[SPZ2-4833] Swappers with unknown inputs from outside platform slows down conveyor belts

[hot hinge]

Sorry for the weird title, but I don't know how else to describe the issue. Please see the screenshot. The platform in question is fed with full miners from the bottom and fed into full trashes at the top. If you have a long conveyor belt that's connected perpendicularly to a lot of swapper outputs, then there is some slowdown on the belt that's touching the swapper outputs. This only happens if the swappers are currently not being fed due to the catchers being disconnected off platform - you can see that the leftmost belt does not have any slowdown despite being connected to the same set of buildings except the catchers being purely inside the platform. The same problem happens if the right side is attached to a space belt that's not being fed at the source.

I'm on 0.1.1 on MacOS. Here is the blueprint in text:

SHAPEZ2-3-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$

[analog igloo]

Thank you for your report @hot hinge !

I seem to be unable to reproduce it on current master build using a very basic producer > BP > trash build as shown in the screenshot below. I also notice some gaps in the right production. Could you check the whole BP to see if there might be gaps in the production speed ?

[hot hinge]

Thanks for the response. By gaps, do you mean the count "142" at the bottom right? That is because of the pushback. It goes up to 180 as soon as I remove the catchers. If I look at the space belt, I can see the right lane stutter but there is never a gap on the space belt itself.

My production side is just a miner platform. The one I used in the original bug report was part of a group of miners about 40 cells away sending output to all 3 levels but I changed it to a single level miner (screenshot attached) attached to a single space belt and I’m still seeing the issue. However, when I tried to compact everything by moving the swapper platform next to the miners, the problem disappeared. Even more interestingly, when I copy and pasted the entire chain (miner+3 extensions, space belt, swapper platform, trash) to a separate asteroid in a more open area, the problem doesn’t happen there. So there seems to be some interaction with seemingly unconnected elements.

I originally saw this problem deep inside one of my milestone factories (I can provide a blueprint of this as well if you think it’s useful). The problem went away when I reorganized my conveyor belts in said factory to not touch multiple outputs of swappers, and I was able to reproduce it with only conveyor belts and swappers as I originally quoted in this bug report so I thought it’s going to be an easy repro for you. But now I can understand it’s not that simple.

This is the entire production to trash chain that I can reproduce the problem with at one location but not at the other. I don’t know what to do next. Do you need my entire savegame? It’s about 15 megabytes.

[analog igloo]

No need, I was able to repro the issue on my side with just the BP, the right catcher seems to be slower than the rest but it doesn't make sense why that would be. I'll be writing a report on this ticket for the team to further investigate what might be causing this 🙂

[SPZ2-4833] Swappers with unknown inputs from outside platform slows down conveyor belts

[hot hinge]

Great, thanks!

[lyric slate]

I had same problem. I've a technical explanation, not a bug explanation:

• when the shape leave the D extractor to go to the vertical conveyer, it's easy for it to evacuate.
• when the shape leave the C extractor to go to the vertical conveyer, it have only to let's the D shape to pass, before go to the vertical conveyer.
• B [...]
• When the shape leave the A extractor to go to the vertical conveyer, it have let's the D shape + C shape + B shape to pass, before go to the vertical conveyer.

But because there are no buffer zone, when the A shape is waiting its turn to go to the vertical conveyer, it obstruct the A extractor exit, then the A extractor mining.

Solution is to put at least one case horizontal conveyer at the exit of all extractors to have a buffer and permit shapes to wait its turn before going to the vertical conveyer, without obstruct extractors.

For me, simple player, it's not a bug.

[hot hinge]

The extractors are not the issue. The screenshot you are responding to is an answer to the question about the whole production line. Please read the original problem description and look at the first screenshot. The four vertical conveyor belts, even though they are fed with identical extractor designs, are acting differently.

[lyric slate]

Yes, I've red, that if you delete swappers you don't have anymore the problem. Just like if crossroad inserts void shapes, or loses time to verify there's no shape from the other entry conveyer.

Sorry if I was just swept up in the second subject.