#math-and-meta

1 messages · Page 558 of 1

median thunder
#

you need to go back and build a module frame factory, and a factory for every item in the tree

wind spade
#

no, you always start from ore

#

you build a separate factory that produces it's own modular frames

lapis bronze
#

i keep commiting suicide because i fall in holes between 4 splitters

median thunder
#

you always start with ore

wind spade
#

yeah

median thunder
#

doesn't that get.... repetative?

wind spade
#

not as much as building 150 constructors because I want to overbuild

median thunder
#

like wow just built a module frame factory, time to build a reenforced module factory.... which means building the entire module frame factory again

wind spade
median thunder
#

if you plan that in advanced

thorn bane
#

i think the problem is the belts connecting the manifolds, you dont need those because thats your bus
belt limitation can be a problem for some things but only 10% of your belts have issues unless youre overproducing alot
but even then my using overflow splitters having multiple belts bet item is not that bad
its in 1 place per factory so if you change something you need to come back to that factory which is alot of walking time
i mean my belt is 6 foundations wide i wouldnt call that alot of wlagking time

but ye i do agree that its using more belts its just easier to place them imo because they are all parallel

median thunder
#

the real meme tho is building constructors manually in satisfactory

#

no matter what construction style you like

#

in a game about automation you think you could be automate the construction of manifolds as well

dusky quail
#

blueprints have been requested for a long time

thorn bane
#

you can do that but its hard to get to 100% resource efficiency that way
which is fine for some i guess but i just follow a planner that i made in advance or have a spreadsheet of all my production buildings

median thunder
#

how do you calculate efficiency, though, like

thorn bane
#

if 100% of the input gets used its 100% efficient

wind spade
median thunder
#

100% of the input is always used. Otherwise it just sits on belts

thorn bane
median thunder
#

there are two resources in satisfactory that matter

thorn bane
median thunder
#

your time, and items per second

thorn bane
wind spade
# median thunder if you plan that in advanced

well yeah, which you can always do, but even if you don't, it's not a big deal to build it again, you can also make each factory look different and none of them is really large, because the production is pretty small and contained

median thunder
dusky quail
wind spade
thorn bane
wind spade
median thunder
#

cause you don't need it? You aren't spending the electricity to get it out of the ground

thorn bane
#

ok maybe ore is a bad example
say youre make 100 iron ingots but only use 20 for iron plates

median thunder
#

then you overbuilt smelters

#

but how much time did you waste overbuilding the smelters?

thorn bane
#

yes exactly so its not 100% efficient
thats what i mean

lapis bronze
#

heresy

thorn bane
#

ok then whats your idea of 100% eefficiency

median thunder
#

the least amount of time needed

thorn bane
#

hm i see

median thunder
#

it's roughly the same amount of work to build a 50 X factory as a 25 X factory

thorn bane
#

thats why i said resource efficency
not time efficiency

median thunder
#

cause design, logistics, whatever, it just scales up

lapis bronze
#

strats for raising efficiency develop over time

  • scaling up or down
  • underclocking
  • overclocking
  • mk.2 resource harvesters
  • etc
median thunder
#

resource inefficiency is if you use stupid alt recipes

dusky quail
#

even though i underclock myself, generally advise against it tbh

lapis bronze
#

scaling is OP, you just find the GCF of all your inputs and outputs

wind spade
#

that's why I do separate factories. I don't have to upgrade anything, because they still work and produce and new stuff can be built with better tech

thorn bane
median thunder
#

wait why do I even play satisfactory

lapis bronze
#

i usually go back and fix my early game stuff, like get rid of doubled up miners and replace it with a single mk.2, etc

median thunder
#

hey, have trains gotten better in satisfactory yet?

wind spade
dusky quail
#

you can just pull whatever you need from existing factories, not sure why it would be considered abandoned

thorn bane
dusky quail
#

distance is the same, except with a main bus you are literally pulling everything

wind spade
thorn bane
#

i think it just scales differently
the more complex it is the more belts you need for a stand alone factory
where for a bus the cost is constant its just the bus+pullign stuff down

wind spade
#

and for new parts you have to expand all belts

#

the bus is exactly where it's not constant, but increases with bus length and size

#

and for the standalone factory I already said I need pretty much same amount of belts

median thunder
#

belts in satisfactory also kinda suck which makes main buses much more pain

lapis bronze
#

can i plant resources into machines to make manifolds work faster

median thunder
#

like if you have a set of stacked belts and you want to move all the resources to the top set of belts

thorn bane
dusky quail
median thunder
#

it factorio it takes zero extra space to do so, in satisfactory the splitters and mergers are absolutely massive

wind spade
wind spade
thorn bane
#

yes so its more work early but less work later

median thunder
thorn bane
#

im just saying that even with them beeing alot worse they are still the best imo just not by alot

wind spade
median thunder
#

yea actually I'm changing my opinion

#

for satisfactory, on-demand is probably the most efficient (in terms of items per sec over dev time)

thorn bane
#

if you get to HMFs / computers its the same work
but if you go all the way to the space elevator its less work

wind spade
lime scarab
#

yea

wind spade
#

and put left belt priority to all

median thunder
#

not my screenshot that's just the first one I found on google
but yea ur right

lime scarab
#

lol

wind spade
#

yeah just suggesting 😉

median thunder
#

ofc :)

dusky quail
#

both have pros and cons tbh, main bus takes a lot more pre-setup

wind spade
#

my main issue with main bus is that it kinda forces you to have centralised setup which is neither recommended nor appealing to me

thorn bane
#

i think alot of people just see the early work and decide a bus is bad
where if you have all your stuff on the bus and youre making complicated things liketurbomotors the benefits just outweigh the initial cost of the bus

dusky quail
#

I've done the work and gotten to turbos and concluded i would never do it again

lime scarab
#

I prefer the initial time consumption of the bus design at the start than the work later

wind spade
#

idk I can't imagine making turbomotors on bus, I'd have to add machines pretty much to all manifolds and the only advantage is that I'm using the "bus", which I need to use just because the manifolds are far away from themselves in the first place

median thunder
#

it's easy to build stuff like ammo and those handheld ore diggers on the bus

#

though, actually, there's a term for a bus dedicated for human-use consumables - a mall

thorn bane
wind spade
dusky quail
#

the length is one of the biggest issues I have with a bus

thorn bane
#

well you can plan ahead when youre making those

wind spade
thorn bane
#

and even if you dont adding a machine to a manifold is easier than making a new manifold because you need to decide where to place it

wind spade
#

not like it's hard to find a buildable space in SF world

thorn bane
#

compared to having a existing manifold and just adding 1 machine?

#

especially with smart mod?

lime scarab
#

Comes down to Time management and planning

wind spade
#

let's keep modding away from this disucssion

#

I could make an argument that I have a mod that produces 10 turbomotors/s from thin air

thorn bane
#

ok even then adding an extra machien to a manifold takes no time
imo you spend most of the time in satisfactory worring about logistics and maybe 5% building machines

#

actually lemme find out

wind spade
#

if it's just adding one machine, then I also add just one machine, so there's no difference

thorn bane
#

but its not just 1 machine
its the 2 belts connecting it and the spot where you place it

#

(i mean in a decision sence not raw materials)

dusky quail
#

"let me just add a few more machines.... and have to pull my bus of 30+ belts again...."

wind spade
#

ok, so I build two extra belts that connect it to the module that's literally next to it. I agree that it's a bit more belts.

Now let's compare the other point - adding new manifold

  • I still build just a few short belts (one for each resource)
  • you need to build belts for ALL resources (increasing bus length)

so here you build more belts than I do. In the end it just evens out

thorn bane
#

i mean in the decision aspect
you need to decide where you place it and how you place your belts

wind spade
#

you don't need to decide where you place your bus?

#

you also have to move all resources to a single spot, while I can just place the factory next to the nodes

dusky quail
#

main bus is a lot easier if you can copy and pasta though, but by hand it's honestly insanity at least for me

thorn bane
wind spade
#

anyway, I'm willing to bet that both systems use very similar amount of buildings and belts, in later stages my system will use less belts

thorn bane
#

less belts but thats not the time consuming factor

#

you spend most of your time figuring out where to place those belts
which you dont with a bus

wind spade
#

how do I need to figure out where to place belts?

#

it's literally connecting two manifolds

#

I just need to find a place to build (which is easy, just plop it near the nodes)

thorn bane
#

you need to figure out where to place the manifolds (and how much space they take)

wind spade
#

production tool already shows a pretty good schema on where they are

#

in relation to each other

#

and I know how much space they take since I know amount of buildings in them

thorn bane
#

imo those things take more time that just pulling it from a bus
but i guess it all depends on how often you have done one or the other

median thunder
#

the game is not manifold simulator

#

the game is design and logistics

wind spade
#

also fyi I'm not trying to force you to stop using belts, I'm just trying to explain you that both systems are pretty much equal in time/building/cost complexity, so I wouldn't say that a bus is better in any way (and we both agreed that it's worse in terms of fps and that you don't have separated stuff [which is both recommended by devs to combat low fps and also useful if you want e.g. to disconnect one factory])

#

I just wouldn't recommend main bus (or any bus) for satisfactory playthrough unless you really know what you're doing and what it brings

#

and even then I think there are better ways to build (but each player can do whatever they want, that's the beauty of the game)

median thunder
#

satisfactory: not good because you build a main bus, but good because you can decide not to build a main bus

thorn bane
#

for me the advantage of having everything in 1 place (dont need to go around to gather stuff/build a storage room) outweights the fps loss but ofc. thats personay preference and i just like building small
but even then imo for complex parts like the last space elevator parts/turbomotors etc. a bus system just takes way less time to build than a factory
i guess i could also be that i just prever a smoother time investment/return curve but thats hard to say

#

btw amelie of the sea made a tutorial where she kinda does both by having a bus in each dedicated factory
for example in the video she made a HMFs factory using a bus
https://youtu.be/I4oysebwPz4?t=10574

YouTube
Amelie of the Sea
Introducing the Tutorial Save File

0:00 The Discord Server has a list of topics for tutorials
0:55 Description of the Tutorial Save File
1:42 Oil Tripling Builds
45:12 Introduction to Bauxite Refinement
54:12 Water Extractor Placement Techniques
55:06 Alumina Solution and Aluminum Scrap Refinery Placement/Clockspeeds
57:59 The Most Common Mistake in Pipe Logistics
01:07:29 Petrol...

▶ Play video
wind spade
thorn bane
#

i just have a container next to the bus of each material beeing made and thats good enough for me

wind spade
#

so it's running around the bus gathering ingredients instead of having them all in one place? 🤔

thorn bane
#

well that is in one place

wind spade
#

bus can be hundreds of foundations long, I don't really count that as "being in one place"

thorn bane
#

its 136 foundations
ofc a storage room would be less but thats good enough for me

wind spade
#

again, I'm just saying that for me it's not "in one place" and I would hate running 130 foundations back and forward to search where is all the stuff I need

wintry fern
#

Whats a reasonable amount of each part to be producing a minute for space elevator 3?
Im looking at the materials required and it's just kind of shut my brain off. I figure it's just a matter of how long I want to wait but I'm dying without direction

#

On second thought I actually dont know how much of anything to produce I've just been kind of winging it up till this point but the game has suddenly become daunting

oblique hollow
#

3rd delivers? so the one that unlocks tier 7 and 8?

wintry fern
#

Yeah the one that's like 2500 versatile framework etc

oblique hollow
#

ive gotten by with 4 assemblers for the frameworks

#

because while they were doing stuff, i was out automating oil

#

i still have to make more than half the engines

#

which is mostly just the rubber that i need

#

so you can very much do it with just 1 manufacturer for the engines and adaptive units and like 2 or 4 for the frameworks, IF you have stuff to do

thorn bane
#

i did 15 frameworks/min 2.5 engines/min 1 ACU/min

#

but i usually overclock them and throw the remaining stuff i have left over in there

#

for example my saved up motors and smart platings

#

and flexible framework/plastic smart platings help alot if you ahve them

feral dew
#

a year has passed since I was last in here, and people are still arguing Manifold v. Bus? I thought we already reached consensus that manifolds are best.

magic shadow
#

oh no here we go again

wind spade
#

there's a debate between manifolds and balancers
there's a separate debate between using main bus or not using main bus

frosty owl
#

Sneeze

thorn bane
#

bless you

wind spade
#

oh no, he must have COVID! runs away

frosty owl
#

More like covideeznutz!

oblique hollow
#

run its the sushi belt guy

frosty owl
muted crypt
dark summit
#

Sushi belts honestly are the best, just for storage have overflow go to a sink

wind spade
#

I personally prefer not to mix items on belt at all

muted crypt
#

you don't play the game

wind spade
#

with exception of having one "sort my crap" container

wind spade
muted crypt
#

fair

wind spade
#

(and what I'll have in my playthrough when I finally start)

versed forge
#

What are the pros/ cons of using belts, trucks and trains for transport?

wind spade
# versed forge What are the pros/ cons of using belts, trucks and trains for transport?

belts:
pros:

  • simple
  • doesn't require power
  • available super early
    cons:
  • is pretty costly for long distances
  • is limited to "one belt", if you want to double throughput, you have to build the whole thing again
  • stackable supports are pain to build
  • long mk4/5 belts lose some of their max throughput, so maxing a long mk4/5 belts isn't recommended

vehicles:
pros:

  • doesn't require any infrastructure on the path (if there's a nice natural road)
  • cheap
    cons:
  • needs fuel
  • is very limited in throughput (one belt per station)
  • vehicles are not reliable (can bump into stuff/each other; they also behave differently if you're near or far)
  • recording paths is pain
    (note that U5 will "fix" a lot about vehicles, so some of these cons will be voided by that update)

trains:
pros:

  • tracks also carry electricity
  • track can support multiple trains
  • super high throughput
  • can have multiple freight cars so you can transport multiple item types
  • can even transport fluid directly
    cons:
  • power hungry
  • unlocked a bit later than what would be ideal
  • are sometimes buggy when multiple trains are in same station
tiny narwhal
#

is there a way on the calc to dissallow overclocking/underclocking

#

mainly underclocking

magic shadow
#

greeny's calc?

tiny narwhal
#

uhh

#

satisfactorytools

is there another calc?

magic shadow
#

that's greeny's
if you need a specific amount, use splitters

tiny narwhal
#

how do you split something into .5

magic shadow
#

if it's not the exact amount, the belts will back up and make things equal eventually

magic shadow
tiny narwhal
#

ok

#

can i split 5 into 3 and 2 somehow

magic shadow
tiny narwhal
#

i guess

#

i was just hoping i could figure out how to load balance and all that

#

and no, if you ask if i know what that means, i dont

wind spade
wind spade
tiny narwhal
#

manifolds?

wind spade
tiny narwhal
#

oh

#

ive never thought of doing that tbh

faint ember
faint ember
#

High-speed input, low-speed consumption gives itself over to input manifolds

#

Low-speed outputs work well with output manifolds

#

The image is compacted coal, which is 1,1:1 ratio

#

So it's literally impossible to provide input slower or faster than the output, so long as you're working with nodes that have no other outputs

#

At the top level, 780 coal/min plus 780 sulfur/min = 780 compacted coal/min, but the mechanism of the factory is really clear in the image.

#

A header for each of the two resources bisecting a twin-row of Assemblers, with a splitter for each pair of Assemblers and Conveyor Lifts to provide space for the second resource

#

Mergers combine the output into a single belt

bleak coral
wind spade
#

indeed

muted crypt
#

another thing to note with vehicles and trains is that they're buggy af to set up when not the host (at least for right now)

odd rose
#

Yoo just a question about train How do I know exactly how many resource one train are going to carry on one travel, is there a formula or something ?

wind spade
#

yes, but you need to know trip time

flat plaza
#

hello you awesome people

can you remind me how many coal gens can a water extractor deliver at 100% eff?

wind spade
#

2.6666 repeating

flat plaza
#

ty good sir

flat plaza
#

and how many slugs does it takes to tweak that number to 3?

#

or 3.2~ ish?

trim abyss
lapis bronze
flat plaza
#

you guys are bloody wizards!

faint silo
#

who hurt you?

muted crypt
#

yes

oblique hollow
#

you heard of maximizing now get ready for Maximining

quaint rampart
#

rock and stone, brother

frosty owl
muted crypt
#

trying to decide if I'm wasting my time idling to hope for cast screws in an effort to simplify my production and cut down on power consumption in a pre-coal base

#

just working on hw in the meantime so it's not a total waste of my time

#

also sometimes alts don't make sense for when they're offered... I think I was offered an alloy recipe when I have yet to finish t2 milestones (need the foundry from t3 to make the alloy)

bleak coral
#

Yeah there's a lot of insufficient prerequisites for alt recipes. I think they only started checking for the machine with U4 (so for blenders only basically), and haven't fixed the others that need that too.

muted crypt
#

It's making it more difficult for me to fish for cast screws lol

faint ember
#

Just get more hard drives

muted crypt
#

Asking now while it's fresh in my mind: do we have any sort of data on general spawn points for lizard doggos? If not, should I try to place beacons down for myself so I can try to collect that? (i.e. wherever I find a doggo, place a beacon)

south linden
#

hey what is the website / discord for calculate everything in satisfactory

muted crypt
#

https://u4.satisfactorytools.com/production - production calculator and flowchart
https://satisfactory-calculator.com/en/interactive-map - very useful map (also has production calcs above, but I'd recommend the above link for it)

Satisfactory Tools

A collection of powerful tools for planning and building the perfect base. Calculate your production or consumption, browse items, buildings, and schematics and share your builds with others!

[SCIM] Satisfactory - Calculator

Satisfactory helper to calculate your production needs. | Gaming Tool/Wiki/Database to empower the players.

south linden
#

thanks men

hollow juniper
#

Do you guys know what is the cheatest way to turn uranium waste into plutonium is?

went on tool and with alternate recipies its ALOt and VERY EXPENSIVE

i just want 2520/m waste gone as cheap as possible

anyone know what recipies i need for it?

muted crypt
#

iirc best thing to do is all uranium alts with all plutonium bases

#

could be wrong, may want to verify that with someone else

hollow juniper
#

i dont want max plutonium though its going in the bin

muted crypt
#

why not max plutonium in this case anyway? that's more points.

#

either way... lemme pull up the calculator

#

Do you have all alts available / should I assume all alts?

hollow juniper
#

The amount of items thatll be going into the factory

doing max clear uranium and this is where i build my final factorys off of

So i dont have the infrastructure for max plutonium

#

im going to be kick starting my factory with power storage ffs lmao

bleak coral
#

You're right, it's all base plutonium recipes, all the plutonium alts give you more rods but need more resources

muted crypt
#

ah, so all base plut recipes then

hollow juniper
#

thats instant cells and fuel unit right?

bleak coral
#

Yeah if you're just sinking don't use those

hollow juniper
#

12.6

guys in questions said i should have half that

bleak coral
#

Yeah I think he's got some math mixed up 12.6 is correct.

hollow juniper
#

aaaaaaaaaaaaaaaaaaaaaaaaa

unknown.png
muted crypt
hollow juniper
#

this doesnt look fun

gloomy palm
#

i wrote this quick script to say how many mk2 pipes working at 600 m3/m flow rate would be needed in order to cover an X amount of coal generators all fully overclocked without there being any water underproduced or overproduced. The least amount of coal generators needed for this would be 375 running from 57 mk2 pipes. And that leaves a water remainder of 0

unknown.png
hollow juniper
#

working at max flow rate never ends well but gl

muted crypt
#

working at max flow rate usually has some rounding issues that come into play, and can cause some problems

gloomy palm
#

or 114 mk1 pipes

gloomy palm
#

what flow rate would you be comfortable with?

hollow juniper
#

idk, i just know pipes are buggy and when its for power its bad thing to have being buggy

muted crypt
#

I don't think I fill mk2 pipes past 450/min flow rate

#

I try to avoid it where possible.

gloomy palm
#

76 pipes each at 450 would be needed

muted crypt
#

exceptions being, as an example, having a 250% OC oil extractor

#

on a pure node

hollow juniper
bleak coral
#

oh you're still on maximize, maximize doesn't solve for efficiency, it only solves for most then stops as soon as it finds it

#

take the number and switch back to items/min

bleak mauve
#

is there an satisfactory player with quite some experience here who can help me out with a problem? i have one cole 240 source going in 8 coal generators each consuming 30 coal per minute.

#

but the last two generators are running low on coeal

muted crypt
#

I assume you're overclocking the generators

bleak mauve
#

yeah

hollow juniper
muted crypt
#

follow along this and see if it helps you out at all

#

it's a general manifolding flow chart I made

hollow juniper
#

maybe it is cheaper and easier it just scary to look oat

bleak coral
#

it is max nuclear, so it's not an insubstantial project

hollow juniper
bleak coral
bleak mauve
#

how can i split this of 120?

unknown.png
muted crypt
#

That's just indicating how much is going through the splitter per minute

#

at that specific step - the very first splitter - all of the incoming resources go through it, i.e. 120/min

bleak mauve
#

do splitters have a limit?

muted crypt
#

their belt limits, yes

bleak mauve
#

ok

muted crypt
#

it balances itself out to split 15/15/90 because the two machines connected to it are full, and thus only pull off what they need (by making room on the input belt)

bleak coral
#

but what are you setting the overclocks to?

bleak mauve
#

one question why does it not work when i put all the generators in one row?

bleak mauve
#

coal

muted crypt
bleak mauve
#

can i call you and show u in a stream

muted crypt
#

246.2289% clock speed, if I did my math right, Lund

bleak coral
#

are they starving all the time? or just drop every once in a while? maybe your clock speed number isn't precise enough

#

cause the game will happily round 30.1 to 30 in the UI

oblique hollow
#

what even did you set them to? 200%?

bleak mauve
#

246.2666%

oblique hollow
#

ah a classic

muted crypt
#

that's why

bleak mauve
#

thats my overclock

oblique hollow
#

exactly enough for double

bleak coral
oblique hollow
#

are the belts the issue or the pipes

muted crypt
#

I got pinged for something that immediately disappeared

bleak mauve
#

ah ok how can i find out the perfect number i need

gloomy palm
muted crypt
oblique hollow
#

reverse engineering the formula

muted crypt
bleak coral
oblique hollow
#

30 = 15* (Clock Speed/100)^(1/1.3)
solve for clock speed xd

bleak coral
#

though the 2x amount is what maroon said, also it's on a couple places on that page, so probably don't need to do math

bleak mauve
#

where is the copy settings button in generators i cant find it

bleak coral
#

ctrl+c

oblique hollow
#

there is none. press ctrl + c

bleak mauve
#

ok

bleak coral
#

then ctrl+v to paste

gloomy palm
#

@oblique hollow hate to ping you outside of a conversation topic, but now that you are online I thought you might be able to give your feedback for some idea I have been drafting just now, i also wrote it out earlier, but what is your thoughts on this script snippet:

oblique hollow
#

570 is a reasonable pipe limit imo.

gloomy palm
muted crypt
#

I'm curious why you're doing 375 coal generators to begin with tbh

gloomy palm
#

it's the algorithm which found that number

#

it exactly divides 60 pipes at 570 worth of water

oblique hollow
#

its kind of insanity to do that much coal xd

gloomy palm
#

yeahhh

bleak mauve
#

@muted crypt how did u calculate this?

bleak coral
# bleak mauve ok

do note 246.2289% is slightly rounded up and gives a bit more than 2x (2.0000001082858x), so every few hours the last two might drop out for a few seconds

bleak mauve
#

should i turn them down to 29?

bleak coral
#

you could just do 246.2288% instead so it's slightly less

hollow juniper
bleak mauve
muted crypt
# bleak mauve <@!267340999498399764> how did u calculate this? 
C = desired clock speed
30 = desired consumption
15 = consumption at 100% clock speed

30 = 15 * (C/100)^(1/1.3)
2 = (C/100)^(1/1.3)
2root(1/1.3) = C/100
2.4622888266898325689987861383355 = C/100
246.22888266898325689987861383355 = C
C = 246.2288 (approximation)
bleak coral
bleak mauve
muted crypt
#

reducing it down to 246.2288 instead of bumping up to 246.2289 prevents it from ever having brief periods of time with no coal in the machine, causing sputtering

muted crypt
bleak mauve
bleak coral
#

I also linked the wiki page which has the formulas on it, so you don't have to that much math

oblique hollow
#

isnt .23 accurate enough

#

246.23

bleak coral
#

no?

#

we get 4 decimal places in clock speeds

bleak mauve
muted crypt
#

That rounds it up even further, and thus will cause more sputtering

#

if you wanted to only use 2 decimal places, going down to 246.22 would be better

bleak coral
bleak mauve
#

thx

oblique hollow
#

2.0000069

#

for .23

bleak coral
#

why do you want it to only have 2 decimal places?

oblique hollow
#

because

bleak mauve
#

Clock speed Coal burn time Coal per minute Energy per coal Generator capacity
10% 23.51s 2.55 300 MJ 12.76 MW 17%
100% (Default) 4 s 15 300 MJ 75 MW 100%
246.2288% 2 s 30 300 MJ 150 MW 200%
250% 1.98s 30.35 300 MJ 151.76 MW 202.35%

#

Are these the Casual numbers everyone uses?

bleak coral
#

mostly people just don't overclock generators

muted crypt
#

^

oblique hollow
#

besides 100%, no

bleak coral
#

it's annoying, I'll underclock for fuel cause I make a target fuel number, but it'd be totally valid and probably better to target a generator count instead

bleak mauve
#

ok thx for help

muted crypt
#

always happy to help

dull bolt
#

Neon green, approved terminal.

#

Oh stuff happen... 👀

gloomy palm
#

i've made a dynamic algorithm which is able to list many possible combinations

unknown.png
muted crypt
#

at least set it in the specific range of a mk1 or mk2 pipe lol

gloomy palm
gloomy palm
muted crypt
#

much better

#

a 2400/min pipe sounds insane tbh

gloomy palm
#

i should be able to make the pipe values in decimal

#

the game lets you have water flow to the decimal point or?

muted crypt
#

I'd check but the save I'm on doesn't have pipes atm

#

but I do believe so... to the tenths place, maybe

gloomy palm
oblique hollow
gloomy palm
oblique hollow
#

yes but thats annoying

gloomy palm
oblique hollow
#

test it . i never did that because why the hell would I

gloomy palm
oblique hollow
#

decimal fluid flow gets practically lost inside the pipes anyway

#

you cant feed something 0.1 alone

gloomy palm
#

ahha

oblique hollow
#

unless your pipe is infinitely small

#

if its any size above 1 it takes forever to fill

#

the lowest usage we currently have by standard is 4.5/min

muted crypt
#

(turbofuel^)

gloomy palm
#

6 of 387 m3/m pipes for 25 coal generators at 202.4% (250% on UI)

#

exact water amount

muted crypt
#

250% clock speed is 2.023520921448593135255414729663x consumption/production, rounding it up by about five thousandths might be a bit much for your calculations if you want perfectly precise

#

generators are very weird in their calculations, and annoying as hell to work with

#

which is why people don't generally adjust clock speeds

gloomy palm
#

my script has this in for calculating water usage at overclock
45 + (45 * 202.4%)

#

lemme see what happens when i plug that huge float in it

muted crypt
#

it should just be 45 * 202.4% if you must use that number

#

as the water per generator

gloomy palm
#

so that means my calculations have been incorrect all this time

muted crypt
#

yes lol

oblique hollow
#

your previous one was basically 45 x ( 1 + 2,024)

gloomy palm
muted crypt
#

then use that

gloomy palm
#

what's a better scenario, that you have more pipes than generators, or less pipes than generators

#

cuz i suppose if you have less pipes, you'd use splitters

oblique hollow
#

doesnt matter really

#

theres no benefit to either of these scenarios

gloomy palm
#

x) right

oblique hollow
#

its a useless thing to think about

gloomy palm
#

well the problem i'm facing is that because of the high precision, the math simply will never reach exactly 0 on water consumption

#

unless we are talking about millions of generators maybe

oblique hollow
#

well theres also the fluid bug so your assumptions are fudged to begin with

gloomy palm
#

xD

oblique hollow
#

what even is this all for

gloomy palm
#

someone was discussing the efficiency of water extractors to coal gen ratio in the satisfactory channel

#

so i thought there should be a way to calculate it with a script

oblique hollow
#

wdym efficiency

#

power loss due to water extractors?

gloomy palm
#

i guess just finding which amount of water flow rate could cover what amount of generators exactly when the gens are overclocked

oblique hollow
#

45 covers one at 100%. there you go problem solved

gloomy palm
#

xD

#

yeah but overclocked coal gens

#

9 of 506 m3/m pipes for 50 coal generators at 202.4% (250% on UI)

#

with the fixed equation

oblique hollow
#

in the most general sense, 1 water extractor covers less than 1,317 generators

gloomy palm
#

hmm thinking_helmet

oblique hollow
#

talking in terms of full pipes is useless

muted crypt
#

Aight gamers I'm leaving for work

oblique hollow
#

pipes are the intermediary. its extractors that count

muted crypt
#

Be back in an hour lmao

oblique hollow
#

have fun at wörk

muted crypt
#

I sit there and answer the phone, it's boring

#

It's why I do satisfactory calculations at my desk lol

oblique hollow
#

i might have that job too soon xd

gloomy palm
#

increasing the precision increases the amounts as i predicted

oblique hollow
#

thats why i said less than 1,317

gloomy palm
#

increasing precision to 5 decimal points (2.02352)

unknown.png
#

i think there's not even enough coal on the map for that lol

#

so essentially it's impossible to get the exact precision of a fully overclocked coal generator in terms of exact water usage for several reasons

oblique hollow
#

1: oc gen bad

gloomy palm
# 
for pipe_flow_dyn in range(570,450,-1):
  pipe_amt = 1
  gen_amt = 1
  n = 0
  while pipe_amt < 100000 and gen_amt < 100000:
    n = (pipe_flow_dyn * pipe_amt) - ((45 * pres) * gen_amt)
    if n > 0: gen_amt += 1
    if n < 0: pipe_amt += 1
    if n == 0: 
      print(f"{pipe_amt} of {pipe_flow_dyn} m3/m pipes for "
            f"{gen_amt} coal generators at {pres*100}% (250% on UI)");break```
#

that's the code btw (tested on python 3.7.8)

bleak coral
#

once you change the clock speed it's impossible to perfectly match the input/output of any generator, you can only get "close enough", which usually means overfeeding slightly

gloomy palm
#

i haven't even begun trying to calculate the coal usage

bleak coral
#

the clock speed formula results in infinite, non-repeating decimals, so the true precision can never be represented by a computer, that's the fundamental issue

gloomy palm
#

but i imagine it's the same problem because of the precision of the multiplier

gloomy palm
bleak coral
#

no, that's just math

#

and the limitations of representing numbers

gloomy palm
#

no i mean, a bug that you have a very long floating point on a game object's properties which are user-facing

#

you want to make it so that they can actually build the system and know that the needs are exact

bleak coral
#

I mean that's like most floating points, it's not hard to get irrational numbers

gloomy palm
#

but why not make it scale to 250% exactly as it says on the interface?

bleak coral
#

and even a lot of rational numbers can't be represented completely precisely, floating points just work off of "close enough"

bleak coral
gloomy palm
bleak coral
#

I have a post on it already

gloomy palm
bleak coral
#

I wouldn't even mind if they capped clock speeds to 200%, and each power shard gave 33.3333% instead of 50%, I just hate the clock speed doesn't equal operation speed thing

gloomy palm
#

yeah i read both posts and all the comments, i definitely agree this is an issue, and one that shouldn't take 2+ years for the devs to have put into consideration

#

it seems like it was done on purpose for some sort of realism sake

#

the power shards are compounding the power boost effect but not amplifying it perfectly

#

as they are mysterious after all

oblique hollow
#

makes no sense since nuclear plants cap out at exactly double at 250%

gloomy palm
#

nuclear might have a different curve because maybe the power shards are affected by radiation, as the game's lore might have us believe

bleak coral
#

I'm pretty sure for non-nukes it's cause they wanted some downside but didn't want to make them less efficient or more efficient when changing clock speeds, and for nukes it's that plus it can't go more than 200% anyway

gloomy palm
#

hmmm thinking_helmet

#

so it was more to reinforce the point that nukes is the ultimate power producer?

bleak coral
#

how did you get there?

gloomy palm
#

get where

bleak coral
#

like that's not what I said at all

gloomy palm
#

they wanted some down-side to non-nukes you said

bleak coral
#

some down side to overclocking

gloomy palm
#

down side to overclocking anything?

#

including nukes

bleak coral
#

yeah

gloomy palm
#

ahh thinking_helmet

bleak coral
#

like overclocking gens is technically less efficient than overclocking other stuff when you consider per shard, you get less work per shard

wind spade
#

on the other hand the only "disadvantage" is that you don't use full power of the shard, but it doesn't have any real effect on fuel efficiency or something

bleak coral
#

for other stuff the downside is they get less power efficient, but since efficiency doesn't change for power gens, the less efficient shard use is it's downside

gloomy palm
#

so overclocking machinery is all linear except for power buildings?

bleak coral
#

yup

#

I mean the power usage on them isn't, but their production rate is

wind spade
#

well overclocking machine means that it has linear speed, but non-linear power consumption

#

I type so slow

bleak coral
#

but the non-linear part doesn't affect your logistics like it does for gens, which is why it's annoying on gens but not on production machines

gloomy palm
#

but in the case of power buildings, it's not just their power output which is non-linear, it's also their resource usage which makes for these problems we discussed

bleak coral
#

exactly

wind spade
#

I agree it's a bit annoying and I'd probably suggest to change fuel efficiency instead of shard efficiency

gloomy palm
#

\o/ i understand now

#

🥳

wind spade
#

(and probably put an exception that underclocking a gen doesn't mean you get more power out of same amount of fuel)

bleak coral
#

like there'd definitely be no reason to do it then

wind spade
#

it would be the same as in production machines though

bleak coral
#

it wouldn't, cause it would affect resource efficiency, changing clock speed doesn't change resource efficiency

gloomy palm
#

the resources consumed should be scaling at the same rate as the power generation no?

wind spade
#

it technically does, if you consider amount of fuel needed to produce X

#

what I mean for example:

  • 100% gen uses 10 fuel/min to produce 100MW
  • 200% gen uses 20 fuel/min to produce 195MW
bleak coral
#

yeah I don't like that

wind spade
#

it's the same as production machines though.

  • 100% machine uses 10 MW and produces 1 cycle/min
  • 200% machine uses 25 MW and produces 2 cycles/min
gloomy palm
wind spade
#

I'm not saying I like it either, but imo it's better than what we have currently

gloomy palm
#

so not only do power shards become less useful at higher overclocks, you actually lose resources for not a lot of benefit

bleak coral
#

as a different solution to what we've got now

wind spade
#

yeah it's my suggestion to what I would do if I was making it

#

I get it isn't ideal but imo it's better than what we have currently

gloomy palm
#

hmm thinking_helmet

bleak coral
#

I'd rather they just change the UI, I'm fine with shards being less useful in generators, I just don't like it being unintuitive

gloomy palm
#

yeah

bleak coral
#

hell if they keep the formulas and fix the UI to reflect what's actually happening that'd be halfway decent

wind spade
#

well my point is that shards in gens don't really have any disadvantage

#

which I don't like much

gloomy palm
bleak coral
#

at least you'd know what's going on

gloomy palm
#

yeah xD

wind spade
#

246.2288% ftw

muted crypt
bleak coral
#

also would be neat

muted crypt
#

So it shows at any given clock speed how much the typical power input/output is adjusted

wind spade
#

well, it was just a suggestion, I think it would be better that way and would be much closer to how overclocking regular machines works, but everybody can have their own opinions 🤷‍♂️

muted crypt
#

I admittedly haven't been watching closely

#

I just look down at red lights

gloomy palm
#

is the red curve how it is now

unknown.png
#

sorta

bleak coral
#

that doesn't look right, no I'm pretty sure that's not right

gloomy palm
#

i made it by hand

muted crypt
#

Brb

gloomy palm
#

m i c r o s o f t

p a i n t

muted crypt
#

more like Microsoft pain

gloomy palm
#

badum tss

wind spade
gloomy palm
#

that looks not very much as exponential as i thought it would be

wind spade
#

x axis is clock speed, y axis is power multiplier

#

(x is the horizontal one)

gloomy palm
#

so it's more like this

unknown.png
wind spade
#

well not really as it should be more flat

gloomy palm
#

it's a subtle curve thinking_helmet

#

yet so detrimental to the math

muted crypt
#

What's the number used for nuclear power

#

1/what

#

For the exponent

bleak coral
#

1.321928

muted crypt
#

ty

bleak coral
#

had to look it up on the wiki, I definitely don't memorize that one lol

sand epoch
#

i find it odd that you knew that so fast..

bleak coral
#

I had the page open

#

copy-paste

#

I don't memorize the number, but I do remember where on the page it is lol

wind spade
#

here you have the two combined

unknown.png
gloomy palm
wind spade
#

x^(1/1.3)

bleak coral
#

^1/x is the same as root of x

sand epoch
#

greeny.. you going to ever join us in actually playing the game.. ? lol (seen it posted many times that you don't actually play)..

wind spade
#

after U5 hits stable, I plan to do a playthrough (and stream it as well for those interested)

gloomy palm
bleak coral
#

there, changed it so the carat doesn't meld into the 1

wind spade
bleak coral
#

or just you expect to be done with tools 2.0 by then?

gloomy palm
wind spade
# bleak coral the good-good changes for vehicles tempt you back?

it's more that I actually want to play, but I want to finish my tools first. I could play even now (as I don't put ALL my free time to tools), but I kinda figured that I may as well wait for U5 (I made that decision back when we didn't know that U5 won't change recipes, but I may as well just go with it to have the complete U5 experience with map changes and stuff)

gloomy palm
#

same

wind spade
# gloomy palm you has a youtube channel?

I do have youtube channel but I'll probably be streaming on twitch instead. I may pud VODs on youtube, but I doubt I'll find the time and energy needed to edit it as a proper video (and also learn how to do that lol)

gloomy palm
#

ahhh, i'm a video editor ^-^

#

if you need i could chop chop 🎞️

wind spade
#

well that's a nice offer, but tbh I'm not even sure if it will last or not 🤷‍♂️ I may decide to abandon streaming/recording it altogether or even stop playing 😄 it's also at least two months for now, most likely U5 will happen around that so we can talk about it then

gloomy palm
#

neat!

#

i have the ability to make videos for gaming but my issue is im too shy to talk on camera

#

👀 and my voice sounds bad

wind spade
#

(also since I have like no followers [around 50 iirc] and I don't know how many will actually watch the content, I probably won't make money out of it, which means that it probably won't have too much money involved in it, or none at all)

gloomy palm
#

hmmmm

#

i have a channel with more than 50 subscribers 👀

#

maybe what i could do is make a multiplayer game

#

and then you record/stream to my channel as a platform with more viewers, and you do all the talking 🤣

wind spade
#

we'll see. I will most likely stream to my twitch and see if it brings new people in (as most of my viewers are from back when I was streaming Dota, so I have no idea how the numbers will be for SF). Also I don't really do it for the money, so I won't try hard to get more viewers or twitch partnership or anything. I'll just stream for fun and see if something comes out of it

gloomy palm
#

yeah no i don't have the channel for money either

#

i was working as a freelance editor before the pandemic

gloomy palm
#

my issue is i can't even start a satisfactory series on my own because i dont like talking on camera 🤣 how does one overcome this problem

#

ironic that i type so much in chats but can't talk much irl thinking_helmet

versed violet
gloomy palm
#

or at least talking and narrating

versed violet
#

If you are spotlight impaired, its ok to just hear the voice.

gloomy palm
#

yeah 😭

muted crypt
gloomy palm
wind spade
# muted crypt what is this graph

straight one: 250% = 2.5times boost (e.g. how speed changes for production buildings with OC)
curved one: how generators work (horizontal axis is clock speed, vertical axis is production multiplier)

muted crypt
#

oh I see

#

I'm just fucking around in desmos rn, red = power consumption (standard machines), blue = power production (generators)

unknown.png
wind spade
#

can add that as well

unknown.png
#

red = straight one

muted crypt
#

nuclear is so close to the blue line here that I didn't even bother including it

gloomy palm
bleak coral
#

oh yeah overclocking production buildings burns a lot of power

gloomy palm
#

😮

muted crypt
#

power consumed is doubled by a production machine at 154.221% clock speed

gloomy palm
#

D:

muted crypt
#

at that same clock speed, a generator would only be producing 1.3955x the power

bleak coral
#

that's why for power tripping tests I just overclock something, usually a manufacturer or assembler

muted crypt
#

you hit 0.5x power produced for a generator at 40.613% clock speed, but a normal machine at that clock speed only consumes 0.2365x as much compared to normal

#

gotta love math

gloomy palm
#

(not mine)

muted crypt
#

(not a meme channel)

gloomy palm
#

(sorry)

muted crypt
#

(:spraybottle:)

gloomy palm
#

(sobbing)

gloomy palm
# muted crypt

does this mean there is a sweet spot where instead of overclocking a generator to maximum, building a second generator with both being default or underclocked would be more resource efficient and produce more power-to-resource ratio?

bleak coral
#

oh overclocking production buildings is always inefficient

gloomy palm
#

oh

bleak coral
#

and gens are always the same efficiency

gloomy palm
#

so better to leave them at 100%?

bleak coral
#

for power/resource efficiency, yeah

gloomy palm
#

fascinating thinking_helmet

bleak coral
#

shards are only ever about space

gloomy palm
muted crypt
#

production building power consumption is more and more efficient the lower the clock speed gets, at the cost of sanity build time and processing resources on your computer... and resource consumption is always the same rate as power production in generators

muted crypt
gloomy palm
#

yeahhhh

muted crypt
#

I don't typically overclock, I underclock the excess

gloomy palm
#

i made everything 250%

bleak coral
#

yeah it's about FPS or covenience, or just saving time by building less, we have virtually limitless space, especially with vertical building

gloomy palm
#

and then realized i had to dial some of them back

bleak coral
#

space is partly not an issue because we have infinite resources, so you also don't care about spending resources on building more stuff

gloomy palm
#

but if we can build less stuff it would be better

#

overall

faint ember
#

Become desensitized.

gloomy palm
versed violet
#

There is no requirment to appear of even talk on video.
Example: Been playing some MMO with built-in audio chat for years, and only spoke to party in audio once

muted crypt
#

taking two machines at 100% and one at 50% and replacing them with one machine at 250% provides the same resource output, but consumes approximately 1.86x the power (if you want a more exact number for this multiplier, 1.8593922817370671883015108822395)

#

so overall if you're willing to put 86% more power into your production at the cost of making it smaller

#

then overclock.... otherwise, don't

gloomy palm
#

_<

bleak coral
#

There is definitely room to overclock some in the energy potential of the map, but you can't do it for everything

wind spade
gloomy palm
#

is it possible that you can calculate the point where a machine consumes the best power to resource ratio because of the non-linear power usage?

muted crypt
wind spade
gloomy palm
#

isn't that like dividing the power used by the input resources

wind spade
#

ah like that

muted crypt
# 
250% x 2 machines = 8.6643105394393299110281481439884x power consumption
100% x 5 machines = 5x power consumption
#

make of that what you will

#

approximately 1.7329x power increase to cut out 60% of machines here

gloomy palm
#

could i request a graphermaker to make a graph

#

actually maybe i can figure it out on my own thinking_helmet

#

so im not bothering peeps

muted crypt
#

I have nothing better to do

gloomy palm
muted crypt
#

it'd just be [base amount] * [clock speed / 100]

#

(100, 20) is five constructors at 100% clock speed, drawing 20 MW total
(250, 34.657) is two constructors at 250% clock speed, drawing 34.657 MW total

#

Both have the same resource output

wind spade
#

the thing is that graph that does [power required] / [materials needed] is useless

gloomy palm
#

😮

#

i was hoping to see something like this and say, yeah this machine has the best resource to power-to-power-usage ratio at 100% for example

unknown.png
#

where the lines intersect

muted crypt
#

and there's going to be an intersect at 100% at all times if you plugged in the equation correctly

gloomy palm
#

oh lol

wind spade
#

because it's essentialy (x * p) / (x * m), where x is number of machines, p is power required by one machine and m is amount of materials needed for one machine. We can simplify that to just p / m and since m is constant, it results in just p, which is clock speed graph 😄

gloomy palm
#

technically speaking shouldn't the power always be higher than the resources line, because then it means the machine is using less power per resource while underclocked than at 100%

#

like where the curve is underneath the linear line

muted crypt
#

trying to calculate it as a per-resource thing is not helpful because of how many recipes there are

#

a manufacturer is always going to use the same power at any given clock speed regardless of what recipe it uses

gloomy palm
#

😮

muted crypt
#

think of it like this:

#

A constructor in the early game making 20 iron plates per minute will use 30 iron ingots per minute, and consume 4 MW as it runs

gloomy palm
#

yuh

muted crypt
#

a constructor in the early game making 15 iron rods per minute will use 15 iron ingots per minute... but will still consume 4 MW as it runs

#

(I only specify early-game because you won't have clock adjustments when you first start)

gloomy palm
#

ooo

muted crypt
#

A constructor making 50 cast screws per minute will use 12.5 iron ingots per minute, but will still consume 4 MW as it runs

#

a constructor making 260 steel screws per minute will use 5 steel beams per minute, but will still consume 4 MW as it runs

#

are you noticing the pattern? lol

gloomy palm
#

yeahhhh

muted crypt
#

calculating it out as power per resource or whatever won't do anything for you

gloomy palm
#

😔

muted crypt
#

"he a little confused, but he got the spirit"

gloomy palm
#

😁

muted crypt
#

lol

gloomy palm
#

thank you for the explanations tho!

#

i appreciate

muted crypt
#

mhm

gloomy palm
#

^-^

muted crypt
#

think I might draw up a flowchart diagram for myself for my planned production lines in this T1/T2 factory

gloomy palm
#

i know a cool production planner for that!

muted crypt
#

I prefer to draw out diagrams in draw.io, and I already have my plan set on greeny's tool

gloomy palm
#

yeah that

muted crypt
#

the link I gave is my plan

gloomy palm
#

yeahh im looking at it now

muted crypt
#

it includes all of these nodes

unknown.png
gloomy palm
#

you need 47.36667 constructors

muted crypt
#

here, in the rocky desert

unknown.png
gloomy palm
#

i tried to plan my factory in such a way that the items were all produced in the same place, so that i could change the workflow whenever i wanted to build a different item, i simply dragged the items i needed into manufacturers or assemblers and then it was all central, i can't see a benefit in processing ores at the site of the miner for this reason

muted crypt
gloomy palm
muted crypt
#

so I actually need 50 constructors 😉

gloomy palm
#

🥺

#

lots

muted crypt
#

it uses up all of the resources available in that little spot

#

and it's producing a bit more than I usually would produce, so

#

significantly more concrete, that's the nicest part

#

normally I start in the northern forest and I'd only be doing 15/min... now here I'll be making 100/min

gloomy palm
#

ahha

muted crypt
#

I'm also making a somewhat permanent setup for smart plating - another thing I don't usually do

gloomy palm
#

do you need it later?

muted crypt
#

it's used for the first Space Elevator unlock (50) and the second Space Elevator unlock (500) directly, then it's used as a component to make Modular Engines for the third Space Elevator unlock (500 MEs -> 1000 SPs), and then the MEs are used as a component to make Thermal Propulsion Rockets in the fourth Space Elevator Unlock (1000 TPRs -> 2500 MEs -> 5000 SPs)

#

so... yes 😉

#

would need 6550 smart plating total, or ~2.73 ISCs full.

gloomy palm
#

;_;

muted crypt
#

Though the fourth Space Elevator package is entirely optional, as it unlocks the ||golden coffee cup||.

gloomy palm
#

😮

muted crypt
#

But I can't not go for it.

#

I've already planned out how I want my six main subfactories to work, i.e. what they produce

#

they'll be from raw to final product, so I won't be sending things between the subfactories - they go from resource pickup to processing to storage

#

I don't think I'm missing anything important here...

unknown.png
gloomy palm
#

woa

wind spade
#

regarding the graph about optimal clock speed, I think interesting point to look at is "power savings per building".

Basically:

  • figure out total power needed at 100%
  • figure out total power needed at [current clock speed]%
  • divide difference of those power requirements by number of buildings

the result is a graph that basically says "how much power will each building save me compared to just building everything at 100%"

and it looks like this

unknown.png
#

the sweet point seems to be at around 46%

#

it's basically a ratio of [effort]/[savings]

muted crypt
#

huh

gloomy palm
#

fascinating

gloomy palm
wind spade
#

it's not better power saving, but it's a decent tradeoff between amount of work needed and total power savings

#

but it's very artificially generated and doesn't represent any actual value, it's just a nice graph

#

since you can't put a value on "amount of work"

gloomy palm
#

oh

wind spade
#

in the end, you can't really use math to determine what's "easiest" way to save power

#

since it's very subjective and depends on what the player prefers

gloomy palm
#

ahh like i was gonna ask about your calculator for example, if there was a switch to bias the algorithm to find the best combination of buildings that take the least physical space, or perhaps the least power usage, or perhaps the least buildings, or perhaps a combination of these

wind spade
#

combination is not really possible

gloomy palm
#

there would be priorities yeah

wind spade
#

but optimisation towards power usage instead of resource efficiency is planned

gloomy palm
#

ahhhh

wind spade
#

optimisation towards space may happen too

#

but combination is not really possible

gloomy palm
#

it would be like having to find tradeoffs for the best power usage in order to bring down the buildings amount

#

the best power usage has 20 buildings, but the 2nd best power usage has 16 buildings

#

maybe player wouldn't mind 2nd best if it meant less buildings

#

^-^ just throwing ideas out there, i once wrote a stamp calculator that was a similar travelling salesman problem, but there were way less variables to consider

wind spade
#

there's no "2nd best" though

#

since it's all decimal numbers

gloomy palm
#

this was in the stamp calculator code, i was 18 and new to programming

unknown.png
wind spade
#

2nd best would be (for example) 0.00....00001 buildings less of the space efficient recipe and 0.00....00001 buildings more of the resource efficient recipe

gloomy palm
#

looking back, this program was a disaster

wind spade
#

that thing just screams for optimisation

gloomy palm
#

yeahp

wind spade
#

repeated code is bad code

gloomy palm
#

yeahp, except i haven't touched java in years

#

i'd rewrite it in python

muted crypt
#

there, that was simple enough.. this is how my limestone is set up currently on my save

unknown.png
#

since it's just T1/T2, it only makes concrete

gloomy palm
#

i should sleep soon, 1am

muted crypt
#

nice and simple

gloomy palm
#

welp

#

gn o/

#

thanks for everyone's help!

muted crypt
#

gn o/

gloomy palm
#

If you had an arbitrary power generator that had enough resources fed into only 50% of the time, and on the same grid, an array of battery backups that took over for the other 50% of the duty cycle, wouldn't that mean you could double your resource efficiency by only running power 50% of the time and running on battery the other 50% (granted there was enough power storage to cover the length of half the duty cycle the generators were off for and assuming the batteries were able to charge fast enough to do this)?

#

I might begin testing this theory today unless someone knows that it's not possiblethinking_helmet

wind spade
gloomy palm
wind spade
#

yeah, so the generators would use more fuel because they would charge the batteries as well

gloomy palm
#

so you'd need double the generator capacity to the battery capacity

wind spade
#

since both generators and batteries are 100% power efficient, it doesn't matter at all if you run a gen at 50% with batteries or at 100%

#

as there's no loss or gain of energy

gloomy palm
#

so what you saying is that if you used 50% of resources one half of the time, you'd be using 200% of resources the next half of the time in order to charge the batteries too

#

which averages to 100% either way

wind spade
#

not sure what you're saying, but if your factory only needs 50% of the power your gens will generate, then the gens will run at 50% and using 50% of coal, no need for batteries

gloomy palm
#

if you timed it correctly, you'd have uninterrupted power between switching states

wind spade
#

sure, but here you go, fixed it for ya

unknown.png
#

so it's the same as if you weren't running power storages at all

gloomy palm
#

ohhhh

gloomy palm
wind spade
gloomy palm
#

so essentially there's no way to get free energy sad 😩

wind spade
#

no, that's what I said like way above. Both generators and power storages are 100% efficient in converting energy to power and back, so there's no gain or loss anywhere in the system

gloomy palm
#

interesting thinking_helmet good to know that before i went and tried testing this idea

#

thanks for clearing that up

#

\o/

dull bolt
gloomy palm
dull bolt
#

I react strongly to bad code (:

gloomy palm
dull bolt
#

Not sure if it beats the function that took an aragument, assigned it to a temporary and then returned the temporary. But that's another discussion :))
and it was called validate_<something> if I remember correctly...

gloomy palm
frosty owl
muted crypt
#

yup

gloomy palm
#

would you say there is only one correct systematic way to build a factory, or is there an element of creative leeway in what constitutes an efficient factory? Is it a case of, if you had 500 highly experienced players and asked them to build a specific project part, the majority would arrive at the exact same setup? Or would all 500 have completely different approaches to the same problem? Do the game mechanics limit what is possible in accomplishing a task in the "most efficient" way, or does it give you a set of options as to what pathway you are basing your "Efficiency" on?

#

and in following to that, what would everyone say should be the main focus in planning a large scale factory in terms of efficiency?

#

for example, should form follow function in the most minimalist way, or should there be an element of 'making it realistic'?

river night
#

It really depends on the exact criteria

#

Efficient how? If you give them an empty world, you could build time efficient (ie. least building time), power efficient, space efficient, or resource efficient

thin cliff
#

Here’s my system for iron, I just got coal power. Sorry for bad handwriting

image0.jpg
river night
#

You could also consider levels of resource efficiency. Iron is pretty ample, so one might not bother to pull in oil to reduce iron use, or whatever, even if its technically more efficient.

signal nimbus
#

Even more ample if you turn all the coal in the world into steel, then use steel alts.

gloomy palm
#

Fascinating thinking_helmet

gloomy palm
unreal osprey
#

At least for me my alclad sheet prod is way faster than any other belt material

tropic pagoda
gloomy palm
gloomy palm
river night
#

not intrinsically, but generally yes

gloomy palm
#

for example, building less buildings should likely mean more space efficient also no?

river night
#

sure those two might fit together

#

but using less resources usually means more power

gloomy palm
#

i think this might be worth writing down

river night
#

and typically also more steps

gloomy palm
#

which kind of steps?

river night
#

more processing steps, eg. more buildings, more space, more energy

#

although space efficient could also be just building with more verticality and just generally cramming everything together, which doesn't necessarily have to mean less buildings

#

but it would take more time to do that

gloomy palm
#

right

gloomy palm
river night
#

resources are infinite, using them to build something doesnt count

#

consuming entire nodes in a production chain is the only thign that matters, really

quaint rampart
#

Ehhh. Buildings are pretty much unlimited - you get so many parts that even in vanilla it stops being a limiting factor pretty fast

#

time investment can be a limiting factor, and bigger factories just plain take longer to build, but not resources really

river night
#

resource scarcity only becomes a real issue on some high-end resources or if you build reaaaaallly a lot of stuff

gloomy palm
#

how does this look

unknown.png
#

they aren't in any particular order

muted crypt
#

build time efficiency is extremely hard to calculate

#

because it will vary for EVERYONE

#

how "compact" something is will also vary

gloomy palm
#

using averages and such

muted crypt
#

averages of your own times?

#

because they may not be the same for everyone, that's my point

#
  • how familiar you are with the game
  • where you plan to build (i.e. distance to materials)
  • what recipes you use
  • how big of a build it is
  • how "nice" you want it to look
  • how much planning tedium goes into preparation
#

so many factors, these aren't even all of them

gloomy palm
#

i see

quaint rampart
#

optimization also looks different depending on scope

muted crypt
#

^^

quaint rampart
#

optimizing a full factory will tend to lead to different results than optimizing, say, two stages in isolation

muted crypt
#

you could optimize for space, FPS, resource consumption.. etc

quaint rampart
#

it's most material-efficient to make iron using iron alloy, for example, but if you optimize for that in isolation, you cut into your copper supply elsewhere

#

if you're optimizing for copper and iron together, you'll probably use pure iron and pure copper, even though pure iron is marginally less iron-efficient

muted crypt
#

ah yeah because you go from 2->5 with iron alloy to 6->13 with pure iron

gloomy palm
#

that's tricky yeah

#

so the game doesn't make a clear path between what is the best solution to a problem

#

there's fuzzy areas and tradeoffs

quaint rampart
#

boring-ass game if it didn't have that 🙂

muted crypt
#

If it was a clear path it would just be the same over and over

#

removes any replayability, removes any need for other options

gloomy palm
#

I recently saw this video (a summarized transcript is in this screenshot) from "Elon Musk's 5-Step Protocol For Successful Engineering" which you can find on YouTube, and it really got me thinking about how it might apply to Satisfactory planning

unknown.png
oblique hollow
#

Overall id just call build time efficiency "time efficiency"

#

Because overclocking falls in there too

gloomy palm
#

would that be like "timing efficiency" ?

#

or is that a separate branch

oblique hollow
#

Not timing, just time

gloomy palm
#

how does this sound

unknown.png
oblique hollow
#

Timing sounds like "doing the right thing at the right time" instead of "doing things in a time efficient manner"

gloomy palm
#

that would fall under play-time productivity

oblique hollow
#

Hmmm fair enough

gloomy palm
#

it's a player specific factor

#

whereas timing relates to the game mechanics

muted crypt
#

so by looking at this, does this mean proper balancing is more efficient than manifolding?

#

Because that's a whole new can of worms in terms of efficiency

gloomy palm
#

i always balanced because it gets everything up and running simultaneously instead of staggered

muted crypt
#

you could pre-fill machines for the same effect with a manifold

quaint rampart
#

balancers are lower latency when ramping production up or down, manifolds are drastically easier and more compact to work with

gloomy palm
gloomy palm
quaint rampart
#

at saturation they have the same throughput

oblique hollow
#

One is building time inefficient and space inefficient, but i guess production time efficient?

The other is small and fast to build and requires some manual prep

quaint rampart
#

only place I really use balancers is nuclear parts, to minimize the amount of crap accumulating on belts and in machines

gloomy palm
#

i want to see parts moving on every belt at any given time

#

if they aren't moving, something is either not efficient, or the excess should be going to storage or a sink

#

technically speaking, i use storage as a buffer in case i need to make a major change which would take something offline otherwise

#

every part received and created is buffered

#

my factory could run a few hours without any new mined resources brought in, entirely on stored units

#

but that's mostly because i'm not using them fast enough, i'm going to build a new factory in U5 to address this issue

#

we're getting somewhere

unknown.png
quaint rampart
#

You could slice that into quantitative and qualitative and probably get a clearer division?

gloomy palm
quaint rampart
#

Navigability is qualitative, for example

gloomy palm
#

ahhh

gloomy palm
quaint rampart
#

Quantitative means there's an unambiguous, numerical metric somewhere that depends only on the optimization choices you're making - building count, power utilization, inputs per output, or whatever else

gloomy palm
#

i'm thinking that i agree that navigability is relative since it's a nice-to-have and mostly helps with long term benefits

quaint rampart
#

Save size is quantitative as well, though most people don't care much 🙂

gloomy palm
gloomy palm
#

foundations, buildings, power polls, train tracks etc

#

like, building a drone port might be more space efficient than running a train line from point A to point B, but is a drone flying using less computational resources than a train would?

#

what if you had 10 drones or 10 trains, which would be using less GPU/CPU resources to move things from one place to another

river night
#

those two things have very different characteristics and you should build what makes sense for the product you transport, they are otherwise not comparable

gloomy palm
river night
#

sure, but thats not what you asked for, you asked for CPU/GPU comparison, seemingly disregarding throughput

gloomy palm
#

would it be possible to match their throughputs in order to compare their compute resource intensiveness?

quaint rampart
#

On a specific version, on a specific computer, in a specific factory, yes

#

In general, no, not really

gloomy palm
#

right

river night
#

probably, but it wouldnt be a very significant test since those are not real-world conditions. trains can move a whole lot more product then a drone can

#

and all that depends on distance

gloomy palm
#

but drones travel faster

river night
#

and trains can be long

versed violet
#

Wouldn't cpu metric be simply a number of active buildings (belts, machines, trains)
And gpu metric a density of objects per area?

gloomy palm
#

hmm tru thinking_helmet

gloomy palm
#

are there any passive factory objects

#

foundations and such maybe

versed violet
#

I do not have in-depth knowledge of how game works, but walls do not add do cpu load, as they do not require updating. Pipes, belts, storage, all the stuff that is calculated with each cycle, does

gloomy palm
#

makes me wonder whether power polls and cables are actively updating or if it only happens once when you connect or disconnect something

versed violet
#

My factory may not be big, but I do notice the cpu load increase as I build more machines and belts.

gloomy palm
#

the power grid though would be a huge portion of updates since the power can be potentially ever changing each cycle

versed violet
#

Hopefully grid just updates as whole and not all nodes one after another

river night
#

indeed the machines would update the power usage on the grid, poles or cables are passive parts in that

gloomy palm
#

it might also be worth noting that when you connect a grid, the whole grid and every single connection is logged into the same json object, whereas if you have separate grids, each json object becomes smaller. so it might be interesting to see whether it's better on the computer to load many smaller json objects (more smaller self-contained power grids) or one giant json object, one single power grid

gloomy palm
#

they merely signify the connection exists and then are aesthetic from that point onwards, as opposed to doing any actual work requiring individual calculating like in pipe segments

river night
#

yes, power is omnipresent, it doesnt actually flow

gloomy palm
#

makes me think though that perhaps pipelines could be pre-calculated after being placed somewhat similar to power grids

river night
#

fluid flow is quite dynamic, i'm sure they optimize as much as they can

gloomy palm
#

since the pipeline won't change unless someone deletes a part or adds an inline part, at which point you re-calculate the whole thing

#

but then that would mean it would slow down adding and removing instead of the live-calculating that it does now

gloomy palm
gloomy palm
versed violet
#

Modding server might be the best place to start, folks there have some in-depth knowledge of game internals.

gloomy palm
#

indeed thinking_helmet

wind spade
#

I don't think that there's big or small json objects

gloomy palm
#

although i'm new there and it seems i come across a little abrupt when chatting there so i try to avoid

wind spade
#

every entity in game is represented as an object and if they use json to save the objects, then every entity gets it's own object, there's no grouping together

gloomy palm
#

the only reason i say this is because of what i saw in the save editor

wind spade
#

but save editor converts save data into a different format

gloomy palm
#

ah thinking_helmet

versed violet
#

save file has little to no bearing of how ingame data looks like

wind spade
#

that as well

gloomy palm
#

would it affect how long the autosave delay would be?

wind spade
#

no

#

that depends on hardware and software

versed violet
#

more objects = bigger save = longer to save in general

gloomy palm
#

if only there was a way to benchmark this thinking_helmet

versed violet
#

you would need to build a test suite

gloomy palm
#

sounds elaborate

versed violet
#

simple engineering

gloomy palm
#

i was thinking of using a stopwatch

#

actually lol.... what about using a screen recorder then counting the time duration of the delay in the video frames

versed violet
#

close, but first you need to prepare several saves with different contents. otherwise, what are you measuring?

gloomy palm
#

yes only 2 saves would be needed

#

or the same save but manipulated

versed violet
#

thats awfully little data points

gloomy palm
#

essentially i want to measure to see whether having X amount of power grids separated is better optimization, or having everything connected to the same grid

wind spade
#

the more saves you test, the more concrete data you can get. You also need to test on different software and hardware to get any data that isn't specific to your configuration

gloomy palm
#

i'd have to first make sure that there's any noticeable different on my setup at all

wind spade
#

not to mention that such a small change (pretty much just one power line disconnected) would most likely have so small effect on save time that other stuff like RAM usage in time of save and how much is browser busy rendering stuff at the background would have way bigger effect

versed violet
#

multiple hardware test is not feasible. Just make sure you have average configuration with no weird shenanigans going around and that no bottlenecks are hit.

gloomy palm
#

1000s of power grids versus all those devices interconnected in the same grid

#

easier to do with mods ofc

#

just make one setup then repeat it

wind spade
#

don't think you'll have any relevant change in save time

#

the game still has to save all objects

gloomy palm
#

thinking_helmet that's why needs testing

wind spade
#

it may also be that the game doesn't really save grids at all

versed violet
#

Build large fuel gen or coal gen array.
Connect several machines to each generator.
measure cpu/gpu.
connect all minigrids together.
Measure again.

wind spade
#

since it doesn't need to, it can just reconstruct the grid data on save load

gloomy palm
#

Im thinking that maybe the save file would actually be larger if there were multiple grids because for each grid you'd have a new parent object

wind spade
#

but... there doesn't need to be any parent object

gloomy palm
#

a

versed violet
#

I expect no difference at all - just a diff number on machines. At most, save will have the extra cables listed.

gloomy palm
gloomy palm
gloomy palm
wind spade
#

why not? all you really need is to save all the entities

gloomy palm
#

Interesting, i just assumed they handled json because thats what the save editors show

wind spade
#

even if they did, why would they bother with tree view if they can just do list of entities

versed violet
#

The save file format doesn't matter for measuring game perfomance.

wind spade
#

what are you trying to measure anyway and why?

gloomy palm
#

surely they'd have a schema thinking_helmet

#

Internal api would be a nightmare without it

versed violet
#

they have objects. UE is object oriented I think

wind spade
#

I assume UE also has some saving capability, they can be using that

gloomy palm
wind spade
#

save file compactness = least buildings

versed violet
#

Build less objects 😉

river night
#

also cut down less foliage