Factorio Forums

What I tried:

I was testing various spacings of inserters and chests interacting with blue belts. I rotated one of the setups 180 degrees.

What happened:

The behavior changed after rotation in the case where inserters putting on the belt and pulling off of it are on the same side of the belt.

What I expected to have happen:

I expected the behavior to be the same after rotating the blueprint.

The attached save shows the setup (should work fine if all the mods listed in the save are disabled). I also made this video of testing the different setups where I discovered the issue: https://youtu.be/YSyl9U029JE

In your reddit post, you referenced this "Not a Bug" bug report, and it too needs to be re-opened / re-evaluated.

To use an archery term, the behavior is sinful. If "inserters can drop items faster than they can pick them up", as Rseding91 said, then I would never expect the "grab" inserters to snarf all the items. But, as you show in your YT video, it's at distance multiples where they will succeed or fail (ignoring the rotational aspect of it); not just "really close to the drop inserter" as the not-a-bug bug implied.

Reddit post in question, in case anyone seeing this wants to see that as well for context.

I think it's a bug too.
With sideloading , this pattern happens:

So the sideloading case behaves exactly like the case where the inserter dropping on the belt and the inserter removing from the belt are on opposite sides. Makes sense, but good to see the confirmation!

The rotational asymmetry was some floating point calculation asymmetry. It was fixed in Version: 0.17.44.

The other issue was also fixed in Version: 0.17.44 for vanilla by making inserters slightly smarter. They are now fast enough to pick up a full transport line off a belt so the timing issue no longer exists.

Brilijant! thanks T.