math-and-meta

[dusky dust]

I'm not trying to dissuade you, just saying that if you're impatient then targetting a 30-minute Elevator delivery is technically working against your timesaving goals. :)

[remote hollow]

i also get your point

[dusky dust]

I mostly just like the thought of trying to construct that graph for myself and seeing where the lines intersect. It's a thought I have basically whenever I decide on Elevator-factory delivery times. :D

[brisk smelt]

continious elevator >> :p

[dusky dust]

Gotta try to quantify how quickly I build factories and such, see if I can actually estimate any of that kind of thing with semi-reasonable accuracy

[remote hollow]

i hate balancing

and the pure iron recipe pisses me off

[mossy halo]

Just over/underclock for more beautiful numbers

[brisk smelt]

just make slightly more than 1200 iron per batch of refineries

thats whati do... plenty of iron in the world

[remote hollow]

See like I would do that but I don’t have many power shards

[mossy halo]

Underclock doesn't need power shards

[remote hollow]

Overclocking does

[mossy halo]

Then underclock

[remote hollow]

I don’t want to find more room for more refineries

[mossy halo]

That's probably 1 or 2 more

[unique cypress]

pure iron is the same as any other recipe if you just ignore having the refineries run 100% of the time

[remote hollow]

What can the highest tier belt carry

[unique cypress]

mk6, 1200/min

[remote hollow]

Bruh

[mossy halo]

Like in this example, except for extractors I underclock and use more machines to have better plain numbers and better handling with 120/m on manifolds without requiring manifold injection.

[unique cypress]

logistics get so much easier when you don't care about machine uptime

I just leave all machines at 100%, sometimes even place a few more than necessary when I have a BP that places 8 and I need 30

[remote hollow]

Hm

Thing is I need it balanced because

I have 3 coal belts with 480 each

Wait

[unique cypress]

the only time where you actually need a balancer is loading trains if you want maximum efficiency

[remote hollow]

I had to make one for my modular motor factory

I probably just need to take a break

I’ve been spinning in circles for a while

[limpid knot]

Underclocking saves more power than overclocking takes, right? Is it more power efficient to have two machines at 75% than to have one machine at 150%? Or is power usage linear?

[unique cypress]

1@150% uses 25% more power than 2@75%

[limpid knot]

OK, perfect. Thanks Kyo 🙏

[median hill]

Is one industrial fluid buffer enough for 40 fuel generators? I just beefed up my power last night and I’ve experienced some flow issues with fuel. I opted to cut everything off and let the lines fill completely including the buffer itself and turn everything back on slowly and so far it seems to be running smoothly. But I’m not sure if it will hold.

[unique cypress]

0 is enough

[median hill]

Expound.

[oblique hollow]

the buffer isnt something for smoothing out flow in normal production

if you have issues with flow not being sufficient and machine starving even after you filled everything or otherwise, you need to check the piping.

[potent pawn]

ok good idea

[pliant briar]

What is next power making thing after coal generators and when do you unlock it

[oblique hollow]

||fuel generators||

[pliant briar]

Oh ok, when do you unlock it?

[oblique hollow]

Tier 6

[pliant briar]

Alr ty

[abstract island]

are fuel generators worth using for power or is it better to use coal until nuclear is unlocked?

[oblique hollow]

coal is more tedious to set up than fuel gens imo

they are both weaker and take more space than fuel gens

and as for nuclear - many stick to fuel gens because nuclear just takes too much to set up

[abstract island]

okok so work on getting the oil fuel gen stuff got it
thank u

[pliant briar]

So wai, it's not worth it to get fuel gens and stick whit coal until nuclear or?

Or get fuel gens as soon I unlock them

[dusky dust]

You'll want to start building Fuel Gens in Phase 3; they scale out much more quickly than coal does

You have a lot of options for how to build out fuel gens. It's possible you'll want to unlock some more recipes before building a big fuel-power area (the "vanilla" fuel gen situation is a little tepid, though still nicer to scale out than coal IMO)

The Phase 3 fuel-power options boil down to: "vanilla" Fuel, Diluted Fuel (still just Fuel but you get a lot more out of your crude oil, with the help of a few alt recipes), Turbofuel, or (for the truly mad), Diluted + Turbofuel.

(There's also technically Liquid Biofuel, but few would want to rely on that for main factory power. :)

[hushed silo]

do yall now any calculator for sati plus?

[oblique hollow]

they have their own SF modeler version on their discord

[queen umbra]

If I have 240 coal and 240 iron I need 5.3333 fondery?

[oblique hollow]

question makes no sense - what recipe ?

[queen umbra]

Steel ingot

[oblique hollow]

then yes

you need 5.333 foundries

[dense sapphire]

Yes, you can find it here:
https://drive.google.com/file/d/1NV_NDdQMVwFH2a06eO3_Hbj__yNOe04h/view?usp=drive_link

[topaz hedge]

does anyone have the diagram to make convayor limiters? ex I have 300/min on one belt and 5/min on another one and I need to limit the 5 belt to roughly that.

the priority meger works, but it'll empty the 300/min belt before taking from the other one.. and I don't like the way it looks lol

[queen umbra]

I don't know but couldn't you do a manifold and underclock the machine

[topaz hedge]

no, not really.

[hushed silo]

🧙‍♂️

my pioneer

thank you so much

[unique cypress]

That's a general belt limiter schematic. b is any belt that has a speed >x, where x is your desired limit

[queen umbra]

How often do creatures respawn?

[deft lichen]

After 3 ingame days unless their spawns are blocked by nearby factories

[queen umbra]

Thanks

@calm flax your beam factory produce only 60 beams/min you think that enough

[calm flax]

yeah it is initially

[queen umbra]

Out of the 60 you make encased industrial beam and versatile framework

How do you split those like for personal usage

How much /min for each?

[calm flax]

You don't make vert framework constantly, so you make a priority of what you need

[queen umbra]

Ok so first stage of my factory I make beam second stage I make encased industrial beam and third and final stage I make versatile framework so its easy to demolish when I'm done building them

[calm flax]

Yeah, Or just turn the floor off with a power switch

Get the smart splitter sorted out... that way when its on it gets sent there, when its off it does not

[queen umbra]

Ok

I only need like one so I'll just buy AI limiter at the awesome shop

[calm flax]

just craft it with bench

Caterium should be a priority for better power poles etc...

[queen umbra]

I won't make a lot of encased industrial beam tho

Like 6 per min

[calm flax]

tripple that when you finish vert framework

[queen umbra]

You're right you're smart how much hour do you have in the game

[calm flax]

suddenly you move from mk3 -> mk4 belts which means your production use changes to encased from normal

[pastel canyon]

is it a good idea to put a buffer in a powerplant? as in, like, putting in like a stack of batteries to store excess/prevent sudden blackouts?

or is it not necessary/a waste of materials?

[hushed silo]

its good to have some and they are cheap but they arent realy necessary

how many to build? u can calculate that quite easly when u compare it to ur actual power draw

set a time u wamt them to last giving u room to fix things and gtg

but the only reason imo to build them in huge quantities is if you would go to lategame with unsufficient power generation and decided to boxfeed nucular pasta at 500%

[pastel canyon]

id say im doing ok for power generation at the part of the game im at

so i dont think ill need the batteries for now

[queen umbra]

I need a heck lot of concrete

[queen umbra]

I'm at phase 2 too and I'm at 1700MW almost going over

[topaz hedge]

Ty, I ended up building exactly that, just without the priority merger. I think it'll be good unless the belt stops ( and it shouldn't)

[lapis sail]

hey, i'm relatively new and want to use trains but it seems a bit more difficult to plan around since i've done nothing with them for my entire playthrough. how do I learn

[unique cypress]

The priority merger is needed in the general case, but most designs do work fine without it

It's only needed when x>b-x, so x>b/2. I.e. the limit is more than half of the speed of the used belt.

I think I've seen someone want such a limiter once, ever. Especially because unless you want more than half of your fastest belt, you can just use a faster belt and use a regular merger. And even if you do, you can just put a limiter on the remainder, and use a smart splitter with overflow. That's how we did this before priority mergers

[slate spoke]

I'm trying to get a train network going, what could cause trains to not reserve a path?

My fail point rn is a roundabout's path signals not registering any reserved paths, thankfully, none are showing an error

[wind spade]

experiment with them

[queen umbra]

Should I put rotor in dimensional depot?
I already have iron plate and concrete in it

[limpid knot]

You should have enough left over to have a dimensional depot for each ingredient in the game, and then some. Start with high use items like concrete and whatever you use for belts etc

[languid spoke]

So question i started at dune desert and im now needing to build my first Fuel Power plant which will be Turbo, my end goal is to have my factories spread out, should i build it at blue crator or closer to home\

[vapid gorge]

wherever is convenient.
though honetly Diluted fuel will carry you to nuclear or the end of the game easily

[half frigate]

I assume since ill be using mk2 pipes, I can feed 2 reactors from 2 water extractors instead of using 1 per reactor, but whats the best way to actually lay out the pipes, im gonna need pumps ofc, but like, in a 3x3 or 5x5 or 10x10. (Im trying to make this easy)

Do i split some of the pipes left and right and have some go down a middle path?

Im gonna have to replace down all 124 reactors anyway, I dont have enough proper space for splitters and mergers unless I do some illegal clipping shit

[vapid gorge]

overclock the reactors. 1 pipe per reactor. Real easy

If you don't , also still feed 1 pipe per reactor. makes your life easier

[half frigate]

Yea i figured

I have an idea of a platform i can make for the pipe supply but it requires any platform and floor pipe holes

[wind spade]

build on top of the ocean, don't lay blocks of pipes, extractor in front of power plant

[half frigate]

Damn, thats a problem then

Ive already got 124 plants down but I guess I will have to re-organize the reactors. sigh oh well, ive got the resources depot'd anyways

[wind spade]

I mean it's obviously one of many ways to do it, but it's imo the simplest, I see people spending days or weeks building huge blobs of pipes from water to plants for no reason

[vapid gorge]

it's probably worth it unless you're really married to the gens being there

[half frigate]

Im gonna give my idea a go first, if that doesn't work well enough, ill re order them

My biggest issue is the belting

I can probably use clipping with elevators so ill give it a try

[vapid gorge]

@main ginkgo

ok so you don't need a loop on the outputs of the water extractors, the risers do nothing

also don't merge the 2x 600 set ups. Keep them split

@main ginkgo oh also 6x extractors at 100% make 720 water, you're bottlenecking yourself

[main ginkgo]

One of those is ouputting less intentionally

Underclocked

[vapid gorge]

ok that's fine I guess, but still get rid of the loops, the risers do nothing and don't merge the two systems.

keep pipes as simple as you can.

I really wish I could stamp out the water tower superstition that keeps raising it's head

[main ginkgo]

Loops including the pipe manifolds at extractors?

[vapid gorge]

looping outputs doesn't do anything

often times, at high flow rates, you may need to loop the input pipes on machines

[main ginkgo]

Give me a minute

I already updated it

[woeful zenith]

There's like only a single scenario that I can think of where it would have some use, but it's pretty niche. The water would have to be lower than your factory and your factory would be variable in height for some reason, by a good amount. So it would just cut down on the use of pumps a bit in that case

But I think in like 95% of scenarios you just won't need it

[vapid gorge]

you'd still have to bring the fluid up ... also pumps are basically free

[vapid gorge]

for actual water towers. This isn't even a water tower though

[main ginkgo]

I love the flow rate

[woeful zenith]

Yep keeping it simple helps #satisfactory-memes message hehe

[woeful zenith]

You can get rid of the pumps btw

[main ginkgo]

Right

[woeful zenith]

Pumps only needed for going up in height

[vapid gorge]

if you're dealing with buffers you want the pumps there. To stop them from wrecking flow for them testing it

there's also some niche and moderately rare cases where pumps can help directionality in a slightly unstable system

[woeful zenith]

Oh yeah I forgot pumps limit flow to one direction and buffers need that since buffers are buffers

-# this is why we don't use buffers unless it's temporary

[queen umbra]

But I don't have many Mercer sphere yet I'm still really early in the game I just completed phase 2

[kindred sky]

I've only ever used buffers for byproduct fluid that I'm still building out production to use, and I still want the other products to keep producing.

[woeful zenith]

Yeah exactly, I did that too for my oil setup recently

Just for flushing it after a bit

[kindred sky]

Yeah me too

[woeful zenith]

Eventually hooked it up to something that uses the byproduct

@vapid gorge What's your theory on the video css made saying "we fixed fluids"(sorta)? Do you think they removed sloshing maybe

[vapid gorge]

eh no idea. If they have they'll have basically turned them to belts and be boring and pointless

Covered belts. Woo.

[woeful zenith]

Lol

I'm so curious man like what could it possibly mean. I doubt they'll touch headlift. Hhhhh I guess we'll just have to wait and see

[vapid gorge]

they'll probably just edit teh settings so that there's effectively no way for fluids to back flow

so, in theory, they'll still be bidirectional, but in reality there won't be any situations where that could happen

meaning they'll be belts

[woeful zenith]

Yeah possibly

[oblique hollow]

Again: no

Bidirectionality alone makes pipes special enough already

And that property, as i have discussed before, is not removed when you tweak them

[vapid gorge]

I'm talking about effective bidirectionality,

If something is allowed but never happens it effectively doesn't exist.

and you've never mentioned any situation that you've spoken about this where back flow will happen

[oblique hollow]

Back flow naturally happens during bottom feeding manifolds

Because all pipes need to rise equally

[vapid gorge]

I was under the impession when yo uwere takling about the edits that machines didn't stutter from bottom feeding back flow

[oblique hollow]

They indeed didnt - that doesnt mean there is no backflow in places where it should be

The pipes still fill somewhat unequally

And in those cases, the pressure differences cause some pipes to stop and reverse in order to ewualize their levels

[vapid gorge]

back flow would stutter the system backwards. There is effectively no back flow

like sure you can go one about how they are technically bidirectional in those circumstances but you look at the overall behaviour and you're essentially discussing belts

[oblique hollow]

Pipes cant magically reverse direction, in case you didnt know

[vapid gorge]

everything you've described about the tweaks has effectively described that as the same management of belts

[oblique hollow]

Except that the way you handle the coal gen 3 to 8 still applies

And literally everything else

[vapid gorge]

depending on how you set it up they're just mergers or splitters

Simpler belts since yo ucan have mergers be also splitters

[oblique hollow]

They are still both at once on the coal gen setup if need be

[woeful zenith]

By the way, it's really sad seeing none of the newbies know about this. Actually I was playing with some decent players a few days ago and they were like "your water is gonna get backed up here btw since they're all connected in one pipe" talking about my 3:8. Then in my mind i was just like bruhhh lol

[vapid gorge]

we now have merger/splitters that are 2 in 2 out like other people have wanted. But for pipes

[oblique hollow]

They have literally always been capable of that

[vapid gorge]

but now they behave as belts

for al intents and purposes

[oblique hollow]

What do you think that even means

Perhaps i am misunderstanding you

[vapid gorge]

don't worry about it. Bed time

[oblique hollow]

Some players have a harder time grasping that the limit applies to pipe segments.

If you sip fluid from a pipe network at one point, then you can add more fluid back to it later on.

[woeful zenith]

Yeah

I mean I guess I can't blame them I didn't understand it for a good bit too

[oblique hollow]

On belts you would think "well i cant add more items later down the manifold if the machines before that point need items"

Except that with pipes you can because of the bidirectionality

[woeful zenith]

True
Yeah I just thought the 3:8 diagram was everywhere on the internet, well it was back when I was learning, so I thought most people would know it from seeing it somewhere

Probably have to post it around a bit more lol

[oblique hollow]

Every day someone else learns of the 3:8

[woeful zenith]

Before I used it though I went 1:4

Which ofc meant overclocking the water extractor

And uh what was the other one without overclocking, 1:2 I think

But ofc 3:8 is most efficient

Wait let me post it in case any newbies are in chat. This is the setup, 3 water extractors and 8 coal gens

(and the positioning of where the water enters the pipe manifold is important)

[oblique hollow]

Alternative view with math

[unique cypress]

No, the most power efficient is 75@1%:2

[woeful zenith]

Oh dang

Mmm yes time to place 75 water extractors for my 2 coal gens

[queen umbra]

But you still can change it

[woeful zenith]

You can yeah as long as it's changed correctly

[oblique hollow]

just gotta do the math

subtract the demand from machines and you get "spare pipe capacity"

if your spare pipe capacity is big enough, then you can add more via another supply pipe

[vagrant lynx]

How's this looking for diluted packaged fuel?

Makes about 120fuel/minute

[unique cypress]

Skip the container and put like 20 canisters in the BP

[vagrant lynx]

The setup uses like 120 canisters/min since its overclocked

[unique cypress]

It doesn't really matter because they just get looped back

[vagrant lynx]

Fair. Now I just to figure out how to actually put multiple together without being spaghetti

It needs 120 water/min so one water extractor per fuel blueprint, and then I can put them all into a mk2 pipe to push down to my fuel gens l

[main ginkgo]

@woeful zenith

Not all extractors are active because this is a test facility not running recycled canisters

Also built around mk.4 belts (480/s), logistics will be less diet spaghetti when I reach aluminum

[woeful zenith]

Looks pretty good yeah

[main ginkgo]

I'm good at everything else, pipes just confused me for some time

[woeful zenith]

I think even mk1 pipes can work here, but as long as the pipes are consistent it doesn't really matter either way

[woeful zenith]

You're not alone there lol

[vagrant lynx]

@unique cypress how's this look with multiple next to each other? Should I reorganize it?

[unique cypress]

I blueprint my DPF setups in rows.

A row of 4 packagers in one BP
A row of 4 refineries in the second
And a row of 4 unpackagers in the 3rd

All of them have the belts/pipes connected and then they autoconnect front to back and side to side

They have foundations included because I do the loopback underneath, which also autoconnects

[fair grove]

hello so im building a turbo fuel plant and im making 760 a min but my pipes only carry 600? how do i get around this??

[languid laurel]

Don’t connect all the pipes together

Sry im bad at explaining

[fair grove]

no no i get it

[languid laurel]

You do? Bro must be good at Pictionary or smth because that was a bad explanation

[fair grove]

nah like i get that i sould not connect them all

but im trying to figure out an layout

[languid laurel]

Gl with it hope it works out

[merry sedge]

have you tried input divided by output?

760/x = some nice amount you can split into 2 or more pipes

[vagrant lynx]

Mind sending se example shots?

[fair grove]

yeah i will try thx

[agile junco]

So I realized something while doing more train math.

You basically only need loading and unloading buffers if the source and sink rates are exactly the same.

Granted, since there is variability in the system with train traffic and machines that produce in bursts, this is actually a bit more complex and a buffer won't hurt. But still somewhat interesting.

[crimson moat]

Not exactly the same, but when they are high or trip time is long you need buffers.

[agile junco]

If the source is greater than the sink, then it will fill up the train and and eventually the buffer.

[crimson moat]

If the source has nowhere to buffer the output items then it may clog and stop producing them between train loads, and you can't transport items which don't exist

[agile junco]

Yes, but that is only going to work if the buffer isn't itself full.

The buffer will be full if the sink cannot get rid of all the source items.

[crimson moat]

If the buffer is full when that happens, the line is achieving max throughput. If there is insufficient buffering, it may not be.

So the buffer may or may not increase throughput depending on the individual system

[agile junco]

No, not neccessarily max throughput, but max fill rate. the throughput of the station could be higher than the source rate.

[crimson moat]

It's still a problem in that scenario

[agile junco]

I'm just clearifying.

And I still don't know what your point is. If the source is greater than the sink the source buffer is full and useless. If the sink is greater than the source, then the sink buffer is empty and useless.

So you only need both buffers when source = sink, which was my original point.

[crimson moat]

Nah you do not understand. If there is less than 1 cart of items on the belt between the output and the stations, especially much less, then the machines won't be able to run and make your items while the train does part of its trip. They will stall and stop producing.

When the train arrives, there's less than a full cart of items ready to load onto it, so it either has to depart partially empty or delay departure. Both lower throughput.

[agile junco]

Departing partially empty isn't an issue if the achived throughput is still greater than your source rate.

[crimson moat]

If that is not the case and you are trying to move up to 1 belt of stuff, then yes, you can mostly get by with a splitter which splits that 1 belt into 2 station inputs WITHOUT the isc being present.

However in that case, 1 belt is not the maximum throughput. In many cases, max throughput is 1.5 - 1.85 or so belts.

[crimson moat]

It's lost train throughput, both for that train and for the overall network. If all of your trains drive half empty then you need to build twice as many lanes and spend nearly twice as much train power.

[agile junco]

Sometimes that's fine.

That's completely seperate from the train station backing up and causing your factory to idle and lose production.

[crimson moat]

sometimes, but it's substantially suboptimal and there isn't really a good reason to take the hit given that building an ISC instead of a splitter at your station is the only change neccesary to delete the problem entirely.

[agile junco]

WDYM, if I want to transfer 40item/min, I probably only need a partially full train...

Making it wait at the station might cause other issues, if I want to reuse the station for other lines.

I often do this to save space.

[crimson moat]

I am not here to argue why you don't mind having your train network be efficient, just to explain why it's a loss to throughput both on the train and on the network in many circumstances outside of what you said

[agile junco]

I'm talking about the buffers.

Yes, if you want maximum throughput you have to run exactly at the right distance and fully loaded. You cannot always control round trip time, but you can set the train to be fully loaded if you want.

[crimson moat]

I'm talking about the buffers.

Yeah. The lack of sufficient buffer room causes the train to depart either partially empty (less throughput per cart and per network) or late (less throughput per cart)

if you want maximum throughput you have to run exactly at the right distance and fully loaded.

You can max throughput at arbitrary distance

[agile junco]

But that limits the train and station to a single item.

The main issue buffers solve is to prevent idling in the source or sink side.

[agile junco]

You can max throughput at arbitrary distance
No. The optimal throughput for a station, for a given stack size and platform belt speed is at a spesific round trip time.

[crimson moat]

That idling is not just caused by mismatched rates, but by fewer than 1 train of items at either the departing or recieving end. Both ends need to have over 1 delivery of items so that they don't lock up before the next train comes.

[agile junco]

I feel like I'm not understanding what you're saying and you're not listening to me.

[crimson moat]

A specific trip time per docking. That means that distance scales linearly with train count.

[crimson moat]

ditto

[agile junco]

If you produce more than you consume, the buffer will be full.

Right?

[crimson moat]

The problem is that you won't be able to produce as much as you want to consume if your production stops before you have produced a full cart worth of stuff, because the belt is clogged.

[agile junco]

"production potential"

[crimson moat]

It will be full, but being full isn't the only criteria that you need to meet. The criteria that you need is that the amount of buffered items between output and train is higher than the storage space on the train cart, which can be tens of kilometers of belt with 500 stack items if you don't use any ISC.

[agile junco]

If I have a 1200 iron node going into a buffer, and that buffer into a station at 2400 (two belts from the buffer). There's no way it backs up... until the train cannot unload, because the sink station is only unloading 1000. Then the miner will idle regardless of the buffer.

If the buffer is full when the train is docked, that's the criteria.

[crimson moat]

Not full, there needs to be enough items buffered so that all input belts stay full until the train cart is full. When the train is docked, it should be recieving 2x1200/min.

that can sometimes be accomplished without an ISC, and sometimes that's unreasonable.

an ISC is bigger than a train cart at arbitrary item stack sizes, so it guarantees compliance

[agile junco]

when you say "input belts stay full" wdym. Like full speed?

[crimson moat]

When the train is docked, it should be recieving 2x1200/min

[agile junco]

Yes exactly. That's what I was saying....

[crimson moat]

If you put a miner into the train station without an ISC, the docked train will only recieve 1x1200/min so the loading step will take twice as long.

[agile junco]

If the storage unit is full at any point while the train is docked then it will not be getting input!

[agile junco]

??????

[crimson moat]

The miner outputs 1200/min.

[agile junco]

If I put a 1200 miner directly into the train station, as long as the train isn't itself backed up with items, I will only get < 1200 throughput regardless.

[crimson moat]

Exactly. The ISC makes it work at 1200.

[agile junco]

No

[crimson moat]

The ISC stores 1200/min of ore while the train is not there, so that the train can be input at 2400/min while it is there. That averages 1200/min.

If you don't have a buffer, the miner will stall and only be able to input 1200/min while the train is there, and 0/min while it's not. That averages less than 1200.

[agile junco]

It will fill up!

then it will be back to whatever the station thoughput is.

[crimson moat]

It will not

[agile junco]

How can that make any sense???

[crimson moat]

you have not outsmarted the entire satisfactory math community here

[agile junco]

If the station throughput is 1000 and the miner is producing 1200. where do those 200 go?

[crimson moat]

They're deleted because the miner is forced to idle because it has nowhere to output the items to.

[agile junco]

EXACTLY!

So remind me how the buffer magically gets you 1200?

[crimson moat]

So your throughput is 1200/min with a buffer, and 1000/min without. Ergo your initial statement is wrong, you said that buffer is not needed.

[agile junco]

No

We're talking about a station with a throughput maximum of 1200!

OK, let's use real numbers then. Assume 2x 1200 miners.

maximum possible single car single train station is 1793.05/min

A buffer does nothing for you here.

[crimson moat]

The buffer stops the miner from shutting down because it always has somewhere to put the ore. The buffer holds the ore during train transit, and then dumps it at double rate into the station.

With input and output 1200, the buffer never fills or empties, and 1200/min passes through the train cart.

Without a buffer, the miner shuts down sometimes because it has nowhere to put the ore, so it mines less than 1200/min, so the train transports less than 1200/min because the items don't exist.

[agile junco]

I'm well aware of this

[crimson moat]

yet you argue against it

[agile junco]

No I do not.

You are just not listening to me.

Again, if you produce more than you can sink, you back up and the buffer becomes useless.

[crimson moat]

You only need loading and unloading buffers if the source and sink rates are exactly the same.

I vehemently disagree with this, and apparently so do you. So we are both arguing against yourself.

[agile junco]

You might produce more because A) you consume less, or because B) your trains throughput is less.

In both cases your buffers would be filled up and useless.

My whole point is that sometimes you just only need one side of them.

Because your system is setup to idle no matter what.

Not that buffers altogether are useless.

Maybe I should have said "You only need both loading and unloading buffers...." but I did clearify above.

[agile junco]

here

I wish I knew how to communicate better to avoid this frustration, but I swear I'm trying to be clear.

[crimson moat]

The loading buffer must not stall production below the target rate

The unloading buffer must not run dry of material before the next unload happens

these are both required even with mismatching rates, or production falls due to transport bottlenecking

Both ends need a "buffer", in the technical sense of the word. It does not always need to be an ISC, but a single ISC is always sufficient

[agile junco]

I mean, like I said; it will never hurt to add a buffer. But it's not always doing anything.

[crimson moat]

In some circumstances, the internal 1 stack buffer of machines on input/output, the belt length etc may be a sufficiently sized buffer.

but it is buffer, and you need it on both ends

[agile junco]

if you are producing 2400 and pushing that through 2x 1200 belts to a single station, no amount of buffer will give you back that lost time during docking of production.

There is no space in that buffer, regardless of how big.

[crimson moat]

It will not, because of the train lock time. But insufficient buffers on input or output will reduce throughput further below the max train throughput, and stall production on one or both ends despite sufficient inputs existing.

[agile junco]

A buffer stops being a buffer when it's fully saturated.

[crimson moat]

If saturated with enough items to supply the train fully effectively (at input), or feed the machines fully until the next train comes (at output) then that loss of functionality does not negatively affect production.

If saturated below those numbers, it does reduce production on one or both ends.

[agile junco]

Again, my point is just that you only need both sides to have a buffer when the source and sink rates are the same. Though in practice there will be variability (like I said) and it's a good idea to cover both swings.

[agile junco]

It makes the buffer useless though!

That's my point.

[crimson moat]

No, it absolutely does not.

[agile junco]

lol dude

[crimson moat]

If you consume 100 plates per minute, have a train arrive every 5 minutes, and have a buffer size of 100, what happens when the train leaves? You produce for 1 minute and then your machines idle for 4 minutes.

You lost 80% of your production for no reason because you couldn't unload the train because the output buffer was too small.

This happens with input higher than output. This happens with input lower than output.

[agile junco]

So you're producing 2400 items per minute. and sending them to an ISB at a station. We agree you'll never get 2400 items per minute through that station... so the buffer is full. Removing the buffer, and you still have 2400 items per minute going to the station and it still cannot push that all through.

[crimson moat]

Input buffer does not do anything if your input is 2 belts per min

in any other circumstance, it does, and 99.99% of the time we're in those other circumstnaces.

[agile junco]

Why are you only talking about one side of the train?

[crimson moat]

Because each side can and will completely independently break flow.

[agile junco]

No

[crimson moat]

Yes

[agile junco]

Sorry, yes, but they are not independent.

[crimson moat]

If you consume 100 plates per minute, have a train arrive every 5 minutes, and have a buffer size of 100, what happens when the train leaves? You produce for 1 minute and then your machines idle for 4 minutes.

You lost 80% of your production for no reason because you couldn't unload the train because the output buffer was too small.

It doesn't matter what you input onto this train or what your network can carry, the output station ALONE will fail for lack of buffer.

Likewise, there are common similar scenarios where the input station ALONE will fail for lack of buffer.

If any one part fails, the entire chain does.

[agile junco]

It will fail if the input cannot produce enough regardless of it having a buffer! The buffer will always be empty!@!!!!

[crimson moat]

You're talking about the output station here?

[agile junco]

we are now, correct.

[crimson moat]

If it can store 500 plates, consumes 100 plates per min, and recieves a train every 5 minutes, then it can run continuously.

If it can store 100 plates, consumes 100 plates per min, and recieves a train every 5 minutes then it WILL idle 80% of the time and cut your production by 80%.

Input doesn't matter for that, it can be higher than output and it doesn't affect things at all. Train network doesn't matter for that, it's more than enough. The only variable here is the output buffer size. It's either sufficient or it's not.

[agile junco]

Input buffer must not become full while train is docked, output buffer must not become empty while train is not docked.

[crimson moat]

That's what i said about output, but the input buffer is a bit more flexible.

Input must not become full at a capacity smaller than that required to fill one train cart. If it fills at a higher capacity, no ill effects of stalled production.

[agile junco]

So if you are producing more than you consume, will the input buffer be full while docked?

If you consume more than you produce will the output buffer be empty after docking?

[crimson moat]

Yes. However a full but insufficient buffer cannot load a train effectively.

A full but sufficient buffer can.

Likewise for output: A sufficient buffer with sufficient throughput will never empty.

An insufficient buffer with infinite throughput WILL empty.

[agile junco]

Wait, don't move the goalposts.

I think you just agreed with my original point.

[crimson moat]

They haven't moved a millimeter

i did not

If you have input higher than output, and output buffer is insufficient, output side will fail.

Sufficient input buffer, insufficient output buffer
is often better than
Insufficient input buffer, insufficient output buffer

[agile junco]

Wait no.

[crimson moat]

A sufficient buffer at both ends, not just one, is always a mandatory requirement. Without it, the system will fail, it doesn't matter what your input and output rates or train throughput is. None of those variables can create success unless a sufficient input buffer and a sufficient output buffer are present. Exactly how large those buffers must be depends on the system, but the required size is never zero. It's at least one, and often many thousands of items large.

These are otherwise undisputed facts which are apparent due to the basic design of the system, and any like it which transfer things in discrete chunks rather than continuously (which is a whole lot of things in the world).

I'm gonna leave it at that since it's just getting very hostile and i don't know how else to explain it

[agile junco]

input being higher than output means that the input buffer is full.

Output buffer might still be needed.

I'm calling input / source, the rate going into the loading buffer. I'm calling sink the rate coming out of the other buffer on the unloading side. For the sake of simplicity I'm assuming we have sufficient throughput, but it's otherwise fill_rate = min(source, throughput) and drain_rate = min(sink, throughput).

OK. Feel free to find a bug in my code if you want. But I don't understand what you're arguing with here.

https://github.com/nixpulvis/sat_is_factory/blob/ab34223ff06e5c69876f63182396546ad1d80809/src/sat_is_factory/train_solver/__main__.py#L265

You absolutely do not need a buffer if you are pushing two 2400 miners directly into a single station. That's just obvious and disproves your final point.

I'm not hostile toward you. I just simply cannot understand what you are talking about when I can make a situation where a buffer is completely usless easily.

And editing (like that) messages makes it a bit hard to have a conversation.

Anyway cool. I feel like an asshole now and I just wanted to share something, which apperently makes no sense to anyone but me.

I can't even tell what messages you're editing anymore.

[oblique hollow]

no clue what the context was here but - hows that station set up?
like.... just 4 or 6 freight platforms or something?

[agile junco]

Single platform. Sorry should have been more clear.

[oblique hollow]

a single freight platform hooked up to 2 miners making 2400/min in total??

[agile junco]

Yes, it's an example of a case where you will get a backup regardless of a buffer.

[oblique hollow]

i believe that comes down to the fact that freight platforms can actually never support 2 full belts of something (if you only got that certain belt mk)

[agile junco]

The buffer will also be full during part of the docking sequence if it cannot unload enough into the platform becuase the trains aren't unloading as much as they are trying to load.

[agile junco]

That is correct.

[oblique hollow]

yeah in that case it really doesnt matter if you buffer or not

i believe the buffer only affects how high of a throughput you can get with something that ISNT 2 full belts

[agile junco]

This was my only original point.

I should have said both loading and unloading buffers, because if they aren't equal you probably still need one.

[oblique hollow]

technically you only need buffers if the belt themselves arent long enough to buffer the 27 seconds

[agile junco]

haha, belt + 1 internal stack yea.

[oblique hollow]

if you just get a 5 km belt, maybe you wont even need a buffer lol

[honest wren]

One message removed from a suspended account.

[oblique hollow]

but yeah - if you want whatever rate you put into the station to be transfered properly, it comes down to how much buffer space the belt has (plus the internal buffers of whatever else is connected)

[agile junco]

I have some code which calculates the amount in the buffers. I'm not 100% sure of the code, but I didn't understand the arguments above.

[honest wren]

One message removed from a suspended account.

One message removed from a suspended account.

[agile junco]

The goal is to make it so the source side doesn't stop, but sometimes that just cannot be avoided.

without increasing throughput, or the amount you sink on the recieving side.

I could walk people through my code, but it's a little messy lol.

[oblique hollow]

i think its a ratio of "how much throughput do you wanna transport vs how much spare capacity do the belts have"
and from that difference you get a "time to unload any stockpiled goods from the 27 second station lockdown"

[agile junco]

But I have fill_rate and drain_rate to cover the amount that actually goes through the system.
https://github.com/nixpulvis/sat_is_factory/blob/ab34223ff06e5c69876f63182396546ad1d80809/src/sat_is_factory/train_solver/train_solver.py#L45

[oblique hollow]

and if that time difference grows too big you lose throughput or something, idk

[agile junco]

Yea, basically: https://github.com/nixpulvis/sat_is_factory/blob/ab34223ff06e5c69876f63182396546ad1d80809/src/sat_is_factory/train_solver/__main__.py#L268-L285

[oblique hollow]

did you ever try to plot some ranges of values for train times and throughput of items inserted?

if i remember right, the wiki has some plots and they kinda have a sweetspot where throughput is highest

[agile junco]

I've done some tests, plots would be interesting, but this is tested against https://satisfactory.wiki.gg/wiki/Tutorial:Train_throughput for a sanity check for optimal RTD and Throughput numbers.

[agile junco]

Yes, this is 100% consistent with that. But it helps in calculating things when you aren't in that optimal spot, since rarely can you control the RTD much at all.

[oblique hollow]

yeah then so far without having seen much of it it sounds "sane"

[agile junco]

I generally use it like train-solver --rtd 3:30 --throughput 3000

Here's an example showing --source and --sink being different:

$ train-solver --stack 100 --platform 2400 --cars 1 --trains 1 --source 100 --sink 200
minimize rtd, minimize throughput >= 100.0

1 train
1 car
partially filled with 124 items in 2 stacks
1 min 14.39 sec per round trip.
2400 items/min total active rate of platform
1526.37 items/min throughput (63.6% platform efficiency)

100.0 items/min source rate (6.55% of throughput)
46 items in source buffer empties after 1.18 sec
100.0 items/min available
200.0 items/min sink rate (200.0% of available)
sink buffer would be empty

EDIT: added defaulted arguments for --stack and --platform (2x 1200 belts).

Unless I'm wrong about something, the sink buffer is useless because it will always be empty.

[agile junco]

Fastest way to get a 5+ min round trip time:

[vapid gorge]

@rocky ravine what exactly is the issue then? the link I shared back doesn't produce containers

[rocky ravine]

yeha but i do want some package turbo fuel

also you're packaging water just to unpackage it

[vapid gorge]

ok so are you already making packages for sinking or yo uwant to turn some of hte resin into containers?

[vapid gorge]

that's part of the diluted fuel recipe you can't avoid that if you want to use it

[rocky ravine]

i alreay have a factory with spare container

[vapid gorge]

ok then that last link is all you need

[rocky ravine]

i'm using the turbofuel alternate

[vapid gorge]

you're making 666.66666 TF pm, decide how much you want to package

[potent pawn]

wwwwwwwwwwwwwwwwwwww

[vapid gorge]

you're also using Diluted Packaged fuel

[rocky ravine]

you using diluted packaged fuel, i use diluted fuel

[vapid gorge]

ah I see what you did

you unlocked EVERY recipe and hit maximise

don't do that

[rocky ravine]

it just put all recipe by itefl when creating a new tab

[vapid gorge]

ok here

https://www.satisfactorytools.com/1.0/production?share=qUpeduM1SwJYxmiy1yOE

[vapid gorge]

if you hit maximise with multiple recipe paths available it basically stops caring about efficiencies

select the specific recipes you want if you hit maximise

OR

after you find out the max amount, change it to Items pm with that number

[rocky ravine]

that it

the maximize just didn't care about my input

what the point of telling it that i have some materials if it just ignore it

[vapid gorge]

because maximise ignores a lot of things. Type in 666.6666

but also what's the current issue with the link I just sent you?

[rocky ravine]

well it just didn't had the packaged turbo

geez thats a lot of refinery

[vapid gorge]

just package whatever you want?

[rocky ravine]

need the mk2 version of every building

[vapid gorge]

that's what over clocking is

[rocky ravine]

yeah but i want it to focus on power

not enough

[vapid gorge]

over clocking doesn't take up much extra power. Power is so easy and plentiful on the map it's a nothing consideration

[rocky ravine]

even with overclocking i had a factory asking me for 50 building + i don't have access to infinite shards

[vapid gorge]

they got rid of mk2 buildings for overclocking

[vapid gorge]

Well that's part of the design choice - you're still basically in a tutorial. You'll get there. Go farm some slugs and sloop them if you want

[vapid gorge]

you could do this instead? less power but way less machines https://www.satisfactorytools.com/1.0/production?share=xabUwVgHPNTLM52pgmpY

don't need coal either

[rocky ravine]

thats what i'm doing in fact

might have shuffled some recipe i don't have

i'm confused on the recipes i have

[vapid gorge]

yeah I think it would help if you just swapped recipes in and out as you're checking

[rocky ravine]

idk why it swap to some heavy and the rest on blend

diluted*

[vapid gorge]

probalby because you're doing weird maximise things

[rocky ravine]

i removed the maximize

[vapid gorge]

link me that?

[rocky ravine]

https://www.satisfactorytools.com/1.0/production?share=uuiCEpW2w6833fpXvCkU

[vapid gorge]

because you hae multiple recipes selected and the tool probably values coal as lesser value

oh and because you can't quite get that much out of the producti9on line without it

[rocky ravine]

or i just skip turbo and go straight to nitro rocket fuel by dragging an azote pipe or some azote canister i have

[vapid gorge]

I turned off Base TF

you can't make that much out of just blend

[vapid gorge]

just don't give it multiple paths

otherwise it'll use them as needed

same plan, just Turbo fuel https://www.satisfactorytools.com/1.0/production?share=zCV6QDYkvLCUWz9SSKCi

[rocky ravine]

i laso need to bring coal, which i can do, but idk if i have the quantity. Even rn idk if i have enough sulfure since they are brough by a dozen of carts

that keep trying to drown

[vapid gorge]

same plan just turbo BLEND https://www.satisfactorytools.com/1.0/production?share=ZFW30wlouSm5bXAUj5fK

you can't make as much with blend though

[vapid gorge]

drown?

[rocky ravine]

idk

[vapid gorge]

if you're using trucks and roads you want it at least 3 wide

for 2 directions

I'd honestly just belt the stuff over early on.

[rocky ravine]

it's fine the game grab them back like in mario cart

[vapid gorge]

I woudln't trust my power station to that personally

[rocky ravine]

it's quite far and not easy access so no train and barely enough space for a truck

also they don't use fuel

they try to leave backward the do a drift XD

now i also need to figure out how many overclocked generator/ machine i need

[vapid gorge]

... you're going to need a lot of factory carts...

[rocky ravine]

i don't even know if my node produce enough coal or sulfur

can probably get around somehow

[vapid gorge]

what does the miner say?

[rocky ravine]

https://www.satisfactorytools.com/1.0/production?share=4PAgzOIQrKHV4mZFo51t
could do that but idk if the efficiency of rocket fuel warrant the change

i can get the nitrogen and it's way less coal and sulfur

[vapid gorge]

I'm not a fan of rocket fuel in general. I think it's fairly dumb.

[rocky ravine]

and machinery

[vapid gorge]

like honestly diluted fuel will just get you ass tons of power

you can finish all the milestones on diluted fuel or get to nuclear on it VERY easily

and it sounds like you're having a hard time tying together a lot of design stuff as it is

[rocky ravine]

cause that particule accelerator is eating every ounce of power in my base

i have a nuclear power plant and i can probably sustain 3 on my setup but i don't have way to deal with waste

[vapid gorge]

just store it, it's easy

[rocky ravine]

omg even the carts are trying to kill each others

[vapid gorge]

are you slooping your PAs or something?

[rocky ravine]

surfur is fine

nope, just climbing to 25MW

or GW idk

[vapid gorge]

if it's using 20gw you've OCed it and slooped it

[rocky ravine]

25k mw

[vapid gorge]

25k mw is 25gw

[rocky ravine]

sulfure is fine, as for coal i can bring it by train

[vapid gorge]

Go rip out the shards and sloops from your PA and your issues will be solved ...

[rocky ravine]

i still need to increase my power production tho

producing 6.4GW for 4.2 use and potential 17 if everything on at the same time

[lethal kettle]

how much oil can a mk2 pipe transfer?

[agile junco]

600

[lethal kettle]

thx

[vapid gorge]

well, look, it seems like you've got a really messy set up going. I'd keep it simple. Go to one of hte oil rich areas and just make a bunch of diluted fuel :\

[rocky ravine]

it's just that i didn't exploit all the oil i have at disposition

i have 4 well and only using 2 and not even overclocked

[vapid gorge]

do exploit it, there's a few spots with 2-3k oil nearby 🙂

[vapid gorge]

https://www.satisfactorytools.com/1.0/production?share=2WxuIMne1JO0vd0hMzQR

[rocky ravine]

oh wait i have drones now

idk how much they can carry tho

[vapid gorge]

as much as you like. If you need more, just add more drones
they are basically just slow teleporters

[rocky ravine]

i've never used them but considering the terrain is hard to go through and i don't want another 10 km conveyor belt that should do, i just need to math out how much i need to use

and idk how to do that

that particle accelerator wasn't sommerslooped

[vapid gorge]

this is a rough estimate #math-and-meta message

but you can allways just add more

[queen umbra]

This area is sooooo good for Mercer sphere and power slug and ressources but it's dangerous as hell

[agile junco]

Just wait until you learn about this area:

less so Mercer Spheres, but hella slugs.

Something like this.

[queen umbra]

I found like 6 hard drive it's pretty good

[queen umbra]

Is it dangerous?

[agile junco]

Somewhat... mostly just lots of elevation. Nothing too scary.

Bring a good jetpack and you're set.

[rocky ravine]

earlier when searching for mercer orbs and sommersloops i got jumped on by 2 Elite gas stinger and another time by 2 Alpha stinger 2 spitter and 4 cliff hog

Thanksfully i had crafted the rifle +1.5k rapidfire ammo

[lethal kettle]

how much can a mk4 coveyor transfer?

[queen umbra]

480

You can see it when you hover it

[vapid gorge]

how much water are yo utrying to fit in those 2 pipes?

[fringe birch]

600 each

[vapid gorge]

why do you have 15 water extractors then?

[fringe birch]

5 for each pipe then tried adding 5 for the water tower

[vapid gorge]

ok get rid of the hump
get rid of the 5 extra extractors
get rid of every single buffer

just use pumps

are you connecting up those 2x 600 pipes at some point? if so disconnect them

keep your fluid systems split into their own sections

also avoid sky factories, there' sno reason for them and it just makes all logistics more annoying , especially pipes

[fringe birch]

so is trying to find level ground

[vapid gorge]

more than 2x as many gens right over the same type of terrain

you could have fit all of them here w/o much worry, barely higher than anything

it just makes life easier

I count 26 coal gens? have 1 pipe feed 13 , in 2 groups

[fringe birch]

meh not at that point in building compact yet.. over time

[vapid gorge]

it's literally just 2 floors

4 rows of 8 on the bottom
4 rows of 8 on top

in any case - do the fixes to the pipes, it'll work

water towers don't do anything pumps do do and pumps are essentially free and simpler
don't merge fluid systems together, that's just asking for trouble
don't use buffers

at best they do nothing at worst they screw your flow

!wikisearch cg

[glad apex]

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

[vapid gorge]

these are extremely convenient coal layouts for beginners and experts. I still use it because it's very convenient

[fringe birch]

water towers can save of using pumps

[vapid gorge]

the 64 gens in my image uses it

[vapid gorge]

yes but then you start connecting up a bunch of fluid systems

which is bad

You can make it work, but when you wind up with problems the first thing to go will be all the mergings you've done

and you've got a ton of issues

pipes are not belts. You can't just slap them around any which way you like

[vapid gorge]

oh and remove the valves, they can throw chaos into the flow as well

just saw those

[fringe birch]

valves were a trial. havnt used them much

[vapid gorge]

like buffers they are a 'at best they do nothing, at worst they cause issues'

[fringe birch]

Think i got it

[vapid gorge]

@fiery ridge pick whatever you think you'll build next, they're all useful

also ask questions in channels people can talk to you in

[flat flint]

hypothetical

how much iron do i need to complete the game (minimum)

[vapid gorge]

1pm could do it

[flat flint]

im currently producing 1480 a minute

[flat flint]

true 😂

[vapid gorge]

don't try to future proof - also 'complete the game' is vague.
this is a sandbox, more like minecraft than factorio. You're done when you can't come up with any more of your own projects to do

[flat flint]

how much copper do i need to producce 5 nucllear pasta

[vapid gorge]

depends on the recipes you use. 😛

[flat flint]

the best ones

duh

[vapid gorge]

https://www.satisfactorytools.com/1.0/production?share=VqynxSVZWOvCGFt3WQez

[vapid gorge]

no such thing as best recipes, anyone who tells you otherwise is a fool or a liar

[flat flint]

my great uncle ferb is no fool!

[vapid gorge]

generally my prefered copper ingot recipe is Base or Copper alloy

but that link shows you 5 pasta pm with base recipes

[flat flint]

it does

[vapid gorge]

that could change though with different wire recipes, or circuit board recipes, that would cut it down too

[flat flint]

how many hours do you have???

im guessin abpve 2k

[vapid gorge]

3500 or so? I've kinda taken a long break though

I did the tiers again when 1.0 hit but just kinda waiting for the game to finish finish and for some mods to update

[flat flint]

cool, cool

[vapid gorge]

this is 5 pasta pm with copper alloy https://www.satisfactorytools.com/1.0/production?share=auoTRemgtOzZzbZ5Ssz6

shrunk down further.

can't go any further really w/o the pure copper recipe, but I hate it. https://www.satisfactorytools.com/1.0/production?share=fqCO5agLlrfujTaePuPg

[flat flint]

why do you hate it

[vapid gorge]

because it's a giant mess of refineries for not really much more savings https://www.satisfactorytools.com/1.0/production?share=fqCO5agLlrfujTaePuPg

3000 copper and 60 foundries
or
2400 copper ore and 160 refineries

[flat flint]

def 3000 coppper

[vapid gorge]

You'd be surprised - TONS of people will just go for pure cause 'min max' or something

But I'd spend that extra 600 copper ore any day to avoid it too 🙂

[flat flint]

im just trying to automate every requierd material a considerable ammount

[vapid gorge]

and this is why there aren't 'best' recipes.
unless you are very specific in your desires 🙂

if you said 'best' is just most copper efficient recipes, you'd pick pure copper plus a few of the ones I checked off to cut copper out

[flat flint]

once ur done with ur break from satisfactory care to come check out my base

[vapid gorge]

that might be a long while xD I'd be happy to look at any images you'd care to post though 🙂

[flat flint]

i might

currently its just a tiny facory automating 4 of the most basic materials

[vapid gorge]

gotta start somewhere 🙂

[flat flint]

oh i have 2k hours im just going big

[vapid gorge]

thats fair 😄
I want to do like 1800 HMF pm at some point.

[flat flint]

hmf?

[vapid gorge]

heavy modular frames

[flat flint]

oh.

thats alot...

[vapid gorge]

it is. I've got some ideas for sushi load balancing some sections of the system, using interesting combos of recipes and foundries

[flat flint]

for parts like that im gonna produce 10 PM for storage and whatever i need for other factories

[vapid gorge]

that'll keep you going well 🙂

[mint dew]

not sure why but my power plant is only makeing 116 gw but i planed it to amke 150 gw i am useing 2700 crude and turning it all into rocket fuel

[vapid gorge]

work your way backwards from generators to the fuel producers

are your fuel producers starving for components or are their outputs clogged?

[mint dew]

so i am amkeing enough fuel but i have 240 full oc gens going and that is 150 gw and i have about 10 that are still filling and i am missing aboujt 30 gw

[mint dew]

.

[vapid gorge]

ok so either you haven't waited for manifodls to fill, which what you'd want to do is under clock a few gens on each manifold to like 50% first to do so

or possibly you have a flow issue if your fuel producers are stuttering. Which you haven't checked yet

[mint dew]

i think its flow bc i have fuel in the refinerys but for the turbo it says that it does not enough
not sure why tho bc they are on the same level

[vapid gorge]

ok so the rocket fuel producers are starving or clogged?

that's the step right before the fuel gens right?

[mint dew]

right before makeing rocket fuel

the turbo refinerys

to the end of the manafold it has almost no flow

do i need to add a pumol

[vapid gorge]

ok but I can't see anything youre doing with your set up or how it's going

you're making rocket fuel for the gens right? I want to know what they are doing first pls

[mint dew]

i am turning crude into rocket fuel

[vapid gorge]

Yes. And the way you trouble shoot a fluid system and work backwards

So are your rocket fuel producers starved of material or are their outputs clogged

[mint dew]

yes

i belive its the turbo fuel i have one line where the last gens pipe has next to nothing in it

[vapid gorge]

Ok go for it

[mint dew]

like add a pump

[vapid gorge]

No idea

Either you’re rf machines are

Clogged
Starved
Working fine

[mint dew]

ik they are geting starved but idk why they are starved

[vapid gorge]

Ok and your tf machines are clogged?

[mint dew]

i would assume so

[vapid gorge]

Check 🙂 it’s important

[mint dew]

uh how do i kinda my first time runing to problems with fluids

it seams to be the flow rate it is bearly geting to 500 and i have 6 makeing 100 a min

[vapid gorge]

Click on the tf machines

Are they starved or clogged? You know what I mean when I say that right?

[mint dew]

um i mean i dont rlly know... 🙃

[vapid gorge]

Ok clogged is when their outputs sections are filled so that the machine can’t produce a new cycle

Starved is when the machine lacks enough input to start a cycle 🙂

If your TF machines are clogged? Flow issue to the RF machines

Starved? Follow the problem backwards

[mint dew]

whats a tf machine?

[vapid gorge]

Turbo fuel

[mint dew]

oh

i mean its that the fuel machines are not about to give the rf machines enoughjt fuel

like i have fuel left over in the machine but it seams that i have no flow and i have idea why

[frosty owl]

Check carefully to spot where exactly the issue is. You need to understand who lacks what and who has too much of what. Then figure out why that is the case.
If you have doubts or issues, take screenshots and ask about them. Sometimes even just trying to describe the issue properly, you might realize the solution

[vapid gorge]

So it sounds like there’s a flow issue if the TF machines are clogged and RF starving , flow issue

Overhead images of how you’re feeding the RF machines

[dapper sable]

hey all, I assume someone here is familiar with SF modeler - I'm trying to make an effort to convert from planning my factories in factoriolab calculator to Modeler but there are 2 things that are driving my nuts about modeler and I wondered if someone has a solution for me:
First isn't big and I suspect it is 'just get used to it' - in the online calculator I quite often use 'Machine' based output - e.g. I want 3 machines of output, short of manually doing the math on how much per minute that is I see no way to add an item and say 'I want 3 machines worth"

Second is the much bigger one, this one is making me so close to going back to the browser based tool:
I have a factory that I need to Net produce and consume an item to make another
online browsers handle this easily, you set the 'desired output' and if something needs item A to make item B it will add more production automatically
I can't for the life of me, figure out how to do this in modeller

Example:
I want a factory to net output 1 machine of Circuit boards and 1 machine of Computers (12.5 and 3.33 per min respectively)
Computers use Circuit boards in production. Unless I manually add the 5/m circuit boards to the output of the CB section it will split the net output of 12.5 into 5 and 7.5 to feed the computer section.
The online tool would automatically increase the production to 17.5 - can modeler be set to do this too when you drag a supply line to an already existing net output?

[mossy halo]

Screenshots can help. I guess you have set a limit on the CB machine, thus it doesn't get increased

[dapper sable]

Sure, I want net output to be 12.5 CBs and 3.33 computers
Default behaviour when linking the 2 items is to split CB production to 7.5--5
Ideally, I would want it to treat the 'limit' as a 'goal' and increase it to 17.5 to retain the same output

[frosty owl]

(1) Should be addressed by setting the number of machines inside that specific node (bottom text field in the node, right below the number of machines currently being used). Do note that that is a max value, wether the calculator can reach it depends on inputs and other factors too

(2) Probably needs you to do the following:

• Find the required output of circuit boards for your production (ie: how much is needed by other machines)
• Add to that what you want (ie: one machine's worth of CB)
  You usually achieve the second point by limiting the output of the CB node, or its number of machines, but that will not update automatically based on how the output is consumed. Ie: if the computer node is modified to take more CB, less will be left for storage
  You can "sidestep" all of this by changing approach in node-making and creating separate nodes for the computer-input production and the "storage production"

[dapper sable]

Thanks Vencam, looks like since it only has 'limit' mode I'll be sticking to the online tool. for large factories I just can't run the risk of forgetting to increase the production on a limit set or forgetting to build extra buildings if I have to have the same item twice simply due to a lack of functionality 🙁

[frosty owl]

Eg: having one CB node without any set limit connected to Computers and one CB node set to output "X/min" will always have a total output of "X/min + however much needed by the Computer production"

[frosty owl]

Wouldn't the issue go away, if you had specific machines set to produce those outputs not (directly) related to other production lines? (See my last example)

[dapper sable]

like this?

[frosty owl]

The limit on the CB node feeding the Computer node can be removed for future expansions

Else, the computer node would be "chocked" by that input if you were to make it bigger

[dapper sable]

that could work, I assume the 1 2/5 is the required total number of machines with that recipe set. Makes things a little less clear as well especially since the 2 nodes don't agree on the total required number of machines.
Possible solution, yes but sub-ideal. I think I'll put in a feature request and keep to the online tool. I have to share this with other builders on the map and anytime that there is a risk of confusion that can be elminated its probably best. Its a shame, I REALLY like the concept, just needs some more polish by the looks of things.
a little tickbox that said 'Net' or 'Goal' that would allow it to auto-increment any 'limit' would be the solution

[frosty owl]

I feel like there might be some misunderstanding about the "1 2/5" part (totally unneeded imo), but given your conclusion it might be meaningless to delve into that further ^^
The tool should have its own discord server, iirc

[dapper sable]

I thought it would have too and it probably does but I had a look around and could only find links for bugs and suggestions and I couldn't find anything on the suggestions page that had a 'discord' link (that I could find...)

[mossy halo]

don't set the limit and it will autocalculate machines needed for the output needed

modeler is mighty, but you have to get used to it

[dapper sable]

then how do I tell it that I want 12.5 of that item as net?

[mossy halo]

somewhere you do have to set a limit, just on the last machine, yes. but on all machines before that, the limit is not needed

[frosty owl]

That would be the node you're supposed to put a limit on, the one sending items only to storage

[mossy halo]

Take a look at this example. I just set a limit 1 on the dirt extractor, because I just want to use a single one for that build. Everything else gets autocalculated.

[dapper sable]

Ah ok, same as Vencam above, have a second node for the same item. I love what I've seen of it, just not that. IMO, multiple nodes for the same item in the same build makes it easy for myself or collaborators to make a mistake
Thanks Vencam, another solution would be to allow a container to 'require' an input, which I could then set as the production target that would be an easy way to do it but I can't see a way to do that either :(.

[frosty owl]

Yep, unfortunately that's not a thing. The closest you could get would be creating a "placeholder" production (replacing the storage container) that would be limited to input a set amount of that item...

[mossy halo]

btw, I always add an dimensional depot uploader (small uploader pictured in the screenshot) in the line were I put something into storage, on the DD uploader I can set a limit of how much I want to get into storage. With the priority splitter before the storages, everything else gets send to the next priority splitter, which sends everything first to the output of the outpost and any overflow to the awesome sink.

[frosty owl]

Or maybe not.. That's exactly what you were looking for @dapper sable ahahah

[dapper sable]

This appears to be the solution! Add a DD requesting that item in the qty needed, a dumb workaround but a workaround indeed!

[frosty owl]

Yeah, I never used DDs specifically 😆

[dapper sable]

haha! I knew there had to be a way! Thanks guys!
I'm still going to put in a couple of suggestions as it would make my life easier as I plan backwards, from final to source rather than the other way around!

[mossy halo]

Most of the time I work backwards too

[frosty owl]

I actually always plan from final to source 😅

[dapper sable]

This one is probably easy, how do I clear only ONE of a connection? I can clear all but I can't seem to get rid of just one...

[frosty owl]

But I also think that it'd be a nice addition being able to put a limit on a specific node-to-node connection rather than on a node's parameters (like their output/min). Add to that the possibility of dragging lines from nodes to "nothing" and it'd be much easier to connect far away points (no need to have both nodes visible if you can drag from one node to the void and then drag that to the other node)

[frosty owl]

Not possible, AFAIK

Which is why I have some "important" connections pass through smart splitters and such, so they're separate from others

[mossy halo]

If the connection was moved manually, a double right-click will auto-place it again. Another double right click on the connection line deletes it.

[dapper sable]

Yeah... just small QoL things, its amazing what it can do but gah-dahmn some of the 'why the eff isn't this a feature yet' things...

[mossy halo]

What are you missing, except of setting limits on containers?

[dapper sable]

Removing only 1 connection as required rather than all of them
Production 'goals' in addition to 'limits'
Unable to set a 'minimum' production that can be updated automatically
I'll put in a suggestion for them 😉

[mossy halo]

Removing one conecction is a double RMB, as told above.
What do you understand by production goals?

[wind spade]

You can also use online calculator to get the numbers and then modeler to plan the build (if you don't want to plan ingame)

[dapper sable]

I think thats the way it will be for now at least. Ideally was looking for a 'one stop shop' offline upgrade but it looks like its not to be for now! Yours and the factoriolab site are the go to for build plans at the moment. 😄

[dapper sable]

double RMB just places a miner for me when used on the link line and opens part limit settings when used on the item icon on either building nodes that are linked. I'm using 1.9.8-Steam
What does work to unlink things is to drag a 'new' link between the same Production and consumer and it unlinks them.
By Limit & Goal I mean that other modes could be introduced, like a 'goal' or 'net' mode where it is not a limit but a tick-box option to say 'this number is not a hard limit but a minimum, add production when needed by other buildings'

[wind spade]

New Tools may be something like you want, but those are far from finished atm

[mossy halo]

You need to double RMB on the part icon/amount of the connection line, not just anywhere

[dapper sable]

yes, that gives me the Part Limit option, I don't always want to 'clear all connections' from that item. Sometimes, its just one of the connections.
If I click and drag that icon to the one I want to remove, it removes it, but 2xRMB just opens that option

[mossy halo]

not on the machine. on the connection line. First double RMB made the line to be auto-placed again, second double RMB deleted exactly that connection

[dapper sable]

Gotcha, that works too!

[frosty owl]

I initially thought you were searching for a way to delete one specific connection out of many by right-clicking on a node's input/output 😆 (which, unfortunately, isn't a thing yet)

[brittle lily]

So, I'm having issues now getting any flow to my coal gens. Everything was fine before I added the pump near the stairs there, but I was having a lot of power fluctuations because 2 of my gens are getting sketchy water input due to the angle they're at, but they were all still flowing. Now I add the pump and nada, it's backed up.

I already flushed the pipe segments before and after the pump, but still nada

[vapid gorge]

how much water are you trying to move through that pipe?
how much can a mk1 pipe move pm?

[vast gust]

@brittle lily maybe turn your pump around

[vapid gorge]

also that

[brittle lily]

I double checked..

[vast gust]

it has an arrow when you place it

[brittle lily]

Yeah I thought I double checked it was facing the right way

[vast gust]

not in the pics its not

[vapid gorge]

but also your pipe doesn't have enough throughput

[brittle lily]

Yeah I see that. I'm overclogging it

Trying to fit 10lbs of water in a 5lbs pipe

[vapid gorge]

!wikisearch cg

[glad apex]

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

[vapid gorge]

convenient coal layouts - 1xmk2 belt of coal, 3x extractors

[vapid gorge]

find a lake with 3-4 coal nodes next to it, and turn it all to power

[brittle lily]

This is my current setup, all mk2 belts

[vapid gorge]

I also highly advise you not mix power production and item production together like that

it's a great way to have unstable systems and you don't want unstable power

[brittle lily]

Not item production. I originally built a truck station and repurposed it to just feed the 2 mk1 belts that split off the long mk2 bus line back into a mk2 belt for the gens

Also yeah, switched the pump around and still nada, so I gotta adjust my flow rates

[vapid gorge]

look at the diagrams - it shows needing 2 inputs

[brittle lily]

So the closest area like that to me is in the basin way away. Obvs that's difficult and a huge waste of resources to run run power lines all the way from there to me which is basically where my current setup is. What should I look into as some potential solutions?

-768,1474 to be exact

[vapid gorge]

... what? that lake is basically designed for early coal power.
it's nearby and a cable back to your base costs... nothing

use distant coal for power, nearby coal for steel. Makes life easy

this game forces you to spread out - you're going to have to get very used to having to have multiple hubs around.

[brittle lily]

I've been trying to stick to one area and build like a central area there, but yeah that makes sense

[vapid gorge]

you'll need resources that just aren't near the starting area very soon. Like oil

[brittle lily]

Are hypertubes the best early game way to fast travel between these hubs I assume?

[vapid gorge]

they are convenient for static travel yup , point and click basically

not great for exploring though

[brittle lily]

Understood. Alright, well, I know my mission for the morning

[vapid gorge]

zip lines and parachutes are great for exploring though , research them asap. Basically research everything in the mam asap

[brittle lily]

Got parachute already but yeah been eyeing the zipline

[vapid gorge]

it's convenient since you can build a power pole on a cliff and next to you and just climb up quickly

[queen umbra]

I got a problem I'm making fuel to make power and by doing so I'm make polymer resin so I also make rubber and plastic but my 2 container for rubber and plastic are full so the fuel power shut down what could I do with rubber and plastic

[mossy halo]

sink it for awesome coupons

[brittle lily]

I need to remember that splitters decrease output multiplicatively

If a 120 output goes into a splitter, it's 60, and if a 60 goes into a splitter it becomes 30

[wind spade]

assuming nothing gets stuck anywhere, yeah

(and assuming you're only using two outputs)

[brittle lily]

Yeah true, if you use 3 now you got a 60 turning into a 20

[queen umbra]

What are you trying to do

[brittle lily]

I'm setting up a new coal plant to exclusively provide power, and I've got two coal nodes I'm running belts for. I was using a splitter to see if I could just converge the new mk1 miners into a 120 belt, then split out that one. But since it's ALSO going into new sets of splitters to feed the gens my predicted 60 coal/min output was halved, so just fixing my strat a bit

[wind spade]

you can also just do this

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

[brittle lily]

Yeah that's what I'm setting up now.

[mossy halo]

Just do a manifold

[unique cypress]

Depends how many belts you connect to it

The total on the output belts is always equal to the input belt

[brittle lily]

Yeah, so this is what I originally had set up. So I'm gonna just bring the two belts down the line seperately and use one belt per row of gens

[brittle lily]

Much better

[queen umbra]

How much coal is coming in?

[brittle lily]

120 to each row, grand total of 240. It's two miners overclocked to 200% (I haven't unlocked mk2 yet)

Which reminds me

Should I be going for upgraded power

Or upgraded miners?

[queen umbra]

Don't you have not enough gen?

[brittle lily]

8 gens, 3 extractors

[queen umbra]

But a gen need 15 coal and you have 240 coal

[queen umbra]

You could have 16 gen if you get enough water extractor

[brittle lily]

I'd need 6 water gens, so 3 more.

Heck it

[queen umbra]

It's not that bad

[brittle lily]

So, I'm a bit confused

Why do these common setups docs show that each 8 gen setup has 120/min going to it if that's in excess?

[mossy halo]

120/15=8

[native mauve]

8 gens use 120 coal, but you have 240 coal. hence the 16 gens

[brittle lily]

I gotcha. The math is getting mixed in my head, but I got it now

[agile junco]

Anyone know if there's an existing tool like https://icemoonmagic.github.io/Satisfactory-Splitter-Calculator/ but which accounts for the fact you can split belts into slower belts to limit as well?

I'm tempted to make one, but I'm also struggling to decide if it's possible to do analytically.

[queen umbra]

??

[agile junco]

Unless I'm mistaken, this tool only splits in halves and thirds.

[queen umbra]

Wdym

You have a mk 2 belt with 120 you can split it to have 2 mk1 with 60

[agile junco]

I ask it to get a target rate of 30 from 190, and it doesn't just split off a 60 belt then split that in half.

It does all the 1/2s and 1/3s to get it that way.

[queen umbra]

Manifold?

[agile junco]

Here, even simpler example of trying to get 60 from 190:

This is not the simplest way to do it.

[queen umbra]

In too dumb for that

[unique cypress]

this calculator rarely produces the simplest solution, even if you don't use slower belts for limiting

[deft lichen]

you can get 60 from 190 with 1 smart splitter

[agile junco]

Yea, that's what I noticed. I'd be curious if there's a better tool.

[unique cypress]

not that I know of

[agile junco]

Exactly, that's my point. Actually why even a smart splitter?

[deft lichen]

overflow setting

[agile junco]

It's a tough problem honestly. At least analytically I can figure it out. I'm not really sure how to optimize recursive functions... but I could simulate it and brute force. Might be fast enough, just less cool.

[agile junco]

Oh yea duh. I was thinking of just merging back in the output. Because I don't want to worry about smart splitters yet lol.

[deft lichen]

I think there should be a computable solution better than brute force, not that I can think of one though

[agile junco]

Yea, it's just tough, because even as a graph search problem, I don't immediately see how to search down the split 1/2, 1/3, or limited belt edges.

I feel like there's a cute mathematical representation here, but I don't see it yet. Very combinatorical.

[mossy halo]

I'd just use adjustable splitters from advanced logistics mod at this point 😅

[agile junco]

Well that's no fun!

[mossy halo]

true

[agile junco]

My end goal is to have a way to solve for the minimum number of belts to carry an arbitrary load with the fewest splitters/mergers to get there.

I have like 6 items in 14 belts to move and I know I can merge these up a bit.

[gentle dew]

I think there was a YouTuber who did something for that... One second

[unique cypress]

If you just want that, you can just use a belt compressor

[agile junco]

Do go on...

Oh using overflow and smart splitters?

[unique cypress]

basically, yeah

[agile junco]

Or maybe priority mergers.

Yea, I guess that would be a lot simpler..

[unique cypress]

depends how you want to set it up, really

I've seen designs with prio mergers but they're not necessary

[agile junco]

I do not want this:

Granted that's a lot larger than I'd need.

[gentle dew]

Found it:
https://youtube.com/shorts/YG_GZqwRHDg?si=yhMAbb0dW9uXxMI6

[agile junco]

I would ideally like to minimize the number of mergers/splitters. which I have a feeling would be possible to make smaller than a compressor like this, but maybe not in general.

[gentle dew]

I'm not sure if that's exactly what you are looking for or not, hopefully it helps

[agile junco]

Yea this is what I was thinking roughly, but he uses a bad example, because he compresses to the same number of belts. But it shouldn't be hard to make it work to compress 4:3 as long as the numbers allow for it.

[gentle dew]

Ah

[unique cypress]

my design would have n*(m-1) smart splitters and (n-1)*m mergers, where n and m are the number of inputs and outputs, respectively

[agile junco]

Makes sense. Though if you had say 3x 180 belts, and only 270 belts, you could get that with just a single splitter and two mergers.

EDIT: could get two 270 belts.

So ideally I'd still have to solve for this whole thing to get a truly optimal solution.

But compressors are a nice trick since they aren't really dependent on the rates themselves and easy to design and construct without a tool.

[unique cypress]

I favour universal solutions that work for all numbers over "optimized" solutions for only one specific set of numbers

[agile junco]

Yea, I get that.

I'm the same way, unless I can have a universal tool that generates optimized solutions. No way in hell I'm working all this out by hand.

I'm also just kinda excited to see if I can use Z3 for more things... but idk if I can make it work for this.

[wind spade]

0, clock machines to match previous rate

[agile junco]

How does that help anything? If I need 3 items each at 180 items/min, I can merge them onto a 270 belt and smart split them later (with overflow to sinks to be safe).

The goal is to minimize belts over a long distance.

Oh I misunderstood, you mean instead of splitting, clock accordingly. I'd still need to merge, though.

[wind spade]

minimise distance = build at source 🙂

[agile junco]

Still I need to know how to clock the needed 1500 to merge into the rest of my production to carry it all. Though I guess this makes the math easier because I can avoid the constraints of solving for the split ratios.

[wind spade]

(and yeah it's kinda a non-answer to your question, but it's the way I'd recommend 😉 )

[wind spade]

the thing is to not merge at all. If one machine makes 20/min, clock the next machine to need 20/min. Or 10/min and build two. Or any other combination. No reason to merge to high capacity belts if you can directly feed it to next machine (again, just one way to play, but I personally like it quite a lot)

[agile junco]

I mean, it's nearby source. It's basically one factory, but I have to get like 7 items on 12 independent belts from one side to the other, and I know I can do it in fewer than 12 belts.

That's slightly different. I'm basically talking about using spare space on 1200 belts to carry lower rate items downstream to the next phase of production.

[agile junco]

The problem with this is that you often have multiple parts being produced and consumed in multiple places, so you have to network them to each other. It becomes a clustering minimization problem to reduce belts, and this is part of optimizing that.

[wind spade]

well, the approach I'm suggesting is "factory makes final product from raw ores", so you don't relly run into those issues

[agile junco]

So you have a reinforced iron plate factory and an iron plate factory separately? For example?

[wind spade]

yeah

[agile junco]

Oh gotcha, yea then you don't have these issues, but that's not how I tend to build.

I like to do all my refinement of ore to ingots in one place... etc

[wind spade]

they don't really have to be separate location-wise, but they do not "steal" resources from each other (so the whole factory is independent, which gave it the name)

[unique cypress]

that's still an issue when you do that. it's not like every item in a HMF factory is only used once

[wind spade]

their issue is "distribute intermediate from central location", which is non-issue in independency

[agile junco]

I mean, you could make it so it's basically a tree... but I wouldn't do it that way.

[wind spade]

yeah, whatever floats your boat 😉 though if you want to read more about it, there's a tutorial on wiki https://satisfactory.wiki.gg/wiki/Tutorial:Independency

[agile junco]

I've read this. It's actually inline with me general build philosophy. This is an independent nuclear production system... I'm just building the inner factory a bit differently. The connection can all be belts, but they still need to flow.

[wind spade]

for sure 🙂 I just wanted to chip in with an option you may have not been familiar with

I personally dislike things like belt optimisation and compression, so I can't really help otherwise

[agile junco]

Independency is a spectrum. Going all the way with it you end up with this, which I think we might agree is too far.

[wind spade]

tbh I'd build this

[agile junco]

I'd build this:

(most of the time)

[wind spade]

oh if you mean how the machines are connected inside one factory, that isn't really covered by independency, that's up to you

[agile junco]

Yea

[wind spade]

indepdendency says that one item's factory doesn't steal products from other item's factory

[agile junco]

This is "one factory" I'm talking about, though technically it's like 5 or 6 closely connected buildings.

[wind spade]

one factory = machines from miners to storage that make a single type of final product

[agile junco]

This is it at the moment.

[agile junco]

That's not my definition of a factory. A factory to me is anything which is all connected with only belts.

[wind spade]

I meant in the context of the independency definition I've given above

[agile junco]

Ah

And I generally think of sub-factories in my larger factories.

[wind spade]

so both of your MF screenshots would be independent, because both would count as "MF factory" for the playstyle

how they are wired inside doesn't really matter

just that the MF output from them are not used anywhere, they go to storage/sink

[agile junco]

Sure. But that was my point about it being a spectrum, since one is internally more independent.

And since my question about belt minimization is about connecting parts of the inside of a "factory", it's relevant to the way I distribute items, e.g. those iron rods.

Basically this: I have these 4 sub-factories and I need to move all their items to the other side of the factory to finish up in another building.

[wind spade]

well then my initial answer is kinda relevant - clock machine to match next step

though since you have "centralised" wire production, it's not easy to do so

[agile junco]

I can't easily do that without grossly increasing the number of belts.

[wind spade]

well, I mean it with "next machine is right in front of previous machine", which you said you don't want to do

[agile junco]

I do this as a design choice and to make building it a bit easier and more aesthetic in my eyes.

Something about having all my wire together feels more realistic and natural, even if it causes logistic issues I might be able to avoid another way.

[wind spade]

yeah, again, valid as anything else 😉 I've given my suggestions, whether you do them or not is up to you. But can't help with belt compressors and such

[agile junco]

Though, I'd still have to get the iron or copper to where the wire is sent instead, so I'm not sure how much it really helps TBH.

[agile junco]

I do appreciate it and it's an interesting discussion. I also just think it's a neat math problem.

[wind spade]

well I usually build in place near the resources I need, or ship the few extra raw resources. I don't really care about minimising logistics in that regard, I just like to have the whole "factory" in one place and have input be raw ore and output final product

[agile junco]

I mean, in that way we're the same basically.

These subfactories are all right next to each other.

Very much still a WIP:

those blocks on the top are the ingots.

[wind spade]

except in my approach ingots are not final product 😉

[agile junco]

They aren't a final product for me either.

This is the top level factory.

Only ingot input is aluminum.

I think I just think of things as subfactories a lot more than you do.

Like to me, my screenshot above is all one factory.

It would be cool to work out the algorithm to generally minimize logistics (either in terms of number of belts, or total belt length, or something like that)... but that's a huge amount of work.

[mossy halo]

How can people live without using autoround/round to fullest machine?

[unique cypress]

I just leave machines at 100% speed 🤷‍♂️

[agile junco]

Why can't I link to cs.princeton.edu?

[mossy halo]

nah, I want my modeler to tell the exact count of machines I use, so I use auto round or round manually to fullest machine heavily

[unique cypress]

because it's not on the whitelist

[agile junco]

🙁

Found a interesting paper titled: "An Efficient Heuristic Procedure for Partitioning Graphs".

A lot of this work is utilized to layout circuit boards.

[wind spade]

why would I want to round?

[mossy halo]

How do you place 7.3 smelters? You don't. You either round that to 7 or 8 smelters. modeler can give you the exact overclock setting for using 7 or 8 smelters using the rounding feature

[wind spade]

but "7.3" has all the info I need. 7 or 8 doesn't, because I still need to look up clock speed

[mossy halo]

You can't place 0.3 of a smelter

[wind spade]

no, but I know I need 0.3 worth of a smelter production

so I can underclock/overclock/match ratios based on the situation

and I don't need third-party tool to tell me that 0.3 of a machine is 30%

[agile junco]

I've never used the auto-round feature. Does it automatically overclock for you?

[mossy halo]

auto-round underclocks, manually rounding can overclock too

[wind spade]

7.3 of a smelter is the same as 7 smelters at 100% and one at 30% or any other combination. Not sure what more info do you get by using some auto rounding

[agile junco]

Oh, what if I want to overclock to near 200%, and just round the last row?

[mossy halo]

Because I can set all machines to the same clock speed, than just underclocking a left over one

[agile junco]

It just helps keep the model more accurate in terms of number of buildings, I don't really feel like it's a big deal.

[agile junco]

I do roughly the same thing, but generally overclocking everything to 200% and leaving the last group "underclocked" to like 150 or whatever.

[wind spade]

I clock based on situation. If a solver tells me to build 7 machines, I don't blindly build 7 machines. I build the amount that makes most sense in terms of previous and next steps, with appropriate clocking

[agile junco]

E.g.

Although in this case the last group was actually over-overclocked to 213%.

Auto-clock would be a lot more useful than auto-round.

[mossy halo]

That's not what that meant

[mossy halo]

Just use the plus and minus buttons at the overclock field and increases/decreases the clock speed in whole machines

[wind spade]

what did you mean then? What use it is to get info "8 machines (rounded)" vs "7.3 machines (not rounded)"?

[agile junco]

Oh shit, that's actually a nice tip. Thanks.

[mossy halo]

I build 8 machines in the end, so I round the clock setting to 8 machines to give them all the same clock speed

[agile junco]

Though I generally overproduce by a hair anyway.

[agile junco]

While this is technically optimal from an energy use perspective and I have unlimited power shards now so that doesn't matter anymore... I still only like to overclock a handful of machines.

[rocky ravine]

Is it really the most efficient way to make plastic and fuel or the site is messing with me?

[wind spade]

I build machines depending on previous/next step, so it may be 7, it may be 10, it may be 4. So for me the information "8" is pointless, while "7.3" is accurate

[wind spade]

it's most oil optimal, yeah (the recipes for rubber and plastic)

[unique cypress]

it's not

[wind spade]

though for more optimal, you can add two more alts

[rocky ravine]

i have a 240 oil that i can push to 600 if needed but idk if the alternate are better

[wind spade]

you can just select all the alts and the site will choose the most resource efficient path

[unique cypress]

600 oil can get you 1800 rubber + plastic (total, not each)

[oblique hollow]

Just need 2 more alt recipes

[rocky ravine]

i reach barely 700 something when selecting maximize

[agile junco]

Yea, this autoround feature is kinda useless to me since it seems to only underclock, and now allow tweaking the results including overclocks.

It's only really useful in doing the math for calculating the correct clock value anyway, when distributed across many machines, which isn't exactly complicated math.

[wind spade]

have you selected all alts?

[mossy halo]

I under/overclock round them to the machines I need depending on the previous/next step. So 7.3 is a pointless info for me

[rocky ravine]

not just about the underclock/overclock but having to do the packaging unpackaging thing to be efficient is uh

[agile junco]

Sorry I'm talking about soemthing else.

[wind spade]

you can't round 7.3 in your head? 🤔

[agile junco]

haha, ceil.

[wind spade]

or that

[agile junco]

But then you have to figure out how to get 7.3 from 8.

Small bit of math.

[mossy halo]

Yeah, auto-round just underclocks, sadly. But than manually increase that into overclock territory with the +/- buttons

[rocky ravine]

idk also why it loops between recycle plastic and recycle rubber

[wind spade]

because that's the most oil-efficient way 🙂

[mossy halo]

If I need to place 10 machines for better downstream handling, then yeah 7.3 to 10 machines can't be easy calculated

[agile junco]

yea, what I would have wanted autoround to do I guess is essentially allow me to enter a value in the count field like normal, but it then sets the clocks appropriately like the +/- do.

Seems weird it's not implemented like that.

[rocky ravine]

i'm running out of shards 😢

[wind spade]

but modeler tells you 7 or 8 as you said

[agile junco]

This also surprised me when I first learned about it.

Gave me a fun sushi belt design though.

plastic and rubber all going in a big loop.

[wind spade]

fun fact, you can actuaqlly build it without loop

[agile junco]

I would make two nodes if belt limits are a concern.

Modeler ABSOLUTELY should have an option to split and maximize to belt limits.

I do stuff like this manually all the time and there's no reason it cannot be done for me:

Though, I guess there'd be a question of even splits, or somehow getting similar number of machine rows like I managed to do here.

[mossy halo]

Yep, did the same thing to create the same nodes multiple times to see it in picture how i would split something up logically when building

[lilac lantern]

Does anyone here have a program they recommend to help me start optimizing my production chains?

[wind spade]

what do you want to optimise for?

[agile junco]

I'm confused, why do you have 2.5 constructors as two groups?

[mossy halo]

what you're referring to?

[agile junco]

you obviously can handle 80+20 items/min.

[agile junco]

the biomass at the start of the production.

[mossy halo]

I can, but it comes from two sources, so it was a bit easier to build it that way at that moment I created it

[agile junco]

Sure, I guess.

[mossy halo]

I'm using the modeler not to just crunch out the numbers, but also to plan build layouts partially

[agile junco]

I guess I'm rarely input limited, and I'm more generally trying to make rows which aren't overflowing belts.

[agile junco]

I do the same.

[mossy halo]

that's a modded early power plant setup ^^

so yeah, i was limited to mk2 belts

[lilac lantern]

I plan do do all the components at some point. Currently reworking all my factors prior to stating phase 4

[agile junco]

This part confuses me. You clearly have a 780 belt, yet you split out a single rounded constructor, why?

[wind spade]

no, I mean "optimal" in what sense? least machines? least footprint? least power? least raw resources? something else?

[mossy halo]

Nah, I only have 120/m belts. So the first row is from the last exploit loop giving me 600/m. the second row is consuming the first three exploit loops. But 500/m is not a nice number to build. so the second row is consuming 3 times 480/m biomass from the exploit loops. Leaves each exploit loop with 20/m biomass, which are combined and processed in the third row

basically I build it grouped like that. But merged the 4 constructors each into a single node.

[agile junco]

Then you can step down the clock of the last group to get the 2 constructors like this (as you showed me)

I don't see how autorounding helps at all here.