Combining, Splitting, and Balancing Large Numbers of Belts

[Koub]

I see that similarly. Any reasons against moving this to won't implement?

I didn't move it to "Won't implement" because despite I think I know what the devs think about this, I'm not one of the devs myself. However, as I said :

Unlikely to ever happen

[ssilk]

I didn't move it to "Won't implement" because despite I think I know what the devs think about this, I'm not one of the devs myself. However, as I said :

Unlikely to ever happen

Ehm, no, of course not because of the devs opinions.
I think this should be moved, because - as Optera said - it makes no sense, because you already can do it, but need to find out yourself which combination of mods to use.

[eradicator]

I think this should be moved, because - as Optera said - it makes no sense, because you already can do it, but need to find out yourself which combination of mods to use.

Can't agree with that reasoning. "Ideas and Suggestions" is for vanilla. So nothing should be moved "because it can be done with mods" as that would defeat the point of this subforum. To me the only reason to move something from here to "won't implement" is when it is clear that the devs don't intend to implement the idea. Or in some edge cases if the @OP agrees that it's better to have it in mods than vanilla.

And for "wider splitters" i feel that the "clearness of dev intent" is given.

[DeadMG]

TL;DR

Splitters, but with any number of inputs/outputs

What ?

There are very few places where players recommend not attempting to design it yourself. One of these is belt balancers. Players accumulate giant books of belt balancer blueprints made by other players. Players who can't find blueprints of them stand very little real chance of making good balancers themselves because they're pretty nasty.

This seems a bit weird because the devs have added helpers for other things that are way easier to make than balancers, and also belt balancers are basically just bigger versions of functionality we already have; the splitter.

Therefore I simply propose that the splitter be extendable so that you can set it to any number of belts wide, with each output having a separate filter, so if you had a 10-belt-wide splitter, you could set 10 (or 9) filters.

Why ?

Belt balancers have always been a pain in terms of size, performance, and finding a working blueprint. With train intersection design becoming relatively trivial in 2.0 I think the equal upgrade for belts is to make belt balancers trivial.

[Koub]

[Koub] Merged into older thread with same suggestion.
Also : #thereisamodforthat : https://mods.factorio.com/mod/Multilane_Splitters

[DeadMG]

Last I checked, all mods that offer belt balancing functionality have some truly horrible UPS problems that makes using them effectively a non-starter.

[Murklak]

Last I checked, all mods that offer belt balancing functionality have some truly horrible UPS problems that makes using them effectively a non-starter.

Is there some reason why a Wube implementation would be similarly poorly optimized? If not, I wouldn't worry about how mods perform when considering a vanilla implementation of this feature.

[mrvn]

Last I checked, all mods that offer belt balancing functionality have some truly horrible UPS problems that makes using them effectively a non-starter.

Is there some reason why a Wube implementation would be similarly poorly optimized? If not, I wouldn't worry about how mods perform when considering a vanilla implementation of this feature.

Wube can do this in C++ avoiding having to spin up the LUA instance on every tick to check if there are items on the input belt and to place items on the output belt. The splitter in C++ can also sleep until the input belt has items or the output belt has space.

So A) a lot faster and B) running far fewer times.

[Illiander42]

Or you could use merged chests and loaders.

But balancers have almost no uses since the introduction of splitter priorities.

The only place they're still the right thing to use over compressors is near trains.

[DeadMG]

The engine implementation of splitters is far more efficient than what any mod could do. All I'm asking for is that kind of efficiency on more than two belts.