Conveyor Belt Router

[Puzzlemaker]

While playing I noticed there isn't any way to separate the two sides of a conveyor belt easily. It would be nice if a "Router" object was added that is similar to a splitter and could route objects based on which side of the belt they entered from. It would also be nice to have a 1x1 object that could funnel objects onto one side or the other of a belt.

[slay_mithos]

There is actually a way to separate both sides, using a splitter, into two "belt to ground" exits facing away from each other.
It's not very intuitive at first, but it works flawlessly.

[Puzzlemaker]

There is actually a way to separate both sides, using a splitter, into two "belt to ground" exits facing away from each other.
It's not very intuitive at first, but it works flawlessly.

That is... huh.

It would be nice if there was an intuitive way to do it. Or even just an option for the splitter, it could pull double duty easily.

Edit: Okay, I just tried it. Why does that work!?

Edit2: Ah, underground belts from the side only let one side of the conveyor belt through... that is useful to know!

[Drury]

I think it's a bit glitchy/hacky honestly. How are new players supposed to know that when you feed an underground belt from the wrong side, it filters out the items placed on it? It doesn't make any sense either, it's just a glitch.

I support the idea that we should have a splitter type that does this in a proper way. It can be done already, but it shouldn't be possible to do it the way it's currently being done.

[DrNoid]

I would also love to see:

• a second mode for the splitter so it separates the streams by side.
• a "device" to merge both halves of a belt onto one side.

The second one (a simple piece of metal really ) can either be a single one with a side-toggle, or two different ones (merge left or merge right)

[just_dont]

While all these designs can be made with inserters and underground belt "glitch", I too wish we had a special compact and deterministic belt sorter. Inserters are non-deterministic, even if you place a couple of smart inserters, they could skip an item due to various circumstances (a factory blackout, for example). And when you try to build a deterministic sorter on smart inserters -- you quickly discover that the goddamned thing takes A LOT of space.

An ideal sorter, in my opinion, should be like that:
1) 1x1 size (or even 3x3 size -- still would be nice);
2) Uses lots of energy (to balance it with smart inserter designs -- somewhere around x5 smart inserters at least);
3) Items won't pass through if sorter isn't working (unpowered, etc);
4) 1 belt of input, 3 belts of output;
5) Configurable per lane. I.e. for 3 output belts you have 6 lanes, and you can explicitly configure which item should go which lane (with the UI similar to that of smart inserter).
6) If one item is configured for several lanes, it'll split it equally.
7) Has the same throughput as the belt.

This would allow all kinds of compact designs, lane switching, sorting, etc. -- at the cost of considerable amount of energy.

But then again, if one day we would get more inserter options -- then a special item for sorting could quickly become redundant.
I.e. if we would get a "90-degree inserter" (you already can, with mods) -- then belt sorting, even the deterministic one, becomes very compact.

a "device" to merge both halves of a belt onto one side.

3 pieces of belt (so 1x3 size) can do that trick. I don't think there should be a separate item just to go from 1x3 to 1x1.

[ssilk]

While playing I noticed there isn't any way to separate the two sides of a conveyor belt easily. It would be nice if a "Router" object was added that is similar to a splitter and could route objects based on which side of the belt they entered from. It would also be nice to have a 1x1 object that could funnel objects onto one side or the other of a belt.

https://forums.factorio.com/wiki/inde ... anes_trick

@rest: I recommend to make a small search about that subject. There have been many tries to introduce such a thing, even by me. In my current opinion it isn't needed, cause there are always (!) solutions without any changes to the current available items possible. Anything what's needed is more items and more of course more space.

I think in, the only point is, that it is often cumbersome to built that. I think this point can be changed, for example if there are some kind of blueprints, which can be built much earlier in the game by the player, exactly for stuff like that.

[Drury]

While playing I noticed there isn't any way to separate the two sides of a conveyor belt easily. It would be nice if a "Router" object was added that is similar to a splitter and could route objects based on which side of the belt they entered from. It would also be nice to have a 1x1 object that could funnel objects onto one side or the other of a belt.

https://forums.factorio.com/wiki/inde ... anes_trick

"Easily" is the keyword here. It's also cheat-ish because you have to utilize a glitch.

[ssilk]

Note, that I added more to my above comment.

Cheat-ish, yes. This is one of those thing, you never find yourself. But most good games have some things like that.

[Drury]

Yep, but it most cases these little helpful glitches aren't as helpful as the game's main features.

All in all it's strange that there are multiple intended ways to place items on both sides of a belt but no intended way to split them up.

[rk84]

The belt-to-undergound graphics could be shifted&scaled to look less glitchy?

Over year ago, it was decided it is an unintentional feature.
Undergroung belt takes stuff from sides.

[Drury]

Well, it's not intuitive and makes no sense, so I'd say, if not remove it, at least make it obvious that it works, because it is HUGE.

You can't let "an unintended feature" like that just sorta exist there as one-of-a-kind key factor of belt distribution system. That's usually what intended and well-documented features are for.

[ssilk]

Yep, but it most cases these little helpful glitches aren't as helpful as the game's main features.
All in all it's strange that there are multiple intended ways to place items on both sides of a belt but no intended way to split them up.

Cause you don't need that very often? I used that now ... hm ... 10 times would be too much...
I understand, but see it from the practical side: I don't see much usage for splitting up lanes, I could find only a handful of situations. In most cases you join them, but splitting...? If you need to split lanes my experience tells me, that this is in most cases broken by design and it is difficult to extend. For me it is a clear sign, that it is now easier to break everything down and rebuilt it, instead of extending it, rebuild it anywhere else with more space, use both lanes instead of one and so on. Designs, where I used splitting lanes don't survive in most cases the next 2 hours.

Well, it's not intuitive and makes no sense, so I'd say, if not remove it, at least make it obvious that it works, because it is HUGE.
You can't let "an unintended feature" like that just sorta exist there as one-of-a-kind key factor of belt distribution system. That's usually what intended and well-documented features are for.

Well, I think you're right with that. But I would replace "HUGE" with "sometimes it's very useful".

[Bleda]

@Silkk: Since I found out about the belt-to-ground trick, I'm using it all the time. It's very useful for large belt-based factories. Of course, if you switch to logistic robots as soon as you can, you don't need any expansions to the belt system. But some people just prefer belts to robots, because they are more interesting to set up.

I also think that there should be a more obvious and beautiful way to split the two lanes of a belt.

[Neotix]

But some people just prefer belts to robots, because they are more interesting to set up.

Belts aren't only more interesting but also more elegant. Bots are flying around like insects, they do more mess in factory than anything else. I can't pick items from ground because of flying bots, sometimes i can't even see what is on ground. Bots management not exist.

That why i also would like to have more machines for belt system.
Also i would like to have monorails system (belts and bots hybrid).

[ssilk]

Belts aren't only more interesting but also more elegant. Bots are flying around like insects, they do more mess in factory than anything else. I can't pick items from ground because of flying bots, sometimes i can't even see what is on ground. Bots management not exist.

I'm sure, this is on the todo-list. The idea I remember was to have switchable layers for that.

That why i also would like to have more machines for belt system.
Also i would like to have monorails system (belts and bots hybrid).

Well, my opinion is, that the bots are a too cheap with using energy. I think they should use up some more, because currently the advantage of having bots vs. belts is really high. We have had plenty of discussions about that some prefer to use the bots for really everything. I would like to have the bots shrinked to smaller areas and the mass transport is made by belts or train or some other kind of new system (monorails, carts, chests on belts, funicular... however it is called).

[Neotix]

I'm sure, this is on the todo-list. The idea I remember was to have switchable layers for that.

It will change almost nothing, only that you will be able to interact with structures on ground without interact with bots (and vice versa). Bots will still flying like insects, making factory look like nest, not like factory.

Well, my opinion is, that the bots are a too cheap with using energy. I think they should use up some more, because currently the advantage of having bots vs. belts is really high. We have had plenty of discussions about that some prefer to use the bots for really everything. I would like to have the bots shrinked to smaller areas and the mass transport is made by belts or train or some other kind of new system (monorails, carts, chests on belts, funicular... however it is called).

Yes, they're chap, small and we need hundreds of them to provide enough items amount. They're flying multiple times between Provide and Request Chests to transport even small amount of items. One big mess.

[Puzzlemaker]

Wow alright I didn't expect this thread to explode.

Just having a mode for the splitters that can split based on which side of the conveyor the items on would be really nice and probably easy to add.

Having an funnel that can move items to one side or the other of a conveyor would also be nice.

It's possible to do both those, but they require weird shinanigans.

That is all there is to it.

[Hyena Grin]

I've been trying to figure out a way to do this that is more elegant and less power intensive than simply using smart inserters to separate two sides of a conveyor belt. I regularly structure my production such that I am using both sides of a belt for different products/resources, and not being able to easily re-separate them onto different belts without setting aside space for multiple inserters is frustrating. This type of function is extremely rudimentary, and I'm surprised that it hasn't been included. More surprised that someone is actually arguing against it, but to each their own.

The underground belt glitch is a good find, but I agree that it's odd that there isn't a built-in tool for this purpose. I'd much rather use a researched device than a glitch.

I also like the idea of the funnel to switch and combine lanes onto one side. I'd use the heck out of that also, for example, when you have mining drills dumping onto both sides of a belt, and you want to shift all the ore to one lane so that you can start dumping coal onto the second lane. Or you have a line of assembly machines dumping products onto an inconvenient side of the belt. Yes there are ways to do this using perpendicular belts, but for those of us looking for more elegant, less space-intensive designs, these kinds of VERY BASIC production line mechanisms would be wonderful.

Perhaps put the funnel into Logistics 1, and the router (whether its own unique object or an optional mode for the splitter) into Logistics 2.

[ssilk]

I also like the idea of the funnel to switch and combine lanes onto one side. I'd use the heck out of that also, for example, when you have mining drills dumping onto both sides of a belt, and you want to shift all the ore to one lane so that you can start dumping coal onto the second lane. Or you have a line of assembly machines dumping products onto an inconvenient side of the belt. Yes there are ways to do this using perpendicular belts, but for those of us looking for more elegant, less space-intensive designs, these kinds of VERY BASIC production line mechanisms would be wonderful.

Well, there are some things mixed together. For the splitting of the lanes, I feel with you. The solution with the underground belt is not very elegant and not obvious. But I won't say "do it so or so", I let the devs decide.

For the rest: Tell me a good reason. Currently I see no reason to have for every function an extra item, cause you need to produce that item, which cludders then your inventory. And all you need is to think and place some belts/splitters/undergrounds. I see a need for having this knowledge better shared, but not in the alpha-version.