math-and-meta

[wind spade]

• wasting time on building factory you don't need, instead of making factory you need at the moment
• not having logistics prepared according to needs of specific factories
• not knowing how much you need in the future (so you may overbuild or underbuild)
• centralising production (which often leads to logistical issues)
  etc.

[outer vale]

easier than having to track how much plastic's being used across every factory that wants a chunk of that plastic and having to distribute it all around

[magic island]

I don't integrate oil or aluminum production directly into the factories that consume them, because that's a huge pain

but I don't massively overbuild my oil/alu outposts either, I keep them scaled to the amounts that I need to ship around

[outer vale]

smaller setups are also easier to keep within belt and pipe limits

[deft lichen]

IsUsingMods: Yes

[brisk shore]

Official Modding Discord

Learn all about modding and join the community here!
<3 @deft lichen

[sand goblet]

?

[unborn ermine]

Its in the crash log

mods are involved

[muted hamlet]

None. I just connected some stuff and voila, it'll work eventually

[white bloom]

after tens of hours spent on my balancer solver I just realized I don't actually fully understand what splitters and mergers do exactly, in any scenario. lol

summoning the splitter/merger experts that tested their behavior thoroughly

1. when a merger's output belt capacity causes a throttle, will it take the same absolute number of items from each of its input belts, regardless of their belt speed or item loads? or will it take them proportionally, i.e. more from the belts with more load / higher speed?

[wind spade]

afaik splitters they are round-robin, skipping full outputs

I assume mergers work the same way (round-robin, skipping empty inputs)

[white bloom]

2. when splitters have varying output belt speeds, will the faster ones get a larger share?

I think it comes down to how quickly the round robin is done

[wind spade]

afaik both advance the round-robin when items arrives

(again this is just my guess based on what I know, but I haven't tested it)

[violet halo]

With a caveat, if the difference is so large that the slot on the slower belt is still occupied, then yes.

[wind spade]

this heavily depends on specific situation

[white bloom]

if a machine accepts from / pushes to a slow belt, that belt is reserved for a short amount of time, isn't it? so if the round robin checks very quickly it might skip that output on multiple cycles

the alternative would be if it really abided by the strict round robin that e.g. a MK1 and a MK6 belt going into a merger that has a MK6 output would cause the MK6 input to throttle to 60 /sec to abide by round robin and hence the sum would be capped at 120 throughput

which I imagine is not what would happen

[white bloom]

like this, yeah

so yeah, I'm really struggling to accurately model their behavior in all situations

sucking in or pushing out an item to a belt takes some time, I think that's simply the inverse of the belt capacity. MK1, 60 items per min, 1/60th of a minute for 1 item (1 second). MK6, 1200 items per min, 20 items per sec, 0.05 seconds per item

[magic island]

it depends not just on the overall rate coming through, but the pattern of gaps etc

[white bloom]

right that makes it even worse, but let's assume it's uniformly spaced to nail down that case at least since it's hard enough

I suspect the merger might be operating with a first-in-first-out queue. when an item is about to dock on one of its inputs it registers in the queue and once there is place to go (i.e. the buffer) it will fill that ones its turn is up in the queue. Since the successive items on the same belt can't enqueue before their predecessor got sucked in, that usually would result in exactly the type of round robin behavior that is typically described

and it would explain what happens at different belt speed inputs, especially why a slower other input can't slow down a faster one

thoughts?

[magic island]

I don't know for sure how the round-robin is handled for varying speeds on splitters

the way I would program it, if I had to, would be: for each item that comes in, check which belts have the capacity to accept it, and then from those, pass to the one that's gone longest without getting an item

no idea if the actual system is close to that

[vapid gorge]

afaik - belts round robbin for even splits as much as they can

[magic island]

sure, but an even split would have to involve compensating for when a belt misses its turn (or takes extra turns because others are full), wouldn't it?

[vapid gorge]

for example

if you have a full mk3 split into 3

2xmk1
1xmk3

the 2x mk1s will have 60ppm
1x mk3 will have 150

[white bloom]

that's for a splitter, right? I find splitters even harder to conceptualize. What about the merger, first?
If I try to turn our description around, I think that would be: whenever there is capacity to push out an item, check which belts are presenting an item, and then from those, choose the one that's gone longest without having an item accepted.

Is that an equivalent phrasing of a fifo queue? If not, how could we test it in game to tell the difference?

[vapid gorge]

mergers are also round robin

but hte use case for that is limited

[magic island]

testing a splitter in-game would require feeding it a supply of mixed items in a given pattern, and seeing what distribution patterns the outputs get

[white bloom]

so 60/min + 1200/min into merger into MK6 out results in only 120/min out and the rest of the 1140/min input back up despite the output belt still having spare capacity? Because that's what strict round robin means. I highly doubt it works like that in game, you can do MK1 belts on machine outputs on manifolds without it causing to throttle down the whole manifold, can you not?

[vapid gorge]

no because the 1200 will have a lot more points of 'input'

it'll round robin the points where there are 2 inputs hit at the same time

[violet halo]

With that amount of difference, the spacing of the items changes the ratio.

[vapid gorge]

is this for sushi belts?

afaik if you're trying to sushi balance it keeps the round robin with multiple outputs but you need a higher mk belt on the output for the split

[violet halo]

The internal buffer compensates for some of it, but at 1200/m, an internal buffer of 9 isn't enough.

[white bloom]

ooh, you mean for tiebreaking when two items register exactly simultaneously! I see now

[vapid gorge]

for example, this person showed you could have a mk1 split into 3x mk1 belts.

2x mk1 belts moved backwards to remerge with the primary belts so it was 2x 20pm belts merging with 1x 60 pm belt, and it did enough round robin that it created a 20pm limiter

the use cases for this sort of behaviour is extremely limited, but it's there

not clear how it would function with a less than 60pm full belt

[magic island]

the round robin is super straightforward for saturated belts with matching speeds

what happens at different speeds / saturation levels is more of a question

[vapid gorge]

and considering machine burst output items there's very few ways to seriously control gaps and patterns

you'd probably want to talk to @frosty owl more about this, probably has run into it a lot with Single Input Sushi Balancing madness before

[white bloom]

would be interesting to predict and experiment how this limiter throughput changes if you change the MK level of various belt segments

[vapid gorge]

I think the main part about it was that the 2 recursive belts were exactly timed to create teh round robin limiter

the 2 remerge belts never got held back but the full 60pm main belt did

you can set it up yourself very easily, just make the 2 return belts symetrical

[white bloom]

can we settle that bit for the merger still first before trying to wrap our heads about splitters?

[vapid gorge]

I think the mergers would be harder to test because you're dealing with different 'beats'

unless you try to merge like 3xmk1s into a 1xmk1 output

you'd probably get an even mix

[white bloom]

nah I mean like different speed inputs and the output is capped at less than their sum

[vapid gorge]

and if it was 2xmk1s +1mk2 to 1xmk2 output I'd expect the output to be 1/2 of the mk2 belt item

[vapid gorge]

better set up some mixed belts and test it 😛

[wind spade]

the problem with mergers is that you're not only dealing with belt speeds, but also with gaps between items and offset of the items and belts from each other

[white bloom]

I might just hop in and put up containers and see how quickly they drain

[wind spade]

having a mk2 belt with 60/min can mean that you have one item every second (same as if you removed every second item), but you can also have 1/2 minute of a full belt followed by 1/2 minute of empty belt

those two behave differently, especially if smart splitters get involved

so with any reasonable software, you either have to do a full simulation over large amount of time and calculate the result, or just assume that all belts have a stable X/min flow and estimate things

[white bloom]

right, I ultimately want to just stay at the stable X/min flow level, but I realized there is still ambiguity about this when bottlenecks and different belt speeds come into consideration

[wind spade]

in my manifold tool I just assume the split is 50/50, unless one of the branches is capped by belt speed to less than that, then it just gets full belt and rest goes the other way

since it's an estimation of fill time anyway, I figured that accuracy at single item level isn't relevant much

[unborn ermine]

where is Ven Cum Sus(hi) when we need him

[white bloom]

I just wonder if you can utilize uneven splitting (or merging) behavior by deliberately introducing lower belt MK bottlenecks at specific points in a balancer network to achieve a certain balancing function with fewer machines than if you had to do it relying on recursive backloops etc alone.

but the end goal doesn't matter much right now I'm just curious how the game actually works in that regard, everybody can see for themselves what to make of it once we know. ^^

[pallid knoll]

Been listening to you guys, made a manifold system where everything gets merged per material and split per machine as manifold. Just need to finish the assemblers and organise it with smart splitters. Won many space with this build. If you have any further optimization or suggestions please let me know!

[white bloom]

like the limiter element Cobalt mentioned one can see how that helps splitting 120 into 100 + 20 for example

(bad example, since a normal 1:6 does the job, but whatever)

[magic island]

yeah, unfortunately I think such a system would get unpredictable very quickly tbh. performance might even throw the numbers off when dealing with higher speeds

[white bloom]

I wouldn't even worry about individual item tiebreaking for my purpose I don't want to create an insane sushi belt system that comes out perfectly sorted without sorting smart splitters on the other side, only worrying about tiebreaking insofar it has measurable impact on the average throughput rates from each individual input/output respectively.

I'll hop in game and see what I can do in creative mode

will report back

[magic island]

yeah I think the use-case for mixed belt testing here isn't so much to see how sushi gets parsed; just a way to tell which belt it came from.

like, send in iron ingots on an Mk1 and copper on an Mk2, see how they each get distributed to learn what the logic is

[fresh mesa]

just to check my math, 300 coal will feed 20 coal gens?

[violet halo]

Yes, 15/m for each one.

[brisk smelt]

what a horrible few hours...

[brisk smelt]

at 100%?

yes

[brisk smelt]

i mean ig space optimization, you can push the splitters two ticks closer to the machine or use lifts

[white bloom]

I'm testing with a setup like this and already got some pretty weird results

with MK1 + MK6 in, MK2 out you get this smooth alternating pattern, i.e. the MK1 can move through unimpeded, getting its full 60 /min through

but when you upgrade the output belt to MK3, increasing its capacity, the MK1 belt starts to stutter

it must have to do with divisibility

[magic island]

so it's affected by how the rate of the belt divides into the rate of round robin ticks? or something to that effect?

[brisk smelt]

yeah because it's trying to take round-robin doesn it

[white bloom]

with MK3 out, input still MK1 and MK6

continuous 1-4-1-4-... alternation

[magic island]

so that would be taking 54/min from the Mk1 belt and 216/min from the Mk6 belt

[violet halo]

So the ratio is the lower of the two, divided by output, rounded down.

[magic island]

assuming the pattern holds and there aren't occasional groups of 3 or 5 concrete

[white bloom]

MK1 provides an item each 1 second. MK3 consumes an item each 0.222... seconds. So it can fit 4 items from the fast belt in, then it's at 0.888... seconds, and the next item completes at 1.111.. then. That would suggest the rate at which the slow-connected container empties should be 60 * 1/1.111... = 54 items per minute. guess I need to video capture this

[white bloom]

yeah, it holds. stable pattern

[magic island]

one thing this would prove is: just because the Mk1 belt gets a turn less often, that does not give it "catch-up" priority.

whereas if the merger was set up to aim for as equal consumption as possible, it would accept items from the Mk1 as often as possible and never block it

(in this specific setup)

[wind spade]

so from math point of view:

mk6 side will always have item waiting (it's faster than output belt)
mk1 side will have item after 1 second

270/min belt can take 4.5 items per second. So after it takes item from mk1, it then waits .22 seconds (until output is empty), alternates to mk6, takes item, goes back to mk1, but it's empty (takes 1 second to fill back), so it goes back to mk6. Same goes for .44s, .66s and .88s. But then at 1s the mk1 fills so at 1.11s it takes from mk1 again and cycle repeats

[violet halo]

That's why the ratio rounds down.

[magic island]

and if one belt fails to supply an item, it'll check the other belt on the same tick. so every missed turn for the Mk1 potentially becomes an extra turn for the Mk6

[patent blaze]

i was kind of considering making all my 800 circuit boards into high speed connectors until i saw the quickwire

[white bloom]

i mean it does get priority in some sense as you expected as per the previous example, the MK6 times are a clean multiple of the MK2 output as well, so MK6 and MK1 input arrive at the same time whenever a MK1 input arrives at all. If it was random or tiebreaking was alternating then 50% of those times another MK6 item should be chosen, but indeed MK1 gets the right of way every single time it is ready.

so it gets preference in the sense you initially suspected that if they are simultaenous it chooses the one who hasn't given something longer, but it's not getting preference to the point that the output would actually block and wait for the upcoming MK1 item. When there is place to go it will always pass on some item

[wind spade]

and it depends how big the gap is

[patent blaze]

24000/min

[white bloom]

ok I will do a timing test to see if the 0.22, 0.44, 1.11 s etc thing actually behaves like this, let's see

57 screws emptied from the container in 1:03.314

that's...

54.016 per minute

[wind spade]

1.1107 🙂

decently close

(it's actually 1.1111111)

[vapid gorge]

that's only 2x what my nuke set up needed, very doable

[white bloom]

yeah, +- one frame error since the video capture and game ticks might be out of sync

[wind spade]

also depends on where in the cycle the first mk1 item gets taken, so that may explain the discrepancy as well

[white bloom]

really nice. okay, can we use it to predict other cases... like suppose I upgraded to MK5 belt (780 per min), what output pattern do we expect to see?

[wind spade]

which belt is mk5? output?

[white bloom]

MK1 + MK6 in, MK5 out

[wind spade]

alright, let me write it up 🙂 sec

[white bloom]

oh that one is nasty

trying to work it out before testing too

[wind spade]

same things still apply (mk6 belt is always available due to being faster than output)

mk5 is 13 per second, so once you take from a mk1 belt, you have 12 items taken from mk6 before it goes back to mk1.

Now, the thing is - that's the theoretical way. In practice, here it fits perfectly (780 into 60 is exactly 13, not something decimal like previously), so it could also happen that due to decimal math, you'll have 13 items from mk6 (once or a few times), because the mk1 belt may have moved at 59.99999/min speed at the tick when checking due to computers sucking at decimal math.

But my estimation is 12 at least in most cases

[white bloom]

12-1-12-1? yeah that makes sense. let's see

[fresh mesa]

whta are you setting up over there? fuel or focket fuel?

[white bloom]

bingo, it's 12-1-12-1. though the 60 divides the 780 cleanly as you said, so it may not have been the best example for something with fractional overlap. the MK6 input doesn't divide cleanly but since it's lower tiebkreaing priority that doesn't matter

can we think of another example to test? maybe something where neither is cleanly divisble, is that even possible with the belt limits being as they are?

[brisk smelt]

plastic and rubber

2400/4800

[frosty owl]

Funfact: I had to include a container in the input belt to feed a machine using single-input-sushi once

Something about a recipe requiring half a stack for one production cycle...

[vapid gorge]

because you couldn't get enough of the right item in a row for it?

[north jackal]

.

.0...................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................................

[glossy wagon]

so if i have 840 capacity then why is my grid crashing with 500 consumption

[pulsar notch]

is fuel running out maybe?

[glossy wagon]

thats what im double checking out and it doesn't seem... it has to be if its dibbing but i have 6 coal refineries going right niw

[fallow siren]

this is what causing ur grid crashing

the production isnt stable

[vapid gorge]

yeah, one or more of your inputs for the fuel gens is unstable

open up the control panels, are they out of coal or water or both?

[glossy wagon]

its definitly not the coal

[vapid gorge]

well check and you'll see 😛 take some over head shots of your whole set up too

[glossy wagon]

is there a limit to how far water can pump level? no lift involved

i don't have any pumps and im thinking that may be my issue? its all level

[hollow hollow]

If water is not going up it shouldn't be a problem

[glossy wagon]

thank you!!!! i had 6 refiners and only 2 pump extractors at 120

[hollow hollow]

Yeah i think you need a bit more than 2

[wary rapids]

I want to say that that 45 rocket fuel per minite i can make off the remainerd of my coal is enogh to run 72 trucks.

does any engeriner know what that MJ / MW dimsional anylisis looks like if i were to properly write it out and cancel units.

[glossy wagon]

need 3 exactly

[wet python]

does anyone have a aluminum setup with the Electrode recipe? that 105/m water output is quite a challenge...

I'm trying to avoid using a VIP junction but seems there is no other way (aside from sinking something made with water)

[hollow hollow]

Since you're already using coke, maybe set up some extra diluted fuel and burn it?

[wet python]

I just had an idea to overclock one of the sloppy refineries and underclock the other, so i keep only this underclocked one fed by extractors

I'll see if it works

[brisk smelt]

js oc them to 250%

[unborn ermine]

So, just confirming I probably dont need the loop here BUT
one side is providing 500/min, the other 200/min.
This should play nice?

Just wanted to ask before I commit to this

Recycled rubber/plastic loop fuel supply, and the next one would be a total of 900/min

[brisk smelt]

yeah probably dont need a loop

but i count 10 refineries, doesn't each one take 75 fuel/min

missing 50 fuel

[unborn ermine]

Ive got the lot clocked so they want 70

[brisk smelt]

ah ic

[unborn ermine]

10 for plastic 12 for rubber

[dark star]

have you tested it yet? im interested to see if you run into throughput issues with the intermittent demand of the machines

[unborn ermine]

its like 233.33~% and 250% for rubber

Still cobbling it together

[brisk smelt]

small amount/machine, not much issues to be found imo

[brisk smelt]

wuoldnt it be easier to do groups of 8x250 that can be perfectly split into being fed by 3 blenders

[unborn ermine]

This is my current plan, 6 blenders atm.

[brisk smelt]

such nice numbers.. my petro factory was filled with repeating decimals and overflow splitter hell

[unborn ermine]

I dont want to me mucking with my water extractors at this point, 5 x 400/min water pipes

[wet python]

i've built something quite along that way

[prisma kraken]

does anyone have any sort of data on the computational cost of smart and prog splitters?

[prisma kraken]

that's a strange mix you have going on there... targetting rates for other production lines?

(its either that or you're just making the machine counts integral, lol

[unborn ermine]

Yeah, I decided to say screw it and do some rubber concrete for my HMF for the computer/ACU factory
End up needing 1080ish rubber + 420 plastic.

havent also accounted if I want to do anything else, but I know if I do its a drop of plastic and rubber for later.

[prisma kraken]

it feels so dirty to use rubber on concrete, but the recipe is really good if you need a lot

[brisk smelt]

there's like 50k concrete w pure recipe on the map... in what circumstance would you need more

[livid turret]

I found out about a set of three refineries in a loop to create 90 plastic or rubber that only requires 90 fuel. Makes it easy to just stamp what I need in multiples of 90.

[prisma kraken]

early i built out 3600 rubber/min and just tossed it on a train to distribute - currently i only have 120/min left of that after it delivers to my concrete, hmf (adhered plate) and cooling system builds

[unborn ermine]

logistics, or lack thereof.

[prisma kraken]

you can never really go wrong with making a few thousand/min of rubber or plastic and just tossing it on a train

pre-1.0, you had to be a little careful with overextending power before hitting nuclear with the recycled rub/plas builds, but now, just build em and stick it all on a train

[latent seal]

can someone help me make sense of the numbers im reading in the game:

can I trust the fpm the stations are giving me?

[prisma kraken]

i honestly don't ever trust them

[latent seal]

should I time the trip when i record it?

maybe thats not how you figure it

i was trying to see if i could math it, like precisely?

[prisma kraken]

don't bother, i'll give you a better way of upper-bounding fuel

each vehicle has a mw rating - for the tractor it is 55mw

[latent seal]

i use plastic to make the fuel packages for the whole system

i want to make sure i don't overproduce plastic

[prisma kraken]

you know how many mj's a fuel item contains (its listed in the codex)

mw/mj = seconds of burn time

[latent seal]

yehh

trying to find in game where it states their mw rating

[prisma kraken]

so a tractor on yellow fuel is 750mj / 55 mw = 13.64 seconds

or roughly 4 fuel per minute

[latent seal]

not sure if that info is available in the game?

or am i missing it somewhere,

[prisma kraken]

i'm looking. the values are on the wiki

[latent seal]

ill take a look at the wiki for sure

bit of a shame if it wasn't printed in the game somewhere the mpg for the vehicles

[prisma kraken]

can't seem to find them in the codex

i thought they were listed in there too

[latent seal]

man i was using my whiteboard and everything writing down all the pieces of info the game was actually feeding me

and like, i knew i didn't have enough information yet

[prisma kraken]

anyway, don't bother to time round trips or anything. the upper bound can be assumed to be the MW rating

you'll overshoot a bit, but that isn't a big deal

[latent seal]

overshooting in my case just means maybe i burn a little more petrol coke

[prisma kraken]

yeah, for me the excess fuel gets sunk

[latent seal]

but hoping to not flood the system

because i have to manually farm biofuel

i am being extra careful 😄

[prisma kraken]

you'll want to get off of biofuel ASAP for vehicle fuel

[latent seal]

i think i have a perfectly viable solution with sloop

[prisma kraken]

btw, in case you overlooked it... vehicles can use coke as fuel

[latent seal]

interesting

i didn't think of it for sure

[prisma kraken]

it actually has a little bit better of a range than coal due to its 200 stack size

[latent seal]

this is the biofuel plant

[prisma kraken]

range really doesn't matter much if you're staing in a biome

i think you may be overdoing it on biofuel

[latent seal]

28,800 m3 biofuel storage with 600 m3 per min biofuel production, all biomass slooped every step of the way

you just said i need to get off biofuel

and now it is clear that I am good for a while 😄

[prisma kraken]

well, i mean, like given enough playtime w/o automating the input, that's going to run dry

[latent seal]

i haven't went biomass hunting in days

but i only have 3 trucks so far, we'll see how it goes 😛

[prisma kraken]

i make the stuff for the jetpack which sips it, but don't rely on the fuel type for any factory stuff

i'm a little miffed that 1.0 got rid of color cartridges

those things would fuel tractors for a really long time - long enough to switch over to batteries and skip oil in them entirely, lol

but flower petals are no more. rip

[vapid gorge]

you almost certainly need a loop - also recommend against merging the 200 down the other side. Though it's not impossoble for that to work

[latent seal]

so math says i eat 37.2727272727 packaged biofuel per minute

[prisma kraken]

i think you forgot to do a minute<->second conversion

unless that's for 4 tractors

[latent seal]

2 tractors and 1 truck

[prisma kraken]

ahhh

[latent seal]

so (750/55) * 2

• (750/75)

[prisma kraken]

btw, with the truck, unless you know you need the extra capacity, save yourself the fuel and use the tractor

[latent seal]

so is it 750/55 ? that means that the tractor has a bigger number

or did i do it backwards, is it 55/750?

no

[prisma kraken]

dimensional analysis 101: MW = MJ/s

750 MJ / 55 MJ/s = 13.64 s

[latent seal]

ohhh wait

13.64 seconds, not 13.64 units per fuel

so its 1 fuel to 13.64 seconds

[prisma kraken]

which is roughly 4 fuel per min

[latent seal]

4.39

got it, okay i can math too

im set now i think, thanks for your help!

[prisma kraken]

np

[latent seal]

14.79

[prisma kraken]

i usually round the number to 4 for the tractor - they use less fuel per minute because they stop and on downhill slopes use less

[latent seal]

so 15 pm

that will be fine

[prisma kraken]

sounds pretty good

[latent seal]

so now let me tweak my plastic machine

[prisma kraken]

what i'm still trying to figure out is how to do napkin math on drone fuel

[latent seal]

and do the dance with the modular frame factory which runs on plastic coated iron plates

[prisma kraken]

well, g'luck with that all

right now i'm trying to get cooling systems and heatsinks home, lol

[warm blade]

Hi if interested in fixing miners alignment please add your vote:)
https://questions.satisfactorygame.com/post/66eef2a6772a987f4a8c6013
https://www.twitch.tv/videos/2305776208?t=01h44m08s

[vapid gorge]

to the world grid?
I hope they remove the grid. Better solution

how do I vote against this now

[warm blade]

I think just dont hide stuff be more transparent with players and dont be afraid of letting go of control. It is not as much fun to play against "automagic" systems who try to predict and prevent me from being stupid - it just ends up being frustrating as you are constantly just fighting and bending "ruleset"

I rather prefer magic systems to be opt-in

[vapid gorge]

what, in great Atheismo's name, does any of that have to do with the world grid?

[fringe pawn]

I'm not sure how I feel about someone randomly bringing up a QA site post fishing for votes if it didn't come up organically in other discussion.

[wind spade]

WG is hell and snapping miners to any kind of grid feels unnatural

[unborn ermine]

Technically speaking, each miner is its own grid

[wind spade]

In multiple channels as well

[fallow siren]

i dont mind nodes arent tied to world grid bcs its a challenge u need to solve

and its fun

[vapid gorge]

your post is nigh unreadable, one of your first issues is 'can't nudge this bp of multiple long belts exactly where I want off foundations'

you know belts in bps don't connect right?

and that foundations exist to create those snapping points

[past reef]

Would be nice to have the miners easily align with however I build, but then I already have natural train tracks I don't mind a small local grid in a random ore node

[vapid gorge]

takes 2 seconds https://www.youtube.com/watch?v=DSNcRGPv0o0

[past reef]

Wat you can already do that? Neber saw it I always did 90 degree straight belt after a miner

[vapid gorge]

have a look at teh vid,

it's very straight forward

[vast jungle]

the method definitely helps to align a foundation to the "grid" of the miner...

[vapid gorge]

once that's done you can use the WG all you like

especially with 'straight mode' belts

[vast jungle]

combine this with a 90° turn and it should be easy to transfer to any other grid

(damn, too slow ^^)

[unborn ermine]

I never did the barrier trick, I usually just added a foundation by eye on top of the miner.
Still was able to have a straight output into a splitter.

[vapid gorge]

yeah lots of ways around it.

[vast jungle]

barriers are another crazy tool the toolbox 😄

[past reef]

Yeah I got it now, straight belt means I just need the method to build foundation from the miner (barrier)

[prisma kraken]

i honestly just use a short diagonal belt to get onto grid and leave it as a small sin

[frosty owl]

I could, but as the machine gets to the end of the production cycles it piles up a full stack of one item, temporarily blocking the belt behind; if the clog got far enough along the beltwork it would have affected balancers, so I slapped down a buffer to... uhm... buffer 😅

I actually got that on video here, it's the manufacturer in the top row. The belt between the container and the manufacturer stops for a few seconds before the machine finishes a production cycle
https://www.reddit.com/r/SatisfactoryGame/s/swHwTZ28qw

I could have just made a long belt (and I tried), but the size was quite ridiculous

[vapid gorge]

Very weird xD

[fallow siren]

i also did this with road barrier, took only seconds

[warm blade]

My point is system got downgraded from EA8 in general, that was easiest MVP demo to show how much worse it is since 1.0 for me. What you prefer is up to you what I prefer is up to me. What devs do is up to them.

[fringe seal]

praise CSS, I don't need to train 5000 coke all the way from gold coast

[warm blade]

takes 2 second to have non-straight belt in OCD focused game 😄

[wind spade]

game becomes 100x nicer if not everything is on global grid with right angles

[vapid gorge]

Down graded what?? Literally nothing you say is making any sense

[tight thicket]

hi everyone, i have a question, how long does it take a truck to consume a stack of petroleum coke?

[vapid gorge]

I’m going to assume your native language isn’t English at this point

[vapid gorge]

the wiki is your friend, praise the wiki https://satisfactory.wiki.gg/wiki/Vehicles#Fuel_consumption

[vast jungle]

all praise (the right!) wiki... 🙂

[warm bane]

Just making sure, but the only rounding in machine interfaces are visual, right ?
Like, my manufacturer at 75% making RCU will make 75%*3.75=2.8125 per minute, despite displaying 2.813

[brisk smelt]

i mean, i think so, but just clock to 100%...

same things are coming out anyways

[past reef]

I recall only clock number is forced into 4 decimal place, other ones can work with floating point error, 2.8125 is well within

[warm bane]

Not really, if you clock things to 100% but they don't get what they need, at some point they starve of input for a while and then they have their startup time when they get enough to start again, and it also make your power graph ugly

[outer vale]

per minute numbers are only a convenience, the ones that actually matter are the cycle time (influenced by clock speed) and the output per cycle

[deft lichen]

Correct, the UI rounds

[white bloom]

Some take-aways after sleeping over the merger experiments yesterday:

• if the belt speed of an input belt exceeds the speed of the output belt. its exact speed doesn't matter. It will always have another item to offer when the merger's output is done churning one out.
• it shows that there is indeed a difference whether its FIFO or choose the longest since last acceptance, disproves my FIFO hypothesis and corroborates the last acceptance hypothesis. The MK3 example loads the fourth item after 0.889 seconds from the MK6 input. At 0.939 seconds the next item from the MK6 belt is ready. At 1.0 seconds the next item from the MK1 belt is ready. At 1.111 seconds a new item is chosen - and it chooses the MK1 belt, which has been longer since the last acceptance, but only got ready more recently. So that settles it.
• if we look for more messy merger behavior we should look at instances where the output is the fastest MK belt, but neither of the two (or even three) inputs divide the output cleanly.
• when we're done with this, we can move on to splitters

[versed violet]

Did you observe the round-robin scenario? Splitters (and mergers) dustribute to each output in turn, unless said outputs are blocked, then it skips the turn.

[fierce vault]

how is the setup for this steel production site?

[prisma kraken]

that oil well there solves coke for all the bauxite in pink forest 🙂

[fringe seal]

• for all the bauxite in the island

[prisma kraken]

looks great for a t2 build 🙂

[fringe seal]

I think I have 120 oil left after that?

[prisma kraken]

yeah, true, there's enough there for all the baux in the world, but i do titan forest and swamp baux on-site 🙂

tf and swamp bauxite has nearby copper, so sheets & casings end up in those places

[fringe seal]

ah

I see, I think I'll just train copper ingots to it but idk

[prisma kraken]

running a train up to pink forest is a little painful. the copper at the base of the rocky desert ramp is what i've used in the past

[molten lintel]

bro what this about to consume a week off me

[prisma kraken]

what's interesting to me is that this copper is now really accessible to pink forest

the sulfur there converts 1:1 to copper as well if that isn't enough

[magic island]

3 generators consume 12.5/min, if that helps you approach it

[molten lintel]

144? generators out of 600/min production poggies

[pallid knoll]

I was thinking of lifts but i first want to complete the whole setup before optimizing 😆

[molten lintel]

would 144 work fine out of a full rocket fuel mk2 pipe? or should I go max 143 or less to give it some room?

[outer vale]

why not just not put it all in one pipe

you could make it work, sure

but it's easier to just not fight it

[patent blaze]

500/120

Only ”nice” ratio

That way you can do 10x12 fuel generator blocks

144 would work but 600 in a pipe is scary

[heavy gust]

Works out for full pipes

[patent blaze]

If you want something in between you can do even increments of 6 fuel generators, 120 126 132 etc.

[heavy gust]

I split it into 2 300

[patent blaze]

That works too

I did 500 in every pipe because i got scared lol

[heavy gust]

Never even connecting the 2

[patent blaze]

👍

[heavy gust]

Half the blenders is an even number

[patent blaze]

Had 4 blenders make 500 for each blueprint multiply that by like 11,something

[heavy gust]

Not even but uhh full, not a fraction?

Whats the word

[patent blaze]

Uuh

[heavy gust]

Whole number

[brisk smelt]

natural number

[heavy gust]

Yes

[patent blaze]

Trancendental number

[brisk smelt]

irrational number

[molten lintel]

integer

[heavy gust]

I would like 3i+5 blenders

[patent blaze]

Im getting the urge to delete my entire rocket fuel plant and redo it with mk6 stuff

And nicer piping so it doesnt run at 90%

[heavy gust]

You using ionized?

[patent blaze]

Nah not yet

[heavy gust]

Dang i would never have bothered without

[patent blaze]

Considering doing that too though since there is excess coal there to make the shards

[heavy gust]

Ionized is crazy resource efficient

If you dont desperately need nitrogen

[patent blaze]

Nah i have excess nitrogen aswell

[white bloom]

I haven't done splitters yet. mergers have only one output so what you wrote doesn't apply to them, or rather applies trivially. the detailed behavior of mergers I found are explained starting at this message here: #math-and-meta message

[patent blaze]

I had tot ake two nodes and only used about half of the second one

[heavy gust]

Im using a full pure node on my nuclear stuff

[patent blaze]

Gonna look into how much coal and quarts it’ll cost

[patent blaze]

Lol same

How many rods are you making

[heavy gust]

16.8 plut

But the water for the plut plants isnt yet fully set up

Still filling rods mostly

[brisk smelt]

build ur nucpp's in the ocean and copypaste them

:D

[patent blaze]

Burning plut is crazy

[fierce vault]

i havent played this before so im learning things as i go, but i hope its not bad haha

[heavy gust]

I did, just not the pipes

[heavy gust]

Fisonium is shit, im either going to mod it till useable or save edit the waste out

[warm blade]

sry - if that helps you. I was not thinking about you when I was writing that post. That post Lists different points of friction between various In game systems. Giving overview where are options to improve QOL. It looks to me you don't understand word synergy (I may be projecting hard). It is also a Joke for people who know how it is used in business.
But your feedback is valid and I edited that post to make a more sense to your point (hope it helps)

[fierce vault]

my full steel production site, is there anything i can do to make it better?

[brisk smelt]

have you perhaps heard of a manifold

also... dont buffer like 1k steel ingots yheres no point

[fierce vault]

i dont know what you mean by buffer?

this is my first time playing

[warm blade]

https://www.twitch.tv/videos/2305776208?t=01h44m08s
Yes Please, add nudge to miners to snap them to nearest coordinate

https://questions.satisfactorygame.com/post/66eef2a6772a987f4a8c6013

[molten lintel]

I don't think there's a reason for buffering ingots (storages), load balance them to make sure they start at the same time but even that is not necessary.

If those are not storages, then my bad, I'm using my phone 😅

[brisk smelt]

hm i also think u could just put the consteuctors right next to the foundries, saves a lot of space

[fierce vault]

They are storages and the only reason for it is so I have access to a stockpile at all times, I have done it with all my recourses so I alwyss have plenty of them and a lot of space to hold them

[fierce vault]

Yea I could’ve done this but I also wanted it to look cool too haha, I’m not going to minmax every setup just yet, I’m just having fun lol

I also just realised I have 4 unused lines of steel 🥲

[molten lintel]

Later on you will eventually remove them but that includes spoilers 😆 Actually not that later on. Some exploring might prove super useful 👀

[heavy mica]

Could someone tell me if this will work? 5 blenders take 240 water each, 1200 total. Two pipes feed them and the middle blender is connected to both pipes.

[brisk smelt]

yeah that works

123 fed by pip1 345 fed by pip2 right?

[heavy mica]

yeah

[fierce vault]

I’m confused on what you mean? Are you talking about the dimension storage?

[modest jackal]

currently doing my first Oil production factory, does this look good?

[worldly granite]

Exploring is always great, u can find parts, harddrives, slugs, sloops and spheres. Also u can build how u like there is no wrong way to play this game.

[brisk smelt]

your crude should always make HOR no matter loop or not

straight to plastic is just horrible effienciy

wait never mind u cant make plastic from hor

[modest jackal]

HOR?

[brisk smelt]

but yeah, you should have 2x the rubber than plastic, general guideline for big factories

[modest jackal]

oh rubber is more useful?

[brisk smelt]

yeah

[modest jackal]

is 120 packaged fuel too much or is that good?

[brisk smelt]

you are nuking this factory anyways when you get tier 7 and blenders, that recipe gives 1 crude -> 3 petro products

[brisk smelt]

its fine, what are u using it for

[modest jackal]

i still don't don't guess as fuel for Jetpack and tractors?

[brisk smelt]

for jetpack, even like 10 is enough

i dont ever recommend using land vehicles.. trains almost always better

[wind spade]

largely depends on your recipe choices. Better to make things when you know you need them, than guessing a ratio in advance

[brisk smelt]

general rule of thumb

almost always applies to the calculator production chain

[wind spade]

general rule of thumb differs from player to player

[brisk smelt]

i have never needed more plastic than rubber in one factory

[wind spade]

many people say that they did, so 🤷

[brisk smelt]

i am just gonna assume they are making an efficient factory if they dont say anything

[modest jackal]

ok i'll also calc other factories together

[wind spade]

choice of rubber vs plastic has nothing to do with efficiency though (not to mention that "efficient" means nothing if you don't specify what kind of efficiency are you talking about)

[brisk smelt]

this one doesnt matter, you are gonna make more anyways later in phase7 with fuel loops

[wind spade]

you can do that in phase 5

[brisk smelt]

with normal fuel recipe? hell no

[wind spade]

with diluted packaged fuel

[brisk smelt]

you can do that?

[wind spade]

it's a recipe that has same HOR->fuel ratio

[brisk smelt]

oh yeah it exists

sounds like a lot more work tho

[brisk smelt]

the calculator almost always gives you 2x the rubber than plastic if you give it p5 spelevator products

[wind spade]

which one?

[brisk smelt]

yours

[wind spade]

depends on selected recipes 🤷

and depends if you're the kind of person that optimises for weighted raw resources

[brisk smelt]

yeah, the recipes it selects gives that ratio

[wind spade]

that's just one way to optimise your production and definitely not the "only correct one" or anything

[wind spade]

nah it's super easy, you put 2 packagers and a refinery in a blueprint, loop the belt around, put 20 canisters in the loop, save the blueprint and then just spam that blueprint (which is essentially the same as placing blenders)

[brisk smelt]

hmm yeah that works

[prisma kraken]

yeah, but the blenders look cooler 😉

i'll use both dilluted fuel recipes depending on circumstances - like early power isn't bad to build with a packaging loop and jetpack/vehicle fuel is easier in the refineries, but i do prefer the blenders overall for their larger per-cycle yield and greater max capacity

[vast jungle]

hmm... I am looking at some fuel consumption rates for drones at the moment... regardless of the distance, the consumption of (in my case) rocket fuel per minute stays below 3/min... did I missed something?
(trying to decide how much fuel I need to play with drones a bit)

[versed violet]

Mergers do round-robin on inputs. Splitters on outputs.
The input one can be exploited to build a belt limiter - merger with 60 belt out and split, route two split belts back to merger -> limits total throughput to 20 (1/3rd of mk1 belt)

[unborn ermine]

Late to the party but never use that number raw, 25/6 is the fraction for the Rocket Fuel burn rate, so you can do real calculations for it instead of funky rounding.
(and suffer much more efficiently later)

[livid turret]

50 canisters, tested, less than that and you get starved waiting for the canisters to loop

[oblique hollow]

depends on belt speed

slower belts waste more of the canister's time in transit

[livid turret]

Also a great BP to stamp, you can even toss 2 OCed fuel gen in a 5x5 BP and get power with less hassle if you are doing regular fuel

[livid turret]

I mean refinery, packing, unpacking and feeding the generators all in one BP

[oblique hollow]

generators is usually better in a seperate BP imo

[molten lintel]

ty

[amber umbra]

I find it amusing that in a packaging loop for fluid transfer, the individual fluid packages are essentially itty bitty trucks/trains.

[livid turret]

All depends, I find it easier to just stamp the bigger BP of packaged diluted to power or the smaller BP just for fuel

[unborn ermine]

np, I loved finding that out.
I ended up making a BP for 4 fuel gens, and 240% clocking worked flawlessly after I looked at the burn rate + pipes divided into modules of 4

(thats if you have pipes holding less than 600)

[molten lintel]

burn rate stays linear when overclocking?

[unborn ermine]

Yep

[molten lintel]

ok that's super good to know

[amber umbra]

Idea is you can think about the amount of cannisters you need to service a packaging loop in the same way as trucks/trains. As the total distance increases, you need to add more packages/trucks/trains.

Guess they're called containers not packages.

[fierce vault]

I’ve explored maybe 50% of the map so far, I like travelling out to find resources I haven’t got yet and researching them in the MAM

[worldly granite]

Nice, i wish i did that too, but there is so much work to do.

[patent blaze]

Is this good

[amber umbra]

yea

[patent blaze]

Main goal was splitting crude evenly, sulfur isnt that nice but its divided by 22.5

Close enough for me

[wind spade]

20 was working fine with short fast belts

[pallid knoll]

@brisk smelt I think my assemblers are tight enough right now? Only need to optimize the rows between smelters & constructors. how would you lift them? I had something in mind but idk

[brisk smelt]

yeah that looks good

verticality is your best friend in this game

[nova mango]

I love the method where you use stackable conveyor poles to keep all input/output directly over the machine. Specifically Krydax, on YouTube, style has been inspirational

https://youtu.be/B-J9QeBpy_k?si=Ku5aqY2dQvYRRpkw

[violet halo]

That's my favorite way to do it, works with manufacturers too.

I usually hang the ones from the ceiling for it though.

Lets you put the assemblers closer together.

[long shuttle]

do refineries provide headlift equal to their height?

or is it no headlift?

[violet halo]

10m for machines that process things.

[long shuttle]

alright thanks

[fringe seal]

deranged idea: unpowered pumps are functionally the same as powered pumps against gas, as headlift means nothing for gas
and they do forcibly absorb gas from one end and expel it the other end, even when unpowered

[livid turret]

Feels like an unpowered pump will work more like a valve

[fringe seal]

test it yourself
in a half-filled pipeline without any incoming input or output, plonk down a pump in the middle of it

[livid turret]

Have you tried using a valve to see if it has the same effect?

[fringe seal]

yup

it left some amount of gas on one side

unlike the pump, which completely emptied one side

[vapid gorge]

except uppowered pumps might still stop back flow

[vapid gorge]

there's also a weird buffer gas trick which may require the pumps to be powered I think

[livid turret]

Pipes are like dark and forgotten knowledge, let's hope one day CSS get a QoL update to help them be more intuitive

[brisk smelt]

i love watching new people tear their hair out though...

[vapid gorge]

they really aren't though, you just can't treat them like belts and they have a few basic rules

people just do really dumb things with them

[gloomy flax]

well I rebuilt my fuel factory (thanks for advice the other day) and as I make it pretty I notice SOME OF IT ISN'T QUITE ON THE FACTORY GRID FLOORPLATE. I think i'll just walk away and leave it

[vapid gorge]

basic pipe rules for reliable systems #math-and-meta message

[livid turret]

Yeah but those are unwritten rules, the novice expects to run a pipe with X amount of liquid and that the consumer should consume X without issue but then you have to mess with flow and fullness to achieve that

Or some other shenanigans to run at 100% efficiency with pipes. Sometimes it's too easy to break a system and the fixes are not conclusive regarding the mistake you were doing

And then, when you think you got liquids, rules change again with gases

[prisma kraken]

I think to that list you should add some variation of 'full pipes are happy pipes'

[vapid gorge]

'flood the whole system' is in there at 3.

[prisma kraken]

maybe bump that up to rule 2 or 1, lol

just because it often helps before reconfiguring a build

[vapid gorge]

ehhh.. I don't think it's unreasonable for some mechanics to need a bit of a dive. Like trains. They look simple and straight forward but how many people need to dig through tutorials?

honestly the only thing I think they could realistically change about pipes and have it be a big impact is just have them be mono directional like belts.
Trivialising the whole system.
I hope I'm wrong ofc, but I'm dubious.

[prisma kraken]

tbh, probably a little bit of a hint about it all when you get to coal power and need 360 in a 300 pipe would be helpful like 'this is a problem that you need to solve and understand, here's a carrot for solving it'

[vapid gorge]

and those rules do make sense once you think about it knowing that fluids are bi directional, it's just people won't think about that and it's consequences

[vapid gorge]

except you don't need 360 in a coal system

only if you choose to make groups of 8 which is convenient for different things

and you could still have groups of 8 fed by a mk2 belt and 2 distinct pipe systems

[vapid gorge]

I think the main issue with that is people jsut not looking at mk1 pipes and seeing there's any flow limit

so many people that have coal issues just don't realise this one fact

[prisma kraken]

ofc, but if you build coal differently than that, you probably are either on your first playthrough and don't want to get the max out of it, or you're on your xth playthrough and trying to do something fancy

[vapid gorge]

or you go 'ok I can fit 6 coal gens on this one pipe, I'll do that, and have another group do 2'

it's just that, as as community, we've developed our own standard

[prisma kraken]

the main issue is that their behavior is non-intuitive and there isn't a published set of rules about it all

[vapid gorge]

of 3:8 groups

[prisma kraken]

after 1000's of hours in the game, i have very few problems with pipes and i know how to diagnose and fix the problems i do have

they, in fact, do work as designed

[vapid gorge]

honestly I would say most of pipe behaviour is intuitive if thought about.

It's things like valves , VIPs, and weird quirks you don't need to involve yourself with. Everything else is logical as long as you know about

Gravity Priority.
Fluid Goes Both Ways.

Maybe people assume there's water 'pressure' when there's not?

[prisma kraken]

yeah, i do think a lot of the non-intuitive is that the simulation imperfectly models real fluids

but it is also complicated by the fact that buffers and valves exist and create even more counter-intuitive behavior

[livid turret]

The flow and fullness is kind of a pressure system, if your pipes are not full, you cannot move liquid/gas fast enough

[prisma kraken]

and i'll add that there is still at least one bug in how headlift works

[vapid gorge]

ehhhhh... except liquids flow to empty spaces, even if on the other side of that empty space is fluid coming from the other direction to also fill

[vapid gorge]

the weird fake head lift thing?

[prisma kraken]

yeah, sometimes pipe networks remember old headlift and a recalculation of it for the pipe network won't happen

[vapid gorge]

yeah that's a weird one. Haven't seen it myself in game yet though

[prisma kraken]

what the order of build steps to trigger that is still elludes me

[vapid gorge]

or at least I properly place my pumps well enough that I don't see the fake head lift and dont' have any negative effectgs from it leaving suddenly

[prisma kraken]

I had an instance of that in a little test system build where i had an extractor 'remembering' the network headlift of 50+ meters and was able to fully fill an elevated industrial buffer without any pumps

wasn't until i deleted and rebuilt everything that the memory of the headlift disappeared

[vapid gorge]

apparently it can just go on new load?

[prisma kraken]

i have no idea, maybe the fluid itself remembers its headlift, idk

there's some problems that i see that i just roll with and move on and toss on the 'don't build it that way' list

but that really is the core problem with the fluid mechanics in the game, expecting someone to have the hours in the game under their belt to know to shrug and move on is actually pretty poor game design

i don't expect we'll see it changing at any point from now on however

[light fox]

is this efficient use of space

[livid turret]

A simplification of fluid simulation would be less frustrating, not fun having to debug something with no clear rules and no other indication than one or two consumers are underperforming.

[vapid gorge]

again, clear rules #math-and-meta message

the more points of input along a manifold the more points of instability
loops and flooding, can, to a point, ignore this

you can also do short systems that don't need loops ofc

and yeah, I think the only 'solution' people could see is just turning pipes into Liquid Belts.

which would absolutely suck

[livid turret]

Those are deduced rules because it worked by trial and error not because there is a clear explanation of how the simulation is working under the hood. Also, you need to reach the forum to get that knowledge or a video or the PDF guide because the game doesn't explain anything besides how much liquid/gas you produce and you consume

[prisma kraken]

i've found that when i try to build extremely compactly and the logistics gets to be a twisty mess, it becomes really difficult to change stuff later, so at this point i often just give myself a lot of space for logistics. your time isn't infinite and space in the game is practically infinite. which is more important to you is a judgement call that no one can make but you

[vapid gorge]

no those have been tested thoroughly. It's just the game doesn't doesn't preset that direcly.

which if that's all you want they could give you a quick image explaining fluid goes both ways I guess?

but I don't have an issue with some aspects of a game needing to dig. Why have everythign be trivial?

[prisma kraken]

as an adjunct rule i've learned based on my experience in the game, trying to be 'clever' often isn't so clever

[livid turret]

Take for example the lady hawk buffer for trains to pull a full 600, the process to achieve that is painfully iterative because you don't have any feedback on what could be wrong, you are just mixing and matching until it works and then you call it a rule, not my cup of tea

While if the game gave us hints about backflow and what could mitigate backflow or why something that should push X is not moving X then you could try to fix it without going in blind

At least headlift is written in pumps (current value)

It should be on producers too

[fringe seal]

btw, prefill + input looping + output looping + watertower made my petrochem factory almost at 600, but not quite

[vapid gorge]

you mea nteh gas thing?
oh yeah but if you just want more info on that its 'gas tries to fill space equally' which is the culprit

[fringe seal]

most of the improvements came from the water tower

[vapid gorge]

depends on the layout.

cut the input looping and water tower

pump does the same job as a water tower with less extra connections

[fringe seal]

it started to work almost at 600 with 200 headlift

50? doesn't cut it

[vapid gorge]

as long as you have enough head lift it doesn't matter. Your layout was probably very hostile to flow

[fringe seal]

input looping improved the throughput but failing far from 600

[fringe seal]

but it mattered, that is the problem

[vapid gorge]

except there were obviously core issues that you were ignoring.

[fringe seal]

and it's just 8 refineries together in a loop

[vapid gorge]

I would have to see the whole layout to comment.

[fringe seal]

your "underflow loop" of course

I can upload the save

[vapid gorge]

and it's extremely common for cause and effect to not be obvious when tinkering with pipes because often 'time' does a lot of work

just over head shots of the system is fine

[fringe seal]

well then

it's this setup

[vapid gorge]

so often people have said 'X fixed my pipes because it worked afterwards' when it was likely something they did 10 minutes ago

actual image sorry, elevation changes and shape can be important.

[fringe seal]

there's several days of delay between the points of time when I tried different fixes

[fringe seal]

sure

[vapid gorge]

assuming the 2 blue inputs don't merge, pumps here are sometimes needed
since fluid flows both ways if 1 branch sucks fluid and there's an empty spot near the junction that can stutter it. pumps can help keep directionality in these situations.

so many pipe issues just come from people not expecting backflow, which even knowing fluid goes both ways is not too surprising tbh

making branches is one of the 'to avoid' on the pipe rules list, but like most of hte rules are flexible if you know how to work around the issue

[fringe seal]

I have not touched this system for at least 10 hours

[vapid gorge]

yup bottom feeding

[fringe seal]

mm

but it's your technique

[vapid gorge]

whats going on over here ?

[fringe seal]

I fully expect people to blame literally everything on bottom feeding

[vapid gorge]

are these systems linked in anyway other than being branched?

[fringe seal]

there are three completely separate loops, each supposed to take 600/min

just it wraps around the building

however, it is completely flat

[vapid gorge]

this looks like you have 2 branches from the same input pipe?

[prisma kraken]

i'm not sure if it helps the discussion, but re: bottom feeding, i build most of my pipes as variations on what is pictured, and i really have very few problems

[vapid gorge]

that's not bottom feeding though

[prisma kraken]

something about feeding downward into machines really solves quite a few problems

[fringe seal]

I did it intentionally, given CobaltOfDoom had good results using it

[vapid gorge]

Well, I'd try pumps right before the start of each manifold and flood it, to help keep directionality.

though I will emphasise that I never recommend bottom feeding. It's just a thing that is possible with care

[fringe seal]

and this random pump is, essentially, removing the water tower

[vapid gorge]

yeah but it's for directionality, not head lift, which is far more important

[fringe seal]

I already flooded the system, but it balanced to sub600

[vapid gorge]

yeah but every time you make a fresh 'reboot'? flood it

[prisma kraken]

one note: if you install a pump, don't just snap it on the pipe, position pumps and then build the sides of the pipe

[vapid gorge]

or place it on the pipe and rebuild the pipes attached to it

[prisma kraken]

yeah, if that helps in the placement, sure

[fringe seal]

initially constructed with loops, flooded it, sub600
added input loops, flooded it, sub600
added water tower, flooded it, sub600
though, each fix showed some improvement over the last

[prisma kraken]

when you snap things onto pipes (or belts), it just invites problems

[vapid gorge]

but yeah - you can do things liek this have have them run, but these are all individual pipes with their own manifolds. And pumps right before hand

[fringe seal]

I will remove input loop on one of the line, and place a pump right before the loop
and flood
knowing fully well what will happen

though

the thing is that even the line that has only 1 pure node also shows sub600

is distance a problem?

but this length of pipeline is bounded by pumps in the front and on the back

[fringe seal]

and of course, remove the length of the pipe, snap pump to foundation, rebuild the pipes

and it's already doing 597/min

just to be sure that it is not the water tower problem, I'll also directly connect the input pipe with the loop

[vapid gorge]

distance, itself, isn't an issue, but the longer you have a pipe the more points where something could go wrong.

[vapid gorge]

"just to be sure that it is not the water tower problem, I'll also directly connect the input pipe with the loop (edited)" - wdym?

[fringe seal]

you told me to remove water tower so I did

[vapid gorge]

oh cool, was just checking.

[fringe seal]

and now we wait

this loop happens to have only one pure node as the input

so, no input loop, just a single pipeline leading to an output loop

with some pumps in the middle

[vapid gorge]

are those floor holes used or decorations?

[fringe seal]

decors

every single one

[vapid gorge]

this is why I don't recommend bottom feeding xD I would have been suspicious of those floor holes

[fringe seal]

I'm waiting for it to dip again

[vapid gorge]

and my bottom feeding loop is tested to have the input pipe be 1 segment

[fringe seal]

it's doing fine for now

[vapid gorge]

don't worry, I too, am naturally suspicious of pipes

[fringe seal]

I went through full 5 phases of grief just because of it

[fringe seal]

as soon as I say it, dips

[vapid gorge]

yeah look, bottom feeding isn't for hte faint of heart

where are you looking for the dip ? the main line?

[fringe seal]

the pump

the pump right before the loop

[vapid gorge]

ehhh... let it run for 10 minutes

see if anything starves

[fringe seal]

sure

[vapid gorge]

like the cat in the box. If you don't check it can't be dead right?

[fringe seal]

I am slowly realizing that I did this to myself

[vapid gorge]

and this is why I don't recommend bottom feeding. It's not super reliable unless you're supre strict about building and I don't consider that a robust system

[fringe seal]

with water tower it's ~95% average efficiency

I just may give up

[vapid gorge]

yeah but a tower doesn't help with back flow within the manifold. Only the loop really does that.

[fringe seal]

you haven't seen the "370 meters of headlift" thread?

[vapid gorge]

Like. It sounds like it altered the flow in some way that made it less unstable. But not nearly enough to actually overcome the core flow issues

[vapid gorge]

I have not

[fringe seal]

#1299098801897738260 message

[vapid gorge]

I am dubious

[fringe seal]

and yeah. the farthest refineries from the loop input are starving

[vapid gorge]

the furthest? that's odd, generally should be the middle ones

yeah if I was in your save I could probably tool around with it but it's very busy visually and hard to see everything

not enough brain for the day and my cupcakes aren't turning out how I'd like

[fringe seal]

ah yeah ig it is the middle one

[fringe seal]

well fair

~597 I can understand, in the similar vein as the 1200 bug
400? something is clearly not working

[jolly furnace]

are these red outputs correct?

[fringe seal]

should be

[vapid gorge]

yeah that means multiple refs are going to be starving at once pretty consistently

[jolly furnace]

splitters and mergers dont effect speed or anything right

[vapid gorge]

not unless you're going over the throughput limit of the output belts. and then that's really the belt effecting the speedds

[jolly furnace]

👍

[vapid gorge]

personally though I'd probably just process each original iron belt and merge the output as needed

[jolly furnace]

wdym process?

[vapid gorge]

do whatever construction step you're going to with the iron belts

and clock them as needed to drop their products on the belts you want

[jolly furnace]

ah okay

[vapid gorge]

clocking is the single most powerful logistics tool you have, don't forget it

perfect for controlling what goes on a belt

[jolly furnace]

I havent unlocked clocking yet 😅 , but I think it should be soon

trying to design my first steel factory to complete the second launch

[vapid gorge]

ah yeah, the MAM is like the most important building you got. research all that asap

[jolly furnace]

oh its within MAM?

ig ill look it up on the wiki

[prisma kraken]

its really worth trying to work at all the available research in the MAM as you progress, everything in it makes life easier and also opens extra recipes in the HD research pool

in the early game, i sort of use the MAM unlocks as the decision making arbitrator to choose what i need to do next

i.e. if there is anything I can unlock in the MAM at any stage of the game, that takes priority over other things

[jolly furnace]

the only things I have are the rebar gun and inhaler rn

[vapid gorge]

you're doing yourself a diservice then, so many goodies in there

even if it was just all for the blade runners

[jolly furnace]

ah okay, ig ill work on it more then

[golden smelt]

@vapid gorge uhhhhh the fuel broke again

[prisma kraken]

yeah, blade runners are such a QoL thing - being able to jump higher and run faster is +1

you kind of don't realize how nice they are until you restart the game and have to start building again without them

[vast jungle]

the only moment they are annoying is when you build in tight spaces and jump OVER the next splitter instead of ON it... of course as soon as you get the hover pack this is kinda moot 😄

[vapid gorge]

show me where the fuel is hurt 😛

although I'm relaxing and having dinner so replies wlil be sporadic

[golden smelt]

dont worry about it, I ended up stabilizing it (for good i hope) so enjoy your dinner

though there are some hiccups every now and then

[leaden cosmos]

Its never for good 😅

[golden smelt]

one can hope

[vapid gorge]

Mine are 😛

[vast jungle]

D(P)F based fuel can be extremely reliable...

[prisma kraken]

my pet rock is as well!

on the subject of fuel power reliability, i'm a little surprised at how stable rocket fuel has been for me. how i thought to build it pretty much worked the first time

[jolly furnace]

I am trying to produce 60 steel beams per minute. I am trying to draw out how to arrange the Foundries I need to produce the steel ingots for the beams.

In the diagrams the 45s represent the Foundries, and the 60s represent the Constructors they are inputted into.

Is it better in this case to take two of the Foundries, split the 90 Ingots/min into 6 - 15 ingots/min and distribute it amongst the other 4 which connect 1-1 with the constructors (left image)

or is it better to sum up all the 45 Ingots/min into one conveyor belt and let the Constructors pick up the ingots from splitters (right image)

[charred saffron]

Collecting and redistributing all ingot is probably easier assuming you have the belts to support those numbers (I think it's a full Mk3, so you should be fine)

[jolly furnace]

yes it is exactly 270

okay thank you

[charred saffron]

Yeah, gathering everything and redistributing it through a manifold is probably way easier, otherwise you're likely to get messy belts trying to load balance everything

[outer vale]

6 to 4 wouldn't be too bad to balance at least. Merge 3, split into 2, and repeat for the other half

[jolly furnace]

Yeah I could do this but I would want to try and keep the waste from this into the rewards machine

but ig if I use the manifold I cant really do that

[outer vale]

why would there be waste?

if anything you'd sink the final product(s)

[jolly furnace]

there is 30 extra

[charred saffron]

They're producing 270 steel ingots and only using 240

[outer vale]

oh, why?

[jolly furnace]

and I have no clocking

[jolly furnace]

wait ig I can just collect at the end

if i is using it perfectly the 30 should just be spit ot at the end

[charred saffron]

(Unlock clocking asap, you'll thank us later, trust me), you can still sink excess with manifolds. Load up every constructor making beams with steel ingots. Once they're carrying a full stack of ingots and the manifold is filling up, you can connect the last splitter to a sink. Only the excess will lead to the sink since everything else will be claimed by the manifold when there is space again

[jolly furnace]

haha yeah you were typing what I just thought

[jolly furnace]

how do you get those purple boxes to unlock it?

[charred saffron]

Purple boxes?

[jolly furnace]

nevermind itll just be easier for me to look it up on the wiki

[charred saffron]

Researching overclocking costs a few power shards, iron plates and wires

and unlocks clocking as a general mechanic, allowing you to underclock as well

[jolly furnace]

so im assuming under/overclocking is just for reaching 100% efficiency easier?

instead of splitting and merging things

[charred saffron]

100% input use/machine efficiency and/or saving on building a few machines at the cost of some power shards, yeah

In the rare case I make screws with the normal recipe, for example, I always clock the constructor making screws to 150%, so it takes 15 iron rods. The constructor making iron rods at 100% outputs 15 rods/min, so that way I have 100% input use in that line

(I rarely do this bc of alt recipes but it's a simple example I know off the top of my head)

[jolly furnace]

oh right for screws ig that is a good example

[outer vale]

it also lets you adjust builds for energy efficiency (more buildings at lower clocks) or space efficiency (fewer buildings at higher clocks).
Or just space savings with power generators
Overclocking pairs well with another thing you'll unlock from the MAM

[jolly furnace]

for the recipes is it random what hard drive gives what recipe choice?

[outer vale]

yes

[jolly furnace]

isnt more building at lower clocks just the same?

total energy

[charred saffron]

It's random based on what you're capable of making and what isn't already unlocked or waiting in a researched drive already, yeah

[vapid gorge]

It’s slightly less but honestly so little less it shouldn’t be an issue

[outer vale]

clocking of production buildings isn't linear in power usage, no

[charred saffron]

According to wiki.gg this is the formula for power usage based on clock speed

[jolly furnace]

oh its not linear? thats cool

[charred saffron]

It's just barely not linearly due to that power, but you need some seriously power-demanding machines for it to make a big difference

For example: a constructor working at 200% clock speed consumes 10MW (according to the wiki), two constructors at 100% each consume 8. That's only significant in the beginning when you're still working with biomass burners, as soon as power is automated those 2 MW are hardly noticeable

[outer vale]

eh, it's still 25% more power usage that's easily avoided

tradeoffs ahoy

[charred saffron]

The 10MW mathematically checks out btw bc 1.321928 is log2(2.5) rounded down, meaning it'd take 2.5x more power to double your productivity

[outer vale]

it comes down to preference in the end

[charred saffron]

It's not a big differences on smth like constructors or assemblers, unless you're on a smaller powergrid, but it's a tradeoff between saving building materialy and space, or power usage

[jolly furnace]

okay yeah idoube ill be building ona scale that requires me to think about that any time soo

but idr know how big factories get

[outer vale]

there are some cases where overclocking is a great idea in general

[jolly furnace]

rn im in like the early-mid game ig

[vapid gorge]

even if you overclock everything max it's only 33% more power overall, it's not a huge dif

[charred saffron]

It's a percentual increase, it becomes a larger difference as the base power requirement grows larger. Luckily, your power grid should grow quicker, cancelling out the growing power requirement of overclocking bigger machines

(for example, blenders cost 75 MW base power. +33 percent that's rounded to 100 MW. That +25 is more than the +3 of a constructor. However, at the time you start messing with blenders you have at the very least a normal oil fuel power plant producing like 250 MW per generator)

[karmic locust]

Is it better to have explosive rebar instead of nobelisks ?

rebar gun has much faster fire rate no ?

[vapid gorge]

haven't checked wiki but they feel like they do a lot less dmg

[karmic locust]

both 50dmg

[vapid gorge]

really? same dmg radius too?

[karmic locust]

no radius info

[vapid gorge]

huh, eh oh well, I never liked the explosive rebar

[karmic locust]

i dont care about dmg tho, explosives are for exploding structures

if fire rate is faster with rebar gun, why bother with nobelisks 🤔

[fringe seal]

so, this is what it takes to process one normal baux node @ 250% w/ mk3

[violet halo]

Same radius too I think. I prefer the cluster nobelisks over the rebar myself.

[fringe seal]

I've heard yes, it is exactly the same as the nobelisk

except one is thrown and one is shot from the rebar gun

[vapid gorge]

weird. I try to avoid clearing flora so don't use much of either

[violet halo]

You can multi detonate nobelisks, so you can lay traps.

[fringe seal]

is the Nobelisk like the Iron Rebar

basic ammo that is overshadowed by its variants

[fringe seal]

not sure on this info though

[violet halo]

Kind of, it isn't really overshadowed until the nuke nobelisk.

[fringe seal]

like, if I wanted to deal burst damage I'd use nuke or cluster
if I wanted to deal AoE, gas
and Pulse is fun mobility option
Default Nobelisk is cheap and that's it

ig that's why it has a niche though

[violet halo]

Gas is for when you don't want to clear the vegetation mostly.

I stopped using the rebar gun the moment the rifle was available.

[fringe seal]

I still have only stun ammo on the rebar gun tbh

probably useless now that I'm on Retaliate

[white bloom]

Concerning merger behavior, there seems to be a huge amount of stuff we don't fully understand yet.

Anyone can help with a fun little experiment. Connect three storage containers as input to a merger - one with MK2, one with MK3, one with MK4. Load each container with a different, easily distinguishable item. (e.g. I did screws, concrete, iron plates). connect the merger output to an industrial container via an MK5 belt and watch.

Q1: What sequence do you expect to see?
Q2: What do you actually see? Can you find any repeating pattern?

It's not as innocuous as you might suspect

[hushed kettle]

oh yeah, merging up to saturation limit causes the weirdest shit

[fringe seal]

not sure how to decorate this but this is a 600 baux to 600 alu in 3x4 foundations footprint

[steel knot]

Could look cool with metal beams or columns supporting the frame foundations

[fringe seal]

the frame foundations are there just so I can snap each module on top of each other

[steel knot]

Should be able to with blueprint build mode

But i like the look anyhow

[fringe seal]

I vaguely remember some sort of jagged geometric design posted in the design channel
and I had that in mind while building it but
alas I don't know how to speak this language of Satisfactory Building

[steel knot]

All my aluminum is refinery modules and smelter modules. Now that I’ve saved the day I might go back and modularize self contained ingots like this

[fringe seal]

I feel the frame foundations clipping to machines look awkward in that case so I removed most of it

but idk could work??

[steel knot]

Ahh I thought blueprint mode snapped to object outlines

Oh nvm

I reread that

[fringe seal]

blueprint mode snap to blueprint outlines yeah
and the top 2 refineries are slightly off vs. the bottom 2 refineries to give room for the chimney

[white bloom]

Merger and Splitter Behavior - A Closer Look

[balmy sandal]

so I'm building a stupidly large power plant (coal generators) 3 pure nodes with 200% MK2 miners resulting in 96 coal generators. all good there. This means I need 36 water extractors all good there too. Now comes the struggle. Get water to generators. I was sitting here doing the math when I realized I was working it out like conveyor belts. splitting and recombining till I had the exact M^3 I needed for each machine. I just had an epiphany though. I'm bad at fluids in this game. I just don't get it. but. If I have everything on the same network and no part of that network needs to exceed 300 M^3 at one time. am I right in thinking it will just .... work. assuming I give it enough time to fill everything up before I turn it on?

TLDR: I'm bad at fluid PLZ help

for instance. |I'm definitely going to need 15 pipes to get it from the reservoir to the power plant (ignore head lift as it's all downhill) but once its in the factory... I just put a pipe on each end of a manifold supplying 13 machines with a pipe connecting it to the next manifold to handle the excess. and so on till I run out of machines?

[wind spade]

run each pipe to separate set of machines

don't connect pipes together

[balmy sandal]

okay so still math it like it's conveyor belts? split it so each pipe has 270 so the math works out even?

[wind spade]

no, use whatever you have in the pipe

don't load-balance pipes

[balmy sandal]

Ooohh okay and overclock the last one so it has the full 300

then I have excess water enough to run the last set of machiens cause the escess dosen't stop anything

Thank you!

[short salmon]

quick question. if you got an industrial storage container with 2 mk 6 belts coming in and out all 4 belts should travel at 1200 right?

[magic island]

yes, full belts in should mean full belts out, if the speeds match

[short salmon]

then why is it i got 2 mk 6 belts coming out of a full industrial container, merging back together once they come out, and forming a single line that for some reason isnt getting above 1050

its traveling a significant distance, but it is the same belt

[magic island]

depends on the setup, I guess. are the input belts always completely full? they're not coming from a train station or anything?

[short salmon]

the container is beingg fed from a train, both lines connected, but both the train station and the container is always full.

basically got the container between so the 1200 line is always moving. but it doesnt seem to be working. basically got 2 lines from train connected to both inputs for the container, and both merge after the container back into a single line.

[magic island]

when a train stops at the station, do the belts run dry? or are the container+belts backing up and staying full all the time?

[short salmon]

the belt going into the container stops, but not the belts coming out. they always show full. but the refineries at the end of the 1200 line arent getting the 1200 promised

its being split between 34 refineries, but barely anything is getting past the 28th refinery

[fossil basalt]

Are you sure it's not just a manifold that hasn't filled up yet?

[wheat lake]

alright 480 bauxite using sloppy alumina requires 192 water/pm I set the extractor to 192 it overflows and stops working. i set it to 190 it sitll overflows and stops working wtf

[oblique hollow]

youre merging the water with byproduct water, arent you

[wheat lake]

yes

[oblique hollow]

thats the noobie trap

it only works if your system never stalls.
undo the merge and consume the water in dedicated alumina refineries

otherwise, if you miss coal or you dont have enough bauxite or your scrap is backing up, it goes to shit again

Very unstable, wouldnt recommend to anyone despite claims from others

[dark star]

I got it to work in a stable state with peiority input valves, it it was def more trouble than its worth. McGalleons solution is better and simpler than what i did or always makinng 100% sure you have cosntant supply and output to keep it going

[oblique hollow]

I dont trust priority jank is all

[dark star]

It hasnt failed me yet, but its usually a pain to set up

[oblique hollow]

I try to keep tabs on what people say about it and, from aggregated comments, i just cant really call them something "intuitive and reliable"

[wheat lake]

ok i see whats happening now it was scrap very slowly backing up

[oblique hollow]

yea and that tends to cause the alumina solution to back up, which eventually clogs the water pipes

[cinder silo]

Diverting the wastewater to concrete production (and switching that to become your primary concrete) is a valid choice.

[oblique hollow]

all fine and well solutions. You just have to know the risks / costs of using each

[cinder silo]

Overflow to sink from the concrete production shoild avoid anything blowing up in your face, past that send the output to encased beams, pipes, the cloud storage, a local store etc.

[vast jungle]

this here is just for delivering empty fluid containers to be filled with Rocket Fuel and shipped back... its as large as the RF factory attached to it

[steel knot]

Also burning the alumina waste water in coal generators was an idea I hadn’t thought of in my first refinery

Seemed like coal was easier to find than limestone in the red forest

But whenever I’ve had issues with VIP junctions and recycling it’s been because of some other math issue elsewhere and not the VIP itself

I know they get a bad rap

[oblique hollow]

coal is quite handy as you already need it for aluminum
that or coke

I know its kinda costly but not as bad as it could be

[steel knot]

Yeah in my case I was using electrode scrap right next to a coal deposit. Didn’t occur to me to use it until chatting here

[prisma kraken]

I've done what you're attempting before, and it is a 'never again' type project. That being said, i did pull away a bunch of lessons. the piping and water supply are definitely the hard part. If you've never done large-volume piping before, yeah it is a learning curve and will get a bit grindy. to help alleviate some of that, i'd highly recommend using mk2 pipes and tune clock rates and water volumes to make it so that an even number of generators exactly consume 600 water

[prisma kraken]

hmmm, in case anyone was wondering, you CANNOT drive a factory cart through the warp portals 😦

[charred talon]

how much power do i need?

like for endgame stuff

[fossil basalt]

More than that probably

[charred talon]

i'm teir 8 rn and i produce x3 more than my max consumption

[fossil basalt]

Are you running any particle accelerators yet?

[charred talon]

not yet, getting those set up rn actually

[fossil basalt]

A fully overclocked, looped particle accelerator making nuclear pasta uses 15000mw

[charred talon]

Interesting. I haven't gotten into overclocking yet

apparently i can overclock my reactors though

[outer vale]

that just saves space on power generators, it doesn't get you more power

[charred talon]

wym

[outer vale]

^

[charred talon]

i understand the context, but how would it not provide more pwoer

[outer vale]

one 200% power gen building (coal, fuel, nuclear reactor) is the same as two 100% ones

still takes the same amount of fuel to get the same power out

[charred talon]

so x2 the power = x2 the fuel?

[outer vale]

yes

[charred talon]

🫥

[outer vale]

like my rocket fuel plant for example. That makes 3000 rocket fuel, which is 180GW of power.
I could do that as 720 generators at 100%, 288 at 250%, or whatever other combination, but it doesn't change that it's 3000 rocket fuel into 180 GW power

[charred talon]

I see

[plush gulch]

Power production and fuel consumption is linear

On normal machines power consumption increases exponentially. 100 machines at 1% is in theory way more efficient

[prisma kraken]

i was able to get through the phase 5 delivery as a rush job with box factories on around 50-60 gw

[plush gulch]

Would be a fun challenge. Every machine at 1% clockrate

[prisma kraken]

the t9 stuff takes a lot of power, but you're really only going to be making 1 part at a time

[charred talon]

I'm still in MW

[charred talon]

I have to get to phase 4 rn. I did all the T8 stuff

Haven't even begun crafting the P4 items yet

[pure geyser]

ok dumb question, planning out a rocket fuel power plant. I know burn is 4.167/m. how's that impacted if I shard the fuel gen?

assuming it would be 10.4175/m at 250%?

[wind spade]

it will be 10.41666666666/min at 250%

[pure geyser]

cool just can't math then, thank ya

[wind spade]

the values you see are rounded ones

so you shouldn't do the math based on them

[pure geyser]

ah ok, yeah that was just in game calc math.

super helpful, thanks. that'll help a ton getting fuel gens down from 600 rocket fuel

huh, I thought 600 rocket fuel would result in more than 36GW. am I mathing wrong still?

[fossil basalt]

Looks right to me

[pure geyser]

I had planned it out previously and had more, forgot that I was planning with diluted fuel, heavy oil residue, AND nitro rocket fuel. 600 crude turns into 2400 rocket fuel. time to go harddrive hunting.

[fossil basalt]

576 generators whee

At least rocket fuel is a gas

[pure geyser]

I would not do this otherwise lol my "perfectly" balanced turbo fuel plant has yet to stabilize.

[latent seal]

has anyone played with merging/unmerging belts

like this

kind of purely for aesthetic purposes

but it seems a bit finnicky, and I'm not certain why

i feel like the belts are behaving like water pipes

[wise ledge]

how many per for maximum throughput? is it 4:1

[pure geyser]

wouldn't it depend on input speed?

[deft lichen]

this ratio has nothing to do with maximum throughput

1:4 is a often used ratio, but you can easily have 1:20 if the track is flat

more about max throughput here

!wikisearch tutorial:Train_troughput

[brisk shore]

https://satisfactory.wiki.gg/index.php?search=tutorial:Train_troughput

[deft lichen]

typo

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

[wise ledge]

what about maximum or ideal speed of trains?

[golden smelt]

@vapid gorge I almost fixed it, and then I made it worse 😭

[magic mango]

i have

[golden smelt]

basically 96/100 gens are running perfectly its just 4 turning on and off

im running 750 turbo fuel and splitting it, there are two pipes, one wiht 450 turbo fuel and the other with 300

the 450 is working perfectly, but the 300 is whats struggling

this is a sample of the system

[amber umbra]

@golden smelt I assume it’s still filling the manifold for the generators. Fast way to check is sloop the production, let it run ~10 minutes. Then de-sloop. Let it run ~10 minutes. See if generators still don’t run at 100%.

[golden smelt]

understood 🫡

[amber umbra]

If it’s a pipe throughput issue, the generators won’t all be running 100% even when slopped.

[golden smelt]

got it

its been slooped

and its running at 100%

all gens are working with no stutter

[unborn dome]

If I wanted to transport a mk4 belt's worth of caterium out of the swamp cave up to the plateau in the titan forest right above it, would that be something a truck could do if I built a nice road for it, or should I be assuming I'll need to just conveyor-lift it up the cliff?

Another thought: have the caterium brought out the cave's back door into the dune desert and have it loaded onto a train, since my main line runs nearby. I haven't been inside the cave yet, so I don't know what it looks like inside to know if this even makes sense.

[golden smelt]

depends on how far you're carrying, because if its close you could just conveyor it, if you think a truck is necessary then you need to think of how to get it fuel, and train is also a decent idea but you have to think of the distance you're taking the item

[unborn ermine]

Distance really isnt an issue, if you need more, just add cars, if you STILL need more, add a new train.

[golden smelt]

me personally I would just conveyor lift that

[unborn ermine]

Yeah a nice conveyor pillar would be simplest

[golden smelt]

im just thinking is it worth the time to set all that up if its not that far away

[unborn dome]

Yeah I'm already running a train to there from the north to bring plastic and rubber, but getting a train out of the swamp nicely didn't look fun lol

[unborn ermine]

Funny enough theres a nice right angle aready built into that cliff

[unborn dome]

For a conveyor pillar?

[unborn ermine]

Just a nice and clean setup

[prisma kraken]

they travel at a nominal speed of 120 kmph. throughput is more about how many cars you have passing a fixed point on a train route than the train's actual speed

[prisma kraken]

in your surveying of that area, is there enough space near the cat nodes for 7 water extractors?

...i'm guessing the swamp there is all fake water

[unborn ermine]

I think you can stuff a few in there, its been ages tho since I tried
This is @unborn dome's project

[prisma kraken]

ahhh

i'll be buiding out that direction soon™️ , just wondering what it looks like if you knew 🙂

[unborn ermine]

OH

I remember I DID check one thing

theres a TON of water at the back end (if thats the front where the nodes are)

[prisma kraken]

I cleared out the cave a while ago, but didn't have in mind exactly what needed to fit

[unborn ermine]

Theres way more water than you will ever need for pure caterium

[bleak ivy]

is there any use to the petroleum diamonds recipe?

[wind spade]

yes

[bleak ivy]

that i can tell, its less oil-efficient than just using the oil one, and it uses more accelerators

[wind spade]

gives you a byproduct

and has more steps if you want to sloop

[bleak ivy]

fair enough

[unborn dome]

Yeah not sure, I haven't been down into the swamp itself yet

[prisma kraken]

haha, g'luck

...there's reason i'm asking instead of checking it out myself 😉

[unborn dome]

I saw a Reddit meme earlier, that the best thing to do in the swamp is...leave

[prisma kraken]

once you have the tech to handle the big spiders, it still sucks, but before you have something that does a lot of damage, man, don't go there

[unborn dome]

I've only got regular rifle ammo...

So I'm hoping to just clear out the caterium cave and gtfo

[prisma kraken]

the spiders down around the turbomotor crash site love ganging up on you

its an endgame area, the swamp, titan forest and pink forest are all full of nasties

[unborn dome]

Oh huh, I've done quite a bit in the titan forest, but didn't realize it was on par with the pink/red forest. I'd assumed that one was harder than all.

[prisma kraken]

i actually think titan forest in places is worse than the swamp

all 3 areas have extra nighttime spider spawns

kind of the way to tackle them if you don't want to spend hours on the walks of shame is to just clear them out at daytime and gtfo before it gets dark

[unborn dome]

Hoverpack has been my solution lol

And dropped a miner connected to nothing on one of the bauxite nodes to prevent spawns near it

Nothing aside from power that is. No belt

[livid turret]

You can cheese the big critters by creating a cage BP

But critters get old fast after a run for drives/sloops/orbs

[unborn ermine]

Or you just drive in a car

[livid turret]

Establish dominance ☝️

[prisma kraken]

i just kill em, alien protein burgers! omnomnom!

[unborn ermine]

I still love how cheesy using the explorer is now, no radiation or gas damage

[livid turret]

With all the ups and downs, driving has always been a no no for me

[unborn dome]

Yeah I've almost never used the explorer, I kept flipping it. 😦

[prisma kraken]

yeah, i rarely use vehicles for anything but short range interfactory logistics

[unborn dome]

(Yes, probably a skill issue on my part lol)

[prisma kraken]

it used to handle a lot better

like it was really screwed up in update 8 after the engine upgrade; 1.0 got it back to driveable again, but it really lost a bit of its utility btw the storage nerf on it and its lesser handling now

it is tolerable to drive though

if i did a start on the north side of the map, i'd be using it for e-w canyon runs

[main shuttle]

kinda wish we could paint the crashed pods so we could easily ID claimed drive pods at a distance

[livid turret]

I always rush for the parachute and the boots at the start

[livid turret]

Build a watch tower next to it

[main shuttle]

i usually just put a mk1 power pole on it

[prisma kraken]

or just clear entire biomes at a time

[unborn dome]

I drop green checkmarks on the map next to them.

[main shuttle]

oh i'm doing a full clear now that 1.0's out

think i just have the deserts left to do

350hrs in

[prisma kraken]

i never really have a problem with it, if i grab the loot next to the crash pod, i can tell i grabbed the drive

[main shuttle]

i keep thinking i may have missed one

[prisma kraken]

i load my save into scim every now and then to check

[livid turret]

It's for when you are looking at it from far away

[main shuttle]

yea, my pea brain keeps looking at pods and thinking "huh, might not have claimed that one yet"

[livid turret]

Watch tower, they are tall and big enough to not be seen

[prisma kraken]

idk, after you have all the recipes, that kind of stops being a concern

[unborn dome]

Same question but with the quartz - truck it or belt it?