[0.9.0] Inserter getting stuck feeding into chem plant

[Toawa]

I've noticed that sometimes, an inserter that is feeding a solid item into a chem plant will get "stuck", where it will stop feeding into that plant, even when there's room for more input. I've noticed this most often in a setup I have for producing sulfuric acid, where I have an inserter feeding in sulfur from an adjacent chem plant. I suspect it has something to do with when the inserter stops feeding in inputs to a factory because its output is full; I've seen the same behaviour in assembly plants, where inserters will stop feeding inputs when the output is full (and being pulled out by another inserter, so the output is 2-3 instead of a full stack). One way to "unstick" the inserter is to either add one of the stuck input manually, or remove the stuck output from the other chem plant.

I suspect this has something to do with an earlier optimization made to inserters that turns them off until their input or output side flags that there might be something for them to do; the chem plant might not be flagging them as such. I think what's happening is, the inserter feeds the acid plant until its acid fills all the way up, and then stops, and even when the acid goes back down to 0, it never gets the signal to start again until I manually change something about the setup.

(I don't know if this is something that can be replicated in a save file, especially considering the other bug that I found, but if you want to see it, my acid plant is in the save file for the train bug I posted about earlier.)

[FrozenOne]

Also had this bug, with sulphur loading inserter indeed.

[doublewohli]

I have made the same experience. With Sulfur as well and with Crude Oil Barrels. How it happens is described by Toawa really good.

[Rockstar04]

I just want to chime in on this as well.

Sound like I had the same setup,sulphur into another chem plant making sulfuric acid. If I removed the stack of 64 sulphur from the first plant and reinsterted it, everything would start back up fine.

[crazybmanp]

i can confirm that this happens with several recipes, but i believe it has to do with the liquid input/output stopping the machine possibly. (running out of petrol, running out of room for sulfuric acid or oil)

[kovarex]

Thank you for the pointers, it was now reproduced and fixed for 0.9.1

Details

There is optimisation, that puts inserter "to sleep" when it has nothing to do.
The problem is with the chemical plant that has full output of the sulfuric acid, so it doesn't allow the inserter to put more, so it is set up to sleep.

But then the acid flows away and the assembling machine doesn't notice that it can process again, so it doesn't tell the inserter to "wake up".

Now the assembling machine checks if the fluid output didn't get under the amount that allows to work on another item in the update and if so, it wakens up the inserter.