Inserter not keeping the lane config when fast-replaced in 2.1.8 and 2.1.9

[leagaleano]

Dear dev team,

I'm testing the Factorio 2.1.X Experimental release, and found a bug in the new inserter's lane mechanics; when upgrading/downgrading, the feature that allows to select a lane in 90° belts, is lost if the inserter type is changed.

How to reproduce the error:
Put a source of materials
Put an inserter that moves the content to a 90° belt
Change lane by clicking V/H (depending if Vertical or Horizontal)
Upgrade/Downgrade the inserter
The lane config is lost.

Video sample: https://1drv.ms/v/c/99f5e1a0eb47f0b7/IQ ... M?e=3sLq66

[dalindes]

This seems to have been flagged as a duplicate, but I'm failing to find what it's a duplicate of, and I'd love to know about status... could someone link the main report, please?

[Rseding91]

I believe this: viewtopic.php?p=695184#p695184 but I might be wrong.

[Rseding91]

Actually I think this may be a bug. Following how splitters work - the inserter should be preserving these values during fast-replace and is currently not.

[Rseding91]

Actually actually, looking into this more: since you can configure the inserter drop lane in this scenario before fast-replacing the existing inserter - it is correct that it always takes what the cursor has been configured with.

[Shadow_Man]

Actually actually, looking into this more: since you can configure the inserter drop lane in this scenario before fast-replacing the existing inserter - it is correct that it always takes what the cursor has been configured with.

I totally disagreed: when you place inserter over existing inserter it acts like manual upgrade (in player's expectations), so newly placed inserter must preserve existing lane orientation.

[boskid]

Its not that inserter is told that it prefers one lane or the other, it is the inserter that is in its normal orientation or mirrored orientation, lane preference is just a side effect. Having inserters keep mirroring of the ghost would create a huge inconsistency in build logic: recycler would have to use cursor mirroring, burner mining drill would have to use cursor mirroring but electric mining drill would have to use ghost mirroring and inserter would have to use ghost mirroring. Common part here is that when mirroring of the cursor is visible through the entity graphics it is "ok" but if its not visible it is being claimed to be a bug. Try building a recycler over a recycler ghost and try to tell if it should keep mirroring of the ghost or mirroring of the cursor.

[robot256]

Common part here is that when mirroring of the cursor is visible through the entity graphics it is "ok" but if its not visible it is being claimed to be a bug.

You may not call it a bug, but it is definitely user-hostile interface design. The user should be able to know what the outcome of an action is before they perform it, without remembering in their head how many times they have pressed each flip and rotate key.

I'm an idiot for posting things without testing them first.

[Filcius]

Actually actually, looking into this more: since you can configure the inserter drop lane in this scenario before fast-replacing the existing inserter - it is correct that it always takes what the cursor has been configured with.

That allows the player to set the lane before replacing, it make sense. It is annoying in most cases, but giving the control of the lane before building seems more important.

[boskid]

The user should be able to know what the outcome of an action is before they perform it.

Player is told what the outcome will be - arrow is drawing on the other lane and yet player decided to build an inserter with arrow in a different spot.

07-23-2026, 19-22-31.png (121.33 KiB) Viewed 149 times

[robot256]

The user should be able to know what the outcome of an action is before they perform it.

Player is told what the outcome will be - arrow is drawing on the other lane and yet player decided to build an inserter with arrow in a different spot.

07-23-2026, 19-22-31.png

Thank you, this is what I was hoping for. I am sorry for my incorrect comment.