Factorio Forums

Hi all,

TL;DR

Should the the item drop position when dropping onto a curved belt be as shown in the wiki (and in the game) or should the drop position be consistently the farthest right quadrant?
https://wiki.factorio.com/Inserters

What?

It is one of those oddities that I have simply accepted up to this point, but now that I have found the forums, I would like to ask why the inserter drop position is the way it is for a curved belt and also ask whether the users find it logical or do you 'trip over' this oddity?
Current drop position on a curved belt shown as '#', my expected drop position shown as '?'
Is the drop position in the third case what one would expect and what are the benefits?

If there is a good reason for it, I would like to understand as it seems 'odd' to me.

Why?

It is, of course, not a big deal, but I believe it to be an inconsistent behaviour that all users will trip over until they get used to it. If the inserter was always to drop on the farthest right quadrant, the behaviour would be consistent and easier for new players to understand (and easier for my simply brain to understand!)

I could be wrong, can’t hop into the game to double check right now, but I believe the inserters have a consistant behavior of dropping to the furthest left quadrant. I believe if you were to reverse that third example so the bottom inserter had the straight and the upper the curve, you should be able to see this. Changing it to the furthest right will just result in the other direction having this same issue.

Edit: In case you haven’t seen it, this mod is very iseful for solving these types of situations: https://mods.factorio.com/mod/Side%20Inserters

FuryoftheStars wrote: ↑

Mon Jun 03, 2019 1:43 pm

I could be wrong, can’t hop into the game to double check right now, but I believe the inserters have a consistant behavior of dropping to the furthest left quadrant. I believe if you were to reverse that third example so the bottom inserter had the straight and the upper the curve, you should be able to see this. Changing it to the furthest right will just result in the other direction having this same issue.

Hi,

If you look in the image I included or the linked wiki page, you can see that, for straight belts, the inserter drops to the right lane of the belt, or the farthest lane of the belt - what I have suggest is that visualising this as "farthest right quadrant" is correct for straight belts, but the inserter swaps to dropping the items on the left when the belt is curved.

SkiCarver wrote: ↑

Mon Jun 03, 2019 1:48 pm

FuryoftheStars wrote: ↑

Mon Jun 03, 2019 1:43 pm

I could be wrong, can’t hop into the game to double check right now, but I believe the inserters have a consistant behavior of dropping to the furthest left quadrant. I believe if you were to reverse that third example so the bottom inserter had the straight and the upper the curve, you should be able to see this. Changing it to the furthest right will just result in the other direction having this same issue.

Hi,

If you look in the image I included or the linked wiki page, you can see that, for straight belts, the inserter drops to the right lane of the belt, or the farthest lane of the belt - what I have suggest is that visualising this as "farthest right quadrant" is correct for straight belts, but the inserter swaps to dropping the items on the left when the belt is curved.

Yes, I understood that.

In your image with the curved belt, you had the belt coming down, then curve right. I was suggesting that if you had the belt go up, then curve right, the top inserter may place the item on the "left" side.

However, I did just have a quick minute to fire up the game and I was wrong. The inserter switches sides and drops on the "right" again.

As such, I would agree with you that this does not seem intuitive and would almost wonder if this could be classified as a bug.

I think the inserter should always drop on the farthest left or right (doesn't matter which but it should be consistent) and that should be from its perspective.

Anyone who votes "It's logical" should explain what the consistent behaviour is. That option should require an explaining comment. Or we can just discard any such vote that lacks it.

Because I see none and wish there was a way know where items will land except just memorising random configurations, testing every time or just never place on any belts except those that have a well behaved far side.

Qon wrote: ↑

Mon Jun 03, 2019 5:44 pm

Anyone who votes "It's logical" should explain what the consistent behaviour is. That option should require an explaining comment. Or we can just discard any such vote that lacks it.

Because I see none and wish there was a way know where items will land except just memorising random configurations, testing every time or just never place on any belts except those that have a well behaved far side.

The way I see it is the inserter places it dependant on how the belt starts. On the LH bend in the example it is going down before turning left (within it's square) so therefore logical to place it on the far side.

Thanks to the posters ... I have improved my understanding .... but I still have a question!
From the examples, I can see that my view of understanding the placement based on the inserter was only partially correct. The placement is ALWAYS on the opposite half of the game tile from the inserter - this is why we do not get a split output in case S in the image.

The part where I was wrong, is that the decision on which side (left or right with respect to the inserter) is based on the belt, not the inserter. The inserter will change the position where it will drop based on the direction of the belt .... this is the 'non-intuitive' part of the placement for me. I assumed that the drop position of the inserter would be a property of the inserter, but it is not. Rather than the inserter always dropping in the opposite right quadrant of the game tile, it is the belt that defines the placement ....

So the remaining question(s);

Would it be better for the inserter to drop in the same place regardless of the belt, or should it continue to adjust to try and be consistently dropping on the right side of the belt (from the belts perspective)?

Are we, the community, willing to put in the time to fix all our games and blueprints that are broken due to changing the inserter drop location (assuming the devs would be interested in such an action!)?

A weird consequence is that inserters inserting on underneathies from the north and south on a north or south going belt will both output to the same side. So there's 3 directions that output on one side and only one that inserts on the other. So even if you know about the weird behaviour you can't easily compensate for it in cramped belt layouts and get unbalanced lanes and half the belts capacity. Same with miners. Which is another argument for never bothering with belts for megabases and use trains+bots only.

Qon wrote: ↑

Tue Jun 04, 2019 12:25 am

A weird consequence is that inserters inserting on underneathies from the north and south on a north or south going belt will both output to the same side. So there's 3 directions that output on one side and only one that inserts on the other. So even if you know about the weird behaviour you can't easily compensate for it in cramped belt layouts and get unbalanced lanes and half the belts capacity. Same with miners. Which is another argument for never bothering with belts for megabases and use trains+bots only.

Also the drop position changes between a tile with no belt and a tile with belt. That makes no sense.

Wisey wrote: ↑

Mon Jun 03, 2019 10:48 pm

The way I see it is the inserter places it dependant on how the belt starts. On the LH bend in the example it is going down before turning left (within it's square) so therefore logical to place it on the far side.

Wrong.
I & P:

SkiCarver wrote: ↑

Tue Jun 04, 2019 12:08 am

Still no correct explanations.

It appears that the inserter treats a curved belt section as moving perpendicular to the inserter. It puts the item so that it will be on the far side of the belt. This keeps items on the same side of the belt as an inserter to the same belt next to the inserter on the curve.

From placements.png this matches:
M: The branch from the left
N: The branch from the right
O: The branch going left
P: The branch going Right

And from Quadrant.png, the inserters for the pipes and underground pipes should be on the same side because they are right next to each other.

I'm not sure I explained this well, but two inserters right next to each other putting stuff on the same belt should put it on the same side of the belt.

SkiCarver wrote: ↑

Tue Jun 04, 2019 12:08 am

Would it be better for the inserter to drop in the same place regardless of the belt, or should it continue to adjust to try and be consistently dropping on the right side of the belt (from the belts perspective)?

I think so. But maybe not for splitters or Q & R would get different behaviour (one would be like S) which would also be a bit weird.

Qon wrote: ↑

Wed Jun 05, 2019 8:20 am

SkiCarver wrote: ↑

Tue Jun 04, 2019 12:08 am

Would it be better for the inserter to drop in the same place regardless of the belt, or should it continue to adjust to try and be consistently dropping on the right side of the belt (from the belts perspective)?

I think so. But maybe not for splitters or Q & R would get different behaviour (one would be like S) which would also be a bit weird.

I think I would be OK with the weirdness ... having a simple rule like; "Inserter drops in the far right quadrant." makes it much easier to predict what will happen and would, i believe, be easier for beginners to pick up.

SkiCarver wrote: ↑

Wed Jun 05, 2019 8:30 am

I think I would be OK with the weirdness ... having a simple rule like; "Inserter drops in the far right quadrant." makes it much easier to predict what will happen and would, i believe, be easier for beginners to pick up.

It's actually very easy to set up, just shift the output position from the middle to any side. I've tried to make it here

In this poster i wouldnl't call situation E F G H I J M N O P, nor logic or intuitive, if you take one of the group EF, GH , IJ, MN, OP.

You can't take one of the letter and assume a physical position in space where the insererter drops the item.

It is not the same exact spot depending on the belts direction which to me is not a logical behavior if we consider a model for physical reality.


But when i started playing i felt like those were the rules you discovers and feel like you've progressed once you can makes tricks using that knowledge.

Yet today i couldn't tell you how one inserter would behave without going into the game, and i've played for couple thousands hours.

But when ingame it feels intuitive, and i've learned how to quick test and remove very fast and when to do it so it doesn't feels a problem at all anymore, just 'character' as the poll suggest.

I didn't tried to have a worded rules that was intelligible such as ' always the furthest right quadrant '. ( which i still don't understand how it explain situation O or M for example, why would have I to think about the direction of the belt being perpendicular to understand the inserter behavior ? except for the coding ).

Qon wrote: ↑

Mon Jun 03, 2019 5:44 pm

Anyone who votes "It's logical" should explain what the consistent behaviour is. That option should require an explaining comment. Or we can just discard any such vote that lacks it.

Because I see none and wish there was a way know where items will land except just memorising random configurations, testing every time or just never place on any belts except those that have a well behaved far side.

That's part of my fun, so i wouldn't really mind a change and redoing blueprints if it was for that clear explanation/behavior, that may be simple enough to fit in a tooltip.

But explaining things in game like "in this situation you need to understand the rules this way" rather than just having access to a picture like the poster above ( which you replicate the part you need in 5 second ), a picture , thousands words .

I explain my vote as 'not in favor of the change' because i like the way it is even 'illogical' , I don't think the new more logical behavior proposed would be much benefit ( for me ).

But it wouldn't be much trouble either, for a +/- average i'd rather see other things get priority.

It isn't intuive, it becomes !

I think its pretty intuitive if you come from the belt perspective

Straight belt: Far side if orthogonal, right side of belt direction if parallel
Curved belt: inner corner

only parallel splitters look weird at first glance but if you think about that far-side thing it becomes clear^^

terror_gnom wrote: ↑

Wed Jun 05, 2019 1:57 pm

I think its pretty intuitive if you come from the belt perspective

Straight belt: Far side if orthogonal, right side of belt direction if parallel
Curved belt: inner corner

only parallel splitters look weird at first glance but if you think about that far-side thing it becomes clear^^

Except that doesn’t really make much sense, does it? Shouldn’t the inserter be dropping its item(s) in the same exact spot, regardless of which way the belt under it is facing?

If the belt is parallel to the inserter, the item is dropped on the right side of the belt (from the belt's perspective).
If the best is perpendicular to the inserter, the item is dropped on the far side of the belt.
Corners are considered "perpendicular," and as such the item is dropped on the "far" side of the belt, which is considered to be the inside of the corner.

In your examples you are forgetting belts curving towards the inserter. While that rarely makes sense, since the belt then stops, it sometimes is useful to prevent the belt running between 2 lines of assemblers from flowing into the next tile where a different belt runs.