math-and-meta

[terse otter]

Sanitywise it's ingots

[cloud shard]

I'm not smart enough to do recipe maths using the best possible combination of recipes, unfortunately

[frosty owl]

I'll go for steel screws any day and anytime, even over ingots~

[cloud shard]

steel screws are the best

[frosty owl]

Greeny's tool can help with that ;)
Easy to make comparisons with that (or SCIM too, for simple setups)

[terse otter]

I'll probably go for whatever is cheapest, and make a factory next to each of my inputs

[frosty owl]

Which ones?

[cloud shard]

dunno, but my hard drives are limited, and I'm already diving into T7/T8 with only 10% of the map explored so far

[terse otter]

Oof

[drowsy flicker]

Steel Screw is literally like, 1 Steel Beam is 52 Screws or some nonsense. 5 Beams for 260 Screws per minute

[frosty owl]

There are enough hard drives to unlock everything 👍

[cloud shard]

and my factory setups are all concentrated at the starting location. they're all ultra-condensed like monoliths

[fringe pawn]

IMO bolted iron plate and copper rotor are the only reason you should consider screws, and those are both alts.

[cloud shard]

that's good to hear

[drowsy flicker]

Basically 1 Steel Screw Constructor feeds a Mk3 Belt's worth of screw output its great

[frosty owl]

Have issues with FPS?
Consider spreading out your production! (Making outposts)

[terse otter]

I'm currently trying to hype myself up to go and grab the rest of the hard drives. I've unlocked all the Tier 1-6 recipes already, but want everything before I dig too deep

[drowsy flicker]

Don't forget your Rifle Champ, Lots of Spooders out there

[terse otter]

I'm going rifleless, lol

[cloud shard]

this is why I encased my setups in walls and built them vertically like a tombstone in a 4x3 footprint

[terse otter]

Gonna ninja all these chumps

[frosty owl]

I killed like... 2 spiders with the rifle... That's how much I used that

[drowsy flicker]

Well, I use the Rifle as they approach me and then give them the smack down, as it should be

[cloud shard]

my factories end up looking like a mini-cemetery

[drowsy flicker]

Xenobasher is just too fun to use man

[cloud shard]

with a storage container in front of each "tombstone" containing the item produced

[fierce ruin]

why smack when you can snack :inhaler:

[frosty owl]

Funfact: buildings spanning horizontally are easier on FPS than towers 😉
Very high factories are the worse FPS wise

[cloud shard]

I wish I knew that sooner lol

[drowsy flicker]

Because the way the game loads things is a radius around you (Including Up&Down), If you're in the middle of the tower RIP Frames

[terse otter]

That's good to know. I was going to build a sky train system, but I'll make sure to build it lower now

[cloud shard]

but since their animation's covered by those walls (no belts sticking out of them) my fps has improved a bit

[frosty owl]

The height at which you build isn't really relevant, it's about how much stuff you cram in that height ^^

[terse otter]

Ah, even better to know

[drowsy flicker]

Also, even if you love trains, take it easy on the amount of train systems around you. One or Two train lines visiting a factory do just fine, More? oh boy.

[frosty owl]

I think the game handles up to 10 in the same place decently enough (so long they have less than ~6 cars each)

[cloud shard]

my pillar factories look like a can of spaghetti on the inside

[drowsy flicker]

Fair, my computer is more Potatoe-sided

[cloud shard]

and a generous amount of clipping

hard to fit a 4x3 setup without clipping

[drowsy flicker]

I accept clipping with pipes, but with Belts Oh hell naw

[frosty owl]

I'm glad I haven't seen a screen of that than 😆
No offense, but that is like one of the designs I like the least 😅

[drowsy flicker]

As long as the pipes aren't clipping with like, buildings or other Pipes I don't mind them moving through terrain

[cloud shard]

I clip the vertical belts with the foundation

mostly because I need to have a solid ground to stand on

[drowsy flicker]

There are walkway shenanigans that let you open small holes through the floor instead of foundation sized holes but I don't remember how to do it.

[cloud shard]

I also do this trick where I build the walkway before the foundation, to space the foundation apart by half, so I could run a vertical belt up

and use 2 splitters to do 90 degree belt turns

[drowsy flicker]

I started accepting 45 degrees in my factories recently, they're a little tough to do but I like em

[cloud shard]

I do those when there's ample room

right now I'm pondering if I should zigzag my belts upward instead of using elevators, because the elevators are bulky in terms of horizontal footprint

[fierce ruin]

zig zag would take less?

[cloud shard]

they're cliptastic

[frosty owl]

I can't stress enough how much cleaner my designs became once I started aligning machines with wall conveyor holes rather then cramming them next to one another
Walking space everywhere, belting space everywhere, decorations go brrrrrrrrr
And also makes it easier on the designing itself

[cloud shard]

the item elevators are more fickle with clipping

my pillar factories used to have conveyor holes, but they end up taking a lot of space too lol

[fringe pawn]

I experimented with feeding all I/O with lifts recently. Being able to walk the entire floor by having everything elevated is nice.

Plus almost no belts have slopes, because for stuff like manufacturers, I just put everything at a different height.

[cloud shard]

my base looks like a literal heatsink, with each heatsink fin maxxing producing one type of item

[drowsy flicker]

Ah yes feeding manufacturers all their stuff, my favorite activity

[frosty owl]

That's why I usually leave the lifts outside the buildings ;)
That also adds a nice decoration to them

[cloud shard]

but fps

[frosty owl]

Lifts are quite cheap on FPS

Cheaper than belts

[cloud shard]

I know, but I'm minmaxing on fps too

[drowsy flicker]

Lifts EVERYWHERE

Okay that's a bit extreme

[frosty owl]

You already kinda F'd it up by making towers, might as well have pretty towers

[cloud shard]

they're pretty in their own way. by being towering pillars devoid of any feature

[fierce ruin]

laughs in copper laptop heat sink

[cloud shard]

if an item is hard to fit into a 4x3, I usually split the components into their separate towers

[faint mortar]

I'm setting up a train to collect 1350 aluminium scrap and 100 alumina solution for my factory every minute. I have built a train with 3 cargo holds and 2 fluid tanks, which should be able to transport 32 minutes worth of alumina sol. and 35.5 mins worth of aluminium scrap.

A train can easily make a round trip from the desert oasis to the waterfall and back again within half an hour, right? They're pretty fast ye?

[drowsy flicker]

Yep, they're pretty speedy

[faint mortar]

Sounds good

[cloud shard]

are trains good on the fps? last I heard, having a train running is like having a second monitor

[terse otter]

How much throughput does a train that crosses the entire map have?

[drowsy flicker]

I wanna say my cross-map Uranium Train that delievers from the Rocky Desert mountaintop to the Water Well in the north of the desert only has like, a 7 minute round trip or something like that?

[terse otter]

Do you eventually need more trains if you are passing enough materials?

[cloud shard]

now that I think about it, putting uranium in trains sounds like a good idea

less radiation

[drowsy flicker]

Wouldn't know, My Uranium Train is only 1 Freight Car and that Uranium node is impure

I have a second Uranium Train for the Swamp Waterfall that's TWO freight cars, but that distance is also much shorter

[cloud shard]

not sure if uranium is worth the trouble, compared to a row of coal plants

[drowsy flicker]

I basically use the desert as a radiation quarantine for all my nuclear stuff

Uranium is worth the trouble. 2500 MW is a Lot, and you don't need many nuclear plants to be basically set for life. Some people go a bit extreme on them imo

Unless you maing all the Nuclear Pasta the world can give you

[cloud shard]

I was planning on sinking all the nuclear pasta and using coal and oil for power instead

[drowsy flicker]

Nuclear Pasta is just an expensive part to make; Its not actually related to Nuclear stuff

[terse otter]

Oil is the best effort->power from what I can tell, but nuclear seems like enough power to be a good final setup

Coal just seems like a pain to try and keep going with

[cloud shard]

haven't tried fuel generators. are they better than coal?

[drowsy flicker]

Yes oh baby

[terse otter]

Very much so, though they are quite large

[cloud shard]

well, I've had a ton of space saved up

maybe I could build horizontally for a change

[drowsy flicker]

A lot of my factories that use oil end up being Fuel Generator hubs too lol.

Made rubber, have Heavy Oil laying around... crap what do I use it for... Oh Right Fuel.

[terse otter]

If I wasn't planning on heavy nuclear, I would be building a fuel megabase

[drowsy flicker]

This factory uses Rubber (Not that much, so didn't feel like doing the whole recycle routine) in a bunch of things, uses some of the Heavy Oil for Coke Steel/Aluminum and the rest for that sweet diluted fuel power generation baby, a lot of my factories are like that.

I don't go exteme into recycled Rubber/Plastic so a lot of my waste HoR gets turned into power which refunds part of the power used on the factory, its real nice

[fringe pawn]

For most people, turbo blend is probably the best power setup. 15 water, 33.75 crude, and 22.5 sulfur for 10 fuel generators. Plus 22.5 resin to be turned into 11.25 rubber if you want to add water. Few machines, few steps, and lots of power.

[fringe pawn]

Going back to the screw discussion, I toggled steel screws to being allowed within my plan, and overall building count and power use went up. The lower building count at a rotor factory and other places can't make up for the quantity of screw constructors, in my case.

[drowsy flicker]

Steel Screw's main benefit is less energy and Extremely small space/building requirement, sounds a bit backwards

Solid Steel also toggled, I assume?

[fringe pawn]

Yep, solid steel ingot with pure iron ingot.

Overall iron use went down, so there were fewer foundries, and fewer factories down the production chain. But that couldn't make up for the high overall machine count of the 62 screw constructors themselves.

[drowsy flicker]

The fuck, 62 Steel Screw constructors would make 16 THOUSAND screws a minute

[fringe pawn]

This is in a chain producing the final space elevator parts in a 5:5:20:20 array.

[fierce ruin]

might as well make it 69

[drowsy flicker]

I think your automated setup may have taken steel screws as permission to use other screw-based recipes?

[fringe pawn]

Actually it's still using adhered iron plate. Turning off adhered plate makes it switch to the default recipe. Forcing bolted plates bumps it up to 93.5 screw constructors. The other screw recipes are copper rotor, bolted frame, and heavy flexible frame. All reasonable choices.

Raw resources barely change. So it's an increase in complexity, machine count, and power use across the board in general.

[bleak coral]

It's heavy flexible doing it. It's a dud of a recipe, the increased resources undoes its speed gains. Switch to encased and see how it turns out.

[fringe pawn]

Toggled, counting

Power is now basically tied (102MW less for screws), and the screw array has a lower machine count, 213 versus 237. A win, but not a dramatic one.

Oof, yeah, heavy flex frame requires rubber, too. PITA recipe.

[bleak coral]

it's a bad recipe, needs something done to it

[fringe pawn]

I'm amazed that even made it into the calculation. Probably because I have a rubber input of 999999 set for simplicity.

[bleak coral]

guess you could try turning bolted frame then plate off, one by one in that order, I know that if you don't have steel screw those have the same problem that flexible does; only steel screw saves them

kinda one-half of the screws are bad equation is they also aren't in any resource efficient recipes outside of copper rotor

[fringe pawn]

Bolted iron plate looked tempting at first

[bleak coral]

and fast recipes are a crapshoot on a macro scale of whether their inefficiency undoes their power/space gains from speed, and can depend on which other recipes are used

[fringe pawn]

For my actual use, it's tempting to introduce copper rotor just to have screws in production. I'll need them for the build gun anyway. And I may even do it with steel rods, just for the sake of the build gun as well.

[glacial hemlock]

Fast recipe sometimes consumes more ingredients, which you have to spend more effort to produce those ingredients, undone the savings from the fast recipe

Copper rotor is nice.

[bleak coral]

we need more wet concretes and steel screws, and less flexible frames

cause if a fast recipe is inefficient, and can't overcome that, it can feel quite useless

[drowsy flicker]

There's also situations like with copper where the refinery recipe eats 10 water/15 Copper to return 37.5 Each, but a foundry will inhale 50 Copper and spit out 100. Refineries are really power hungry...

[bleak coral]

I mean trading something for extra efficiency is pretty bog standard, whether that's complexity, power, or space or some combination of it, that's fine

all the recipes are careful balances of speed, efficiency, and complexity, my issue is when something doesn't do any of them well

[drowsy flicker]

Simplicty of Caterium Computers and Oil (Rubber+Coke) Circuit Boards saved past me's soul in the past

[fringe pawn]

People hate on electrode circuit board for some reason, but there's something to be said for the option to make circuit boards purely from an oil well. There are map locations where it makes sense due to nearby caterium.

[bleak coral]

oh I think it's bad, but I see the appeal in just making it from one resource

it has some sort of redeeming quality, as weird and niche as I personally think it is

[fringe pawn]

Silicon circuit board blows it out of the water for most people. That's an excellent recipe.

[bleak coral]

well, everything else does a better job if you consider any sort of efficiency

[fringe pawn]

Electrode circuit board and heavy flex frame seem like they accidentally left the output quantity field blank, defaulting it to 1?

[bleak coral]

idk, maybe electrode, but I don't know what to do with flexible, I don't think it's efficiency is terrible but how do you make it faster when it already needs that many screws?

I guess you could make heavy flexible the efficient one, and speed up heavy encased

like reverse their roles, heavy encased already has simplicity going for it too so I think it'd be fine

[fringe pawn]

Heavy encased is basically a straight upgrade of the original

[bleak coral]

it is, the original actually is worse than flexible, and maybe flexible wouldn't look so terrible if encased wasn't such a superstar lol

[fringe pawn]

I'm also realizing that you may not even be able to use copper rotor effectively when you unlock it, as you may only have MK2 belts.

[bleak coral]

bolted frame and plate have the same issue

bolted frame and plate should be saved until mk3/4 and steel screws anyway so meh

[fringe pawn]

Looks like I'm also going to switch to the default beacon recipe. Mostly so I have cable in the system again.

[glacial hemlock]

There are many reasons to use other alternative recipes as long as resources efficiency is not the top priority. For min/maxer, there is only 1 choice.

[wind spade]

I wouldn't say so, it depends on what are you maxing

[cloud monolith]

Thank you for the link, interesting stuff!

I did some experiments myself. I stripped down my factory logistics to the bare minimum and came up with the following test setup:

So it seems, that also connections between belts and elevators are reducing the overall throughput . In my setup, the amount of connections (in)between belts and elevators reduce the throughput by ~5.5 items/min.

I'll continue testing...

[fierce ruin]

i didn't test anything about elevator, i just assumed they would probably act like belts (but it's good to have someone check that; thank you. i also like your testing setup: simple and effective).

[frosty pawn]

I thought we already figured this out last week? The problem is simple: when an item reaches the end of a belt (a single section of belt) there is a timing error where it has to sit there in between ticks for a small amount of time before it can be moved to whatever is connected to it. If you have a long series of belts, the problem compounds into a much more noticeable difference between the number of items going in at the start and the number coming out at the end.

Basically items progress along the belt section by a set amount per tick depending on how long and how fast it is. So, for example, if one tick progresses the item along the belt by 9% it takes 10 ticks to get 99% along the belt and then one more tick means it reaches 108% so the belt had to stop for 8% of the total time before the item is transferred to the next section (or building). In practise the percentage is much more tiny than 9% but the result is similar - the item has to sit on the belt for a fraction of a tick so you actually get less than 780 per minute.

[fierce ruin]

That is your claim (and that's a decent explanation), but that doesn't hold well with my testing, or not entirely: mainly because using splitter/merger mostly fix the problem (just some back up on load/save and not during game).

so yes; discussion carry on with new results since last week.

[frosty owl]

The more data, the better~
All seems going according to what we noticed so far though, right? ^^

[wind spade]

that may be because of the splitter/merger internal inventory

[frosty pawn]

They do not completely solve the issue. Splitters and mergers have an internal stack of 8 items for a buffer. if you let the system run long enough, it will have the same problem but the number of items lost per second will be less

[fierce ruin]

i guess it's the reason, yes (but i filled their buffer first, so it's not just that they "hide" the back up).

[frosty pawn]

also if you do not fill the belt completely (e.g. put 700/min onto a mk5 belt) the spaces between items also act as a buffer so you have less loss of throughput.

[frosty owl]

I don't see the discrepancy between that and what @frosty pawn said...
Adding mergers not only takes away the direct connection between belt segments, but also adds buffering
Note: slapping the splitters/mergers on top of an existing conveyor does nothing for the issue, proving the segment-segment connections are the issue

[fierce ruin]

read the tests i have made, the link is just above; i filled their inventory first.

[frosty pawn]

this is why splitting a mk5 belt at the start into 2 mk4 belts and then merging it at the end is helpful

[wind spade]

well in general if you're not maxing belts you're fine, right?

[frosty owl]

Not maxing them for over 1 belt segment

[frosty pawn]

yes, not maxing belts is helpful. same with pipes.

[wind spade]

isn't the pipe max issue resolved by doing a loop?

[frosty owl]

So balancers can still go brrrr

[fierce ruin]

sure. it's the best solution. But that mean you can also max a short belt from a miner to a splitter without problem.

[frosty pawn]

doing a loop means the pipes arent maxed any more, so yes

[fierce ruin]

from my test, i don't have flow problem on normal pipes, just with cross-junction. So yes, a loop on the manifold should be enough of a fix.

[frosty pawn]

balancer vs manifold is moot because after the first section of the belt, you have a splitter taking some items off the main line directly into a machine

i dont make loops to solve flow problems, i use verticality to control the direction of fluid flow and have never noticed a problem (but that doesnt mean there isnt one)

i only put junctions horizontally when they are merging fluids because i want each input to have equal priority, but i dont do that for outputs

i dont want inputs and outputs to have equal priority

[wind spade]

no I mean that a long maxed pipe with a loop at the end for machines works as well, as opposed to belts, which break when long and maxed
e.g.

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

[frosty pawn]

this is because none of the machines take fluid out at a constant rate, so you need some amount capacity remaining in the pipes as a buffer. the first splitter effectively creates the option of using the 2nd longest pipe section as a buffer

you could remove the loop and put a fluid buffer just before the manifold part of the pipe and you would again have no issues

except the first piece of pipe after the buffer would need to able to support more fluid throughput that the total of all the machines, because the amount of fluid passing from end to end does not stay at a constant rate, it goes up and down and averages at the sum of all the input numbers of the machines

[topaz hedge]

Long pipe at max feeding a loop will break when it enters the loop

I see my mk5 belt testing has continued far beyond what my small brain could come up with lol

Well not really my testing but

[frosty pawn]

if you build a manifold like this, the only section of pipe that will cause a problem is the one between the buffer and junction because fluid required at all of the outputs combined fluctuates between 0 and infinity, averaging at the number you probably expected

            |
            +--
            |
            +--
            |
--[buffer]--+
            |
            +--
            |
            +--
            |
            +--```

[topaz hedge]

With pipes and buffers.. from what I've seen, if input doesn't exceed consumption by a small margin, the system will run low after enough hours/save/load cycles..

[frosty pawn]

yes, exactly. As long as you havent calculated to push 600/min through a pipe, you will be fine

you dont really need to build a fluid buffer though because each machine has its own internal buffer which is actually quite big when you add them all up across the whole manifold

[topaz hedge]

Even if you're just using 150. If you're producing 150. After a couple hundred hours, you'll start to notice empty/half full pipes

I think they're currently looking into that issue though

[fierce ruin]

i think it can happen because of save/load. and i didn't notice that while the game was running

[frosty pawn]

well yes, because the average of 0 and infinity is half of infinity, which is more than 600

all the machines have to power up when you load the game and that takes a moment

even if theyre not supposed to have a warmup time, the game has to calculate stuff

[topaz hedge]

Yeah. Hopefully they'll come up with some kind of workaround for that.. nothing quite like loading your game and your whole nuclear plant drops out and your fuse blows lol

[frosty pawn]

that's what power storage is for 😄

[topaz hedge]

Sure, but it still shouldn't happen regardless

[frosty pawn]

there are many things in the real world that shouldnt happen, but they do and we have to account for them

[topaz hedge]

This is game though. Lol

[frosty pawn]

there are also things in the game that shouldnt happen, like lizard doggos finding slugs and stuff when they are trapped in a box O_o

[topaz hedge]

This is true. I think the whole power plants now consume 100% is more of a headache for them than us. Because these are things we didn't really notice before, or not to this extreme and now they have to fiscit

[frosty pawn]

yeah and like snutt said on stream, they set the tick rate as low as possible to make these rounding errors and such more visible so that they can be found and fixed before they raise the tick rates

[topaz hedge]

By far the worst is when you don't put a floor on your box, and the poor doggos fall though the ground lol

[frosty pawn]

or when your vehicle falls through the ground and it had all your stuff inside D:

[topaz hedge]

Is the tick rate adjusted for us on EA/ex or on their internal testing? O.o

[frosty pawn]

i think its the same everywhere, but its just deliberately low

[topaz hedge]

Sad moments. And ah alright. I've been on factory burn out lately so.. I haven't really been keeping up as much lol

[faint mortar]

Ok heat-fused frames are defo worth it

[frosty owl]

You started it all better take responsability

[cloud monolith]

Did some more experiments. As it turns out, the amount of belt-belt interconnections is not really relevant for this issue to appear. However, I observed a higher variation in smaller setups, e.g. with only 5 belt-belt-interconnections, the loss in throughput varied between 2.2 and 12.6 items/min.

[fierce ruin]

from my test, more belts mean more back up, but i tested just with larger amount of belts than that (min 20). I will make some other checks on that.

[sand epoch]

I have yet to see any item loss, regardless of belt size.. no clue what everyone keeps talking about.

[fierce ruin]

it's just if you use belts at max throughput, and specially for high speed belt (mk5, maybe mk4); other than that it should be barely noticable.

but if you try to push 780 item/min on a long mk5 belt, you will find that the real throughput is a bit lower than that.

(and your fps probably also play a role here, the lower the problem the bigger the problem is; probably)

But all of that is easily avoidable: just don't try to push max throughput on a belt. (that's the important thing to remember)

[fringe pawn]

Are low FPS an issue if you're across the map outside of a belt's render range? I've got a 1.5km feed from a pure bauxite node that I'm curious about. It immediately spits into an ISC, then splits to to MK5 belts. It's in the top middle of the bamboo forest, and travels West to the ocean, going down the cliffs to refineries.

[fierce ruin]

I don't know a lot about how FPS affect that. But if you split directly into 2 belts (not full) i think you are fine (all tests that i know showed that miner->ISC->2 belts works fine if you have just 1 belt between miner and ISC)

[cloud monolith]

Very helpful, thank you all! How much does low fps affect the throughput, are there any numbers?

And another thing: Is there any recommendation about chosing the right belt, e.g. regardless of required throughput, just pick the fastest one versus picking the one that fits the required throughput nicely? Thinking about manifolds and so on...

[oblique hollow]

not really, just estimates.

and even then its unsure

[fringe pawn]

Never use minimum spec belts, as it limits your option to overclock. This has happened to me many times where I wondered why machines weren't performing well after I overclocked them, but I noticed the input belts were the bottleneck. I think a 50/minute input item went up to 125, but I had only been using MK1 belts. Once they're down, it's easy to forget about low level belts that you've built, so now I always do max spec.

[terse otter]

My counter opinion, never overclock a system past what you design it for. Encapsulating everything and closing it up, and just building more if you need more, will save on throughput headaches

[cloud monolith]

My impression is, that manifolds run smoother, when you choose min. spec belts 🤔

[terse otter]

Manifolds don't really care, since it'll run suboptimally until all the machines are full.

You can sorta force splitting logic with throughput limits, but I think it's better to just give them ample throughput.

Too many bugs related to maxing out belts

[fierce ruin]

it's true when you start the manifold; but it shouldn't be a problem when machines are full enough. A good solution is to first fill the machine entirely, then let it run.

[fringe pawn]

Yeah, I like to see inputs jammed before I allow a system to flow. Sometimes watching how it unjams can reveal problems, as well.

[terse otter]

An empty belt is the best way to tell what is missing, yeah.

[fierce ruin]

hi

[fringe pawn]

I think deleting a belt link in the final merged output flow is the best way to perform a final factory inspection. You can check for empty belts in the entire system that way.

[fierce ruin]

thats very true

[bleak coral]

only makes a difference if the belt speed = machine input speed, in which case it'll not need to warm up, otherwise it's the same as long as there's enough throughput"

once it's warmed up it doesn't change anything at all

oh and I guess it might have more output while it's warming up, but that doesn't really matter

[frosty owl]

You can see the effect clearly when you check out a miner trying to output 780/min
Unless you split the belt immediately, the miner will slowly but surely back up over time (items will pile up in its inventory instead of reaching your factory)
Not a big deal unless you're an efficiency freak, I think

[cloud monolith]

Not a big deal unless you're an efficiency freak -> It's driving me crazy! 😬

[frosty owl]

I like to use min spec belts, moslty because it helps me remembering the throughput I want/need on a certain belt later on and because I don't like see the item move by steps on fast belts

Though I also make sure to ALWAYS make the belts upgradable (not hidden between splitters or the like)

[terse otter]

If you pretend like it's not happening the inefficency will actually flow through the entire base fairly evenly. So everything will run slightly slower, but not in an extreme way.

[frosty owl]

To get that perfectly smooth power draw, you should either not rely on maxed belts or keep them always at 1 segment max ^^ (and all the suggestions that have been given already)

[cloud monolith]

Agreed... I know what I'll do the next 50 hours in game 😂

[fringe pawn]

You can also plan a small excess at every step. In planning my final space elevator parts factory, I rounded turbo motor production up to the next 100% factory production increment by telling Greeny's tool to produce 2.5 spare turbo motors, which created a corresponding increase in machines in the rest of the system. Then I moved downstream to pressure conversion cubes, and asked the sheet to create .65 extra of those to round up the machine count. Then I told it create 2 extra fused frames to round that machine count up. So I'll have a little bit of excess at every step. Part of this is also for build gun use.

[cloud monolith]

That's exactly, what I do. However, my current challenge is to max out production with exactly 1 belt per item type - see how far you can get. (Main bus, no injection)

[fringe pawn]

Even ores? 😱

[cloud monolith]

Jup 😁

[frosty owl]

Warning: this method could still encounter issues if you were to use maxed belts
Eg: Trying to output 780 copper sheets from a factory ending up in some machines slowly piling up items in their output, eventually losing efficiency (and of course, not providing enough sheets for the next step)

[fringe pawn]

I guess there's some nice simplicity in running everything out of one miner.

I'm guessing no screws for you?

[cloud monolith]

Oh yes, and it's really fun to optimize even the last bit. And then I discovered this max-belt-thingy.

[frosty owl]

I hope you're not including 500 stacks items in the 1 belt limit (Exept for concrete ofc)

[cloud monolith]

Well, 780-x screws/min 😅

[cloud monolith]

Oh yes...

[frosty owl]

Can't do that much production with just 780/min wire and quickwire :/
Unless you keep them off the bus for local productions ofc

[fringe pawn]

I guess you could do direct feed from screw constructors to copper rotor assemblers.

[cloud monolith]

Currently at 2 Turbo Motors / Min, .4 Thermal Propulsion Rockets, .8 Assembly Director Systems

[wind spade]

main bus 🤢

[eager solar]

you sure you want to try going for 780 pasta/min? is that even doable with existing nodes

[cloud monolith]

Don't think so 😂

[oblique hollow]

Factorio practices in Satis, eeeew

[sand epoch]

to storage only.. lol

[wind spade]

then it's not main bus but just stacked belts

[sand epoch]

everything on 1 belt is not main bus?

[wind spade]

main bus is when you have stacked belts going through your base, splitting items to factories and merging products back

it's something that's used a lot in Factorio, but in Satisfactory it's crap and not really useful

[fringe pawn]

Curiosity is getting the better of me, let's see what happens when you tell the tool it can only use 770 of each raw resource.

[sand epoch]

error screen

[frosty owl]

Try making a big overclocked factory in a canyon and that statement loses a lot of power 😉

[wind spade]

no matter what, main bus is crap in Satisfactory. It's much better to use directly what you have on belt than merge stuff onto a bus

[frosty owl]

I think that just means taking forever to fill up the space elevator

[cloud monolith]

We are currently experimenting with different designs after 1500+ hours in the game. We never did a main bus, nor did we play Factorio, so...

[frosty owl]

If your machine density is high enough (OC, or working on an area producing only items needing high throughput), you can easily build around a central bus
Eg: bus going from input freights to output freights from one end of the factory to the other, buildings take unloaded items from the bus, unload the output on the same belt to be carried to the loading stations or used by other buoldings along the way

[wind spade]

if it's just a few stacked belts transporting items from A to B, then it's not a main bus. If otherwise, why merge belts just to split them again, when you can use directly what you have on belt from start?

[cloud monolith]

That's true, we really had fun when we set the goal to produce as many fuel rods as possible just using only 1 central belt that way (in update 3), facilitating smart/progr. splitters.

[frosty owl]

If belt A has items for for 3 buildings, you can either split the belt and feed each building or have the belt run along them splitting along the way AND getting the output back
If buildings are all around the bus (which depends on how dense the area is with machines), you can then use the bus' belts to manage all the outputs/inputs from one place (and do belt shenanigans with mixing if one wants, allowing for great ease of expansion along it
Eg: not needing to route belts through 3 buildings to go from A to B, if you haven't planned for B in advance: just have B close to the bus

[wind spade]

OR you can build the buildings at the source, save materials, space and fps

[terse otter]

The reason main busses don't work here is lag. building everything up close in a typical mainbus fashion is liable to drop a lot of frames.

Otherwise it would be my personal go-to

[frosty owl]

Ahahaha, I can see that
I had an alluminium setup working only with 1 mixed belt + 1 belt of bauxite
Even the scraps where managed on the mixed belt :jacelul:

[sand garnet]

well that, plus it just being unnecessary due to unlimited nodes.
It's also a pain to manage

[terse otter]

With verticality, a main bus would actually be really convenient.

[cloud monolith]

Yeah, sometimes it's really fun to try some custom challenges!

[frosty owl]

That's an "or" you could place after anything 🤷‍♂️
And OC generally IS FPS friendly

[fringe pawn]

Limiting yourself to one of each pure node, The Sheet claims it's possible to get .15 : .15 : .75 : .75 of each final space elevator part per minute. https://u4.satisfactorytools.com/production?share=iqgnCFfknO1ztdZ7Bteh

Copper is the closest to the limit at 731.3. It has a claimed peak power consumption of 2682, which would be easy to run from geysers. The Sheet doesn't include infrastructure like water pumps and water extractors, but it also generated a scenario requiring only one water extractor shared by alumina solution and cooling systems.

[wind spade]

still if the belt has to transport stuff for three machines from A to B, then it's not a main bus

[terse otter]

I imagine the build becomes more efficent if you add more of the common nodes to the mix

[frosty owl]

"If your machine density is high enough [...], you can easily build around a central bus"
With that I'm assuming you have buildings filled with machines on top, to the sides and below the bus, on at least one floor/building in each direction
Otherwise it's pretty much a waste of belts

[terse otter]

It's belt expensive, but the trick is in how you set it up. A lot of resources are end-products, so you don't have to belt everything.

And in terms of verticality, I was mostly referring to stacking belts for better scaling throughput

[frosty owl]

I don't quite get what you mean (I was referring to brining items for 3 BUILDINGS)
If the bus feeds multiple buildings and take items from them, it sounds in line with your description of one (main bus is when you have stacked belts going through your base, splitting items to factories and merging products back)

[frosty owl]

I think a bus not using stacked belts would span quite a long way horizontally

[terse otter]

The core benefit of the central bus approach is simplicity and extendability. By branching crafting away from the bus in a straight line, you ensure that any part of the base can be easily expanded, and by putting all resources needed into a central line, you ensure that everything is available when you unlock new items

[fringe pawn]

I'm an idiot. I forgot to enable alts in the scenario I just posted. So that's with all default recipes, too 😛

I guess that's cool in its own way, especially only needing one water extractor.

[wind spade]

sure, but comes with so many disadvantages (for SF), that it's not really worth 🤷

[terse otter]

In SF, lag will prevent it 100% of the time

[frosty owl]

Agreed, though I think it can be useful without being overly wasteful only in areas very dense with machines
Eg: refineries usually don't make it worth it due to their low output/(space taken) ratio

[terse otter]

The biggest non-lag related problem to the idea is probably pipes. The throughput on liquids is too low to afford adding everything to the main bus, so building the crafting facilities for related stuff in the ocean is much easier.

[wind spade]

not just lag. The fact that SF belts are much slower (fastest SF belt is slower than slowest Factorio belt), items required in large quantities, etc. make it not really worth as you need tons of belts. Another point is that nodes are fixed and there's a sink, so production and consumption are constant. And finally, the need to constantly keep track of how many available resources are on the bus...

[frosty owl]

You still need to keep track of the items going through your factories one way or another, having them all in one place makes things handy (especially with efficiency checking mods)

[terse otter]

For the large item count issue, there is a similar issue in factorio when dealing with high item count things such as copper wire. The solution used is to belt the more compressed resources before/after the item of note. A similar approach could be taken here in many cases.

For the slow belt issue, stacking belts is incredibly easy, and mass producing them is as well.

[wind spade]

you don't if you build modular factories that make ores -> products 🤷

[wind spade]

and losing fps as well 🤷

[terse otter]

Never going to disagree with the fps issue

As I said, fps is the only reason I'm not doing this

[wind spade]

"you need a lot of belts" is a complaint about fps, not complexity of build or materials needed

[frosty owl]

That's just a small percentage of factories though

[wind spade]

can be done for everything

I usually recommend that playstyle as well

[frosty owl]

Ofc it can... Doesn't mean it gets done or is the most fun thing to do always

[terse otter]

Hmm... What would happen if you belted resources into the distance? It would be stupidly expensive for little gain, but hiding all the production away could save fps, right?

[frosty owl]

Eg: the canyon factory I mentioned earlier could be managed very nicely from its bus. And a the bus' belts still had to be done one way or another since there was no space to place train stations close-by, only in certain areas along the bus

[frosty owl]

A train would be the most FPS friendly way imo

[terse otter]

Actually here's my "main bus" idea for this game. One major line of resources. Train stations coming away/to the line into the extensible crafting facilities.

If it's just the belts carrying stuff, and they decrease in number as you shuttle stuff away and bring in lower quantities of crafted stuff, it should be a lower fps cost

It also act's as a neat hub for trains, simplifying all the builds

[frosty owl]

I'm not sure I get this correctly...
Isn't that the same of the bus I described but with trains connecting to buildings rather than belts? :thinking_helmet:

[wind spade]

this is #math-and-meta , here we talk about most efficient stuff 😛

[terse otter]

Could be the same idea, not sure on the comparison

The main goal here is to compile your production and ensure extensibility at all points.

[frosty owl]

I see no mention of that in the channel description ||Really, no "monkey covering eyes" emote? ||

[frosty owl]

The best way I found to do that (in high density production areas) was leaving a belt space where I could route all the belts to different buildings
But having them on a bus make things much more orderly
Unless one builds everything 1:1 or similar ofc~
(Huge buildings with refineries are still best dealt with using stations usually, though)

[terse otter]

If I ever get my base to the size where it's obvious what the main bus build looks like, I'll post the map

Right now, I'm still trying to finish unlocking the last couple things so I can scale up power for all of this.

[fringe pawn]

I enabled alts in the 'one of every node' idea, and it's fascinating to watch Greeny's tool work by using both coke steel ingot and solid steel ingot to deal with the coal bottleneck. 1.4 : 1.4 : 7 : 7 plus 90 turbofuel per minute is possible.

The challenge of precise distribution in executing that would be quite an undertaking.

[frosty owl]

Those are decent enough numbers ✅

I mean, most people finish the game with less, so "decent" might be a little weak

[fringe pawn]

Machine count would be somewhere in the mid 800 range, factoring water extractors. The 20 turbofuel generators might allow you to overclock most things and get the machine count to 300~

[terse otter]

That's only 10ish hours of production to win?

[fringe pawn]

Slightly under 12 hours.

[terse otter]

It would work, but I'd rather build a megabase that can do it 10-20x faster, even if it's unnecessary.

[frosty owl]

The "One Node Man" challenge gets real

[fringe pawn]

It makes me wonder about the tightest place to set it up? What's the smallest map footprint that has one of every pure node, plus allows 17~ water extractors?

I think you'd be limited to the desert and swamp.

[terse otter]

I think one potential spot is east of the western beach. Dunno about the eastern half of the map

[frosty owl]

Why need to have everything there? Can't use a train?
I mean, at that scale a single train could handle all resources :hehe:

The "speed runner spot" has a decent variety of nodes while being low on building space close to the ground

[fringe pawn]

A tight footprint is just an addition to the concept. Definitely unnecessary. Though it looks like it only needs 305 sulfur, so at least you're not limited to by the 3 pure sulfur nodes on the map.

[oblique hollow]

@sand garnet

the external connections "A, B and C" operate the internal "switches" A, B and C

if A has a signal, then the switch is closed

if you wanted to you could add an extra indicator for where the "power" for the internal circuit comes from

[sand garnet]

see my first instinct when reading that image is
A turns into B turns into X
B and C (AND gate) turn into Y

[oblique hollow]

a redstone OR gate is literally just a redstone dust connection like the B and C one (where that diamond is the connection)
and an AND is just 2 redstone "switches" in serie

this is a more common way

[sand garnet]

the thing I could see being an issue is that this is just.. 'less fun' to set up, even though it's more in the theme of satisfactory

[oblique hollow]

as if logic is ever fun to set up xd

[sand garnet]

I dont play MC, but from watching Ethoslab, it seems a lot of fun

[oblique hollow]

they could just as well make funny little icons and basically have Factorio in Satis

just no factory building. only logic

[sand garnet]

logic simulator 2021

[oblique hollow]

the best game of the century. FICSIT-Approved

[fringe pawn]

I eliminated some recipes to make it a little simpler, but I feel like there's some room for improvement in the 'one of every node' concept. It's using heavy flexible frame! https://u4.satisfactorytools.com/production?share=GKGqKS2CGK2Ch6YECbsX

[bleak coral]

Do still have rubber set really high?

[fringe pawn]

No fake inputs are set. Just 770 for all ore nodes and 600 for crude.

This is different than what I'm actually building, it's a theory based on a discussion that took place earlier.

With strategic overclocking I'd say a building count between 350 and 400 is possible. 1.4: 1.4 : 7 : 7 space elevator parts and 90 turbofuel.

[frosty owl]

There's still some quartz left
Couldn't you squeeze in more production by making that into silica to increase alluminium production?

[fringe pawn]

I tried pushing toward maxing normal alu ingots, but it's not enough to make a difference.

[toxic wadi]

so i jsut finished my class in linear algebra and im wondering if anyone has found any applications for lin alg in satisfactory?

This is because one of the example questions in the course was on refineries and I saw a few parallels with how the game works but i haven't wrapped my head fully on how i can apply that kind of math to the game

[versed violet]

So far the "hardest" problem I had to tackle with math was following question:
Given belt limit of 780 at miner and 780 at truck station,
What proportions of Quartz crystal and Silica can be produced from 780 raw quartz to fit into 780 limit on truck station belt?
37,5 raw quartz = 22,5 quartz crystal
22,5 raw quartz = 37,5 silica

[fierce ruin]

Linear algebra can be used to solve optimisation problems like "what the minimum ressources i need to produce some given items" or the other way around; basically satisfactory calculator like greeny's probably use some form of linear algebra.

[toxic wadi]

im kinda interested to see if i can find some differential equations problems in the game

[versed violet]

Minimum power needed to mine X units of Iron, assuming minimum miner clock of 100% to avoid weird areas?

[lucid geode]

The awesome sink points per ticket increase rate of change is a differential if i remember correctly. You can integrate it to get the total points required over a given ticket range too, which might be more useful.

[fringe pawn]

I need to merge 2 normal (600/minute each) and 1 impure (300/minute) sulfur node. Can I have the two normal nodes feed to an ISC, and the impure node split and merge with those two feeds to the ISC? Then two MK5 streams come from the ISC towards production.

[versed violet]

Put a splitter onto each of three incoming belts, pull 2 outputs of each, then merge in two mergers going into target?

[fierce ruin]

differential equation are classical for study some movements, even if it's not the main point of the game, you have some problem that can use differential equation if you want (like you have a train with less engine than needed for a slope, but some initial speed; will he get on top of the slope before stopping; or simply studying trajectory from jetpack).... i wouldn't say that's really interesting to study these exemple in satisfactory; but if you want to find application it's easy to find some.

[versed violet]

Basically this, just with one merger less. (or can build the third merger in case you want to connect 3rd thing later)

[versed violet]

Find the best shape for train tracks, so the train arrives at second station on same height in shortest time 😉
-> https://en.wikipedia.org/wiki/Brachistochrone_curve with height recovery and friction.

[steep jewel]

ifi want to make two different things for example screws and metal rods should i take a portion of my metal rods to make screws or start a whole other line of things to make the screws?

[frosty owl]

You should leave at least 1 rods constructor free for your own use (storage), so plan based on that ^^

[versed violet]

depending how far in game are you - there is an alt recipe that allows making screws without making rods

[wind spade]

for example my whole tool runs on linear programming (simplex algorithm)

[naive quarry]

^ if you’re interested the DPV textbook for algorithms is available online for free, check the chapters on linear programming and on network flow problems for cool linear algebra applications. Network flow especially should interest you probably

[fringe pawn]

Ugh, these recipes that extend out four decimals like heavy encased frame. 2.8125 per minute, which pretty much locks you into increments of 45/minute with 16 factories if you want even numbers.

[wind spade]

underclocking

[fringe pawn]

Or overclocking, but either way it's a kink in the system to deal with

[glacial hemlock]

The best recipe for making screws still requires one to make rods first.

[thorn pollen]

is anyone good with pipes

im wondering how many walls does one pump need before i need a new pump

[fierce ruin]

1 wall = 4m; basic pump = 20m headlift

mk2 pump is 50m headlift

other machine have 10m headlift

[placid compass]

its around 4 i think to do it comfortably but it should just snap to the end of the headlift

[fierce ruin]

and after putting a pump, you will something that show you on the pipe where to snap the next pump

[thorn pollen]

do valve cut back flow?

[fierce ruin]

yes

[thorn pollen]

thanks for your help

[versed violet]

depends on your definition of 'best'
Casted screw eliminates need for rods, one machine less.
For really big volumes I prefer steel screw -> beams into load of screws.

[wind spade]

for really big volumes I prefer not to make screws altogether, for anything else I'd say steel screws, casted is used for like two hours

before you switch to screwless recipes

[frank mesa]

Just processing 480 raw quartz, only 13 manufacturers needed

Remembering is constructors you need, an configuring the first one before you autoconnect with Smart

[oblique hollow]

Pull out the pump holograph and slide it along the pipe. It snaps to the limit of the previous pump

[bleak coral]

unless the pipe goes horizontal/diagonal for too long, then it fails. it works perfectly if you're just going straight up though

[glacial hemlock]

If you are looking for volumes, then use the steel ingot to craft rod first then screw will create more screws

[wind spade]

Or just

[wind spade]

Doesn't belong here

[bleak coral]

<@&387163995947270144> & #off-topic-media too

[wary tulip]

Steel screws work the best from what I’ve seen. It’s easier to transport steel around and make it into screws, then to mess with any other systems.

[fringe pawn]

I'm only I'm going to be using screws for copper rotor. I'm going to use steel screws to feed the rotor assemblers directly from the screw constructors, no merging or anything. So it's more like I'm changing the recipe for copper rotors to use steel beams.

[fierce ruin]

steel rod into screw does the same (with more machines but more ressource efficiency); but i'm with greeny on this one: a good screw recipe is to never use screw.

[loud heron]

The less complexity the better

Unless it's an extremely hard to make resource.

I'll always take classic battery

[fringe pawn]

Classic battery uses fewer resources overall anyway. You add a little bit of oil to cut down significantly on sulfur and bauxite.

[fringe pawn]

Back to screws, assuming solid steel and pure ingot recipes where applicable, copper rotor with steel screws has a lower building count, building footprint, and power use in the entire supply chain than steel rotor with caterium wire.

For the whole production chain, excluding miners, but including water extractors, it's 43 versus 53 buildings. Copper rotor has a negligible win in power use in the whole chain.

If you underclock the screw factories and match them up to be 1:1 with the rotor factories, copper rotor still wins on building count and factory footprint for the rotor factory itself, as it's 9 constructors and 9 assemblers versus 20 assemblers to produce 100 rotors.

[bleak coral]

I think steel rotors one and only purpose is a simplified motor setup so you can use the same materials for rotors + stators

which to be fair, is pretty neat

[fringe pawn]

There is some nice simplicity with steel rotor and the default stator and motor recipes. If you want to throw building count to the wind, a single pure iron node and single pure coal node can support 75 motors per minute, which is plenty most people. Using pure iron, solid steel, steel rods for screws, iron wire, and steel rotor. The catch is, that's 324~ buildings.

[versed violet]

Check-my-math moment:
I have 3 rows of 13 ISC (=39) for my waste storage. Currently using 2xNuke @ 250% and planning to double/quadruple that.
SCIM says I have stockpiled 315,245 units of uranium waste, so I'm still good to go for couple tens of playtime? Or should I look into recycling system before doubling my power generation?

[fringe pawn]

At one point I did the math and determined that millions of plutonium rods would only reach 500m. You're fine.

[versed violet]

Getting worried, as it has reached the "complaint management booth" now

[fringe pawn]

Nice.

Can you just move it further out into the ocean? If you go out the the death border you'd need literal years of runtime for the circle to expand enough.

If I'm ascertaining that location correctly, you've got lots of additonal space to the West.

[versed violet]

The thin green line is the Death Border

[fringe pawn]

That's the damage border, not the death border.

[versed violet]

It was the farthest convenient location from the nuclear power plant, while still being able to drive on the ground to avoid border damage

This one is more worried about having to drive back there to put more ISC on top of the pile while suffering the existential damage.

[fringe pawn]

IMO empty your inventory, grab some stacks of concrete and inhalers, then build foundations SW until you hit the instadeath border. Then put up a sign that says:
DO NOT CROSS LINE
(no, not even then)

Otherwise my understanding of the radiation growth is that the radius of the circle grows more slowly as time goes on. So I'd say you're fine as is. Someone explained to me that if you think of it as continually adding rings, because each new outer ring has a larger area than the last, it takes longer to fill. I can't actually be sure that's true, they also gave me the caveat that radiation isn't necessarily 100% understood.

[frank mesa]

In real life it's a sphere, so it spreads even more

as a rule of thumb, 2x more distance is 4x less radiation

[versed violet]

My actual concern is the capacity of the stoarege area

[frank mesa]

i'm planning to use a mod that enables the use of very expensive shielding boxes, still end up with a pile of nuclear waste, but a bit more manageable

[fringe pawn]

Is there a particular hold up with the waste processing? With alts it's only 13 buildings to deal with the waste from 36 uranium plants.

[versed violet]

Only holdup is building it. And planning the location, thats important too!

[terse otter]

I'm trying to figure out waste processing a little. what resources do I need to funnel into it, optimally?

[frank mesa]

planning a waste disposal, building it 10x larger than you think you need, also include the option to do uranium waste reprocessing, and then deal with Plutonium waste (or sink the plutonium fuel rods)

[fringe pawn]

It looks like there's plenty of space remaining on that island, I'd say do it all right there.

[fringe pawn]

With or without alts?

[terse otter]

With, of course.

[fringe pawn]

https://gyazo.com/689810fff1ac50edd7a4b5085be4673f

IMO that's the best setup for uranium waste. That assumes 36 uranium plants.

You could then power 16 plutonium plants with the rods, or sink them.

[versed violet]

I Think I'll make the waste into non-fissile uranium first, then ship that to my secret missile base to convert into plutonium
I have a very good storage space ready for the plutonium rods!

[wooden pond]

what ya storing there uranium or plutonium waste?

[versed violet]

Non-standard question:
Which recipe chain ensures maximum usage of nuclear waste with least products?
Is it the default non-fissile uranium recipe -> default plutonium pelled -> plutonium cell?

[terse otter]

I think the picture a few messages up is it

[versed violet]

Uranium waste for now.

[wooden pond]

looking at the numbers and yeah the defaults are the best for eating the waste

[fringe pawn]

My setup is to maximize total nuclear power per unit of raw uranium ore. It may look different with another goal in mind. The underlying numbers stem from using the impure uranium node in the desert. 300 uranium -> 36 uranium plants -> 16 plutonium plants is the best you can do with that in mind.

[wooden pond]

uses less water to

but the trade off is that you use silica instead of uranium

[versed violet]

I plan on maxing the swamp uranium node (nearly, will leave 20 unused uranium, as this one is too lazy to build fractional buildings). Then I want to use maximum amount of waste to make minimal amount of products that need utilization.

[wooden pond]

How many power plants you gonna get?

[versed violet]

It looks I can supply 60 total, whch is 30 at full OC

[wooden pond]

600 waste/min then

thats exactly 16 blenders to process the waste

*just for the fissile stuff

[fringe pawn]

I believe all the default recipes do that.

[versed violet]

I need to split it somehow for the pluto pellets manufacturing step?

[wooden pond]

2 blenders supply one pellets machine

with the fissile stuff

then ti's 25 waste/min for the pellets

[versed violet]

The uneven number is my fault, as I have a 2 blenders with defualt recipe supplying my current 2 nukes that I don't want to touch.

[wooden pond]

Actually greenys calc should do the math

[fringe pawn]

Underclock or overclock buildings to make the uranium waste distribution split nicely.

[versed violet]

Funny, the two steps use exaclty 100 waste.

So 12 belnders for non-fissile + 6 axeletarots for pellets will utilize the full 600 waste.
Cringing at my power bill with the accelerators.

[wooden pond]

i got the math

12 fuel rod factories

[fringe pawn]

Double the building count and halve the clocks if you're tight on power.

[wooden pond]

12 blenders and 6 particle

2:1 ratio

[versed violet]

I assume with the power I'll be making from that uranium node, power will no longer be a problem 😂

[wooden pond]

i assume your gonna sink the fuel?

[terse otter]

I guess the good news even with all the particle's draining power is that the reactors will be running full tilt at that point.

[versed violet]

For roleplay reason, I'm going to store them in the "Rocket"

[wooden pond]

lol

i wonder if CSS will make the power plants realistic

So if water doesn't come it it asll goes kaboom

[versed violet]

building damage is denied

[wooden pond]

lol

I wonder why we didn't have to process the waste tho

Actual real life nuclear waste needs to be cooled down first

[versed violet]

Maybe thats what the water is for??

[cloud shard]

tell me if I'm wrong, but isn't fused quickwire a bad recipe?

[wooden pond]

thats for the power plant

[fringe pawn]

The numbers work out nicely. 150 waste per minute going to plutonium pellets, 450 waste per minute to non-fissile. Feed all waste to an ISC, and split it into 4 even streams from there. Split one stream to pellets, and the rest to non-fissile.

[versed violet]

for cooling the waste 😉

[wooden pond]

Once the graphite rods are removed they are so raidioactive they still give off heat

[tardy moth]

Actual nuclear power plants recycle water

[cloud shard]

I mean, it uses 5 times the copper so that you'll get a bit over double the caterium output

[wooden pond]

I don't think they do

[versed violet]

depends - if you are short on caterium, anything that pads the caterium is good.

[wooden pond]

Those gigantic cooling towers are spitting out the steam

[tardy moth]

It's literally just steam

[cloud shard]

and uses an assembler and more energy and more piping

[tardy moth]

That's for excess

[fringe pawn]

Is there a story behind the auto-emojis on this server?

[tardy moth]

It still feeds it water but recycles most of it

[wooden pond]

ah cool

[cloud shard]

yeah, but copper becomes more valuable late game, I heard

[tardy moth]

I could also be speaking absolute bs i read on the internet but i'm pretty sure that's what's happening

[wooden pond]

for MOST people the amount of caterium and oil is more then enough

[versed violet]

Caterium is rarer.
My private sheet says Fused is 10% better than base recipe in power usage

[cloud shard]

and a tradeoff between 5 copper to 2.1 caterium material-wise sounds like a bad deal

[wooden pond]

did my 2 seconds of googling, found a paper on it of a power plant in canada

[tardy moth]

was i right?

[fringe pawn]

I'm using caterium for basic wire going forward just because I know I won't be maxing the planet. If I were that'd be a terrible decision, but for QoL if you're just building in the NW corner, high caterium use generally gives you excellent QoL perks.

[cloud shard]

my caterium smelters are all sitting pretty doing a whole lot of idling

[versed violet]

There is 11010 minable caterium on the map vs 28860 copper. Also mind that for caterium 3 ore = 1 ingot, so we are looking at 9 times less ingots with default recipes

[terse otter]

I'm doing default caterium in my meta-facilitys (building materials only)

[cloud shard]

ok that one makes sense

*makes more sense

[terse otter]

Gonna focus on optimization once I've got enough of everything to just build as much as I want.

[versed violet]

Base recipe has better power usage if we assume non-overclocked miners and not maxed nodes

[cloud shard]

still, practically speaking, it's a lot more work for (at best) less than equivalent exchange

[fierce ruin]

it's more work for a very good exchange

[terse otter]

With copper alloy as well, you can put iron into caterium

[fierce ruin]

considering all alt, if you are aiming for 60 quickwire a minute, you trade 26 caterium ore for 10 copper ore; considering caterium is rarer; that's a good trade.

[cloud shard]

iron into caterium sounds like a good deal

[fierce ruin]

what do you mean iron into caterium ?

[versed violet]

[don't tell him about iton wire!]

[fringe pawn]

Copper alloy ingot -> fused wire and fused quickwire

[tardy moth]

fused wire isn't that good though, is it?

fused quickwire is much better

[fierce ruin]

fused wire can be practical; but it's not ressource efficient

[fringe pawn]

If you're going to be tight on copper, you still want pure copper ingot, but copper alloy ingot is best QoL.

[terse otter]

Fused wire is pretty compressed production rate wise

15 resources/min in, 90 out

Probably the best compromise if you are looking to build something fast.

[versed violet]

Caterium wire!

[terse otter]

It's more compressed, but quite expensive

Can't really ignore the extra cost in really big bases

[versed violet]

it all depends whether you maxed the nodes yet. it is very convenient for packaged wire

[terse otter]

I mean, caterium ingots are WAY more expensive then copper ingots, both in ore rarity and recipe efficiency.

Almost 3x the ore, and 4x better recipes

[fringe pawn]

And then there's iron wire, the ultimate anti-efficiency recipe 😛

[frank mesa]

But there's loads and loads of iron nodes

[fringe pawn]

Don't get me wrong, I use iron wire. In particular, I used it to make a beacon factory over a single iron node. But it's a convenience recipe versus an efficiency recipe.

[frank mesa]

True

[terse otter]

It's cheap, but a lot of constructors 😅

[fierce ruin]

considering weighted points, that's the most ressource efficient

[wind spade]

iron wire is more efficient tho

[fringe pawn]

I'm primarily using caterium wire now for my 5:5:20:20 factory (that's still in progress). That would wreck me if I decide to expand those numbers, though.

[fierce ruin]

caterium wire is the most inefficient one, tho...

[fringe pawn]

Not if you remove the weighting for resource values

[terse otter]

At 15 in/120 out, it's a small footprint

[fierce ruin]

weighting for ressources value matter a lot.... iron is clearly not important to save, unlike some other ressources; and even with that, caterium wire cost a lot of caterium

[wind spade]

but that comparsion is weird, recipe using e.g. 1 uranium would be less efficient than using two iron

[fringe pawn]

In my case, I gave myself 3 pure and 2 normal caterium nodes to work with, with the goal of 5:5:20:20. It makes sense to use every bit of that ore if it's also the most power, space, and logistally efficient option.

[terse otter]

The weighted point matters entirely based on factory scale. If you are building small, it's more about what's nearby.

If it's a world factory, it comes more into play

[fierce ruin]

i'm prettyre some caterium heavy recipe later in production line would be better than caterium wire (i can totally understand fused wire... but caterium wire is really expensive)

[wind spade]

you always want to weight resources somehow, doesn't matter how. Pretty much never does one iron ore equal one copper ore in terms of weight

[cloud shard]

I'm planning on sinking some of my extra items. are Assembly Director Systems the best way to convert materials to points?

[wind spade]

thermal rockets

iirc

[cloud shard]

how about supercomputers? I haven't gotten to thermal rockets/ADS just yet

[terse otter]

Just take your space elevator part facility and funnel it into a sink?

[fierce ruin]

!wikisearch sink

[shadow prairie]

https://satisfactory.fandom.com/wiki/Special:Search?query=sink

[fierce ruin]

you will get each part sink value on the wiki page

[wind spade]

supercomputers are 8th best item

[cloud shard]

cool

[frank mesa]

Space elevator items with no further use are sink candidates

[fierce ruin]

Don't you have to take how many you can make into consideration as well as points per item?

[wind spade]

that depends on what you're making and what are you aiming to do

the question about best item to sink has just one answer - thermal rockets 🙂

[cloud shard]

I'm taking into consideration material use, then power use

[frank mesa]

You do, but when you have a large amount in storage, it's worth setting up an overflow into the sink

[cloud shard]

I mean, I could sink every single excess item, but that's just going to end up forcing me to build more power plants. it's inefficient

[fringe pawn]

Amusingly, my use of caterium wire makes quickwire stators worthwhile for me.

[cloud shard]

quickwire stators are efficient?

I thought caterium's rare

[wind spade]

not so much, you just have to be careful what you use it for

[fringe pawn]

For my planned factory, I've told Greeny's tools I've got 3500 caterium ore to work with. It'll be budgeted where it makes sense.

[frank mesa]

Although there is the caveat: when using nuclear power, do consider sinking the plutonium fuel rods to get rid of uranium waste

[cloud shard]

I guess I could trade away fused quickwire and quickwire stators and break even, material-wise

[wind spade]

why so, waste is almost never a problem

[fringe pawn]

I've continued to waffle over my waste strategy, and I've decided to (eventually) build 16 plutonium nuclear plants in case I ever start wanting to use the waste rods.

Otherwise the rods go to containers with an overflow sink that will need an unthinkable number of hours to ever be activated.

[cloud shard]

I'll just yeet myself off the map when it comes to plutonium waste disposal

[wind spade]

that can create death crate on place where you last stood

[cloud shard]

but not always?

[wind spade]

not sure, haven't tested this myself

but I would not use non-automated solution for getting rid of waste anyway 🤷

[fierce ruin]

and if you want to just abuse the game to dispose of waste, might as well do that with SCIM or a mod (i guess a mod do that)

[cloud shard]

there's always the lizard doggo genocide route

but I'm not a monster

[frank mesa]

Automate trucks to drive into the abyss?

[fierce ruin]

you will end up with truck into the void that hurt your game performances.

[wind spade]

they don't despawn and eat your fps

[fringe pawn]

The wiki says the safe Z height is 1997m. If you're extra paranoid about radiation, that's an option.

[frank mesa]

Could consider that a bug and edit out the abyssed trucks in the savegame

[wind spade]

if you want to "cheat" by save editing, why not remove the waste directly? 🤷

[fringe pawn]

Yeah, trying to despawn the waste in the first place is already trying to glitch the game.

[terse otter]

The buggy method of clearing waste caused a bug, so I cheated to fix the bugged bug and now my bug is bug free.

[cloud shard]

I have one pet doggo, but I'm afraid to get more doggos because they can give me a small chance of a nasty surprise

[wind spade]

just set up a container a few hundreds of meters away, the radiation will never reach you

[cloud shard]

also, their old home is mostly parking lot now. I doubt they'll respawn if they get the rifle

[fringe pawn]

Pick one that still has a valid respawn area?

But I agree with Greeny

[calm gale]

they need a way to do something with the radioactive waste in game

[cloud shard]

they did. and it's only available late game

[wind spade]

there is, turn it into plutonium

[frank mesa]

Wouldn't the doggos respawn at any vacant spawnpoint?

[calm gale]

i mean a way to either make it inert aka safe or delete it

[wind spade]

the whole point of the waste is to be non-deletable lol

[terse otter]

Yes, plutonium rods can be put into a sink

[cloud shard]

this is such a troll recipe

[wind spade]

this is such a great recipe to automate stuff you can't otherwise automate

[terse otter]

If miners were more useful, it would be good

[cloud shard]

just because you can, doesn't mean you should. lmao

[calm gale]

id kill for that recipe lol

[wind spade]

in a game about automation, you should automate everything you can 🤷

[frank mesa]

You would only need 1 manufacturer producing miners

[calm gale]

i mostly just wanna trry out drones or such

[fierce ruin]

except screws; because fuck screws...

[calm gale]

casted screws ftw

[cloud shard]

I mean, if I'm hurting on caterium (fat chance of that happening), I could maybe plop one of those down every second and spend 100% of my time filling up storage containers of the stuff

[terse otter]

Screw% runs, I'm telling you

Maximum screwage

[fringe pawn]

I believe the most nuclear power you can muster results in 252 uranium plants feeding 112 plutonium plants, creating 112 plutonium waste per minute. If you build a high tower out in the sea, there's no realistic scenario where the radiation becomes a problem. It would takes literally years of the game running 24/7.

[cloud shard]

steel screw > cast screw

[calm gale]

id need to find more HDs

and TBH even with the rifle and 240 ammo yea i dont trust my luck

[frank mesa]

If I go exploring, I take 1k ammo minimum

[terse otter]

Rifle is for the weak, just ninja dodge everything in melee

That's what I'm doing so far 😄

[cloud shard]

I automated ammo for the rifle

now I could spam all the bullets

[calm gale]

same

[terse otter]

I probably should do the same, but I'm too lazy

So far, I've only died to fall damage, so it's not a big deal to me.

[calm gale]

i have a full crate of ammo but since its 1-1 on ammo idk if its worth

[fringe pawn]

The real killer in all open world builders

[frank mesa]

I automated beacons, automating ammo was just 1 manufacturer away at that point

[cloud shard]

I thought about not automating the boolets, but incorporating call of duty into satisfactory is just too tempting to pass up

[terse otter]

Just make it gears of war and add a cover system, should fit perfectly

[cloud shard]

cover system is terrible against those charging pigs

I just choose a direction and spam everything moving with hot lead for a solid 10 mins

the leaves everywhere are annoying. I can't shoot them if I can't see them

[frank mesa]

Just toss a couple dozen nobelisk pipebombs, and kaboom they're all gone

[cloud shard]

I should get around to automating nobelisk sometime

but I don't use them too often. I'm having way too much fun using the chainsaw

[frank mesa]

When you have automated ammo, automating nobelisk is quite trivial

[cloud shard]

that is true

[frank mesa]

ACME Munitions and Explosives Inc. ftw

[cloud shard]

yeah. I'll add automating nobelisk on my to-do list

[frank mesa]

Automate everything, stay effective! 😋

[cloud shard]

I still refuse to automate portable miners tho

[frank mesa]

Not a lot of time involved in setting up an equipment workshop and quickly manually crafting a couple portaminers, true

But just for the because you can, I would automate producing them

[fringe pawn]

Until they stack, the recipe doesn't make sense.

[cloud shard]

^ this

also, the motor requirement too

[fringe pawn]

If you want to make a factory for them, you're going to spend more time building the factory, then managing your inventory to make room for them, versus just bringing a stack of plates and rods with you to manually build them. And you should be bringing a stack of plates and rods with you everywhere, anyway.

[cloud shard]

idle power draw could be mitigated with a power switch, but it's a whole lot of backbuilding for an item that takes too much inventory space, when an alternative of building it when needed, on-site, is more efficient

that is, unless, there's an item that needs portable miners as an ingredient

now that would really require portable miners

unless that item can be easily made by hand

in which case, the recipe is still useless

[cloud shard]

but yea, I should automate portable miners, at least for the lols

[thorn pollen]

@oblique hollow that doesnt always work for me

[versed violet]

They work nice for padding storage containers in personal storage, for when you don't want 24 slots of turbomotors sitting iddle when you only need 2 for personal use. But I don't have dedicated factory line, just using my configurable pick-a-brick manufacturer to build a bunch when needed.

[cloud shard]

lol

that reminds me

when are we gonna get limit storage slots mode

[wind spade]

just put a bunch of other stuff in slots you don't want to use 😄

[fierce ruin]

And remember to not use the sort button.

[wind spade]

why not, it will just sort stuff 🤷

[fierce ruin]

If you are using single stackable items to fill slots in a container then sorting will leave a bunch of empty slots, I didn't think of filling the whole slots with something cheap like limestone.'

[wind spade]

portable miner

[fierce ruin]

Yes, that was how the thread started.

I was thinking they were expensive if automated, but I guess since they don't stack they'd only use a stack or so of motors per container, so not really a big deal.

[tranquil oasis]

editable speed 🙂

[tardy moth]

Didn't you know?

[tardy moth]

LMAO

[tranquil oasis]

why is it a possiblity

[tardy moth]

I didn't know that was possible

Thought it was only numbers

[tranquil oasis]

Didn't you know?

[tardy moth]

:(

lol

[tranquil oasis]

lol i stole ur joke

[sand garnet]

it doesnt actually register anything else than numbers within a set range

so while you can type etc, it doesnt do anythi

[glacial hemlock]

re-open the building overclock UI after you type something to verify if it is actually saved into the machine

[sand garnet]

spoiler alert: it's not.

[bleak coral]

you can type more than numbers now, because you can type equations in there now, and I guess they didn't bother restricting what non-numbers you can type

be warned though: just like in the quick search menu it evaluates right to left, so if they're more than one step they may not be whet you expect without parenthesis

[sand garnet]

pretty sure you've always been able to do that though, at least the typing of letters etc

[magic shadow]

so i can tell it to overclock to banana but it wont

[fierce ruin]

typing what we want was here, but the fact that the game takes into account math formula came with update 4: "• Overclocking now supports math formulas when entering a number (similar to the Quick Search)"

[sand garnet]

ahh

thanks for letting me know 🙂

[naive quarry]

Brb making a penis% motor factory

[fringe pawn]

Make sure it uses steel rods and screws.

[versed violet]

Try 1e3 or other weird numbers.

[fringe pawn]

What are the uses people are getting from the ability to enter a formula in the overclock settings?

[versed violet]

100/3

[sand garnet]

im guessing it's just 'amount divided by machine count' usually

[versed violet]

Odd values like 33,(3)

[fringe pawn]

Nice

[oblique hollow]

or "i have this much and now i need one fifth more"
so i just enter " [X] / 5" and get my number

entering fractions is simply easier than a lot of decimals

[fringe pawn]

What's messing with my head as I play with it is that 2/3 gives .667 instead of 66.667

[versed violet]

Its in percent. Multiply by 100.

[fringe pawn]

Right, it's just different from my usual weapon of choice (Excel)

[versed violet]

Not once and not thrice I have got 0.33(3)% instead of 33.3(3)% I wanted.

[thick reef]

++++++++++++++++++++++++++++

+++++++++++++++++++++

whops guess i had something resting on the + key. Sorry about that

[silver steppe]

just spreading some positivity right?

[true minnow]

equations?? Gauss & Newton liked this.

[frosty pawn]

you can easily underclock to get round numbers of output when you dont care about the input or make the output based on limited input. for example say you have pretty much unlimited steel ingot supply and the steel rod alt but you only need 45 steel rods per min. the recipe gives 48/min and you want 45/min so 100*45/48 automatically figures out the clock speed should be 93.75

similarly, say you have 10 leftover steel ingots from a manifold and you want to use them to make some extra rods: 100*10/12 sets the clock to 83.3333 (recurring)

or you have 15 ingots and you dont know how much to overclock: 100*15/12 = 125

sorry i arrived late to the conversation and just realised you guys already explained this

[glacial hemlock]

What happen if the formula is entered, closed the UI, and open the UI again? Will the clockspeed show 93.75 (or whatever decimal) instead of nearest integer?

[bleak coral]

yes, we have 4 decimals to play with now

might actually use more than 4 decimals too and only round up the UI when you use an equation too or enter an amount in ppm instead, though that's hard to prove since 4 decimals is pretty precise already

[open zenith]

Does this look like an efficient storage unit?

[wind spade]

it looks like way more storage than you'll ever need and is overkill

[open zenith]

Well

It looks cool

[wind spade]

you asked about efficiency 🤷

[open zenith]

Yeah you are right

Is 96 slots enough?

Just 2 Industrials

[wind spade]

one ISC is already more than you need

with maybe exception of concrete

[open zenith]

Concrete is a pain in the butt

Ima go 96 just in case

[wind spade]

when have you ever needed 96 stacks at once?

of anything?

[open zenith]

At once is a key word

Yeah I guess

Never

[wind spade]

bigger storage is only for when you need a lot of materials at the same time. If you need a lot of materials over time, you need bigger production

yes, storage can provide a buffer, but if you empty that buffer, it's still about your production speed

so rather than overkilling a storage, go overkill your production 😛

[open zenith]

Can I do something to this?

Its really bugging me out

[sand garnet]

i dont see an issue

[open zenith]

They dont really snap

[sand garnet]

foundations look fine to mie

[open zenith]

Nah the walls

[sand garnet]

oh the little corner missing

nope, cant do anything about that

[glacial hemlock]

You mean the size of building not exactly being the multiples of 8 meters? Nope, you can't do that in vanilla.

[glacial hemlock]

Look for ficsit mod micromanagement

[brave jetty]

so is it just me or is it more efficient to package fluids and put it on a train than to put fluid directly onto the train?

[fierce ruin]

fluids have various packaged/fluid ratio, so it may won't be true for all of them (but that's also my impression).

[fierce ruin]

When you refuse to have plastic and rubber issues again

[upbeat tide]

only need the bauxite from the swamp and max aluminum ingots is done

[frosty pawn]

packaging and unpackaging fluids costs power

[brave jetty]

Nothing costs power if you sacrifice space to under clock🧠

[glacial hemlock]

pump also cost no power if you know how to skip pump, hehe

[fringe pawn]

I have 3 pure bauxite nodes, each immediately feeding an ISC, the splitting into 2 belts , for a total of 6 streams. Then I have a normal bauxite node, which I will split 6 ways, and add it into the other streams. Then I will have 6 sloppy alumina refineries clocked at (100*14.5/6) handle the bauxite. I will directly feed single refineries to process up to 2900 bauxite.

Each solution refinery will feed directly to two electrode scrap refineries. But here's where my head is exploding. That would normally take 19.33~ electrode refineries, so I'm already starting with a third, and trying to reduce that number of buildings to 12. What formula to do I need to enter for the game's logic to overclock those buildings appropriately?

[upbeat tide]

@fringe pawn do what I did. Create a single aluminum ingot loop per node of bauxote

3 sloppy, 4 electrode, for a normal

3.9 sloppy and 5 (1 at 120%) scrap for pure

[fringe pawn]

I have existing factories I'm using

[upbeat tide]

Ah

[fringe pawn]

Though all the bauxite is streamed to one location, so I could still use the principles

Currently it's processing 2 pure nodes with 8 slop/16 electrode refineries. I'm hoping to add nodes by overclocking and reducing machines.

[upbeat tide]

16 is too much. You can cut that down

Its 5.2 scrap refineries for a 780 bauxite supply

[fringe pawn]

I can easily reuse my existing infrasutrure if I keep the 1 -> 2 setup for processing. Not having to replumb and rebelt anything would be a big bonus 😛

[upbeat tide]

This is mine. 7200 ingots setup so far

[fringe pawn]

Pure ingots?

I'm considering dabbling in the normal recipe. For a fraction of the production.

[subtle elk]

if i have every alt recipe should i just turn all of the oil on the map to residue because i cant think of any recipe that needs normal oil that wouldnt be better with residue

[upbeat tide]

Yup, once you have the HOR alt, recycled plastic/rubber, etc than your use for basic oil recipes ends

[fringe pawn]

How advanced can overclock formulas be? Parentheses allowed?

All I know is that I'm not entering the right formula to reduce 19.333 buildings to 12

I think I may have finally done it with excessive parentheses: 100*(19+(1/3))/12

Then I just need valves from water extractors to feed 145 additional water to each bauxite processor and I think I'm good!

Yeah, perfect symmetry. 6 solution lines running out of the bauxite processors, then 6 water lines going in, 2 from the ocean, 4 from scrap wastewater.

[frosty owl]

Power plant with balanced rods input for you, @wind spade
The rods are 1:3 from manifacturers to generators, they're practically asking to be balanced
#screenshots message

[fringe pawn]

Building my first recycling loop later. Any words of warning or wisdom? https://gyazo.com/0d6ccc6dc2d4ec10307769f0ececc9b9

[oblique hollow]

if you want more output, recycle the residual rubber too

buuut that would need more fuel...

do you want to use those 180 extra fuel for power? or is it free to use

[fringe pawn]

It's for fused frames

[oblique hollow]

ah ok

then this is the best that can be done

[fringe pawn]

I'm going to occasionally siphon some production for jetpack fuel and filters. But there's so much room in the rubber and plastic production that it's fine, both are 500+ over my actual needs.

[oblique hollow]

thats a lot of coke.... aluminum?

[fringe pawn]

Yep. Technically it's 10 more than I need, but I like a container of everything for possibles.

[oblique hollow]

coke is literally useless outside of automation xd

[fringe pawn]

Something something future content, hopes, dreams, maybes 🤓

Worst case scenario, there's wiggle room to just redirect it with clock changes. Though if something changes, the odds are it's going to be something that can't be solved by shifting that sliver of coke production, and I'll be tapping new nodes.

The larger impetus is I needed more plastic for batteries, for drones.

I think I'm going to underclock the blenders such that I'll have 64, and they'll each feed 2 refineries directly.

[upbeat tide]

I have dedicated 900m3 of oil to petro coke production and gonna use 80% - 90% of it in my aluminum plant 🙂

[glass dock]

What is the ratio of water extractors to pipelines to coal generators?
If I understand it correctly:
Water Extractor produces 2 m^3/s
Pipes Mk1 can carry 5 m^3/s
Coal Generators use 0.75 m^3/s

Which leads me to the ratios of
3 Ex to 1.2(?) Pipes to 8 CoalGen

My Question now would be how to solve the issue of the 0.2 pipes/how to join the extra pipeline that carries 1m^3/s back into the water grid at the coal plant?

[sand garnet]

!wikisearch coal generator

[shadow prairie]

https://satisfactory.fandom.com/wiki/Special:Search?query=coal generator

[sand garnet]

click that @glass dock

[upbeat tide]

So dont count production in terms of seconds.

Will drive you nuts

[bleak coral]

Also no one will be able to communicate with you

cause it's all per minute in this game

[glass dock]

Sorry force of habit.
But I landed at basically the same numbers as wiki.
The one question that remains for me is do I have to be careful of anything when joining the second pipe carrying 60 m^3/min back into the water network at the coal plant(assuming that I use 8 Coal gens)?

[upbeat tide]

Just connect it after the 8 coal gens have been connected

Manifold loop it

[glass dock]

nice.

[upbeat tide]

And its easier just to have two water extractors on one end of the pipe and the third feeding from the back

[bleak coral]

each section can only hold 300m^3/min, so if some gets used up by machines you have more room
e.g. if you put the 1 extractor on the other end, use two pipes, or inject it somewhere in the middle
the wiki tom linked has pictures

[upbeat tide]

https://steamcommunity.com/sharedfiles/filedetails/?id=2124946046

[bleak coral]

also don't forget pipes are bi-directional and junctions are joiners for containers, not splitters or mergers

[upbeat tide]

You can find charts and diagrams here in my Steam guide

[glass dock]

meh I was just gonna upscale to 40 coal gens 15 extractors and and 6 pipes. Then the ratios match up nicely again And I have the coal production to support this venture

[bleak coral]

if you're going that big, I'd advise actually doing injections and build on top of the water so you don't have a bunch of pipes

[upbeat tide]

Just so you know, coal gens do not match the 300 pipe limit well

Such as 300 water can supply 6.666666 coal fens. That is why the 3:8 ratio is used alot.

Unless you have mk2 pipes

[glass dock]

Might as well. I was bootstrapping on 6 coal gens but now want to move away from bootstrapping and built a factory producing 1 mk2 band of excess T1 products (screws rods etc)and 1 mk1 belt of excess t2 products(reinf plates beams rotors etc)

That thing is way too powerhungry for my 6 coal gens.

[upbeat tide]

For big projects its best to break it down into blocks. Based on the more limiting resource, like water in this case

[glass dock]

Or I can spend hours engineering a blackbox that takes in a bus of ore belts and then spits out products at near perfectly balanced ratios.

[upbeat tide]

Factorio style busses are not that useful here

[glass dock]

I know

was still fun to build ¯_(ツ)_/¯

[upbeat tide]

Overflow manifold is design king over direct balancing.

[oblique hollow]

recently decided to try factorio myself.
immediately had to install an item rate mod that shows the rate machines produce / use things because heck no i wont guess or "just flood it all"
im really glad satis has this inlcuded in everything already

also items / second would get so small in both games it would be really cumbersome

[glass dock]

meh, in factorio it is easy to calculate. Recipe duration (t) is given and you have a crafting speed multiplier(x).
So items per second is just 1/(t*x)

[upbeat tide]

Or really big. Like if you translated my aluminum ingot factory to /s, it makes 120/s aluminum ingots. Aka 7200/min

[oblique hollow]

yeah, no thanks. the streamlining of that makes it easy.
Factory planning fun and all but some things really dont need to be

[bleak coral]

yeah it's annoying to calculate /s time in factorio every time, instead of just having the numbers myself

[oblique hollow]

it doesnt really add any fun or depth to it all either

and it gets funkier with modules

doing percentage calculations myself? i got better things to calculate

[glass dock]

Modules just change x :^

[oblique hollow]

yes, still no thanks

[upbeat tide]

The only thing I wish SF had ingame that Factorio does is production stat charts

[glass dock]

I can see that. I do math like this all the time every day so im used to it. But to be fair you calculate the ratios once and then expand by stamping down more blueprints of your balanced design

[oblique hollow]

since we dont have that we have to redo math every time

[glass dock]

Which is imo annoying

[oblique hollow]

or just.... have to write down our plan

devs are debating whether or not to include blueprint-ish tools

its a really heated topic in CSS

[upbeat tide]

I am actually on the fence, leaning no blueprints

[glass dock]

Also one of my biggest pet peeves with satisfactory. The inability to place large areas of foundations at once. And the no copy pasting recipes

[oblique hollow]

satis is more a factory builder, while Factorio, for all intents and purposes, is a factory planner

you do funny ghosts and machines build stuff

buuuut having the ability to copy small things in a box area WOULD be very useful

[upbeat tide]

I am on the fence for BP’s because of the vertical layer. Or at least if the BP just placed machines that would work for me

[wind spade]

late to the party I guess, but I usually recommend this setup

  G  G  G  G
  |  |  |  |
E-+--+--+--+
E-+
E-+--+--+--+
  |  |  |  |
  G  G  G  G

[glass dock]

Oh I absolutely agree but for being about building satisfying factories Satisfactory makes the building itself too much of a annoying chore

[bleak coral]

I definitely think we'll at least get mass foundation build of some sort, maybe not as extreme as Smart! but something

cause if they're considering doing something for mass build, that's the most obvious place to start

[oblique hollow]

i think the idea was to go the minecraft route: yes, building an entire village is annoying by hand, but feels great once you are done

buuut seeing as things are a bit.... lets say demanding, we do need some fast building assistance

its an odd mix of Factorio meets MC

[glass dock]

Also why do lights look so utterly atrocious?

[oblique hollow]

the range? or everything

main reason is the performance

[glass dock]

https://i.imgur.com/L66NIA0.jpeg

[oblique hollow]

overlapping lights are very performance-intensive

[bleak coral]

they're not real lights for performance reasons

[oblique hollow]

also, you can adjust their luminosity with the control panel

so its not so hideously bright. i recommend like 25 % Luminosity indoors

[bleak coral]

yeah default luminosity is a bit... much

[oblique hollow]

aaaand change the color to a warm yellow-ish white

[glass dock]

I would just prefer them to light up the room instead of an area slightly more than 1 foundation

[oblique hollow]

#announcements message
watch that

Ben explains why lights be like that in that video

[glass dock]

moving this out of the math channel

[fringe pawn]

What's the preferred layout of a recycling loop for rubber and plastic? It seems like you could get away with a shared belt if you had an even number of plastic and rubber refineries within the same module.

I'm going to have 64 fuel refineries feeding 128 recycling refineries, so I have easy numbers to work with.

[supple belfry]

Last I looked at optimizing that, it involved lots of packaging and unpackaging, but it has been a while

[fringe pawn]

The need for packaging has been removed. You can make diluted fuel by feeding water extractors and heavy oil directly into a blender.

[eager solar]

don't know if that's still the case in u4 but optimal setup in u3 allowed you to get 900 rubber or plastic/min for a full mk1 pipe, so 1 shared belt wouldn't be enough

[fringe pawn]

Pipe bandwidth isn't a bottleneck anywhere, I can break it down to more pipes as needed.

[eager solar]

I'm talking about the belt, current max can do 780 items/min so if you go optimal one belt won't be enough

unless I didn't understand your question

[fringe pawn]

Right, so modules of eight refineries, perhaps? Four plastic and four rubber. No bandwidth bottlenecks anywhere theoretically.

[oblique hollow]

i do one belt rubber, one belt plastic. feed them to the inputs and slap an overflow splitter on it

[versed violet]

My personal setup is:
rubber belt from residual rubber refineries
feeding a row of plastic refineries
who feed 2 rows of rubber refineries (because who needs plastic anyway?)
I also siphon couple rubber pieces from the poutput back to plastic, in case residual rubber refineries stopped.
Chain another module to rubber utput of first one, repeat as fuel lasts.

[fringe pawn]

Hm, hadn't thought to work my residual refineries into the mix. Nice.

[versed violet]

I'm worried to ask what did you do with residue

[fringe pawn]

I was just going to immediately put that rubber into the master output channel. Occasionally siphoning for fabric.

[versed violet]

I just use it for bootstrap/plastic feed and redirect overflow into output. ¯_(ツ)_/¯

[fringe pawn]

Here are the actual numbers I'll be working with, though I'll clock the blenders to 60% so I get good symmetry. Coke production is 10 higher than it needs to be, and rubber and plastic are at least 500 higher than they need to be. Pretty much the only reason I'm making plastic is batteries. Though I'm also going to do steel coated plate for kicks. https://gyazo.com/d1586f7686e4da541e577615c5ef9247

[versed violet]

Blenders at 100% are perfectly symmetrical => 12 of them uses up an mk2 pipe of residue and outputs 2xmk2 pipe of fuel

[fringe pawn]

64 blenders -> 128 recycling refineries

[dawn pebble]

I was wondering what is the best way to get a line of 45 items per minute

[versed violet]

mk1 belt -> splitter -> splitter -> merge output from 1 and 2 splitter (45), remainder (15) is sent elsewhere

[dawn pebble]

okay thatnks

[fringe pawn]

Thanks to the nice numbers, I think I'll build half now, then use shards to double production later. I'm not building space elevator parts yet, so I'm not in a rush. 8 modules of 4 blenders, 4 plastic recyclers, and 4 rubber recyclers should work.

I really enjoy that when you paste building clocks, it automatically grabs shards

[upbeat tide]

Finally got the final bauxite nodes collected into a train starion, took a while for the asthetics

[fierce ruin]

How can i seperate 5 incoming into 3 and 2?

[oblique hollow]

5 incoming what. Belts? Or items / min

[fierce ruin]

items

someone in #old-questions-and-help said over/underclocking

which i havent done yet so ill have to figure it out

[oblique hollow]

That or building a ridiculous load balancer or filling both machines up by hand and then letting overflow take care of splitting

[fierce ruin]

i guess ill try the overclocking

sounds like it could be interesting tbf

[oblique hollow]

Underclocking is usually recommended vs overclocking

Because of non-linear power consumption

[fierce ruin]

yea thats what i meant over/underclocking

why the hell is my power still fluctuating

[oblique hollow]

I blame pipes

Probably water problems

[signal nimbus]

Hold on... 5 incoming? As in 5 per minute?

[oblique hollow]

Sounds like Reinforced iron plates or something xd

[signal nimbus]

Yeah, but like... that's just a split in half, split one of the lines in three, loop one line back before the first splitter.

If he's trying to split five lines into 2 and 3 lines, depending on /minute, that would be different.

[manic locust]

Hey all, I wrote a program to minmax points/min, determining the best alt recipes/power setup/etc. My reddit post didn't get much exposure but I thought people might still be interested: https://old.reddit.com/r/SatisfactoryGame/comments/nz6aqa/satisfactorylp_optimal_recipe_solver_for_update_4/

Here's the spreadsheet, it's got some pretty detailed info: https://docs.google.com/spreadsheets/d/1q4dvdzhLXdbDdV1lxKM27TFOuI2TariiiItFIuygWEw/edit#gid=1025182137

[fringe pawn]

I love the jump in machine count at 100%

[fringe pawn]

I did a ctrl+f for geyser and see that you may not be using geothermal power. Is that relevant at MP0 or other places at all?

[manic locust]

It should be there

[fringe pawn]

Huh, does ctrl+f not work in Sheets?

Either way, cool stuff

[bleak coral]

I mean the geysers aren't making that much of a difference lol, all less 1% combined

thought I guess it matters for actual most points

[fringe pawn]

Yeah, looking closer at the numbers I'm not that surprised, but if it were ever going to be possible, an all 1% clocked factory is one way.

[fringe pawn]

Okay, here's a module of 4 recycled plastic and 4 recycled rubber factories to be clocked at 200%. The right 4 refineries make rubber and send it to the upper loop, where the rubber circles until it either goes into the left 4 refineries, or into the overflow smart splitter in the bottom corner. The left 4 refineries put plastic into the lower loop, with the same sort of setup. Am I good to go once I feed them?

[glacial hemlock]

Bolted frame and coke steel... nice

[wind spade]

hi, you may be interested in my tool repo, particularly parser of Docs.json into much more readable and useful data.json 🙂
have a look:
https://github.com/SatisfactoryTools/SatisfactoryDataParser (note - API doesn't work yet)
https://github.com/greeny/SatisfactoryTools (bin directory)

[manic locust]

That would've been very helpful, I'll look into it

[wind spade]

also I'm using pretty much the same approach as you for calculating stuff, and with future update of the tools I plan to add sink points as well, so I'm interested to see if the solution will be similar/same/different 😉

[wind spade]

and finally - can I ask you how you did overclocking? afaik you can't have exponential functions in LP model

[manic locust]

Multiple columns with different clocks, for manufacturers I used 1/25/50/75/100%, and for miners I used 100%/conveyor limit (162.5/250%)

[wind spade]

hm

[manic locust]

it's not perfect, but close enough

[wind spade]

I see... I wouldn't let it underclock tho, I'd go for "maximum with machines @ 100%"

which is imo more realistic

(and you can throw out the machine penalty that way)

also, for the theoretical maximum, run the numbers again, but with 0.0001% underclock, 1% is still wasting a lot of power 😉

[manic locust]

I believe the game won't let you clock under 1%

it's mMinPotential in the json

[bleak coral]

yeah it won't, changes anything under 1% to 1%

[wind spade]

hmm, I thought I've seen somebody posting 0.0001% underclocks

w/e

[bleak coral]

you can enter it, but when you exit the UI and reenter, it's been changed

[wind spade]

well that I know

[bleak coral]

I said it for a while, cause I forgot to check if it was changed after I exited the UI

[manic locust]

as for the machine penalty, I considered it a more general tuning approach, avoiding 1% clocks being one specific example

because the real motivation behind avoiding 1% clocks is the crazy machine count

[wind spade]

Finally, I was a bit surprised to see that it didn't use Fuel Generators at all. I had thought it might avoid nuclear power due to the processing cost (especially as it was forced to process all waste), but still the algorithm always used nuclear power.
I've always been saying that people should use nuclear more 😄

[bleak coral]

super-low clocks are funny (they can go below idle power even), but ultimately impractical for wide-spread use

did the algorithm decide to sink plutonium, or was that a you decision cause you didn't want to deal with waste?

[manic locust]

I did not allow it to accumulate waste

if it couldve, I'm sure would've done it

(use the plutonium fuel rods for power that is)

[bleak coral]

I'm curious what the difference is if you let it accumulate waste, cause fertile uranium can save some key resources

[manic locust]

I can add an option for that

[bleak coral]

also I'm not surprised it didn't use fuel at all, the amount of oil, sulfur, and maybe coal needed to keep up with nuclear is massive

[wind spade]

is massive in terms of raw numbers, but is minimal in terms of items per MW

1 unit of turbofuel has 2 GJ, while one uranium fuel rod has 750 GJ

I'm guessing sulfur saving may be key, with addition of using way less oil

[bleak coral]

I'm not quite getting your point? are you agreeing with me you're not surprised it couldn't keep up with the power demands for max points or what?

[wind spade]

ah nvm I can't read

[bleak coral]

oh ok lol

did you take it the opposite of what I meant?

[wind spade]

I read it as "to keep up nuclear"

[bleak coral]

ah

that with is important 😛

I'm also thinking when you're maxing out stuff, that oil products are super important in stretching out the number of parts you can make by using oil alts in addition to the other recipes.

[wind spade]

750 GJ with nuclear

750 GJ with turbofuel

(ignoring power cost)

though looking at power usage tabs, it seems that power cost for nuclear is ~550 MW + miners, while for turbofuel it's ~900 MW + miners, so it's even worse

[bleak coral]

that surprises me a little, but I guess for turbo that's a lot more machines in total

plus I bet more refineries

[wind spade]

turbo

nuclear

[bleak coral]

interesting

[wind spade]

anyway for this also - if you use my datasource, fluids are normalised (so you no longer have to deal with ingame units, but you get the normal values, divided by 1000)

[manic locust]

https://tenor.com/view/thumbs-up-great-agree-awesome-cool-gif-15385195

[manic locust]

I added an MP2000;Waste sheet for a run that allowed nuclear waste accumulation

as you expected it used Fertile Uranium and let a bunch of uranium waste accumulate

and then it produced just enough plutonium for its energy needs

[bleak coral]

wait so it's accumulating plutonium and uranium waste? like it's making more uranium waste than it needs to make plutonium rods?

is it cause making uranium is always more resource efficient than making plutonium at all?

[manic locust]

I think it's saying that it would rather make uranium power than plutonium power

like it will use all the uranium ore to make uranium fuel rods, but it won't use all the uranium waste to make plutonium fuel rods - it will only use a fraction of uranium waste that it thinks is sufficient for power needs

[wind spade]

which makes sense since there's no cost for "wasting" waste

[bleak coral]

interesting, though I can't imagine it's very practical considering how much faster uranium waste is made

[wind spade]

waste is not a problem 😛

[bleak coral]

plutonium waste isn't, uranium waste could be

[wind spade]

would still take years to become one

[bleak coral]

cause it's made 10x faster

[wind spade]

depends what problem are we talking about. Radiation? nah. UObject limit? well yeah

[bleak coral]

oh well that's already an issue we're ignoring for max map anything 😛

[wind spade]

sure, but then again - radiation is not an issue 😉

[bleak coral]

I'm more worried about storage getting full, or needing too much, though maybe that's arguing semantics when you're talking about max points haha

[wind spade]

since you maxed everything, you have nothing else to do apart from building storages

[bleak coral]

guess it's something to while the points (and nuts) pile up

[manic locust]

you could be multiplying your buildings by 100 and setting 1% clocks

:)

[wind spade]

where do you get material if everything is sinked

[manic locust]