mojira.dev
MCPE-161163

Shulker boxes broken by a piston have weird, unreliable physics

When a shulker box is broken by a piston, the entities have weird, unreliable physics.

 

To reproduce simply break a shulker box from the side with a piston, as seen in the image.

Expected result - The broken box entity gets pushed by the piston into the hopper on the left reliably.

Observed result - about 1/30 shulker boxes shoot up and get stuck on top of the surrounding blocks around where the box previously was.

 

This prevents bedrock players from having small reliable shulker box loaders. There are designs that help partially prevent this, but even with their massively larger size this bug makes it nearly impossible to make a 100% reliable, tileable shulker loader.

 

The second image is a popular design thats made specifically to counter this bug, yet it still losses nearly 1/5 boxes.

Attachments

Comments 3

All blocks gain random momentum when they drop as items, and items can be pulled backwards if they land on a retracting piston. I was able to reproduce the same behaviors shown in your two example screenshots in Java Edition. So I think this is probably working as intended.

It would be interesting to know what percentage of Java boxes got stuck. Also it's worth noting none of these boxes landed on the piston.

In the 1.19.70 Previews hoppers collect items from the full block space above them (fixing MCPE-55824). That should make shulker box collection after breaking very close to Java Edition. Does that address the issue raised here, or do you still feel that there is a bug?

AdverseLemon2

(Unassigned)

Unconfirmed

Multiple

1.19.20

Retrieved