Combining, Splitting, and Balancing Large Numbers of Belts

Post your ideas and suggestions how to improve the game.
Carl
Inserter
Inserter
Posts: 45
Joined: Mon Mar 07, 2016 10:25 pm
Contact:

Combining, Splitting, and Balancing Large Numbers of Belts

Post by Carl »

Sorry for the awkward title but i only have a rough idea of how maybe the point i'm going raise could be addressed. The core issue i'm getting at here however is when you have a large number of belts feeding into somthing, (usually ore coming off the train/miners for me though some people prone to even more insane sized factories may have similar issues with the iron and copper portions of their main bus), and balancing or compressing down/up the number of belts gets very troublesome, i can do it fairly easily from a setup PoV for any number which is both even on input and output using splitters, (i imagine everyone can duh), but if you want an odd number of outputs or inputs i haven't the foggiest how to do it without a LOT of space usage, and above 4-5 belts it gets really, really space consuming even with even numbers on inputs and outputs.

Conceptually what's needed is a method of expanding the basic splitter concept to encompass as many input/output steps as you want. The problem is i'm not really sure how well the game support such modularity. The alternative method is to be a bit cheekier and resurrect the loader/unloader concept, but make it only interact with a specialised storage machine, that would have to be modular, but based on liquid tanks it looks like a unified storage inventory across multiple seperate but connected blocks is possibble, For UI reasons it would have to be inaccessible to the player, (in the same way you can't access the inventory of an inserter), and the total capacity per block does not need to be high, but if done right it should meet the needs.

Now some might say you won't normally need this, and i'd say yes and no, it depends a great deal on how big your final form is and at what point in the game your talking about, small to moderate final builds will probably just about bump into this issue IMO, however large and above builds will start running into it late game, (or earlier for bigger builds), and i imagine as new ever more expensive techs and machinery are added you'll find the definitions of small, medium, large, e.t.c. increasing which make the problem more prevalent and occur at earlier stages. Still i'd say any such tech should be mid game at the earliest. If anyone has any other idea's on how thi8s might be achived, (or even a technique i haven't discovered), feel free to but in.

Mehve
Filter Inserter
Filter Inserter
Posts: 316
Joined: Sat Aug 06, 2016 9:12 pm
Contact:

Re: Combining, Splitting, and Balancing Large Numbers of Belts

Post by Mehve »

I suppose it's a matter of opinion how big a problem this is. While current belt balancer designs are obviously more than a single space solution, you can usually combine/split quite a few lanes before things get really awkward.

mattj256
Fast Inserter
Fast Inserter
Posts: 203
Joined: Sun Mar 27, 2016 7:25 am
Contact:

Re: Combining, Splitting, and Balancing Large Numbers of Belts

Post by mattj256 »

I like the game's design as it stands. If you're obsessive and have a lot of time on your hands you can figure out how to do this with nothing but splitters and underground belts. If you're like me you can look it up online and try to memorize it. Or you can take the newbie approach and do a simple design with splitters that works but not optimally.

I like the elegance of the circuit network designs but they can be cumbersome to set up. This will be easier once 0.15 comes out and we can store and share blueprints across games.

If the developers want to add anything, it should be a priority splitter. Here's a mod that does what I'm talking about: https://mods.factorio.com/mods/msu320/compoundsplitters First it outputs as much as possible onto one belt. If that belt is full it starts putting things onto the next belt, and so on. It would be nice if this was in vanilla. (There could be a "mode of operation" on the existing splitters. No need to create a new entity.)

Carl
Inserter
Inserter
Posts: 45
Joined: Mon Mar 07, 2016 10:25 pm
Contact:

Re: Combining, Splitting, and Balancing Large Numbers of Belts

Post by Carl »

I suppose it's a matter of opinion how big a problem this is. While current belt balancer designs are obviously more than a single space solution, you can usually combine/split quite a few lanes before things get really awkward.
Well i'm specifically talking about the point at which it does get awkward as all hell. As an admittedly extreme example my current setup for copper roe has to feed 10 lines of miners down into 8 whilst balancing and compressing. My Irone ore is 13 into 20. (both those outputs go through a belt->storage->belt sequence that cut the final numbers in half). But really over 8-10 combined input+output lanes gets to be a real pain.
I like the game's design as it stands. If you're obsessive and have a lot of time on your hands you can figure out how to do this with nothing but splitters and underground belts. If you're like me you can look it up online and try to memorize it. Or you can take the newbie approach and do a simple design with splitters that works but not optimally.
Thats an interesting setup, but the bigger of those is starting to push the practical limits of the concept before it becomes hugely unwieldy. My own design is quite a bit different and smaller sizes a touch more unwieldy but if it scales as i think it should, (i really need to put it through a test in sandbox ;)) it's probably better at big sizes.

This is specifically about those cases where people are building larger factories, like you need to, to beat some of the achievements, the amount of copper and iron you'll consume require a lot of ore to be processed, so it's going to give you issues like this. And i imagine in the long run as noted resource requirements are going to go up for basic gameplay which is going to extend the problem down to more typical factories.

Mehve
Filter Inserter
Filter Inserter
Posts: 316
Joined: Sat Aug 06, 2016 9:12 pm
Contact:

Re: Combining, Splitting, and Balancing Large Numbers of Belts

Post by Mehve »

Carl wrote:Well i'm specifically talking about the point at which it does get awkward as all hell. As an admittedly extreme example my current setup for copper roe has to feed 10 lines of miners down into 8 whilst balancing and compressing. My Irone ore is 13 into 20. (both those outputs go through a belt->storage->belt sequence that cut the final numbers in half). But really over 8-10 combined input+output lanes gets to be a real pain.
That's kind of the point, though. It's easy to get jaded by some of the megafactories out there, but 8+ similar belt lines IS extreme. By the time you have that many belts moving the same material in parallel, you're into the realm of single-digit minutes between rocket launches, and equivalent. I'm not sure we should expect shortcuts at that point.

Carl
Inserter
Inserter
Posts: 45
Joined: Mon Mar 07, 2016 10:25 pm
Contact:

Re: Combining, Splitting, and Balancing Large Numbers of Belts

Post by Carl »

Mehve wrote:
Carl wrote:Well i'm specifically talking about the point at which it does get awkward as all hell. As an admittedly extreme example my current setup for copper roe has to feed 10 lines of miners down into 8 whilst balancing and compressing. My Irone ore is 13 into 20. (both those outputs go through a belt->storage->belt sequence that cut the final numbers in half). But really over 8-10 combined input+output lanes gets to be a real pain.
That's kind of the point, though. It's easy to get jaded by some of the megafactories out there, but 8+ similar belt lines IS extreme. By the time you have that many belts moving the same material in parallel, you're into the realm of single-digit minutes between rocket launches, and equivalent. I'm not sure we should expect shortcuts at that point.
Not even close to that often. Bear in mind the way i'm running it it only comes to 4 lines of feed output via the belt->storage->belt setup and i plan for re belts as the prime mover so the throughput is a lot less than it seems. That many inputs is to balance the expected uncompressed state they'll run in for much of the time between trains once i inevitably have to move to that. Don;t get me wrong i'd still classify this on the upper end of the large scale, but you don;t need 8 before it becomes a monumental pain, anything more than about half that size is one. Big builds just make it especially noticeable. But it';s there on anything above a medium sized factory.

User avatar
ssilk
Global Moderator
Global Moderator
Posts: 11093
Joined: Tue Apr 16, 2013 10:35 pm
Contact:

Re: Combining, Splitting, and Balancing Large Numbers of Belts

Post by ssilk »

Another splitter. :)
I point to viewtopic.php?f=80&t=518 Smart Splitter (-Suggestions will never End) and
viewtopic.php?f=80&t=24061 New Types of Belts (And why they are NOT a Good Idea!)


Btw: This subject is addicting; even me couldn't resist to write some splitter suggestions:
viewtopic.php?f=6&t=23616
Cool suggestion: Eatable MOUSE-pointers.
Have you used the Advanced Search today?
Need help, question? FAQ - Wiki - Forum help
I still like small signatures...

Carl
Inserter
Inserter
Posts: 45
Joined: Mon Mar 07, 2016 10:25 pm
Contact:

Re: Combining, Splitting, and Balancing Large Numbers of Belts

Post by Carl »

Well i would have posted this there, except this isn't about splitting for the most part. Yes you can use splitters to do compression or load balancing across multiple belts but it's not what they're designed to do, thats just another use people have figured out ways to use them to do. And by and large i don't have any real issue with that. Unfortunately beyond a certain point such ad hoc solutions become completely unworkable due to the scale involved. At those scales the game really needs a (suitably high tech tier, expensive, e.t.c.), dedicated solution to the issues of compression and balancing. Slitting is quite simple and the least space consuming part if all inputs are load balanced allready. I just don;t see any reason a dedicated load balancer or compressor shouldn;t handle the splitting aspect integrally. But it's very much the compression and especially balancing which is a giant pain in the posterior beyond a fairly modest scale.

topica
Manual Inserter
Manual Inserter
Posts: 1
Joined: Thu Jul 05, 2018 5:53 pm
Contact:

Wide Splitter (3 belt splitter)

Post by topica »

TL;DR
Wide Splitter (3-belt splitter) makes cargo wagon loading and unloading better.
What ?
Wide Splitter is 3x1 entity and has 3-belt input and 3-belt output.
This splitter can combine 2 or 3-belt input and divide them to output belts equally.
Why ?
When I use cargo wagons, I use 3 to 1 belt balancer or 6 to 1 balancer that is implemented with conventional splitters.
(for 4 cargo wagon unloading, four 3 to 1 belt balancers -> 4 to 4 belt balancer)

Well-known 3 to 1 belt balancer, however, has a complicated shape and its throughput is not perfect.
This makes loading or unloading logic obscure.
I think belt logistics should be more useful.

4-belt splitter or more are unnecessary because (2^n)*(3^m) belt balancer can be implemented if there are 3-belt splitters, and other belt balancers have low demand in the game.

Koub
Global Moderator
Global Moderator
Posts: 5572
Joined: Fri May 30, 2014 8:54 am
Contact:

Re: Wide Splitter (3 belt splitter)

Post by Koub »

Won't happen :)
Koub - Please consider English is not my native language.

User avatar
bobingabout
Smart Inserter
Smart Inserter
Posts: 7198
Joined: Fri May 09, 2014 1:01 pm
Contact:

Re: Wide Splitter (3 belt splitter)

Post by bobingabout »

would love
Creator of Bob's mods. Expanding your gameplay since version 0.9.8.
I also have a Patreon.

User avatar
darkfrei
Smart Inserter
Smart Inserter
Posts: 2658
Joined: Thu Nov 20, 2014 11:11 pm
Contact:

Re: Wide Splitter (3 belt splitter)

Post by darkfrei »

It can be 3x3 entity:
3 mini-loaders (input), one chest 3x1 and 3 mini-loaders (output).

User avatar
eradicator
Smart Inserter
Smart Inserter
Posts: 4064
Joined: Tue Jul 12, 2016 9:03 am
Contact:

Re: Wide Splitter (3 belt splitter)

Post by eradicator »

bobingabout wrote:would love
would love 1x1 splitter that splits to all three sides :x.

crysanja
Long Handed Inserter
Long Handed Inserter
Posts: 73
Joined: Sun Aug 03, 2014 4:47 pm
Contact:

Re: Wide Splitter (3 belt splitter)

Post by crysanja »

this is a complicated solution, the easy one would be a 1x6 chest.

User avatar
Gergely
Filter Inserter
Filter Inserter
Posts: 547
Joined: Sun Apr 10, 2016 8:31 pm
Contact:

Re: Wide Splitter (3 belt splitter)

Post by Gergely »

Is this moddable?

User avatar
darkfrei
Smart Inserter
Smart Inserter
Posts: 2658
Joined: Thu Nov 20, 2014 11:11 pm
Contact:

Re: Wide Splitter (3 belt splitter)

Post by darkfrei »

Gergely wrote:Is this moddable?
Just take this two:
https://mods.factorio.com/mods/Atria/WideChests
https://mods.factorio.com/mod/miniloader

Three loader to load, three loader to unload chests 3x1 or 1x3.

User avatar
Optera
Smart Inserter
Smart Inserter
Posts: 2637
Joined: Sat Jun 11, 2016 6:41 am
Contact:

Re: Wide Splitter (3 belt splitter)

Post by Optera »

Has anyone ran an ups comparison for n to m balancers made up from splitter, loader and miniloader (hidden inserter)?

User avatar
eradicator
Smart Inserter
Smart Inserter
Posts: 4064
Joined: Tue Jul 12, 2016 9:03 am
Contact:

Re: Wide Splitter (3 belt splitter)

Post by eradicator »

I have to admit i expected splitters not to properly split when there's not enough input to compress all outputs. I expected that the first loader would grab all the input and the other two would go empty. But alas it magically works:
balance.jpg
balance.jpg (134.66 KiB) Viewed 2577 times

User avatar
Optera
Smart Inserter
Smart Inserter
Posts: 2637
Joined: Sat Jun 11, 2016 6:41 am
Contact:

Re: Wide Splitter (3 belt splitter)

Post by Optera »

I just ran 3 different 4x4 balancer builds through 10k ticks benchmarks and the results where surprisingly definite.

The setups:
2018-07-07-16-02-03-2543866.jpg
2018-07-07-16-02-03-2543866.jpg (184.06 KiB) Viewed 2576 times
Results:
splitter: 0.315 ms/tick
loader: 0.260 ms/tick
miniloader: 0.171 ms/tick

Conclusion:
Miniloader are nearly twice as ups friendly as splitter based balancers with loader somewhere in the middle.
For larger balancers the advantages of loader and especially miniloaders will only increase. Next time I need a 16x16 balancer slapping miniloaders next to a wide chest is the way to go.

Sidenote:
Only the splitter build is count perfect. The loaders distribute according to build order and chunk update.

User avatar
eradicator
Smart Inserter
Smart Inserter
Posts: 4064
Joined: Tue Jul 12, 2016 9:03 am
Contact:

Re: Wide Splitter (3 belt splitter)

Post by eradicator »

Optera wrote: Results:
splitter: 0.315 ms/tick
loader: 0.260 ms/tick
miniloader: 0.171 ms/tick
Not too suprised about splitters being last, after all they have to do per-lane logic and count-precision splitting etc. Loaders being that much slower than inserters ... i guess they never get any optimization love.

Post Reply

Return to “Ideas and Suggestions”

Who is online

Users browsing this forum: No registered users