Half Belt Splitter

[robhol]

I hate when suggestion subforums turn into a circlejerk about how completely perfect the status quo is, and damn anybody who has any idea that changes the game. Sigh.

[Gammro]

A level of conservatism is often found in gaming communities that have been established for a while. This can be seen by the "at least 100 hours" comment made by ssilk, I don't support this. A new player can often bring a fresh look on things that we've accepted as-is. However, I truly believe that this idea in particular is not something I want in my game, and rather have the developers spend their time on things I do like.

In the end I don't make any decisions, but I can try to influence them, exactly like you are doing. That's why I voice my opinion here.

[Garm]

I hate when suggestion subforums turn into a circlejerk about how completely perfect the status quo is, and damn anybody who has any idea that changes the game. Sigh.

A level of conservatism is often found in gaming communities that have been established for a while. This can be seen by the "at least 100 hours" comment made by ssilk, I don't support this. A new player can often bring a fresh look on things that we've accepted as-is. However, I truly believe that this idea in particular is not something I want in my game, and rather have the developers spend their time on things I do like.

In the end I don't make any decisions, but I can try to influence them, exactly like you are doing. That's why I voice my opinion here.

Can I suggest you to look at this thread or this one?

Just because this suggestion has some resistance doesn't mean whole community is conservative. It means that this suggestion has people who like it and people who hate it.

This is the point of this forum - to find ideas that accrue universal support, as well as to polish controversial ideas, and in doing so help developers.

[ssilk]

A level of conservatism is often found in gaming communities that have been established for a while. This can be seen by the "at least 100 hours" comment made by ssilk, I don't support this.

I want to comment this. I mean it not so exact! But when I see some suggestions which are obviously made, because not knowing differently, then I would really say: go on playing and when you really understood come back. The 100 hours is 5-10 more or less normal games. 100 is good to memorize. Maybe 50 would be better... But that's not the point.

A new player can often bring a fresh look on things that we've accepted as-is.

Sure! I really want it. This is important.

[Hyena Grin]

There's two or three people who you will never be able to convince on this subject.

Even though posts suggesting funnels and belt separators seem to get suggested on a regular basis, the same two or three people will make it seem like the entire community is against this idea. From what I can tell, a lot more people seem to be on board with the idea. What the developers think is another story. I have no idea.

Personally, I think that funnels and half belt splitters would vastly improve this game. A large chunk of this game revolves around moving resources around on conveyor belts, and tools in order to do this more efficiently would remove an unnecessary clunkiness and inefficiency (and yes, 'hackiness') that is virtually required in the game right now. Clunkiness that is ugly and messy to look at, clunkiness that detracts from design, and frankly detracts from its appeal to people who aren't already invested in how the game already functions, and thus know how to exploit unintended mechanics.

Either this is a game about building factories, or it's a game solving puzzles on how to accomplish incredibly simple things - such as separating a belt - with a set of tools that does not appear to want you to accomplish them. You need a frigging underground conveyor belt to separate two halves of a belt. An underground conveyor belt to separate two lanes. How the hell does that compute?

How someone can keep a straight face while suggesting that this is good game design is beyond me. There is basically no way to figure out this solution in-game except by accident. I want to separate two lanes on a damn belt. It is a simple and reasonable task to want to accomplish, given a game whose mechanics revolve in large part around belts. Give me a metal sheet and I could literally solve this problem in seconds in real life. I shouldn't need to consult a frigging article with helpful step-by-step pictures to do this in a game about designing factories.

If you don't want to use a new tool, don't use it. Stick with your wildly over-engineered, glitchy, messy designs. I don't see why other people ought to have to use them too.

Every argument that has been suggested against it seems arbitrary, stubborn, and silly to me. But whatever. Hopefully the devs will consider adding something which removes the need for absurd solutions to simple problems. Whether it's a new object or a 'smart' version of the splitter. Personally I'd rather have a single tile object which allows one lane to continue forward while ejecting the other lane out the side (where it can continue on another belt). There would be a basic UI when clicked that allows you to change which lane is being ejected. I would use this literally all the time.

[GotLag]

Seconding everything Hyena Grin just said.

I think it would be more useful and flexible to separate the functionality into another item, though. A 1x1 entity that's a conveyor belt with one side blocked off so items can only flow along the other side. That doesn't reduce the need for splitters, and also provides other interesting possibilities. for example:

Code: Select all

┌─┐
│F│
└─┘
^ v
█▀█

A steel furnace (F) fed with iron ore and coal by an inserter (^) and emptied of iron plates by another inserter (v) on to the same belt the iron ore came from. The iron ore is unable to flow past the furnace as the bottom lane is blocked while the top lane (with the coal) is able to continue.

[AndaleTheGreat]

ok i'm done reading this thread, people just keep reading more into it than there is.Also, to hell with you for insulting me because I haven't played enough in your opinion.
I'm done with this shit thread. Delete the thing. Screw you guys. There was ZERO reason to make all this argument bullshit in here when it is JUST a suggestion.

[khh]

Ok, first off, whatch this video on depth vs. complexity: https://www.youtube.com/watch?v=jVL4st0blGU
This is needed to understand what I mean by depth and complexity, and how my opinion formed.

The way I see it, this is adding a part that is specifically designed to do something you can already do in the game, except it's now made into one block. This does increase the complexity of the game, because you now have to learn the function of yet another building block. But it also makes the previous option obsolete, removing the depth added by using the base game mechanics to your advantage. I know some people think this is an exploit because it's not intended, I don't. I think it's a solid piece of engineering and using the mechanics given to us by the game to our advantage. I think non-intended game mechanics are signs that there's depth beyond what's explicitly given, and that it's a good quality a game can have.

Your argument is undermined by the very video you posted. In it it is stated that depth isn't depth if it isn't available to the players, if it isn't something the player is able to make a concious choice about. And can tell you right now that the splitter+underground solution, while clever, is unavailable to the vast majority of players. It simply doesn't occur to them, because those elements aren't designed to function that way. The only available solution is using smart inserters as a filter, and that's a very ugly solution to such a simple problem.

I am totally for implementing something which can do this. It could be argued that a large part of Factorio is logistics, ie finding the best way to move stuff around. In early and mid game belts are really the only way of doing this, and yet not being able to split the belts feels like there's a basic operation missing. It's like working with bitmanipulation but being denied xor. I mean, sure, ((a&b)&(~(a|b))) works, but it's just not as stylish as a^b.

[Garm]

Why everyone is concentrating on underground hack when smart inserters can perform similar function in legit way?

[khh]

Why everyone is concentrating on underground hack when smart inserters can perform similar function in legit way?

They do work. In fact, it is the only purpose I've really found for them. From a pure depth-vs-complexity view I'd rather add this item and remove smart inserters. I think that would cause more interesting gameplay than the reverse situation we currently have.

[Garm]

Are you serious?

smart inserters have a humongous amount of uses.

They can allow you to send 3+ items on a single belt. In fact my personal record was 20 different items on a looping belt.

They can let you mine mixed ore fields and then separate ores by type onto separate belts

They can work with logic network allowing on demand boiler systems.

They can allow extensive routing systmes with 100% precise custom splits.

They are huge blessing in mixed cargo trains allowing not only custom loading and unloading, but loading/unloading at same time.

They can fill a single line of a belt with custom pattern of items

They are logic gates!

and this is just tip of the iceberg....


That is depth!

[Nahon]

Why everyone is concentrating on underground hack when smart inserters can perform similar function in legit way?

I don't have a solid opinion for or against this idea of one-belt split, the need haven't arose for me to use a device like that yet, but I do understand why it is demanded so often.

The problem with smart inserters, why they don't work reliably all the time is clogging. If you want to remove all items (of the same type) from one lane of a belt and keep the belt going with just the other side (with one or more different type of items), you can set up one or more smart inserters to do this. But if the other belt they're putting the stuff clogs, you'll have a flow of the items on the original belt going past the inserters. You can figure out ways to overcome this, but it won't ever be fully reliable. I wouldn't ever use the solution based on the underground belt glitch. Whenever the devs decide they'd need to change some hitbox, bounding box or other parameter, your system is done for. That's why I consider that unreliable, despite it works flawlessly in the current version.

I wouldn't bury the smart inserter though, it's a very useful device on its own right.

[Hyena Grin]

If your intent is to fully separate two lanes of a belt, the only way to do this without the potential to gum up the works is to use a loop and two smart inserters. That way anything that is missed will end up cycling around to be picked up again. Unfortunately, there isn't an easy way to transfer both lanes of a linear belt onto a loop and still maintain the two-lane setup. So you need to start your 'system' with a loop, or use an additional two smart inserters before the loop, to separate a dead-end belt onto the loop in the appropriate lanes. On the whole, this becomes a rather large, energy inefficient, and problematic design.

So no, I don't think that a loop + 4 smart inserters is a good stand-in for something which should be easily accomplished with one tile and little to no energy cost.

Maybe there are other designs using smart inserters which are error-proof. But that's not really the point. If someone can show me a design which uses two or less smart inserters (not underground belts) and requires no more than 3x3 tiles then they might have an argument. Anything more than that is, in my opinion, not really a solution for such a simple problem. It's an over-engineered, energy inefficient mess that cannot be replicated frequently without a serious drain on energy and space, for performing a task which should be incredibly easy to do.

That is, I think, where the frustration lies. There is this enormous gap between how simple a task belt separation ought to be, and how complex and over-engineered the proposed solutions are. You cannot walk into a factory without seeing a belt separator in some form or another. Sometimes they are complex things and sometimes they are literally just a metal sheet which divides a belt into halves, but they are never as ridiculously over-engineered as are required in the game.

[Sedado77]

Why everyone is concentrating on underground hack when smart inserters can perform similar function in legit way?

They do work. In fact, it is the only purpose I've really found for them. From a pure depth-vs-complexity view I'd rather add this item and remove smart inserters. I think that would cause more interesting gameplay than the reverse situation we currently have.

Are you serious?

smart inserters have a humongous amount of uses.
They can allow you to send 3+ items on a single belt. In fact my personal record was 20 different items on a looping belt.
They can let you mine mixed ore fields and then separate ores by type onto separate belts
They can work with logic network allowing on demand boiler systems.
They can allow extensive routing systmes with 100% precise custom splits.
They are huge blessing in mixed cargo trains allowing not only custom loading and unloading, but loading/unloading at same time.
They can fill a single line of a belt with custom pattern of items
They are logic gates!
and this is just tip of the iceberg....
That is depth!

Agree 100%. Smart inserters are the core of the game.

If your intent is to fully separate two lanes of a belt, the only way to do this without the potential to gum up the works is to use a loop and two smart inserters. That way anything that is missed will end up cycling around to be picked up again. Unfortunately, there isn't an easy way to transfer both lanes of a linear belt onto a loop and still maintain the two-lane setup. So you need to start your 'system' with a loop, or use an additional two smart inserters before the loop, to separate a dead-end belt onto the loop in the appropriate lanes. On the whole, this becomes a rather large, energy inefficient, and problematic design.

No it's not. There are other ways. For example, entering another belt from the side, using 2 belst run in paralell and smart inserters, (If your factory clogs, you have an entire different problem, which in turn can be fixed by some chests acting as buffers)

So no, I don't think that a loop + 4 smart inserters is a good stand-in for something which should be easily accomplished with one tile and little to no energy cost.

That's what you and the people here aren't getting. Factorio is not an easy game, and it wasn't meant to be so!!!

Maybe there are other designs using smart inserters which are error-proof. But that's not really the point. If someone can show me a design which uses two or less smart inserters (not underground belts) and requires no more than 3x3 tiles then they might have an argument. Anything more than that is, in my opinion, not really a solution for such a simple problem. It's an over-engineered, energy inefficient mess that cannot be replicated frequently without a serious drain on energy and space, for performing a task which should be incredibly easy to do.

Exactly my point. It is not meant for you to have a 2x2 factory. The idea of this game is to FORCE you to expand, and to find the way for everything to work automated and without clogging.
Would you also like a preset blueprint that can create the whole factory for you? "This would make the game easier so you don't need to think so much on how to fix a problem.

That is, I think, where the frustration lies. There is this enormous gap between how simple a task belt separation ought to be, and how complex and over-engineered the proposed solutions are.

Wellcome to Factorio! THIS is what the game is all about! That's the beauty of it, and if you spend a few extra hours playing it (IDK if 100h, but 20 or 30 maybe) you'll come to realize it!

You cannot walk into a factory without seeing a belt separator in some form or another. Sometimes they are complex things and sometimes they are literally just a metal sheet which divides a belt into halves, but they are never as ridiculously over-engineered as are required in the game.

How many of those factories are in another world, and have lasser turrets around and get attacked by aliens?

I agree that this item could make the game a lot easier, and that's why I strongly oppose to the idea.
What nobody seems to get, is that when ssilk says

[...]when I see some suggestions which are obviously made, because not knowing differently, then I would really say: go on playing and when you really understood come back. The 100 hours is 5-10 more or less normal games.

He means that you'll realize that there are a handfull of situations where you'll use only one side of the belt for one thing, and the other for another.
My 1st game, I was trying too hard to use each side of the belt for a different thing. I did the "coal/ore" for the smelters, the "iron/copper" as main belt... and around the 10 or 15 hours of gameplay i realized that it was never gonna be enough with only one side.
Since then, i don't use 1 side of the belt EXCEPT in very specific places: Labs, or late tier things that need lots of time to produce (ex: "batteries/plastic" or "electric engines/robot frames"), because I already know that if i do use one side of the belt, sooner rather than later, I'll have to remake the WHOLE thing.
Even worse, if the throughput of only 1 side of the belt works for you, you have a deeper problem in your chain, but you probably haven't realized that yet.
Don't warry, if you follow ssilk's advice and play a few more hours, you'll see it. If not, please, PLEASE! share your factory design, because I have a lot to learn from you.

But again, The main issue here is that devs are ruling against this at least for as long as I am in the forum. They did time and time again, so you're wasting your time and breath, which i don't really care.
But you're also wating the devs and mods time, and that is really valuable.

Again, I wasn't against this before, I even sugested it!: In THIS THREAD and also in THIS OTHER.
Then I played Factorio, watched some LP's (OfficialStuffPlus is GREAT) and discovered to my surprise that I WAS WRONG
Just go and play a little bit more. If in a week's play you still think the same, come and say it again. But you probably won't.

[BurnHard]

The very first mod for this: https://forums.factorio.com/forum/vie ... =14&t=3657

[Garm]

Again, I wasn't against this before, I even sugested it!: In THIS THREAD and also in THIS OTHER.
Then I played Factorio, watched some LP's (OfficialStuffPlus is GREAT) and discovered to my surprise that I WAS WRONG
Just go and play a little bit more. If in a week's play you still think the same, come and say it again. But you probably won't.

And that is why I think if you look at age of "Pro" and "against" crowd (which by the way are equal in size more or less) One will notice that predominantly "Pro" group has registered mostly in May 2014, some in April. While "anti" supporters have registered earlier.

This is not because since we are so old we are conservative - It because we had similar ideas is the past and we realized, that we were wrong....


P.S. if your smart inserter setup is clogging up - you design is probably lacking something. Either appropriate termination of the belts to enforce limitation, or insufficient amount of inserters to move all available items.

[DrNoid]

That is, I think, where the frustration lies. There is this enormous gap between how simple a task belt separation ought to be, and how complex and over-engineered the proposed solutions are.

Wellcome to Factorio! THIS is what the game is all about! That's the beauty of it, and if you spend a few extra hours playing it (IDK if 100h, but 20 or 30 maybe) you'll come to realize it!

Exactly my point. It is not meant for you to have a 2x2 factory. The idea of this game is to FORCE you to expand, and to find the way for everything to work automated and without clogging.

Then you should actually be in favour of removing the current splitter. It is a very complex, specialised thing yet its function can be implemented by chaining 2 very simple operations together:

• The lane splitter (a simple piece of metal that pushes one lane off the belt)
• The lane re-balancer. (takes the input lanes, and outputs the items neatly distributed over the output lanes)

Individually those two operations are easier to understand than the current splitter, yet by combining those two you can cover all functionalities of the current splitter.

... philosophy of Factorio when you are supposed to work out your solution from simple components...

I think a lot of the splitter discussions that keep popping up are fundamentally caused because the current splitter violates the "simple tools that do 1 thing, and do them perfectly" idea that Factorio is based on, yet we miss the simple tools that are actually the basis of its functioning. And in most cases the current splitter doesn't actually perform quite as we would like it to. Lets look at the current use-cases for the splitter (if I miss one, please post, with how you would solve it with the above two mentioned simple tools):

• A belt with 1 lane filled needs to be split in two: 1 re-balancer to put stuff on both lanes, followed by a lane-splitter. This doesn't even need the complex "lift items over a lane to the other belt" functionality of the current splitter.
• A belt with 2 lanes, filled with the same stuff, needs to be split in two belts, each with only 1 lane: 1 lane-splitter, done. In this case it's even simpler than the current splitter, since that puts stuff on two lanes.
• A belt with two lanes of different things needs to be split in two belts with two lanes of different things: 1 lane-splitter, each side followed by a re-balancer and another lane-splitter. Then 1 underground belt to swap the places of the middle two belts, and finally you have the left-two merge and the right-two merge. This is the use-case the current splitter was made for, yet it is still pretty easy to do with simpler tools.
• A belt with two lanes of different stuff, need to be split into two belts, each with only 1 type: 1 lane-splitter. This use-case requires the ugly underground-belt trick at the moment.
• A belt with stuff mostly on one lane (that fills up too much) needs to be spread better over both lanes: 1 re-balancer. For re-balancing purposes the current splitter only works well if one lane is completely empty. If both lanes have things, the split-and-merge trick doesn't work very well.
• Merge stuff from two belts onto one: A T-junction of belts, with a re-balancer if need be. I have never used a splitter for this.

[Sedado77]

That is, I think, where the frustration lies. There is this enormous gap between how simple a task belt separation ought to be, and how complex and over-engineered the proposed solutions are.

Wellcome to Factorio! THIS is what the game is all about! That's the beauty of it, and if you spend a few extra hours playing it (IDK if 100h, but 20 or 30 maybe) you'll come to realize it!

Exactly my point. It is not meant for you to have a 2x2 factory. The idea of this game is to FORCE you to expand, and to find the way for everything to work automated and without clogging.

Then you should actually be in favour of removing the current splitter. It is a very complex, specialised thing yet its function can be implemented by chaining 2 very simple operations together:

• The lane splitter (a simple piece of metal that pushes one lane off the belt)
• The lane re-balancer. (takes the input lanes, and outputs the items neatly distributed over the output lanes)

Individually those two operations are easier to understand than the current splitter, yet by combining those two you can cover all functionalities of the current splitter.

... philosophy of Factorio when you are supposed to work out your solution from simple components...

I think a lot of the splitter discussions that keep popping up are fundamentally caused because the current splitter violates the "simple tools that do 1 thing, and do them perfectly" idea that Factorio is based on, yet we miss the simple tools that are actually the basis of its functioning. And in most cases the current splitter doesn't actually perform quite as we would like it to. Lets look at the current use-cases for the splitter (if I miss one, please post, with how you would solve it with the above two mentioned simple tools):

• A belt with 1 lane filled needs to be split in two: 1 re-balancer to put stuff on both lanes, followed by a lane-splitter. This doesn't even need the complex "lift items over a lane to the other belt" functionality of the current splitter.
• A belt with 2 lanes, filled with the same stuff, needs to be split in two belts, each with only 1 lane: 1 lane-splitter, done. In this case it's even simpler than the current splitter, since that puts stuff on two lanes.
• A belt with two lanes of different things needs to be split in two belts with two lanes of different things: 1 lane-splitter, each side followed by a re-balancer and another lane-splitter. Then 1 underground belt to swap the places of the middle two belts, and finally you have the left-two merge and the right-two merge. This is the use-case the current splitter was made for, yet it is still pretty easy to do with simpler tools.
• A belt with two lanes of different stuff, need to be split into two belts, each with only 1 type: 1 lane-splitter. This use-case requires the ugly underground-belt trick at the moment.
• A belt with stuff mostly on one lane (that fills up too much) needs to be spread better over both lanes: 1 re-balancer. For re-balancing purposes the current splitter only works well if one lane is completely empty. If both lanes have things, the split-and-merge trick doesn't work very well.
• Merge stuff from two belts onto one: A T-junction of belts, with a re-balancer if need be. I have never used a splitter for this.

IMO, you're not giving any new functionality AND you're replacing only 1 entity with 2... there (again, IMO) is nothing of what you say, you can't do with splitters as they are: (I'd post screenshots if I could, but I'm @work and internet security doesn't allow me)
For Reference: I put "side A/B" for the output sides of the splitter and "X/Y" for the input.

• A belt with 1 lane filled needs to be split in two(a. belts/b. lanes):
  a. Splitter. Done.
  b. Splitter + a belt rejoining in the empty side.
• A belt with 2 lanes, filled with the same stuff, needs to be split in two belts, each with only 1 lane:
  Splitter, then make a "T-junction" for feeding side A of the output of the splitter into a side of the belt and the side B feeds an(x number of) inserter that places items on the other belt.
• A belt with two lanes of different things needs to be split in two belts with two lanes of different things:
  Splitter. Done.
• A belt with two lanes of different stuff, need to be split into two belts, each with only 1 type:
  Use the Underground belt "hack"+ splitter.
  Or, even easier, smart inserter(s).
• A belt with stuff mostly on one lane (that fills up too much) needs to be spread better over both lanes:
  b. Splitter + a belt rejoining in the empty side. Also you can cascade more of these as you seem fit (I have 4 belts away from my iron furnances to cope the throughput.)
• Merge stuff from two belts onto one:
  Splitter with only one way out, or T junction as you said.

As you see, in 6 cases, you only made 1 simpler.
For the rest, we already have the tools for doing them.
IMO this is not necessary. It doesn't add to the game and doesn't help in any way.
Oh, sorry, It does, In only 1 way.

I'm in favor of Splitters. I like them. If i could I'll marry one

Edit: Quoteception! xD

[DrNoid]

(... A lot of repeating what I said...)

IMO, you're not giving any new functionality AND you're replacing only 1 entity with 2... there (again, IMO) is nothing of what you say, you can't do with splitters as they are: (I'd post screenshots if I could, but I'm @work and internet security doesn't allow me)
For Reference: I put "side A/B" for the output sides of the splitter and "X/Y" for the input.

(... More mostly repeating what I said ...)

As you see, in 6 cases, you only made 1 simpler.
For the rest, we already have the tools for doing them.

It's not about using less tools, it's about using simpler tools with a much clearer behaviour. The current splitter is a black-box with some weird behaviour, which is why you can not see what happens on the inside. Combining simpler tools is better than using a complex tool, as stated by a dev.

You ignored the fact that I wrote that lane-rebalancing only works well with the current splitter if the input only has 1 lane filled, but here is a much better example of splitter weirdness:
A belt with 1 item-type on it, perfectly balanced on both lanes (left-right-left-right, etc.). Split that and I would expect two belts with perfectly balanced lanes, yet that is not what happens. But if you split a belt with two item types on it (f.i. metal left, copper right), in the exact same pattern (left-right,left,right, etc.), you do get the expected output.

Using simpler tools makes all use-cases simpler (where simpler means more predictable & easier to follow), because you can actually follow the individual items.

[Gammro]

It's not about using less tools, it's about using simpler tools with a much clearer behaviour. The current splitter is a black-box with some weird behaviour, which is why you can not see what happens on the inside.

Using simpler tools makes all use-cases simpler (where simpler means more predictable & easier to follow), because you can actually follow the individual items.

Let me just explain the simplicity of the splitter. All "weird" behaviour you claim happens, stems from this basic behaviour, making them very logical results:

The splitter tracks where the last item was sent to. If it was left, the current item will go to the right. If it was right, the current item will be sent to the left. All items stay on the same side of the belt.

That's it.

There's only one addition, and that is when one of the outputs is full. Then all items will go to the other output.(Which to me would seem logical behaviour)
Now let's go through your examples:

A belt with 1 item-type on it, perfectly balanced on both lanes (left-right-left-right, etc.). Split that and I would expect two belts with perfectly balanced lanes, yet that is not what happens.

If there's time between items that get inserted(so no items at the same time on the input), this is very clear. This is a result of the above stated rule. If the 1st item is on the left side of the belt, it would be sent to output 1. The next item, regardless of on what side of the belt it's on, gets sent to output 2. So when you send items in a perfect ratio, all items on the (in this example) left side after that are at the odd(1,3,5,7 etc) point where the inserter will send it to output 1. And all the items on the right side, will be input at all even(2,4,6,8 etc.) points where it will be sent to output 2.

But if you split a belt with two item types on it (f.i. metal left, copper right), in the exact same pattern (left-right,left,right, etc.), you do get the expected output.

This is not possible, the splitter doesn't look at what is inserted. So with the exact same input pattern, the same output would happen. My suspicion is that the belts are filled in this example, versus empty(ish) in the previous one.

Behaviorally, the splitter is VERY simple.