#❗-1-18-experimental
1 messages · Page 28 of 1
(The entire thing, not log but all the watchdog lines)
One moment, Remote Desktop is being an asshole
...remote desktop? 🤢
Didn't know you could Remote Desktop to a Linux machine, the only type of machine capable of running servers without the equivalent of 10 tons of bloat!
I have to interact with the server somehow when I’m away from it physically
!ban 688097543753564161 Cheap eta troll
:raised_hands: Banned Vorpal_Flame#1776 (Cheap eta troll) [1 total infraction] -- NotMyFault#3732.
👍
Damm that was quick
Heres the most recent recorded https://paste.gg/p/anonymous/1070f6ae96ec479b9c4ee2e7f32f235e
why do villager trades no longer get cheaper after 1 cure
It was a config option I think?
Hmm I’m not the master at this, but it might seem like ur storage device is slow?
@cunning tiger @vague falcon is correct. It's in the paper yml I think
In order to keep the height at 0-255, I have taken the following steps on a test server:
- Launch Paper 1.17.1
- Create a superflat world
- Load a datapack that contains the following overworld.json https://pastebin.com/TGDkfVDq
- Stop server and launch Paper 1.18.1
This approach seems to work for superflat worlds, even with newly generated chunks. In case of a normal world, the server still attempts to generate depth on existing chunks, causing these exceptions by the hundreds and making the world unusable: https://pastebin.com/NFEvDMvf
Pastebin.com is the number one paste tool since 2002. Pastebin is a website where you can store text online for a set period of time.
ty
Is there a way to get the extra world height in the nether and end?
what is it called
like where would I use ctl F to find it
Trading exploit something.
If you type in exploit you should find it
ty
I’m currently using an SSD. Could it be an issue with the server just trying to load too much?
At launch it was running smoothly with 25 people (pregenned out to 10,000 blocks
Well this seems to be when reading disk stuff, so yea it may have something to do with that?
i cant get it running on windows
i grabbed myself the latest jar, tried to run it, did this: Minecraft 1.18 requires running the server with Java 17 or above. Download Java 17 (or above) from https://adoptium.net/, but it still says Minecraft 1.18 requires running the server with Java 17 or above. Download Java 17 (or above) from https://adoptium.net/
What is your java -version output?
java version "1.8.0_311"
Java(TM) SE Runtime Environment (build 1.8.0_311-b11)
Java HotSpot(TM) 64-Bit Server VM (build 25.311-b11, mixed mode)
ur using java 8
oh
java 8 is in path
You need to change your env path to 17
Either explicitly launch your server with 17 or just uninstall 8
griefprevention flags
how
do i launch it with 17
If you don’t need 8 the fastest way to fix this is just uninstall 8 reinstall 17
it installed when i installed mc java (client)
It doesn't
The client had Java bundled in I believe
I think they're there already?
not 8_311
oh god
'java' is not recognized as an internal or external command,
operable program or batch file.
if u uninstalled java 8 just reinstall java 17 so it creates the path
Don't we have something on the paperdocs for that
Performance is just a lot worse than it would with these patches, and it seems like the chunk loading Patches are not there yet, which is noticable. It's not like we're in a hurry though, but would be nice to upgrade of course.
Hi, do you know what's causing this?
Version: 1.18.1 - Paper 82 Build
Plugins: AdvancedNMotd, AuthMe, CMI, CMILib, CoordinatesHud, CoreProtect, DoubleShulkerShells, dynmap, EasyWhitelist*, FastLogin, InvisibleItemFrames, LuckPerms, PL-Hide, PlaceholderAPI, PlugMan, ProtocolLib, SkinsRestorer, TAB, Vault, WorldBorder, WorldEdit
I'm about to update a server to 1.18. Trying to see if there's anything specific I should be aware of.
set to 0 to disable
Thank you
Ewww why the f did they add nether ceiling void damage are they trying to piss off old veterans?
wait is that a recent update or something
my entire villager trading hall is on the nether roof

That’s a complete garbage patch that’s not worth it.
It's been there for years, disabled by default
Clearly it’s enabled by default
it's not.
Explain video kk
ty for clarification
Where is the patch for this?
yeah all of my paper servers have it set to 0
server patches, Configurable-top-of...
Nvm I’ll find it in GitHub and see.
no?
default is 0, if you come from an old version of the config, if will use 128 if you had it enabled before
The nether roof is free real estate lol
Guys I don't know why it was on either, now it's off.
Maybe someone turned it on, I don't know.
[Server thread/ERROR]: Remove callback for map threw exception
i was on build 79, fixed in build 82?
Pastebin.com is the number one paste tool since 2002. Pastebin is a website where you can store text online for a set period of time.
?
That's an out of memory error
Yes but I have xmx/s10G and a 64 gb dedicated server
Start.sh from the ssh shell
How many other process are there
Without docker or something else
Other 2 servers with 8GB allocated
what cpu do you have
Ryzen 5 3600
Thsts not relevant
At the moment of the crash i had 37gb out of 64
The error you get is out of memory and if you ain’t running inside container. The only thing left is to find out what else is taking up memories
the heap ran out of memory
to work out why it crashed, you'd need a heap dump, there is a JVM flag to create a heap dump when it runs out of memory
do note that the full heap dump has to be loaded into ram, so, if you wanna have something you can analyse you'll need to note that
Hello! Would appreciate some advice - previously I was able to copy overworld regions into the DIM-1 folder to create a spooky "upside down" world, but doing so in 1.18.1 creates several de-serialization errors and ultimately regenerates the regions. Is such a thing no longer possible or is there a different method now?
Uh, what?
DIM-1 is where vanilla stores the nether, the only thing putting anything there should do in a CraftBukkit-derived server is make its migration kick in and move the contents to the world_nether directory
🤯
Correct. Previously, in every iteration of vanilla (and Paper) I've been able to put the overworld regions into the DIM-1 folder to create a spooky "upside down" overworld in the nether. It's not an unusual practice.
Whatever it is you're doing isn't a feature, so it won't be supported, moreso a coincidence I'd imagine.
and mojang changed the world heights across worlds
Ah world height. Didn't think of that
you'd basically need to create a datapack to set the world height or whatever to be the same as the place you pulled the regions from
Yup. Thanks.
my mind blowing learning this 😄
I'm pretty sure they're just saying they copied world to world_nether but with extra steps by making CraftBukkit move it a second time
Which is not so mind blowing 😛
The "upside down" part isn't something anything will do automatically though, no idea what they're talking about there
upside down would just mean everyone falls in to the void
Oh, they mean like Stranger Things
yeah the upside down part is freaking me out hahah
I think "overworld but with Nether lighting", with "upside down" referencing ...yea
Mirror world with a creepy atmosphere
y=-y?
lol no
ya i guess xD
just mirror world, yeah
So yeah, it's just copying world to world_nether
You could probably make a plugin do that automatically
except for that worldheight problem
You'd just have to start the nether world with a datapack that expands it to overworld height
I can't give much information right now since I'm on my phone, but I ran into this weird bug yesterday
I found a supposed stronghold, and I did /locate to make sure and everything, but there's no actual stronghold in the ground
But the eyes of ender and /locate and everything says that there's a stronghold
Why does this happen, do the mobs chase you but don't attack?
I did a small amount of research by searching in this server and apparently it's something to do with the structure seed
Sorry for not doing research on this, I would have completely forgotten if I didn't say anything now
You should probably wait until you aren’t on your phone so you can provide more info.
Yeah that's totally my bad, I just had to get this down
structure seed does not have stronghold options. Has the map seed been changed before?
Not once before

is leaf still asleep
Hey I'm not sure what the issue here is, can anyone help? https://pastebin.com/2RDB8tMv
Pastebin.com is the number one paste tool since 2002. Pastebin is a website where you can store text online for a set period of time.
That's just another form of asking for an ETA, please just don't
post log + timings
a stacktrace of a zombie is useless, thats just the particular moment your server froze
nooooooooo
:>
Oh thanks. I'm going to go through and make sure all the plugins are working properly and are the correct version. We just updated today. I started looking and saw several that need some config edits
nice! you can go through eternity's optimization guide, it has been updated for 1.18 https://eternity.community/index.php/paper-optimization/
don't forget to run with aikar's jvm flags too
in world edit, how would I replace a torch facing a certain way aka facing north
Use debug stick maybe
have you tried asking on worldedit's discord?
What should I provide to get help to debug some issues on a new server I'm setting up for 1.18? Timings report and the console log? There seems to be a lot of times the server lags behind, and can't keep up as well as just straight crashes. Can be when builders do some worldedit stuff, but also when using Chunky to pregenerate chunks and other times I'm not sure if there was anything specific happening
Are you running Aikar’s flags?
full crash logs and timings report when you are getting lag
true!
vouch!
staring over the edge of the unloaded chunks into the void is when lots of introspective thought happens
Last tick was 4.3 seconds ago...
Is the server overloaded?
yeah
the server is lagging so bad here
didn't happen with 1.17
I guess it due to people loading chunks that are extra large at this point
chunk generation and loading is definitely more sluggish on paper 1.18 than 1.17 was, definitely.
i think the larger chunks definitely must play a large part
yeah
we need some kind of y level generator lol
people go like 50k blocks
and the server suffering with 5 tps
world generation became much more complex and therefore slow in 1.18
i suppose there's not much that can be done to optimise the world gen algorithm without tampering with vanilla behaviour
well, not therefore, but in this case it was the consequence
I mean, paper will be faster than vanilla due to chunk system changes and starlight, and other various optimizations
why isn't mojang implementing multi threading
in worldgen itself we have some optimizations but not too many, you can't deviate from vanilla behavior at all
and focusing on caves lmfaoo
for example though the end generation is night and day in vanilla vs paper due to the noise caching
Is there a kind of build tools options for paper?
What are you trying to accomplish?
build the paper jar locally
Yes, see https://GitHub.com/papermc/paper
Just note that there are a lot of different types of jars that you can build
And depending on what you want to use the jar for/why you want to build it locally that'll change
Ahh so i would have pull it via git and run the patchs feature.
so i can access nms and other features and even maybe add my own patchs
If you want to write a plugin using nms see
i have seen userdev and meh
If you want to maintain your own patches, you'll want to maintain a fork of paper via paperweight patcher
i rather just use the build local to use nms
Well... Good luck
I do it now, its not hard to understand the obfuscation.
for 1.17 and below that'd be a more acceptable thing to do/thing you could do
you WILL run into issues doing that with 1.18
and you will not get help here with them pretty much
No issues yet
well, glad it's working for you
anyway thx for the info on building it, might create a auto build for jenkis and then link it over to my own repo's so i can access what i need in the jar.
thx again
just note the reobf jar isn't valid to compile against
it won't work as you expect in the long run
but, if it works for you I guess it works
Yea never had issue doing it this way forever (years). If it becomes a issue i can re-visit this idea and re-do something to make it work, always a way to have stuff work just have to find a work around what i have found coding for 15 years. 🙂
it wouldn't have been an issue in the past, it will be in 1.18. But hope it works!
why do shears not work in dispensers
they do? if you can't reproduce a problem in vanilla, make a bug report on github
I'm having an issue with stronghold generation after the 1.18.1 update, I have a video of it here: https://youtu.be/oWAhJbe9M84
Same here. I ran locate stronghold but nothing is there
If I /locate a stronghold that has coordinates that have been loaded in what seems like everything before paper-82, the coordinates are incorrect and do not bring me to a stronghold
I went to a completely random area and /locate'd a stronghold, and it brought me to the correct coords
If I went to pre-1.18.1 chunks in that have been loaded in anything before paper-82, it also has the incorrect coords to a stronghold
Have you open up an issue officially?
Should I?
With all the good info you provide here. As discord tempt to get washed away
I see
Just with steps to recreate and with your own world download if possible
I had similar issue but it was odd as only one of stronghold was not found
try disabling seed based feature search in paper.yml
wasnt that removed in 1.18 anyways?
I'll look for both of these
The "everything before paper-82" thing is just a theory, would something potentially untrue like that be ok in an official issue? Or should that be omitted
yeah both of those config settings were removed in 1.18. the patch is sitting in the no-longer-needed folder
yeah I had forgotten mojang implemented something similar/the same, thanks
Should I include this or no? @long edge
let me watch your video, one moment
you can stick it in there, although no recent commits touched that. It sounds a bit more like your seed has changed or something, but that doesn't quite make sense either
I ran /seed and it's the same seed that it's been before the 1.18.1 update
This also affects eyes of ender, and I believe it was the cause of my nether portal linking issue earlier
This
nether portals are unrelated
and yes, /locate and eyes of ender use the same mechanism
Oh ok
nether portal linking often borks, the logic is fragile and it even breaks in vanilla in amazing ways
Ok I made the issue https://github.com/PaperMC/Paper/issues/7158
so is 4
ho my god... I'm also facing that issue: players try to find strongolds and... eyes of ender leads to place where there is no strongold... the seed hasn't been changed
any workaround? has someone tryed regenreating thoses chunks?
if you revert to a pre-82 build, does it go back to working?
You will test on paper to report a paper issue 
Can anyone see what I should work on to reduce lag? This 3rd time trying to load chunks with Chunky it seems to actually work...https://timings.aikar.co/?id=2c8bbede15c94f2dae5814d91b29555e#plugins I know I will fix the Aikar's flags, but anything else to look at?
Aikar's Timings Viewer - View Timings v2 reports from Paper and Sponge
You could try disabling Dynmap while pregenerating chunks with Chunky. Once you are done pre-generating, then run /dynmap fullrender
That's what worked for me
will do, it actually got to 15% now before crashing
any idea why I cant use this aikar flag? I do have UnlockExperimentalVMOptions enabled.
[Server] INFO Error: VM option 'G1MixedGCLiveThresholdPercent' is experimental and must be enabled via -XX:+UnlockExperimentalVMOptions.
is it just shockbyte not putting it in correct order when enabled from a gui like this?
20.12 10:39:23 [Server] INFO Error: The unlock option must precede 'G1MixedGCLiveThresholdPercent'.
20.12 10:39:23 [Server] INFO Error: Could not create the Java Virtual Machine.
20.12 10:39:23 [Server] INFO Error: A fatal exception has occurred. Program will exit.
order is important here
Reach out to shockbyte if all you are doing is checking boxes. More than likely you need to uncheck them all then check them in the right order probably
Also update paper
ok 👍
Any input on this timings report? I enabled all aikar flags except the one that failed. Disabled dynmap. Generation seems fine in chunky now, been running from 16% to 30% this restart. https://timings.aikar.co/?id=4be84c5739b24864ac7da2556416e945#timings 20.12 11:57:38 [Server] INFO Task running for world. Processed: 120599 chunks (30.68%), ETA: 3:41:10, Rate: 20.5 cps, Current: 174, -16. Also 1.18.1 now
Aikar's Timings Viewer - View Timings v2 reports from Paper and Sponge
Youre missing a lot of the flags, so you haven't enabled them at all tbh, G1GC comes standard with java 17.
don't do drugs
😄
you should run aikar's flags
follow the guide here, but ultimately your cpu on the host sucks as well
also you cant fit 200 players in 8gb of ram in 1.18.1
ok, yea Im reading that guide, very helpful. Wont be 200 players, but yea, will increase if its needed, at least that is easy there
oh it's that shit? 😛 I mean its not even mentioned in the plan
SB is living in 2012
exactly
also gc every 3 seconds?

timings cost is low-ish because there is fuck all to even time lol
I was looking into hetzner, and this seems like a good option, you just miss the dashboard etc that makes it so easy for beginners
tf's that
hetzner is fine, hear things about their network
so idk
a very good host
Bloom and DedicatedMC are the recommended ones around here
ok
that's one expensive machine wtf
oh its dedicated, nevermind
64GB ram
you will hit a cpu bottleneck before 64GB
yup
you will almost never need more than 12 gb of ram
hetzner seems like such an edge case option for servers lol
its designed for people who dont need a super strong network and 100% uptime, but want performance at budget
same as ovh
there are some videos of their datacenters on youtube
there is no 16GB servers on the performance plan on Bloom available🤦♂️
and essentials is only 8gb
essentials?
just go for performance
??
yea only thing not on waitlist is performance+ and essentials
then check out other hosts like DedicatedMC or EnviroMC if you can't wait to be in a waitlist
really liking the demo of bloom
hello
Fulltimehosting is also cool. Had a GREAT experience with both fulltimehosting and dedicatedmc
xd hablo espanish
Had to leave bc my map was humongous
si no q toy aburrido
Yo también but please speak English here
no enter
necesitas ayuda con paper 1.18?
I would recommend HeavyNode over bloom. They don't restrict cores and allow you to enable Aikar's flags
no pa
entonces #general
waitlist wasn't terribly long for me
might be anecdotal but if you want it it's worth the wait
I have a question about this commit. By skipping the close() on an exception, isn't the bad data still written out and it's just leaving the file open? It seems to me this would leave file open and probably locked.
https://github.com/PaperMC/Paper/commit/8c5be166861b662d0b4aa775f0e38c8acb9a30f1
Assuming the person making the PR knows what they are doing, I'm guessing that #close() actually flushes the buffer to disk. Seems hairy though, depending on the volume of data I would just pass it through a different output stream first
Write a comment on it maybe?
Playing paper 1.18.1 and there is a problem with our wither skeleton farm. It makes use of the mob-spawn-range towards the player but no mobs are spawning at all. We can put the spawn platform lower and than mobs do spawn suddenly. What setting in paper could influence this?
We changed the mob-spawn-range from 6 to 8 (default in MC Vanilla) already
But still same range with mobs spawning or not. They only spawn when the afk spot is 25'ish lower
Well there is not any mobs that are spawning
within 128
but 96m there is mobs spawning. That should be the paper setting mob-spawn-range right?
Also extra point to notice. When the mobs that do spawn when I am in 96m range go through a portal to the overworld they immediately despawn, however when a player fully loads the chunks they do actually come out of the portal
Time to get better hardware for 1.18 ._.
8700K w/ 53 players not really handling it anymore
Anyone else having issues with breaking bedrock after enabling the feature?
Doesn't paper prevent bedrock breaking exploits?
Yes
network or single server?
It says bungeecord in the SS
Bungeecord yeah, but only a couple other servers running on the machine at the same time.
It's basically entityTick eating most of the TPS, villagers mainly
:(
they also prevent tnt duping
i feel like if mojang acknowledges and doesnt attempt to fix a bug then its a feature now
butttt idk
Yes but theres also a config file to bring it back
Other way around, really. If Mojang acknowledges something is for sure a bug that's when the Paper folks will come and patch it
Or if it's especially egregious like a dupe
For most server owners this is desirable
We always just punched a hole in the ceiling at 0,0 and called it a day
iirc we also limited the area in the ceiling you could build on so people didn't build 100 grinders
time to check 0,0 in all my servers 👀
any advice on the 1.18 portal linking issue? Any patches or pull requests we might try to get existing portals linked properly again?
I have not done a whole lot of research just yet, but on our Paper server, the "jeb_" sheep does not appear to work. Anyone else ran into this issue?
Isn't that not a thing anymore in modern versions?
Jeb sheep -> #paper-help message
ah, good ol' upstream
is it close to done?
lol
Updates to Paper do not have any sort of estimate for when they release, ever. Any and all updates will arrive when they are ready, and the only thing to do is wait for them patiently along with everyone else.
lol even a command
its just its been 20 days
startlight is in
track patch progress here
im not a server owner and the owner wont put it on because its early builds
Well, they should make backups and test it
60k+ servers run these builds, so you most likely fine
The last few game breaking bugs are only related to world upgrading so take that as a sign to start new world
What would you call game breaking?
The memory leak one - Fixed
The random overworld biome in The End - Fixed
The stronghold /locate not working - apparently always been in upgraded world since 1.13
Oh you included closed ones in that
I dont think any issue has shown up yet for server with brand new world ic
Paper is patched to log4j 2.17, so all that's left is zero-days again
and those are always a risk
So what problems exist with existing game world upgrading? How serious of a problem is upgrading an existing world?
not much anymore
well you are outdated!@
the vuln in 2.16 is "just" a 7.5/10 DoS, not a 9.0+/10 RCE
but yeah, keep patching
can just support latest 
(to be clear, the DoS is a full JVM crash)
as a basic fix ..going through the roof portal reconnects the overworld portal in our experience.(on 1.18.1)
Multi-threaded, yet is a paper fork... Okay
are you talking about minestom?
clearly not since he's mentioning a paper fork
It's possible in a limited scope
If you need to support every setup with stupid plugins it quickly becomes impossible
understandable
Paper generates another world-settings category under settings?🤔
feature seeds? i didnt know that was a thing
all anti-xray is flawed basically unless you're talking about something like orebfuscator with gigantic overhead
wondering what that does 🤔
if you find the seed you can bypass antixray
seed crackers exist too :>
as you know where all the ores are because you have the seed - unless you change feature seeds to stop this/generate ores differently
ty I'll try that
what
for Xray one, it doesnt, the feature seed is normal
but you should only have 1 world-settings
Thanks bud. We indeed use a plugin which adds colors to renamed items.
I mean it shouldn't, but
world-settings:
default:
world_the_end:
world_nether:
@stoic depot ^ it should look like this
with sub setting below each world
Yeah, I'm trying to figure out which setting broke it
the error message should tell you which line
its usually the line above it
so if it says 50 go check 49
I'm not seeing any errors tho, the world-settings section just.. keeps getting there :P
I'm currently copying values to a fresh config to see which is causing this
Is there someone that have test server with WorldGuard?
And can test one thing for me
make sure you are grabbing latest dev build
but WG has their own discord for support
Probably I found a desync bug
I just don't know what event is cancelled
But bug is in paper or upstream
Create a region with player 1, place water blocks and spawn axolotl. Then try to grab it using water bucket via player 2 that not added to owners/members of region
You got You can't do that there and event was cancelled
But axolotl disappear on player 2 client side
And still living on player 1 and server side
I think that PlayerInteractEvent (or what) isn't cancelled properly
Or PlayerBucketEntityEvent
Can someone test and check this please
If it confirms I can create bug report on GH
pretty sure that's already known
(I found this on 1.17.1 and reproduced on 1.18)
well, either way, that sounds like something to investigate from the WG side, so you should talk to them first
it's a server issue afaik
Like, WG cancels events
if the server ain't handling event cancellation properly, that's on the server
coughsnowbuckets
/gamerule doFireTick false?
Hey, theres an open PR to fix it. You should review it 😛
I'm need to create bug report or it's already exist?
i've seen it, what am i supposed to do with it tho. it's like a half line fix.
ping cat and tell him to review it
probably. but for now, for whatever your use is, the only exception is invisible armor stands it seems
Figured this one out. It was our anticheat which is supposed to automatically enable anti-xray and and was modifying a wrong path in paper.yml
Alright, so I've had this issue where the difficulty keeps getting reset from hard to easy. I've used the difficulty command, as well as the server.properties file. Am I missing something?
EntityTakeDamageEvent? Maybe?
if using plugins like multiverse, check their settings
mine gets reset by multicraft (yeah yeah, host isn't my choice) anytime i edit server.properties from the panel (even before Paper)
is there a file in the world folder I need to edit? None my plugins should affect it
don't mention
it's set with the difficulity command per world
no need to edit files manually unless you have a plugin messing with it
i wasnt looking at docs when I said that, it may or may not exist but it definitely existed somewhere according to my brain
EntityDamageEvent is called for item entities
but, you can't set the health of the item until https://github.com/PaperMC/Paper/pull/6514
bazinga but also not bazinga :(
note, that the item entity will still disappear for clients because of a client bug
sorry mb, that's the issue, it resets it even after using the difficulty command
then you've got a plugin doing smth
replicate it resetting without plugins
or a datapack
despawn and respawn the item entity this could not go wrong
yeah, you could do something like that
alright I'll try it thanks
hmm, I think this is a plugin thing... PlayerInteractEvent doesnt fire for entity interactions, PlayerInteractEntityEvent does, and its not cancelled by default for a region
I don't know enough about how worldguard protects against these things to say for sure.
well, it actually is an issue. https://github.com/PaperMC/Paper/pull/7162 should be a fix
this worked great. For anyone else trying to fix nether portals:
- Take the overworld coords, divide by 8, write this down
- Travel to these exact coords on the nether roof and create a portal
- Go through it
- Go back to the nether roof, and remove this portal
After doing this, the original pair of portals should now be fixed
or if trying to reconnect portals between roof and overworld, go to the nether roof portal ..and go through it ..simple
(my server mainly works on roof to overworld portals)
there is an issue wher going through an overworld portal doesn't connect yes
it seems to nudge it in a way that it fixes it
they don't "link", the game searches for them every time and creates a portal when it can't find one
so if you go through an old portal on the roof it links to the overworld portal so thats a quick fix
portal linking is something I've never been able to fully understand lol
there is no permantent linking thing
are you saying the word 'link' is incorrect?
Yes
I think vanilla portals are just unreliable, not sure this is a paper problem
the thing searches for portals each time you walk through
there is no permo link between the two
Ok, thats cool, I just use 'link' as shorthand for 'works', sorry
the logic has been janky as shit for years, increasing the search radius might help but theres issues going back years over this
please don't mention
ofc, increasing the search radius might have its own set of caveats
curious how the fix works then. Something in there is stateful in a way that that worked. Or maybe not
Maybe portals should just be linked but then people wouldnt be able to play around with placement or put a ton of portals next to eachother
well, there is a cache which is persisted for a little while
theres also many caveats to the game, I wouldn't be surprised if there was some oddities around chunk upgrades and what it deems to look into, etc
as a philosophical discussion ..I would love to continue, ..but its probably not appropriate in this channel?
"it's not a bug it's a feature"
so perhaps we move this to general?
I mean, I got no much more to say about it
If portals were linked then if you put a ton of them next to each other then they would be smushed together in the nether 🤔
Or something
tbh I just suggested something that works ..
it works across restarts I assume?
you guys discuss all you want
yes
only thing I can guess is that loading the chunks helped or something odd
..in our experience across ..7 days ish of restarts
good to know. doesn't make a whole lotta sense if the cache is in memory, but if it works then it works
like, as said, there is 0 perma linking, it's all based on the search logic
all I can guess is that some chunks weren't loaded in 1.18 or something and the search logic tripped up
esexo :O))))
we have very few plugins, didn't use chunky, none of that stuff ..we just vanilla upgraded
ohh that's a good guess. Maybe the search logic tries to load a 1.17 region and then craps out
How can I reduse ram usage? It is idily using the entire allocation, and I can't really allocate more as I'm using other software
allocate less
982mb is being devoured idle, how do you expect it to run with less then that
Java will allocate all the RAM you give it. Even if it doesn't, you don't want to have other stuff eating that since when it needs the memory it's going to OOM or aggressively GC
The latest build with the tripwire patch, is this a fix for the string dupe bug?
oh is this on a Pi or smth
No my windows laptop
What have I done.... suddenly minecraft:plain biome in the end world
update papermc, this issue is fixed
as in, it won't happen again, but you still gotta fix your existing end biomes with worldedit //setbiome
according to the bug report it usually only affects a "small part" of the end https://github.com/PaperMC/Paper/issues/7125
you'll have to do some guesswork and fly around
it happens when initially loading your 1.17 world, so only areas you've explored in broken 1.18 versions will be affected
ideally that's the type of thing somebody would write a plugin or something for to try to patch afflicted worlds
I just used //setbiome on the whole island + the black zone and... my server is currently... unreachable since 2 minutes i just hope worldedit will finish the job i should have used async world edit but i've heard it's buggy
FAWE or asyncworldedit should be fine for just this job. Just make sure to take backups first
or if you turn off watchdog, worldedit operations will complete eventually without crashing
WE ticks the watchdog by default now iirc
oh that's good
i need to expandvert because biomes are 3D even in the end right?
still running at 800% cpu on 8 core hahaha 36 players in the network still
are chunk loaders disabled in paper?
Is there a recommended batch line for executing 1.18.1 paper?
Ok thanks
can I undisable them
Nope
r u sure
just wondering if those crashes are coming from my server or paper so anyone else getting frequent restarts without anything specified in console ?
You'll have to check the logs to see the cause
we don't have any complementary information in the logs either :/
how to make glow squids despawn at a hard 50 blocks
Well, what do they say? If there's really no info, it could mean out of memory if your server is configured to autorestart
20.12 21:18:32 [Multicraft] Restarting crashed server in 2.0 seconds
20.12 21:18:32 [Multicraft] Looks like a crash, check the server console. Return value: 137```
that's all it says everytime and tps is good, we're not overloading anything
137 generally implies killed by host due to memory usage
(i.e. the host ran out of memory)
oooh
are we sure there is no way to enable chunk loaders
anyone are having old gc issues caused by entity tracking?
How did you reach the conclusion that it's caused by entity tracking?
analysed with spark
Vanillatweaks had a datapack for chunkloading blocks. This worked for 1.17.1 paper, no idea if it still works for 1.18.1.
Would advise against allowing chunkloaders on a public server.
not sure this is the right place to ask but what kind of issues are people having most with 1.18.1 builds, crashes, world corruption, or just random non-critical probs
Those have been fixed: touching a lectern would cause you to get stuck, random small bits of the end would turns to a plains biome
The ongoing issues rn are mostly performance related and not specific to Paper, like chunks taking a long time to generate and players with bad connection getting timed out, falling through the floor due to chunks taking long to load (I personally haven't experienced the latter)
^^
This of course doesn't cover anything not found yet.
The time out issue is one of the largest issues i run into with my test setup. Over half of players get connection issues.
That just seems to go back to 895 with some exhaguration due to the change in 1.18
Like, mojangs "patch" for that was literally something I did years ago
exaggeration
And ive noticed that since the log4j fixes got rolled out my server keeps freezing every few minutes for 5-10 seconds. It happens more often for longer periods if more players get on.
That's probably coincidental.
Well, i downgraded to pre- log4j fix versions, and they didnt have the issue
i can shove nearly 100 maggots on #20, but #66+ i get the freezing issues
players >.>
Oh.
interested in how No tick view distance will be ported given the new simulation dist
for now you just use vanilla's no-tick thing
to my understanding there are deeper changes in the impl of it
So, the concept will stil be there, but, fstr
Yea between #20 and #66 a bunch got fixed, i never bothered getting those builds as #20 ran stable as heck. But i had to upgrade due the log4j thing. So now im stuck on #66+ builds which... underperform to say the least.
always update
send timings when it happens
i'm willing to bet you have severe memory issues / some memory leak
We can't really help when you jump several dozen odd builds in one go
like, people can look at timings, etc, but in terms of spotting a regression between builds, glhf essentially
Eh, dont agree with the always update thing. Too many things tend to break if every update is followed. I prefer to check if everything keeps working if paper or any of the used plugins update to ensure i dont kill the server by exchanging a jar.
I rather run stable than have to deal with random errors due something not being coded right on whatever end.
Most of the times its plugins breaking with version upgrades, but i remember between ~290 and 300 ish on 1.17.1 paper the spawnerspawnevent got changed, and a lot of stuff broke overnight.
i don't mean updating every build
because you shouldn't auto update
but do upgrade every 10 or so builds
And generally have a test instance to test the upgrade on to make sure nothing busts.
Yea normally i cycle updates every 2-4 weeks depending on how busy i am
only things im 100% sure of wont break anything ill update whenever i notice new versions. like Luckperms, Essentials. Stuff that barely ever breaks.
shit happens
you can always revert back a build, that's why we recommend testing
when you jump several dozen odd builds and expect support for a regression, you're basically SOL
theres 46 builds there to run through and look for regressions
who knows which one introduced whatever issue you're facing
Eh, 1.18 still "experimental/unstable" so i dont bother for support at all, just got the test enviornment up atm, was aiming for xmass to go public with 1.18. But i've shifted that to 2nd week of january now.
I considered for a few hours to grab the log4j fix and applying it to build #20 so i could run that "stable" build with the fixes. But my brain cant comprehend what i gotta copy paste 😛
Yeah that's what I'd do. If you're seeing a big issue then I'd start going up from #20 until you find the issue.
Then you can say "between build X and Y this issue pops up"
I would put $5 on it not being the log4j bit.
It's just bumping the log4j version up.
Yea wouldnt suprise me if its something else, but i dont have enough braincells to understand what it would be if i managed to find it anyway.
Grab the first build after #20 and test with that. Keeping increasing the builds until you encounter the issue.
Not to sound unappreciative of all you guys hard work. But im a old boomer that needs to do boomer stuff. Dont have the time to troubleshoot. Im sad to say i'm one of those people that waits until the "OMG THINGS NOT WORKING HELP" spam dissipates before doing anything without contributing much to the testing of versions myself.
I think in this case you're the only one we've really seen with this issue.
So testing helps you and no one else, really.
You could just wait but if it's a specific thing impacting you then swapping versions will find the cause and let you make decisions on what to do about.
You could just wait but you might get months and many builds from now and it won't have fixed your specific issue.
Makes sense, when i see the builds cross into the #90's this week ill do another update and see if it persists, if it does, ill run through some builds to see where it starts existing.
I built a auto update script so it will pull the newest build every morning (checks the last build i have and if the newest one is greater) and run it on a dev env (docker /rancher). If everything goes fine then I just update the image for the containers for the other servers (Update folder location and reboot the servers to update the paper jar). If not I look into why it failed and try again the next day, if it was a major change, fix the issues and update the assets.
This sounds rather juicy, ngl
Is that a complicated setup?
Nah
And would it be possible to adapt such a script to check for plugin updates instead of a paper update? assuming its just crawling their githubpage?
Having my test server autoupdate would be... quite efficient >.>
It depends how they have there repos setup and if you can access the changes, all and all it in theory yes.
Sounds like a useful tool, i really need to learn how to code, but at this point, ill be dead before i understand any of it >.>
kinda related but man would I kill for a maven-like repo for plugins. having some trusted central source to auto-pull them from? 🔥 we use version control for our server so it would be killer
the simplicity of a self-hosted repo but the reliability of central essentially?
Im kinda hoping that Velocity will "push" into this direction. Although im not seeing it at this point..
we already mirror a handful of stuff, not on the level you're wanting though
Well, that's what hangar will offer, just boils down to adoption
Would anyone call DBO actually "trusted" anymore?
i mean that'd be the minimum
i mean regardless of your opinions on dbo it's a) more than spigot does and b) an example of what might be necessary to have that
I thought it was largely abandoned and submitting stuff basically went into a void now?
i mean even things like maven central aren't like, "free of bad things". it's only "trusted" insofar as you have to proof you own the domain which you publish to
What DBO used to be is definitely a good example.
Yea, am not 100% sure if we've decided what we wanna do there as we don't wanna be spigots premium queue with the throughput of the free stuff
uh, no? some of the largest plugins still publish there just fine
not sure why you think it's abandoned
auto scanning stuff was defo a thing we wanted but that would only catch the more obvious stuff...
dbo is still alive, just, more like a lingering corpse
aka, a curse project
yea i mean overwolf's upcoming api changes are controversial but it's by no means dead yet
really only controversial since NEW projects wont be in the api by default i think
Any known hopper bugs?
Please explain in more detail
Please explain in less detail*
i will try to get the details. the people in my server dont like to give me anything more then "YoU BrOkE iT, StOp TaKiNg AwAy My FuN REEEEEEEEEE"
Most of the time is player being dumb
also make sure hopper-amount is 1 in spigot.yml, might break some setups otherwise
Use default amount yeah
Some guide would want you to change it
You can reduce lag by impose a hard limit of X hopper per chunk. Player can adopt to it.
do datapacks still cause problems on servers?
Wdym? @onyx apex
As long as it doesn’t have repeating functions it will generally be ok
ive heard that datapacks break when used alongside spigot and paper
One with custom biome generations yesh
good to know, thanks
To be fair, they also break in vanilla
But, there are some challenges around datapacks in bukkit stuff
I think the one that attempt to generate biome that doenst belong to X dimension breaks. I recall that popular Better End datapack is brought up many times to not work on paper, spigot, and probably bukkit
It used to crash outright on paper now I think a patch was made to just ignores it so it doesnt crash 
Would Paper enable 2b2t to update to modern minecraft versions?
iirc at some point DBO basically just started autoapproving plugin updates
from what I've heard they eventually plan to, and would probably continue to use papermc if so
2b2t has its own set of challenges that aren't very similar to any other server
namely the fact that people are constantly trying to destroy not just the blocks in the server, but also the server itself
They already use Paper iirc, just the 1.12 version
Would anyone know why my server's log just doesn't exist at all?
It still prints a log file but other than that nothing displays in the command prompt
wait nvm I think I might have found the issue
My files keep on resetting to be read only for some reason
Nope, still doesn't show any logging in the console nor does the server actually start correctly
Wait hold on
My game says localhost has a server running on it
Ok theres a ghost server running
No console, but its there and it has the plugins I had installed
The log doesn't appear still, but the server loads again
stuff like that generally screams that the server /restart'd, either through the command or a crash, etc, and the restart-script pointed to a valid shell script and so it started off in an unattached process
really boils down to how are you running the server, even in worse case the thing should just be able to write to the console as any app, otherwise it tries to use fancy features for some stuff
Looking back on the latest.log I can see an error related to the terminal
This is the error, not sure if it's related though
error.txt by @trail mural: https://paste.gg/77caa8a72bc241de93168451c649fca4
The full log also seems to end after [Server thread/INFO]: Timings Reset with nothing related to me joining the server or closing it using the /stop command in game
Oh actually tahts because another fake server loaded
Yeah I'm using windows
make sure that you start the server with a bat file otherwise you won't have a console to dump info to
Yeah I got a batch file.
Could it be that I'm using a java version located in my documents folder?
:Loop
"C:\Users\user\Documents\Java\jdk-17.0.1\bin\javaw.exe" -Xmx3G -jar paper-1.18.1-84.jar -nogui
goto Loop
use java, not javaw
also, if that's named start.bat, make sure that restart-script in spigot.yml points to a dud file
tysm I'm an idiot
yeah its called start.bat
oh I ned to chang thtat then
if the server crashes or you /restart, it will start that bat file, and so you'll have 2 servers tryna start up
if it's MapLike stuff, yea, typical mojang update
corrupt image Im guessing
Is it possible to stop all logging? Im working on a test build with very limited disk space
if the few byte of file is an issue, you really want to fix storage issue 
Its more out of curiosity :) do logs get deleted after a while or are they just staying in the log folder until i purge it?
You wanna adjust the log4j xml if you wanna play with that
I am actually not sure If Mojangs config is setup to delete old files
I wanna say spigot changed something and there's an upper limit now
Ahh thanks then thats what im going to do next. Its a rasoberry pi with a micro sd. I d like to keep the sd as clean as possible because its not very big
Makes me wonder if it would be beneficial to store my backups on a external drive 
You always want offsite backups
Yeah youre probably right. Ive got one more question. How would ou approach scheduling server stops and restarts?
Would you use a server plugin or run some crontab/ tmux scripts?
later is better
Thanks again 👍
what kill signal do you have to pass to paper for it to gracefully shut down?
Nothing yet. Im still in the middle of setup basically. Isnt a simple stop command not doing the trick?
Sig term should be respected with a graceful shutdown
Im still a noob and i think i couldnt quite follow your last two sentences. Im pretty much learning as i go along. What exactly is sigterm? The wikioedia article in my language is not very good :/
It wasn't directed at you
i was asking about something else
sigterm is one of many signals that linux can send between different programs running in a computer
But basically, in Linux you can send special signals to an app and it would do special things
Sig term, signal terminate or whatever, tells the app that you want it to gracefully terminate
Oh my bad then
Interesting... Ill probably watch a tutorial about this 🤭
don't go too deep down the rabbit hole
We do what we must
Ngl i was pretty annoyed teaching myself linux at first. But the more you get into it the better it gets :p
Hello, just a quick chat I found something weird that monster (Zombie, Creeper, Skeleton) was spawn in The End (Spawn cause: Nature) on latest Paper version
BTW. I need to testing on other Server on new fresh world
Yep. I have plugins install on this server. But its has nothing to do with mob spawn
Hi - this was a bug in older versions that got fixed.
Check the biome - it's likely not an end biome.
You need to grab the latest version and update the biomes with WorldEdit.
Or if it's a fresh world regenerate it on the new build.
OMG! Biome plains
Thank you I didn't notice that
No worries! Hasn't impacted everyone but has hit a few people.
that makes a very op farm. Dont get me wrong 
Haha
Would it be possible to delete the chunks and then pregen them again using chunky?
No really. This is my solution. Quick fix plugin [WRONG CODE!. SEE BELOW 👇]
Working code: #❗-1-18-experimental message
I was gonna say imma borrow this, but ill borrow the correct one if you dont mind :p
something like this (WRONG CODE!. SEE BELOW 👇 again)
Working code: #❗-1-18-experimental message
That hooks into WE/FAWE i assume?
EDIT! WRONG CODE!. SEE BELOW 👇 super again
Will this not override real nether biomes?
No it only replaces plains with normal end
unless im reading it wrong with my -1% java knowledge, but i dont think i am
In whih line is the check for plains?
It checks the chunks on if its a end biome or not, if it detects a non end biome, it flips it to a end biome
Since Environment is THE_END. and method #isTheEnd(Biome) will check biome. //Ahhh ok
Yeah sry its still early 🤭
Although Tacklezaza can probably explain with more detail 🙂
Quick note!. This is untested code. 🤣
Test it and send me the plugin pls 😋
Well it's might be a good idea to send ".jar" file in here? //Maybe -..-?
//I'm gonna testing code first to see is it working...
We generally don't like people sending jars to folks.
Someone could send a jar with malicious code in it and we don't want to get people into the habit of grabbing random jars and using them.
We tend to tell folks not to use random jars people DM them, etc.
You could toss the code up on Github.
^
Yep. If you're the developer. You know what to do. Just copy & paste code to your plugin 🎉
//And yes the code is working. ❌ Take my word back. not really working in somecase
Final Code: #❗-1-18-experimental message
Sure! ❤ No copyright 🤣
Well. More quick note, That code not work in some case for example. if world is loaded before plugin initial then ChunkLoadEvent is not being called? maybe
And stay tuned for more update 😑
//See update below
#❗-1-18-experimental message
mmh... Plains biome in the end? Take My Money 😄 😄 😄
I can't get the stonecutter recipe buttons to link to the recipes they should - does this warrant raising an issue on the tracker?
Haven't been able to get a straight answer from anybody about whether it's just my own bad programming, or an actual bug with the 1.18 version of paper.
Not that I expect to be answered, it just makes it harder to tell whether the issue is genuine or not and I don't wanna waste anyone's time.
Have you tried testing on a fresh Paper build from the website with no plugins?
Well, that's what I've been originally working on, yes
I'll quickly try with the latest update
(cuz I assume new builds are being released all the time)
Often-ish
But then I'd test on a vanilla server.
See if it's a 1.18 Minecraft bug or a 1.18 Paper bug, basically.
Oh.
(maybe I'm missing something here)
Sorry, I didn't realize this was a plugin you made.
I suggested to try testing with no plugins.
Ah, all good
Yeah, see, I can't tell if it's me doing something wrong, or an actual issue
because the recipes appear
and the buttons appear
but they produce the outputs of... different recipes to what the item on the button shows
idk if that makes sense
If it's some kind of custom recipe then it's likely not Paper.
But hard to say for sure.
definitely feels broken though, given that you only give the registration function a single output itemstack, yet the itemstack on the button and the itemstack produced are literally different
Best you can do is asking in #paper-dev when folks are around to answer.
It's kind of a quiet time right now
I tried the other day
multiple others were on but the question was overlooked
I am concerned that given the nature of this, it's actually not just my bad programming skills lol
Are the recipes using different amounts of the item ?
I think I recall someone having this issue a while back
(once again, I don't have any sense of entitlement, it's just that if most people don't know much about the stonecutter stuff they're not gonna answer it)
With different amount of light blocks
no, each one simply has the same input and different enchanted books on output
let me send the source
Please send large files/logs to a pastebin
A sensible, modern pastebin. Share text and source code snippets with no hassle.
Tho as your friendly neighborhood moderator already said, let's move to paper-dev
oh, okay!
Sorry, didn't take that as a hint to move
lol
(I've placed the code there now)
This is final code for fixing Paper 1.18 The End Biome Error @acoustic yacht
https://paste.gg/p/anonymous/5896403caa484984a4127fba79a4554c
** No copyright ❤
hi does anyone know how to write plugins in 1.18.1 because the 1.18.1 spigot.jar does not have the packages that trhe 1.16.5 used to have
Use a proper build tools like maven or Gradle
If you need to access internals (you really shouldn't), use paperweight
i just downloaded it from getbukkit
ok
how to fix the fog?
update to .1
I've had this same issue - any ideas?
Hi
Did paper remove the starlight mod's patch in the new versions?
Oh wait - spigot.yml had mob-spawn-range set to 6?
(by default)?
Well we lowered the spawning platform with 32 blocks
Yeah but I changed that to 8 and it still behaves the same
In spigot yml?
Yes
hmm
Well you could try maybe if it works I did something wrong then
No it didn't
Maybe a paper bug then
Here's my config files https://paste.md-5.net/dejugiyamu.coffeescript
huh
That could be it
@true bloom yeah that's the issue
Is there any risk that build.gradle.kts in Paper-API still says log4j 2.14.0 ?
a) no that is just the API and b) even that was fixed https://github.com/PaperMC/Paper/commit/df8d28a450fe391b9f0c2dc4a64543cf9b84266f
Oh this just got patched
Yea
it is fixed lol after i provided 1.17 backup world which got affected on 1.18
Is this supposed to be the default? It seems to be the default here and in spigot (spigot.yml) - whereas I thought the default used to be / was meant to be 128 blocks / 8 chunks
wow that's a name I haven't heard in a while. Definitely do not go to that website, it's had infected minecraft jars for years
if you type in downlod spigot thats the first website that comes up
figures. google search results for mc related things are always loaded with minecraft malware and rehosting websites, they don't care to clean up the problem
You don't download spigot, you download buildtools to build spigot, hence the discrepancy
help
also what was it infected with?
what type of malware
It literally says the error
delete banned-players.json
ok
I'm not sure, there was someone who investigated the jars from that website a while ago and found fishy stuff but I don't remember the specifics
the site was a bit sketchy as well so that explains a lot
full of ads and stuff
and the fact that its called download bukkit when i am downloading spigot
bukkit is a real thing, but has to be manually compiled, and usually you're using a fork like Paper or Spigot
ik it is real
is serverjars.com safe?
i also downloaded from there in the past
No get your jar files directly from the source
If you want spigot, spigots site
Paper, papers site
ok
Anything else is unverified
how to put a certain seed
in the server.properties file
So far so good!
what line?
it's always different, key is level-seed
in my case it's line 4
enable-jmx-monitoring=false
rcon.port=25575
gamemode=survival
enable-command-block=false
enable-query=false
level-name=world
motd=A Minecraft Server
query.port=25565
pvp=true
difficulty=normal
network-compression-threshold=256
require-resource-pack=false
max-tick-time=60000
use-native-transport=true
max-players=200
online-mode=true
enable-status=true
allow-flight=false
broadcast-rcon-to-ops=true
view-distance=10
server-ip=0.0.0.0
resource-pack-prompt=
allow-nether=true
server-port=25565
enable-rcon=false
sync-chunk-writes=true
op-permission-level=4
prevent-proxy-connections=false
hide-online-players=false
resource-pack=
entity-broadcast-range-percentage=100
simulation-distance=10
rcon.password=
player-idle-timeout=0
debug=false
force-gamemode=false
rate-limit=0
hardcore=false
white-list=false
broadcast-console-to-ops=true
spawn-npcs=true
spawn-animals=true
snooper-enabled=true
function-permission-level=2
text-filtering-config=
spawn-monsters=true
enforce-whitelist=false
resource-pack-sha1=
spawn-protection=16
max-world-size=29999984
yea idk, add it manually
GC
G1 Young: 211 avg(4.35ms every 616.94s)
G1 Old: 0 avg(0ms every 0s)
✓ Using Aikar's flags
Amazing GC
Hello how can i add more attackers in the raid
Hi, any solutions? Is there any way to kill them?
UUID of added entity already exists: EntityPig['Pig'/11786, uuid='9b29f828-294a-4ccc-b44b-7bbc2a463238', l='ServerLevel[world]', x=-1797.04, y=64.94, z=-555.97, cpos=[-113, -35], tl=3691, v=false]
That usually happens when server closes unexpectedly
you don't have to worry about it, it gets resolved automatically
Okay, thank you
there was this hostflow malware thing recently
1 month ago or smth
I am using the latest version
@torn portal if you mean to remove fog. Run fabric with sodium extra
how is the Ryzen 7 3700x for 1.18 servers?
It’s ok
ok
i'd say its quite decent
Good afternoon,
I heard that the server properties have changed somewhat with the introduction of Simulation distance,
is this something I should change inside my server config file?
😛
ahh, that's neat,
Thank you both!
so wait we can use simulation-distance as old vanilla ticking view distance and
view-distance as no tick view distance?
Mhh seems like my server.properties doesn't even have the view-distance in it
no wait it does I just can't read :p
I believe so
They aren’t exactly the same
why did they add the differator of simulation distance and view distance to the client anyway, was it for optimization?
singleplayer
1.18.1 is just broken and cannot be fixed?
not on the server no
it's not borken. It's intended 
so you can see the ground from the mountains,,, and the mountain-tops from the ground
( i watched a video where someone did demo of this on youtube )
1.18.1 pushed the fog back some and changed how it works but the client fog still respects the server view distance
So if you want the fog to go away you either need a client mod or a large view-distance in your server.propreties
Is there a reason why its started raining in End Planes biome and bees have stopped producing honey? (paper-1.18.1-81)
The end plains things has been fixed, you might need to regenerate that chunk
cool ty, regenerating might be an issue lol.
I think WE had some tooling to change biomes in a chunk, no idea how that works these days
But a bug in the upgrade process changed the biome of some/a lot of end chunks
Easiest would be to roll back to a 1.17.1 backup of the end or wipe it entirely
end biomes had an oopsie doopsie problem with the dataconverters, you do need to use a backup for those chunks
(it's also why you don't ever forceUpgrade, if there's a bug, the affected chunks is obviously maximized)
Pog leaf finally woke up

