I've confirmed this behavior on 4 different setups with pretty different hardware configs. 18w50a, windows 10, openjdk 11
@Patrick Are you relogging between changing the graphical settings locally? Once the chunks are "waiting" there doesn't seem to be a good way to make them come back other than relogging.
Unless you're talking about just general performance issues, in which case... That's probably a compound problem from at least 1.13.
Edit: The problem can appear again after the server and client have been up too long. At the very least, it is an improvement.
Setting my render distance higher than the server's seems to reliably fix this for me.
Can anyone else test this?
I noticed it specifically at 15 chunks locally and 10 server side I reliably get no problems.
At 8 locally and 10 server side I get problems very quick. It's actually so consistent I did several tests where I flew back in forth in creative until there was a problem.
4 local/ 5 server: 15 seconds before chunk error
5 local/ 5 server: 30 seconds before chunk error
6 local/ 5 server: no problems in 10 minutes
If this is a good workaround, until it gets fixed, I'd recommend people set their render distance 1 chunk higher than the server so you benefit somewhat from the chunk fade of the local setting.
As a note, I am running openjdk 11 on all clients and server but I don't think it is related.
Yeah I should have commented in the other thread instead. The fix of making sure server side chunk distance is smaller than client is for the chunks never loading bug. The chunks that go "waiting for chunk" when you stand in them.
This bug of things just randomly taking forever to generate or always being last to load when going through a netherportal, etc is probably related but not the same.