Factorio Forums

The current means of unmerging two halves is complicated but most of all completely non-intuitive to a new player.

If all it takes is belts to merge, it should only take belts to unmerge. I see it as simple as a 5th rotation option that only appears when a belt is placed at a T-intersection. Or even a context-sensitive auto placement if the constituent belts are placed a certain way.

Short answer don't merge two belts, if you later want to separate them.

Longer answer: I understand that building something to separate 2 belts isn't obvious to new players, but it can be done with just a few belts. (And if you only want one of the output lanes, then you don't even need the splitter).

I would not call it complicated, but it is not obvious. Here is a palette from my creative world for splitting and merging.

It's not so much complicated as it is unobviousness, and an ugly hack.

ManaUser wrote:It's not so much complicated as it is unobviousness, and and ugly hack.

I fully agree.

Related thread:
viewtopic.php?f=5&t=48047 Underground belt lane splitting, anyone else hate it?

The underground belt graphic doesn't help, by looking at it you would agree that it should block all lanes.

I don't really care that it is not a popular opinion, but I love this trick. I would hate to see it go and I would hate it for new parts to be introduced to replace its functionality. Ideally we have a small set of simple components that interact in lots of interesting ways (which boosts creativity) instead of dedicated solutions to specific problems.

Yes, the graphic could be altered to make this functionality more obvious.

Caine wrote:I don't really care that it is not a popular opinion, but I love this trick

I agree that being able to split the individual lanes of a belt is an important feature.

Caine wrote:I would hate to see it go and I would hate it for new parts to be introduced to replace its functionality.

We agree that we disagree in that respect.

Caine wrote: Ideally we have a small set of simple components that interact in lots of interesting ways (which boosts creativity) instead of dedicated solutions to specific problems.

I fully agree with this statement. However, I would add that these simple components should always act in an intuitive manner. This is certainly not the case with underground belts, when used as belt lane splitters. It is rather an ugly hack.

Tekky wrote:We agree that we disagree in that respect.

Yes we do. Although I am also open to the introduction of a new item that handle all the cases in an interesting way. I like that the underground belt solves the problem, but it is not a clean solution. It takes space forcing you to be creative with your belts.

In the underground trick hate thread linked above there is this example:

I see at least two new components here and at least a third one is missing (which would allow the inner lanes of two belts to pass through). To me, these are examples of one-trick ponies which are just too boring of a solution.

I have no idea what a proper solution would look like to me. The underground has two issues:

• It is not obvious that they can do this, the graphics even suggest that this will not work. I say, fix the graphics.
• You cannot have them back-to-back as they link up and thus must have the same direction. I am not sure I want this addressed. It adds to the challenge.

Would an alternate solution be to stop underground belts from splitting off the end of a belt, but add a new entity which does exactly that. Seems like a step sideways maybe when the cleanest solution (as I think as a few others do) Is to rework the underground graphic, to make it obvious that in X rotation a new interaction occurs.

Also, I'm pretty sure I saw Rseding91 mention that the way undergrounds split like this is part of the suite of tests they auto-run? Please correct me if I am wrong here.. So I can't see this feature being changed? Though I would myself say that it does need a cleaner implementation.

I see no need for a new device; belts could just have a context sensitive option to zipper out at the end of a T, or even at just a right angle.

It's easy to make coffee if you have hot water and roasted, ground coffee beans. Not so easy to split that coffee back into the dry grounds and the hot water. And if you know you want pure hot water later, you probably shouldn't use all of it to brew coffee.

Has anybody ever suggested a filter splitter?
A, lets say, purple spltter with which you can select up to X items you want to direct to the other belt.
The rest will just pass through,

Similar things have been suggested before. But since you can already achieve the same thing with filter inserter why should the devs add it? Just making it easier to so something we can already do with a couple of pieces of belt or a few inserters isn't necessarily a good argument for a new item. Things like introducing new and interesting game mechanics or reducing tedious or annoying aspects of gameplay are better reasons. If separating belts is annoying to you or you don't like using inserters or undergrounds to do it, you can plan always your designs so you don't need to separate belts.

A filter splitter seems to be quite easy object to add to the game, but I might be wrong.
Yes, filter inserters can technically do the same thing, but their throughput is very limited.
You will need multiple filter inserters to achieve the same goal.
This is an ugly solution and a filter splitter might be a fun addition.

I have to admit, any sane person would never dump all sorts of items on one belt with the intention to seperate them later.
It is very inefficient.
Unfortunatly I happen to be insane. Sometimes I truly want to make organized chaos.
Kind of like obfuscated code when programming, only now we have an obfuscated factory.

5thHorseman wrote:It's easy to make coffee if you have hot water and roasted, ground coffee beans. Not so easy to split that coffee back into the dry grounds and the hot water. And if you know you want pure hot water later, you probably shouldn't use all of it to brew coffee.

Without using analogies, unmerging belts reduces the entropy of your factory, thus It can't be an easy thing to do. This is a consequence of the laws of belt-dynamics.

m44v wrote:

5thHorseman wrote:It's easy to make coffee if you have hot water and roasted, ground coffee beans. Not so easy to split that coffee back into the dry grounds and the hot water. And if you know you want pure hot water later, you probably shouldn't use all of it to brew coffee.

Without using analogies, unmerging belts reduces the entropy of your factory, thus It can't be an easy thing to do. This is a consequence of the laws of belt-dynamics.

I am not certain that a fundamental law of nature is as fundamental in a digital world.
A digital world where (almost) only imagination is holding one back.
An item which minimizes entropy can cetrainly be created.

Lextreme wrote:I have to admit, any sane person would never dump all sorts of items on one belt with the intention to seperate them later. It is very inefficient.

Please enlighten me on your definition of efficiency, as I do not see this.

E.g. I have an early game green circuit belt where I do this to share iron input and circuit output on a belt which avoids the need for long handed inserters.
See: viewtopic.php?f=202&t=21922&start=20#p324199
I use this belt sharing technique whenever I can to simplify. There is no efficiency loss here in my opinion, just a way of balancing throughput ratios.

Caine wrote:

Lextreme wrote:I have to admit, any sane person would never dump all sorts of items on one belt with the intention to seperate them later. It is very inefficient.

Please enlighten me on your definition of efficiency, as I do not see this.

E.g. I have an early game green circuit belt where I do this to share iron input and circuit output on a belt which avoids the need for long handed inserters.
See: viewtopic.php?f=202&t=21922&start=20#p324199
I use this belt sharing technique whenever I can to simplify. There is no efficiency loss here in my opinion, just a way of balancing throughput ratios.

Two is not "all sorts"

Also, you know how to easily split a belt carrying 2 items into their own belts. I doubt you'd have started a thread similar to this one.

I think my biggest issue with this is... it just doesn't make logical sense.

Yes, I agree that building creative setups out of a minimal number of pieces is part of the appeal of factorio. But, let's imagine you're an engineering designing a product line of belt pieces to sell.

Would you:
A) Design in the piece that splits lanes, because all it takes is adding a single curved divider between the two lanes to direct them to different belts?

or

B) Tell all your customers that they don't need the 1 tile solution because a more expensive 6 tile solution already exists?


From another standpoint, if you're actually building this, would you rather:

A) Place a little metal divider on a single piece of belt

or

B) Place an expensive splitter, then dig two tunnels underground so that you can use a non-intuitive approach to a simple problem?

OK, Factorio is a fictional world. But a key element of fiction is "plausible suspension of disbelief", and Factorio is also all about efficiency - in the real world, the lane splitter is such a painfully simple, obvious, and intuitive piece that no real designer would even remotely consider skipping it in favor of the underground belt solution.