mojira.dev

Nemes

Assigned

No issues.

Reported

No issues.

Comments

Not creating a new issue as we ran into this on a modded server, so not exactly reportable to Mojang, but leaving this here nonetheless for anyone who stumbles upon it (or whether a dev wants to look into it on their own accord).

 

1.18.2 with a few mods, but one (maybe important?) observation we made is that if the player who was causing the affected chunks to be loaded logged off, the issue would immediately resolve itself, server lag would stop and would not return upon the player rejoining. However, we can't verify whether the chunk actually saved or got regenerated as they were (old but) untouched chunks.
Started happening more consistently after building a few large farms, so it may be tied to tick lag or generally lower tick-rate.

Can confirm in 1.17 here too with more screenshots. Also causes incorrect pathway fillers (planks) to be generated over water where they shouldn't be while generating the usual plank pathways with standard path blocks.

Seed: -7131751027343848337
Coords: 7275 100 6775

Does not seem to happen every time I generate the world, might be related to which part of the area is first loaded and triggers the village population (eg. from where the village is first approached), ran into a similar issue with a desert temple and a village generating inside of each other which was also chunk population order dependent.

[media]

[media]

[media]

[media]

Can also confirm in latest (as of commenting) snapshot.
Snapshot: 21w06a (EDIT: found naturally without the datapack)
Seed: 4850489050736750011
Coords: -19865 -72 3787

[media]