You Don't Want Precise Splitters in Satisfactory
HTML-код
- Опубликовано: 14 янв 2025
- One of the oldest requests for Satisfactory is a precise splitter or a ratio splitter. But in this video, I'm going to tell you that we don't need those, and that we can already solve all of our problems with the tools in the game.
Follow me on twitter: x.com/Satisfac...
Similarly, it blew my mind that coffeestain tried to make trucks only pick up the fuel they should need for a round trip of the route they are running. They will only ever *use* what they need and absolutely no more, so when would it ever matter how much they actually pick up as long as it's enough?
There is a case like this, when a truck takes all the fuel from the station and the next one is stalled, making a huge traffic.
@@nati0598that will get fixed by the time the trucks do the route a second time
@@captainblacktail8137 literally false. If there is a lack of fuel, that's because the input can't satisfy the demand. The next one is going to stall too...
@@captainblacktail8137what if you have more trucks than you actually need? Let's say the fuel can only support three trucks, but your items produced/required need four trucks.
@@zimzimph That's just a lack of fuel. Give the station more fuel or set up a station dedicated to only providing fuel. Simple as.
I never needed such splitter, but it would be nice to have a priority merger.
There is actually a way to do that with an industrial storage container that doesn't leak items from the low priority belt. ruclips.net/video/SchAH4nAhY8/видео.html
"There is a mod for that"
A perfect example of where a "counter splitter" would work.
Lets say you make turbo fuel.
You bring in canned fuel, unpackage it, remake it into turbo fuel then package it back into containers.
That means there are more cans than needed so you need to sink those.
Imagine how nice it would be to have a "counter splitter" that just straight up counts, "9 cans in this direction, 1 in that" and then loops the functionality.
And please don't use the argument "this can be done by doing x or y" because the same can be said about other things in the game, like the pipe valve for example.
@@unconnectedbedna how does this have anything to do with my comment? Also for that scenario you can just use smart splitter with a normal output for the packager and overflow for the sink.
@@K21040 So you couldn't hold yourself, you had to use whataboutism even though I pointed out it is invalid...
If what you type is true, the valve is also unnecessary because you can just use other techniques.
Or in your case, why would a "priority merger" be needed? You can achieve it with container rotating tricks instead!
It has to do with that what YOU request is ALSO not in the game, do I really have to point out the obvious?
And no, an overflow in this scenario would start throwing out to many cans if the turbo fuel packager were to pause for some reason. A "count splitter" would not, it would pause the packager from unpacking more fuel until the turbo fuel is starting to get packaged again.
Players:"We want to just plug in numbers."
Satisfactory: "Lol do math, nerd."
We would still need to calculate the ratios
Factorio is way "harder" yet gives player these quality of live features.
Using manifolds is not complicated math.
One big improvement you can make to a manifold is using different tiered belts.
For example, if you have a manifold receiving 300 items per minute, so you will need to use a Mk.4 or Mk.5 belt for the main belt. However, don't use this belt on the belt that goes into the production building. Instead use a Mk.1 or Mk.2 belt to limit the speed. It's not an exact rate limiter, but it gets you the right order of magnitude. Now you don't have to wait for the production building to saturate, but only need to wait for the lower tiered belts to saturate, which happens a lot faster.
I use different speed belts and sometimes feed the manifold from the center or even from both ends.
Very true! Good strategy.
@@stanleyclark923that's how I have been doing it as well. I seriously thought it was a no brainier. Generally due it making the most sense for me to not use the best belt for all machines if I want to guarantee the higher volume goes to the most machines. I also use a smart filter to go to the lower needed belt and prime those machines if I can, so the higher belt gets what is needed all the time. I remember doing this in update 4 on my coal belts and have done both ever since.
When I make a long manifold I use the lowest belt possible for every connection.
This works only if in the "main" belt are passing more items than the capacity of a Mk 1 belt and even in that case you may not achieve precise control over the flows. The only way to precisely split items is using a carefully tailored load balancer.
Honestly I feel like 99% of convery belt issues in the game would be solved if we could set machines to have a max self-inventory size, or just reduce it from one full stack to say 1x the amount it needs per cycle. If my machine only needs 6 steel beams per minute, I don't need it to fill up to 200 steel beams before it starts overflowing the belt. Just let it max at 6.
You do want to let the machine have at least a small buffer, but it doesn't need to be a full stack. When a machine finishes its cycle, it instantly consumes the resources used in the recipe, and checks whether it has enough buffered to start the next cycle. If it doesn't have enough, then it shuts down, and remains idle until the resources are inserted. It then "warms up" for a very brief period before it starts producing again. The end result is that your production would be way below the displayed numbers per minute if the machines had no excess buffer at all (they would run out of materials every cycle, and have to do the whole shut down/warm up routine before they could continue producing). The smallest buffer where this doesn't cause issues is twice the recipe's input requirement (which is still way smaller than a full stack, at least in most cases).
Captain of industry has this kind of system. The machine can store 1 stack as long as it is not connected to a belt. Then, it's 2 cycles of input products. It is very efficient.
@@scpWyatt
*_Factorio does this thing where it'll store two times the normal required amount for a recipe before it moves on. Like 6 Steel Beams would be buffered at 12._*
Just initialize the machine with a stack. This probably isn’t possible in the very late game with the rarest materials but throughout the rest of the game this should be very doable.
This would solve my gripe and excuse why I do load balancing overall, when later I want to delete 20 constuctors, I don't want to see floating boxes of steel beams :/ or one box 20 stacks of steel beams.
I am a programmer. You are way off-base about the difficulty of making a ratio splitter. You can implement them a lot like normal splitters. If the splitter is set to 2:1:1, then it just outputs in the pattern AABC on repeat. 5:3:7 can naively be done in the same way as AAAAABBBCCCCCCC, but you can do round-robin dropping the outputs that have reached their value to get a much more even result: ABCABCABCACACCC (2:1:1 just becomes ABCA, which is the same pattern starting at a different time)
"It can be done other ways" is not a very strong argument. Those other ways have other factors beyond "they get the right item amounts". I have *never* used load balancers outside of the basic 1:2, 1:3, 1:4, and even then only in my starter setups. Load balancers eat so much space that they have always been out of the question. I use manifolds because they're compact, but I do it begrudgingly since in their current form, they take *forever* to prime unless I'm manually inserting items into machine inputs, and I need the manifold to be primed before I can check my work and see if I built this one corner of the factory correctly. Similarly, while I have started to warm up to underclocking a little bit after 5 *years* of owning the game, but I still feel the lost space pretty intensely. I don't like to make flying buildings, so I'm often running into space constraints. And I don't tend to go power slug hunting much, so underclocked/overclocked pairs are especially obnoxious as solutions.
People want to have ratio splitters because we don't *enjoy* trying to use those other techniques.
A "heijunka" sequence solves the even balancing of something like a 5:3:7 ratio, and keeps the output rates as close as possible to what is desired rather than potentially sending to the same output several times in a row.
I also hate making floating buildings, but putting down some base foundations/pillars/walls can help a lot :) Eventually it becomes a bit necessary to build high above ground with the amount of space needed, and building this way makes it way easier to expand rather than having to constantly adapt to terrain. Construct, explore, exploit, and automate! ✅
There is a lot of "THIS IS HOW YOU ARE SUPPOSED TO PLAY" commentary in the Satisfactory community. Its a shame.
That thing you want? You dont actually want it.
@@ColossusBall So true. As soon as someone wishes QOL features, there is also someone jelling "Satisfactory is about exploration and building". For that reason we still have a too restricted and clunky blueprint system.
Well explained.
I got the perfect example for you.
Convert packaged fuel into packaged turbo fuel. You end up with more cans than needed.
Imagine how nice it would be to just program a splitter to let out exactly the % that needs to be sinked.
You have 2% too many cans? Set the splitter to send 98 in one direction and then 2 into the sink, done.
The argument that "other methods exist" is invalid. Should the pipe valve get removed because you can achieve the same functionality without it?
But I kinda do agree that maybe the game should keep every counter as "per minute", and a "counter splitter" would not work that way, so in a programming point of view, there might be other reasons NOT to introduce something like that.
As a LAZY programmer,
what this guy said
Yeah, I totally want to have to build 20 more splitters to get the ratios right for that recipe that has .125 in it instead of overclocking and slooping 60 other machines to output exactly what that one obscure recipe needs. I want precision, I don't want excess machines to do a simple job.
🤓
I haven't payed much attention to the community, but from everything I've seen when people talk about precise splitters, they mean ratio splitters. And people absolutely do want ratio splitters, which is what a load balancer is, just with out the complex mess of normal splitters.
I've seen plenty of both, but specifically there are many references to basically making it a valve for belts.
Even ratio splitters would take a lot of fun out of the game. Do some math or don’t and wait and sink the excess…
Ratio splitters are my dream since day one. Even as an advanced player, I still use multiple combinations of splitters and mergers to change the ratios. My factories look like a mess just because of this. There is no solution except filling up the machines with crap that could be making precious coupons.
@@FAQUERETERMAX Of course there's a solution! :) Breathe in, breathe out, let it go. Let it go. It's okay. Nobody is taking those tickets, they're yours, it just takes time. 300 foundation long belt backed up with turbo motors, it is okay. :D
But on a serious note smart splitters with overflow solve most of this. You can fill up one machine at a time in a manifold, about halving startup time and can sink anything that's not needed immediately at the end of the line. Constant flow, constant produciton - and the stack in the machines' input is a buffer ensuring constant production when you inevitably rearrange the belts.
@@WillyBotson That's essentially the difficulty of game design. Players tend to optimize the fun out of their game, and as a game developer, the hard part is to keep making it fun after the player optimize the fun out of existing gameplay
It's funny, I have played 100's of hours of Satisfactory and always felt it is missing something. I recently picked up Foundry and fell in love with it. It has whatever the hell is missing from Satisfactory for me. So now whenever I watch videos on Satisfactory it just makes me want to go play Foundry more. :P
I really like the modular buildings. Building a blast furnace should be a massive undertaking and only Foundry does that well. It does need a lore more content though.
A much better improvement would be to have belts not fill up machines all the way and instead limit the input slot to like 2-3 items. That's how Factorio does it and everyone uses manifolds (unless they're doing trains, then balancers to split between wagons).
As a bonus, it means much less inventory clutter when you deconstruct things.
A lot of other things would have to change for that to work. Some recipes require dozens of screws for example. If only 3 items can enter at a time then the recipe is literally impossible. Not saying it could not work this way, but the way machines work would need to fundamentally change.
@@SatisfactoryNews I think in Factorio items are only put in when there's less than 3 crafts worth of items in it. So for things that need thousands of items it inserts thousands, but for things that need just one it only has a few buffered.
@@15Redstones Yes, Factorio has the perfect solution for manifolds.
@@15Redstonesyes, having full stacks stored in each machine really sucks. Also when stack sizes are so widely different it can mess up a sushi belt after a truck or train trip. You get 100 rubber, then 50 computers, then 500 quickwire, etc. it’s not ideal. And there’s no easy way to limit a belt speed unless you use the native belt speeds and a system of mergers and splitters to throttle it. These things are necessary when there’s a truck or train or drone trip in between, when you can have a firehouse of materials hit the belt when it should be a trickle. I know I wouldn’t have this problem if I didn’t use sushi belts-but I want to save space and keep things simple-so this is a big deal. If you’re running separate belts, then this isn’t an issue outside of saturating manifolds.
My argument for a precise splitter is more about using more advanced technology unlocks to solve problems you were otherwise dealing with before. If I have to wait until I unlock programmable splitter or later to get a precise splitter so I don't have to worry about load balance or priming a manifold, that to me just feels like a reward for having learned and dealt with those problems up until now.
Yeah, it's not like the entire gameplay loop is built around this exact premise right? Nope, this one specific thing has to remain just as difficult for the entire game, because "its fun" for players other than you...
You're using early game things like steel beams as examples when you're talking about small numbers, but the endgame is where small numbers are important. Nuclear power for example. You aren't making fuel rods in bulk and precise delivery would be nicer than waiting *days* for buffers to fill so your power is all online. Basically anywhere you want a load balancer instead of a manifold and have a weird number of outputs, a ratio splitter would be beneficial.
In my previous videos about load balancers vs manifold, I mention that nuclear is maybe the only place where a load balancer is absolutely necessary (and it's the only place I use one in my playthrough). But, as always, you can get rid of any worries about splitting by dedicating a single fuel rod machine to a single nuclear power plant, if you're really concerned about that.
Direct connection of fuel rod production to reactors in small groups works well (i.e. not putting all your fuel rods on one belt). One uranium fuel rod manufacturer will supply either 2 reactors at 100% or 1 at 200%. The downside of this is if you want to have your fuel rods made somewhere other than where your reactors are.
Waste can benefit from some actual balancing. Using the standard recipes needs sending 3/4 to the blenders and 1/4 to the particle accelerators. That's a pretty trivial case to build though, you should be able to work that out by the time you reach the nuclear stage. Alternatively do your reactors in groups of 8 - connect 3 to each blender and 2 to the particle accelerator in a group recycling group.
You can make your own ratio splitter by chaining splitters and mergers. e.g. split into 3, merge 2 of the outputs, you now have 1/3 and 2/3. spilt, merge 2, split the 1/3, merge two, merge the two merged, you have 1/9 and (2/3+2/9)=8/9.
@@Derzull2468the issue is that that's SIGNIFICANTLY harder than just putting one splitter down and setting a ratio in it. it also eats up FAR more space.
i jusr throttle the urarium output and over/underclock to match consumption and put a 4to1 on the uranium waste...nukes are the only thing i don't just wing ( i tried but a few times when i got back from hunting/exploring i started glowing or had a plant go on and off 😂 fuel is more my jam anyhow 😂
0:32 - Implementation of a precise splitter shouldn't be difficult. Regular splitters follow an ABAB or ABCABC pattern; for a precise splitter, it could simply follow the defined pattern. For example, a 2:3:1 splitter could go AABBBC, or follow a more consistent "heijunka" sequence such as CBABAB. The ratios could be calculated based on the requested items speeds, i.e. a 80:120:40 splitter would translate to 2:3:1.
2:44 - This solution becomes less attractive when you have items going long distances. Say you have a 100/min belt that you want to split into 20, 50, and 30/min, while all still going to the same factory/area. Sure, you can underclock machines to produce at these ratios, but this means that you also now need three separate lines of item transfer. Therefore you would need triple-stacked conveyor belts or triple the train cars since you now need to keep these flow rates separated. It's much more clean and flexible to transfer one item stream and split it appropriately once you need to.
4:02 - This is a great point! I think I and many players often worry about 100% resource usage (not to be confused with 100% efficiency). There are in fact almost always more resource nodes, and if not, you can optionally overclock one of your existing miners and split off of that.
6:03 - I don't think precise splitters would make expansion more complicated -- we already have to re-balance and allocate resources each time we expand a factory. Having a precise splitter would only simply change this process from creating a new splitter array to just changing the numbers in the splitter settings.
All this said, I understand the devs' perspective; splitter arrays and ratio balancing is a huge part of this game, and adding a precise splitter would eliminate that puzzle element entirely. In general, I think the best course of action is only using manifolds near machines (so it doesn't look like factories are clogged up), and creating splitter arrays and sinking excess materials elsewhere.
If the game had precise splitters, I think I'd actually use manifolds. But since they don't, I use load balancers. I just hate stopped items on belts.
Yeah, but everyone here prefers to play a different way than you, so your opinion is wrong!
Thats what I'm gathering from reading comments anyway... its "just their opinion" though so... I guess that means you're wrong...
It would help fasten the "overflow" type of splitting, so it would be more of a QoL change than a game fixing one
An intelligent precise splitter shouldn't be too complicated to program.
I am using a MOD called "Throughput Counter and Limiter". Problem solved.
It adds a conveyor belt attachment that counts and displays the actual item throughput per minute and can optionally limit the number of items allowed per minute. It is unlocked via a Tier 2 Milestone in the HUB.
Loser
A percentile splitter would be enormously handy.
When would you use it where a Smart splitter wouldn't do the exact same thing just a few minutes later?
While I do agree that you do not need precise splitters at all, implementing them is not that big of a hassle. I actually got a piece if code running already that perfectly splits up values in any arbitrary ratios while basically shuffling the output orders in a consistent matter. Just using some modulo and gcd maths, it's pretty good.
Nice! I love that. Like I said, programming is beyond me, but I don't doubt that some version of this can exist.
You just count the number of items with modulo some number. And every time its 0 you do the lower ratio output.
How would that code work if one of the belt i backing up ?
@@clnetrooper probably just skip the belt and retry. If needed you could also keep a "debt" for each belt, so they get priority when space is available again. I just doubt this would be helpful though, because there are usually good reasons why a belt backs up if you already split the items perfectly.
@@clnetrooper i'd imagine that if the ratio was something like 4:1:1, then if the 4 output is backed up, the other two just does 1:1, which is an even split; if it's one of the ones backing up, it would do a 4:1 split on the other two.
I prefer the items on belts to never stop, only matters for visual aesthetic, therefore I would still want the precision splitter.
These things "can" be done technically but CSS prefers to make the toilet flush..
It's actually trivial to program, just alternate outputs as it currently does but have it stay on each output for some number of items.
I make all of my factories at 100 percent efficiency. Every building, even the miners. I have thought about this kind of thing before, and I still agree with what you said. The overclocking and underclocking help a lot. The only use I could have for this is when I need a weird number like 96 ( I needed this in a versitial framework factory the other day) and when I set a miner to make 96 items per minute, it becomes really difficult to is this node for anything else, because splitting 96 off from another number would be very difficult. That’s about the only use I would have for it, but I think that they could accomplish just about the same thing with variable speed belts. Say I set one belt out of a splitter to have a limit of 96 and the other to a limit of like 60 if I needed that for something. If the splitter did it have the total amount of material coming into it, it would work more like a normal splitter. Anyways that’s just my idea, it would be cool if they added variable speed belts in like caterium elctronocs in the mam with smart splitters and that stuff.
I build overflows out from miners so that spare minerals are sent to other floors. I do end up with a lot of extra stuff, but i'd rather have my factory slowly saturate across the board than to never have any extra i can grab.
Which measure of efficiency are you using as your primary goal?
There are over a dozen different types of "efficiency" in Satisfactory and many of them oppose each other.
@@5353Jumper the efficiency it says in the machine menu, if it’s on or idle
@CoasterCam-k3x that is a tough one for sure, gets so finicky late game.
@@5353Jumper definitely
I think what people want isn't a precise splitter, but a quantity filter or gate we can put on storage containers
Preface: I agree with you, it'll eventually even out, but I wanna answer this
1:00, programmer here. There are for sure 1000 ways to solve this, how I would approach is: define the output ratios, in this case 1:3 (30/30 : 90/30), for three output it could be 1:2:4, for example. Create a "bucket" for each output (note: does not withhold items, only counts them). Each time an item goes into the splitter, chose an empty bucket in a round-robin fashion and add one item into the bucket. Do this always skipping full buckets. When you receive an item and all buckets are full, empty them all and start over. Problem with this approach: towards the end, the 1 belt will receive no items, easy to solve: choose the bucket with the least % filled instead of round-robbin. Here's how'd it go:
0 (1) 0 (3)
Item Input
1 (1) 0 (3)
Item Input
1 (1) 1 (3)
Item Input
1 (1) 2 (3)
Item Input
1 (1) 3 (3)
RESET
0 (1) 0 (3)
Item Input
1 (1) 0 (3)
...
You get the idea
And not only underclocking and overclocking would be a possible way. You can also use different kind of Belts (Mk.1, Mk.2 etc. etc.), which have different kind of speed.
Yup I will typically use an MK3 or MK4 belt for many basic builds as their main input and output source lines that goes down the rows of splitters and mergers. Then I'll use either an MK1 or MK2 belt, lift coming out of the splitters going to the machines so that the first 2-3 machines don't hog the resources from the next say 5 - 9 machines on that line. As for the output, it doesn't really matter to much as it is based on how fast the machine can produce its given output. Now if you overclock your machines and they can exceed that of an MK1 belt then at least an MK2 belt for output to the merger would suffice, however I digress. I want my outputs to be at least MK3 or MK4 not because of the rate of what the belt can handle per say, but more or less in how fast the material can travel to its destination. So yeah, I'll typically use MK3,4,5 belts as the main belt going to the splitters and then typically use an MK1 or in some cases an MK2 coming off of those splitters, but as for the outputs feeding the mergers, I typically don't go any less than that of the main merged belt. If I'm using an MK4 then I'll either continue to use an MK4 or at least an MK3. It does kind of depend on the build or current factory that I'm working on. I rarely use belt balancers. I mostly use manifolds, under and overlocking measures.
I get around needing to split anything by just having dozens if not hundreds of belts of that item XD
"What do you mean there are containers? Don't you see this "network" of belts going back and forth, easily 1000 items on each of them..." xD
There have been multiple mods that have done this, and they can be quite satisfying, also. Just because you don't see their value doesn't mean they shouldn't exist. Heck, there's someone that's trying a biomass-only run, is that wrong because there's other tools in the game to do better?
There is a difficult balance between "feature that solves a problem and makes the game more fun" and "feature that optimizes the fun away".
I think a precise splitter strays too close to the second one. It doesn't solve a major pain-point of the game (at least not for the majority of players) and actually makes it so that a major part of the game (solving the puzzle of how to distribute your resources) is optimized out. Why would I ever try to do math and balance my factories properly if I can just determine an exact number to go through a splitter?
Contrast this with the new Dimensional Depot. At first I thought it might fall into the category of optimizing the fun away. But then I heard Snutt describe it a solving the pain point of running back and forth between your factories when you run out of materials. Which makes sense, that is a common annoyance in the game that every player experiences, and it interrupts the flow of normal gameplay.
I'm not sure where the comment about biomass came from. I don't care how people play the game. This is a game design question, not a question about imposing limitations on yourself (I love doing playthroughs with those, btw).
@@SatisfactoryNews "I don't care how people play the game."
Then why do you care if people have access to a such a feature? If you actually didn't care there would be no problem right? I doesn't affect you in any way. Nothing would keep you from playing the way you do now just because there is an additional splitter in the game.
@@Allenschezar4890 People can play the game however they like. It's not illegal, and I'm not upset if someone plays differently than me. There are thousands of ways to play this game. I also don't care if people use a mod to achieve something they desire. In fact, I don't think in this video I ever said that I care that people are playing with a precise splitter, or that I am somehow affected by it. This video is certainly my opinion, but it's not an opinion piece necessarily. It doesn't have anything to do with the way I play. This video is from the perspective of game design and I repeatedly say that you don't need a precise splitter and I stand by the fact that having one is poor game design and optimizing the fun out of the game. My only opinion here is that I think that if you're using a precise splitter you're optimizing away the problem-solving aspects of the game, which is a core component of what makes the gameplay compelling and keeps you sucked in. I'm sorry if this video made you feel attacked or like it keeps me up at night knowing people are playing with mods, but neither of those things was my intention.
@@SatisfactoryNews The thing is that people looking for said feature are not enjoying the game the same way you do.
To you said act of "balancing" may be a fun factor or even, as you say, a core aspect. To others it is not. And that is i think the problem with doing videos like that.
You may have felt like it was necessary to explain why one does not need this feature but the issue to someone advocating for it is that this just comes off as you telling them they enjoy the game wrong even if that was not your intention at all (and honestly the title and Thumbnail don't really help the matter either).
@@SatisfactoryNews Your stating your opinion as a fact that applies to everyone. For you the puzzle solving is compelling, but that may not be the case for everyone. You state it's a fact that having one would be poor game design, which again is actually another opinion. I'm certainly not kept sucked in by the annoyance of having to balance things.
Advanced logistics mod already handles all of this perfectly by specifying the ratios. I don't use the splitter much, but i love the merger.
It would be better for clean builds and could make load balancing have the structure of a manifold. Perfect efficiency without the startup time
It's actually not that hard. a phone game about factories actually did it, except instead of doing items per minute, you had a fixed count it reached until it swapped. This means it will go in a direction as long as it takes to reach the count meaning left in scenario 0:55 could take 30 seconds and right could take a minute and a half to reach. This means it falls onto the player to further refine the output and make this work. In the game i played, there was no max amount a conveyor could handle, so i actually found myself in need of this type of splitter often since in that game space was a big premium. In satisfactory space isn't nearly as expensive to have, so it's not nearly as big of a problem using multiple conveyors instead of one super conveyor that splits on specific time to allow for cheaper "smart splitters" which is essentially how i used count splitters.
Isnt the only reason you really need to worry about precise splitting is belt efficiency? In the early game, time is basically your biggest obstacle. You can't mass produce the resources you need to make more efficient technology, so belting your machines efficiently cuts down that time immensely. But once you have belts that can handle more resources than your machines need per minute, you can eventually outgrow your production until you need resources elsewhere.
Rather than a precise splitter what if we could get precise conveyor belts instead? A smart belt where you can lower the speed of it to be exact in how many items it can transport every minute
What you mention is a fair point about under/overclocking the machines for rate limiting. But there is also another way to do it that’s closer to exact rate splitting. You can use smart splitter with different belt rates. So you can think of it like extracting resources off the belt at whatever rate the belt will allow. For example if you need 30 of something you put a smart splitter with 60. (Mk1 belt) where the overflow is the throughput line. Then you just divide the mk1 belt by 2 and merge the second half back into the overflow line. Make it into a BP and you can cover all the common numbers pretty easily (like 15, 20, 30, etc…)
In terms of computer science based on containers and their respective algorithms, we can think of various types such as vector arrays, lists, stacks, queues, etc. The actual algorithms that can be performed on them themselves aren't the main focal point. It pertains more towards the nature of the data structure itself. We only really need to consider one specific type of container, dispenser and that is the priority queue. Not exactly the same because within programming, a priority queue is basically a list of elements where each element has a specific priority value and depending on implementation but typically those with the highest priority are sent first. We can use this as soft of an analogy with smart or programmable splitters and by setting specific outputs to have priority. With that we can direct the flow of material based on priority. Outside of computer science itself but more towards hardware engineering, this is akin to the properties and uses of Decoders, Multiplexers, Demultiplexers, and comparators. We can determine which bit line or bus becomes the active input or output based on a control signal. The only difference here within Satisfactory that I think is missing that if it was present would allow Satisfactory to be considered Turing Complete is the ability to toggle or control these priorities from an outside source. We can set them during construction within their properties, but these are predetermined default fixed values. This is more comparable to ROM than it is RAM. We do not have a mechanism to change this dynamically from some kind of logic distribution or control center. For me, it's not really a necessity or must have for this to be a complete polished and enjoyable game. Yet I do feel that either adding or having this ability would change the entire playing field opening up a lot more creativity and flexibility throughout the community and all of their amazing builds and creations. It would really be interesting to see such a mechanism implemented. I'm not necessarily referring to controlling exactly how many items can go in or out of a splitter or merger. By design and intent, they take on the rate of the belt that is attached to its inputs and outputs respectively. I don't mind this and I don't think this needs to be changed. However, when using Smart or Programmable Splitters and possibly Smart or Programmable Mergers while already being able to set a specific lane as a priority, another as an arbitrary or don't care output, and the last as an overflow is a static system. With adding in the extra capabilities to change these from elsewhere remotely should in theory allow pioneers / engineers to build a remote-control center, a logic distribution center to where they can turn on and off various factories or supply lines instead of having to travel to that factory directly and changing each device manually. I think this would be more useful, beneficial and practical. I don't need exactly 1 item to go left while 99 go right. What I do need is to control when or not this line pushes through or not depending on on-demand needs and requirements.
I can't say I understood all that, but I do appreciate the thought that went into your comment. I'm just a dumb marketing guy, don't know shit about programming.
@@SatisfactoryNews It's all good. I just threw it out there for any who wants to read it, maybe they can take or get something out of it.
Great read. If this game becomes Turing complete, someone will make a simple factory emulating CPU
Ill always recommend building for your current miner tier. Because as soon as you unlock a new one you should have gained considerable experience making any new factory far better than the old ones
I feel like the "how should it work" would be that it effectively counts the items coming in and then sorts them in a given direction based on where it is in its count. As an example, if the splitter is set to 10-20-30, then it would do 1 on the left, then it would alternate middle and right along the lines of m-r-m-r, then repeat the right output before splitting the next to the left. The whole sequence would go something like l-m-r-m-r-r-l...
My opinion on ratios on programmable splitters is that it gives them a reason to exist. For a lot of people, sushi belts are already just not something that's used very often, and it's even less common that you need to split one sushi belt into two or three outside of big central storages (which I personally never used in early access despite beating it, and had never considered since it never seemed worth the hassle). As such, programmable splitters are basically just more expensive smart splitters to me, and I already basically only use smart splitters for overflow purposes and extremely small hand fed setups like converting plants into biomass or remains into protein.
As such, being able to program ratios into a splitter would be something I could make use of, rather than it just being an additional buildable I cycle through when placing down splitters and mergers.
Precise Splitters might not be a good idea; but what about belt limiters?
The splitter does not need to care about ratios. We could limit it by effective belt speed.
Actually underclocking all the belts would be a PITA too, but a simple splitter-format building with just one input and one output and a slider for max items/min could solve quite a lot of problems some people would want to use precise splitters for.
Precise belts is a "better" solution but both solutions would be a nightmare to adjust if you change your factory, haha
For my vanilla playthroughs, I use both load balancing and manifolds depending on the situation. Any excess resources go straight into a sink, which is why there is always at least one sink in every factory.
In my modded playthroughs, I use a mod which adds in ratio splitters for two key reasons, the first being that it keeps everything nice and neat. Sure I could split easily into 16 by having 4 rows of splitters going into 2 outputs per splitter, or I could just have one line with the ratio splitters with ratios such as 15:1, 14:1 etc. Keeping everything looking nice and tidy. The second reason is that it stops me from going overboard with production setups, as sometimes you'll need an obscure number, and rather than overproducing something to reduce one or two of the outputs later on, I can just ratio where I need to.
Overall, the most important thing is that everyone can play this game how they want. If you don't want ratio splitters, don't download the mod. If you do, add the mod to your game. So long as you're having fun with the game, it doesn't really matter. As I've mentioned, I do both modded and unmodded, so I know that things can be done without ratio splitters, I just like to keep my factories looking clean and tidy. Again though, it's up to you how you choose to play the game, choose what's best for you and go with it.
I don't worry about perfect splits or 100% efficiency. I tend to build more machines than I actually have materials to feed them with. This way the source node can be maxed out, but the downstream production machines are never maxed out under any circumstances. This way, I will always have excess production capacity, but my power demand tends to fluctuate all the time.
I think the most important argument against ratio splitters in the base game is the fact that you can build it yourself. Honestly its quite an interesting challenge to make a blueprint for some weird splitting ratios like 5 or 7, and then even cooler would be to make it an adaptive one (i.e. a splitter with 16 output to which you connect howewer belts you want and it splits evenly). So yeah, adding ratio splitters is just a shortcut that you can make yourself already
I'd like to see you make a 1 to 37 splitter. Because that's the usecase that ratio splitters would be realistically for.
I mean it's gonna be bulky, but still not very hard to design. And I imagine such weird splitter you are only going to need once so just make it. I can eyeball that it will fit in 4x4 space and that is not a lot when you are dealing with such numbers.
Also duh, use an overflow gate
What I did was feed multiple items onto 1 belt (8-10 different items) so it's 1 line going in, a manifold smart splitter system that will filter items into the production, an overflow system for the smart splitters at the production line, and then the overflow system takes all secondary items to storage. once it hits storage I did the same smart splitter manifold system with overflow, and if it fills up at the storage it gets sinked off.
Production > storage > sink . this enabled 100% of my machines to always be producing, the only shortfalls are supply (which i can increase), none of the lines are EVER backed up, and i generate tickets like crazy. it's all about how it's managed, but a stopped belt is a non-producing belt.
Is this really that different than setting the input of a constructor to way above the true input?
The splitter will split as if it were getting 120/minute, which would be equivalent to the ratio, like you said.
Yeah, but that's a solution, and the problem is everyone here prefers you to play the game their way, and when you really think about it... you dont want to play differently than them right... right?!
I would like two things to help with "debugging" issues, that comes from merge and splitting belts. 1 a little gate you can set over the conveyour to monitor the actual throughput on the belt, like the valve can monitor for you. This will help greatly when debugging a factory. And a small change to the valve, make it show the average of the past minute instead of only actual numbers, they almost constantly change making it less usefull to monitor :)
devs probably think they're smart bc they made a problem out of nothing and now we have to deal with it
The best part of the game is solving problems. IMO, a precise/ratio splitter goes too far to optimizing problem-solving out of the game without addressing a major pain point unlike the depot and zooping do.
@@SatisfactoryNews this is just the problem I don't want to solve, bc the solution is flawed either way, not satisfactory
you either have clunky monstruous splitter/merger setups for balancers
or you have the drawbacks of manifolds
the perfect solution is to have precise splitters!
and its not like we don't have to solve a problem when setting up a precise splitter, we still need to calculate the ratio
@@SatisfactoryNews another thing is, for the sake of problem being artificial, it could exist and be not perfectly solvable until you get to late game tiers, like many people suggest that precise splitter could be the programmable one
So the problem exists until you unlock it, and by the time it WILL feel so satisfying replacing manifolds/balancers with programmable(precise) splitters!
First time I unlocked them I thought THIS what they were supposed to do, not be a glorified smart splitter that has a slim use cases
I mean... its not like they have ever added features to the game that solve problems they created though... right *ahem* (conveyor lifts) *ahem* so why would they ever add features that solve the problems they created?
@@SatisfactoryNewsBut there is no "problem" that ratio splitters would "solve". For advanced builds, everybody is using manifolds right now, because load balancing something like 23.154 steel beams just is not going to happen, it's effectively not a factor. By adding ratio splitters, no challenge would be lost since absolutely nobody is engaging with the challenge in the first place.
Ratio splitter makes perfecf sense to me. The current splitters already work this way with the ratio just being 50/50 or 1/3 each. Just allow us to choose the ratio of the outputs. Would be cool if they just added it to the programmable splitter to give it a bit more use.
I would want a dynamic entrance splitter, so you can change output to input gate
In terms of programming, implementing a ratio splitter actually doesn't differ that much from a normal splitter, but you're right with stating that we don't need and probably also don't want them :)
(Normal splitters clearly store data to check what output was used last time and use the other one, so for ratio splitters, we'd just have to store some more data and round here and there - not difficult, but yeah, also not practical)
Bit late to the party on this one, but when I first heard "precise splitters" I kinda just pictured something like a regular smart splitter, just with an output you can set to a priority value -- let's say 24/min -- and then an "overflow" line (or if there's two "overflow" outputs then split them evenly like regular splitter outputs). Then, if you have an input of 30/min, you'll get your 24/min belt and a 6/min overflow belt. And of course you can add more input to increase the overflow line, or if you wind up decreasing the input below 24 then 100% of the input will still go to the priority line.
After having read some of the comments though, I do also agree with the idea of a ratio splitter. Either way, I really don't think it takes away from the idea of "do the math," it just leads to less annoying situations that REQUIRE either more machines to underclock/overclock, which isn't "fun" or "do the math," it's just annoying; same with the idea of having 20 splitters to get the right number, multiplied by the number of times I need that number in different areas.
Realistically, yes, it doesn't come up often, and ultimately it's not a huge deal, most recipes usually have a nice ratio where just adding more input can keep the numbers clean. But it's weird to see so many quality of life things get dismissed in this community just cause the masochists want to keep things needlessly complex. I'm convinced if the game didn't already have the smart/programmable splitter then the community would be opposed to it being added simply on the basis of "just add more belts and keep the items separate" or "do the math and figure out the overflow yourself"
So, my reason for wanting a splitter with output control is more that right now the programmable splitter is the most worthless buildable in the game. They literally are just super expensive smart splitters
Yes! I made this *exact* argument on reddit like 2 years ago, and the entire comment section basically was this video telling me there's already ways to do what I want in game, and part of the fun was making those work for me and blah blah blah, all ehile ignoring that the programmable splitter is a useless feature that can be achieved with a row of smart splitters and mergers, which is just... their argument turned around on them, but *they* have the programmable splitter, whereas those of us that want something else that may be useless *to them* are told it's just... not really necessary and we don't want it anyway rightm?
my issue is you're looking at it too small, if I need 512.7 going one way, and 207.3 shipped of to a different factory, building more machines will take up too much space and a manifold will never fill enough to be efficient
Are you talking parts or ore? I'm trying to grasp where over and under clocking wouldn't immediately resolve this? Is it like when you're trying to get a factory started?
Your numbers end in a decimal. You need two extra machines at most? One to solve each decimal. Or overclock a touch and you can use a few less machines. Clocking literally is the percentile/decimal splitter everyone is asking for. The game is based on real mechanics and physics. A splitter no matter how smart can predict flow. Flow can be controlled…with your machines.
Thank you... oh my lord THANK YOU. Now I can link this every single time someone keeps asking and bringing up their 'own idea' of such a setup. As if typing this one such request came up not 5 mins ago on the qna site. What I find mind numbing is people cant grasp is its not just 'send X amount of items' you're working in items per minute, it needs a constant flow just as the default game gives us already without ever needing to tell a splitter such. As is, once I got to the stage of the game of unlocking 480 belts and such, I used smart splitters to future prof my builds. As in I start building and calculating all my production lines to account for when I unlock 780 belts and Mk 3 miners. And I use a smart splitter at the section of this production line that could account for current 480 flow rates set to overflow, so resources wont pass it and try to fill machines that the belt speed don't yet support. This way when I go unlock 780 belts all the machine are already in place and resources will flow fully just by upgrading belts and miners.
1:22 you can make it put an amount of items to 1 side and then an amount of items tot he other side (example: it puts first 2 items to the left, and then 3 items to the right as soon as they enter it)
yeah it can literally just go A B B B A B B B if you set it to 1:3, it's literally the simplest implementation of this, tho as a factorio player I don't see much use for this anyway becuase like said in the video it will fill up eventually if you provide more than you use and it won't use extra resources
@panzervpl9406 well if you do your math, then they won't really fill up, but yeah, now that I have more experience with building factories in this game I realise they are kinda useless (except for people who like to make everything absolutely perfect, or those who don't like waiting until the factory spins up)
The way to implement this is obvious. Right now splitters have a certain order. Left, middle, right, left, etc. All you need is the ability to edit that order. So I could make a splitter go left, left, middle, left, right, repeat, if I wanted. That would not only solve the issue of the splitter not knowing the future, but it would be a more interesting and hands-on approach that would be more fun to work out. Of course, it's still not really needed.
It's not very hard to code it at all. I don't quite understand your point about the splitter not knowing what is coming. Does it have to? It would just be a dumb counter.
This is how I'd implement it:
Each output would default to 1:1:1, which is literally how it works now.
You'd be able to edit each port by setting a value from 0-5. Maybe even higher, idk. This value would be the number of items it allows through before switching to the next output.
_What_ the items on the belt are would not be relevant. It could be a sushi belt for all it knows.
As far as smart splitters go, it would work like this:
Each output would default to 1:1:1, as it is now.
They could potentially be edited to the same values of that of the normal splitter. However, you actually wouldn't be able to edit the values from 1 _unless_ you have 2 or more outputs set to the same item. And once you do set it, it will only count that particular item when it passes through, ignoring anything else being sent through the remaining output.
You have to understand, this isn't about what is _needed_ to solve a problem. It's about adding another avenue. Giving you something new to experiment with.
Providing other ways to avoid having to use a ratio splitter is not a valid argument against having a ratio splitter. Otherwise you could just as easily say "Why do we even need trains when vehicles and belts work just fine?"
Oh, they understand, adding the option though, would mean they either have to do things the inefficient way, the more time intensive way, or a way that cuts all that out, and since they don't have fun that way (I mean, outside of things like conveyor lifts, floor holes, trains, drones, etc.) you'd be forcing them to play the game in a way that doesn't require them to spend the time to calculate how they split things, and remerge them, so they don't want the temptation!
Ummmm. How do you think splitters work currently? You don't need to read the future to alternate outputs. Nor do you need to read the future to alternate outputs but swap after 3 then 1 then 3 then 1. Trivial programming really. This is the kind of thing you learn in the first week.
Counterpoint: Biofuel.
These are limited in the sense that you only get more as you collect more, so I don't want to just wait for a manifold to fill up, because there's not an unlimited amount coming in. So a system where I can say... have 1 in 100 remain as solid biofuel for my chainsaw, while 99% of it goes towards liquid biofuel for my jetpack is actually a good idea. I could build this with a bunch of splitters and mergers, but a precise splitter would do it so much better.
Hmm, true, but I guess that begs the question of, if you only get biofuel when you collect it, does your factory really need to be perfectly balanced if it's already not running all the time?
@@SatisfactoryNews It's because it's not running all the time that I need that balance.
If I'm fresh out of solid and liquid biofuel, when I dump my wood/leaves/etc into a sorting bin, without selective sorting, about 50% of that will go into the bin for chainsaw fuel. Which is really unnecessary. I really only need like 1% of that to remain as solid biofuel for the chainsaw.
Sure, I could do a more complex splitter setup, or I could create separate drop off points for chainsaw fuel and jetpack fuel, or go pull a bit of solid biofuel out from the process and keep it aside, but this is precisely the point of a quality of life improvements. To allow us to make things the way we want. A precise splitter means I can just dump everything in one box, and just the right amount will be siphoned off for chainsaw fuel.
the same applies to animal protein, how much will go to DNA, to biomass, or to inhalers. It's these limited resources that truly need it.
I agree with the priority merger person in the comments. But to speak to the ratio splitter idea: this item already exists in a mod called "Throughput Counter and Limiter." It works pretty well, too. Has over 2.5M views and 114K downloads as of this comment. You use splitters for each machine in a manifold setup as usual, and before each item input, attach a throughput limiter. Set the items / min, and forget it. When you need to update it, update it. I used it for my computer factory, and it worked pretty well (default recipe). It's a handy item for speeding up manifold saturation time. Because of that mod, (in update 8) I no longer had any issues with manifold setups, and 100% prefer them to load balancers. You say it's probably difficult to implement, but one person already did it.
About the only times I really wanted to be precise with inputs has usually been regarding power systems. Waiting for coal plant #1 to fill up when I need the power from #2 and #3 to keep the factory running in order to give the plants time to fill up. There was a pre-1.0 mod I used that allowed it without much issue as long as you remembered to set it right. If I had 5 plants I needed running, then splitter one was set to 1:4, the next was set to 1:3, then 1:2, then 1:1. This let me load balance on a single feed line without going through the mathematical logistics or needed a 3x5 foundation space to work the load balancing.
Outside of power plants, I'm fine with overflow manifolds as I usually make mini-factories bent toward producing a specific output.
This is why I only use manifolds for all my production lines. It takes less time to set up, takes up less space, and all the machines work at 100% just fine once the belts get saturated. In the meantime while I wait for production to get to 100%, I can add details to the build or continue the factory if it isn't complete yet. Also I turn on parts of the factory that I just completed and then they are ready by the time I finish the next part of production.
I don't think "You can do thing but in harder way" is a good argument against a feature requested for convenience. Literally just have an interface like smart splitters except it displays numbers. "For every one item that goes through this output, put three through this one." Let's say from left to right, you input 2, 0, 3. Every time it puts 2 items through the left output, it will send 3 through the right output, and never touch the center output. Or maybe even set the center output to "Overflow".
Yeah, but that's the only argument you'll ever get, that or "there's a mod for that" never-ending whether you like mods in satisfactory or not, either play the game the way they like, saturate the belt all to hell, or build huge load balancers.
I've been arguing for an actually programmable splitter for years at this point and I always get the same replies no matter how many times I shout that "the hard way" isn't a sufficient argument for withholding a feature.
I think it would be cool to have some 2 story splitters and merges that allow for up to 7 splits/merges. Adding a few more primes and starting with 2,3,5,7 would be super nice for setting up blueprints, and I think it's much more in line with what the game is trying to do.
And because they are 2 stories high, they still have a trade off which means they wouldn't just immediately replace all of your current splitters/mergers as a pure upgrade (plus I'd assume they would require more advanced parts)
In my builds I mostly use manifolds. There's only a few situations or occasions where manifolds aren't the better option. The only two cases that I can think of is with specific aluminum builds where you do have to manage or handle byproducts and more importantly in nuclear builds, where saturated belts with radioactive materials leak into the environment radiating everything. Other than that, manifolds are easier, faster to setup, require less trouble shooting and are very modular, very repeatable and scalable. They also require a significantly smaller footprint in terms of area helping to make your factories more compact. I would be more concerned with how this game handles fluid flow with its bidirectional implementation and requirements of head lift in conjunction with backflow and sloshing than I would with belt balancing. Now there is a third technique that is soft of a hybrid of the two. It is still a manifold at its core where you do have a main source line that is being divided by splitters and partially down the line just add in a merger to that main line before the next series of splitters. You may have to either introduce another supply line coming in from the outside world, or you could take the end of the line and feed it back into the merger for any materials that went past the line. These hybrid types can help to reduce or minimize the accordion trickle effect that basic manifolds exhibit. I also incorporate buffer storages before and after each production line. This way if there happens to be less of a demand for a given product, the pre storage buffer will back fill and continue to do so all the way back to the original source even back to the miners themselves. When the demand picks up and the storage buffers are full for that line, there's less of a need to wait for any previous or intermediate staged production times. The materials are already there waiting to be used. It's sort of the same kind of concept when it comes to train stations with their cargo loaders / unloaders and having a buffer storage before and after it so that the loading and unloading of the materials appears to be seamless. Cool video by the way!
First and foremost, solving problems like this inside the game is what makes it fun for me! I'm a proud nerd that owns it, I like solving weird and wonderful problems like this!
The thing that popped into my head around 1:20 is that you just described what a fluid valve does. As a programmer, this should be easy to implement (the integration tests would be a bit of a PITA tho. My point is that it "may be easy/simple to program into the game", but the ticket schedule is full until Q3 so it's not "that" easy). As a project manager, my last statement made me physically wince! Lol
That said, I think the game is better without a "stupid-smart splitter". Let the belt overload, the machines on that line will run at 100% whether it's the 24/min they need or 100/min are feeding into that line.
The system I have theorized and want to try is using a manifold with say a 480 belt, but each output belt is only a mark 1 belt. So if a construcor takes 60 items per minute the mark 1 belts will automatically lower each constrctor to 60, so everything is perfectly load balanced, but it's designed like a manifold taking up less space. And even if it's like 45 the mark 1 belts will still help reduce the wait (or for 30 you could do a manifold into a 2 way splitter)
Honestly implementing a ratio splitter in a useful way would be super straight forward. Don't try to use actual item/min values, but just attach a value to each output which determines how many items are sent out that output before moving on to the next one
Industrial storage can having dual in and dual out is really useful for levelling supply/demand.
I understand what you say but for a lot of end game recipes it’s not as easy as you make it out to be. But after having played with a mod that ads precise splitters i can say it’s handy in some cases but not necessary most of the time. Like you said just let it run and if you really want a machine to run more turn of the other machine until it’s filled up. Btw the way the mod did precise splitters is you could specify a number on each side and it would split that exact amount of items then move to the next. For example if you had 60 coming in then set it to split 90 and 30 it would first put 90 items on the first belt then 30 once it was done with the 90 which worked fine in my opinion.
Yeah that is my assumption of how it would work other than doing ratios. Which presents its own problems that you have to account for especially if there is a deficiency of resources.
Definitely seems like a mod worth trying out! That implementation seems really ineffective though, because then you're getting huge batches at a time. Something like a "heijunka" sequence would work way better, e.g. a 2:3:1 splitter could follow a CBABAB sequence and make the output rates as consistent as possible.
@@zain374 that is true but the mod allows for big numbers and splitting that way doesn’t work if you have numbers like 32/45/98 although i myself never really used it like that I presume it’s the main reason it’s splits like it does.
The reason it would be immensely useful to have this is so that you can set up manifold style splitters and not have to wait for the first machines in the line to "fill up" before the remaining machines get product. You could just set the splitter to output what its machine needs and leave it at that. It's really not any more complicated than this.
This video summed up "it's not a problem for me so it's not a problem for you"
It never occurred to me I didn't need to use precise splitters. I waste so much space and mental effort splitting and merging lines to get the exact right ratios (or as close as reasonable) and then I saw your example of a filled up line and the rest going where it needs to and I feel so dumb lol. Honestly at this point the time saving of not needing one line to fill up is completely offset by the time I spend setting up significantly more complex lines.
OK…so what you argue is great…until I moved products by truck or train to a second factory that I want to get items at a particular rate…like especially when using a sushi belt. You can use buffers, sure, but you may need that item in multiple disparate places. Soooo…I could use a bit more precision in some of my setups. Especially for small numbers of things…like electronic components.
No. Ficsit does not waste. Splitters just alternate between outputs round-robin, like dealing cards. A rate limiting splitter would work by simply "dealing" more than one item to a given output port. Instead of giving 33% to each by counting 1 2 3, counting 1 1 2 3 would mean that the outputs are now split 50%/25%/25%. It would be so easy to implement, and it would make load balancers more compact. You could get a ratio of 1 to 5 with only two splitters and without needing a loopback, for instance - just count 1 1 1 2 3 and split 1 three ways. Boom.
Agreed, it would be easy to implement definitely make these systems way easier, but I think they *want* that part of the game to be a bit complex. When you're not using manifold, splitter arrays are a big puzzle element of the game, and having to take up more space for more complicated things is kinda the whole flow of this game.
I like to use manifolds. I always make sure the first machine loaded is the last to unload on the exiting manifold. Helps keep all the machines buffered withhold the last machine starving for materials
1:00 this is the way it would do it, it can’t predict the future but it can know the past. It simply takes a count of what it has already sent through and it sends the next item accordingly. Say it knows it already has sent three through the right so it will know to send the next one to the left. That’s how it would work.
I think I would want a priority splitter. That's the only kind of splitter I'd want more of
I think maybe adding the option on advanced splitters not to set a ratio but limit the max output on one side to specific numbers maximum could be nice to have for aesthetic purposes
What we actually need is the ability for trains to recalculate their route so we can have parallel queues of trains
I'm on the fence about this because people who actually want to use items like this, will actually use them and apperciate the addition. Also, they wouldn't be difficult to implement, just difficult to learn to use optimally.
However, what makes me sit on the fence about it is that I feel like it would be overwhelming or daunting to a lot of people. It would probably create a situation for a lot of people where they feel like they have to or should be using them.
I think this is something better left to the hands of modders. Those interested enough to use them will most likely be willing to mod their game to get the desired effect, while keeping more casual players from feeling like they're not playing as efficiently as they should be or becoming overwhelmed.
Wild. Your entire argument is "We can do what you want, less efficiently, eating up either drastically more space and power, or more finite items". That is a nonsensical argument, to me.
"go use another node" is also comical. Go make a new setup, that requires more resources, and more power.
Further, this wouldn't need to be a "new" item. Make it part of programmable splitters. Make it part of the filter. It would be trivial.
Dimensional storage just makes this argument even weaker, now. It makes all the sense in the world to pull a smaller percentage out of each end product run and shift it over to the dimensional depots.
The only real thing here is the devs don't want it. Which is a mistake, to my mind, but it's there game and it's not the end of the world.
At the end of the day, I still think it optimizes the fun out of the game and helps you ignore an important problem-solving part of the core gameplay loop, and I know that opinion is shared by the devs. But either way, there are mods for this, so you can still have it if you absolutely want it.
@@SatisfactoryNews it's your condescending attitude that made me feel like making this comment. "You don't actually want this, I know better than you, you're opinions are wrong" just ... Why did you need to take that track?
Yeah, it's bizarre to me, that "optimizing the fun out of the game" wasn't an argument that was made when conveyor lifts were added, or the floor holes allowed for not clipping through floors or forcing you to run belts and lifts outside the factory or up through the middle.
I've been barking up this tree for 2 years now, and keep getting the same arguments back, all of which amount to a condescending "play how i play and shut up."
Personally i'd like to have something to regulate the exact rate on a conveyor belt, mostly because it would help understanding how a system it supposed to operate if you ever run into problems, but i understand that other people disagree. What i really would like to havr though is a statistics interface for splitters and mergers, or at least for smart and/or programmable splitters. This would tell you the average rate over the last x time and maybe even show a graph. It would help you optimise factories because you can see how much of something you're actually getting
Im always down for more content and options. I might be to lazy to use it though. Im one of thise players who oversizes most segments by a bit and then use buffers. Then its all manafolds and go do something else while it fills up
I built a sorting system that could benefit from splitters that I could program in the amount
Yeah, but thats obviously trying to have fun that the game doesn't want you to have so... build it the way other people say you should. /s
While i do agree that we have ways to already do this ingame and this feature is more of a "shortcut", i do want to defend how it can be implemented (as an actuall programmer).
As you mentioned earlier, the spliter goes in a round robin sequence (A->B->C->repeat). If an output is blocked by items, the output is skipped, so if B is clogged, it will skip it and check C, which isn't cloged, so, output there.
Now, my solution wouldn't really redo what is already made, this system works and there are also propably events that the belts call when they free up, and the end of the algorythm is perfect. What i instead suggest is changing the begining, which the end can handle.
Each output is set to a certain intager. Lets say A = 1, B = 0 and C = 2. From this, we can now construct a sequence: A C C. Or A = 3, B = 1 and C = 1 to AAABC. The algorythm will now use this new sequence to iterate over, which in theory, should make the spliter behave as desired.
It works for any input rate, as it is based on item handeling, doesn't change the mechanic and would be fairly easy to use.
I started playing Satisfactory a few weeks ago and a programmable splitter is not something that ever occurred to me. What I instead found lacking in the game was a more advanced calculator and a better note-taking system to keep track of all the math and project requirements. There’s plenty of 3rd party apps, but I still find myself writing down basic schematics on paper. Also, being able to re-visit saved notes would be immensely helpful when updating/rebuilding a factory.
Have you discovered the note pad yet? When youre in any menu, click on the right side of the screen. It's very basic but it does exist.
@@SatisfactoryNews I did and that’s what I was referring to. The character limit is quite small but I can see what the developers might be going for - keeping players from drowning themselves in too many self-imposed tasks. What I would prefer is a system that allows me to save each set of notes in its own file (kind of like blueprints) that I can access later or delete if redundant. Recently I have started using Obsidian which supports Markdown editing so I can have my factory notes and calculations saved in one place. The obvious downside is having to Alt-Tab every time I have to use it (or buy a second monitor or use my phone).
Satisfactory doesn’t need as complicated of a system, but having players leave the game to access 3rd party tools to set up a factory defeats the purpose of being a complex and cerebral game. I enjoy the complexity but it becomes overwhelming without the necessary tools to navigate it. Especially as a new player.
Ficsit does not waste! 4:29
And now you need a precise splitter for iron bars...
This is just about the dumbest most condisending videos I've ever seen.
I'm currently making a max coal factory (12000MW)
But I'm scared of the coal not arriving to the last machine due needing to be a manifold system. 600 coal going into 40 machines on a row where they consume 15 each should mathematically work.
It should work but it will definitely take a while. I would recommend starting the miners and the water extractors, but DON'T give power to the coal generators yet. Let the belts and pipes completely saturate with materials, THEN turn the coal generators on. That way you're not starting the system completely cold.
Do a few splits load balancing and then break into smaller sets of manifolds, 4 or 8 machines. This speeds things up a ton.
Pretty sure someone has already created a mod that does just that. An adjustable splitter or merger that sends/take X amt of items going to one direction. It's actually useful in filling up a manifold because I can send the exact amount to all of my machines in one go. Like if I have 4 prod building and it's being fed by a manifold, I can do a 1 to 3 split. Where the one goes to the first building then the rest will go down the line quite nicely. Although after the setting cpy & paste feature came out and this mod not updated to that, this mod became tedious and I ended up not using that anymore and just like my manifolds fill up. Then I discovered the modular balancer mod and it is awesome!
One could say that you really should need mods, but then Coffee Stain themselves would rather let the amazing modding scene handle that while they tackle for bigger things
It is realy simple:
Just make a belt valve.
You can set a smart splitter in front of it, and call it a day.
This is actually, kind of a good idea!
It's easy to code one. The devs just want folks to make load balancing setups to get their desired ratios.
What i want is logic systems. This throughput based system drives me nuts when it comes to anything being shipped where i have only the clunkiest tools possible to manage how much goes were.
The lack of any automated control at a higher level than throughputs and overflows just drives me batty after creating smart train systems in factorio where stuff only gets delivered if it's actually needed.
That’s the fun of it. There’s a multitude ways of doing it in satisfactory
The slider is there. It’s on the machines.
What I do is put the max resources in the lower production machine and than set up a smart splitter so the overflow goes to that lower machine and the higher is always running at peak.
I agree, the that splitter isn't necessary. A lot of issues that come from not having a splitter like gets fixed by oversaturating like you said. If you really care about ratios, you could use lower tier conveyor belts. As for coding a precise splitter, it wouldn't be difficult to do.
It’s been awhile since I thought I wanted something like this and I’m trying to remember the specific use case, and there’s probably a pretty good chance that I’m being lazy and could’ve done it better but in my lazy set up this would’ve been helpful
I think what it was is I had electromagnetic rods being produced somewhere and I wanted to send some off down a long conveyor to be part of a nuclear set up in a cave , and I had a splitter sending one to storage from one to the nuclear cave
The problem was they were then going onto a sushi train built with several other things or being sent down the one belt to the nuclear cave and when they came out of the nuclear cave there was way too many electromagnetic rods going down there than I actually needed , but the problem was the only way I had to send them was a splitter that said to send every second rod to the nuclear cave and every first rod to storage
Now maybe if I had a custom splitter set up where it’s sent outputs one and 2 to storage and then output three went back in and one and two went to storage again and and that way it was only sending you know every second one or something like that, maybe
But in that very specific use case my brain said I wish I had a splitter where I could just say send every fifth one out output two
As a computer programmer myself, I would think that it would be pretty easy to make a ratio splitter.
ie. if you set output 1 to 33.33% and output 2 to 66.67%, the splitter would repeat putting 1 item in output 1, then 2 items in output 2. Although, I could see the problem with setting very precise percentages, as then, the splitter would put a ton of items through one output before slightly more or slightly less through the other.
I am very early in my first playthrough and this helped so much lol I was trying to use splitters to get the right ratio of items so that my lines would keep moving. At first I assumed "smart" splitters would be based on the ratio function and was stumped when I realized that wasn't a thing. But this video helped me understand that I do (or will shortly) have the tools I need. I haven't unlocked overclocking but I was only thinking of it as a speed it up function which didn't feel sustainable so early in my game. But underclocking to achieve the ratios I'm looking for would be way easier than the splitter/merger mess I keep trying to make work xD thanks for explaining things in a way that made sense to my brain!
Glad it helped!
Personally, I would like a smart splitter that allows for Priority. The best example I have atm is storage.
In a basic Smart Splitter storage I use frequently, I have 3 destinations: Dimension Depot, Container, and Sink. Currently I need 2 smart splitters, 1 to send to Depot then overflows into a second that sends to container which overflows into my Sink line.
What I would like is to set a primary output, then overflow, then a secondary overflow. This way I can utilize all 3 outbound ports, and also reduce the complexity of my footprint.
I agree, precision splitters are not a solution and i also see no need for them. However, what we need is precise speed for parts injection, especially from buffers / containers. Even more, when we work with sushi belts. You can do something with lower tier belt, splitter/merger constructs. But this is limited. There is a mod that provides exactly this - a valve for belts: Throughput Counter and Limiter. And i use it a lot! ... B.t.w., what i also would like to have:
- limiting the available slots of a container
- pre-assigning truck/train/drone slots to a specific item (when you want to run mixed transport of items)
- as train schedules have load / unload item type only already, add a max amount value to those items.
Imagine, you have a factory for AI Limters and Control Rods. One train car is more than enough for volume. Train Shedule like this:
- 1st station unload max 1 stack rod, 1 stack limiter
- 2nd station unload only 2 stack limiter
- and so on ... this would be efficient. ... I not even ask for precise numbers, precise to a stack (or cargo slot count) would already be a win.
Honestly, yes, I absolutely do want that.
For transport.
Putting things into a drone or truck or train, let's say I'm making 40 heatsinks per minute and need 30 for a manufacturer at the same site and need 10 somewhere I need to drone it over to. Let's say the drone is already carrying another small quantity of something else. In this scenario, I cannot simply let the small end saturate and have it balance out, I need to either put in a sink at the far end to keep the two items from locking the other out or run two drones. Now, if I go with two drones, I could allow it to saturate, but that's quite a bit of overkill on the drones, doubling the fuel requirement. As it stands, I either need to up my production to 60/m on the heatsinks and kiss away 20/m to the AWESOME sink, or set up a load balancer so I can run both productions from the existing source.
Now, that's a fairly easy load balancer, but wouldn't it be simpler if I could put in a rate limiter instead? Something like a fluid valve but for belts? Maybe a belt brake? I only want this belt to run at 10/m instead of 60/m...this solves my issue cleanly, and honestly it makes no sense whatsoever that we don't have this option at all without convoluted splitter/merger tricks that you'll probably forget about when you ramp production and wonder why you can't use the extra 10/m you just added to the line, not realizing you're sending 2.5 of them to a sink halfway across the map until you find and remember the load balancer you set up.
Anyway...I guess my point, at the end of the day, though, is why are people so against having options that they won't use? If you won't use it, it doesn't affect you, but if you will, then it'll be a big benefit to you.
Like trucks, and jump pads, and stun rebar, and gas noblisks, someone finds them useful (okay, maybe not gas nobs, those things are completely pointless and terrible, but if you like them I'm glad you have them...just, why? the animals are immune to gas as far as I can tell...), just not me. Why can't this be like that? Something that some like and use, and others don't bother with. Hell, plenty of people think programmable splitters are useless...we still have them and some people find good uses for them. Let us have belt brakes, dammit.
Honestly, the only interesting thing you say in this video is showing how you can actually programme outputs by programming the machines rather than the splitter itself. It amounts to literally exactly the same thing, I would personally find it easier to be done on the splitter rather than down the line, but basically the option is there.
A quick note: I don’t particularly care for a programmable splitter, I never had them in any other game, so there. But it is absolutely TRIVIAL to code in. This is more about restrictions breeding creativity than any difficulty to implement.
Different speed belts, lifts, and under/over clocking is a good way for using load balancing and manifold.