provide a reason why this change is to be made? why are we avoiding retiring the scheduler of players who have been in the world but are in the change state?
#folia-github
1 messages ยท Page 2 of 1
provide a reason why this change is to be made? why are we avoiding retiring the scheduler of players who have been in the world but are in the change state?
It will cause a crash if the player in configuration state and doesn't reply the keepalive from the server for more than 30s.
Like this(offline mode):
C -> S handshake
C -> S loginstart
S -> C loginsuccess
C -> S login acknowledged
S -> C Keepalive
C: Do nothing and keep this state and connection active(send something e...
Reducing the required player spread is a great direction in my opinion! Previously single wandering player would connect regions and cause lag for them from time to time.
Does it have to be based on view distance though? The ticked simulation distance is set lower than view on almost all servers. If players were able to see further, they would feel less restricted and more encouraged to build. This is especially noticable on recent minecraft versions with huge mountains and valleys.
Even w...
Please re-test with https://github.com/PaperMC/Folia/commit/e89a107ef0d4a6f009ad81a2bf329b312c5e6cea
Due to the issue I described earlier, the actual retire() call needs to be in the common listener to ensure that the scheduler is retired in all cases
Is your feature request related to a problem?
Folia seems to be very suitable for loading far lands. But I can't find a teaching mechanism to restore this on it.
Describe the solution you'd like.
I want to change some code to restore far lands, or even change the location where it appears.
Describe alternatives you've considered.
I want to use farlandsagain to restore this mechanism, but this plugin does not support folia.
Other
No response
I don't see what folia has to do with the far lands at all? If you want to re-add them you'll be on your own
Expected behavior
when executing the command ./mcrcon tps, i expect to get CLI output of the tps report given by the ingame command or the server console command "tps"
Observed/Actual behavior
the server console/log spits out the following error.
[18:33:38] [RCON Listener #1/INFO]: Thread RCON Client /192.168.0.100 started
[18:33:38] [Region Scheduler Thread #3/WARN]: Unexpected exception while parsing console command "tps"
java.lang.UnsupportedOperationException: Dispatching c...
I think even if we had a "fire and forget" option that would be great, as already mentioned, even today with Paper you can not really rely on RCON anymore for request<>response operations, but it is useful to send commands. We usually then read the results from the DB (most plugins will write to the db, but they do not action of 3rd party writes to the db)
Expected behavior
No errors :D
Observed/Actual behavior
[21:43:41 ERROR]: Thread Region Scheduler Thread #0 failed main thread check: Accessing entity state off owning region's thread
java.lang.Throwable: null
at io.papermc.paper.util.TickThread.ensureTickThread(TickThread.java:78) ~[folia-1.20.4.jar:git-Folia-"4d78338"]
at org.bukkit.craftbukkit.v1_20_R3.entity.CraftBee.getHandle(CraftBee.java:25) ~[folia-1.20.4.jar:git-Folia-"4d78338"]
at org.b...
Stack trace
https://paste.gg/p/anonymous/7eb2fec54a9442b781e1cebc262d44b3
Plugin and Datapack List
[21:02:15] [Region Scheduler Thread #2/INFO]: Server Plugins (31):
[21:02:15] [Region Scheduler Thread #2/INFO]: Bukkit Plugins:
[21:02:15] [Region Scheduler Thread #2/INFO]: - AquaLimeAdmin, ArmorStandEditor, Brewery, Chunky, CoreProtect, CoreProtectTNT, Essentials, EssentialsSpawn, GSit, HeadDatabase
[21:02:15] [Region Scheduler Thread #2/INFO]: InventoryRollbackPlus, LiteBans,...
Stack trace
Plugin and Datapack List
[08:06:33 INFO]: Server Plugins (31):
[08:06:33 INFO]: Paper Plugins:
[08:06:33 INFO]: - CreeperRecover, HuskHomes
[08:06:33 INFO]: Bukkit Plugins:
[08:06:33 INFO]: - AntiPopup, AuthMe, BetterRTP, BlockLocker, Chunky, CMILib, CoreProtect, Essentials, EssentialsChat, EssentialsSpawn
[08:06:33 INFO]: FoliaFlow, FreedomChat, GSit, LibertyBans, LuckPerms, MiniMOTD, PlaceholderAPI, PlayerSkullDrop, Residence, SimpleWhi...
Looks like you're running a fork of Folia. Reproduce the issue on Folia and ask in Discord or open a new issue.
Sorry, we do not support offline mode servers.
Stack trace
https://paste.gg/p/anonymous/7eb2fec54a9442b781e1cebc262d44b3
Plugin and Datapack List
[21:02:15] [Region Scheduler Thread https://github.com/PaperMC/Folia/issues/2/INFO]: Server Plugins (31):
[21:02:15] [Region Scheduler Thread https://github.com/PaperMC/Folia/issues/2/INFO]: Bukkit Plugins:
[21:02:15] [Region Scheduler Thread https://github.com/PaperMC/Folia/issues/2/INFO]: - AquaLimeAdmin, ArmorStandEditor, Brewery, Chunky, CoreProtect, CoreProtectTNT, Essentials, ...
Please post the full log from startup to crash. You can run it through a Minecraft log paste site to redact IP addresses.
SERVER IS RUNNING WITH VELOCITY(PROXY MODE)
The server is allowing offline mode players to join, and thus this statement is effectively a cover-up to the fact that you're running an unsupported setup.
However, there is a valid bug report here.
actually, nope, not a valid bug report, the errors are being thrown by plugins doing stuff which is basically inheriently unsafe of folia
sorry, I need time to find this log, wait please, ok?
I find this error, you can reopen this issuess? @electronicboy
the log file you showed is not a Folia issue
Entity threw exception at world:1331.7986503388208,62.52326431274413,-1235.555351045593Entity threw exception at world:1331.7986503388208,62.52326431274413,-1235.555351045593Entity threw exception at world:1331.7986503388208,62.52326431274413,-1235.555351045593Entity threw exception at world:1331.7986503388208,62.52326431274413,-1235.555351045593Entity threw exception at world:1331.7986503388208,62.52326431274413,-1235.555351045593Entity threw exception at world:1331.7986503388208,62.52326431...
These are plugin issues, not a Folia bug. You're also running an offline mode server. This issue has been closed and won't be re-opened.
These are plugin issues, not a Folia bug. You're also running an offline mode server. This issue has been closed and won't be re-opened.
Where you are see plugin error? YOU check my new replaced log error?
[10:17]electroniccat[no-ping]: but, spigot (through being a fork of craftbukkit) has changes which often need to record block changes in order to play them back, which is what causes 90% of the oddities with redstone vs vanilla
I'll let leaf reopen this if he so desires
I'll let leaf reopen this if he so desires
who is this?
Stack trace
https://paste.gg/p/anonymous/ae271a7892344bda9d44ccbc004e4c48
Plugin and Datapack List
[21:02:15] [Region Scheduler Thread https://github.com/https://github.com/PaperMC/Folia/issues/2/INFO]: Server Plugins (31):
[21:02:15] [Region Scheduler Thread https://github.com/https://github.com/PaperMC/Folia/issues/2/INFO]: Bukkit Plugins:
[21:02:15] [Region Scheduler Thread https://github.com/https://github.com/PaperMC/Folia/issues/2/INFO]: - AquaLimeAdmin, ArmorStandEditor,...
Please stop re-opening this issue or you will be banned from the Paper org.
Stack trace
Plugin and Datapack List
[07:31:22 INFO]: Server Plugins (30):
[07:31:22 INFO]: Paper Plugins:
[07:31:22 INFO]: - CreeperRecover, HuskHomes
[07:31:22 INFO]: Bukkit Plugins:
[07:31:22 INFO]: - AntiPopup, AuthMe, BetterRTP, BlockLocker, Chunky, CMILib, CoreProtect, Essentials, EssentialsChat, EssentialsSpawn
[07:31:22 INFO]: FoliaFlow, FreedomChat, GSit, LibertyBans, LuckPerms, MiniMOTD, PlaceholderAPI, PlayerSkullDrop, Residence, SimpleWhi...
You appear to be using a fork of Folia, and are running in offline mode. If you can reproduce this issue on Folia without any plugins and in online mode, please post a new issue or come to the Discord for help.
Expected behavior
hopper suck up item instantly when item goes through its above
Observed/Actual behavior
hopper occasionally does not suck up item
Steps/models to reproduce
(translated)
1 place flowing water above a hopper
2 place another hopper below the first hopper and use redstone signal to lock it
3 put 5 set of required item or 1 set of required item in the hopper and fill up the other slot with random full item in the hopper below to make a hopper filter
4 place 1...
I made a workaround for that, I made a script that send a command like /tps in a server running in tmux or docker. And if I don't get an answer, of that a value for the mspt is higher than x amount. Then I kill the server process. (It will look like a little rollback for players, but help in having a resilient server)
I think this issue can be closed. send-proxy-v2 is already supported via paper (and folia) when 'proxy-protocol' is enabled. I also can confirm it via a test setup that it works.
global
log stdout format raw local0
defaults
log global
timeout connect 5000ms
timeout client 50000ms
timeout server 50000ms
frontend minecraft_front
bind *:25565 transparent
mode tcp
default_backend minecraft_back
backend minecraft_back
mode tcp
...
Expected behavior
Ignited TNT will be 100% able to be pushed by piston
Observed/Actual behavior
Some ignited TNT cannot be pushed
Steps/models to reproduce
- use piston to push TNT to anti explode area
- TNT that cannot be pushed successfully will destory the redstone contraptions
Video
https://youtu.be/-oU9bg7o3S8
Plugin and Datapack List
Bukkit Plugins:
AnnouncerPlus, AntiPopup, Attollo, AutoTreeChop, AxInventoryRestore, AxShulkers, BetterRTP, BlockLocker, Blue...
Can you reproduce this without any plugins installed, please?
Expected behavior
Expected behavior for ticking or warning as overloaded?
Observed/Actual behavior
Currently what's happening is Folia seems to be missing a few ticks ->
schedule start and schedule end are exceeding the "10ms" mark causing those regions specifically not to tick which then makes the MSPT metrics be very high with spikes
Steps/models to reproduce
no idea how
Plugin and Datapack List
none
Folia version
master
Other
Is this a configuration pro...
What's your output of /version?
If you're running on the master branch you're on 1.20.2, not 1.20.4. You'd want to be on the most recent build to test any issues/bugs.
Does the region scheduling work based on the global executor and or so how is the scheduling of the regions like activity or inactivity, that might be how the task is getting scheduled, I am running a small custom version that only changes the Async event where i'm passing some data, I can implement debugs to look at the issue further just lmk if you have anything in mind.
(1.20.4 is not the master, I know i'm on 1.20.2)
I kind of added a debug here:
(scheduledEnd-startTime)
for the times it does not tick the problem is -4396147
values are getting to be negative
final long scheduledStart = this.getScheduledStart();
final long scheduledEnd = scheduledStart + TIME_BETWEEN_TICKS;
private long scheduledStart = DEADLINE_NOT_SET;
public static final long DEADLINE_NOT_SET = Long.MIN_VALUE;
the issue is parkNanos on the mechanism of the ticking scheduler.
here are my debugs from prod:
[13:12:15] [Region Scheduler Thread #4/INFO]: [STDOUT] [ca.spottedleaf.concurrentutil.scheduler.SchedulerThreadPool$TickThreadRunner] disable 3 for 26.812526 took 139.299752 1533645036816609
[13:12:15] [Region Scheduler Thread #4/INFO]: [STDOUT] [ca.spottedleaf.concurrentutil.scheduler.SchedulerThreadPool$TickThreadRunner] break 3 on 1533645037444055
[19:29:57 INFO]: This server is running Folia version git-Folia-"60b1508" (MC: 1.20.2) (Implementing API version 1.20.2-R0.1-SNAPSHOT) (Git: 60b1508)
output of /version (there's some local git commits i did so probably will not really match with the master) - none of the commits touch any region ticking.
Expected behavior
block 1 is directly stacked on top of block 2
- player stands on block 1
- player breaks block 1 below themselves
- player falls to block 2
- player does not move/stutter from block 2
this is from the perspective of the player
Observed/Actual behavior
when a player breaks a block below themselves and falls to the next block, the player sees themselves briefly stutter back on top of the block that they already broke and back to where they fell after br...
could not reproduce on latest papermc 1.20.4 build
also tried running the server on a windows 10 machine, repeated all testing, same exact issue persists
repeated all testing but reset config files in-between, same exact issue persists in latest 1.20.4 build but not in previous 1.20.2 version
Expected behavior
The expected behavior is so that when the client sends packets, they get processed "immediately", currently what Folia does for every single packet is schedule each packet as a new task, for 1L. If you notice a simple debug, (I have done this) - The time between the packets arrive and get processed should be "instantly" - this is not really a vanilla only thing, but this affects PVP overall and other actions that players would notice a 50ms latency on their action, SPEC...
For knowledge this issue turned out to be two things:
- Overloaded system which caused the #parkNanos to wake up after the long diff, causing the tick to endup running a "long" time later and then not completing the wanted 20tps.
- After also switching to a low latency kernel this helped a bit, as well as fixing some small thread switching overhead also helped. (This is nice and should probably be in the readme too - overallocating threads wil lcause tps issues due to overloading the JVM).
Expected behavior
Throwing pearl short distance should show and player should be teleported in the same facing direction.
Observed/Actual behavior
Short distance pearl throwing are not visible, player is teleported doing a 180 spin.
Steps/models to reproduce
- Throw a pearl short distance
Plugin and Datapack List
[00:21:21] [Region Scheduler Thread #2/INFO]: Paper Plugins:
[00:21:21] [Region Scheduler Thread #2/INFO]: - MiniPlaceholders, MiniPlaceholders-Player-...
Asked on Discord also, but can you reproduce this without plugins?
I don't see lots of actual solutions for this, since the new threading is pretty much tasks scheduled and shared threads across regions, I feel like possibly having sub-ticks to process incoming packets on the same region scheduled threads would possibly help. (I implemented this).
Expected behavior
nether portal under normal circumstances can be loaded centered on itself 3*3 chunks
Observed/Actual behavior
nether portal chunk loading centered on itself 3*3 chunks seems to have failed
Steps/models to reproduce
Whenever an entity is teleported through a nether portal, the chunk at the linked portal gets load ticket with load level of 30, meaning that it is fully loaded and can process entities. This load level also spreads to adjacent chunks but they get...
I'm running into this bug as well,It can be seen that this plugin has a bug on Folia 1.20.4
Expected behavior
No crash
Observed/Actual behavior
12:47:58] [Region Scheduler Thread #2/ERROR]: Thread Region Scheduler Thread #2 failed main thread check: Cannot move entity off-main
java.lang.Throwable: null
at io.papermc.paper.util.TickThread.ensureTickThread(TickThread.java:71) ~[folia-1.20.4.jar:git-Folia-33]
at io.papermc.paper.chunk.system.entity.EntityLookup.moveEntity(EntityLookup.java:526) ~[folia-1.20.4.jar:git-Folia-33]
at io.papermc.paper.chunk.system.ent...
This happens without any bugs, it's a Folia issue (It happened on DonutSMP too)
Can confirm, moving the project somewhere else which made the path shorter fixed the issue.
Thanks!
To OP: Try moving your project to Desktop, it should work.
Expected behavior
Players are able to join
Observed/Actual behavior
After running Folia on 8939611bb53d6d1d98fd76e91c320d55a28700ad for a couple of hours (varies), players are unable to join. Either stuck in "Encrypting" or "Joining world" screen until connection times-out.
No Exceptions / No Errors. But logs are full of:
[12:38:44] [Region Scheduler Thread #4/INFO]: Foobar (/xxx:20758) lost connection: Disconnected
[12:38:51] [User Authenticator #90/INFO]: UUID of pl...
Expected behavior
A player will be able to enter after being added to the whitelist.
Observed/Actual behavior
After a player entered the server and got a kick with the reason that he is not on the whitelist. If after being added to the whitelist or re-entry, the player can no longer enter the server and is stuck at the entrance until 'Joining world...'.
Steps/models to reproduce
- Enable whitelisting.
- Try to enter the server and get kicked by the whitelist.
- Repeat th...
If you are also experiencing this without whitelist (which I also had issues with independently), then its a duplicate of #200 - I also was able to mitigate by downgrading version (if that is what you mean by Kernel).
Is your feature request related to a problem?
Repeaters etc tend to cause a lot of mspt spikes when farms are ticking
Describe the solution you'd like.
Currently I know that the big issue is likely related to how the wire handler is done, which is always looks for states to update, and that would cause crashes.
Describe alternatives you've considered.
Limiting Redstone a lot is an option but you know.. at some point Folia would need to implement this, because it's a very go...
The server needs to be restarted in these cases.
In reality, this only happens due to the lack of a watchdog - which should automatically kill the server and warn when these events are happening.Is there no other way but to restart?
Imagine other players playing normally, but one or two players getting stuck due to a region bug and needing to restart the entire server
Can we restart bug region only?
You should be able to modify the Folia code and implement yo...
Global region cannot access any region state, which includes entities.
Any mechanic that relies on precise timing of these tickets cannot work on Folia. The ticket is added before the entity is teleported, but Folia adds some number of ticks of delay before the entity arrives to the target location. This doesn't take in account the fact that regions tick in parallel and as a result there is no synchronisation in the ticks between them.
Folia additionally enforces that the chunks in the entity's bounding box is loaded, which may add yet more delay after the tick...
@Spottedleaf Towny has adapted already to the change so it's not a pressing matter any more. I am curious as to why it was working before the 1.20.4 versions. Did your team close a loophole?
This can be closed - and rather create specific ticket? To me I see no difference in usage to Paper: it works "as expected" (e.g. those async issues also exist in paper).
I actually can't see what's going on here, what should I be looking for?
I actually can't see what's going on here, what should I be looking for?
yaw/pitch are ending up to be wrong values, so whenever you throw a pearl, sometimes it'll screw the head rotation and position likely too, so when you're tping even tho you're looking directly to lets say south, on the other POV it'll be NORTH or other direction (north is just example not static).
Yeah this looks like an entity tracker problem that Folia has made worse, do you also see this causing position desyncs as well?
81cc5d2 Switch to experimental SectorFile branch - Spottedleaf
[PaperMC/Folia] New branch created: experiment/sectorfile
Spottedleaf Towny has adapted already to the change so it's not a pressing matter any more. I am curious as to why it was working before the 1.20.4 versions. Did your team close a loophole?
more checks are added to protect the server from entering an unknown state, for which it needs to crash to escape safely without any data corruption
The more there are such checks, the less crashes caused by random plugin actions will happen
Expected behavior
I don't know why I see this.
Observed/Actual behavior
Steps/models to reproduce
don't know
Plugin and Datapack List
pl
[01:59:11 INFO]: Server Plugins (34):
[01:59:11 INFO]: Bukkit Plugins:
[01:59:11 INFO]: - AquaLimeAdmin, ArmorStandEditor, Brewery, Chunky, CoreProtect, CoreProtectTNT, Essentials, EssentialsSpawn, GSit, HeadDatabase
[01:59:11 INFO]: LiteBans, LuckPerms, Maintenance, Mclogs, NewbieFilter, OfflineManager,...
I need a world which showcases the issue so I can produce locally.
e97b1ee Move Connection tick to after world tick - Spottedleaf
The player that reported this issue gave me this map, which is on 1.20.2.
The machine that shown on the video is here, if the spawn point is wrong, the machine is the one above
https://cdn.discordapp.com/attachments/1201178996767993916/1206811602175655936/TEST.zip?ex=65e69863&is=65d42363&hm=6b8846d792b3ea15acde308d096a31fc689dc0d05f7af25ae9d4f3445a041192&
[20:46:21 ERROR]: Not a string |
[20:46:21 ERROR]: Error loading saved data: map_1407 |
java.lang.IllegalArgumentException: Invalid map dimension: null |
at net.minecraft.world.level.saveddata.maps.MapItemSavedData.lambda$load$1(MapItemSavedData.java:156) ~|
[?:?] ...
@Spottedleaf Has anything changed since then? Is it possible to fix the server hang when loading a large number of chests with things with the appearance of Sectorfile?
Are you able to replicate this issue with latest branch?
@Spottedleaf Has anything changed since then? Is it possible to fix the server hang when loading a large number of chests with things with the appearance of Sectorfile?
No
Some prodding at internals with a debugger seems to indicate this is some sort of networking issue? The state I can observe in the server w/ a debugger doesn't match up with what I'm seeing on a client.
This is probably the same problem as https://github.com/PaperMC/Folia/issues/197 given this commentary.
Are you able to reproduce this issue without any plugins?
I don't think it's plugin related as it happens randomly
Expected behavior
When using an EntityTransformer it is expected that the entity (not spawned entity -> not existing yet) can be modified without the need to schedule a task in another thread.
Observed/Actual behavior
Server crashes because the entity access (of a not spawned entity -> not existing yet) because of "Accessing entity state off owning region's thread".
Steps/models to reproduce
- Listen for the AsyncStructureGenerateEvent
- Add EntityTransformer to it
- Jus...
Expected behavior
max-players=x allows x players to join the server.
Observed/Actual behavior
only x-1 can join the server.
> list
[23:17:16 INFO]: There are 0 of a max of 2 players online:
[23:17:23 INFO]: UUID of player Dynam1cNET is e85f9f08-ef80-4523-b363-cf426fbe7d31
[23:17:23 INFO]: Dynam1cNET joined the game
[23:17:23 INFO]: Dynam1cNET[/xxx.xxx.xxx.xxx:49630] logged in with entity id 100 at ([world]150.5, 67.0, -102.5)
> list
[23:17:28 INFO]: There are 1 of a ...
This seems to stem from the check here https://github.com/PaperMC/Folia/blob/e97b1eee63b802568447d344bac5a66bf90c6c71/patches/server/0003-Threaded-Regions.patch#L15976 as the connection of the player trying to join is included already.
Expected behavior
Collection chunks = world.getForceLoadedChunks();
It should get all force loaded chunks of this world.
Observed/Actual behavior
It seems worked on main world, but failed when it comes to world_nether, and throw null dump below.
[17:26:50 ERROR]: Thread Region Scheduler Thread #0 failed main thread check: Scheduling chunk load off-main
java.lang.Throwable: null
at io.papermc.paper.util.TickThread.ensureTickThread(TickThread.java:71) ~[deerfol...
The respawn and teleport APIs are very awkward since the target region for the respawn position or teleport location are not actually owned during the event invocation, which may make modifying the event outcome or even reading from it very challenging. It requires a different event system to properly facilitate this, which is not a priority currently.
EntityPortalEvent doesn't work either
Expected behavior
Players keep raising animals in the world and using portal teleportation to bypass the maximum number of animals in the chunk. This allows for unlimited spawning of animals, causing server lag. I want to ban some animals from teleporting through portal, but in reality,EntityPortalEvent does not work!!!
I want to prevent some entities from entering the portal
Observed/Actual behavior
EntityPortalEvent does not work in Folia
I want an alternative solution
S...
There have already been discussions for a solution for this which involves a new event system concept which has already been discussed and is part of the todo list formulation.
There have already been discussions for a solution for this which involves a new event system concept which has already been discussed and is part of the todo list formulation.
Thank you, I will wait for the new event system. But is there an alternative solution before that? As long as entities can be restricted from passing through the portal, any method is possible
if you just want to prevent entities from portaling in general, can't you just use the enter portal event?
oh, that's not cancellable; not sure what the best approach here would be
Expected behavior
entity spawns normally.
Observed/Actual behavior
[15:46:08 WARN]: [MCCcore] Global task for MCCcore v1.0-SNAPSHOT generated an exception
java.lang.NullPointerException: Cannot read field "captureDrops" because the return value of "net.minecraft.server.level.WorldServer.getCurrentWorldData()" is null
at net.minecraft.server.level.ServerLevel.addEntity(ServerLevel.java:1879) ~[?:?]
at net.minecraft.server.level.ServerLevel.addFreshEntity(ServerLev...
You can't spawn entities at GlobalRegionScheduler. It needs to be in a RegionScheduler.
https://jd.papermc.io/folia/1.20/org/bukkit/Bukkit.html#getRegionScheduler()
sorry for the bad code GitHub messed up the code bud anyway thanks.
Expected behavior
A hopper item filter collecting an item in a waterstream while moving at the appropriate speed, with the 8gt hopper cooldown considered.
Observed/Actual behavior
Items passing hoppers and occasionally failing to pick them up
Steps/models to reproduce
Construct or use a schematic with a waterstream that has hoppers being able to pick up items every game tick. Attached is a farm that uses hoppers to filter out non-compostable items, yet they fail to do so rel...
Is your feature request related to a problem?
--forceupdate doesn't work multithreaded and it's too slow for updating big worlds. updateing 15k*15k world took 30 minutes.
Describe the solution you'd like.
make the world updater update chunks parallel not sequentially.
Describe alternatives you've considered.
there is no alternatives.
Other
have a good day.
It already is threaded, but the math for determining cores could maybe be less restrictive
int threads = Runtime.getRuntime().availableProcessors() * 3 / 8;
Stack trace
[19:20:52] [Region Scheduler Thread #21/ERROR]: Entity threw exception at oneblock:53585.58354723191,203.0,24006.97783524526
java.lang.ArrayIndexOutOfBoundsException: null
[19:20:52] [Region Scheduler Thread #21/ERROR]: [io.papermc.paper.threadedregions.TickRegionScheduler] Region #2052 centered at chunk [3361, 1506] in world 'oneblock' failed to tick:
net.minecraft.ReportedException: Exception ticking world
at net.minecraft.server.MinecraftServer.tickChildren(...
the new build in question
This server is running Folia version git-Folia-"399c242" (MC: 1.20.4) (Implementing API version 1.20.4-R0.1-SNAPSHOT) (Git: 399c242 on ver/1.20.2)
Unknown version
Previous version: git-Folia-"c76cca5" (MC: 1.20.4)
After updated to the latest version, we get the same issue
The actual cause exception is missing, try running with -XX:-OmitStackTraceInFastThrow
[03:24:43] [Region Scheduler Thread #18/ERROR]: [io.papermc.paper.threadedregions.TickRegionScheduler] Region #71730 centered at chunk [-1799, -250] in world 'oneblock' failed to tick:
net.minecraft.ReportedException: Exception ticking world
at net.minecraft.server.MinecraftServer.tickChildren(MinecraftServer.java:1830) ~[folia-1.20.4.jar:git-Folia-"ef2c697"]
at net.minecraft.server.dedicated.DedicatedServer.tickChildren(DedicatedServer.java:447) ~[folia-1.20.4.jar:git-Folia-"ef2c6...
is this issue still activated? I got the same problem as him ;(
is there any way to re-enable them? i know they're unstable but i would like to try some things out to see how broken it actually is
No there is not way to enable them. However it is possible to make plugin with same functionality of commend block. That would be the better approach.
user reported on discord that their server is using patches in order to load unsupported plugins into their environment
Expected behavior
I am sending a message to a player with player.sendMessage(Component) and sometimes (not always) the player is kicked because it can't sign the message.
For instance the message sended contains clickable command text. It only occus when the component built contains click events or hover events
Observed/Actual behavior
The player is kicked on the message reception and the console shows that :
[11:13:47 WARN]: Failed to validate message acknowledgements from S...
Are you using a proxy?
Yes ! Velocity with geyser and floodgate, here is the start logs
[09:35:28 INFO]: Booting up Velocity 3.3.0-SNAPSHOT (git-3bce4d85-b374)...
[09:35:28 INFO]: Loading localizations...
[09:35:28 INFO]: Connections will use epoll channels, libdeflate (Linux x86_64) compression, OpenSSL 1.1.x (Linux x86_64) ciphers
[09:35:28 INFO]: Loading plugins...
[09:35:31 WARN] [floodgate]: en_ is not a supported Floodgate language.
[09:35:31 INFO] [floodgate]: Took 130ms to boot Floodgate
[09:3...
Expected behavior
Running server
Observed/Actual behavior
Error: Unable to initialize main class org.bukkit.craftbukkit.Main
Caused by: java.lang.NoClassDefFoundError: joptsimple/OptionException
Steps/models to reproduce
git init
git pull https://github.com/PaperMC/Folia
bash gradlew publishToMavenLocal
Plugin and Datapack List
Could not be compiled
Folia version
SNAPSHOT
Other
No response
Follow the build steps on Paper's README
Follow the build steps on Paper's README
Ah, working! Thanks ๐
Had used the wrong .jars ๐คฃ
Can confirm. Happens also when client disconnects while enchanting.
Little precision, the player is kicked only when he clicks on the message event.
It happened on my server too, a player lost connection while enchanting a bow, and both the anvil and the bow disappeared
Expected behavior
string dupe machine uses Vanilla Minecraft mechanic. By using Tripwire Hook, String, Flowing Water, Lever with appropriate setup in Vanilla Single player world (or servers), Strings can be duplicate to unlimited numbers.
Observed/Actual behavior
Folia unable to do the same thing in version ef2c6971891b7571261591cd22d27ff555c5d3f2 .
I upgraded from e97b1eee63b802568447d344bac5a66bf90c6c71 to ef2c6971891b7571261591cd22d27ff555c5d3f2 10 days ago, I don't receive a...
Bug not working is a bug, huh? "vanilla mechanic" lmao
Bug not working is a bug, huh? "vanilla mechanic" lmao
It depends on how you view a thing, it can be a feature or a bug.
We all know that beds will explode in the nether and the end since the nether and the end dimension was implemented. Base on your logic it should be a bug, but Mojang decided to added the "Intentional Game Design" message during 1.13 in case someone don't know it is not a bug, it is a intended behavior.
Folia do allow duplication as a feature in case you don't k...
This exploit not working is not a bug, folia pulled in paper's patch for the string dupe. The next time paper is updated folia will also include the option to bring back vanilla behavior.
I will need some world data to produce locally
we cannot provide exact world data due to the nature of the server. however, i've provided a test world with a farm that does not function properly due to the hopper issues. there should be no non-compostable items entering the hoppers above the composting array after the filters.
folia test.zip
This is one of the few issues we still face with Folia in comparison to Paper.
Is your feature request related to a problem?
My server is bungeecord, the core used by server 1 is purpur, server 2 is folia, I have enabled the big ender_chest feature in purpur, but my sync plugin gets an error message ender_chest the size should be 27, but the incoming data is 54
Describe the solution you'd like.
If you can, I'd like to add settings to enable the big ender_chest
Describe alternatives you've considered.
I thought about adding an extra ender chest plugin, b...
There is generally little interest on our part in breaking vanilla compatibility, and such a feature would not belong in folia anyways
Expected behavior
The Conversation API working correctly.
Observed/Actual behavior
This bug already has been reported under:
https://github.com/PaperMC/Folia/issues/24
After nearly a year I'd like to ask if there are any news regarding the broken Conversation API of Folia. Beginning a new conversation with a player results in the chat being broken. No message can be send anymore upon restarting the server.
`
[17:40:57 ERROR]: Chain link failed, continuing to next one
java....
That API is currently in a weird state due to the fact that it relies upon legacy behaviours, it's not currently planned on what we want to do for it on paper, and so I'd imagine it's at a much lower priority for folia to resolve this as part of the 20 other dozen APIs which are currently missing
Thank you for the fast reply @electronicboy! Thats unfortunate but Folia's standpoint on this is understandable.
Expected behavior
In vanilla Minecraft, a placed anvil block will fall down if there is no blocks below itself, it will then stopped on a solid block, or it will turn into an item if the block below is not a solid block such as slabs.
Observed/Actual behavior
In folia, when anvil is placed at Y=0 and it fall down, the anvil will turn into item.
Steps/models to reproduce
Place anvil like this, left side works normal, right side have problem.
 ~[folia-1.20.4.jar:git-Folia-"2986554"]
at net.minecraft.world.entity.Entity.setRemoved(Entity.java:5602) ~[folia-1.20.4.jar:git-Folia-"2986554"]
at net.minecraft.world.entity.Entity.remove(Entity.ja...
We cannot accept stack traces from software that isn't ours
I tried to reproduce this issue on full empty server, only with plugin with PlayerQuitEvent, tried to teleport player to different worlds, but there is no crash.
But I can still reproduce this crash with my plugins on clear folia. This stack traces doesnt contains my methods, that is why I created this issue
This server is running Folia version git-Folia-"25ee657" (MC: 1.20.4) (Implementing API version 1.20.4-R0.1-SNAPSHOT) (Git: 25ee657)\nYou are running the latest version
[14:...
player.teleportAsync(loc)
player.getInventory().clear();
player.getAttribute(Attribute.GENERIC_MAX_HEALTH).setBaseValue(20);
player.setHealth(20);
player.setFoodLevel(20);
player.setSaturation(0);
player.setExhaustion(0);
player.setFireTicks(0);
for (PotionEffect effect : player.getActivePotionEffects()) {
player.removePotionEffect(effect.getType());
}
these actions on a player quit are kind of "dumb", you should schedule those actions for the new player join event instead
UPDATE: It happens occasionally even with LaggRemover removed.
There comes a level of "it's very unlikely to be folia inducing this", if you know what you're doing, modify the constructor for that class to throw an exception if null is passed in instead of actual data
There comes a level of "it's very unlikely to be folia inducing this", if you know what you're doing, modify the constructor for that class to throw an exception if null is passed in instead of actual data
ok im trying
Try to update for 1.20.6.
Known problems:
kyori:adventureversions doesnt match, and4.17.0is not available;MapItemAPI changes;ChunkStatusAPI changes;
btw
remapper - net.fabricmc:tiny-remapper:0.10.1:fat -> net.fabricmc:tiny-remapper:0.10.2:fat
decompiler - net.minecraftforge:forgeflower:2.0.627.2 -> org.vineflower:vineflower:1.10.1
I think the changes are minor, but it still seems necessary
https://github.com/KyoriPowered/adventure/releases/tag/v4.17.0 was just released
I found it. But upstream paper using 4.17.0-snapshot in api, and 4.16.0 in server. I noticed in server build script have a commnet saying "Keep in sync with adventureVersion from Paper-API build file".
So I'm not sure if it should be modified in folia patch, I think maybe its belonging to paper's patch?
Are you sure? It has been updated an hour ago or so.
btw
remapper - net.fabricmc:tiny-remapper:0.10.1:fat -> net.fabricmc:tiny-remapper:0.10.2:fat decompiler - net.minecraftforge:forgeflower:2.0.627.2 -> org.vineflower:vineflower:1.10.1
I think the changes are minor, but it still seems necessary
Yes, UR right. It vital for some class to be decompiled.
https://github.com/KyoriPowered/adventure/releases/tag/v4.17.0 was just released
I found it. But upstream paper using
4.17.0-snapshotin api, and4.16.0in server. I noticed in server build script have a commnet saying "Keep in sync with adventureVersion from Paper-API build file".So I'm not sure if it should be modified in folia patch, I think maybe its belonging to paper's patch?
This has been fixed by [Paper](https://github.com/PaperMC/Paper/commit/5729b294eef2ffd...
Migrations from folia's patch are done, there still a syntax err in net/minecraft/core/dispenser/ProjectileDispenseBehavior.java line 48 DispenserBlock.eventFired should use get() to fetch bollen type. I leave it alone, because it doesnt changed in folias patch, I think it's upstream's work.
Anyway, patches can be compiled allreday, but I found server stucked when preparing level (I think I may need some help).
Server crashed when prepareing level...
crash-2024-05-09_15.09.31-server.txt
Greetings, maintainers. I'm done with 1.20.6 update. Could you review it please?
Fine, I got new problem.
Any commands send by client will cause "Internal Exception: java.lang.UnsupportedOperationException", and then kicked.
Commands send my console is ok.
This fault has no stacktrace, just "Internal Exception", I dont know how to debug it.
enable debug in server.properties or attach a profiler which can do exception recording
but, large PRs should generally be cleared by leaf first, not sure how much acceptance there is towards pulling in an update from somebody else
enable debug in server.properties or attach a profiler which can do exception recording
but, large PRs should generally be cleared by leaf first, not sure how much acceptance there is towards pulling in an update from somebody else
Is he on vacation now? There is no information when we will be able to get a review of this PR or update to 1.20.6?
Fine, I got new problem.
Any commands send by client will cause "Internal Exception: java.lang.UnsupportedOperationException", and then kicked. Commands send my console is ok.
This fault has no stacktrace, just "Internal Exception", I dont know how to debug it.
You can attach exception catchers on the netty threads, that'll be usually it unless you're testing with plugins iirc you're not it seems, I'll get on development in your branch too, to help test out and get it ready.
Fine, I got new problem.
Any commands send by client will cause "Internal Exception: java.lang.UnsupportedOperationException", and then kicked. Commands send my console is ok.
This fault has no stacktrace, just "Internal Exception", I dont know how to debug it.You can attach exception catchers on the netty threads, that'll be usually it unless you're testing with plugins iirc you're not it seems, I'll get on development in your branch too, to help test out and get it rea...
Are you missing something on ProjectileDispenseBehavior? it doesn't build, it's missing .get()
Are you missing something on ProjectileDispenseBehavior? it doesn't build, it's missing .get()
Yes I leave it alone, because I thought that's upstream's work. It doesnt caused by folia patch, so I keep it out of folia's patch.
Yes, that's a folia change, it should be changed, on paper that's a boolean, on folia for thread safety that's a thread local. IT should be on the patch
Yes, that's a folia change, it should be changed, on paper that's a boolean, on folia for thread safety that's a thread local. IT should be on the patch
Ok , got it. I may have a missing.
Things should be looks better now.
this all lgtm, seems to be working already tested the majority of the features, including experimental.
[PaperMC/Folia] New branch created: dev/1.20.6
load check should be before chunk load
did not account for this new field
should not be double-loading the data
we shouldn't be changing the visibility unless required
this patch is needed, as the logic was just moved elsewhere
In general we (Paper included) don't take external work for updates, since there's a lot of things involved that can go wrong
Expected behavior
regenerate chunk file on 1.20.6 version, but maybe structure can't generate or what
Observed/Actual behavior
stopping is server by UnsupportedOperationException
Steps/models to reproduce
I removed the r.4.4.mca fully for fully regenerating on 1.20.6 on folia
Plugin and Datapack List
pl
[12:37:24 INFO]: Server Plugins (41):
[12:37:24 INFO]: Paper Plugins:
[12:37:24 INFO]: - CarbonChat, FoliaWorldLoader
[12:37:24 INFO]: Bukkit Plugins:
[12:37:24 ...
Problem still exist:
This server is running Folia version git-Folia-"25ee657" (MC: 1.20.4) (Implementing API version 1.20.4-R0.1-SNAPSHOT) (Git: 25ee657)
Expected behavior
When spawnRadius is set to 0, players on the server will spawn at [0,0] on their first join or respawn if they don't have a spawnpoint set.
Observed/Actual behavior
Players always spawn at [0,0] no matter what the world spawn is.
Steps/models to reproduce
- Set spawnRadius by
/gamerule spawnRadius 0 - Set a custom world spawn by
/setworldspawn - Get killed or join with a new account
- You still standing at [0,0] instead of the custom world spawn
...
Is your feature request related to a problem?
Release 1.21 coming soon
Describe the solution you'd like.
Version 1.20.6
Describe alternatives you've considered.
No
Other
Hello, could you please update Folia to version 1.20.6? There is a little less than a month left until update 1.21 and I would like to get the latest version 1.20. I saw that it is already in the dev branch, but not sure if I can use this version. Do you have any TODO, what is ready in the dev branch an...
Try to reproduce this using the original Folia. You are using a fork of it.
Best way you can just try and see dev branch, test, and report issues, developers will know what left, of the current state is not ready production for use as they said there no eta.
Expected behavior
normally use plugin like HeadDataBase for custom Heads, I think plugin compitable with paper, but what with Folia idk
Observed/Actual behavior
while player try use HeadDataBase, catch warn:
net.minecraft.ReportedException: Saving entity NBT
and after a time server stopped while 'WorldEdit' try use HeadDataBase, this plugin will crash server with this reason:
Caused by: java.lang.IllegalStateException: Player name contained disallowed characters: 'Head Datab...
The plugin created an improper itemstack, later versions of paper, when rebased, will prevent such items being created in general
The plugin created an improper itemstack, later versions of paper, when rebased, will prevent such items being created in general
faulty plugin is 'HeadDataBase'?
if that's what created that itemstack, then yes
Stack trace
[21:32:28 ERROR]: Error sending packet clientbound/minecraft:set_entity_data (skippable? false)
io.netty.handler.codec.EncoderException: Failed to encode packet 'clientbound/minecraft:set_entity_data'
at net.minecraft.network.codec.IdDispatchCodec.encode(IdDispatchCodec.java:53) ~[folia-1.20.6.jar:1.20.6-DEV-78f49e0]
at net.minecraft.network.codec.IdDispatchCodec.encode(IdDispatchCodec.java:20) ~[folia-1.20.6.jar:1.20.6-DEV-78f49e0]
at net.minecraf...
That error is pretty much always down to plugins creating malformed packets
Stack trace
[12:16:30 INFO]: Preparing start region for dimension minecraft:overworld
[12:16:30 ERROR]: Encountered an unexpected exception
java.lang.IllegalArgumentException: Source: 78
at io.papermc.paper.chunk.system.scheduling.ThreadedTicketLevelPropagator.setSource(ThreadedTicketLevelPropagator.java:42) ~[folia-1.20.6.jar:1.20.6-DEV-78f49e0]
at io.papermc.paper.chunk.system.scheduling.ChunkHolderManager.updateTicketLevel(ChunkHolderManager.java:568) ~[folia-1.20....
While trying to use Conversations, I noticed that it was still using the old Bukkit schedulers.
Stack trace
[21:04:13] [Region Scheduler Thread #3/ERROR]: [io.papermc.paper.threadedregions.TickRegionScheduler] Region #2528 centered at chunk [8434, -2210] in world 'world' failed to tick:
net.minecraft.ReportedException: Exception ticking world
at net.minecraft.server.MinecraftServer.tickChildren(MinecraftServer.java:1830) ~[folia-1.20.4.jar:git-Folia-"25ee657"]
at net.minecraft.server.dedicated.DedicatedServer.tickChildren(DedicatedServer.java:447) ~[folia-1.20.4.jar:git-F...
Expected behavior
on paper
You are running the latest version
Previous version: 1.20.6-DEV-78f49e0 (MC: 1.20.6)```
### Observed/Actual behavior

didn't...
This is covered in the documentation
Expected behavior
patch.bat working as expected
Observed/Actual behavior
+ bash patch.bat
> Configure project :
paperweight-patcher v1.7.1 (running on 'Linux')
FAILURE: Build failed with an exception.
* Where:
Build file '/root/Folia/Folia-Server/build.gradle.kts' line: 3
* What went wrong:
Plugin [id: 'com.github.johnrengelman.shadow'] was not found in any of the following sources:
- Gradle Core Plugins (plugin is not in 'org.gradle' namespace)
- Included B...
Delete the -Server folder
On Fri, 31 May 2024 at 15:36, Andrea Cavalli @.***>
wrote:
Expected behavior
patch.bat working as expected
Observed/Actual behavior
- bash patch.bat
Configure project :
paperweight-patcher v1.7.1 (running on 'Linux')FAILURE: Build failed with an exception.
Where:
Build file '/root/Folia/Folia-Server/build.gradle.kts' line: 3What went wrong:
Plugin [id: 'com.github.johnrengelman.shadow'] was not fo...
Still failing:
> Configure project :
paperweight-patcher v1.7.1 (running on 'Linux')
> Configure project :paper
paperweight-core v1.7.1 (running on 'Linux')
> Task :getPaperUpstreamData FAILED
FAILURE: Build failed with an exception.
* Where:
Build file '/root/Folia/.gradle/caches/paperweight/upstreams/paper/Paper-Server/build.gradle.kts' line: 3
* What went wrong:
Plugin [id: 'com.github.johnrengelman.shadow'] was not found in any of the following sources:
- Gr...
Had the same error when running applyPatches, fresh clone of the repo resolved the issue
Might also work if you just delete the Server and API directories.
either you didn't delete the -Server folder as said, or your repo is in an inconsistent state. worse case, just delete the thing and reclone it.
Currently, apart from using the dev-bundle, we have no way to check the TPS. Therefore, I propose implementing it with either the Location or the Chunk as an argument
Wouldn't it be better to at least store the time in nano seconds? As i remember it's not the fastest method to be called so many times
I don't think this is a good behavior, and it's totally undocumented, maybe change to Optional<double[]>/double @Nullable [] or throw and error?
I think there is already some internal storage of this, because typing tps will show it from the top regions. Would propose to consult that if possible.
I modified it with a Nullable, which I think is the best option.
Wouldn't it be better to at least store the time in nano seconds? As i remember it's not the fastest method to be called so many times
Yes, calling it every time is not a good idea, so I preferred to store it in a variable (like /tps)
Expected behavior
When I try to log in my Minecraft folia server, I have an exception thrown about some crafts I think. I am not using NMS, only using folia-api
Observed/Actual behavior
The exception from the console
pacifista-survie | io.netty.handler.codec.EncoderException: Failed to encode packet 'clientbound/minecraft:update_recipes'
pacifista-survie | at net.minecraft.network.codec.IdDispatchCodec.encode(IdDispatchCodec.java:53) ~[folia-1.2...
For instance, that code was working on folia 1.20.4
Air is not a valid recipe ingredient anymore, see https://github.com/PaperMC/Paper/issues/10748. Try using a space instead as a possible workaround
Expected behavior
Observed/Actual behavior
Steps/models to reproduce
I took a map from a ship underwater (treasure)
Plugin and Datapack List
[19:40:19 INFO]: Server Plugins (19):
[19:40:19 INFO]: Paper Plugins:
[19:40:19 INFO]: - AdvancedServerList, EpicGuard, FairyChat, H...
Expected behavior
see log
Observed/Actual behavior
Steps/models to reproduce
Protocolibplugin
Plugin and Datapack List
[19:40:19 INFO]: Server Plugins (19):
[19:40:19 INFO]: Paper Plugins:
[19:40:19 INFO]: - AdvancedServerList, EpicGuard, FairyChat, HuskHomes
[19:40:19 INFO]: Bukkit Plugins:
[19:40:19 INFO]: - CommandWhitelist, FancyHolograms, FarmControl, IllegalStack, Liberty...
Not a Folia issue, since you are running offline mode please seek support for plugin crashes elsewhere.
Expected behavior
Tamed mobs such as wolves should teleport to their owner when their owner teleports or flies away and lands
Observed/Actual behavior
Tamed mobs do not teleport long distances (presumably across region)
Steps/models to reproduce
Steps with random example coordinates:
/tppos 17000 100 -2000- spawn wolf
- tame with bone
- fly like 1000 blocks away or
/tppos 18000 100 -2000, touch the ground, walk around etc. - wolf doesnt teleport.
- go back to `/t...
This is normal minecraft behaviour, once after you teleport the chunks you were in will unload/no longer be ticked, which prevents the wolf from teleporting to you. Once another player causes the chunk to tick again it will teleport like normal.
Stack trace
paste your stack trace or a paste.gg link here!
Internal Exception: io.netty.handler.codec.EncoderException: Failed to encode packet
Plugin and Datapack List
antiRedstoneClock, BetterRTP, BetterTeams, Chunky, ClearLag, DiscordSRV, Essentials, EssentialsSpawn, Geyser-Spigot, Graves
[21:38:31 INFO]: GriefPrevention, GrimAC, Hibernate, HibernateX, LoginSecurity, LPC, LuckPerms, *OnePlayerSleep, Orebfuscator, PL-Hide
[21:38:31 INFO]: PlaceholderAPI, Protoco...
Wrong repository you're running Paper not Folia, this should be in https://github.com/PaperMC/Paper. Or preferably use our discord for support.
I also noticed this. As I understand it, the logic is embedded in the ensureRunningOnSameThread code located in the PacketUtils class.
gamePacketListener.player.getBukkitEntity().taskScheduler.schedule(
(net.minecraft.server.level.ServerPlayer player) -> {
run.run();
},
null, 1L
);
Why do we need to call packet.handle after 1 tick? Can't we perform this op...
The earliest you can schedule a task on an entity is for the task to run the next time the entity ticks. You can't perform the operation immediately because you need to run the packet logic within the entities context.
The earliest you can schedule a task on an entity is for the task to run the next time the entity ticks. You can't perform the operation immediately because you need to run the packet logic within the entities context.
The behavior here is going to be different because on a vanilla server where there is only 1 thread, you don't need to worry about dispatching packet handles into the correct thread context, there only is 1, and so you can just use the shared task queue which allows pa...
The only current mechanism for ensuring that you get into that context is the entity scheduler, the thing is that the scheduler can only promise to run a task on the next tick. as regions do not spin wait between ticks, there is no mechanism to "immediately" post a task to run against an entity in a fast manner, and you have to remember that entities are not strictly held to a region, and so anything region based is also going to be precarious
The only current mechanism for ensuring that you get into that context is the entity scheduler, the thing is that the scheduler can only promise to run a task on the next tick. as regions do not spin wait between ticks, there is no mechanism to "immediately" post a task to run against an entity in a fast manner, and you have to remember that entities are not strictly held to a region, and so anything region based is also going to be precarious
Thank you for the clear response. Purely h...
Yes. That's what's being done on DonutSMP. (Not 1ms though, since that's not very CPU-wise and you'll likely need low latency kernels for that as well usually). 200 TPS. 5ms subticks.
Yes. That's what's being done on DonutSMP. (Not 1ms though, since that's not very CPU-wise and you'll likely need low latency kernels for that as well usually). 200 TPS. 5ms subticks.
Good solution, glad to hear it works, btw thanks for info.
Are there any news about un-/loading worlds? I also would need it during runtime.
From what I saw it is planed to add world loading support. Does this also include while running?
Try to replicate without any plugins.
Is your feature request related to a problem?
no
Describe the solution you'd like.
I've builded a little tool for my friend that was strugling to find how to get the folias build.
Here's it is:
https://github.com/Izocel/GetPaper
Describe alternatives you've considered.
a static markdown table versions.md
Other
No response
We intentionally don't provide downloads for Folia.
If you can't build Folia yourself, you are not technically skilled to operate a Folia server in its current state.
I can build it but my firneds can't :)
Log sample:
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโฌโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโ
โ (index) โ Values โ
โโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโผโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโโค
โ ----------| 1.20.1 |---------- โ '----------------------------...
Try to replicate without any plugins.
still with no plugins
the exception is gone after setting anti-xray to false
I can confirm this to still be a bug in the latest version of Folia (1.20.6 Patch 5).
Here is a quick recording I did regarding that: https://youtu.be/aquvqg6Yeyc
Is your feature request related to a problem?
have 1.21 updated?
Describe the solution you'd like.
update 1.21
Describe alternatives you've considered.
i dont know sorry.
and i dont speak english so if you dont mean my idea,i'm sorry.
thanks and please :)
(@_@๏ผ
Other
No response
Close the issue, this is not happening any time soon paper is still working on it.
Stack trace
pacifista-hub | [18:11:41 INFO]: [PermissionHandler] Le joueur FunixGaming a รฉtรฉ configurรฉ avec les roles : Admin, Lรฉgendaire, Joueur
pacifista-hub | [18:11:41 ERROR]: Could not pass event PlayerJoinEvent to PacifistaMain v1.0.0
pacifista-hub | java.lang.IllegalStateException: Player is already removed from player chunk loader
pacifista-hub | at io.papermc.paper.chunk.system.RegionizedPlayerChunkLoader.removePl...
Teleporting players inside of the PlayerJoinEvent is unsupported behaviour
Expected behavior
[16:51:02 ERROR]: Thread Region Scheduler Thread #0 failed main thread check: Scheduling chunk load off-main
java.lang.Throwable: null
at io.papermc.paper.util.TickThread.ensureTickThread(TickThread.java:71) ~[folia-1.20.4.jar:git-Folia-"e97b1ee"]
at net.minecraft.server.level.ServerChunkCache.getChunkFutureMainThread(ServerChunkCache.java:293) ~[?:?]
at net.minecraft.server.level.ServerChunkCache.getChunk(ServerChunkCache.java:223) ~[?:?]
...
Stack trace
https://paste.gg/p/anonymous/dbddd2a74a454583b92841c0266fa35a
Plugin and Datapack List
Chunky
Actions to reproduce (if known)
Its random, but always the same error PID
Folia version
1.20.6
Other
No response
Ask your host provider to run a memtest in your ram, you likely have a faulty one.
the thing accessed illegal memory inside of code that the JVM has compiled, general, update java/use the LTS/check memory
gotcha, i will do that. Thanks
Is your feature request related to a problem?
1.21 please
Describe the solution you'd like.
1.21 please
Describe alternatives you've considered.
1.21 please
Other
No response
@xKumorio Please do not open Github issues for estimates on releases.
Stack trace
[23:56:58 ERROR]: Entity threw exception at world:-5676.567540324623,-27.78994546376124,-574.1961417528591
java.lang.IllegalStateException: Accessing entity state off owning region's thread
at io.papermc.paper.util.TickThread.ensureTickThread(TickThread.java:79) ~[folia-1.20.4.jar:git-Folia-"ec6b172"]
at org.bukkit.craftbukkit.v1_20_R3.entity.CraftAxolotl.getHandle(CraftAxolotl.java:22) ~[folia-1.20.4.jar:git-Folia-"ec6b172"]
at org.bukkit.cra...
Is your feature request related to a problem?
update1.21
Describe the solution you'd like.
update1.21
Describe alternatives you've considered.
emm because paper has been already 1.21
Other
No response
@tjw123-tjw Do not open issues for version updates.
Profile link
https://spark.lucko.me/iSSY6h69pZ
Description of issue
Folia 1.21 missing ๐
Paper 1.21 works ๐
Folia 1.21 is urgently needed, Paper has been updated months ago, why does it take so long to update? ๐
โโ๏ธ
Please respect minecraft community and do your job, provide 1.21 Folia fast so all servers such as DonutSMP can update. Currently we lack new features, this is unacceptable.
UPDATE NEEDED ASAP! โ
Also why does it take so long at all? Possible reason is that you...
Expected behavior
Possible reason is that you rewrite big portions of the game such as chunk system every time! Even worse, only one person works on all of this... this is BAD, such software can not rely only on one person, other maintainers don't even have a clue about changes.. Stop rewriting everything! Give us normal update!
Observed/Actual behavior
Folia 1.21 is urgently needed, Paper has been updated months ago, why does it take so long to update? ๐
โโ๏ธ
Please respect minecra...
Is your feature request related to a problem?
1.21 ASAP needed. Do not rewrite the whole game everytime just because you want to... it takes too much time to update because you spend time rewriting and rewriting for nothing! Nobody can help with update because nobody understands what that one person decided to rewrite AGAIN this time!
Describe the solution you'd like.
Abandon chunk rewrite and integrate multithreading better way to make DonutSMP run on 1.21
Describe alternativ...
Is your feature request related to a problem?
When is Folia 1.21 estimated to release??! Seems like it's a DEAD project. Paper was updated months ago but Folia is unable to catch up! Maybe Folia 1.21 will come after 1.22?
Also, new implementation, ShreddedPaper already works better!!! How come??? Better multithreading, huge performance boost even for small worlds without "YoU nEED 128 CORES TO RUN IT" crap!
Sadly, Paper rewrote 50% of the server again just for the sake of it so Shred...
We don't provide any estimates on updates because people who work on the project work on it on their own free time. We also provide no warranty for any of our software.
Folia is still experimental so there should be no expectation of updates or stability.
Can you stop?
Are you okay? There is 0 communication from developers about FOLIA! Nothing! No mention about solving current issues, no future plans, paper discord is just "don't ask this, don't ask that, we don't know anything because we don't understand how folia works" LOL! How can this be a public project? Just hide it!
Are you okay? There is 0 communication from developers about FOLIA! Nothing! No mention about solving current issues, no future plans, paper discord is just "don't ask this, don't ask that, we don't know anything because we don't understand how folia works" LOL! How can this be a public project? Just hide it!
show your true face, add me on discord pedrojava
Folia is not production software, it's basically a single person personal project, there's no guaranteed updates, no support, and no official builds. If you want it updated, update it yourself, or wait.
If a server uses Folia such as 2B2T or DonutSMP it's their choice to use software which may never get updated again, they are able to update and build it themselves.
Stack trace
[11:43:43] [Region Scheduler Thread #2/ERROR]: [io.papermc.paper.threadedregions.TickRegionScheduler] Region #2 centered at chunk [5, -2] in world 'world' failed to tick:
net.minecraft.ReportedException: Exception ticking world
at net.minecraft.server.MinecraftServer.tickChildren(MinecraftServer.java:1901) ~[folia-1.21.jar:1.21-DEV-388cdac]
at net.minecraft.server.dedicated.DedicatedServer.tickChildren(DedicatedServer.java:473) ~[folia-1.21.jar:1.21-DEV-388cdac]
at net...
The 1.21 branch has just been created. There are likely to be issues. Please be patient, and do not open Github issues for stack traces on brand new dev branches.
13e5373 Move chunk lists and nearby players to region data - Spottedleaf
thanks you for fixing and smart :) enjoy
Stack trace
[19:11:47] [Region Scheduler Thread #7/ERROR]: [io.papermc.paper.threadedregions.TickRegionScheduler] Region #1410 centered at chunk [-167, -34] in world 'world' failed to tick:
java.lang.ArrayIndexOutOfBoundsException: Index 6 out of bounds for length 5
at me.lucko.spark.paper.common.monitor.tick.SparkTickStatistics$TpsRollingAverage.add(SparkTickStatistics.java:181) ~[spark-paper-1.10.84-20240720.204128-1.jar:?]
at me.lucko.spark.paper.common.monitor.tick.SparkTic...
The same problem. The reason for the appearance is not known.
lucko spark when is integrated in paper doesn't fully support folia, so you just can disable sparkl in paper config for fix this issues for while-time
#253
It's different.
Isn't it the same thing?
the build is the same in the new build, it seems to have been fixed
No, it's an entirely different error. First one is some internal fields relating to ticking, this is spark
Expected behavior
Need normal seen all near entity(mob, player and etc) and default flying on elytra.
Observed/Actual behavior
https://youtu.be/U6ELRImRkXE - bugs elytra. For fix elytra - needed rejoin.
https://youtu.be/M54H7hwU6rY - bugs with entity.
https://www.youtube.com/watch?v=7ee1KrqaIM4 - secondary case bug with entity.
https://youtu.be/cgHILv6xtMQ - third case bug with entity.
Steps/models to reproduce
This faults work with and without plugins only on clean Folia....
I confirm the problem with the entity
Expected behavior
Normal player behavior
Observed/Actual behavior
Invisible and frozen players
Steps/models to reproduce
use folia 1.21 (61dcc07) without any plugins through velocity (3.3.0-SNAPSHOT-413 with tcpshield plugin only) under tcpshield protected domain.
But probably it's folia issue.
(https://youtu.be/7nUHzfhuekk)
Plugin and Datapack List
Folia version
61dcc07. 1.21
Other
No response
what about elytra? this will be fixed too?
Is your feature request related to a problem?
It may be, technically speaking, there's a small "gap" in knowledge, basically from what I see, there's two types of ticking, you have a global ticking per region, I am not sure if it's per region but from my debugs it was running a lot more than just once, I am not sure if this will be per level.
basically, on the current task grabbing, the thread basically parkNanos until the deadline, the problem is, a newer task with a lower deadline (...
I am not experiencing this exact issue, due to the custom implementation on https://github.com/PaperMC/Folia/issues/196, the issue kind of diminishes itself, but on a high count of regions and "20 ticks" of thread per sec, meaning around 50ms gabs every task, this issue will most def appear a lot more.
This is already an invariant maintained by the scheduler (SchedulerThreadPool). For a thread pool of size n, the earliest n tasks are the tasks being blocked on.
This is already an invariant maintained by the scheduler (SchedulerThreadPool). For a thread pool of size
n, the earliestntasks are the tasks being blocked on.
Yes, but that's earliest based on current data, if you have 128 regions, you may also want to have more than 64 threads otherwise you can run into scheduling problems?
It's earliest based on whichever tasks are scheduled.
Stack trace
https://mclo.gs/8jBvH6y or https://paste.gg/p/anonymous/adca91c1f59d411a95ca1b514edeea74
Plugin and Datapack List
spark-1.10.60-bukkit.jar
Actions to reproduce (if known)
When I launched about 350+ bots and suddenly closed the application with bots, they all suddenly left the server and the server turned off and crashed.
Folia version
folia-bundler-1.21-R0.1-SNAPSHOT-reobf.jar
Other
No response
Expected behavior
Players get effects from scheduled task via my custom plugin
Observed/Actual behavior
Effects did not apply and the console logged an error
`
Cannot add effects to entities asynchronously
at ca.spottedleaf.moonrise.common.util.TickThread.ensureTickThread(TickThread.java:59) ~[folia-1.21.jar:1.21-DEV-2b8c879]
at net.minecraft.world.entity.LivingEntity.addEffect(LivingEntity.java:1167) ~[folia-1.21.jar:1.21-DEV-2b8c879]
at net.minecraft....
don't mutate entity state from the global scheduler
Stack trace
https://paste.gg/p/anonymous/1ce0ae5effd64dc2abda2911d05b07cf
Plugin and Datapack List
Actions to reproduce (if known)
No response
Folia version
Other
Stopped the server on Folia 1.21
Stack trace
paste your stack trace or a paste.gg link here!
Another Fork https://spark.lucko.me/SnZ2v6VqBq ๏ผRun 13h๏ผ
Folia https://spark.lucko.me/0IPMgOFZdp ๏ผRun 10m๏ผ
Folia https://spark.lucko.me/tPQbINjmur ๏ผRun 20m)
Folia https://spark.lucko.me/v8nIIqWVGB (Run 30m)
Plugin and Datapack List
plugins
[11:42:20 INFO]: Server Plugins (36):
[11:42:20 INFO]: Paper Plugins:
[11:42:20 INFO]: - MiraiMC
[11:42:20 INFO]: Bukkit Plugins:
[11:42:20 INFO]: - ajLeaderboar...
Stack trace
[paste your stack trace or a paste.gg link here!](https://pastes.dev/P5FlpIJyMF)
Plugin and Datapack List
plugins
[14:35:20 INFO]: Server Plugins (36):
[14:35:20 INFO]: Paper Plugins:
[14:35:20 INFO]: - MiraiMC
[14:35:20 INFO]: Bukkit Plugins:
[14:35:20 INFO]: - ajLeaderboards, AntiPopup, ArmorStandEditor, AxInventoryRestore, BetterGUI, ChestProtect, Chunky, CommandWhitelist, CoreProtect, CreeperConfetti
[14:35:20 INFO]: DeathMessage, Essentials, GSit,...
I can confirm this too, because this is very strange vioalitons old object. My spark: https://spark.lucko.me/yxaIfmfKol
Expected behavior
I hope to fix this bug as soon as possible because it is really causing my server to malfunction
Observed/Actual behavior
This is the crash log I captured, and every crash is a similar log
[20:03:16] [Region Scheduler Thread #4/ERROR]: [io.papermc.paper.threadedregions.TickRegionScheduler] Region #112963 centered at chunk [3043, 13069] in world 'world' failed to tick:
java.lang.IllegalStateException: Accessing entity state off owning region's thread
at ca.sp...
Stack trace
paste your stack trace or a paste.gg link here!
Plugin and Datapack List
Actions to reproduce (if known)
The server crashes every time the player enters the bug and logs in to the server.
Folia version
Other
No response
The previous folia version appears on Mclogs, but after the crash, I updated it to the latest folia version and the same situation still occurs. The player was forced to change his account.
Expected behavior
The damage should go to the dragon - that's right.
Observed/Actual behavior
Damage to dragon does not go through correctly.
Steps/models to reproduce
Steps to reproduce the bug:
- Create a new map and go to the world_the_end to the dragon.
- Restart server command /stop
- We launch the server and try to hit the dragon with a sword, almost no damage is done. Doesn't register hits
- When the dragon descends to the portal and is above it, it cannot be h...
The player also reported that when he killed the ender dragon, he should have flown to the frame with the portal before dying, but he didn't do it.
I played in this server, and I would like to provide more information on the player side.
As we observed, The server will need to run for more than six(6) hours to present a sudden TPS drop (as low as <1.5). For players in the server, the TPS would drop to 1~2 for few seconds, then backs to normal(~19.5 TPS) for around 15 seconds, and repeats the malfunction. As about 15 minutes pass, the frequency of this malfunction becomes higher and the normal TPS time shortens to about 5 seconds. Duri...
Expected behavior
Fix bugs
Observed/Actual behavior
https://github.com/user-attachments/assets/ab1b876b-3926-4790-af0b-c841ad81b165
Steps/models to reproduce
This server is running Folia version 1.21-DEV-dev/1.21@61dcc07 (2024-08-01T02:15:43Z) (Implementing API version 1.21-R0.1-SNAPSHOT)
You are 3 version(s) behind
Download the new version at: https://papermc.io/downloads/paper
This is the version I use
Plugin and Datapack List
:
@EventHandler
public void onRespawn(InventoryCloseEvent event) {
Player player = (Player) event.getPlayer();
if (event.getInventory().getType() != InventoryType.CRAFTING || !player.isDead() || !player.isConnected() || player.getHealth() > 0)
return;
// do stuff
}
To listen and/or cancel entity entering a portal, looks like only [EntityPortalEnterEvent](...
Expected behavior
Applying trim to the armor
Observed/Actual behavior
The trim pattern does not apply to the armor, as a result it takes away the trim and the armor remains default
Steps/models to reproduce
A simple attempt at applying trim to the armor
Plugin and Datapack List
Server Plugins (16):
Paper Plugins:
- FancyHolograms
Bukkit Plugins: - Chunky, Essentials, JBWM-Plots, JbwmLib, JbwmShop, LuckPerms, Matrix, PlaceholderAPI, ProtocolLib, Survival
Vau...
Stack trace
Plugin and Datapack List
plugins
[14:50:19 INFO]: Server Plugins (45):
[14:50:19 INFO]: Paper Plugins:
[14:50:19 INFO]: - CarbonChat, FancyNpcs, FoliaWorldLoader
[14:50:19 INFO]: Bukkit Plugins:
[14:50:19 INFO]: - ArmorStandEditor, AutoWhitelistRemove, BreweryX, CoreProtect, CoreProtectTNT, Craftorithm, DeluxeMenus, emotecraft, Essentials, EssentialsSpawn
[14:50:19 INFO]: GSit, HeadDatabase, InventoryRollbackPlus, InvisibleFrames, InvS...
Server crashed when if player has one of two effects bad omen or hero of village
Expected behavior
After killing the Ender Dragon, jump into the portal, and when the End Poem is playing, press the ESC key to return to the main world.
Observed/Actual behavior
After killing the Ender Dragon for the first time, jump into the portal and use ESC to skip the End Poem, which will show that the player has been falling in the End.
Steps/models to reproduce
After starting Folia, enter the End and kill the Ender Dragon, then jump into the portal and press ESC to ski...
Most of the time you get stuck when you enter the End portal and it doesn't get fixed without restarting the server.
Expected behavior
Need Update I think
Observed/Actual behavior
Steps/models to reproduce
Plugin and Datapack List
Folia version
Other
No response
Its fault. Looks like plugin problem
Stack trace
paste your stack trace or a paste.gg link here!
https://mclo.gs/XBazwWN
Plugin and Datapack List
Actions to reproduce (if known)
No response
Folia version
Other
It is predicted that the problem is caused by Folia. It gives the same error in some plugins. Th...
[PaperMC/Folia] New comment on issue #266: glow item frame invisible after 1.21 folia restart server
Sorry, I updated the latest version of folia-1.21 and the display was restored, not a plugin issue
[PaperMC/Folia] New comment on issue #266: glow item frame invisible after 1.21 folia restart server
Sorry, I updated the latest version of folia-1.21 and the display was restored, not a plugin issue
[PaperMC/Folia] New comment on issue #266: glow item frame invisible after 1.21 folia restart server
I'm unable to reproduce the issue on newest Folia.
Is your feature request related to a problem?
1.21.1 update is needed ASAP! Hopefully it won't require another chunk rewrite 50K lines patch and 32 more cores minimal requirement.
Also open issues need a FIX, such as raid / ender dragon crash and memory leaks.
Please clarify whether we should expect 1.21.1 FOLIA at all or not.
Describe the solution you'd like.
Solution is to release working 1.21.1 Folia without bugs and memory leaks.
Describe alternatives you've considered...
So many woos, so little yeahs
[PaperMC/Folia] New branch created: dev/1.21.1
[PaperMC/Folia] branch deleted: dev/1.21
you haven't ran save-off have you? otherwise, something is keeping millions of entity instances around, would really need to see a heap dump to see where it's coming from, not that I can offer much
you haven't ran save-off have you? otherwise, something is keeping millions of entity instances around, would really need to see a heap dump to see where it's coming from, not that I can offer much
I think nobody run /save-off. On my server only /save-on always
Stress tested by using https://github.com/PureGero/minecraft-stress-test/ on the latest version but cannot reproduce.
Stress tester start commandline: java -Dbot.count=200 -Dbot.ip=127.0.0.1 -Dbot.port=18750 -Dbot.login.delay.ms=5000 -Dbot.radius=10000 -jar minecraft-stress-test-1.0.0-SNAPSHOT-jar-with-dependencies.jar
Plugins:
Server start commandline: ./zulu22.30.13-ca-jd...
you haven't ran save-off have you? otherwise, something is keeping millions of entity instances around, would really need to see a heap dump to see where it's coming from, not that I can offer much
I think nobody run /save-off. On my server only /save-on always
Someone will always use a command if it's available. For example save-off and save-on during backups makes sure that the server does not make any changes to whats going in the backup while it is being done.
Stress tested by using https://github.com/PureGero/minecraft-stress-test/ on the latest version but cannot reproduce. Stress tester start commandline: java -Dbot.count=200 -Dbot.ip=127.0.0.1 -Dbot.port=18750 -Dbot.login.delay.ms=5000 -Dbot.radius=10000 -jar minecraft-stress-test-1.0.0-SNAPSHOT-jar-with-dependencies.jar Plugins:  bytes. The memory is accumulated in one instance of java.lang.Object[], loaded by <system class loader>, which occupies 23,189,861,312 (32.88%) bytes.
22GB by Spark, huh?
Go to plugins/spark/config.json and change backgroundProfiler to false.
Expected behavior
Well, everything had to be crafted
Observed/Actual behavior
nothing was crafted
Steps/models to reproduce
[06:33:33 ERROR]: Failed to handle packet net.minecraft.network.protocol.game.PacketPlayInWindowClick@69e6ff76, suppressing error
java.lang.IllegalArgumentException: Recipe cannot have an empty result
at com.google.common.base.Preconditions.checkArgument(Preconditions.java:143) ~[guava-32.1.2-jre.jar:?]
at org.bukkit.inventory.Craf...
One instance of me.lucko.spark.common.SparkPlatform loaded by org.bukkit.plugin.java.PluginClassLoader @ 0x22b4fbfe848 occupies 23,189,898,336 (32.88%) bytes. The memory is accumulated in one instance of java.lang.Object[], loaded by <system class loader>, which occupies 23,189,861,312 (32.88%) bytes.22GB by Spark, huh? Go to
plugins/spark/config.jsonand changebackgroundProfilertofalse.
I tried, but seems it is still leaking
What if you remove spark completely?
The player got into a bug again and the server crashed when he entered the server. The player had to create a new account.
What if you remove spark completely?
Yeah
@MineSunshineone Leak suspect reports are generally useless, but, the spark plugin you've got installed is eating like 22G of ram
@MineSunshineone Leak suspect reports are generally useless, but, the spark plugin you've got installed is eating like 22G of ram
I know this. When I saw this report, I removed Spark and started the second test. There was still a memory leak. I guess it was because I didnโt completely remove the built-in Spark.
Same issue If the server is up for a long time. TPS drops suddenly, usually around 5. It recovers after 10 seconds and then drops suddenly again. This problem does not get fixed until you restart the server. When you look at /tps, it shows meaningless values.
me too, think its because its still on development and the recipes are bugged
Expected behavior
Completing the "This boat has a legs" achievement
Observed/Actual behavior
I can do the "Feels Like Home" achievement, but the previous achievement "This boat has a legs" doesn't work.
Steps/models to reproduce
- Take "warped fungus on a stick."
- Put the saddle on Strider.
- Sit on the saddle and try to control the Strider.
Plugin and Datapack List
Test...
having same issue, really waiting for a fix
https://github.com/LuminolMC/Luminol/blob/dev/1.21.1/patches/api/0006-Try-fixing-folia-273.patch Might be helpful for this issue
No...
did you tried to complie whit that pach?
Next update upstream with Paper fix the problem.
Next update upstream with Paper fix the problem.
This patch is not completely present on last upstream of Folia
okay, i will try to complie that branch and if that not works i will try to just only add that patch to master branch
Next update upstream with Paper fix the problem.
https://github.com/PaperMC/Paper/blob/f97aff74b6f7fd4940c0c0b6cca3f8ac6e1afdef/patches/api/0469-Fix-issues-with-recipe-API.patch#L51C10-L51C105
This patch is not completely present on last upstream of Foliaokay, i will try to complie that branch and if that not works i will try to just only add that patch to master branch
EDIT: realized thats papermc patch, so i will try to add it to foliamc as 0006 and complie it
...
Next update upstream with Paper fix the problem.
https://github.com/PaperMC/Paper/blob/f97aff74b6f7fd4940c0c0b6cca3f8ac6e1afdef/patches/api/0469-Fix-issues-with-recipe-API.patch#L51C10-L51C105
This patch is not completely present on last upstream of Foliaokay, i will try to complie that branch and if that not works i will try to just only add that patch to master branch
EDIT: realized thats papermc patch, so i will try to add it to foliamc as 0006 and comp...
Next update upstream with Paper fix the problem.
https://github.com/PaperMC/Paper/blob/f97aff74b6f7fd4940c0c0b6cca3f8ac6e1afdef/patches/api/0469-Fix-issues-with-recipe-API.patch#L51C10-L51C105
This patch is not completely present on last upstream of Foliaokay, i will try to complie that branch and if that not works i will try to just only add that patch to master branch
EDIT: realized thats papermc patch, so i will try to add it to foli...
Stack trace
paste your stack trace or a paste.gg link here!
https://mclo.gs/XTdEwoV
Plugin and Datapack List
Actions to reproduce (if known)
No response
Folia version
Other
No response
In 053b04c7a1541cca6cda6f0bd1fb70cd2525c407 the problem is still present.
this will be fixed in 1.21.1?
This is caused by some plugin on your server.
Perhaps should add a check that the world is null and have the method behave differently.
// Paper start - more teleport API / async chunk API
@Override
public java.util.concurrent.CompletableFuture<Boolean> teleportAsync(final Location location, final TeleportCause cause, final io.papermc.paper.entity.TeleportFlag... teleportFlags) {
Preconditions.checkArgument(location != null, "location");
location.c...
This is caused by some plugin on your server. Perhaps should add a check that the world is null and have the method behave differently.
// Paper start - more teleport API / async chunk API @Override public java.util.concurrent.CompletableFuture<Boolean> teleportAsync(final Location location, final TeleportCause cause, final io.papermc.paper.entity.TeleportFlag... teleportFlags) { Preconditions.checkArgument(location != null, "location"); ...
I found the problem, yes there is more than this log. It is caused by the BadOmen effect. This error and crash still continues in version 1.21.1.
#264 #268
I doubt it's because of vanilla mechanics. Specifically here the call happens on the API.
I found the problem, yes there is more than this log. It is caused by the BadOmen effect. This error and crash still continues in version 1.21.1.
#264 #268I doubt it's because of vanilla mechanics. Specifically here the call happens on the API.
Check the logs on the number I tagged. The log here is misleading. It comes entirely from badomen.
If you clear the player's effect with the /effect clear command, you will get rid of this crash. You must do this very quickly when the player enters the server.
This PR fixes flickering of entities without passengers when teleporting them with teleportAsync.
Before:
https://github.com/user-attachments/assets/116a5598-c2e3-4208-b4fd-f7a77bcdb073
After:
https://github.com/user-attachments/assets/5514329c-1244-4c77-a79f-946968c7ce75
Used a modified version of https://github.com/TheCymaera/minecraft-spider for the examples above
i test in 9203a7a887eecdccb93a96ca460ce13364842884
The portal from the endlands into the endlands that I tested last night looked like this:
It went straight through the portal and dropped into the space below, without returning to the main world as in the single player game
TPing through the portal was instantaneous, and then the player went into a weird state:
He can't see anyone else, they can't see him, he can't TP away, he's always online, and he can't be taken offline by kicks or b...
Experiencing the same. when Raid.createOrExtendRaid is called, server crashes.
#264 - duplicate
I think this issues is not have full logs all errors. So I think mine issues is not duplicate
Expected behavior
Fine work Brewery plugin on Folia, but
Observed/Actual behavior
Steps/models to reproduce
install Brewery plugin on folia.
Plugin and Datapack List
[13:48:49 INFO]: Server Plugins (46):
[13:48:49 INFO]: Paper Plugins:
[13:48:49 INFO]: - CarbonChat, FancyNpcs, FoliaWorldLoader
[13:48:49 INFO]: Bukkit Plugins:
[13:48:49 INFO]: - ArmorStandEditor, AutoWhitelistRemove, BreweryX, CoreProtect, CoreProtectTNT, Craftorithm, ...
Its a plugin issue
do you have idea how fix it?
com.dre.brewery.Barrel.isSmall() must be called on the region scheduler
Stack trace
https://paste.gg/p/anonymous/58cceef8a2d3449381d518592ed5a308
Plugin and Datapack List
No plugins
Actions to reproduce (if known)
Unknown
Folia version
[16:45:06 INFO]: This server is running Folia version 1.21.1-DEV-dev/1.21.1@9203a7a (2024-08-19T16:38:24Z) (Implementing API version 1.21.1-R0.1-SNAPSHOT)
You are running the latest version
Previous version: 1.21.1-DEV-053b04c (MC: 1.21.1)
Other
Temp fix: /gamerule disableRaids true
Similar issue with villager passes through a nether portal, same exception for Accessing poi chunk off-main:
#folia-help message
The raid omen effect issue is a duplicate of #264, please create seperate issues for other issues
Expected behavior
As of 1.21, ender pearls can be teleported to each other in other dimensions, such as when the player is in the Nether and the Overworld has a pearl to teleport, they can be teleported directly, and they will work in the same way in the End
Observed/Actual behavior
As of 1.21, ender pearls can be teleported to each other in other dimensions, such as when the player is in the Nether and the Overworld has a pearl to teleport, they can be teleported directly, and the...
Expected behavior
-high speed elytra flying above nether ceiling while chunk generating
-logout
-login normally
Observed/Actual behavior
-high speed elytra flying above nether ceiling while chunk generating
-logout
-unable to login, chunk being locked
Steps/models to reproduce
-high speed elytra flying above nether ceiling while chunk generating
-logou...
Expected behavior
In Minecraft version 1.21 and later, when an entity uses an End Portal to enter the End, blocks placed on the Obsidian Platform should turn into item drops rather than disappearing. This ensures that valuable resources are not lost during the portal transition.
Observed/Actual behavior
When an entity uses the End Portal to enter the End, blocks placed on the Obsidian Platform directly disappear instead of turning into item drops.
Steps/models to reproduce
Pl...
Expected behavior
Citizens plugin's NPCs of type Player cannot be displayed
Observed/Actual behavior
Citizens plugin's NPCs of type Player cannot be displayed
Steps/models to reproduce
picture: has Int2ObjectMap K
but in folia there didnt have Int2ObjectMap K
but Citizens plugin need it
how can I do to fix it
Citizens code:
@Override
public EntityPacketTracker getPacketTra...
Please do not download the above file.
Please do not download the above file.
ik, ty
why the swash sticker cross hair?
Folia has to make a lot of changes internally to work properly and safely, re-adding that map is not possible. Plugins have to adapt to folia, not the other way around.
Profile link
not necessary
Description of issue
Currently there's a method that used to be called on ticking of entities, trackUnload
This used to be the method called on 1.20.6 and below:
which would clear internally on the method, the ref list toPRocessTrackingUnloading.
ON 1.21 and upfront the method was completely forgotten, and you ended up using another ref list, called trackerUn...
if anyone is interested:
https://github.com/PaperMC/Folia/issues/283 here's the cause.
I totally agree with you. I have this problem with 1.21.1. I never had this problem with 1.20.4. I migrated my server to 1.21.1 and my server is essentially a vanilla server that aims for high entity without compromising on the experience. I use the same plugins but with the latest version 1.21.1 I can't keep my server up for more than 7-8 hours because all my ram is empty and I have to restart the server.

/datapack list unknown or incomplete command
Folia version
1.20.6-DEV...
How many blocks is enough to notice an effect? Could you maybe record the video for example?
Is your feature request related to a problem?
It's kinda hard as a plugin developer to tell what the do's and don'ts are of writing a folia plugin
Describe the solution you'd like.
A guide for plugin developers, also a mode with strict thread checks everywhere, ensuring we are doing things correctly to avoid race conditions. For example thread checks EVERYWHERE where cross region access occurs, maybe plugin scanning for non thread-safe collections (silly idea maybe), stuff like th...
I'm pretty sure your runOnEachRegion example would require locking the regions while it runs to prevent them from merging or splitting in the middle, all regions would have to stop ticking to process this.
Also, this is like 3-8 issues in one, you should split it up in to API requests, documentation requests, and server feature requests (the flag).
I am not sure how to describe. If for example you have a chest full of
blocks and you mine the chest you would expect all the blocks that where in
that chest to be there but that is not always the case. Same thing happens
if you are mine a lot of blocks quickly but not picking up all the blocks
while another player observes. The observing player can see some of the
blocks that where not picked up will randomly disappear in a short period
of time
On Sun, Sep 1, 2024 at 2:59โฏPM xymb...
Profile link
https://github.com/PaperMC/Folia/tree/dev/1.21.1
Description of issue
Description: When exploring and loading new chunks in the minecraft:mangrove_swamp biome, noticeable lag occurs. This causes a significant delay in world generation, affecting the overall gameplay experience.
Steps to Reproduce:
Start Minecraft and generate a new world.
Locate an area with the minecraft:mangrove_swamp biome.
Load fresh chunks in this biome (e.g., by moving or teleporting to u...
suddenly, in a new world, no more lags
Is your feature request related to a problem?
a
Describe the solution you'd like.
a
Describe alternatives you've considered.
a
Other
No response
Is your feature request related to a problem?
a
Describe the solution you'd like.
a
Describe alternatives you've considered.
a
Other
No response
Is your feature request related to a problem?
a
Describe the solution you'd like.
a
Describe alternatives you've considered.
a
Other
No response
I'm pretty sure your
runOnEachRegionexample would require locking the regions while it runs to prevent them from merging or splitting in the middle, all regions would have to stop ticking to process this.Also, this is like 3-8 issues in one, you should split it up in to API requests, documentation requests, and server feature requests (the flag).
done
Cracked server, no support
not sure what this has to do with thread locking?
Cracked server, no support
not sure what this has to do with thread locking?
Expected behavior
Player performs the command successfully.
Observed/Actual behavior
[15:23:05 ERROR]: [ca.spottedleaf.moonrise.common.util.TickThread] Thread Region Scheduler Thread #0 failed main thread check: Dispatching command async
java.lang.Throwable: null
at ca.spottedleaf.moonrise.common.util.TickThread.ensureTickThread(TickThread.java:58) ~[folia-1.21.1.jar:1.21.1-DEV-88e0907]
at org.bukkit.craftbukkit.CraftServer.dispatchCommand(CraftServer.java:1032) ~[folia-...
commands need to be dispatched in the correct context, i.e. the players tick thread (see the player scheduler)
Can't I just call Player#chat("/" + command)?
Upon trying, looks like there is a bug in Folia:
[15:56:54 WARN]: java.lang.IllegalStateException: PlayerCommandPreprocessEvent may only be triggered synchronously.
[15:56:54 WARN]: at io.papermc.paper.plugin.manager.PaperEventManager.callEvent(PaperEventManager.java:42)
[15:56:54 WARN]: at io.papermc.paper.plugin.manager.PaperPluginManagerImpl.callEvent(PaperPluginManagerImpl.java:131)
[15:56:54 WARN]: at org.bukkit.plugin.Simp...
that stuff expects to be called within the context of the player. That event is a sync event, after all.
Expected behavior
Have almost no lock overhead that would trigger a possible lag source due to lock await across multiple regions.
I don't know exactly how this could even sort of work, but I feel like this a very bad limitation for the design of Folia currently. if not the most limiting issue. outside just making sure that code is optimized in here, I wonder what other sort of design could be approached.
Observed/Actual behavior
Ok, to start off, the current lock is acquired wh...
Here are some examples of the issue, which will lead to those regions that were locking the lock,to instead take 400ms lag spike each
I got a similar thread stack dump like this:
Profile link
Description of issue
Ok, I am getting this weird stacktrace here
PID: 194 | Suspended: false | Native: false | State: WAITING
Stack:
java.base@21.0.3/jdk.internal.misc.Unsafe.park(Native Method)
java.base@21.0.3/java.util.concurrent.locks.LockSupport.park(LockSupport.java:371)
ca.spottedleaf.moonrise.patches.chunk_system.scheduling.ChunkHolderManager.loadPoiChunk(ChunkHolderManager.java:1170)
net.minecraft.world.entity.ai.village.poi.PoiManager.getOr...
"dev 1.21.1" is not a valid version.
I'm trying to help improve the software. but Since some people seem to care too much about every detail of the pattern, i'll close this.
Why would you care about random users comments? If some "maintainer" closes an issue then OK, that's their business. Otherwise I don't see why you would want to close own issues if there were valid reasons to open them at the first place.
Why would you care about random users comments? If some "maintainer" closes an issue then OK, that's their business. Otherwise I don't see why you would want to close own issues if there were valid reasons to open them at the first place.
Because that's from the Paper team.
Why would you care about random users comments? If some "maintainer" closes an issue then OK, that's their business. Otherwise I don't see why you would want to close own issues if there were valid reasons to open them at the first place.
Because that's from the Paper team.
The person who has replied to you on both Discord & previous Github issues, is in fact, not part of the Paper team.
Expected behavior
The variable DISABLE_SNAPSHOT is kept as a static boolean on paper, but since paper is ticking single threaded, it doesn't really run into issues, every Folia call however, does not take this into consideration,
I think this should be changed to
ThreadLocal, otherwise this could possibly lead into issues?
public static final ThreadLocal DISABLE_SNAPSHOT = ThreadLocal.withInitial(() -> false);
this.snapshotDisabled = DISABLE_SNAPSHOT.get();
...
"dev 1.21.1" is not a valid version.
This is complete nonsense. How do you expect the dev version to reach the official version 1.21.1 without people reporting bugs and issues on dev? With magic that will find and fix bugs?
"dev 1.21.1" is not a valid version.
This is complete nonsense. How do you expect the dev version to reach the official version 1.21.1 without people reporting bugs and issues on dev? With magic that will find and fix bugs?
I think he mainly meant that we don't know the build of 1.21.1 used, is it on Folia or another fork?
The fork is using the latest commit 88e0907cadf2358e1d3e06222c0a97722ca4ce12 from the dev 1.21.1 branch. and there's no patches that are touching any of the chunk system therefore not causing this. Let's focus on what's causing or maybe a solution.
Expected behavior
The fact that the /vanish command will throw the player into invisibility
Observed/Actual behavior
an error occurs with the region flow
Steps/models to reproduce
sheduler
Plugin and Datapack List
- CoreProtect, GSit, LuckPerms, PlasmoVoice, SuperVanish, TAB
Folia version
This server is running Folia version 1.21.1-DEV-dev/1.21.1@88e0907 (2024-09-13T01:58:52Z) (Implementing API version 1.21.1-R0.1-SNAPSHOT)
You are running the latest version
###...
[ca.spottedleaf.moonrise.common.util.TickThread] Thread Region Scheduler Thread #0 failed main thread check: Accessing entity state off owning region's thread
java.lang.Throwable: null
at ca.spottedleaf.moonrise.common.util.TickThread.ensureTickThread(TickThread.java:58) ~[folia-1.21.1.jar:1.21.1-DEV-88e0907]
at org.bukkit.craftbukkit.entity.CraftSkeleton.getHandle(CraftSkeleton.java:43) ~[folia-1.21.1.jar:1.21.1-DEV-88e0907]
at org.bukkit.craftbukkit.entity.Craft...
This is not a folia issue, report this to the developers of the supervanish plugin
Also happening on old versions
Expected behavior
When lightning strikes, it should ignite both entities and the surrounding blocks, as per vanilla Minecraft behavior.
Observed/Actual behavior
In the version of Folia from this branch, lightning successfully ignites entities on hit, but it does not set blocks or the surrounding area on fire as expected.
Steps/models to reproduce
- Spawn a lightning bolt either naturally or using the /summon lightning_bol...
Update Folia to last commit of Paper
Hello ! That's my first contribution for Folia. I really want to help that project, I alaways wanted to manage a Minecraft server with multi threading.
I hope that I am doing right ! I have managed the conflicts with the upstream Paper repo
I don't think it will be accepted, see this comment: https://github.com/PaperMC/Folia/pull/220#issuecomment-2104933723
Up! It would be great if folia updated the paper more often
I am planning to use that branch on my production server next week ! I will keep you updated for issues I will find
Closing this, as Leaf has stated that we don't really accept PRs for updating Paper due to the nature of the complexity, and we'd rather not have people "bumping" PRs or Github issues.
I recommend use Paper over Folia until it gets stable to use.
Expected behavior
With Paper config unsupported-settings/allow-unsafe-end-portal-teleportation set to true, the gravity blocks duplication via end portal is supposed to work.
Observed/Actual behavior
No blocks are appearing on the end side when the machine is running, the overworld behavior seems correct.
Steps/models to reproduce
Build a duper (i used a 10gt for the tests but it should be the same with all types of dupers)
Run it and put a second account on the end side
###...
Stack trace
Caught previously unhandled exception :
java.io.IOError: java.io.IOException: Input/output error
at org.jline.keymap.BindingReader.readCharacter(BindingReader.java:170) ~[jline-reader-3.20.0.jar:?]
at org.jline.keymap.BindingReader.readBinding(BindingReader.java:110) ~[jline-reader-3.20.0.jar:?]
at org.jline.keymap.BindingReader.readBinding(BindingReader.java:61) ~[jline-reader-3.20.0.jar:?]
at org.jline.reader.impl.LineReaderImpl.doReadBinding(LineReaderImpl....
Your terminal was killed; generally, either a "you forgot to run it in a multiplexor over ssh", or, was killed by the host
Is there any update to this? On a community server we are playing this now happened Twice that Mulitple map art got corrupted by being replaced with other maps or are just gone by not showing anything at all. the first time it happened the server crashed but the second time it was a normal restart.
First suggestion would be to reproduce this on the 1.21.1 branch.
Lots of commits since 1.20.4 and we don't backport those updates.
i dont have proof rn but this has happened on 1.21.1, its been pretty frustrating
Haven't figured out what is happening yet, but this has happened on my 1.21.1 server. Don't know what to do about it.
Stuff like this is often due to things like map IDs being reused, used to be fairly common when the limit was muuuuch lower; wonder if there is a race or something here
same issues on folia 1.21.1 .
bump
this is correct, when I moved from master 9a19e4286c04bc01b9d7bc005cf66f8cc2dd49ff to dev/1.21.1 06d719d57d7ca9df0bda9f881d330b05aca99ebd the duping does not work.
Not sure which commit caused this problem
edit: logic looks fine, make sure that you've not overrun IDs
i very much doubt that anyone is hitting the map limit. since AFAIK the limit is 4,294,967,296 maps
Is your feature request related to a problem?
I know that Spark was disabled because it wasn't compatible with Folia. Since then, lucko has created a repo here: https://github.com/lucko/spark-extra-platforms which enables support for Folia. So I was wondering if it could be re-implemented?
Describe the solution you'd like.
I think we can use this repo, but I don't get the impression that it has been published, so that might still prevent re-implementing it.
Describe alternativ...
Expected behavior
Hello,
I noticed that we have a problem with the commands that allow us to select the world we want. For example, let's run this command:
_Euphyllia issued server command: /execute in minecraft:the_nether run kill @e[type=!minecraft:player]
I'm supposed to kill all entities in the nether world besides the player. But it does it on all worlds.
This happens through the console as well as through the execution of the command by a player.
Observed/Actual behavi...
Expected behavior
Wind Charges work as expected as per vanilla behavior
https://streamable.com/dtefua | Video of Vanilla Behavior
Observed/Actual behavior
Wind charges have inconsistent behavior, only working occassionally with seemingly no rhyme or reason.
https://streamable.com/qs3n2i | Video of Behavior
Steps/models to reproduce
- Obtain Wind Charge Item in Game
- Use Wind Charge
Plugin and Datapack List
 ~[folia-1.21.3.jar:1.21.3-DEV-c937c72]
at net.minecraft.server.level.ServerChunkCache.syncLoad(ServerChunkCache.java:100) ~[?:?]
at net.minecraft.server.level.ServerChunk...
I was unable to reproduce it on my server in 1.21.1 and 1.21.3.
Pure vanilla?
https://github.com/user-attachments/assets/519e4019-0edf-4e58-ad5c-0fb8f39a25db
https://github.com/user-attachments/assets/f32c543c-2aa0-4237-9bdb-fbfbdde4768d
Not flickering, but definitely made teleportAsync smooth. ๐
Please try on e50d09b
Assuming resolved
Any mechanic that relies on precise timing of these tickets cannot work on Folia. The ticket is added before the entity is teleported, but Folia adds some number of ticks of delay before the entity arrives to the target location. This doesn't take in account the fact that regions tick in parallel and as a result there is no synchronisation in the ticks between them. Folia additionally enforces that the chunks in the entity's bounding box is loaded, which may add yet more delay after the tic...
Does this still happen on 1.21.3?
Moved to https://github.com/Tuinity/Moonrise/issues/81, as this affects Moonrise and Paper
Teleporting during a quit event will cause a crash, it cannot be done
All usages null check the input, so I can't see how this is happening. Please re-open if it happens on 1.21.3
eb5ec0b Adjust shutdown logic to close player inventories - Spottedleaf
Please run paper debug chunks --async from the console (note: --async only works from the console), and provide the console output and the chunk debug. Please also note that the chunk debug will show all online players locations and all loaded chunks.
Will try that soon !
Stack trace
[04:57:02 ERROR]: Error sending packet clientbound/minecraft:recipe_book_add (skippable? false)
io.netty.handler.codec.EncoderException: Failed to encode packet 'clientbound/minecraft:recipe_book_add'
at net.minecraft.network.codec.IdDispatchCodec.encode(IdDispatchCodec.java:53) ~[folia-1.21.4.jar:1.21.4-DEV-d4bc338]
at net.minecraft.network.codec.IdDispatchCodec.encode(IdDispatchCodec.java:20) ~[folia-1.21.4.jar:1.21.4-DEV-d4bc338]
at net.minecraf...
- Updates checkout and setup-java to v4,
- Replaces the old Gradle job (
gradle/gradle-build-action) with new variant. (gradle/actions/setup-gradle)
wrapper-validation action is no longer needed, setup-gradle v4 validates the wrapper
wrapper-validation action is no longer needed, setup-gradle v4 validates the wrapper
will remove.
Skipped this cuz origin repo did not reflect the v4 update in readme.
Stack trace
A fatal error has been detected by the Java Runtime Environment:
EXCEPTION_ACCESS_VIOLATION (0xc0000005) at pc=0x00007fffd4e34e9a, pid=498576, tid=312608
JRE version: OpenJDK Runtime Environment Zulu21.32+17-CA (21.0.2+13) (build 21.0.2+13-LTS)
Java VM: OpenJDK 64-Bit Server VM Zulu21.32+17-CA (21.0.2+13-LTS, mixed mode, sharing, tiered, compressed oops, compressed class ptrs, g1 gc, windows-amd64)
Problematic frame:
V [jvm.dll+0x3e4e9a]
Core dump will...
- You seem to be running something that isn't Folia.
- Your Java version is outdated. Use 21.0.5.
- Folia doesn't have any effect on how native methods are resolved or executed.
- You seem to be running something that isn't Folia.
- Your Java version is outdated. Use 21.0.5.
- Folia doesn't have any effect on how native methods are resolved or executed.
I'm sure that I'm using Folia. It isn't relevant to what version of Java but server core, cuz I used such JNIC obfuscated plugins on Spigot/Bukkit without any errors.
Only by using Folia can I reproduce the error, one more thing, I can run the same plugin without native obfuscation on Folia correctly.
Your dump indicates you are using a modified version of folia, which we do not support. Earlier builds of jdk 21 had a bug that would cause jvm crashes when some tooling tries to access jvm internal data.
Liminol != Folia. It's a fork, ask them for support. Who knows what changes they've made.
Windows is a pretty rare deployment OS. Use Linux if you can, it's way more popular. Popular = more chance someone had the same issue as you in the past.
If you have to stay on Windows, try different Java runtimes. Maybe Oracle Java will work? Maybe GraalVM will work?
Expected behavior
[14:11:12 ERROR]: Command exception: /time set 9
java.lang.IllegalStateException: World mismatch: expected world_nether but got world
at net.minecraft.world.level.Level.getCurrentWorldData(Level.java:841) ~[folia-1.21.4.jar:1.21.4-DEV-d4bc338]
at net.minecraft.world.level.Level.getGameTime(Level.java:1844) ~[folia-1.21.4.jar:1.21.4-DEV-d4bc338]
at net.minecraft.server.MinecraftServer.synchronizeTime(MinecraftServer.java:2033) ~[folia-1.21.4.j...
- You seem to be running something that isn't Folia.
- Your Java version is outdated. Use 21.0.5.
- Folia doesn't have any effect on how native methods are resolved or executed.
I'm sure that I'm using Folia. It isn't relevant to what version of Java but server core, cuz I used such JNIC obfuscated plugins on Spigot/Bukkit without any errors.
Only by using Folia can I reproduce the error, one more thing, I can run the same plugin without native obfuscation on Folia corre...
Liminol != Folia. It's a fork, ask them for support. Who knows what changes they've made.
Windows is a pretty rare deployment OS. Use Linux if you can, it's way more popular. Popular = more chance someone had the same issue as you in the past.
If you have to stay on Windows, try different Java runtimes. Maybe Oracle Java will work? Maybe GraalVM will work?
- You seem to be running something that isn't Folia.
- Your Java version is outdated. Use 21.0.5.
- Foli...
Folia makes 0 changes that would break native library loading; some native library called into the JVM internals which blew up, but there is no information on what it did to trigger that, and folia doesn't make any changes relating to natives afaik, nor is that version of folia supported, so, even if there was a breakage, it would not be fixed
Liminol != Folia. It's a fork, ask them for support. Who knows what changes they've made.
Windows is a pretty rare deployment OS. Use Linux if you can, it's way more popular. Popular = more chance someone had the same issue as you in the past.
If you have to stay on Windows, try different Java runtimes. Maybe Oracle Java will work? Maybe GraalVM will work?
- You seem to be running something that isn't Folia.
- Your Java version is outdated. Use 21.0....
Folia makes 0 changes that would break native library loading; some native library called into the JVM internals which blew up, but there is no information on what it did to trigger that, and folia doesn't make any changes relating to natives afaik, nor is that version of folia supported, so, even if there was a breakage, it would not be fixed
All right, but the problem only exisiting on Folia, no errors of native method loading were found on Spigot/Bukkit. We have to find another core ins...
We have tested java 17-21, which is Oracle version, and they still failed.
As I said, try Linux. Or at least WSL. There is not enough information, it can be pretty much everything. ๐คท
Hell, it could even be that RAM is damaged in one specific place that JVM likes or something. I had this happen to my own machine (system almost working, but just crashing on one specific program... just because of how the memory layout happened).
So memtest86 could also be a good idea.
Emmm, Could we talk on QQ?It's unconvenient to discuss a issue that doesn't caused by folia here()
3058877423
We have tested java 17-21, which is Oracle version, and they still failed.
As I said, try Linux. Or at least WSL. There is not enough information, it can be pretty much everything. ๐คท Hell, it could even be that RAM is damaged in one specific place that JVM likes or something. I had this happen to my own machine (system almost working, but just crashing on one specific program... just because of how the memory layout happened).
So memtest86 could also be a good idea.
I made a su...
This happens in Vanilla and Paper.
This looks rather intentional, as the entity selector parser will only limit it to the current world context under a few conditions. Your best bet is to report it to Mojang.
I can't produce this locally either on 1.21.4.
94d65c3 Update old position/rotation when teleporting e... - Spottedleaf
This may be a fix on 1.21.4: https://github.com/PaperMC/Folia/commit/94d65c398cf3015bafcaf46a09b2ea42cec66d24
If it still happens, please provide the world data
I really don't know what that code is doing. You are using some library and haven't said what it is.
I've run /recipe give searchndstroy * on https://github.com/PaperMC/Folia/commit/94d65c398cf3015bafcaf46a09b2ea42cec66d24 locally and it works.
I am going to need code that I can run to look at this further.
get vanilla recipes is just a getter
public ArrayList<NamespacedKey> getVanillaRecipes() { return this.vanilla_recipes; }
this is the only function that adds to vanilla_recipes
private void loadVanillaRecipe(String recipe_name, ItemStack result, HashMap<Character, RecipeItem> dictionary) {
NamespacedKey recipe_key = new NamespacedKey(plugin, recipe_name);
// TODO really fucking sloppy, only works with one ingredient shapeless
if (config.contai...
Folia does not allow entities to be ticked after they teleport, which fixes the sand duping. It has to be this way as a removed entity is considered off-region.
This works for me locally on 1.21.4... I suspect this was due to the early Paper build that Folia 1.21.1 was based on for far too long.
1c0ecd3 Use ThreadLocal for CraftBlockStates#DISABLE_SN... - Spottedleaf
Yeah there is some potential gain to be made by using something like ReentrantAreaLock here, however using that would greatly complicate the implementation. It's sort of on my TODO list, but not very high.
You can alleviate some of this overhead by using a larger grid exponent (like 4 or 5).
The thread checks in Folia are pretty aggressive and are enabled by default. If you think there are any that are missed, you can open an issue and I can look at it.
There's not really anything I can do about plugin development beyond that.
This works for me on 1.21.4. If this still happens, please provide a video so I can look at reproducing the issue.
[PaperMC/Folia] New comment on issue #266: glow item frame invisible after 1.21 folia restart server
I cannot produce this on 1.21.4.
This also happens on teleport commands and such. On vanilla it doesn't happen.
Will close -- appears to be a setup issue on our end which wasn't confirmed properly. Thank you for the assistance<3
is it still happening on latest? or can you close the issue
Can you try on the latest version and let us know if it still happens?
This is probably fixed on the latest versions already. Because I have not seen this on 1.21.1 already.
I assume this is fixed on the latest. With the ref count changes? @Spottedleaf
Expected behavior
I am trying to create a executable server version for folia and I have an error when I try to execute the gradle task createReobfPaperclipJar
Observed/Actual behavior
> Task :folia-server:compileLog4jPluginsJava
Note: Processing Log4j annotations
Note: Annotations processed
Note: Processing Log4j annotations
Note: No elements to process
Note: /Users/funix/Documents/tmp/minecraft/folia-git/Folia-Server/src/log4jPlugins/java/io/papermc/paper/logging/DelegateLogE...
update your JDK, there is a bug in earlier releases of 21
Do not use createReobf... for 1.20.5+, use createMojmap...
When I was using getForceLoadedChunks() it throw:
[21:44:53 ERROR]: [ca.spottedleaf.moonrise.common.util.TickThread] Thread Region Scheduler Thread #2 failed main thread check: Async chunk retrieval
java.lang.Throwable: null
at ca.spottedleaf.moonrise.common.util.TickThread.ensureTickThread(TickThread.java:51) ~[deerfolia-1.21.4.jar:1.21.4-DEV-2bd64a2]
at org.bukkit.craftbukkit.CraftWorld.getChunkAt(CraftWorld.java:373) ~[deerfolia-1.21.4.jar:1.21.4-DEV-2bd64a2]
...
@Override
public Chunk getChunkAt(int x, int z) {
net.minecraft.world.level.chunk.LevelChunk chunk;
try {
Objects.requireNonNull(((ServerLevel) world).regioniser.getRegionAtUnsynchronised(x >> 4, z >> 4)).regioniser.acquireReadLock(); // Folia - region threading
warnUnsafeChunk("getting a faraway chunk", x, z); // Paper
chunk = (net.minecraft.world.level.chunk.LevelChunk) this.world.getChunk(x, z, ChunkStatus.FULL, true);
...
For multiple versions CraftChunk is just a wrapper around coordinates (like CraftBlock) without any real information or reference to the chunk. Just create craft chunks in getForceLoadedChunks() with the coordinates without loading them (and they're most probably already loaded anyway), there is a SeverLevel,int,int constructor that you can use instead of the one that takes LevelChunk
This may be a fix on 1.21.4: 94d65c3
If it still happens, please provide the world data
Still happening
Still happening
As asked, please provide a download of the world data so we can test/verify.
Still happening
As asked, please provide a download of the world data so we can test/verify.
I did have this problem i tested with fresh folia server no problem even when apply my server settings to it no problem.
The cause that i have kinda found to make that happen 100% all the time is that have the
Gamerule spawnChunkRadius 0
When set it to 0 and you fly away unloading the chunk with piston it cause ghost block
When set the spawnchunk back to default 2 (i think) this problem havent happen again.
Expected behavior
TNT explosion event is the same as vanilla
Observed/Actual behavior
In the latest version 1.21.4 of folia and paper, TNT explodes earlier than in vanilla
Steps/models to reproduce
Ignite the TNT, it will be a few tenths of a second earlier than vanilla.
Plugin and Datapack List
[22:51:34 INFO]: - HuskHomes, HuskSync,
[22:51:34 INFO]: - BetterRTP Chunky, CMILib
Essentials, EssentialsSpawn, GSit, LuckPerms, PAPIProxyBridge, PlaceholderAPI
PlayerChat,...
I did have this problem i tested with fresh folia server no problem even when apply my server settings to it no problem.
The cause that i have kinda found to make that happen 100% all the time is that have the
Gamerule spawnChunkRadius 0When set it to 0 and you fly away unloading the chunk with piston and the fly back to it and console spam with that error and it cause ghost block
When set the spawnchunk back to default 2 (i think) this problem havent happen again.
...
[PaperMC/Folia] New branch created: ver/1.20.6
[PaperMC/Folia] New branch created: master1_20_6
[PaperMC/Folia] branch deleted: master
[PaperMC/Folia] New branch created: master
[PaperMC/Folia] branch deleted: master1_20_6
Sorry about that - can you reopen a PR targeting the "new" master branch?
Sorry about that - can you reopen a PR targeting the "new" master branch, if it's still relevant?
Sorry about that - can you reopen a PR targeting the "new" master branch, if it's still relevant?
Expected behavior
Piston normal operation
Observed/Actual behavior
The console will generate the following error:
[13:28:48 ERROR]: [ca.spottedleaf.moonrise.common.util.TickThread] Thread Region Scheduler Thread #1 failed main thread check: Cannot asynchronously load chunks
java.lang.Throwable: null
at ca.spottedleaf.moonrise.common.util.TickThread.ensureTickThread(TickThread.java:51) ~[luminol-1.21.4.jar:1.21.4-DEV-27a2e2a]
at net.minecraft.server.level.ServerChunk...
Expected behavior
no errors
Observed/Actual behavior
[13:49:51 ERROR]: [ca.spottedleaf.moonrise.common.util.TickThread] Thread Region Scheduler Thread #1 failed main thread check: Cannot asynchronously load chunks
java.lang.Throwable: null
at ca.spottedleaf.moonrise.common.util.TickThread.ensureTickThread(TickThread.java:51) ~[Folia-1.21.4.jar:1.21.4-DEV-1c0ecd3]
at net.minecraft.server.level.ServerChunkCache.syncLoad(ServerChunkCache.java:100) ~[Folia-1.21.4.ja...
After testing, it is reproduced in the latest version of folia
------------------ Original ------------------
From: Euphyllia @.>
Date: Mon,Dec 16,2024 5:05 PM
To: PaperMC/Folia @.>
Cc: YIIYI @.>, Author @.>
Subject: Re: [PaperMC/Folia] Folia1.21.4 When the piston pushes the block infront and there is a lot of creatures in the block in front, will producemoving pistons that cannot be cleared normally (Issue #311)
You ...
This is suspected to be a Moonrise problem. I can reproduce it in single-player 1.21.4 Fabric + Moonrise mod.
Build failed
Perhaps it's because it's not compatible with 1.21.4?
Sorry about that - can you reopen a PR targeting the "new" master branch?
will be reopened under new branch.
This except 1 and 2 are already in master, 4th change isn't and Ubuntu will be frozen to 22.04 for stability reasons.
from some testing on 1.21.4 it seems 1 way to corrupt a locked map is to teleport before interacting with it in your inventory.
We tested by creating a locked map, not touching it after its creation and then teleporting, which corrupts it. If you click the map before teleporting the map remains ok.
This was tested on 1.21.1 and 1.21.4
Build failed Perhaps it's because it's not compatible with 1.21.4?
I fixed it on my side, but I can't compile it strangely, it says there is an error even without this PR
Build failed Perhaps it's because it's not compatible with 1.21.4?
I fixed it on my side, but I can't compile it strangely, it says there is an error even without this PR (on Windows)
I think there's no problem with mine
can't reproduce without plugins
After I uninstalled all the plugins, I can still reproduce this bug
After testing, it is reproduced in the latest version of folia
โฆ
------------------ Original ------------------ From: Euphyllia @.> Date: Mon,Dec 16,2024 5:05 PM To: PaperMC/Folia @.> Cc: YIIYI @.>, Author @.> Subject: Re: [PaperMC/Folia] Folia1.21.4 When the piston pushes the block infront and there is a lot of creatures in the block in front, will producemoving pistons that cannot be cleared normally (Issue [#311](https://github.com...
After testing, it is reproduced in the latest version of folia
โฆ
------------------ Original ------------------ From: Euphyllia @.> Date: Mon,Dec 16,2024 5:05 PM To: PaperMC/Folia _@**._> Cc: YIIYI _@.>, Author @.**_> Subject: Re: [PaperMC/Folia] Folia1.21.4 When the piston pushes the block infront and there is a lot of creatures in the block in front, will producemoving pistons that cannot be cleared normally (Issue [#311](https://github.com/Paper...
Expected behavior
In spectator mode, the hotbar transforms to a menu, which lets you navigate between players on the server, and teleport to them.
In the tested use-case: Spectator is in Nether, Player "To Be Spectated" is in overworld.
When using this menu, you just expect to get teleported to that player, including a smooth dimension switch.
Observed/Actual behavior
When trying to use the spectator mode menu, to teleport to a player that is in another dimension, the server cras...
Maybe is because of Paper Teleport API is broken on Folia... I don't know.
Current broken API
Most API that interacts with portals / respawning players / some player login API is broken.
ALL scoreboard API is considered broken (this is global state that I've not figured out how to properly implement yet)
World loading/unloading
Entity#teleport. This will NEVER UNDER ANY CIRCUMSTANCE come back, use teleportAsync
Could be more
Maybe Help You
Teleporting should still work, only the teleport() method and the teleport event are removed/broken
https://github.com/user-attachments/assets/80a209cf-089e-4932-b56c-b4efc09e981f
Might be a issue of the chunk system.The entity's chunk status turned TRACKED when the machine is out of the player's simulation distance like that video and entity won't be ticked once before the moving piston block push it and the oldposition is still 0,0,0.
 ~[folia-1.21.4.jar:1.21.4-DEV-8af1aef]
at net.minecraft.world.entity.projectile.Projectile.getOwnerRaw(Projectile.java:99) ~[foli...
An ender pearl was thrown through a nether portal. That's not nothing =p
Stack trace
[01:25:14 ERROR]: [ca.spottedleaf.moonrise.common.util.TickThread] Thread Region Scheduler Thread #4 failed main thread check: Accessing poi chunk off-main
java.lang.Throwable: null
at ca.spottedleaf.moonrise.common.util.TickThread.ensureTickThread(TickThread.java:51) ~[folia-1.21.4.jar:1.21.4-DEV-8af1aef]
at net.minecraft.world.entity.ai.village.poi.PoiManager.getOrLoad(PoiManager.java:89) ~[folia-1.21.4.jar:1.21.4-DEV-8af1aef]
at net.minecraft.wo...
Expected behavior
Items from broken trees, as well as mobs and other entities, should be properly tracked and send to the Player's client so they can be displayed.
Observed/Actual behavior
Occasionally, regions will stop sending these entities to players, resulting in a confusing desync where you can pick up items but never see them on the game / hear them get picked up.
Steps/models to reproduce
Unsure, though this is occurring at high player counts (we've been running at 130 - 400 players)
Plugin and Datapack List
[19:21:46 INFO]: Server Plugins (28):
[19:21:46 INFO]: Paper Plugins:
[19:21:46 INFO]: - CarbonChat, FancyHolograms
[19:21:46 INFO]: Bukkit Plugins:
[19:21:46 INFO]: - AxInventoryRestore, BlueSlimeCore, ChangeSlots, Chunky, ChunkyBorder, CMCLinkManager, CombatLogX, CoreProtect, DatapackLoader, Essentials
[19:21:46 INFO]: FreedomChat, GrimAC, helper, helper-sql, LiteBans, LuckPerms, NBTAPI, packetevents, PlaceholderAPI, PlasmoVoice
[19:21:46 ...
Just commenting here to say I've had this happen a couple times as well with player counts of ~60 up to 130 on a not too modified SMP, I've also been able to teleport into these regions and no players were visible or any entity (item frame, mobs, leads)
This server is running Folia version 1.21.4-DEV-master@8af1aef (2025-01-01T16:22:18Z) (Implementing API version 1.21.4-R0.1-SNAPSHOT)
You are running the latest version
Previous version: 1.21.4-13-4803421 (MC: 1.21.4)
[03:31:54 INFO]: Server Plugins (21):
[03:31:54 INFO]: Paper Plugins:
[03:31:54 INFO]: - FairyChat
[03:31:54 INFO]: Bukkit Plugins:
[03:31:54 INFO]: - AntiHealthIndicator, Chunky, ChunkyBorder, CoreProtect, DiscordSRV, Essentials, FreedomChat, GSit, JustLootIt, LuckPerms
[03:31:54 INFO]: OreAnnouncer, packetevents, spark, squaremap, Stateful, SuperHarvest, TAB, Terra, ToolStats, voicechat
I'm a bit suspicious that Chunky may be doing something strange that's upsetting Folia somehow with how it's generating/saving these chunks on occasion. @ItsSyfe Was your world pregenerated on 1.21.4?
During my testing on 1.21.3, I had not pre generated any of the land and I don't believe anyone encountered any issues similar to this. The player counts were much lower so the regioning wasn't super active, but nevertheless I don't recall the issue occurring. It's also possible this is some sort of 1.21.4 problem in particular, but I'm unsure.
Yeah we pregenned with chunky but this is happening in chunks that have had a lot of player presence (it happened at a player's base for me), and unloading then reloading the chunk also seems to fix it for me as well.
Though like your case I haven't been able to reproduce it reliably as it just happens at seemingly random, I haven't been checking the chunk x y though.
After thinking about it, I've had similar circumstances as well (player bases randomly becoming glitched out) so I doubt it's that. Could you post your paper configs in here? I'd like to cross reference with mine and see if there's any similarities. Censor anything sensitive of course
@Spottedleaf do you see any solutions for this that apply for the current design outside of this "hacky" way?
I will need some kind of setup to verify this behavior.
If you can make one, please open an issue here
POI data doesn't run through main, so I would hesitate to say it is fixed.
I'm not sure what caused this. There's unfortunately a significant lack of profiling data available from the chunk system. Guess I can put that on my wishlist for that.
Expected behavior
Ender pearl to unload on player logout then reload on player login
Observed/Actual behavior
Ender pearl stays existent (loaded) even after the player is no longer on the server
Steps/models to reproduce
Setup a stasis chamber and then have an alt logged in to check on the pearl and simply logout on the account that threw the pearl.
Plugin and Datapack List
[00:05:10 INFO]: Server Plugins (22):
[00:05:10 INFO]: Paper Plugins:
[00:05:10 INFO]: - FairyChat
[00:05:10 INFO]: Bukkit Plugins:
[00:05:10 INFO]: - AntiHealthIndicator, ArmorPoser, BlueMap, Chunky, ChunkyBorder, CoreProtect, DiscordSRV, Essentials, FreedomChat, GSit
[00:05:10 INFO]: JustLootIt, LuckPerms, OreAnnouncer, packetevents, spark, Stateful, SuperHarvest, TAB, Terra, ToolStats
[00:05:10 INFO]: voicechat
Folia version
[00:26:50 INFO]: Checking version, please wait...
[00:26:50 INFO]: This server is running Folia version 1.21.4-DEV-dev/hard-fork@4e7607d (1970-01-01T00:00:...
Stack trace
[02:18:01 ERROR]: [ca.spottedleaf.moonrise.common.util.TickThread] Thread Region Scheduler Thread #7 failed main thread check: Accessing poi chunk off-main, world=season4_world, chunk_pos=[12, -333]
java.lang.Throwable: null
at ca.spottedleaf.moonrise.common.util.TickThread.ensureTickThread(TickThread.java:66) ~[folia-1.21.4.jar:1.21.4-DEV-4e7607d]
at net.minecraft.world.entity.ai.village.poi.PoiManager.getOrLoad(PoiManager.java:89) ~[folia-1.21.4.jar:1.21.4-DEV-4e7607d]
at net.minecraft.world.entity.ai.village.poi.PoiManager.exists(PoiManager.java:338) ~[folia-1.21.4.jar:1.21.4-DEV-4e7607d]
at net.minecraft.world.entity.ai.behavior.GoToPotentialJobSite.lambda$stop$2(GoToPotentialJobSite.java:50) ~[folia-1.21.4.jar:1.21.4-DEV-4e7607d]
at java.base/java.util.Optional.ifPresent(Optional.java:178) ~[?:?]
at net.minecraft.world.entity.ai.behavior.GoToPotentialJobSite.stop(GoToPotentialJobSite.java:45) ~[folia-1.21.4.ja...
Folia 1.21.4 has removed that feature,players now won't store the end perals that they threw
Expected behavior
Zombie Piglins will only fall from the Hell Gate into the Death Passage and be killed normally, instead of some monsters jumping out of nowhere.
Observed/Actual behavior
Zombified Piglins will spawn in places around the farm, such as in water, outside the execution platform
Steps/models to reproduce
- Build a regular Zombie Piglin Farm and start it
- Wait 5-10 minutes
Plugin and Datapack List
[22:51:34 INFO]: - HuskHomes, HuskSync,
[22:51:34 INFO]: - BetterRTP Chunky, CMILib
Essentials, EssentialsSpawn, GSit, LuckPerms, PAPIProxyBridge, PlaceholderAPI
PlayerChat, PlayerTitle, ProtocolLib, QuickShop-Hikari, Residence, Vault, WorldEdit
Folia version
Luminol version 1.21.4-DEV-dev/1.21.4-hardfork@efd5589
Other
No response
This kernel is not monitored by the PaperMC Team
this bug can be reproduced under the latest folia๏ผ4e7607d๏ผ
------------------ Original ------------------
From: De1fy @.>
Date: Sat,Jan 25,2025 8:03 PM
To: PaperMC/Folia @.>
Cc: YIIYI @.>, Author @.>
Subject: Re: [PaperMC/Folia] In a Zombified Piglin Farm, Zombified Piglinsspawn in places they shouldn't. (Issue #320)
This kernel is not monitored by the PaperMC Team
โ
Reply to this email directly, view it on GitHub, or unsubscribe.
You are receiving this because you authored the thread.Message ID: @.***>
also notice that 2d goldfarm there sometime piglin walkign aroudn the area outside the killchamber
But it could just be a vanilla feature
https://minecraft.wiki/w/Zombified_Piglin#Reinforcements
This feature is exclusive to Java Edition.
On hard difficulty, zombified piglins have a 5% chance to spawn as leaders. Leader zombified piglins can spawn other zombified piglins as reinforcements.
Folia 1.21.4 has removed that feature,players now won't store the end perals that they threw and server won't unload them because the server won't trace ender perals
This is correct. Cross region entity handling is not something that's going to work.
Stack trace
https://mclo.gs/OpybgLh
Plugin and Datapack List
0 plugins
Actions to reproduce (if known)
This is happening in the infernal dimension
Folia version
This server is running Folia version 1.21.4-DEV-dev/hard-fork@92efc63 (1970-01-01T00:00:00Z) (Implementing API version 1.21.4-R0.1-SNAPSHOT)
You are running the latest version
Previous version: 1.21.4-DEV-8af1aef (MC: 1.21.4)
Other
No response
Expected behavior
Everything is OK
Observed/Actual behavior
The server crashes when certain players join
And issue the following error
[io.papermc.paper.threadedregions.TickRegionScheduler] Region #32 centered at chunk [437, -61] in world 'world' failed to tick:
java.lang.NullPointerException: Cannot read field "entity" because "trackedEntity1" is null
at net.minecraft.server.level.ChunkMap.addEntity(ChunkMap.java:909) ~[luminol-1.21.4.jar:1.21.4-DEV-7604000]
at net.minecraft.server.level.ServerChunkCache.addEntity(ServerChunkCache.java:747) ~[luminol-1.21.4.jar:1.21.4-DEV-7604000]
at net.minecraft.server.level.ServerLevel$EntityCallbacks.onTrackingStart(ServerLevel.java:2802) ~[luminol-1.21.4.jar:1.21.4-DEV-7604000]
at net.minecraft.server.level.ServerLevel$EntityCallbacks.onTrackingStart(ServerLevel.java:2745) ~[luminol-1.21.4.jar:1.21.4-DEV-7604000]
at ca.spottedleaf.moonrise.patches.chunk_system.level.entity.EntityLookup.entityStatusChange(EntityL...
We don't accept crash reports from software that isn't ours
also notice that 2d goldfarm there sometime piglin walkign aroudn the area outside the killchamber
But it could just be a vanilla feature
https://minecraft.wiki/w/Zombified_Piglin#Reinforcements
This feature is exclusive to Java Edition. On hard difficulty, zombified piglins have a 5% chance to spawn as leaders. Leader zombified piglins can spawn other zombified piglins as reinforcements.
After observation, not only zombie piglin farms, but other monster farms may also appear. It will not appear in vanilla
Stack trace
paste your stack trace or a paste.gg link here!
Plugin and Datapack List
pl
[19:41:07 INFO]: Server Plugins (54):
[19:41:07 INFO]: Paper Plugins:
[19:41:07 INFO]: - CarbonChat, EnchantedBook, FancyNpcs, MoreFoWorld
[19:41:07 INFO]: Bukkit Plugins:
[19:41:07 INFO]: - ArmorStandEditor, AutoWhitelistRemove, BreweryX, CoreProtect, CoreProtectTNT, Craftorithm, DeluxeMenus, emotecraft, Essentials, EssentialsSpawn
[19:41:07 INFO]: GSit, HeadDatabase, InventoryRollbackPlus, InvisibleFrames, JoinMessenger, LevelledMobs, LiteBans, LuckPerms, Maintenance, Mclogs
[19:41:07 INFO]: MilkyPixelart, MinerTrack, NBTAPI, OreAnnouncer, PlaceholderAPI, PlasmoVoice, PlayerHeads, ProtocolLib, pv-addon-discs, pv-addon-groups
[19:41:07 INFO]: pv-addon-lavaplayer-lib, pv-addon-sculk, pv-addon-spectator, RestoreChunk, rpchat, SkinsRestorer, spark, squaremap, squaremap-worldguard, squaremarker
[19:41:07 INFO]: TAB, Terra, TWSAdmin, TWSCheck, TWSC...
Expected behavior
The items should flow normally without lag or rollback
Observed/Actual behavior
Items sometimes flow normally, but sometimes start to have strange lags and rollbacks.
Steps/models to reproduce
Build an item conveyor using packed ice and water, throw items on it, and wait until they start to have strange behavior, like small rollbacks. It takes around 20 seconds for it to appear, and it disappears periodically.
Plugin and Datapack List
None
Folia version
This server is running Folia version 1.21.4-DEV-dev/hard-fork@92efc63
Other
I made a small video about this
But issues what you link os outdated on 1.21.4 not a hard-fork
I have tested on non hard fork branch but it still happens()
But issues what you link os outdated on 1.21.4 not a hard-fork
I have tested on non hard fork branch but it still happens()
okey
But issues what you link os outdated on 1.21.4 not a hard-fork
I have tested on non hard fork branch but it still happens()
And I pointed it to the wrong issue.Sorry for that,I am currently thinking it is related to https://github.com/PaperMC/Folia/issues/311 after I compared the code between 1.21.4 and 1.21.1 of folia()
But issues what you link os outdated on 1.21.4 not a hard-fork
I have tested on non hard fork branch but it still happens()
And I pointed it to the wrong issue.Sorry for that,I am currently thinking it is related to #311 after I compared the code between 1.21.4 and 1.21.1 of folia()
The old position of entity was not updated correctly and the method "applyEffectsFromBlocks" depends on the old position of that entity
But issues what you link os outdated on 1.21.4 not a hard-fork
I have tested on non hard fork branch but it still happens()
And I pointed it to the wrong issue.Sorry for that,I am currently thinking it is related to #311 after I compared the code between 1.21.4 and 1.21.1 of folia()
The old position of entity was not updated correctly and the method "applyEffectsFromBlocks" depends on the old position of that entity
So problem in 1.21.4 is taking place?
But issues what you link os outdated on 1.21.4 not a hard-fork
I have tested on non hard fork branch but it still happens()
And I pointed it to the wrong issue.Sorry for that,I am currently thinking it is related to #311 after I compared the code between 1.21.4 and 1.21.1 of folia()
The old position of entity was not updated correctly and the method "applyEffectsFromBlocks" depends on the old position of that entity
So problem in 1.21.4 is taking place?
Yeah and 1.21.3 is also taking place()
at CrazyCrates-4.3.4.jar/com.badbones69.crazycrates.api.ChestManager.closeChest(ChestManager.java:37) ~[CrazyCrates-4.3.4.jar:?]
at CrazyCrates-4.3.4.jar/com.badbones69.crazycrates.tasks.crates.CrateManager.endQuickCrate(CrateManager.java:1449) ~[CrazyCrates-4.3.4.jar:?]
at CrazyCrates-4.3.4.jar/com.badbones69.crazycrates.listeners.MiscListener.onPlayerQuit(MiscListener.java:126) ~[CrazyCrates-4.3.4.jar:?]
55-57 lines string says what is this cause of plugin.
Does this happen on Paper?
Can still reproduce
Using spectator mode and fly more than 1024(make the final position in other tickregion) and switch to creative mode immediately then it will throw that error
Can still reproduce
Using spectator mode and fly more than 1024(make the final position in other tickregion) and switch to creative mode immediately then it will throw that error
how do you switch to creative mode? from by plugin or operators?
Can still reproduce
Using spectator mode and fly more than 1024(make the final position in other tickregion) and switch to creative mode immediately then it will throw that error
how do you switch to creative mode? from by plugin or operators?
/gamemode creative ()
If this isn't a vanilla feature, then it's not a Folia issue specifically and it's something upstream in Paper as we're also seeing this on Paper version 1.21.4-117-main@a35cfe9.
Will get our server upgraded to the latest hard fork build to see if it's still occurring & try it in vanilla just for good measure before reporting it to the Paper team.
Stack trace
paste your stack trace or a paste.gg link here!
https://mclo.gs/reuyr5i
https://mclo.gs/vAky7kf
Plugin and Datapack List
[21:17:01 INFO]: Server Plugins (54):
[21:17:01 INFO]: Paper Plugins:
[21:17:01 INFO]: - CarbonChat, emotecraft, EnchantedBook, FancyNpcs, MoreFoWorld
[21:17:01 INFO]: Bukkit Plugins:
[21:17:01 INFO]: - ArmorStandEditor, AutoWhitelistRemove, BreweryX, CoreProtect, CoreProtectTNT, Craftorithm, DeluxeMenus, Essentials, EssentialsSpawn, GSit
[21:17:01 INFO]: HeadDatabase, InventoryRollbackPlus, InvisibleFrames, JoinMessenger, LevelledMobs, LiteBans, LuckPerms, Maintenance, Mclogs, MilkyPixelart
[21:17:01 INFO]: MinerTrack, NBTAPI, OreAnnouncer, PlaceholderAPI, PlasmoVoice, PlayerHeads, ProtocolLib, pv-addon-discs, pv-addon-groups, pv-addon-lavaplayer-lib
[21:17:01 INFO]: pv-addon-sculk, pv-addon-spectator, RestoreChunk, rpchat, SkinsRestorer, spark, squaremap, squaremap-worldguard, squaremarker, TAB
[21:17:01 INFO]: Terra, TWSAdmi...
Can still reproduce
Using spectator mode and fly more than 1024(make the final position in other tickregion) and switch to creative mode immediately then it will throw that error
how do you switch to creative mode? from by plugin or operators?
/gamemode creative ()
Can I have an updated log?
[09:46:55] [Region Scheduler Thread #4/ERROR]: [ca.spottedleaf.moonrise.common.util.TickThread] Thread failed main thread check: Cannot asynchronously load chunks, context=[thread=Region Scheduler Thread #4,class=io.papermc.paper.threadedregions.TickRegionScheduler$TickThreadRunner,region={center=[4800, -1392],world=world}], world=world, chunk_pos=[4859, -1394]
java.lang.Throwable: null
at ca.spottedleaf.moonrise.common.util.TickThread.ensureTickThread(TickThread.java:88) ~[folia-1.21.4.jar:1.21.4-DEV-873afea]
at net.minecraft.server.level.ServerChunkCache.syncLoad(ServerChunkCache.java:99) ~[folia-1.21.4.jar:1.21.4-DEV-873afea]
at net.minecraft.server.level.ServerChunkCache.getChunkFallback(ServerChunkCache.java:152) ~[folia-1.21.4.jar:1.21.4-DEV-873afea]
at net.minecraft.server.level.ServerChunkCache.getChunk(ServerChunkCache.java:320) ~[folia-1.21.4.jar:1.21.4-DEV-873afea]
at net.minecraft.world.level.Level.getChunk(Level.java:1107) ~[folia-1.21.4.jar:1.21.4-DEV-873afea]
at...
Can still reproduce
Using spectator mode and fly more than 1024(make the final position in other tickregion) and switch to creative mode immediately then it will throw that error
how do you switch to creative mode? from by plugin or operators?
/gamemode creative ()
Can I have an updated log?
Might related with a fix of mojang: https://bugs.mojang.com/browse/MC-92875 And updated logs are above()
I freeze to 22.04 in case stability (or other) problems happen, this change is present on velocity repo.
Folia works fine on 24.04. Zero reason to use a version that's already 2.5 years old when a newest LTS is fine.
Ps. Velocity also works fine on 24.04. If anything change that repo to 24.04. ๐
Is this a world being generated with Terra?
Still can reproduce on latest version
[09:46:55] [Region Scheduler Thread #4/ERROR]: [ca.spottedleaf.moonrise.common.util.TickThread] Thread failed main thread check: Cannot asynchronously load chunks, context=[thread=Region Scheduler Thread #4,class=io.papermc.paper.threadedregions.TickRegionScheduler$TickThreadRunner,region={center=[4800, -1392],world=world}], world=world, chunk_pos=[4859, -1394] java.lang.Throwable: null at ca.spottedleaf.moonrise.common.util.TickThread.ensureTickThread(TickThread.java:88) ~[folia-1.21.4.jar:1.21.4-DEV-873afea] at net.minecraft.server.level.ServerChunkCache.syncLoad(ServerChunkCache.java:99) ~[folia-1.21.4.jar:1.21.4-DEV-873afea] at net.minecraft.server.level.ServerChunkCache.getChunkFallback(ServerChunkCache.java:152) ~[folia-1.21.4.jar:1.21.4-DEV-873afea] at net.minecraft.server.level.ServerChunkCache.getChunk(ServerChunkCache.java:320) ~[folia-1.21.4.jar:1.21.4-DEV-873afea] at net.minecraft.wor...
also notice that 2d goldfarm there sometime piglin walkign aroudn the area outside the killchamber
But it could just be a vanilla feature
minecraft.wiki/w/Zombified_Piglin#Reinforcements
This feature is exclusive to Java Edition. On hard difficulty, zombified piglins have a 5% chance to spawn as leaders. Leader zombified piglins can spawn other zombified piglins as reinforcements.
After looking at this other related issue (https://github.com/PaperMC/Paper/issues/12057). this looks to be the case to me, this is not a bug.
Expected behavior
The method isInWall is not doing a thread check for the whole range. this is weird because this causes players to not be removed from the server due to doTick() getting called.
java.lang.Throwable: null
at ca.spottedleaf.moonrise.common.util.TickThread.ensureTickThread(TickThread.java:88) ~[folia-1.21.4.jar:1.21.4--293dd35]
at net.minecraft.server.level.ServerChunkCache.syncLoad(ServerChunkCache.java:99) ~[folia-1.21.4.jar:1.21.4--293dd35]
at net.minecraft.server.level.ServerChunkCache.getChunkFallback(ServerChunkCache.java:152) ~[folia-1.21.4.jar:1.21.4--293dd35]
at net.minecraft.server.level.ServerChunkCache.getChunk(ServerChunkCache.java:320) ~[folia-1.21.4.jar:1.21.4--293dd35]
at net.minecraft.world.entity.Entity.isInWall(Entity.java:2869) ~[folia-1.21.4.jar:1.21.4--293dd35]
at net.minecraft.world.entity.LivingEntity.isInWall(LivingEntity.java:4360) ~[folia-1.21.4.jar:1.21.4--293dd35]
at net.minecraft.world.entity.LivingEntity....
Expected behavior
ShulkerBox Should open properly after right click.
Observed/Actual behavior
There is an error: at java.base/java.lang.Thread.run(Thread.java:1583) ~[?:?]
[03:22:26 ERROR]: Could not pass event PlayerInteractEvent to NekoPlugin v1.4-SNAPSHOT
java.lang.IllegalStateException: Thread failed main thread check: Cannot queue block even data async, context=[thread=Region Scheduler Thread #2,class=io.papermc.paper.threadedregions.TickRegionScheduler$TickThreadRunner,region={center=[-89, 17],world=world}], world=world, block_pos=BlockPos{x=0, y=0, z=0}
Steps/models to reproduce
1.Write a plugin using the following code: https://pastes.dev/Xs6McEXrCv
2.compile plugin put and start server
3.add Shulker Box to you inventory and right-click on air
4.got error
Plugin and Datapack List
Plugin: Neko-Plugin(by me)
Datapack: Not have
Folia version
This server is running Folia version 1.21.4-DEV-dev/hard-fork@3af04d9 (1970-01-01T00:00:00Z) (Implementing A...
Stack trace
[03:37:50] [Region Scheduler Thread #2/INFO]: TT_Candy[**********] logged in with entity id 407 at ([world]767.7788530463216, 63.0, 790.3219713260949)
[03:37:53] [Region Scheduler Thread #3/INFO]: TT_Candy issued server command: /tp 0 150 0
[03:37:53] [Region Scheduler Thread #3/ERROR]: [ca.spottedleaf.moonrise.common.util.TickThread] Thread failed main thread check: Cannot asynchronously load chunks, context=[thread=Region Scheduler Thread #3,class=io.papermc.paper.threadedregions.TickRegionScheduler$TickThreadRunner,region={center=[50, 34],world=world}], world=world, chunk_pos=[0, -1]
java.lang.Throwable: null
at ca.spottedleaf.moonrise.common.util.TickThread.ensureTickThread(TickThread.java:88) ~[folia-1.21.4.jar:1.21.4-DEV-3af04d9]
at net.minecraft.server.level.ServerChunkCache.syncLoad(ServerChunkCache.java:99) ~[folia-1.21.4.jar:1.21.4-DEV-3af04d9]
at net.minecraft.server.level.ServerChunkCache.getChunkFallback(ServerChunkCache.java:152) ~[folia-1.21.4....
Cannot reproduce on non-hardfork branch master
1.21.1 1.21.3 1.21.4 reproduced
1.21.1 1.21.3 1.21.4 reproduced
I try to reproduce, but can't
This server is running Folia version 1.21.4-DEV-dev/hard-fork@5558184 (1970-01-01T00:00:00Z) (Implementing API version 1.21.4-R0.1-SNAPSHOT)
You are 1 version(s) behind
Download the new version at: https://papermc.io/downloads/paper
Previous version: 1.21.4-DEV-873afea (MC: 1.21.4)
This is tried on old fork, and how I remember I catch same crash
Error seems disappeared when set grid-exponent to 2
Error seems disappeared when set grid-exponent to 2
Just like a default?
Error seems disappeared when set grid-exponent to 2
Just like a default?
Default value is 4 in folia1.21.4()
Error seems disappeared when set grid-exponent to 2
Just like a default?
Default value is 4 in folia1.21.4()
Really? I'm set very long time ago to:
threaded-regions:
grid-exponent: 2
Error seems disappeared when set grid-exponent to 2
Just like a default?
Default value is 4 in folia1.21.4()
Really? I'm set very long time ago to:
threaded-regions:
grid-exponent: 2I will try with a default value (4)
In older version it is 4 but I don't know when it was changed to 4.I tested with new generated default config and it'4
So I still can't reproduce on new version of fork and with a default value
Error seems disappeared when set grid-exponent to 2
Just like a default?
Default value is 4 in folia1.21.4()
Really? I'm set very long time ago to:
threaded-regions:
grid-exponent: 2
I will try with a default value (4)In older version it is 2 but I don't know when it was changed to 4.I tested with new generated default config and it'4
Ok, the technical answer is, if the value is 4, regions will merge if they're more distant from each other.
I got a similar issue, but i also don't know why it happens
[13:56:37] [Folia Watchdog Thread/ERROR]: [io.papermc.paper.threadedregions.FoliaWatchdogThread] Tick region located in world 'world_the_end' around chunk '[-61, -6]' has not responded in 5.264367338s:
[13:56:37] [Folia Watchdog Thread/ERROR]: ------------------------------
[13:56:37] [Folia Watchdog Thread/ERROR]: Current Thread: Region Scheduler Thread #0
[13:56:37] [Folia Watchdog Thread/ERROR]: PID: 83 | Suspended: false | Native: true | State: RUNNABLE
[13:56:37] [Folia Watchdog Thread/ERROR]: Stack:
[13:56:37] [Folia Watchdog Thread/ERROR]: java.base@23.0.2/sun.nio.ch.UnixFileDispatcherImpl.write0(Native Method)
[13:56:37] [Folia Watchdog Thread/ERROR]: java.base@23.0.2/sun.nio.ch.UnixFileDispatcherImpl.write(UnixFileDispatcherImpl.java:65)
[13:56:37] [Folia Watchdog Thread/ERROR]: java.base@23.0.2/sun.nio.ch.IOUtil.writeFromNativeBuffer(IOUtil.java:137)
[13:56...
Expected behavior
The event should either not get called, or when cancelled it should properly handle the location, currently
it seems to be glitching.
Observed/Actual behavior
The blocks call relativeTeleport on the BLock for farms if you jump on them.
and if the event is cancelled, the teleportTo becomes out of the region.
Ideally likely to remove the call of the event
Steps/models to reproduce
Just jump on a block and cancel the event
Plugin and Datapack List
Folia version
dev/hard-fork branch.
at least until 3af04d9c6a98d24032bf8c99ad5446b0f381e320
Other
No response
Can you please show the full stacktrace?
If this was to be added, what has to be reworked in general @Spottedleaf
tryMarkTicking would probably become a region clock?
how would this work in terms of the ReentrantAreaLock