw from it to build the Phase 3 SE production.

[wind spade]

I mean you can just build motors for storage

And later if you need motors for next step, you build them separately

[primal flicker]

And I already have HMF, computers, and motors automated in DD, with full ISCs.

[wind spade]

Then you don't need to plan for future

[prisma kraken]

my invented problem was that i tore down my motor production for using steel for MF's and HMF's in order to move motors from iron glenn to the NF bluff, and just sort of jammed myself up on doing a massive build with it that i'm still working on

first-world problems 🙂

[primal flicker]

I'm using a lot of steel alts so I'll probably wind up with satellite steel production and then handle the logistics based on what that looks like.

[prisma kraken]

phase 3 is tricky if you build a lot in it, very easy to be constantly needing more power

[primal flicker]

That's why I got to 40GW before starting a build materials depot.

Step 1: automate HMF, motors, and computers
Step 2: upgrade power by an order of magnitude, vs existing coal
Step 3: materials depot
Step 4: automate Phase 3 SE parts, proportional to their Phase 4 requirements

[slim matrix]

is this ok?

30 cables, and then 20 sheets and 10 wire

too much cable no?

[primal flicker]

Straight manifold works fine. No need to balance, because the machine speeds will do that for you (as long as total ore used ≥ miner capacity)

You need a lot of cable for power infrastructure. And sheets will be used heavily, for coal power and throughout later stages of the game, so make sure to have plenty of storage for those.

[slim matrix]

wasnt sure, thank you

[prisma kraken]

by the same token, it is ok to balance... for intro copper, i'll set up 2 smelters for it and 2-way split into them and then 2-way split into 4 construtors for wire just because its easier to belt

(also b/c i'll start it running before i unlock splitters and that's about the only way you can sanely splice splitters into things after-the-fact, lol

[primal flicker]

Also, chasing down every under-Mk belt and lift in a poorly laid out 280-machine power plant is a godawful nightmare, but it makes the resulting flat power graph SO satisfying!

[wind spade]

manifold will work even if ore used < miner capacity 🙂

[prisma kraken]

SCIM is very helpful with that

<=

[primal flicker]

I MEANT ≤
It's still early, for me. 🥱

[wind spade]

basically, no matter what is the ore used and miner capacity, it will work

[primal flicker]

Not tech ically true, but I've been staying up too late optimizing my fuel power build. 😆

[wind spade]

if you don't have enough materials, then it won't work at 100%, but it will still work at max possible capacity

[primal flicker]

Unless ore < demand, in which case one or more machine at the end of the manifold starves.

[wind spade]

but a balancer would starve as well, so it's not property of manifold, but property of not having enough resources

[primal flicker]

So it throws off ratios if you are doing more than one thing downstream from that manifold.

A balancer would starve each output proportionally.

And clock speed adjustments solve all of that, once unlocked.

[wind spade]

and a manifold reduces your max power usage 🤷‍♂️

but point is that efficiency is unchanged no matter if you use manifold or a balancer

[primal flicker]

I always strive to make that a non-factor

[wind spade]

I just don't like people saying "manifold will work if you have enough resources", because that makes it sound like it won't work if you don't

[slim matrix]

decent?

ill probably unlock miner mk2 soon so will have to change everything

but its actually a learning experience

[wind spade]

you don't have to "change everything", you can just put a splitter after miner and build new part of factory

[slim matrix]

oh right

would be nice if all ore nodes were snapped to 4/2m

well i guess this one would be fine if i built it in 2m

but kinda stupid that it wont allow 4m by just 0.5m

[primal flicker]

I tried to keep that in mind on my current save. Everything I built before unlocking them, I built as if I already had a Mk2 miners feeding it.
Same now, for Mk3 mining and Mk5 belts. I'll be able to return and upgrade those in place, and the machines are already on standby to receive the increase in materials.

[primal flicker]

IDK what the vertical limit is, but you can cover the node completely with foundation and still snap a miner on top of it. You'll just have some local grid alignment error to deal with.

[slim matrix]

tahts what i did as you can see on screenshot, the issue is i have it clipping at about 3.5m cause it wont allow to actually build it at 4m cause the node is not properly placed in the map

well i imagine it is not properly places cause the other nodes i tried were properly snapped to 2m or 4m

this one is at i would say 1.5m

[ornate echo]

i have 3 x 90 belts and neeed to split it into 6 machines, how

[snow dove]

split each belt into two

[ornate echo]

wait no i am wrong

i have 3 x belts giving 120

[primal flicker]

360 total? 60 per machine?

split each belt into two

[ornate echo]

i got it now, i messed up the math so many times during building

[fierce ruin]

I want to have pipes just going straight down this reactor. Should I just put a valve in front of each I/O. I'm just worried about too much liquid building up at the bottom

ah I think I've got a plan

A valve here, no?

[wind spade]

never use valves

[fierce ruin]

why?

[wind spade]

they have tons of issues and pretty much 0 uses

[fierce ruin]

ah alright. Any suggestions?

[wind spade]

yeah, prefill pipes and don't worry about exact flow, fluids will handle themselves

full pipes are happy pipes

[fierce ruin]

got it I know how to do that one 🙂

[prisma kraken]

valves do a few things... 1) they stop backflow which can actually exacerbate sloshing 2) their resolution for limiting flow is very crude & often is so inexact that it causes issues 3) they just generally complicate troubleshooting

and 4) they violate the rule keep pipe networks simple

[wind spade]

@granite swan see #math-and-meta message for alt recipe choices

[fierce ruin]

So what’s the difference between a manifold piping system and what I consider load balanced? Does anyone have a good example of a manifold plant?

[granite swan]

k thx !

[wind spade]

manifold

--+--+--+--+--+--+
  |  |  |  |  |  |

since pipes don't really work as belts, if you have a pipe junction, you can get pretty much any split, 100/0, 50/50, 60/40, etc. And that's not even mentioning that pipes are two-way, so you can also have some fluid flowing "backwards".

Pipes also kinda need to be prefilled to work properly (in most cases) and also looped so that backflow doesn't hurt your system (again, in most cases).

For these reasons, building a "balancer" for fluids is kinda pointless and goes against best pipe practices

here's example of looped pipe manifold:

--+--------------+
  |              |
  +--+--+--+--+--+
  |  |  |  |  |  |

[fierce ruin]

Thanks that was a really nice response. Are there any videos on the pipe mechanics in satisfactory. I confuse my irl understanding with satisfactory.

[pale ravine]

I don't fully understand the principle behind this.

Would this work for something like aluminum scrap where you're using and also producing water?

--+--------------+ | | +--+--+--+--+--+ | | | | | | 1 1 1 1 2 2

Say that 1's are doing alumina solution and 2s are doing aluminum scrap.
Would that work?

[wind spade]

for that it's better to have separate loops

[viral ravine]

For recycling water in alumina setup you'd do best to build a VIP system

[wind spade]

or even better - recycle the water into something else

[median heath]

Better is subjective.

[pale ravine]

Interesting, thanks

[viral ravine]

https://cdn.discordapp.com/attachments/932761153703149659/1179617352413876284/VIP.png?ex=6583a9c5&is=657134c5&hm=2db9371fb62eb377d688213105d5b6a8ede1c7c8d94be04e44386df98ccbf257&

[median heath]

Both methods equally work.

[viral ravine]

bottom pipe has recycled fluid

[wind spade]

given VIP is considered black magic even by McGalleon 😄

[median heath]

It still works.

[viral ravine]

ye it does

[wind spade]

for now

[median heath]

Yes. And if it stops working then better will no longer be subjective in this context.

[viral ravine]

no reason not to use while it still works, once it stops then it will be back to the drawing board

[pale ravine]

Thanks a lot

[vocal adder]

Hi !
I'm working on a tool that automatically find the optimal recipe combinaison to produce a chosen part. You choose the part, a production rate goal and the criterion used for optimisation and it outputs the optimal list of recipes to use and the resources needed.
Is it worth it to make it publicly available or are there other calculators with the same features? Those in the pinned messages seems a bit more limited.

[median heath]

https://www.satisfactorytools.com/

[deft lichen]

This is cool!

SFTools can't optimize for other targets than resource efficiency

[median heath]

Pretty sure greeny added other metrics..

[deft lichen]

Only in the beta which is now outdated

[primal flicker]

Being able to optimize for different qualifiers seems like it could be a unique feature

[primal flicker]

Does it find solutions that split production of a part between 2 or more recipes to reach the user-defined goals?

[deft lichen]

Also, can the tool handle loops? Like aluminum or UFR?

[median heath]

||Also the last time I asked simple questions about an optimizer it got taken extremely personally. So this time I figured I would skip all that drama and shoot straight to linking Tools.||

[primal flicker]

||How very dramatic of you... (/j)||

[wind spade]

yet

[vocal adder]

I don't see a case where this is needed, do you have an example?

[vocal adder]

Loops are the main issue I had. I'm still looking for a clean way to handle it. I added some "uber recipes" manually to avoid loops in the meantime

[wind spade]

rubber or plastic production with most resource (oil) efficient way come to mind at first, using both recycled rubber and residual rubber

and then obviously things like "limiting this raw resource to X/min", which I'm not sure you do in your tool (in Satisfactory Tools it's the "items/input" section)

[vocal adder]

Didn't think of rubber/plastic because that the main cases that I did manually because of the loops

[wind spade]

yeah, loops are usually a dealbreaker for most people when it comes to making their spreadsheet/tools. Took me a while to figure them out as well

[vocal adder]

I think my algorithm would choose other recipes rather than split production between recipes

[wind spade]

tools pretty much always pick the best option, doesn't matter if there's loops or splits or whatever. And the mentioned beta with other optimisation targets can also do building count or power efficiency. But the path to it was horrible, had to rewrite the whole solver like 5 times (and even before Tools, the old calculator I made had a different algorithm as well). Basically I had to give up classic tree traversal super early 😄

[vapid gorge]

I mean I did when I needed both base and fused QW recipes so I didn't have to import material as I was just a bit short of copper

I'll have to again soon with multiple steel recipes too

[primal flicker]

My current SFTOOLS planner has a mix of coke steel and solid steel because I wanted to limit both iron and oil usage, without using sulfur.

[hybrid star]

Am I reading the wiki correctly on this, a normal nuclear power plant on Uranium fuel rods produces 10 Uranium waste a minute? I could hook up the belt and find out for sure but I'm building towards everything turns on at once including with some delay the disposal services

I felt like it was more once upon a time because I used to have 2 belts leaving the powerplant with the waste

[primal flicker]

Yes, 10/min at 100%.
Most people seem to clock those at 250% though.

[hybrid star]

Any reason why? The power plants are built so I'm probably not going back on that without a big reason. Cheers on the confirmation, now to build a factory to sink the waste

[primal flicker]

I would guess (1) save space, (2) simplify water logistics, and (3) expensive build materials

[hybrid star]

I was away from the game for a while, this is a wildly annoying factory to build, control rods which means fucking stators, Alu shit, Steel. Guess it will be a few days before the swamp glows green.
The water seems simple enough. 300 for 6 powerplants = 1 OC water machine

My last nuclear powerplant back in update 5 or 6 or so was really veiny though so I can see that

[vapid gorge]

it's a lot simpler with the base recipes

[hybrid star]

It just a sigh here we go again thing. I built a megatower to produce all the precursers to Uranium rods and a train system to get them over to the swamp. I'll get the nitric acid stuff tonight and then figure out the rest tomorrow and later

[ashen stirrup]

It takes 8.4 seconds to burn 1 compacted coal. So does that mean to fully utilize a belt of 260 compacted coal, I need 260/8.4 coal generators? I think the conversion from seconds to minutes cancels out

[random charm]

maybe check this https://satisfactory-calculator.com/en/planners/power/index/powerOutput/2700/generator/Build_GeneratorCoal_C/fuel/Desc_CompactedCoal_C

257.14 Compacted coal can supply 36 Coal generators

important is second parameter, requirement per minute

260 / 7.14 = 36.41

[prisma kraken]

the wacky there is easier to see if you realize 50 CCoal/min exactly fuels 7 coal gens

whenever you see a repeating decimal with 142856 in it, that's our good friend 7 popping up

[gloomy palm]

Do you guys think the Sink Point Calculator I made would be useful if I brought it back again?

[ashen stirrup]

Thank you for this 🙏 you've saved me quite the inconvenience 😂

[cold crescent]

I need to split a line of 25 rods per minute into 5 lines of 5 rods per minute without smart splitters. Ive been thinking about how to do it for the past 10 minutes and I just cant figure it out.

[median heath]

Manifold.

[wind spade]

--S--S--S--S--S
  |  |  |  |  |

[arctic willow]

you can't divide by 5 directly with splitters alone; you'd have to divide it into 6 and feed the sixth output back into the input

[median heath]

Manifolding allows you to divide by 5 directly 🤷‍♂️

[prisma kraken]

manifold is simplest, but you can do the 6-way and loop back as Kinitawowi said, or clock the consumers to 125% and turn it into a 4-way split

[cold crescent]

Ill try manifold, if it doesnt work Ill say

[median heath]

They always work 👍

[cold crescent]

Ill be the judge of that

[median heath]

Nothing to judge. Over 4 years of evidence.

[cold crescent]

Im trying to ensure that all my machines never shut off unless the grid fails, thats why its so complicated for me

[median heath]

Manifolds always work. You just have to either wait for them to fully spool up, or pre-feed them to skip that step.

[cold crescent]

Im an idiot, it works like a dream

[random charm]

yep, but question was, 260ppm belt, how many generators he can feed. And 260/50 is again integer with decimal. Therefore 260/7.14=36.41, so my answer was 36 generators can be fed with 260 belt(257.14ppm)😎

[snow dove]

except you can underclock generators

[ashen stirrup]

I know I can underclock generators, but 36 is a nice number and I don't mind overfeeding each one a bit

[ashen stirrup]

I should be good to place one mk1 pump every 5 wall segments going straight up, right?

Or would there be a benefit to doing it on every 4th segment/16 m?

[deft lichen]

no, not at all

the pumps give 2m extra headlift over the recommended

so 20m is perfectly fine

[ashen stirrup]

Thanks!

[primal flicker]

I would personally make it 38 or 40 and slightly underclock some or all of them.

[rustic patio]

i love load balancing

4x600 to 3x780 and 60

[median heath]

https://tenor.com/view/no-nooo-luke-skywalker-starwars-gif-5054968

[rustic patio]

sith load balanecrs

jokes aside i didnt know how to do it better without building differently

i was too lazy to redesign my blueprints 😩

4680 steel

no idea what to do with it just yet

i hope the priority merger stuff works

[heady vine]

i'm quite sure 2 years ago i figured that its best to put some batteries per each geothermal station.
Question: any math? Cause i dont remember it was purely random or i did calculate that i need 2-3 per each?..

[deft lichen]

0.5 for impure, 1 for normal, 2 for pure

[upbeat tide]

Are the pipe bugs still present? Eg mk2 pipes unable to handle max volume properly?

[mystic moon]

That has never been a "bug" as far as I'm aware

[upbeat tide]

It was a rounding error. One of the reasons they changed nuclear reactor OC water uages

[heady vine]

buut if normal produces 100-300 and charge rate is 100 per storage isn't 2 storages per normal the answer? i don't get the math

[deft lichen]

(100+300)/2 = 200 MW average
deviation from average is 100 MW at both extremes

you can have a factory consuming 200 MW at most, charge when it's producing over 200 and discharge when under 200

[heady vine]

uh, right. i don't care about its base production - only fluctuating part.
therefore its + or - 100 tops

[prisma kraken]

I still see it now & then when i get a little fancy with mk2's at 600/min, when it pops up, i just switch to mk1's and feed from both sides of the manifold

you know, its nice that they fixed trucks to actually be able to climb ramps, but you'd think that something that hefty would have an engine that would support climbing at a decent speed, lol

[toxic current]

i just looked at how you can produce just 1 heavy modular frame per min and that is like, made me sick

[wind spade]

it's super easy compared to lategame productions

[keen adder]

in order to seperate 7 constructors to make a total of 100 rods/min, i need to underclock to 14.(285714, repeating), but that isnt a clean number, so what underclock percent do i use?

that isnt clean either shit

is there another way i can use to distribute the iron to the rods evenly?

maybe 7 is a bad number

i just dont want to underclock too much or waste a slug

can also use 8 and have 12.5 per

but then if i want to not underclock i am basically wasting a slug

[wind spade]

95.2381% 🙂

or 6x100% + 1x66.6667%

[hybrid star]

I've got a math issue and I can't pin down where the issue is, bear with me. A train is bringing in 2 cars of quickwire. The boxes they deliver into never empty so I know I don't lack for total quantity of quickwire.

They are feeding into 10 manufacturers, 9 of which eat 187.5 quickwire per minute. I have them in 2 blocks of 4 and then a half block with the one that is fully overclocked and the last manu consuming 112.5 quickwire. The total consumption then being at 1800 quickwire.

The freight boxes have 2 exits. From Box 1 I have the first exit feeding 4 manufacturers, which demand 750 per minute. And then the second exit is doing the same. So in a manifold those should create a backlog.

The thing is those machines never fully fill and eventually run out every so often. I don't get. I've even been shoving overflow from the second box into the other lines and they still fail to fill the machines.

What am I missing? I feel like its obvious but I'm scratching my head

Jesus Christ. I think I just saw it happen. The train delivering stops the outgoing wire until it finishes and there isn't enough wire on the lines to hold the supply steady. Didn't see that coming.

[wind spade]

yeah, you need a buffer

connect both outputs into an ISC and then have one belt going from the ISC to your factory

[vapid gorge]

there's a 27~ second pause when the station is in use you need the buffer to compensate for. It's why you can never get 2 full belts throughput per platform

[hybrid star]

I think I'm going to cut off the machines until the buffers fill. That should smooth it out a lot. The other option would be a third car but I like the design I have and don't want to mess with it

[vapid gorge]

how much are you trying to move on each platform?

[hybrid star]

1800, to feed the 1800 demand. Maybe some bonus for buffer on the other end. Its a ride over to check what I did

[vapid gorge]

on each platform, not total. Do you have mk5 belts?

[hybrid star]

yeah

each platform should get an evenish split of the wire. There was some messing about with smart splitters

[wind spade]

doesn't matter, it will naturally fill itself

[primal flicker]

Resolves to 20/21.
Use any convenient multiple of that.

[wheat olive]

Man I'm just having a realization that not all factories need to be based around 100% productivity. For example I have a small factory just for keeping my personal sorting containers always topped up on constructon resources. The buildings only come to life after I've taken some stacks out of the containers in order to top them up and I simply can't build fast enough for the rate of re-filling to ever be an issue. But I just realized that means I don't need 1:1 ratios for everything. Like if I have a line making copper sheets and a line making copper wire, I don't actually need enough copper production to run them both. I can only have enough copper prodution to run one at capacity and that's fine, I'll never need more. There's always enough of a buffer of resources in the smelters to run everything for long enough anyways, and then the smelters just have time to re-fill their buffers once the fabricators are done.

Which means that like a single copper mine and a few copper smelters can safely run machines that would normally need double or triple the production to keep them going consistently at 100%

[hybrid star]

Maybe its perverse but I like going back and checking my work and seeing that it works. Also right now its a nuclear power thing and it has to work when all the bits go online because I can live without toxic waste build ups

[prisma kraken]

use 8 constructors making 12.5/min

i'm not saying that to be smartypants, 7 is just a really icky number in the game that never ends up being even with anything

[keen adder]

yea fair

though, personally id rather not have to use 25/m if i want to do it at a rate of x2 because then that really eats at my slugs

[prisma kraken]

well, 8 at 12.5/min is underclocking and it'll save a mw or two

also nice number for a compact balanced split which is always for me

[rustic patio]

everyone here seems to love trucks but i tried using them for the first time just now and got a "vehicle deadlock" warning

and they refused to move. do they just keep moving after a while or am i f'ed?

this was my initial plan

how would yall move it?

thinking of just doing belts

[vapid gorge]

I'm doing belts hidden by rail in that location 😄

[rustic patio]

hmmm

i need to redo my rails anyway...

[upbeat tide]

Embrace the belt! My entire world is belt fed

That said its 2k hours old and that plan started when trains were bad

[vapid gorge]

There's rail going back and forth and I was vaguely tempted to make it functional but meh, running them underneath saved a lot of station space, and it looks good!~

[rustic patio]

waow

lots of circles

[vapid gorge]

What a shock right 😛

[rustic patio]

oh
ur username 😭

ugHHhhh im really struggling with like

on one hand i want stuff to be pretty, on the other hand i dont want to spend a billion hours trying to get stuff to not just be floating in the air

[vapid gorge]

You can do similar w/o all the circles 🙂

[upbeat tide]

My map

[rustic patio]

opposite of circles

[upbeat tide]

Yup 🙂

[vapid gorge]

Ah yes, my old nemesis, the world grid

[upbeat tide]

And nothin quite like 252 working nuclear plants

[rustic patio]

max nuclear?

[upbeat tide]

Yup and max plut just no reactors for those

[vapid gorge]

ah neat, the image I shared before was all 50.4 uranium rods made on location

well it's about 70% functional anyway

[upbeat tide]

My nuke plants are in the west but rods made in the swamp

Dont ask why I dont even know

And yes it took a eon to build that

[vapid gorge]

it's easy to make rods somewhere and burn them in another part of the world 🙂

[snow dove]

<@&387163995947270144>

[lethal fable]

ty

[vapid gorge]

@stark spire didn't want to bring it up in the post and cause confusion - but could you, in theory, have 2 stations right in front of each other, set the train to stop at both, and get teh platform to load onto a second fluid car too?

I'm not sure if there's a situation where this would be a good solution though

[stark spire]

I have had really bad things happen with more than one station controlling the same platforms. They had a build lockout on that for a while, iirc

[prisma kraken]

that land bridge is pretty tricky for vehicles, you have to keep their paths extremely far apart and its not fun to drive the path w/o spinning out, etc

i recommend if you decide to keep with the vehicle idea, that you lay some street lights along the path to partition the path in half to help you stay on one side

[rustic patio]

The issue is I have two different routes that both need to go over the bridge

[prisma kraken]

in general though, vehicles do sorta derp out whenever their paths get too close

i've tried doing that before and it didn't end well, lol

[errant ocean]

Does anyone have a balancer for 5 to 6?

[heady vine]

last time i heard "eon" was from asgardian in stargate....
anyway, how much power to power all that?

[heady vine]

doesn't staisfactory-caclualtor has bps for them?

[heady vine]

do you have lanes?

[errant ocean]

Unfortunately not

biggest they got is a 3:5

need a 5:6

[heady vine]

ok than can't it be done with simply split-all-merge-all?

[errant ocean]

im already at max loads on the belts so it wont rlly work well like that

[prisma kraken]

split one line into 2 to result in 6 lines; merge 2+1/2 together, then split into 3

if you're at max on the belts that won't work, you'll need a proper balancer

that's the ugly to make, you pretty much need to do a 6 way split on each line and then merge the 1/6th's back pairwise

think it probably can be built without clipping in a 4x4 foundation space

is it for cheap silica's limestone?

[errant ocean]

i tried figuring it out but couldn’t figure out how to do it cleanly so i think that’s a tomorrow problem

[prisma kraken]

i know how to build it, but don't really feel like spending the time on it, lol

this is a 4:3 balancer

what you'd be doing is a 6 way split instead of a 3-way on 5 lines stacked instead of 4

so you'd have 30 lines forking out of the back end

what you'd do for each column coming out is a vertical merge down like so:

its then an exercise in trying to pack it tightly together

(as i said, it's the ugly to make)

[errant ocean]

Yea i mean i was able to do it a little bit ago, it just ended up being huge and ugly so ima tackle the problem in the mornin when i am fresh

ill prolly make it into a blueprint bc ik ill never wanna do it again lmao

[prisma kraken]

yeah, well, you may also want to step back and ask if there's a better solution to your problem like an injected manifold

or changing the factory around to eliminate the need

[errant ocean]

it’s a coal plant bringing in 1350 coal to 36 gens all 250% overclocked so not much i can do to eliminate the need of this splitter i don’t think

[prisma kraken]

sort of an odd number, when you OC generators to 250, 16 of them use a full normal node's 600

[onyx nest]

bro i litterally joined the server for this and instantly find it. mind sharing a blueprint? i cant figure it out from the picture. thanks in advance

[ashen stirrup]

Have you tried looking online? For example, https://satisfactory-calculator.com/en/blueprints/index/details/id/2936/name/4%3A3+Load+Balancer

[onyx nest]

i didnt even know there was a non fandom wiki

tysm

[ashen stirrup]

!wikisearch Load+Balancers

[brisk shore]

https://satisfactory.wiki.gg/index.php?search=Load+Balancers

[ashen stirrup]

This is the non fandom wiki

The first link was SCIM, which is a save editor, interactive map, blueprint hosting site, etc.

[prisma kraken]

i don't have a blueprint for that one

[thorny cedar]

https://www.youtube.com/watch?v=QkukZSqHl4o

[prisma kraken]

this picture may help you see how to belt it better however

[ashen stirrup]

When the fellow was asking for a blueprint... 🤷‍♂️

[thorny cedar]

oh u dont want to see it u want it hand feeded into your mouth im sorry

i wont do it again

[prisma kraken]

i gotta see what the one in the vid looks like, have a feeling mine is nicer

[thorny cedar]

its without lifts yea

not as compact but you can make it its just to see the theory of splitting 4 into 3 each and merge them into 3 again

[prisma kraken]

yeah, its not that difficult to understand the premise of... you just split the x input lines into y parts where y is the number of outputs, then you merge one piece split off from every input line into the output

[thorny cedar]

yep

[prisma kraken]

they're much easier to belt if you split horizontally and merge vertically

in creating that one, i found a new technique in doing the vertical merger of multiple lines a little more compactly

[wind spade]

why are there balancers in #math-and-meta when they are not meta

[thorny cedar]

its called math u learn it later 😄

and who decides what meta is

the ppl i guess and if the ppl want balancers it becomes meta >_>

[prisma kraken]

i just answer the questions that people ask in the room they're asked in 🤷

[wind spade]

most people "want" balancers, because they are unaware of alternatives

[prisma kraken]

i tend to agree, usually they solve the wrong problem

that being said though, i enjoy belting them to figure out how to do it cleanly

[thorny cedar]

i cant agree bcs i dont know what "most people" even think so i dont know if they want balancers bcs they dont know better or they want them just because

[wind spade]

I'm basing that on large majority of people who after being explained what a manifold is said "oh wow I wish I knew that earlier"

[prisma kraken]

most people want balancers because its how you solve the problems in factorio

[wind spade]

(or something similar)

(even in factorio it's usually not 😛 )

[thorny cedar]

every factorio thing is a manifold

[prisma kraken]

well, whereever the group-think came from, it is group-think

[thorny cedar]

the grapplers grap stuff from a belt into a machine until its full

[wind spade]

until it's full it has enough to run for given amount of time

[thorny cedar]

i tend to balance in factorio aswell but most ppl dont do that at all

[wind spade]

the only balance I do is miner -> smelter

after that there's no reason to balance usually

oh, and train stations

[prisma kraken]

in SF, when i can easily construct a balanced split, i'll do so

not having to wait 5 minutes to start debugging a missed belt, etc i think is a good reason

coming off of a train, that's really the only place i get tempted to use a full-on balancer

[wind spade]

in SF? I treat trains as belts

X goes in, X goes out
no reason to balance

[prisma kraken]

ensuring that all train cars/stations empty at the same rate is desireable

[thorny cedar]

thats why i use single car trains in low quants high qualities

i dont want to deal with this stuff 😄

[prisma kraken]

i don't think that necessarily needs a balancer, however

[wind spade]

if I put 500 into first car and 300 into second car, I get those amounts at the other end and hook them to two manifolds that eat 500 and 300 respectively

[prisma kraken]

there really is a quantum leap with building scale though when you start pulling a lot of product into a factory and it doesn't fit in a single train car or on a neat number of belts anymore

[wind spade]

one miner = one car

(or two miners = one car if impure)

[prisma kraken]

well, that doesn't work for pure nodes

[wind spade]

why not?

[thorny cedar]

and u want to unload them onto 2 belts with the same destribution to set the train to wait until empty

so u balance the belts

[prisma kraken]

the round trip needs to be less than ~4.5 minutes

[wind spade]

why?

I add more trains if one is not enough

[wind spade]

I don't. Since each belt eats exactly as much as I'm putting on it, it will unload naturally the way I want

[prisma kraken]

3200/780 = ~4.1

[wind spade]

(uh, was supposed to be reply to "balance the belts")

[prisma kraken]

and yeah, you can double up the trains

which i do in some places, in others i split the miner between two train cars, etc

where it gets a little more tricky is with plastic/rubber or aluminum goods that kind of want to be sized in a yield of 450 or 900

[wind spade]

I mean then put 450 in a car 🤷‍♂️

[prisma kraken]

you absolutely can split 900 between two cars of 450/min, but for a 200-stack item, you can easily ship that in a single car

[thorny cedar]

so u balance the imput then 🤷‍♂️

even if it means you clock the machines to get 450 onto a belt its still balanced

[wind spade]

no, I put 450 in a car (meaning I gather 450 from machines that make 450)

[thorny cedar]

yea its still balanced you know

[prisma kraken]

you're essentially putting the balancer on the sending side

[wind spade]

it's not balanced, it's at most direct input

balance = taking X inputs and producing Y equal outputs (where both X and Y is bigger than one)

[prisma kraken]

that is a balancer

[wind spade]

5:1 is not a balnacer, but a merger

[prisma kraken]

900->2x450, i count 1 input and 2 outputs

[wind spade]

what is the 900?

that's not on one belt

so it's not one input

[thorny cedar]

no u use 450 into lets say 2 lanes it the same balanced as 1 belt with 900 items split in 2

its a balance

you can argue its "direct fed" but you still balanced it beforhand

[wind spade]

two lines of 450 are two separate things.
if it was a balancer, it'd have to be balanced between the two lines

[thorny cedar]

you deliberate set up the machines to be 2 belts to balance them yea

[prisma kraken]

well, in the case of oil, you're actually doing the balancing while its still oil in the structure of the factory

[thorny cedar]

or multiple not 2 its just an example

[prisma kraken]

i think we kind of need to codify what is considered a balancer

[wind spade]

the (pretty commonly used) definition of "balancing" is
"taking X inputs, produce Y equal outputs, no matter what the inputs in X are"

so if I applied it to two 450 belts, the balancer would have to produce two 450 belts if both are full, and two 225 belts if one is full and one empty (which is not the case, since they are separate)

[thorny cedar]

so you put x oil (no matter how much its manifolded as long as it is enough) into the machine and out comes 450 per belt thats per your definition a balancer

[prisma kraken]

we are arguing semantics

i'm happy if we accept that some balancing is being done implicitly somewhere

[wind spade]

(amount of inputs and outputs can be whatever, but at least number of outputs needs to be bigger than one)

and this ^ needs to happen immediately and at all time, so manifold is not a balancer (even though it eventually behaves as one)

[thorny cedar]

looks like a balncer

[wind spade]

you can't balance pipes, it's a manifold and that isn't a balancer (as per my message above)

[prisma kraken]

there's a thin like you're treading on there with pipes always being a manifold

[heady vine]

why?

[wind spade]

because miners run out 🙂

[thorny cedar]

is a balancer aswell

[wind spade]

it's not if it relies on overflow mechanics (with oil)

[thorny cedar]

no there is more

[wind spade]

if the thing after A is 1:5 belt balancer, then it's indeed a balancer

[thorny cedar]

bcs if u compare oil to plastic there is a multiplier but that does not matter

[wind spade]

pipes can never be balanced, so if we're talking balancers, we are talking solids only

[thorny cedar]

bcs 3*1/5 is oil input multiplied to plastic output

[wind spade]

that's irrelevant

if A is oil, then it's not a balancer, because fluids by definition can't be balanced

if A is solid, then it needs 1:5 balancer to reach the amounts you've posted

[thorny cedar]

a i oil b is plastic still a balancer

[wind spade]

how do you balance fluids

[thorny cedar]

the input is a manifold

[wind spade]

so... not a balancer

[thorny cedar]

why would you balance it you balance it behind

[heady vine]

i fear to ask why you need to balace plastic

[thorny cedar]

its just the topic

[heady vine]

so not a real use case?

[wind spade]

no, they are trying to somehow convince me that manifold is balancing or idk

[heady vine]

manifold is a balance of sanity and time investment
||screw that comment, i tried to smooth the argument by finding common ground but one of the sides is not behaving||

[supple belfry]

Don’t get caught up on semantics

[thorny cedar]

i dont try to convice you i just show u the facts and hope your brain is as big as your ego

[wind spade]

why do you have to result to personal insults when I'm just trying to understand your point

not sure how converting oil to plastic after being manifolded is in any way related to balancing

and having 60 somewhere and 60 somewhere else without any direct connection is not "two balanced lanes"

[thorny cedar]

you dont try to understand my point you try to argue into ridiculousness

just the sheer fact of the assumpion a balancer is made out of splitters and mergers only is not correct

[heady vine]

i think the argument would benefit from revising the definition of balancing and balancer

[wind spade]

never said that

balancer = a contraption that takes X inputs and produces Y equal outputs, no matter the state of inputs (they can be full, empty, partially empty, etc.)
it needs to work immediately (within reason) and at all times, not rely on overflow mechanic (then it's a manifold, not a balancer)

[thorny cedar]

thats one part of balancer yea

there are ratio balancers and total balancers

they exist in satisfactorie

i can balance a 60 item output belt even if the input varies between 60-780

and i can make a ratio split out of the variation of 60-780

[wind spade]

ratio balancer is basically a normal balancer where you just merge the outputs to get the ratio

(can be sometimes simplified ofc, but I don't think that distinction is necessary for definition of balanacer)

[wind spade]

wdym "balance 60 item output"?

[thorny cedar]

you can make a balancer so the delta output of one belt is always a total number not a ratio

[wind spade]

that's not a balancer, that's "rate limiter"

(or something else)

[thorny cedar]

.

[wind spade]

if output is always 60 no matter the input, that doesn't fall into the definition of balancer

and we may be arguing semantics, but if you use a term that's pretty clearly defined (especially in the context of factorio, where this started), then I'm following that definition

[thorny cedar]

so this is no "balancer" this is an overflow than

its a mixed thing i would say

[wind spade]

well you'd most likely achieve this with one or two smart splitters with overflow setting, so yeah, that's overflow

[thorny cedar]

i mean its just because the number of belt is 60 if the output would be 45 you still ratio balance it

aswell

[ashen stirrup]

What's the math on what fraction of input is sent to the nth output of a one-sided splitter manifold?

[wind spade]

before filling 1/2^n
after filling it's the amount that machine needs

[ashen stirrup]

Yeah, I get that after filling

I did the napkin math yesterday while playing with a friend, and it was fun to say that we pre-fill because without doing that, the last coal gen gets one coal every 500 years

[heady vine]

isn't it just simple splitter with 2 mk1 belts?..

[wind spade]

the number and how you achieve it is irrelevant imo
what matters is the behavior

if you get X amount and split into Y outputs, if any output is different from X/Y, then it's not a balancer

[thorny cedar]

thats why i clarified its just an example with the numbers who happen to fit an mk1 belt

[heady vine]

wont agree because we are in the channel where naming matters.

[thorny cedar]

wont agree i dont know the language that well that it would matter dont be full of yourself

and its not the topic at all what started it

but we can continou in my mother language so we can argue about semantics or not

[prisma kraken]

just playing devil's advocate, balancers can be used to feed unequal outputs, but draw at equal rates from inputs... I don't think that fits your definition of a balancer, yet it is an inherent property of them

[wind spade]

unequal outputs? either it relies on overflow (and then it's not a balancer), or it's a ratio balancer (which I clarified above that it's achieved by merging some of the outputs together and isn't too important for the actual definition of a balancer)

for example, if you want to split 150 into 100 and 50, you'd do a 1:3 balancer and then merge two outputs

(the whole thing can be called 1:2 ratio balancer)

[prisma kraken]

lets say i have 4 lines of 600 and i want it split into 190, 260 and 150, a balancer will do that

yes, it will overflow on the 150, then 190 and then distribute to the 260

[heady vine]

than the whole conversation just doesn't matter.

[prisma kraken]

but it will do so will all the input lines never waivering in their 600/min speed

[wind spade]

if it overflows, then it's a manifold, not a balancer

[prisma kraken]

so this is a manifold because one side backs up?#math-and-meta message

[wind spade]

yeah

[prisma kraken]

so when is something not a manifold?

[wind spade]

when it doesn't rely on overflow mechanics (waiting for one or more belts to back up to start distributing in desired manner)

[prisma kraken]

so lets say we just have a belt splitting 3 ways

into 3 machines making rod

even split...

[heady vine]

btw, can a manifold implemente with splitter be counted as manifold? considering that every splitter is not olbiged to fill its production building before passing to the next in line

[prisma kraken]

is that a manifold?

[wind spade]

no, because it doesn't need any belt to be backed up to work in desired way

(assuming all the machines are clocked evenly)

[prisma kraken]

now what if i change the clock rate on one machine to 150 and another to 50

[wind spade]

then it's a manifold

[prisma kraken]

is it a manifold, acknowledging that the topology is the same

so topological classification can be different for the same topology?

[wind spade]

because now the desired split is 3/2/1, instead of 1/1/1

but it doesn't achieve that until belts are full

if you'd replace the splitter with 3:2:1 ratio splitting thingy, then it would be called a balancer

[prisma kraken]

ok, a thing is not a thing if it has a specific property

[wind spade]

well, I guess it also depends how you want to look at it

the splitter does 1:1:1, so it's technically a balancer
however what the factory needs is not a 1:1:1 balancer

so I guess it depends if you look at the contraption in vacuum (what it does), or if you look at what the contraption should do (in context of factory)

a 5 machine manifold can be called 1:1:2:4:8 ratio balancer, but that's not what the factory needs 🤔

[prisma kraken]

i mean, i get what you're saying, but perhaps we're both learning that these definitions are a bit more elusive than you'd think

[wind spade]

I'm looking at it from the POV of your factory - if three machines need things in 3:2:1 ratio, then you either build a manifold (one or more splitters in a row), or a balancer (1:6 balancer, where you merge some outputs; or equivalent contraption)

manifold will rely on belts/machines filling up, balancer won't

[prisma kraken]

even the word manifold is a tricky one, lol

[wind spade]

and I think that ^ is how pretty much everyone is looking at it as well

[prisma kraken]

i do think there's a nuance that your missing though

let's say we take something simple of a 3:2 split/balance/contraption

you can belt that up as a split of one line that merges into the other two

and the lines will all be balanced, but that isn't a balancer

the reason why is that if one line backs up, one of your inputs will stop

[wind spade]

3:2 meaning 3 inputs to 2 outputs?

[prisma kraken]

with a true balancer, all inputs will keep getting consumed at a lower rate

[prisma kraken]

yes

a true balancer, all inputs feed all outputs

[wind spade]

so, if it's a "balancer", then all of this must be true:

• if only one input is getting resources, both outputs have half of that
• if all three inputs are getting resources (equally or unequally), both outputs have half of sum
• both of the above must work immediatelly, without any belt having to back up
• if one output is backed up, all resources go to other output

if it's a split of one line merged into other two, it isn't a balancer, because it doesn't do any of the above (except when all inputs are getting stuff equally)

[heady vine]

*immediate is a big vague, but certainly without any interaction with whatever the balancer feeds into

[wind spade]

immediately = within reason (belt speed and feedback loops on prime splitters)

[prisma kraken]

quibble: 'half the sum' -> 1/nth the sum

[wind spade]

well I'm talking about the 3:2

[prisma kraken]

i'll agree with it

i do think there's something very specific topologically that defines a balancer vs manifold and it doesn't have to do with overflow vs no overflow (or consuming rates), but more with the any-to-any relationship of the inputs & outpus

[wind spade]

any to any is a required property for a balancer, yeah

arguably less important in satisfactory given that our inputs don't change over time (in most cases), but still part of the definition

[prisma kraken]

i mean, this is actually something i recently learned in working through some of the balancing problems that i've been tackling

what i'll also assert is that where balancers do have a use is not so much in equalizing the output rate but on equalizing the input rate

i.e. where i'd use them is to make sure multiple cars from a train are draining at the same rate

i've run into that problem before, and opted to just toss more trains at the problem instead of building a balancer

[heady vine]

topologially is too abstract for the game. in game we have concrete implementations of certain tools with their own properties. but we can mimic properties of concepts, resulting in different outcomes.
and in the context of the game, there's little need to balance because there's limit in machines' storage and belt feedback (limit), so manifolds can do good job.

[prisma kraken]

well, we're well out of the realm of what you practically need to do to solve the game's problems and into more the theory side of things, but its an interesting topic

what i'm actually trying to figure out is what makes a manifold a manifold, and kind of the venn diagram of balance and manifold and splitter

like if they all do the same thing, there's a circle you can draw around them and say that they're all widgets that move stuff from n inputs to m ouputputs

but are they disjoint/super/sub/equivalent sets of each other?

[heady vine]

the crucial difference i see is in how they behave over time, when belt feedback kicks in

balancer will simply clog and property that "each free output receives same amount of items" will remain, so output to remaining items will change but remain even.
while in manifold you will get a much less noticable change, that is uneven across all.

there's really any point to use balancer in place where you expect to get belt feedback?

[rustic patio]

it works 🎉

four way remote controlled hypertube intersection

idea stolen from here (i think, forgor where i first saw it):
https://youtu.be/WIrn5mvMepA?si=62JkpXeYhC5XZew0

oh wow hes even on this server

[tired viper]

I've 33.333 Fuel generators working, plus 2 more with 3 powershard at 375, plus all the coal factories, yet I'm only recieving 5.6KW Energy. Is there some kind of cap?

[stuck ingot]

i would say that the fuel powerplant isnt connected with your world grid

[true junco]

5.6KW? do any of the generators produce that low a wattage? Did you mean GW?

Assuming you meant 5.6GW...

"Is there some kind of cap?"

No. The game doesnt have a power production cap except that you eventually runnout of resources. Max power is somewhere over 1TW (1,000,000MW)

33.3333 Fuelgens should give you 4,999.995 MW

That plus the 2x375MW is 5749.995MW.

so it looks like its almost all accounted for. If you meant 5.6GW, then you are only short about 150MW of what you should have according to what you say you have built.

[tired viper]

Thing is we also had the old Coal Factories connected, so we we're lie 2.4GW short. We reseted the server and it works, don't know what caused it.

[true junco]

Multiplayer or Dedicated server? If so, then that was the cause. MP/DS for SF are "experimental" on their own.

[vagrant holly]

So i have a factory that makes 1440 steel ingots per minute. I am currently doing tiers 3 and 4 of science what should i create with said steel in your opinion

[swift robin]

just shove it into a ton of contructs making pipes and beams 🙂

[wind spade]

wait until you actually need it, don't make stuff that you don't need yet

(in the same way you probably shouldn't make random amount of steel ingots before knowing how much you need)

[swift robin]

well the factory already exists so too late for that

[vagrant holly]

I had a plan

But i overshot the math a bit

[wind spade]

keep it for the future then

[swift robin]

first you need more beams than pipes, later you will probably need more pipes than beams

[vagrant holly]

yeah or just make a lot of HMFs

[wind spade]

no reason to randomly make beams or pipes or anything, if you don't know how much you need

make as much as you need now and leave rest for the future

[swift robin]

if you just split it to go to a buttload of constructors making pipes and beams you can let your downstream demand allocate the steel for you i guess

[vagrant holly]

I have a plan!

its better to overproduce than to underproduce things

[wind spade]

it's better to produce as much as you need

[vagrant holly]

call it future proofing hahaha

[wind spade]

future proofing is almost impossible and better to not do

[prisma kraken]

you're a little on the low side with a 1 TW figure; nuclear alone with burning plut rods is ~1.1-1.2 tw, and that's not looking at burning oil or coal

(i will concede though that its just not practical to go that high, you run out of resources to do stuff with that power pretty quickly)

[shadow ember]

Does anyone here have an idea how i can make three 180 belts (540 total) into two 270 belts? when i only have mk3 belts?

[wind spade]

split one and merge each into one of the main

[wind spade]

or just use the 180 belts

[oblique hollow]

180 = 90 x 2
270 = 90 x 3

[prisma kraken]

yeah, 180/2 = 90... 90 + 180 = 270

[prisma kraken]

speaking of our 3:2 problem, lol

[median heath]

420 = 90 x 4.66666666666666666666666

[ashen stirrup]

A splitter for the "middle" belt that goes into two branches, merge each branch with a "side" belt.

[lime wadi]

Making a lil tool that tells you the max extraction of a particular biome based on what tech you've unlocked so far. Going to use this to do an analysis of every alternate recipe, at all stages of the game and at every location.

My reasoning for doing so: The WP formula on the wiki does not take specific location or unlocked milestones into any consideration, lending a lot of bias to alt recipes that are best suited for endgame.

[lime wadi]

Making fuel?

[proven fossil]

The issue is the crude outputs at 675 a min

and i only have t2 pipes, so is there a cheaper way to make it?

[lime wadi]

If you're willing to add water, you could try the Diluted Packaged Fuel or Packaged Fuel alt recipies as well as the Heavy Oil Residue.

You'll have to use the Packaged version before Tier 7

[rancid sapphire]

blud only haves t2 pipes ur trolling

[proven fossil]

Oh you can use packaged water with refineries

Is there any way to do it without water though?

[rancid sapphire]

oh thats cool

i did not know that

[lime wadi]

Yeah its a great alt if youre willing to tackle the challenge of dealing with packagers and adding water