#math-and-meta
1 messages Β· Page 112 of 1
Only time I've ever used exact splitting was in my coal power plant, when I needed all the machines to start running immediately and not have gaspy power for a bit
Sure, I was just thinking that everything that canβt go into the first machine (due to the low speed belt) will go to the next one, and so on. Thus, more machines would fill up in parallel instead of focusing on the first one π€
It seems like a rather small difference anyway π
Is it. Which why there is nothing wrong with using lower tiered belts to the side.
(I do it often)
Belt balancing is when multiple belts, possibly at different rates, are split and merged to give multiple equal outputs, maybe not useful. Your situation is splitting a common resource output to match required inputs and is usually referred to as load balancing and is very useful.
If they are only using a manifold and relying on innate belt throughput behavior to balance it, then it is not load balancing.
Using a bunch of splitters and mergers to send each machine exactly what it requires would be input load balancing.
You donβt need all coal power to start at once and even if you did you could have prefilled the machines
The only time that it is 'needed' is maybe for radioactive materials, and even then, you can manifold it, its just gonna be a very spicy area to walk around in
filling more machines at start means that more machines are processing the input, which means that less input is being "backed up"
see f.e.
red = mk5 side belts
blue = mk1 side belts
bottom is minutes, left is production
when splitting onto a lower-speed belt, it's preferable to have a steady input. ie, if you split off an Mk1 belt from something higher, you want the source items to be coming through steadily enough that the splitter can constantly allocate 60/min to it
but manifold input belts are choppy, with items being pulled off at asynchronous intervals, creating unpredictable gaps the further along you go. as the source belts alternate between full/empty, the side belts alternate between getting something/nothing. so even if there's enough passing through on average to meet capacity, they may not actually receive that capacity until the system fully stabilizes
whereas splitting onto same-speed belts in manifolds means that any given branch has the capacity to take the entire source input if other branches are backed up
once filled, yes. But mk5 side belts will get there faster
that's not the reason mk1s are slower though
Interesting, thanks for sharing ^^ now I wonder how exactly you calculated these values π guess I will look into this in more detail tomorrow π
this should also depend on the relative speed difference between the source and side belts, right? If I have a mk2 source belt, then there should be no point in using mk2 side belts. The splitter is giving me not more than 60 items/min anyway
this is mostly just simulation running on code I wrote and then outputting values every 10 seconds or so
Yeah, I will probably try different manifold setups and different belt speeds. Mk5 belts are nice, but you need to unlock them first. I build a lot of manifolds with mk1 side belts long before I get to mk5 π
it's valid for pretty much all manifolds with slower side belts
at best it's equal to faster belts
any other case it's slower
the only case when slower side belts make the manifold fill faster is if the side belt is exactly the speed that machine needs (e.g. machines that need 60/min with mk1 belts), because then there's no filling time and all machines work immediately
True, I guess. Itβs just fun to solve these things π it would also be interesting to find the absolute time necessary to fill different manifolds with different length and belt speed
like this? https://satisfactory.greeny.dev/machine-fill
A powerful Satisfactory calculator, filled with features, supports overclocking, alternate recipes, live map, power calculations, etc.
(press the red button, ignore rest of the site)
I know, I know π the fun part is to solve it yourself π
the problem is that there's really no formula for it, and you have to simulate it π€·ββοΈ
Thatβs okay π
well there were some attempts at making the formula, but it ended up being some horrible thing that would take longer to solve than for the manifold to fill
ππ
kinda wish the radiation tool here was up to date, love all the work you do!
U2 values:
ore: 15
uCell: 7.5
NFR: 60
NWaste: 20
U8 values:
ore: 15
uCell: 0.5
UFR: 50
UWaste: 10
PFR: 250
PWaste: 200
you can use that to convert values, so f.e. if you're looking to know how much radiation does 1000 UWaste produce, you multiply it by it's U8 value (10, which means total is 10000), and then divide by some of the U2 values of any item (e.g. NWaste is 20, which results in 500, meaning that 1000 UWaste in U8 is equal to 500 NWaste in U2)
In that context, no.
But the question was about which stabilizes faster.
If you prefeed they are both equal.
So it truly doesn't matter from that methodology.
I'm beginning to realize that no form of transporting goods is as fast or efficient as just running a long conveyor bus, is this the place where people smarter than me about the math and numbers involved would have an opinion on that?
I just tried trains for moving 6 pure iron nodes and the factory they are feeding can't run the Mk5 belts from the train depot at 780 ppm. I ended up pulling up the whole track and just used the floor tiles they were on for conveyors instead.
Even putting in storage buffers to account for the pause when transferring cargo that small delay in loading and unloading seems to mean that the resource will always come out slower than using just belts over trains.
Are drones any better?
Nothing can go faster than belts.
But other options can be better for your FPS and more fun. Plus, setting up a drone or truck is easier than running belts if the endpoints are far apart.
For all the trouble I ran into sorting out the train path and block signals I needed to allow two bi-directional trains to share one track between two depots I feel like running the conveyor bus was simpler to do and less work overall. The drones I can see being easier than connecting two points but if they aren't at least as efficient as the belts they load and unload to I would probably still prefer the conveyor bus for how I design my factories.
This is why a lot of folks only build double tracks.
I do think the trains are neat, I originally built a train line making my way to the iron nodes and that was great I had all the building resources I needed in the freight car and I just moved the train up the track until I got where I was going which spared me a lot of back and forth. But now I'm thinking that's the only valuable use case the trains have for me
I don't care for them. But for some people, the "value" is in having the trains, not how well they perform.
what heavy modular frame recipe is the best?
there is no best, see #math-and-meta message
I personally use encased frames
There isnt one. in my opinion
Encased if you want to use a ton of concrete.
Flex if you want to offset concrete with rubber (and know how to use steel screws "properly")
Default.. well i dont see why anyone would use default. Since it uses both more steel and concrete than encased apparently.
Fyi "conveyor bus" is something different than you think
A bus is if you have belts going through your factory, split from them to nearby machines and merge back products from those machines.
If you jave just a few belts transporting resources from A to B, you have stacked belts, not a bus
And for your question, one train car should in most cases be able to handle even more than full mk5 belt worth of throughput
Greeny: I suspect they were trying to get a pair of saturated mk5 belts thru one freight segment. Which we know is impossible.
@north ocean
There are known maximum thruputs for stations because of the loading/unloading lockout periods. Trains work fine if you are moving less than that per station. There are equations and graphs that show the thruput plotted over RTD (round trip duration) with a different curve and max transfer rate depending on the stack size of the item being transported.
Generally tho if you feed 1 belt of 780/min into 1 dock, then you will get 780/min out at the destination. Just put that 1 belt into an ISC then 2 belts into the station, and then 2 out from the destination into an ISC then one from there.
TLDR: You can always use more freight docks and freight cars to get your intended thruput.
wdym "one freight segment"?
The buildable freight loading unloading segment of a train station.
one platform?
Yes
one platform can do more than one mk5 belt worth of throughput
You cannot move 1560/min.
more than one mk5 belt worth
Like i said. Pair of saturated mk5s
or in other words, more than 780
which I said here
I think they were trying to move 2 saturated mk5s per platform.
Just for me to understand, this is not working because of train driving speed and time for loading/unloading?
because of the fact that the train station locks for 27 seconds when the train is docking
so for those 27 seconds, nothing can get in or out the platform
Ok got it
Thanks
This was the meta reply I was hoping for, although I'll take the terminology correction on board as well I guess.
Is there a spreadsheet somewhere with those throughput curves for round trip times?
I did have 3 depots with two full mk5 belts going into industrial storage containers, but if I understand your response I can get back up to two 780 belts after the train unloads if I spread those across more depots?
there's a formula on wiki
The theoretical throughput of a Freight Platform can be calculated, by taking the amount of items carried between stations, and dividing it by the time it took the train to complete one round trip:
The Round Trip Time must be measured, as it depends not only on the distance, but also on the grade on the trip, weight of the train and the amount ...
but if you don't want to deal with that, easy option is usually just "one platform/car per belt"
(you still need the buffers - put industrial container in front of the platform, connect it with two belts to the platform and connect that one belt you want to transfer into the container - both for loading and unloading)
I personally don't see much reason to optimise it "close to max possible throughput", because then if your train gets stuck in traffic or something, it fails
Ok so this page was helpful thank you, it still kind of confirms my assessment that pulling conveyor stacks is better for what I'm designing around. Trains are neat but I'm not willing to suffer the throughput loss even if it mean putting up with the tedium of long conveyor pulls
what throughput loss?
trains do not lose you any throughput
it's just that one car/platform has smaller max throughput than two full belts
they just have a throughput limit per platform, thats all
but what you put in on one side should go to other side in the same amount, if you build it properly (add buffers as I said)
While I understand what greeny's being pedantic about I mean I would prefer to endure the tedium of pulling long conveyor stacks for the sake of their simplicity vs what it would take to get the same throughput utilizing trains to be most clear
if its pretty decent terrain you can just use a truck
I mean it's just "put each belt into one platform with a buffer" and then you're only building one track (or two, if you plan to have more trains later)
depending on the item type, you can have one truck being fed with multiple mk 5 belts from multiple stations and then drop off stuff at multiple stations
trucks dont have the same throughput limit as trains
done right, belt to truck is 100% the same throughput
Trucks are equal in terms of speed π€·ββοΈ
So thats why people like trains so much? Hm. Interesting.
People like trains more mainly for 2 reasons:
- trains are fun and people just like trains.
- people suck at driving, so rather than spend an hour learning to do it properly they just permanently swear off trucks and use only trains.
Fair. I like truck highways just for the look of it tbh
trains have several advantages:
- tracks carry power with them
- they use power as fuel directly (easier to deliver than any solid fuel)
- once built infrastructure can be reused for other trains
- one train can carry insane amount of resources (having enough cars)
- they can easily serve as a people movers as well
I will never say that trains are bad.
Nor can it be stated they are better/worse.
They are different.
There is an argument to be made that drones are better than either trucks or trains 
cause they go 'whiiiir' through the sky obviously. No other logistic does that. Trumps everything
I recorded a truck route without crashing but the truck keeps crashing on autopilot
it goes between the recorded points, drive slower to have them closer together and have a more accurate route
Also, using
to clear the path is very helpful.
As my username states, best is relative. you could use factory carts for transport as they need zero fuel. they are as inefficient as all get out, but it works.
As defined, best is objective. The word you're looking for is preferred
Best is objective if there is criteria stated. if you say Best with no context, then it is relative.
yeah best needs context
if you are looking for something objective. opinions need no context
No. Best with no context implies that it is superior in every way.
I had a specific case where i was able to drive up a slope while recording but the autopilot couldnt. The autopilot doesnt like to floor it apparently.
no, because there are factors that directly conflict with each other, therefore no recipe can satisfy all the requirements.
Best with no context, WOULD mean that. But that usually doesnt exist. (That and context is often implied irl so even then there is a context)
So I made a crazy turbofuel array and it's more efficient than I predicted... wondering whats going on
the array should only power 500 fuel gens
nothing overclocked
1200 sulfur perfect efficiency input
but i'm at 561 gens and counting
I just keep building more and they keep working
How much Turbo are you making in total?
2250/min
50 blenders at 100% efficiency
inputs plugged in at 105% to make sure no downtime
How long has the system been running more than 500?
Mods? Multiplayer? Dedicated Server?
dead stock
i use a lot of storages
10-12 large fluid tanks in the array
but that's just a backup buffer
I am unsure whether you answered my question or not.
Were the buffers full?
they fill up in about 5-10 mins and stay full

only thing I can think is that I accidentally overclocked a blender
but then my inputs would start failing
and I'd see an empty belt, which I do not
cant complain just oddddd
Will report the # where the fuel gens stop working lol
what is the best way to attach blender manifolds? I feel like my splitters and pipes arent efficient
meaning say 5 blenders on each side of a central aisle
manifolds are manifolds. Really it's just layout though you probaly want to keep your pipes simple.
You could have the pipe loops drop from above and the belts below or on level?
yeah youre right I should be using multi-level, pipes above. smart
I do pipes from below too but you have to be more diligent and careful with pipes like that
you can also save belt space by doing sushi belts to feed them
oh yeah true, pipes below works
if the recipe only needs 1 fluid 2 solid items you can feed it all with 1 belt and 1 pipe
that sort of thing should be done by planning each step together - so knowing what groups of machines are sending items to the next group and what clockings you need
oh definitely, no worries I'm getting quite adept at that
always building backwards from the final machine
and it's fun feeding a manufacturer 4 items with 1 belt π
A lot of my manifolds involve maxing out belts so I havent tried this yet
but that is sooo satisfying, smart
it can be really convenient - you just end up making more smaller manifolds. And it can work out really well if you plan it out well. Avoid high throughpput items like screws or use steel screws and send beams to a constructor that just feeds directly into hte manufacturer
it's also a bit of a work around where you might have 1 item type that needs a lot of throughput but the 3 other items kinda waste the space in possible throughput?
I've got 1 semi sushi set up with 3 items on one belt and another belt just does wire to them
yessss so smart
I have so many smaller setups where I shoulda done this π¦
the big turbofuel manifolds use a lot of fluids, so not much help there.
it just depends how you like setting things up, a lot of people don't like getting into the nitty gritty of that much planning and that's also fair
@twilit wing See #math-and-meta message for alt recipe choices
Does anyone use the encased industrial pipe recipe?
seems like it could be handy seeing as more of them are made per ingot than beams
yeah, it saves resources at the cost of more machines/space/power
wait more machines?
I thought it would've been LESS since pipes are easier to make
well besides, I just lost it rip
I Think I just accidentally discovered how to reroll a hard drive
but the EIBs themselves are made slower
in the end, it ends up being slightly more machines
ahhhhh alright
damn, all of these seem super handy
it's not like you lose the other two anyway
got 4 more drives to try and get it with so
taking a break from factory building, stocked up on rebar and on an adventure rn collecting stuff for inhalers and hard drives
imo the screw one is probably the least niche use. At least from my experience
bruh the swamp ones are so bad
im running round here eating berries and they think I got 3 supercomputers or a quantum computer to open them lol π
5 superposition Oscillators
think im done exploring the swamp for hard drives lol
Yeah swap exclusively has the super late unlocks
because the swamp is the most hostile biome
icl its great for loot, im just running around stealing all the computers and batteries and shit, easy coupons
I'd keep the computers for my first train but I need to unlock the stuff to craft the computers with to actually get them, so...
doesn't help that ||the quantum computers and superposition oscillators are ficsmas calendar-only items and can't actually be crafted||
lmao
Change the name of this channel to meth and meta
no
Ooh, imagine if that was a feature... What if you had to choose one to keep AND one to lose permanently...
then rerolling would actually have a value
Then i would immediately start using mods if there was a mod that eliminated such a design change. π
Exactly.
fun fact, if you have alts you don't like, just hit ESC, and it will give you the HDD back, and you can reroll it
this is why you use SCIM to see what you need before you go on site
picking one and getting another HDD in meantime is just better imo
Ok Frame of reference is 2,100 uranium ore pm 100 of that is used in alt. Fissle uranium making what's supposed to be a 5:1 for pellets and that's using 200 UWpm
i mean true, but when two of your alts are for making coal from other stuff, and the other one is kind of useless to you, i would rather just reroll it
Out of supposedly 1600 leaving 1400 left and that is using the normal recipe with the same 5:1 into plutonium
if you pick them, you remove them from the pool, increasing chance of "useful" recipe being there
true. but im looking for specific alts and don't have a lot of spare HDDs. im just gonna leave it as this is what I am doing, and its not necessarily most optimal, but its my system
the link i sent is a plan for a max 252 plant and making PFRs after that, so you can probably tweak it to represent your situation
I wish there was something practical to do with the extra drives...
they could be driving a belt rollercoaster
make mini pedastals and position them to be on them, try and do a mini build and flavor them as external hard drives, or whatever you dream of
I've tried that but since I don't know what amout of the other materials I need it adds them in and therfore clutter it to were I can't tell what's what
so you don't like that a tool that tells you exactly what you need, and a general idea of where those items are going is telling you all that?
if you want to just be told what to do, but don't want to actually spend time to see exactly what you need, then why are you playing this game in the first place?
So make manual adjustments.
Add PCCs for plutonium processing, and make a separate tab for the PCC factory, etc. I recommend making one just for oil products.
I'm just trying to figure out how many plutonium fuel rods I can get are I don't need to know how much concrete or how much steel I'll need is
do you know how to read the output of the tool?
So you manually input the waste from your UFR and set PFR production to "Max"
And make sure to remove the uranium already in use.
That's the thing according to it making 50 UFR pm takes 2,0** some oddball amount of uranium when my number was an even 2k
its making 50.4 UFRs as that is the max capable in the game
I sent you my nuclear build
its using the map max of 2100 Ur/min
So I'm trying to find an even number of uranium so that I can have an even number of waste/fissle according to wiki 5 waste and 5 uranium make 20 non fissle and it takes 25 waste and 100 non fissle to make 30 plutonium pellets which is a 5 non fissle to 1 pellets
what do you define as an even number?
5Γ5=25 25Γ4=100 non fissle uranium
Which is 100 waste and 100 uranium
Don't mix numerology with engineering.
All numbers are happy numbers.
Clearly not mine mine are confused
Not happy
84.1394 is just as happy a number as 80.
No that is a number with an identity crisis it doesn't know if it's a 84 or a 85
you want to know why the numbers in the map i sent are 'weird'? because 1: thats not including my map that has increased production lines to make the numbers more pretty (as i like whole machines divisible by two) and 2: thats not making nice numbers along the way, its maximizing production for the raw resources
and it sounds like you need to go back to math class
No numbers just break my brain as you can clearly tell
Can't have no .28282828287347 no no has to be whole
oh thats interesting, the server doesn't like me using discord markdown to make text bigger. imma have to tinker to figure out how they did that
Then my one question for you: WHY ARE YOU PLAYING A GAME ABOUT NUMBERS AND MATH THEN?
Because A. It's fun B. Once I figure out the math it's simple from there
It's just sometimes my brain and the calculator don't see eye to eye
im gonna stop helping you because if I try to do so, i swear im going to go crazy
I derived the equation for how many PFR/min you can make after producing X UFR/min.
And used a graphing calculator to find whole number intercepts. You can, too.
Newsbreak I'm already crazy
Cuz my math isn't mathing
Wait I think I figured it out to the calcmobile!
<cheesy 80s music>
people freaking out at the sight of numbers like 5.625 even though its just 45/8 in disguise
oooooh careful its 2.8125 uuuuuuh spooky
Even though you set the number, so you can make it whatever you want.
Oh no not 45/16 please anything but that aaaaaaaaahh
Almost all the decimals in this game are just all the "common fractions" that inches are divided into.
Ie, set of numbers A = BΓ·(2^C)
where B can be any whole number, and C can be 0, 1, 2, 3, or 4
45-81 rule.
Besides. The game doesnt actually do anything in fractions and decimals anyways.
All recipes require whole numbers of ingredients, and produce whole numbers of products every whole number of seconds.
The decimals only show up in the "per minute" values, and they are just a measurement of whats happening per minute. Not the actual things that are happening.
I remember when per min wasn't on the UI.
Better times... π
mind posting it again? it is such a gem
Everyone above has read it or was involved in its creation though? π
(i may have lost my notebook i wrote it down in)
In the search bar, type "the premise of the" and it pops right up.
oh yeha the 81 40 ratio thing
45*
close enough
Am very interested to see if it remains true in 1.0
I will never deal in "close enough."
How about in 'ish'?
Acceptable.
I'm tempted to do away with residual recipes entirely and replace them with recycled only, just sinking all my polymer or exporting them to a dedicated factory. I've had polymer back up my factories twice now because of the rounding to 0.01/min
how bad of an idea is that? Or how have y'all approached HOR residue + residuals
why tho? they work great together
residual rubber for kickstarting recycled loop
just throw in a stack when building it. recycled is great, but residual is giving me a lot of trouble
Using Residual with Recycled is how the Recycled Loop has no byproduct.
it gets rid of the resin byproduct
and that's exactly the problem I'm having - because the math I'm running requires, for example, XX.XX5 per minute, I can only round up or round down
or you can use exact number
45-81 rule
the game will round for you
it won't
it won't?
the rounded values are rounded just for your convenience, but it internally doesn't use them at all
game works on cycles, not per minute
Aw hell. guess i should go reset my recipe %s then
I remember when per minute wasn't on the UI.
Better times...
my one issue with satisfactory keeps coming back to feeling like i'm fighting the game to do what I want
Change what you want π
maybe play the game how it is designed rather than trying to play it some weird way π€·ββοΈ
Yes, how dare I design my HMF factory for exactly 2 manufactories at 100%
how weird of me
you can tho
That's like saying "I'm frustrated that I can't play Witcher 3 like I play Call of Duty."
game doesn't fight back if you do that
sevrahn i'm just gonna ignore that you tried that hard to do mental gynmastics lol
if you actually think that's anything close to this then you can just move on
I can, I won't π
and sure greeny there might be a combination of recipes where i don't have this problem, but then i'm trying to fix a problem within the game - if the game says this recipe works for these products, then it should be able to do it
Because the analogy is perfect.
all recipes work for every product in any amount, only limitation is world nodes
4.8435 is same kind of number as 8
don't be afraid of decimals
of course - but as a player, I have no way of knowing that when I type in 3.1415926536 and it changes it to 3.14 that it's still tracking everything
you do
so in my mind there's no reason to ever type in more than 3.14
that's bad UI if they aren't doing for example 3.14...
the UI clearly shows you per cycle values
per cycle means nothing
per cycle is how the game works
that's why the cycle is shown on the machine
Per cycle means everything because it's the only thing the game actually cares about.
per cycle means nothing because a closed loop cares about input and output
not in what chunk size those items are input and output
You deciding to care about different things than what the game is hard-coded to care about is a you-issue.
Not the game's.
It does NOT know more than two digits of per/min rate. You can see that in the clock speed update.
It DOES know the clock speed to 1/10,000th if a percent.
the game is the one that cares about max stack sizes @median heath
how recipes work:
- every X seconds (recipe time), it takes Y amount of input and outputs Z amount of output
the fact that you can calculate rate/min and game shows you that is a different thing, nothing to do with actual production
Which have nothing to do with per minute values?
ok sev i'm just done with you - you're arguing like a high schooler on a debate club who's using it to just try and get a feeling of winning than actually learn anything
green I agree entirely
Wow.
but the game isn't designed around balancing cycle times
Cannot accept how the game functions so instead diverts to personal insults. π
the game is designed around input and output to each machine
that's not what I am saying
although we had that for pretty long time
then tell me how i "should" be playing the game, cuz i've apparently been doing it wrong for like 2 years
so it kinda was designed around balancing cycle times
the entire idea of overclocking and underclocking is adjusting the cycle time to blance the average flow rate
in PER MINUTE metrics
Again, the game does not care about per minute in any respect.
Except that it didn't start with any per/min info in the UI
or to adjust how often the machine outputs items π€·ββοΈ you can look at it both ways
sure, which is why i don't get your hold up on why per minute is an inherently useless thing to look at when designing a factory
that's not what I said?
.
that wasn't about cycle times? that was just a generic statement that if a game fights back, it's usually a sign you're doing something weird and there's easier way
which again, means i'm fighting the game to do what it wants me to do and not what I should be able to do mathematically
π€¦ββοΈ
sandbox means you make tools and let the user decide how to use them. as long as all the rules are followed, there should be no problem. but in this case, the rule is buried behind unclear UI
that rule being "hey I'm actually tracking exactly what you said you wanted but I told you I was shortening it"
"I told you I was being imprecise but I'm actually not"
oh and becuase I just now found a way to word this, the game very much cares about per-minute. it's one of the intended interfaces between the value we both agree the game cares about (cycle time) and our desires for a clock %.
"The game cares about per minute, I am supporting this statement by declaring the very things that prove it does not care about per minute."
the ENGINE doesn't care about per-minute
- you take the amount you want (e.g. 2/min)
- you calculate how many buildings produce (e.g. if one machine makes 0.8/min, you need 2/0.8 = 2.5 machines)
- you calculate the input the machines need (e.g. if each machine needs 20, you multiply it by 2.5 to get 50, so you need 50 in total)
- you repeat this for the input, and then for input for that, etc etc., until you get to raw ores
number of machines is how many you build. If you get decimals it's fine. For example 2.845276 constructors means 2 @ 100% + 1 @ 84.5276%. If you get more than 6 decimals on the machine count, round up to 6 decimals (because game can do up to 4 decimals of clock speed). That makes the machine starve once in like 50 hours for a second or two, but that's hardly a problem. Or you can use some specifically crafted numbers to evade 7+ decimals in machine counts, known as 45-81 rule (but again, most people don't bother with this).
I agree with you on every single point. MY issue is when you put in 84.5276 and the game turns around and shows you 84.53, you as a player have no reason to believe that it is actually tracking those extra digits
and you have no reason to believe that it's discarding them either
so you turn to wiki or community and ask
game is in early access and not as polished as it may be in the future
????????
in what world is changing a value when it's echo'd back not presumed to be a changed value
if the UI returned, in our above example, 10 as a number, you'd presume that to be different, right?
you don't really change X/min anywhere, so idk where it "echoes it back"
the UI literally says "target production rate ______ per minute"
the only thing you change on machines is clock speed %
on the low level yes
If you put 84.5276 parts per minute, the game rounds it to 84.53 and displays the clock speed accordingly.
If you put the clock speed at 84.5276% it will stay exactly that.
on the high level, no
on the high level, yes
no. both values can be changed. ultimately they're both used to calculate a clock %, but I can change the target production rate instead
Do. Not. Manually. Enter. Target production rate.
then don't let me
if you put in items per minute, it calculates clock speed, puts that calculated (and rounded) clock speed into the machine, and then calculates per minute production again from it to show you.
If you put 20/min in the field, you may get something different than 20/min
Then talk to the legions of players who bitched about not having it until they finally added the box so they could shut up and feel better about themselves even though the game doesn't care about that box at all?
and now we're getting to the core of the problem. the game is giving me two apparently equally viable options, and one of them is changing my value while the other is apparently perfectly precise
Sometimes you do things just to shut the playerbase up π€·ββοΈ
but I have no idea of knowing that 1 that's even a thing or 2 which one to trust
again, if you have no idea about something, maybe consult wiki or community?
why would I have no idea about it? again, when two things are presented as equivalent, why would I assume they aren't?
why would you assume they are?
because it's facially the obvious interpretation
like f.e. minecraft as a sandbox has almost no information about anything and would be pretty hard to play without wiki
not really
it's quite easy. there's minimal mechanics
and they don't depend on riding a knife's edge
yeah good luck figuring out quazi connectivity without wiki
that's a fair point, lemme think
quazi connectivity is something common enough that the average player will run into it as a problem, and they can generally get around it by simply making their build larger. I can see that as paralleling to here, but the fix is basically "change recipes" or "sink the excess" which the latter is probably going to be my approach and it'll work just fine, so ultimately this is pretty moot
was just one random example out of like 100000
ehhhhh
most don't cause the problems quasi-connectivity do
unless you just mean indirect power
minecraft doesn't give you info about anything, many things you figure out yourself, many you don't and have to turn to wiki
yeah and figuring out for yourself is very different from being presented with false information
which you could easily figure out yourself by doing the calculations manually
again, that's not the problem
checking the recipe speed and per cycle values, seeing it doesn't fit, so something is wrong, one of the two displays
the problem is that apparently the per-minute number is treated differently than the clock % number
it is, because it was added later as a quick patch because people wanted it there
and why would I know that
You know that now. Why would you keep acting like it's someone else's fault you were inconvenienced by being ignorant of it?
if you play any game with assumptions that any of your "I think it works like this" is always correct, then you're gonna get burned like this a lot
jackie chan has a famous quote that feels relevant:
Whatever you do, do the best you can, because the film live forever. "No, because, you know, that day raining and the actor don't have time". I said "Would you go to every theater to tell the audience?" No! The audience sit in the theater -- "good movie", "bad movie". That's all.
good UI, bad UI, that's all
early access game, that's all
and after it's released?
then check the UI at that point π€·ββοΈ
or better - make a thread on QA site to let devs know that you have issues with UI and suggest how they could make it better. Don't cry here where they don't read
that's like the whole point of early access - you get the incomplete game and you get to suggest new stuff or fixes
if you use any computer program WITHOUT the assumption that "this number is showing me X idea which tells me the game thinks X idea" then I legitimately have no idea how in the world you read words on a digital screen because you don't know if they're the same as what I typed in
the game's like 95% done - this has been in here since... what, update 6? 5?
it's not lol
I bet we keep getting updates after 1.0
For a while
Mix of polish and bugfixes plus dlc
since update 4 (or 5? idk) there has been a dedicated part of the team that were working on features that we haven't seen yet. If you want to tell me that two years of work will result to only 5% of game, then idk what to say
(This is the root of why I refer to the time before they added ppm to the UI as "better times")
that's news to me - in Snutt's roadmap video he said there was a branch but not an entire dedicated group
People complaining about having to do math in a game about math?
well somebody worked on that branch π€·ββοΈ
"dedicated" may not mean the same people at all times, but there were people working on it
Game giving you false hope about how things work instead of forcing you to learn how they actually work.
there's no indication of how much is on it, so why are you assuming it has dedicated people instead of tickets that the normal devs do
that's generally what the word "dedicated" means...
oh crap I assumed it meant the same thing here as it does elsewhere. that's on me
sorry for not being native english speaker, will never happen again
IDK I thought the answer was really obvious the first time I put a rate in and got a clock speed of 83.1667 instead of 83.3333
live and learn /shrug
Deductive reasoning is not a skill everyone posesses.
honestly if it means anything you're fluent enough that I held you to the standard ofa a native speaker lol
I'm not even smart enough to speak a second language so I can't give you crap beyond what I already did
yeah google translate is crap at my language so π€·ββοΈ
Bing translate though π
Bing has never led me astray π€·ββοΈ
What language?
czech
YOU BETTER CZECH YOURSELF BEFORE YOU WZECH YOURSELF
I tried it twice and it completely missed the context of what I was going for so I joined in on the memes of it
I don't know a lick of it. But I like watching this YouTuber series about Prague and how to visit without getting scammed.
If that counts for anything π
honestly if you're worried about getting scammed, evade buying anything in centre of prague
Seems like it.
the costs there are just inflated because tourists do not know about how much should things cost
since our costs are generally pretty cheap compared to e.g. EU countries, they can charge similar costs as in germany or even more and get away with it because tourists will pay it without thinking too much about it
and then you go 15 minutes away from center to some side street and have 3 times cheaper lunch with better quality
Honest guide!
(I know this is offtopic, last msg, if you want more, ping me in #off-topic-general ) but I personally like Czechia as a whole and if you want to come, I could show you around lol π
I think ultimately, my lesson learned here is to stop underclocking. Let the machines starve and self-balance. clear out any excess secondaries during a factory's warm-up, then let it go
have a few batteries and lean on duty cycles to balance my average power consumption
um, that's not what I've suggested, but π€·ββοΈ
I didn't say it was
what's wrong with the way I suggested then? π€
you mean set the clock % directly?
yeah, since that's based on the machine count you have anyway
because that's making the exact same assumption that I incorrectly made about parts per minute - that the value you put in is the value that matters. and it doesn't. the thing that matters is resource availability and output buffer availability
(talking about this apporach)
clock speed matters tho
so if I lean on that, which is a valid assumption to lean on, then I'm guaranteed to have my flow be balanced
sure it matters, but we agreed that's not what the game operates on
no
it operates on cycle time. and i can't set the cycle time.
which is a function of clock speed
no, that's function of cycle time and recipe
game calculations go this way:
- take clock speed
- multiply it with recipe time
- get cycle length
you can't change recipe time, but you can change clock speed, hence you can change cycle length
with the assumption that there's no pre-calculation to the clock speed between steps 1 and 2
there isn't
which is the assumption I made about items per minute, because I thought it was doing exactly that with items per minute
game saves clock speed and recipe for each machine, nothing else (well things like position, items inside, etc., but you know what I'm talking about)
but if that's the case, then cycle time doesn't exist anywhere
recipe
no current cycle time
recipe has duration
like if a recipe has a 4s duration and a 50% clock, 2s exists nowhere
like current progress on current cycle? that gets saved as well obviously
the game only tracks 4s and 50%
may be in RAM π€·ββοΈ
that's not what was said to me earlier, but ultimately, the fact that this is even having to be discussed means it's not a safe assumption
the only safe assumption is 1) items are present to start the recipe and 2) there is room in the output buffer to complete the recipe
etc
I mean if you want to be exact, you'd also have to track if the machine is running or not for those shutodwn and startup animations and delays
obviously I'm simplifying things to just the relevant parts
which is: changing clock speed is the most accurate you can get to changing "average input/output per minute"
ah it was a sev comment
sure, but again, that's only becuase you already know it is
that approach doesn't even sidestep the assumptions I would be leaning on by starving machines
and the only cost to starving machines is a wonkier power graph
which gets us again back to "game can't show you all info" and "some info can only be gathered from wiki/community"
and "if you feel like this could be improved and know how, please make a QA site post"
and it gets ME back to "the game is showing you incorrect info, so screw you for not knowing that it was wrong, trusting it was right, and not asking someone who knows already that it's wrong"
i'm doing so rn lol
trying to boil it down
the info is technically correct, just rounded. Practically "items/min" don't exist in the exact form anyway, it's just aproximation that you get close to over time
I personally think that for most cases the info is enough and once you start doing a lot of math, you figure yourself pretty quickly that the info is rounded (it's possible to figure it out just from one machine's UI π€·ββοΈ )
I have a silly example for you that shows the game being inconsistent in the way I mentioned greeny
but it demonstrates what you were saying so it's not to show you as wrong or anything
Fun times in M&M today i see. π
heavy encased frames is 2.813 items per minute at 100%. if you set the target production rate (which again we said not to do) to 2.813, the clock will not to go 100%, but to 99.9111% and the target rate will be 2.81
to reiterate, setting the items per minute to 100% does not set the clock to 100%.
fun fact, it's 2.8125
EVEN WORSE LOL
It's 3 every 64s
you can use e.g. my tool to see accurate numbers π https://www.satisfactorytools.com/codex/items/heavy-modular-frame
3/64*60 = 2.8125
Why? Its just 2 13/16
Which is again proof that the per minute shit on the UI is just there to make you feel better about yourself.
What the game cares about is 3 every 64s.
I guarantee you that it doesn't. it actually cares about 3 every X instants, where the net result is 3 every 64s
...
you tried to do that BS on me earlier
Believe what you will.
π
unfortunately, that's what got me into this mess π¦
All machines only ever deal in whole numbers of items consumed and produced per cycle. And all recipes take a while number of seconds to complete per cycle. All we can do is adjust that rate with clocking and clocking is restricted to increments of 1 ten-thousand of a percent.
3 per cycle.
64s cycle time.
If you change the clock speed to 50%, you now have a 32s cycle time.
Still 3 per cycle.
Cycle times and clocks speeds are what matter.
Per minute is irrelevant.
I let it mix recipes and that's where this thousandth precision issue came up
There is no "net result."
Cycles either happen or they do not.
tools should be precise enough to use the numbers in the game without issues π€
the game doesn't calculate things every second. it calculates probably more often than the frame rate of the game
it calculates % of current cycle
https://www.satisfactorytools.com/production?share=FLJZDmp3qJNe3krOCnud
I backed up polymer at the residual plastic refinery because I was, again incorrectly, inputting items per minute and not clock%
tip: tools show you the clock speeds when you hover over them π
sorry backed up polymer at the crude oil refineries which starved the residual rubber
yeah i know that NOW lol
remember i know more now than I did when I built this 2 weeks ago. it just now shut down
for the second time
Its not BS. its EXACTLY what the game actually does. Heavy encased frames takes 64 seconds and then the items in the input and output buffers increment by the exact whole number of items the recipe says.
the game doesn't calculate things in 64 second increments. that's the BS joke I was referring to. the game has no notion of "64 seconds" but some time increment, probably milliseconds, that it knows it can add a bunch up to to get 64 seconds
just like the game has no notion of "items per minute" but it knows it can calculate a clock% from that
I was being intentionally obtuse when I said that
I mean you can go further and further with that and end at "computer has no notion of the game it just does stuff with electric pulses"
(or continue even further to atom level)
but for useful discussion, saying "game counts in recipe cycles" is good enough
Ah, so instead of getting along you want the convo to be unconstructive. Got it.
Sounds like you should learn about this thing called "irony" and this other thing called "rhetoric" and maybe jf you're feeling frisky, "a "joke" instead of getting upset about a discussion you came in halfway on and tried to get snarky about
But why is "the game works in items per minute" NOT good enough? This is the hair splitting we've been talking about.
They're both derivative values
The issue is merely precision
because IPM is derived of the cycle
And the cycle is derived of the instant
not relevant to #math-and-meta
One is something the game actually tracks (whether a cycle completed or not).
The other is something players injected into the game because they are too used to how other factory games work.
just to be clear:
it's completely fine to use IPM (and majority of people do)
I'm just talking about what info you read from UI. In machine UI, the clock speed % and recipe time are exact values. Anything other is derived and rounded. So for the purpose of reading stuff from game UI, cycle time is a better choice
For the purpose of calculation, feel free to use IPM (it's generally easier), but beware of where you get the values from
What's IPM
items per minute
Oh
Hahaha now we get to the real fun stuff. "Exact value" isn't exact, because it's data stored in a variable. There's always a precision limit. (I don't expect us to debate that out but it came up when I was drafting my Q&A submission)
(Before I had to bail to wake my kids from nap)
#math-and-meta π
You're not discussing Satis anymore with that rabbit trail.
You're purely talking computer science.
Maybe I don't understand what the word math means
given that both values are within precision, they are exact
Which gets us right back to IPM having the exact same issue. Provided your input is within precision, it's exact
Imagine having to announce to the world you're blocking someone, as if we care
The case study of just how round circular logic can be.
Bingo!
cycle time is always in seconds (integer, so 100% precise)
clock speed % is rounded to 4 decimals, so 100% precise to those 4 decimals given floats have usually like 8 digits of precision
Doesn't the game handle common fractions nicely like 1/3?
IPM is precise if calculated from the exact values, not if you read it from machine UI
which is why I posted the message above - #math-and-meta message
I don't know how many other ways you have loaded up to explain the exact same concept.
But let me know if they eventually get it π
there's no 1/3 in clock speed or cycle time
Ehhhh I think maybe you're right but only by a technicality
By everything rounding in the way it should when you truncate values
cycle time is integer
clock speed is float with 4 digits of precision max
neither of those can have 1/3 in them
(and yes, I know computer can't do 1/3 exactly, so technically nothing can have 1/3 in it)
So computers can do 1/3 but only by assuming that if you get close enough, means 1/3 similar to the hard coding of "infinity" in floats
I doubt satisfactory is doing that, but for all intents and purposes, the player is supposed to think it does, because 1/3 truncated to approximately 1/3 then multiplied by 3 will round back to 1
π€¦ββοΈ
but again, you're not counting with 1/3 anywhere
And none of us have any way of knowing that. 33.3333% with a displayed value of 4 when the base is 12 tells me 1/3
all cycle times in the game are integers
all possible clock speeds have max 4 digits of precision
any calculation involves these two numbers
You keep thinking I'm disagreeing with you. Stop it lol
I'm agreeing with your base assumptions
I have a way of knowing that because that's how the game works. From code persective
Great, then write up the code for the end users so we know too :p
what
both clock speed and cycle time is shown exactly
there's no need to do any code
But ipm is not shown exactly, because you know it isn't
Insert sev's circular logic joke
which we've cleared like 3 hours back, why are you bringing it up again?
(and again, you can prove ipm isn't shown exactly by just messing with the clock speed slider/input)
It's not a joke.
You keep circling.
Because my whole point has been UI clarity and how the burden is on the user to know that they should trust this part of the ui but not that part by way of the game being early access and there being a discord of people who have looked at the code to translate the game into real world rules
Game does not care about IPM.
This isn't a difficult concept to grasp.
Sigh...
in many parts of the UI, game only shows cycle, input and output
IPM is shown in just a few parts
Willful ignorance with no intent of self-evaluation.
+1 to the mute list.
Both are derivative values. That's my point. Ipm is merely a lower resolution derived value
Again, imagine anyone caring about your block list
they are not, since cycle + clock speed are saved
This isn't twitter
BASE clock is saved
there's only one clock
Sorry base cycle
well no, base cycle is part of recipe
there isn't.
that's why what is saved is "recipe" and "clock"
Oh I misread what you were saying. I thought you were saying the game tracks the current cycle, not the base cycle
You just said cycle and clock, not recipe and clock
Thats my bad
ignoring the fact that game saves current progress of cycle and other stuff, the two important things for our case is "what recipe is set" and "what clock speed is set"
Yeah ignoring that stuff
so how is 64 seconds derived, if it's litearlly read from the recipe?
@wind spade , pls, either troll or someone so deep into themselves they cannot be reasoned with.
friday night, nothing to do π€·ββοΈ
also listening to this new banger album in meantime
Work on the next tools update π
even the blocked message msg is getting annoying xD
I'm leaving tomorrow morning, so I don't want to start anything that would take more than like a hour
fair
Because 64 seconds is a derived value from the actual time unit, but because it's easier for the devs to work with (!!!) they store it as 64s
Who said you could leave?
actual time unit the game uses is seconds though
tickets to a ball we bought
yes, like any game, delta time
(Also entirely possible 64 is the default value set in the data table for this recipe, meaning it is derived from nothing)
How elegant. 
it is π
Data tables are fun. π
Not as fun as napping though, which I am about to do. Enjoy your evening.
I lost track of my train of thought while reading that cuz i got jumped on by a toddler, but it's clear the dogpile is here so I'll stop and chat with you next time greeny. π have a good ball!
Ty for always remaining respectful when it's very easy to insult others while hiding behind a keyboard. Good to know there are still adults around.
Welp, I for one am glad this discussion happened as I've now learned not to enter the IPM field in the machine UI but change the clock speed instead. Thanks all!
And I agree with ERRORMONSTER that it would appear to most people that changing either field should give the same result/accuracy. That certainly was the case for me.
(again, things were better when the IPM field didn't exist)
Especially since when I rely on Satisfactory Tools calculator, where you input the IPM you want... then it would seem to me that you can just put the same IPM values in the game as well.
Yes, as I said in my previous message. π
In light of another recent test:
If you have a mk 2 input manifold at 600, use a mk 2 pump at the feed input. Even if the manifold is absolutely flat.
On an output manifold, having the mk 2 somewhere in the middle is what seems to work
You're saying having the pump negates the need for looping?
Circumstantially, yes
Has to be a mk2?
Even with looping mk 2 can help
Will you finally accept VIPs if they still work in 1,0?
No, but i wont erradicate them from the manual
I dont like em but that doesnt mean they should go undocumented
I don't think it's such a big deal that it's worth noticing 
As in:
- if you don't care about wether a machine will produce one extra or one fewer item than expected every few million cycles or so, it doesn't matter which field you use;
- if you care about such precision, you're likely to know enough math to know that (eg:) 10/min x 33.3333% won't be exactly 1/3 of 10 but rather 3.3333/min as shown in the output/min field (it doesn't even show as many decimals as the calculation would lead to!)
That isn't a lie
Tools give you the clock speed when you hover over the node
You can't get 20.
It isn't possible.
As Ven just pointed out, it's close enough that no one really cares. (Which is why it is simpler for Tools to just say 20 instead of 19.99998 or 20.00001).
You can. By limiting the machine input
pretty sure if there's a recurring decimal there every number after that is going to be out by a really small margin
Grumbles about uneven power draw
You are talking about how much the machine makes if it works at 100% capacity. Tools give you solution with minimal possible downtime to achieve 20 exactly
go use 45/81 rule then
Indeed.
The point is that tools do not claim "if you clock to this, you make 20 exactly", but "because you're inputting only 20, you are only making 20"
See above
makes sense
Is about 50 batteries a minute enough to bring all uranium in the world to dune desert?
theres only 2100 which is barely 5 mk 4 conveyors
you can do that with 5 or 4 drone ports
So you are saying it is enough?
definitely
Too much*
i had some spare resources and satisfactory tools told me i can do about 50.4 batteries a min
just do 50 and if theres excess batteries: who cares
i want to make a 40 per minute computer factory. but having a tough time deciding which circuit board recipe
ive got quartz and caterium at equal distance. but the quartz one just seems easier + 1 row of machines less
both are about equal
ive done factories using both and it doesn matter really, they are both a pain in their own way
Equality in Satis?
Freedom of choice?
Never that... 
but both are good right
well, if you match it with the computer alts, you can import either a lot of caterium or a lot of quartz @swift whale
yes, both are good
normally 1 recipe is just totally superior. but for this one its a tough pic
there rarely ever is a "bad recipe"
idk how you define superior but every recipe offers a trade
Wait what?
i tend to use Caterium Circuit board when i do caterium computers
cause i dont like mixing up that many resources
Ct Comp + Electrode CB π
I know its just oil but isnt the price in oil quite steep, relatively speaking?
ive used that recipe once but it wasnt a very serious factory
Everything is a tradeoff. π
ik
That said I'm fairly certain it will be adjusted in 1.0
I cant really pick electrode CB until i got most of the recycling recipes
the crystal oscillitor recipe looks very intensive tbh
the oil cost is a bit too painful without recycled for me to use Electrode CB
Fair.
not that much
you need a manufacturer either way
if its not for computers its for the oscillators
Even without. Base Rubber gives you HOR to make Coke.
yeah but at an inconvenient relative ratio
Ratio smatio.
I still do Ct CB + Crystal Comp almost always.
But El CB + Ct Comp is slept on imo.
oh wait
it uses half the coke
you get from the HOR of the rubber recipe
i never noticed
Combine with HOR alt + Residual Rubber and you're cooking.
(Can't remember if HOR Alt or Poly Resin alt was the one I used)
30 crude to 3.333/min CB
aka 90 Cude/min to 10/min CB
oof, that would need 945 oil in my 15/min computer factory
Or you could turn the Islands into a 200/min.
and that 945 is just the rubber for E CB
The fact you can make 45 Comp/min without oil on Speedrunner's is just.. fun.
honestly the polymer resin alt might be the better choice if you dont use recycling if you wanna do ECB
65 Rubber + 60 Coke
..... or HOR alt + Poly....
hold on, lemme calc something real quick
HOR + Poly alt = 150/min Resin and 60/min HOR
which is 75 Rubber and 180/min coke
enough for a fully overclocked ECB assembler
plus some more coke to spare (67.5/min)
so 90 Crude to 12.5/min CB
25% efficiency gain
compared to using just the Rubber Recipe
How much oil to HOR and how much to Poly?
Nice.
i never noticed how both recipes compliment each others numbers
HOR alt turns Poly's 130 to 150 Resin
and HOR alt bumps total HOR up from 20 to 60
Hey, there's been plenty of times I find gems in recipes when I go over them another time all these years later.
i still think the oil cost for poly is way too high for what it offers
We'll see what 1.0 changes.
what to do with 22.5 HOR to spare
Like Fluid Trains were agreed upon to never, ever be used until last year when I checked the math again and realized they are better than Freight.
could do a 50% overclocked Coated Cable Refinery
or just make Fuel or something idk
I assumed we got stuck with the idea back when their capacity was off (50 and 500)
Aye.
That's what I'm getting at.
We get stuck with things. And when we reevaluate, we find stuff like the above concerning HOR + Poly.
in what circumstance again?
I mean I'm still stuck underclocking π
They have higher throughput.
due to the canister loops?
The only circumstance where Freight wins is if you're not looping cans.
If you're just creating on one end and destroying on the other, for example.
gas still wins out i guess?
Drones exist, i know
Always package gas, yeah. 4x compression ratio OP.
Buffers not keeping throughput up doesnt help either
Buffers with a window so you can see fullness is all I want π
No buffers best buffers
That translates effectively to "no trains best trains" π
Given that is the usecase of buffers.
Only if you do fluid trains
There are non-fluid buffers too.
Those are called containers
did you know theres kinda a mod for that 
"Kinda" isn't good enough.
it turns the entire buffer into a window or rather fullness gauge
I don't want the ENTIRE buffer changed.
Ew.
thats all ill say on it, rest of it would be for the mod discord
When and where?
Zappers only.
No Jetpack.
Blade Runners.
You choose location and time of day.
On top of the manta at noon
Each party gets 2 bacons for heals.
hi, i got a questin about coal generators. can someone help real quick?
No no, we help fake slow around here π
just ask your question
As opposed to asking someone else's?
i have 128 coalgenerators and 4 normal coal veins and 1 pure. how do i split the coal veins perfectly?
Manifold.
?
!wikisearch manifold
you dont need a perfect split
this is efficient
It will be.
ok
Manifolds and Balancers have equal efficiency.
What do you not get?
im new to the game, so dont hate on me. i dont know what they are trying to tell me
Ok what's your highest mk belt?
4
And you're still making Coal Power??
Ok so a mk4 can feed 32 Coal Gens.
why not oil
im at phase two of the spaceelevator
Put them all in a line, one splitter in front of each.
If you have mk4s. You're on phase 3.
if you have mk 4 conveyor you have access to Oil
yeah i need to complete phase 3 sorry
i have mk4 assembley lines
i dont have computers so i cant build this with oil
@proven raptor mk4 feeds 32 gens.
Put them in a line. Single splitter in front of each.
mk4 miner?
No such building exists.
Is that what I said?
I didn't say anything about miners.
At all.
This is about belts.
oh yeah i have mk4 belts
mk4 feeds 32 gens.
Put them in a line. Single splitter in front of each.
yes
Is that part clear as to how it works with 100% efficiency without you needing to do anything else?
i think so yes
What are you doubting?
myself?
Elaborate.
What about the above system is unclear in any respect.
i put 32 coal generators in a line and put splitters in front of it
i have done that already
And it runs without issue.
Because splitters will self-balance.
So "how to feed equally" is just manifold it.
That is too variable to answer.
Miner mk + clock speed.
Find out the exact amount being given, then you can know how many generators.
put a miner on it
and then pick a clock rate
then divide the rate you get by 15
thats your number of coal generators
I'm trying to build a drone networks that feed itself with battery
from a manufacturer which is fed with drone ports
if drones are loaded enough it must be working right ?
(pushing more batteries than it consume to feed the battery factory)
The maximum number of batteries each drone can consume is 4.2/min so as long as you feed that amount you should be good. The only real limit is the total number of batteries each drone can carry determines its travel distance. I recommend connecting the drone ports to the same power grid. There is a bug where if they are on disconnected power grids, the game will not correctly calculate the battery consumption of the drones.
what powerplant should i focus on? Nuclear seems to complicated to get clean with so many different resources. Thinkin about some turbofuel layout but the sulfur and crude oil nodes are so far away, so might setup a drone thingy for that
ive never tried setting things up with trains since i dont know yet what/where i need to transport
if you get the heavy oil residue and one of hte diluted fuel alt recipes that can do you for the whole game if you like and is simple
Nuclear.
Whatever is more fun for you. You could do entirely coal gens if that's your thing.
But that would be a lot of coal gens.
I think you can set up 180 coal generators in the blue crater (at least WhatDarrenPlays calculated... Haven't seen enough of his vids to know if he does it, but he's pretty good about his calcs) but yeah, agree, do whatever feels fun for you. If nukes feel like a chore, explore other options! Eventually you might be in a spot where nukes naturally become something you want.
Should be 160, and that's anywhere with 4 normal coal nodes. You could do that there, and at the snake tree forest lake, and also at NE RD "Coal Bay". And that's not even factoring in the crater lakes coal or anything else.
Hell, burn petroleum coke for the memes.
thanks for answers, yea i dont want to setup and plan a huge train line for nuclear right now, i dont wanna have to redo the trains once i do them. My world is mostly empty right now with one megafactory that im workin on getting rid off cause its ugly
I've been chewing through just planning nuclear power, for a while now. First attempt and I'm aiming for the stupid goal of 300/min plutonium waste.
I think I can do it all with drones.
Lol true... Or cut your burn rate in half to feed 300+ plants with compacted coal.
If the sulfur is convenient. Which it rarely is.
The nice thing about coke is it's so simple to produce. You just need to do something with the resin.
But fuel still gives almost 3x the power as coke from the same amount of crude.
how do i make an overflow system for fluids?
i have to send 2100 fuel/min from 21 blenders to 70 fuel gens max overclock
Fluids cannot be overflowed. They can be packaged and sunk. And you don't want any overflow if your supply = demand.
? They can be overflowed.
That's how you know how much to package and sink π
i split the gens into 10 rows of 7 gens each. now i gotta figure out how to evenly distribute
In this case, the answer seems to be zero because 2100 = 2100
Don't "figure it out"
Do not think about pipes in terms of belts.
yeah i know
You cannot split or merge.
You cannot direct specific amounts.
Fluid goes where it can. So build a system that accounts for that.
2100/600 is between 3 and 4 so you need 4 pipes at least
this is what i got so far. 5 belts 420/min each
Alternatively each group of 3 blenders makes 300/min fuel which would supply 10 gens @ 250%. Do that 7 times.
hmm yeah that might be the way to go tbh
would save me a lot of effort and headaches
Im fairly certain that the pipe manual that is in one of the pinned posts of this discord has a good diagram for a pipe overflow device. And a bunch of other useful things.
take each pipe and put it into exact amount of gens that it needs
why does it need to 'over flow' ?
do what greeny said
#math-and-meta message
The last link there should help
If I have all the math figured out, do you know why this is not at full efficiency
because the efficiency meter is full of lies. look at the colour of the lights on the machines to see if any are stuttering
the ones on the top level are, but the pumps say they can pump to that height
You would need to look at all the machines that are fed by the extractor - do their lights stay green?
you can also just stare at the extractors light for 30 seconds - if there's something wrong it'll likely stutter within that time
like some say they arnt getting enough fuel, but the pipes say they can carry enough, and the extractor isn't running at full capacity
ok so trouble shooting 101:
Look at the machine - is the problem too full of output? or too empty of input?
In this case empty of input
Action - follow it back
the machines making fuel - are their outputs full and stuttering? or are their inputs empty and starving?
starving
so the machines making fuel aren't getting the inputs to MAKE the fuel? Cool, is the input just oil?
yes they arnt getting enough oil, but the oil extractor at max capacity
ok so you followed it back to the oil extractor - is the oil extractor's output getting blocked up and stalling sometimes?
there is no where for it to get blocked up except at the refineries?
does it's output get full of oil and make it stall
oh yes
control panels are a gold mine of info π
ok so the situation is this
Fuel producers aren't getting enough oil - Oil extractor is backing up.
There's a few different issues this could be - can you think of what some might be?
Well 1 main one is most realistic
The pumps are not at max head lift so I wouldn't think it to be how much to moves the oil upwards
if you're getting flow at the top and most machines are working? it's probably not headlift
there's sometimes very edge cases where you're just ad the max headlift and some machines will work but most won't but it's fairly rare.
are most machines up there working?
no, they are out of oil
ok so the answer I was looking for for the 'main reasons for this' was 'flow issue'
Something in your set up is stopping flow. Could be head lift, could be a buggy floor hole quite a few things.
The only way to really diagnose that is with some over head shots of the pipe layout from extractor to refineries
Maybe make a post in #1038092680493801533
that way we don't flood this channel
ok ty
second image is quite frightening
im not surprised that this doesnt run well at 600/min
its too complex
please but those 3 pipeline pumps on the ground
not 3 m off the ground
I'm sure that'll be part of the qna post discussion π
for sure
I was mostly trying to establish some basic trouble shootings skills π
for relevancy:
#math-and-meta message
the pipes here are raised and all
which likely leads to backflow into the junctions
"If you have a mk 2 input manifold at 600, use a mk 2 pump at the feed input. Even if the manifold is absolutely flat.
On an output manifold, having the mk 2 somewhere in the middle is what seems to work"
Are you saying something has cropped up that sometimes means you need a mk2 pump before the start of a manifold now?
mk 2 seems to like it
looping and flooding not enough?
you can check my QA post on pipes issues
this is one more safety net to looping or splitting
if you dont do looping or splitting, then even mk 1 vs mk 2 pump makes a difference
and by "On an output manifold, having the mk 2 somewhere in the middle is what seems to work" you mean a pump somewhere along an output manifold?? that's weird.
yes
weird
same reason why you put it before the input manifold and not "in the middle of the input manifold"
the mk 2 needs enough pressure to not be affected by the first row of junctions when its at 600 flow
point is you need the 50 m for the 600 flow to survive the very first (for input manifold) or last junctions (output manifold) to not get interrupted
Huh, the only times I've really had to add some pumps in for something like that is when I've branched a pipe into multiple manifolds. Usually on different elevations (not that I do that frequently)
if you dont do branching or looping, this absolutely makes the difference
That's a good heads up, thanks π
I can't seem to search the qna forum by author, do you have a link to your post?
i kinda feel like i should ping laura with this
cause i remember her post with the bottom feeding manifold
and how mk 2 somehow helped
https://questions.satisfactorygame.com/post/60352c10fedfd12dd739b885
second save file / issue report
oooohh the main qna not discord, gotcha sorry
Also do you have an idea of what might be the differences in a flat manifold whether it might need a pump or not?
ive known for a while now that head lift really is pressure
its just that for the flow math, there is a cutoff point
at like 1.5 m
wdym 'cutoff at 1.5m' ?
above 1.5 , it doesnt directly affect flow math between pipes
but turns into the "all or nothing" head lift we know
below 1.5, head lift is the pressure that dictates how much flow there should be between pipes
as in , if it doesn't have to rise more than 1.5 post the junction nothing happens?
1.5 m is the pipe diameter
this is literally a 0 m height difference manifold
a flat pipe has a head lift of 1.5 or 1.3 m
cause thats the pipe volume itself
right ok, you know I never considered where the machines started calculating head lift from
center point of the pipe, but then head lift is still considered for the full pipe diameter
would issues be solved if they just calced it from the top of the pipe?
no that just shifts it
once a pipe is full, it has an extra volume called "overfill"
overfill allows pipes to sustain max flow (cause then the normal volume doesnt drop - it gets pulled from overfill)
overfill also affects how pump head lift gets transmitted
and i believe it is the interaction between the overfill portion for sustaining max flow and passing on head lift thats important here
if that portion suddenly varies, you get a major reduction in passed on head lift
buuuut if the applied head lift is big enough... then you dont lose as much due to flux
Wild.
I guess just to be extremely specific, what I was asking before was if you took 2 flat manifolds (not identical) what do you think might be attributing factors that 1 might need a pump at the start and the other wouldn't?
I don't deep dive like you, I mostly try to focus on what rules can be bent/broken and what might be possible causes
Conditions
- Its a Mk 2 manifold at 600/min
- the Sidelines are not Mk.1 Pipes
- it is not split or looped
Then you probably benefit from a Mk. 2 Pump at the input
ah ok. Maybe I misread, I thought you suggested using the pump on a looped single manifold
it may not be needed, but it definitely wont hurt either
You know I don't know if I've ever gotten a mk2 system flowing at 600 w/o a loop. And I mean ones that are just 1 junction being split to only 3 machines or something
open that test save from the QA report then
and have a look around it
the test build right in front of the HUB demonstrates what i just talked about
Oh I believe you - I very likely just design layouts that jsut do not play friendly w/o loops π
I like my pipes to look certain ways and it very quickly got to the point of 'ok just loops everywhere, all the time, loops'
this is funny btw
"Dark Purple: violently slamming the limit slider of a Valve from 0 to 300 multiple (around 30) times results in the Valve generating head lift. This manifests in the form of the Valve in the attached save displaying a flow greater than 0 mΒ³/min on an empty pipe. The oil stops at the roughly Z = 0m. This does not happen above Z = 0m. Unfortunately, this permanently applies to all Valves in this save, even newly built and dismantled ones."
That's quite the fluid behavior 
was fixed
shadowfixed
i noted that in a comment below
what pictures should I show?
Angled from above - from the oil extractor to the refineries
what made you even think of jsut ramming the slider back and forth?
just for funsies
oh so now it works. I feel its just toying with me
remember - don't trust the meter. Look at the the light
its constantly at 20 in there now
Did you change anything
wait no its still broken. I gtg now but I just removed the valves to regulate flow I added after it started not working
when you have time make a post with the images I suggested and I can take you through proper pipe design if you like. Also remove all valves and buffers
ok ty.
That reminds me. I had a fluid freight platform that was full of flyid but the sight glass in the interface showed only half full. Is that a known bug?
If you have all a hoverpack, all of the necessary alt recipes unlocked, and have a lot of power shards you can use my nuclear power blueprint set.
Its known
I just made my first coal generator setup witch consist of 8 Generators and 3 Water Extractors. But now the last 2 cut out because they run out of coal. I have done the math and it should be working becuase its running on a pure vein and fully MK2 Belts and Lifts. Anyone that could help?
are they all filled with coal?
No the last 1 is some how not getting enough
You did a manifold. It won't fill unless the others are full. Are they full?
Turn off the first generator or two until the last generator is full or at least half full.
Thank you got it working
Anyone have any spreadsheets
for what?
