[Rseding91] [2.0.28] Mixed fuel ends up in trash slots

[Quintuple5]

As reported on the Pyanodon Discord, one user somehow ended up with this:

afbeelding.png (412 KiB) Viewed 1124 times

You're seeing a Coal powerplant, a building that takes different kinds of coal and heats up molten salt (one input, one output). The input belt is a mixed belt of 3 types of fuel (all visible in the picture), which is a common scenario in Pyanodon. Somehow, the fuel sometimes ends up in what appears to be the trash slots of the building.

I've attached the save provided by the reporter. If you load it up, click on the building, you will see the trash slots indeed slowly fill up. Every few seconds you can see the number going up, sometimes it takes a bit.

I have no idea how to reproduce it from a new save however. If I make a Coal Power Plant in a new save and give it a mixed belt, the inserters who don't "win" patiently wait for the fuel type to be used up before they insert their item into the fuel slot. If I had, I would have made a save with as few mods as possible to make debugging easier, sorry

To fulfil the stickied post: I would not expect the fuel to end up in the trash slots of the building, but instead for the inserters to wait for the fuel slot to become empty (which is the behavior I see in a new save).

Note that all feature flags are on due to the "enable all feature flags" mod, but none of the items involved here spoil.

As a side note, I noticed that hovering over the fuel slot when it's empty shows a set of fuel inputs that do not include "crushed coal", which as can be seen in the save clearly is a valid fuel for this building. Should I make a separate post over that?

[Muche]

All Coal powerplants in that block are affected by the issue.
If an inserter is holding something, it will drop it into the machine, into a new trash slot, if needed (even if the held item is totally unrelated to fuels, e.g. a belt).
As there are multiple inserters feeding the various fuels, when the fuel slot is empty, all inserters activate, pick up whatever is on the belt and drop it. If multiple fuels were picked up, a trash slot is created/inserted into.

However, I have no idea how the machines got into that state.

Placing a new coal powerplant (even in the same map) does not exhibit the issue, inserters are waiting when holding incompatible items.
With an inserted Coke I was able to create one trash slot (as it is a spoil result of Red-hot coke - see viewtopic.php?p=647189 for similar behaviour).

The multiple trash slots symptom reminded me of 123664 Crusher adds more and more trash slots, but I did not manage to get a machine into the broken state by using similar actions.

I can see Crushed coal in the list of fuels for Coal powerplant (but also see the note below).

125212-Py-CoalPowerplant-Fuels.jpg (19.95 KiB) Viewed 1059 times

I should also note that my mods configuration very likely does not match savefile's one. To even get spoiling into Coke, Py Alternative Energy's Enable decay startup setting has to be enabled (disabled by default).
I have no idea which other settings might be different and are needed to reproduce the issue in a new save.

[Bacoid]

I'm getting some similar issues, may not be entirely relevant but may also shed some light on what's happening. I'm also in Py but I've been able to reproduce my issue in base game + Space Age.

My original issue, in Py, was that my copper smelting stack occasionally had a piece of coke on it's output belt. Coke is just another fuel in Py and I was fueling said smelting stack with 4 different fuels (including Coke) on the same mixed belt.


(Note the piece of Coke towards the right hand side of the output belt)

Was it just coke polluting the output belt and not the other fuels on my mixed belt? Yes.
Why just coke? ...because it's a spoilage result.

How is it happening?
1. Furnace runs out of fuel
2. Two different inserters grab two different fuels, then attempt to put them into the same furnace at the same time
3. The inserter that happens to drop it's fuel into the furnace first, drops it into the fuel slot as expected
4. The inserter that ends up being second to drop it's fuel into the furnace, if that fuel is a spoilage result... the furnace will accept it and immediately put it into it's trash slot.
Note: this requires there to be at least three different fuels present in the situation, one in the furnace which runs out, and two more, one coming from each of the two inserters.

https://imgur.com/4wWNBCw

If I'm interacting with the spoilage mechanic then I should be mindful on spoilage things right?
I'm not interacting with spoilage at all actually. In Py, Coke is the spoiled product of Red-Hot Coke. I cannot make Red-Hot Coke, nor is it even researched! I am simply producing Coke from Coal and using excess to fuel my furnaces.

Are other items/fuels affected? Yes. Any fuel that is a spoilage result will act exactly like this... which means... I can reproduce this in base game Space Age, since the 'spoilage' item is a spoiled result AND a fuel.

https://imgur.com/IJBtZGO

Again, I'm not technically interacting with the spoilage mechanic. I imagine if I have a Space Age base on Nauvis and I have some spoilage laying around and I decide to send that to my smelting as fuel, if there happens to be the three different fuels and two inserter input requirement, then spoilage can and will pollute the output belt!

For anyone else having this issue, my temporary solution is to filter my input inserters to: one inserter filtered to the item to be smelted and the other inserter filtered to *not* that item, aka the fuel inserter. This insures that only one inserter can input fuel into the furnace.

[Quintuple5]

It also works with mixed smelting setups if Space Age is on (and no other mods), because then iron and copper ore are both spoilt results and therefore may end up in the trash slot:

afbeelding.png (532.9 KiB) Viewed 855 times

I synchronized the two inserters to insert at the same time by removing the substation, cleaning the setup, and placing the substation back down (being careful that the inserters are not mid swing). Two things can happen:
- if stack size of the inserters is 1, one of iron and copper will win and be turned into a plate, and the losing item ends up in the ingredient slot to be used the next smelting cycle
- if stack size of the inserter is 2, the losing item will end up in the trash slot and will be taken out by the output inserter.

[Rseding91]

Thanks for the report. This is now fixed for the next release. Due note though: the inserters will get stuck not being able to put the differing fuel into the machines.

[AwesomeReally21]

It seems the problem is still there.

I updated the game to 2.0.30 this morning. Still saw the fuels in the trash slots after few hours.

01-10-2025, 16-45-39.png (330.39 KiB) Viewed 230 times

[Loewchen]

It seems the problem is still there.

I updated the game to 2.0.30 this morning. Still saw the fuels in the trash slots after few hours.

Post your complete log file and a simple save file showing the issue please, see 3638.

[Bacoid]

Update from me after more testing, furnaces with fuel that's also a spoiled result all seem to be fixed, but it appears that was indeed a different bug than the mixed fuel one in Py's Coal Powerplant. This mixed fuel bug has nothing to do with the spoilage mechanic from what I can tell.

It seems it's possible for assemblers to be put into a 'bugged state' where they accept more fuel than they should and put excess into their trash slots instead of having the inserters wait patiently for space. The only requirement other than the bugged state is having inserters put more than one 'slot' of fuel into an assembler at once, either by putting two different fuels in at once (in the case of Py), or... by putting more than one stack of nutrients into a biochamber at once. Yes I have once again made it a base game + Space Age bug!

https://i.imgur.com/dkXwkv9.gif

Also, when I say assembler, I mean assembler prototype. Biochamber and Py coal powerplant included!

I have found one manual way to put the assembler into this bugged state but there appears to be a second, natural way, as the people I have spoken to are insistent that they have not ever done this manual method.

The manual way I have found is pretty simple:
1. Attach a wire from the assembler to literally anything else
2. That's it, it's done

It should be in it's bugged state now, the circuit conditions don't seem to matter and *you don't even need to keep the wire*. You also cannot undo the bugged state once it's bugged... unless you copy a non-bugged assembler onto the bugged one. Yes the bugged state *can* be copy-pasted, and yes you ALSO can blueprint said bugged state, so here, have this blueprint. (base game + Space Age blueprint)

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

Copy blueprint

You should notice right away that the left Biochamber forces it's inserters to wait patiently, while the right Biochamber just accepts everything and throws excess into it's trash slots! The left Biochamber is indeed the non-bugged one, while the right Biochamber has been put into it's 'bugged state' manually with a wire, with the wire removed afterwards.

[Rseding91]

That case is unfortunately working as intended. If it didn’t work that way, changing recipes on assembling machines would cause inserters to deadlock if they were mid swing when it changed.

If you’re going to use circuits with assemblers that accept fuel then you need to not feed them mixed fuel.