❗-1-18-experimental

[peak pulsar]

We're on build #74, so almost 🤞

[neon shard]

how much backups

[peak pulsar]

one per hour

[neon shard]

as if I have that kind of storage

[peak pulsar]

Just download more storage

pffft

It's easy. Same as downloading ram

[neon shard]

Yeah good idea

[peak pulsar]

[potent pilot]

is it possible to get chunkcaching back (thats not working since 1.16?)

[vapid prawn]

What?

[feral spire]

I am confused

[potent pilot]

in 1.12 - 1.14 (at least that are the version where i know it works)
if the server have viewdistance of 4 and u walk around then u can see like 10 viewdistance if your client is set to 10 viewdistance. since 1.16 or something u cannot see longer than serverviewdistance

[vapid prawn]

That's a client mod

[feral spire]

are you talking about optifine caching?

[vapid prawn]

There is some dedicated mod for it too

[feral spire]

bobby is latest hot shit I think

[potent pilot]

maybe i thoughts it was vanilla caching

[vapid prawn]

Do you have any idea how freaking hard it is to find a mod called "bobby" for a modding platform called "fabric"

Ok, not that hard if you include minecraft and/or mod

https://www.curseforge.com/minecraft/mc-mods/bobby

[potent pilot]

couldnt this made serverside if u tell the client a fake viewdistance packet or something

[vapid prawn]

Looks like optifine removed it in 1.17

So that lines up with how you thought it was

The server side version of this is setting view-distance higher than simulation-distance

[feral spire]

optifine has been lagging behind a lot now. They have not have any "stable" release since 1.16

[potent pilot]

but this is sending chunks still... thats not the same

[vapid prawn]

How could you do it without sending chunks? 😛

[potent pilot]

the client is holding them and not updating them like dead chunks, but would be pretty usefull for this new mountain terrain

[vapid prawn]

I mean, you might be able to do something clever with not sending unload chunk packets but I feel like this would end up killing clients

[feral spire]

bobby does that (kinda) they just save old chunks and not update them until you are actually close

[vapid prawn]

Right but it actually saves them to disk too, doesn't it?

[quasi panther]

looks like it yeah

[feral spire]

yes

[vapid prawn]

A server version either needs to send all the chunks or only works if you walk around the area first

Bobby works after the first time you walked around the area, server version you have to do it every time you enter the area

If you could even do a server version without breaking shit

tldr just install bobby

[potent pilot]

wait is this a fabric mod

[feral spire]

yes

[vapid prawn]

While you're at it replace optifine with sodium

Then you don't need to worry about optifine/fabric compat

[devout fox]

yeah use sodium

[restive marsh]

would yall say 1.18 is production ready

[rocky nest]

please read the channel name

[peak pulsar]

No, but it's semi-stable atm

You shouldn't be running it on production just yet

[restive marsh]

ok, i'll continue to viaverion for now. thanks

[inner crypt]

It's fine to run in prod if you are careful

Literally over 50 thousands servers do so

Just make sure your backups work

And that you did your own testing

[feral spire]

https://kennytv.eu/secretgraph/

I think paper kills it this update

[dapper lotus]

jesus fuck

[river parcel]

anyone have a working version of NexEngine for 1.8

[peak pulsar]

graph is slightly biased, but still, damn

[feral spire]

1.8 isn’t supported

[peak pulsar]

1.8 or 1.18?

[river parcel]

1.18

[vapid prawn]

Does https://minidigger.github.io/paperstats/#1.18 include 1.18.1?

[upbeat coral]

[rocky nest]

based off kenny/secretgraph, yes it seems to

[inner crypt]

Yes

You can even append 1.18.1 in the url

It's just the UI that isn't updated

[rugged rampart]

lithium

starlight

[inner crypt]

Gotta install half the periodic table

[feral spire]

Yes

[eternal briar]

is this number bukkit/spigot servers

[inner crypt]

Yes

[hidden light]

when is the ETA for 1.18

[half walrus]

__There Is No ETA__

Updates to Paper do not have any sort of estimate for when they release, ever. Any and all updates will arrive when they are ready, and the only thing to do is wait for them patiently along with everyone else.

ETA means 'Estimated Time of Arrival'!

[hidden light]

b r u h

[peak pulsar]

1.18 is already out

[hidden light]

srry

[peak pulsar]

[hidden light]

1.18.1

is the world gen bugs fixed?

[peak pulsar]

what do you mean?

[hidden light]

like will there cunck errors

now?

or has that been fixed

[peak pulsar]

Latest one is the most stable

[dapper lotus]

I mean, theres the vanilla memory leak

the vanilla "we don't clean up after ourselves" errors

but, it works

[peak pulsar]

does it exist in paper?

[polar quarry]

yep

[rough shell]

would paper forks say that it runs paper?

[inner crypt]

No

Generally not

[peak pulsar]

Depends on if they override bstats server metrics, right?

Not the server brand

[rough shell]

I think there should be an option for it to also include paper forks

[feral spire]

They ain’t technically paper so idk why it should

[inner crypt]

PRs Open

[vapid prawn]

The memory leak is when you upgrade an old world to 1.18

If you make a new world you're probably fine at this point

[quasi panther]

lectern bug is still open, isn't it?

[vapid prawn]

That's not really a big issue

It's just the client and server having a different idea of what the hitbox should be

[normal pecan]

build 76 fail?

nevermind

cache was bein weird again

[split steppe]

it's now fixed yeaaaaaaaaaa boii

[vapid prawn]

lol 28 minutes ago

That build probably just showed up on the website

[split steppe]

yes

bad english from my part

[languid harbor]

So the upgrade memory leak is fixed? Hell yeah

[shut kiln]

neat

[wheat spoke]

well that's useful

[vapid prawn]

Time to close down the channel

[timber ridge]

Haha

[crude orchid]

any idea why i get invisiblity 2 effect whenever i switch worlds / /invsee a player on 1.18

its most likely a plugin obviously, but any ideas by chance anyone

[dark wigeon]

vanish

invisibility 2 is usually related to vanish

we can't really guess at which plugin is causing it without a list of them though

[crude orchid]

i dont see anything about it in super vanish plugin, unless essentials vanish re enabled or something when upgrading

ill look more ty

[rough shell]

I'm unsure why I have the lectern box marked down as different (my version is correct) than what MCP and paper show, what's going on with this bounding box?

[dark wigeon]

lecterns are just too strong

[rough shell]

public static final VoxelShape SHAPE_TOP_PLATE = Block.box(0.0D, 15.0D, 0.0D, 16.0D, 15.0D, 16.0D);

this is likely the issue. The client simply phases through this box because... I don't know??

why do paper thin bounding boxes get ignored by the client but not by paper

[cloud swan]

anyone else getting a shit ton of lag?

[dark wigeon]

what

[crude orchid]

on what

lol

[cloud swan]

the new paper build

[timber ridge]

Which build?

[dark wigeon]

"12"

[cloud swan]

most recent one, #76

running a ton of ticks behind

[crude orchid]

with 20 players + 100 plugins

oh im on 71

hm

[cloud swan]

yeah like 100 ticks behind for me

and i have a 30GB ram server dedicated only to MC

only happens when the server starts

[timber ridge]

Windows is generally slow. But also don't give 30GB of RAM to Minecraft.

[steel dagger]

I think someone said it already, more ram is not nessecarily good

[timber ridge]

You generally only want like 10GB.

[prisma wing]

Why would you use windows server for MC hosting

[timber ridge]

If you're having lag, run and send a timings report.

[cloud swan]

i'm giving it -Xms16G -Xmx18G

[timber ridge]

xms and xmx should both be 10GB.

They should match and don't need to be more then 10.

(Generally)

[prisma wing]

But they're hitting 14 gb of ram usage

[timber ridge]

Unless you have like 100 people playing

[cloud swan]

at a time?

We have like 600+ players that cycle through

[timber ridge]

At once, yes.

[cloud swan]

hmmm

I guess i'll try lowering it

[dark wigeon]

i think the important thing here is that this is only happening on the latest one (i assume?)

so it should be related to the latest change

[cloud swan]

Yeah it was okay in the past

Like Java keeps going to Very High randomly and that's when the lag happens

[timber ridge]

Fix your RAM settings and send a timings.

[cloud swan]

What do you mean by timings?

[timber ridge]

Start the server, do /timings start and let it run for ~10-15min, then /timings paste

Post the link in here

[cloud swan]

alrighty just started it

[vapid prawn]

100 plugins 🤣

[cloud swan]

Our server runs fine with like 30 100 plugins is crazy

[feral spire]

Yeah as oslet said. The ram you had was unnecessary.

[crude orchid]

does anyone have geyser setup on 1.18 paper yet

[clear steeple]

In theory I don't know why it wouldn't work Geyser is pretty good at updating

[void viper]

Anyone else with wrong-linking portals?

[void viper]

yes.

[vapid prawn]

Are the portals really deep?

[quaint lodge]

It's working so far for me from my limited testing

[real bolt]

Yeah, it's a confirmed bug https://github.com/PaperMC/Paper/issues/6995

[void cipher]

is anyone else having crazy high ram usage on 1.18.1?

[tropic hawk]

Is that just with mulitverse or all servers?

1.17.1 world Gen memory leak is fixed now right?

[gusty trench]

Yep

[tropic hawk]

I have been tracking that issue. Anyone other issues I should know about before porting my server over

[vapid prawn]

You tell us

[real bolt]

It depends on what plugins you're using

So maybe it's an issue with a plugin you're running and it's just not known yet

[quasi panther]

that bug you linked doesn't have the status: accepted tag, that's not "confirmed" yet

at least not as a Paper bug

[tropic hawk]

what im wondering is if it has been replicated in stock paper

[quasi panther]

it doesn't sound like it

[vague falcon]

...stock paper?

[tropic hawk]

i mean paper no plugins

[long edge]

I haven't, but just reading that issue there's a lot of stuff that doesn't line up

Some of it could be caused by changing configuration

But it's also important that portals don't actually 'link', that's just a term people use

And then stuff like that guy saying that their portals are going to the same coordinates in the overworld/nether, that's not possible without plugins and doesn't line up with the rest of the stuff there

[molten sentinel]

Do any of the newer builds incorporate the second patch for log4j or is it still mitigated by just the flag being removed?
https://www.zdnet.com/article/second-log4j-vulnerability-found-apache-log4j-2-16-0-released/

[half walrus]

All the info we can provide is pinned in the #paper-help channel

[somber escarp]

God its nice to not have that memory leak lagging out my server now

[normal pecan]

I mean technically? Bungeecord instance is running geyser, all servers are 1.16.5, and one is 1.18.1, no issues

[crystal nacelle]

how stable is the most recent build? im going to assume not stable enough for a relatively big server yet

[tropic hawk]

i will let you know in a few days @crystal nacelle i got a server of about 10 people im about to migrate

[crystal nacelle]

sweet, thanks

let me know about the backup situation as well, seeing as in my server its a bit complicated to backup

[tropic hawk]

im gonna back it up, download the map, trim it in mcaselector, run it in 1.17 to test my changes, then update to 1.18.1

how big is your world and playerbase @crystal nacelle ?

[crystal nacelle]

not sure how big the world is, i know the playerbase is around 70+

[idle light]

https://github.com/PaperMC/Paper/commit/51d168752bc22bc37ade1d4577355cad6146ba56

from build 76 onwards :)

👍

[tired whale]

I just had a nether portal bring me to the top of the nether bedrock. I think it's because I was outside of the nether chunky border range in the overworld when I tried to portal into the nether.

[neon shard]

The scale titled overnight

[peak pulsar]

cool

[drowsy umbra]

Is paper also gonna get a fix for this now? I know paper just used the same patch from log4j 2.15 in build #64 but what about this now?
https://www.reddit.com/r/java/comments/rgg5hg/apache_log4j_216_released_to_address_missed_edge/?utm_medium=android_app&utm_source=share

[thorny jewel]

Which is faster, vanilla 1.18 or paper 1.18 (this is for vanilla survival)

[drowsy umbra]

paper

[pulsar dagger]

paper will be a lot faster yeah

[pulsar dagger]

we've updated to 2.16.0 for 1.17.1 & 1.18.1, do note the part about the missed cases being edge cases however, as far as we are aware they are not applicable to Paper, we've just updated in case

[drowsy umbra]

Yeah makes sense

And only just noticed the commit

[void nebula]

3rd day on early builds, still no bugs noticed in vanilla mode. But yes theyre exist im sure

[vivid iron]

message.txt by @harsh idol: https://paste.gg/6fad1e5824594b59b3f438ebab788c99

[peak pulsar]

help

[tawdry spoke]

Should I update my server to paper 1.18.1? There is usually 2 or 3 of us playing at once currently

[mortal token]

full log

[vivid iron]

fml.txt by @limber cove: https://paste.gg/109ce837591d4bdf98cacb554bab22d2

[fiery dock]

is paper 18.1 stable?

[timber ridge]

It's not marked stable yet, no.

[fiery dock]

is .0 stable?

[latent parrot]

.0 Is not further maintained no

And wasn't stable when .1 dropped

[ancient sage]

Is the memory leak fixed now? Or is the latest build not enough to fix it?

[feral spire]

the found one yes

[uneven pagoda]

Ye it's fixed, now waiting for merge #7115 to get every more performance increase..

[tropic hawk]

does 1.18 have anti xray?

[peak pulsar]

Yes, its in paper.yml

[tropic hawk]

thanks

[flat sable]

Glad the leak fix was merged, nice job

[tropic hawk]

so nether goes above default?

what about other preworld configs?

[still crane]

My server people will be happy that the memory leak is fixed.

[languid lodge]

What is the status on the disappearing villagers

[river cosmos]

We are still waiting for an answer from Hogwarts

[languid lodge]

?

Is that a person or are you referencing Harry Potter?

[feral spire]

Yes

[languid lodge]

So clear EterNity, thx

[vapid prawn]

Disappearing villagers is an issue open in the Mojang bugtracker for years now

[languid lodge]

Ah

[vapid prawn]

Sometimes they just randomly go away

[languid lodge]

So Mojang, not paper

[vapid prawn]

Mojang added some logging to try to at least see when it happens but otherwise no one knows

[languid lodge]

Could there be a possible paper patch for disappearing villagers?

[somber escarp]

Sometimes the villagers just really need to go out for a carton of milk and some smokes

[languid lodge]

I feel those vibes though

[vapid prawn]

Could Paper fix it? Sure, if someone could figure out why it happens. I wouldn't count on it.

[languid lodge]

Interesting

Time to throw Log4j statements at everything lol

[vapid prawn]

There might be other causes from both vanilla and Paper that lead to this and those would be more likely to be fixable but I haven't heard of any

[languid lodge]

Fun

[vapid prawn]

If you figure out a way to reliably reproduce the problem a lot of people will love you

[languid lodge]

Just checking because from what I heard, it was a Paper 1.18.1 exclusive issue

And it isn't

As I just found out

[vapid prawn]

I've not heard of any Paper-specific issue causing villagers to suicide

Might be one though

[languid lodge]

Well at least i have something to do now

[still crane]

Wait. There are disappearing villagers?

[inner crypt]

Always have been

[vapid prawn]

https://bugs.mojang.com/browse/MC-153904

And if you think that one is bad check out https://bugs.mojang.com/browse/MC-161823

[uncut pelican]

that one happened to me when my server corrupted last year

half of a friend’s wizard tower was split from the rest and rotated 90 degrees

as well as that, a random chunk generated like 50 blocks higher than the terrain around it and a different biome too

[brittle aspen]

Didn't work. Its going to 8gb same as before. Maybe little slower, but not much.

[peak pulsar]

Update to latest paper, there was a memory leak

[brittle aspen]

Im updating to 1.18.1 now

Updated, hoping it worked

[upper shuttle]

@peak pulsar Barry: Invalid Request!
Please follow the guidelines listed here: https://helpch.at/faq/request-free
Your Request: https://paste.helpch.at/kowutaceyi 😦

@peak pulsar ?

[peak pulsar]

Sorry but im not gonna help you with that

If you can write, you can also read

[junior axle]

the anti xray works below height 0? are a new config or something?

[peak pulsar]

it does, it should be configured correctly by def

[agile notch]

Hi When does a stable generation of what kind wake up the pain ?

[inner crypt]

Wat

[dapper lotus]

wat

[inner crypt]

Try rephrasing that please

[peak pulsar]

When it is ready

[warped sleet]

He probably asks when a stable build will be there. But still... 🔁

Check the announcements and don't expect an ETA

[noble quest]

"stable generation" could be talking about world generation, which might mean upgrading 1.17.1 worlds, which has recently been updated.

[neon shard]

^

[neon shard]

I love this

[coarse gull]

Man asked that and then vanished

[noble quest]

It's confusing because there's a few almosts of what the sentence could be referring to.

Left us with a conundrum.

[steel dagger]

some people just want to set the fire...

[iron portal]

https://www.lunasec.io/docs/blog/log4j-zero-day-update-on-cve-2021-45046/ welp, time to wait for another paper update

[vestal veldt]

it’s already updated

[iron portal]

Yea scrolled a bit up

[noble quest]

Build 75 updates Log4J to 2.16.0, which has all fixes. That article is a post mortem.

[next owl]

part of end became plains

and creepr skely spawning

[lapis belfry]

Any idea when 1.18 will be considered at a somewhat stable state?

[half walrus]

__There Is No ETA__

Updates to Paper do not have any sort of estimate for when they release, ever. Any and all updates will arrive when they are ready, and the only thing to do is wait for them patiently along with everyone else.

ETA means 'Estimated Time of Arrival'!

[hot topaz]

Hey! From any reason I am getting kicked from my server when I join/teleport

I use the latest paper jar (1.18.1)

[next owl]

it is rn

not stablw

but out of experimental

[lapis belfry]

When did it leave experimental?

[warped sleet]

Does that matter?

[lapis belfry]

Was just wondering sorry

[devout fox]

no-tick view distance doesn't seem to be working

[dapper lotus]

doesn't exist in 1.18 yet

[devout fox]

like the server is just using the tick view distance

ah

[acoustic burrow]

[honest ruin]

i thought it was renamed :
view-distance: simulation-distance (server.properties)
no-tick-view-distance: view-distance (server.properties)
am i wrong?

[languid lodge]

Somebody send the image

[dapper lotus]

wasn't renamed

the paper patch hasn't been updated for 1.18 yet

but, vanilla added their own mechanism for it

[slate monolith]

https://is.paper.sexy/ViewDistance.png

[honest ruin]

ah ok, because in paper optimization guide i read this:

[steel heart]

question thats been asked a thousand times.... if a client has a render distance of 10, but the server is 20 for view distance.... does the server still load 20 chunks for the player?

[dapper lotus]

the paper optimistation guide

bruh

that image that was just linkerono'd was at the top of the optimisation guide

[next owl]

[feral spire]

It’s up-to-date now so you can just follow it and yes it makes no mention of no-tick now because as zcat said it no longer exist

[vapid prawn]

Pretty sure it does right now, that's a part of what's missing

[vapid prawn]

iirc the rate limiter isn't in 1.18 yet so paper is probably loading and sending chunks faster than your desktop can handle

Server is flooding you out because it's too fast 😄

[crude orchid]

how long would u say until fixed? i dont mean exact date but like month, 2, 3

if you can say

[vapid prawn]

¯_(ツ)_/¯

[crude orchid]

could that be whats destroying my geyser players connections?

[vapid prawn]

Maybe

[crude orchid]

anyone on bedrock has terrible network lag

[vapid prawn]

This would only happen on join/teleport and you might notice a little something when moving quickly

[crude orchid]

yeah on join

look if u have a sec

[vapid prawn]

You can mitigate it by turning down your view distance

[crude orchid]

https://youtu.be/4LPiYJLh6YQ

bedrock client is unresponsive for like 2+ minutes / 15 join attempts

[vapid prawn]

Turn down your view distance and see if it gets better

[crude orchid]

view distance on server or client

[vapid prawn]

Right now afaik there is no rate limiter and the server doesn't respect the client view distance

So all you can do is turn down the view distance server side

[crude orchid]

ahh ok ok gotcha ill try that

[uneven pagoda]

(one of my friends just wants to understand this moment)

<Assaq> but anyway i dont know why instead of porting the remaining patches/fixes they just add completely new features like schooling fish api

[vapid prawn]

There is like one person who knows how to do the remaining patches

Could someone else do them? Maybe, after enough time. The person who wrote them can probably do them in a couple days though and wants to be the one to do them

So it's just a matter of waiting for them to have some free time and feel like working on this stuff again

Or giving up on them and probably dropping half those patches and someone else figuring out how to update the others

[zealous idol]

prob has finals rn lol

[bold breach]

Does this basically mean that if I set my view distance client side to 6, and the server is set to 20 that my client would be forced to try to run at 20?

[vapid prawn]

Yes

Well, it'll get sent 20 radius worth of chunk packets

It'll only try to mesh and render them out to 6

[bold breach]

So it's rendering 20 and basically fogging it out

[vapid prawn]

No, it's rendering 6

It's downloading and decompressing 20

[bold breach]

Ah

Would that affect the performance or the connection more?

[vapid prawn]

...at least I think the client is smart enough to skip meshing the extras

It definitely doesn't render them

Well, the bandwidth and all the CPU cores being used up to decompress all that data and then mesh all the stuff it actually wants to render can overwhelm your computer and/or network

Especially since Paper loads chunks so fast so they're probably coming in more or less all at once

[bold breach]

Alright

Thanks

[vapid prawn]

One person mostly solved their problems by changing view-distance from 10 to 8

Client still got laggy on join/teleport but only for a moment and it wasn't timing out

I think they use a circle now so the number of chunks sent is different but the old setup would be 441 chunks for 10 and 289 for 8 so there is a big difference

[bold breach]

It got changed to a cylinder

[vapid prawn]

Those numbers will both be smaller in 1.18 but the ratio should be similar

[bold breach]

iirc 1.18 renders more blocks vertically and the same horizontally, and has less fog

*1.18.1

[feral spire]

fog technically has nothing to do with it as it is just there to hide the ugly edges you can remove it with mods

[bold breach]

Tradeoff is depth of field

[brittle aspen]

It does seem to help a bit, but it does not completly. It was on about 7/8gb ram with 7players, 3 left, but the ram stays the same. It isnt freeing it up. (I upgraded my plan btw to 10gb (Its free for me) )

[peak pulsar]

what are your flags?

[vapid prawn]

Depending on how you're measuring RAM that's normal

From an OS level the JVM will never use less RAM than whatever the max it hit was

It doesn't give RAM back to the OS

[brittle aspen]

https://timings.aikar.co/?id=07be70bbb0014b438798687b354c9cc4

[dapper lotus]

?

java using the memory you csaid it can use is literally a non issue

[brittle aspen]

Yes but is using a lot.

[dapper lotus]

which is normal for java

[brittle aspen]

I first had 8gb and it crashed alot

[dapper lotus]

that memory is generally not actually being "used"

[vapid prawn]

If you set Xmx to 10GB eventually it'll be using around 10GB

[dapper lotus]

minecraft allocates a lot of crap fast

[brittle aspen]

I know, but it crashes when more that 8gb is allocated

[vapid prawn]

So it can delay doing GCs longer or do them slowly in the background

[dapper lotus]

because your host is misconfigured

[vapid prawn]

How much RAM does your computer have?

If you get an 8GB VPS then you can't tell Minecraft to use 8GB

[dapper lotus]

if using X amount of ram causes the thing to crash, you shouldn't be allocating X

[brittle aspen]

Im hosting at a provideer

[half walrus]

-Xmx controls the amount of heap memory assigned to the JVM, this does not include other memory used by java, or native memory used by other libraries such as netty (for networking) or SQLite. Please do not allocate all of your memory!

[vapid prawn]

The OS needs some RAM too

[dapper lotus]

if you are having issues past that, you get into the territoriy of needing heap dumps, etc

[brittle aspen]

it dropped 1gb now btw

So i set my xmx to 7gb when I still had 8gb m=but didnt make a diffrence

[peak pulsar]

@brittle aspen tell them to set 7gb in your startup flag but in pterodactyl to allocate 8gb or 8.5gb

[brittle aspen]

I have 10gb server now and atm its xmx is 7gb

So I change that to 8?

[dapper lotus]

2G should be fine

[peak pulsar]

so in pterodactyl they allocated 10gb?

[brittle aspen]

yes

[dapper lotus]

as said, if ithe jvm actually runs out of memory, you need heap dump, basically

[brittle aspen]

Im not sure what that means. Just lowering my xmx?

[dapper lotus]

heap dump?

you basically "dump" the contents of the JVMs heap to the disk so you can load it up and see what's using memory, etc

[brittle aspen]

And how?

[dapper lotus]

depends on how it's crashing

[brittle aspen]

Its just allocating 8.1gb and than it crashes

[dapper lotus]

if the server logs start screaming about it being out of memory, there is a JVM flag to make it dump when it runs out

You mean, the memory usage hits 8.1G in the panel, and it crashes?

[brittle aspen]

Yes

[dapper lotus]

crashes how exactly as there are many means of failure there

[brittle aspen]

just not responding

we get timed out and I need to kill the server

[dapper lotus]

But, if you're telling us that it's crashing at 8.1G on the panel, but, you have 10G allocated to the container, that screams like a host issue

[brittle aspen]

to make it work

[brittle aspen]

now 10.1 but before 8.1 because I just upgraded

[polar quarry]

it doesnt matter if your client shows "timed out". how is the server crashing? what is it saying in the console (or logs)?

[vapid prawn]

If you get a java.lang.OutOfMemoryError your Xmx is too low, if you get Killed your Xmx is too high, if you get random "server stops responding" that doesn't sound like a memory thing

[dapper lotus]

that's docker pausing the process when it runs out of ram, basically

[vapid prawn]

Unless your GC flags are bad and it's pausing for several seconds to do a full old gen

[dapper lotus]

when it pauses like that you're generally SOL

[vapid prawn]

I've never heard of that

[dapper lotus]

you can use /paper heap to create a dump but then you need to get into analysing it

but, Xmx just covers the heap, which then leads into many complex areas

try without plugins, I'd maybe guess

beyond that, no idea what to recommend

if you disable the OOM killer on docker it basically just causes the process to freeze instead of being killed

[brittle aspen]

I have a heap dump npw

now

it actually dropped to 5.5gb now

with the same players online

there is actually 1 more

[real bolt]

I haven't actually checked if the portals don't line up, but all I know is that portals used to work one way in 1.17.1, and now they act differently in 1.18.1

I assume it's a bug because it usually spawns a new portal in the overworld 5-50 blocks away from the intended portal, where "linking" would be a better option

[feral spire]

do you includes the difference 1.17 vs 1.18 on the original issue on tracker?

[real bolt]

I didn't make the issue

[feral spire]

there is one I believe if you have additional info to add on to it

[real bolt]

Oh, right, my bad

I might, I can't add right now though

Thank you for telling me this

[feral spire]

yeah the more info you can giff the better

[radiant gust]

What's ur setup actually

[umbral saffron]

Thanks to all of paper for fixing WorldGenRegion leak 🥳

[agile notch]

20:37:26 ERROR: The server has stopped responding! This is (probably) not a Paper bug.
20:37:26 ERROR: If you see a plugin in the Server thread dump below, then please report it to that author
20:37:26 ERROR: Especially if it looks like HTTP or MySQL operations are occurring
20:37:26 ERROR: If you see a world save or edit, then it means you did far more than your server can handle at once
20:37:26 ERROR: If this is the case, consider increasing timeout-time in spigot.yml but note that this will replace the crash with LARGE lag spikes
20:37:26 ERROR: If you are unsure or still think this is a Paper bug, please report this to https://github.com/PaperMC/Paper/issues
20:37:26 ERROR: Be sure to include ALL relevant console errors and Minecraft crash reports
20:37:26 ERROR: Paper version: git-Paper-76 (MC: 1.18.1)
20:37:26 ERROR: ------------------------------
20:37:26 ERROR: Server thread dump (Look for plugins here before reporting to Paper!):
20:37:26 ERROR: ------------------------------
20:37:26 ERROR: Current Thread: Server thread
20:37:26 ERROR: PID: 22 | Suspended: false | Native: false | State: RUNNABLE
20:37:26 ERROR: Stack:
20:37:26 ERROR: net.minecraft.advancements.AdvancementProgress.hasProgress(AdvancementProgress.java:67)
20:37:26 ERROR: net.minecraft.server.PlayerAdvancements.save(PlayerAdvancements.java:228)
20:37:26 ERROR: net.minecraft.server.players.PlayerList.save(PlayerList.java:584)
20:37:26 ERROR: net.minecraft.server.players.PlayerList.remove(PlayerList.java:630)
20:37:26 ERROR: net.minecraft.server.players.PlayerList.remove(PlayerList.java:591)
20:37:26 ERROR: net.minecraft.server.network.ServerGamePacketListenerImpl.onDisconnect(ServerGamePacketListenerImpl.java:1973)
20:37:26 ERROR: net.minecraft.server.network.ServerGamePacketListenerImpl.onDisconnect(ServerGamePacketListenerImpl.java:1953)
20:37:26 ERROR: net.minecraft.network.Connection.handleDisconnection(Connection.java:712)
20:37:26 ERROR: net.minecraft.server.network.ServerConnectionListener.tick(ServerConnectionListener.java:221)
20:37:26 ERROR: net.minecraft.server.MinecraftServer.tickChildren(MinecraftServer.java:1636)
20:37:26 ERROR: net.minecraft.server.dedicated.DedicatedServer.tickChildren(DedicatedServer.java:480)
20:37:26 ERROR: net.minecraft.server.MinecraftServer.tickServer(MinecraftServer.java:1470)
20:37:26 ERROR: net.minecraft.server.MinecraftServer.runServer(MinecraftServer.java:1264)
20:37:26 ERROR: net.minecraft.server.MinecraftServer.lambda$spin$0(MinecraftServer.java:317)
20:37:26 ERROR: net.minecraft.server.MinecraftServer$$Lambda$3929/0x00000008014022d0.run(Unknown Source)
20:37:26 ERROR: java.base@17.0.1/java.lang.Thread.run(Thread.java:833)
20:37:26 ERROR: ------------------------------
20:37:26 ERROR: Entire Thread Dump:

?

[umbral saffron]

Use PASTEBIN DUDE!

[dapper lotus]

your server took too long to tick

.paste logs

[half walrus]

Please send large files/logs to a pastebin

[wheat spoke]

woah

so, whatever bug fix that stopped the memory leak? It caused my spawn to respawn

well thats... weird.

[feral spire]

Wut

[wheat spoke]

that square is my spawn, it was totally different before i applied the most recent experimental patch

had lots of stuff

[umbral saffron]

looks like different seed to me tbh

[feral spire]

That’s a known issue I think

[wheat spoke]

oh, i know its a different seed

[feral spire]

In game looks fine right?

[wheat spoke]

nope

totally nuked and replaced

hard edges and all

[feral spire]

Ok that’s not known issue

thought you mean renderer

[umbral saffron]

heh I was just about to upgrade xD

[agile notch]

how to fix it ?

[wheat spoke]

i'm gonna roll back i hope

[dapper lotus]

Don't mention.

I already told you what we needed to see.

[wheat spoke]

fuck, my last rollback is before i regenerated the entire damn map

MCA selector doesn't work with 1.18 gen either

(yet)

[long edge]

it does?

[wheat spoke]

can't get it to load, it crashes every time i attempt

[long edge]

what does it crash with

or, go to https://discord.gg/JAawR9se but it definitely does work

[wheat spoke]

it just gets to near 100% CPU utulization and then just... hangs

there's a big report for it, apparently happens if you have pre 1.18 gen and 1.18 terrain on the same server?

[feral spire]

Don’t tell me it leaks on program too

[wheat spoke]

oh wait

its actually loading now

i hope it doesn't crash again, i'm gonna go shower and see where its at when i'm done, if i can just replace that chunk my life will be a lot easier

[wheat spoke]

nope, crashed again

[wheat spoke]

sweet, got it to work long enough to use a backup to replace the chunks that got replaced when i upgraded for some reason, lets see if it this sticks

[wooden rover]

does world gen use significantly more resources (like ram and cpu) in 1.18 compared to 1.17?

[dapper lotus]

yes

[uneven pagoda]

but if every map are pregenerated, these is no diff?

[feral spire]

There is still the task of sending the chunk to player

[dapper lotus]

theres a wider column to deal with

i.e. chunks are larger in memory

there is more area of a chunk to tick

etc

[crude glacier]

2 more weeks

[wooden rover]

any idea like, how much more ram it uses

my server will probally crash and burn

[dapper lotus]

idk

[wheat spoke]

I was using 12-14 gigs or so when using chunky

i do have a couple other plugins though

[wooden rover]

yeah i probally should pregen

[mossy wigeon]

Can we all agree that paper 1.18 is stable and production ready on New Years? Like imagine having multiple servers doing New Years resets

[wheat spoke]

It'll be stable when its stable

after what happend to me this morning, i'd say it needs some work.

[wooden rover]

what was the world size, like in gigabytes and blocks

[mossy wigeon]

No no, I’m not asking for a ETA, I’m saying that what if the developers literally holded it off from being stable until New Years just say they can say new year, new stable version or something

[half walrus]

__There Is No ETA__

Updates to Paper do not have any sort of estimate for when they release, ever. Any and all updates will arrive when they are ready, and the only thing to do is wait for them patiently along with everyone else.

ETA means 'Estimated Time of Arrival'!

[mossy wigeon]

Are you reading what I’m saying or am I describing what I mean too poorly?

[wooden rover]

ohh just re-read it

im stupid

that would be cool tho

[mossy wigeon]

Sometimes I word things poorly, my bad

[mossy wigeon]

I agree

But it is what it is, exams week is a struggle for some people

[wooden rover]

my bad, i read the bottom half of the message u sent

[rugged rampart]

see, the problem is that holding off on updating isnt really a good thing

you want things to work properly as soon as possible

[mossy wigeon]

But PR

[feral spire]

There are still issues needed to be fix

[mossy wigeon]

True though some owners need as much time as they can get

[feral spire]

And big patches not reapplied

[mossy wigeon]

Regardless, that’s just a thought I had since this whole New Years for Minecraft thing in my YouTube feed

[next owl]

e

[bold breach]

That should be something server owners can decide on their own

No reason to withold patches which make the server stable, especially with many servers already currently running experimental versions

[dapper lotus]

wat

[half walrus]

__There Is No ETA__

Updates to Paper do not have any sort of estimate for when they release, ever. Any and all updates will arrive when they are ready, and the only thing to do is wait for them patiently along with everyone else.

ETA means 'Estimated Time of Arrival'!

[crude glacier]

https://tenor.com/view/walter-white-shocked-fall-4k-4160p-gif-21623835

[bold breach]

It's not an eta thing, it's deliberately witholding the final patch until new years

Also they already did /eta

[next owl]

karens of github

[dapper lotus]

they're more talking about the status, not the withholding of patches

[bold breach]

No reason to call it unstable if it isn't tbh

[next owl]

[eager flame]

is it that hard for people to just sit and wait?

[gusty trench]

I don't think there's anyone withholding patches and calling 1.18 unstable

[feral spire]

I have hard time understanding what Jaxpo is saying. I hope I am not the only one

[bold breach]

I'm talking about the previous conversation

[bold breach]

This

[feral spire]

You know he’s not part of the project developers right?

[bold breach]

I'm not saying he is

Just giving a comment on his suggestion

It's a hypothetical

[next owl]

This ain't mojang a company to do that kinda new yer stuff

also many servers like mine reseted with eaarly builds

[bold breach]

I already reset to do early testing

[crude glacier]

2 more weeks trust the plan

[hidden cargo]

Why run early builds straight in prod

[bold breach]

?

[bold breach]

Server isn't released yet, it's shutdown and the whitelist is cleared so only some people have beta access

[hidden cargo]

Ah kk that's dif

[bold breach]

It's basically to see if there are any issues with plugins and stuff that aren't due to the experimental build

[hidden cargo]

@next owl was talking about him tho

[bold breach]

Ah

Probably to jump on the hype but let's see if they respond

[hidden cargo]

Lol

[sullen heron]

I think no tick view distance doesn't work. My no tick distance setup on 12 chunks, but I see only 4 chunks from server.properties/spigot.yml

[peak pulsar]

well no-tick view distance was removed in favor of vanilla's simulation-distance

[vapid prawn]

https://is.paper.sexy/ViewDistance.png?1

[sullen heron]

in paper.yml it's doesn't removed, I think Paper don't remove old strings?

[peak pulsar]

correct

[long edge]

there probably should have been a config migration, but, it's pretty much too late for that now and not enough was known at the time to determine how it would end up working

[sullen heron]

okay
If I setup 4 real chunks and 12 simulation
4(active chunks)+8(no tick chunks)?

[long edge]

but new servers won't get that added

[vapid prawn]

Other way around

[long edge]

simulation is ticking chunks

just normal view-distance is non-ticking

[sullen heron]

Same feature from paper no-tick-view-distance?
I did not really like the simulation of the distance from Mojang, because for some reason they spawn mobs outside the real chunks as before

[feral spire]

a bit different

[sullen heron]

Understood thanks

[vapid prawn]

I think the mob spawn thing was a Paper bug

[sullen heron]

But I tested it in 1.18 default minecraft server
My fellow developers also noticed this problem in the simulation
Most likely this is a minor flaw in Mojang

[half walrus]

__There Is No ETA__

Updates to Paper do not have any sort of estimate for when they release, ever. Any and all updates will arrive when they are ready, and the only thing to do is wait for them patiently along with everyone else.

ETA means 'Estimated Time of Arrival'!

__There Is No ETA__

Updates to Paper do not have any sort of estimate for when they release, ever. Any and all updates will arrive when they are ready, and the only thing to do is wait for them patiently along with everyone else.

ETA means 'Estimated Time of Arrival'!

[prisma wing]

Guys we're close to 5k

[worldly holly]

eta

[normal pecan]

345 in this channel, but i assume 1/2 of those aren't asking for one

[sullen heron]

Mojang added Per player spawn mob?
It's added in 1.18 snapshots?

[worthy thunder]

Are there any differences feature-wise between simulation distance and no-tick?

[silent quest]

eta

[feral spire]

Not the same

https://is.paper.sexy/ViewDistance.png

[worthy thunder]

Ugh I'm tired, between the new view distance and old no-tick someone told me it messes with some stuff differently than no-tick, but I couldn't really verify it so I assume it's cap

[peak pulsar]

anyone can help me with my timings? https://timings.aikar.co/?id=3dd01b853cfd48048e3b7ae3aecc32f0

[rocky nest]

purpur

[quiet pine]

ask purpur

[peak pulsar]

Ten bucks paper gets marked as stable within next 5 days

Protip: Don't bet on this with the core team

O_o

they can decide the outcome ;)

[rocky nest]

@.kashike want an easy 10 bucks?

😛

[peak pulsar]

I take back statement

[feral spire]

Empty your brain and just follow this

[half walrus]

https://eternity.community/index.php/paper-optimization/

[pallid merlin]

Paper Chan nendoroid when

[vapid prawn]

Supposedly no-tick worked better but all the things I've seen have been Paper bugs

[worthy thunder]

I hereby declare this guide biased

[dapper lotus]

There is a behavioral change between the two mechanisms but there is basically a few cases where paper now refs the wrong value somewhere

I'm still under the assumption that no-tick is coming back, but, idk how much or what

[feral spire]

Leaf said something along the line of just overwriting it but using the Mojang name but he’s still busy

[dapper lotus]

Imma go pee on his rug

[feral spire]

(Ask him how many diamonds he got last night) he will understand the reference

[worthy thunder]

Is it anything game-breaking 🤔

[stuck mural]

this was a change they reverted during the snapshots. vanilla minecraft does NOT have per player mob spawns

[sullen heron]

it's bad :(

[dapper lotus]

the entire mob spawning logic just needs rewriting

[feral spire]

We noticed some inconsistency

I don’t think Mojang revert it do they?

[dapper lotus]

idk

I'd revert you if I could

[feral spire]

[arctic oracle]

can anyone give me a hand with this? beyond the obviously missing flags i need to bug the host for?
https://timings.aikar.co/?id=721abe4594bb41e4af704c40016bdc6f

[timber ridge]

Basically Aikar's flags, set the xms and xms to the same, probably drop the xmx a little bit depending on players, but a lot of it looks like chunks and villagers.

Pregen the map if you haven't

[arctic oracle]

map has been pregened and i am working to discourage 1x1 villager breeders, anything else in there?

[dapper lotus]

Generally, you just have too much going on

[crude glacier]

has anyone updated their 1.17.1 paper server to the experimental builds

[arctic oracle]

good to know, i know that there used to be a plugin to nerf villager ai without breaking them, is that still a thing?

[crude glacier]

cant be farked to wait any longer just wanna play

how stable is it

[arctic oracle]

somewhere between unstable and known stable

[crude glacier]

https://tenor.com/view/walter-white-shocked-fall-4k-4160p-gif-21623835

[arctic oracle]

i mean i have been running in prod for a week if that helps

[dapper lotus]

only 50k servers running it right now

[eager flame]

i mean whats wrong with just waiting

[timber ridge]

You can update if you want - just make sure to backup everything so you can restore if there's an issue. Maybe warn your players that it's experimental still.

[normal zephyr]

except chunk generation and goats, performances seems better on this 1.18 ? 🙂

because when I check some timings in this channel, I don't want to update from my 1.16 and stay outdated but with more performances

[solemn blade]

just wondering if anyone else has seen a significant increase in general chunk loading lag from antixray (engine 2) in 1.18 compared with 1.17? ig it’s definitely expected with the extra height but i’m wondering if there’s anything i can do to mitigate it

[feral spire]

Lower view distance

[normal pecan]

anyone have that link to the todo list?

[autumn comet]

!kick 231958939024097282 Annoying troll; asking daily for etas is not gonna change anything

[wheat heart]

:raised_hands: Kicked Big Musk#3895 (Annoying troll; asking daily for etas is not gonna change anything) [1 total infraction] -- NotMyFault#3732.

[normal pecan]

i noticed lecterns & memory leak are now patched, just curious what's still on there

[dapper lotus]

It's still linked in the 1.18 channel, where it always has been

[normal pecan]

not seeing it in #❗-1-18, is it somewhere else?

[dapper lotus]

near the top

[normal pecan]

ah i see

i was looking at the links

not the message

[lavish sable]

I just updated to 1.18.1 and this happened , kinda funny , but wondering if anyone else had the same problem? https://pastebin.com/6H1RUyHd

[vapid prawn]

I just always assume these are https://bugs.mojang.com/browse/MC-153904

But Paper does change things around chunk loading so it might be a Paper bug, if you load from a backup is it reproducible?

[peak pulsar]

Since I switched to 1.18 several of my players have trouble connecting to the server, like they connect and immediately get disconnected. One of my players was able to solve the issue by switching to the vpn, this makes no sense to me, because that should increase his latency, not decrease it, any ideas?

I would says this affects about 1/5th of my players.

I assumed it was the antixray plus larger chunk sizes bogging down their connection, but this guy fixing it with a vpn breaks that theory.

[floral blade]

Wasn't that an issue fixed in 1.18.1?

[peak pulsar]

I'm using the latest patch.

These are players with 100s of hours on the server who now connect, say something in chat, can't see chat and are disconnected D-:

[native hornet]

You need to post logs, nobody can help you otherwise

[peak pulsar]

just my console log?

[dark wigeon]

bad route, maybe

i had a similar thing happen on my server, a guy complained about bad connection (on 1.17.1) and using a vpn fixed it for him

[peak pulsar]

hmm okay

I have a vpn I use privately, installed on the server itself, I wonder if that would fix it.

(letting trusted people use it)

[astral berry]

When you say 50k servers are running it right now, does that mean 50 000?

[dark wigeon]

there could still be a bad route to that vpn

[dark wigeon]

no there are 50 special "k" servers

yes, 50,000

[astral berry]

I didn't even know there were 50 000 servers lmao

I suppose there would be

[dark wigeon]

there are many more than that

[vapid prawn]

It's probably a routing thing but what's your view-distance set to? If the network is kind of okay but can't handle a burst of traffic that could impact it

[peak pulsar]

my server is played by people who want a server with high view distance, so it's set to 15, I've been recommending that players with connection issues lower their view distance thinking it's the same thing.

[vapid prawn]

The server doesn't currently honor the client view distance, afaik

[peak pulsar]

ooh shit, that would explain it

[vapid prawn]

Or have the chunk send rate limiter

[peak pulsar]

chunk send rate limiter?

[vapid prawn]

Yeah, it turns out Paper got so fast at loading chunks it was killing clients by sending them too fast

[peak pulsar]

that is a setting I can change then?

[vapid prawn]

I don't think that patch has been updated for 1.18 yet

[dark wigeon]

@Spottedleaf#0001

[peak pulsar]

okay, well this sounds like it is likely the issue

[vapid prawn]

The only thing you can do right now is turn your view-distance down on the server

Heck, the VPN might be helping by being slower so the client has some time to process things 😄

Or it might just be a bad route that can handle a little traffic but not the flood of chunks you get on join/teleport

[peak pulsar]

thank you very much for explaining it in a way I can pass on, I'll be temporarily lowering my view distance from 15 to 10.

[crude orchid]

do you use geyser?

i think the same shit is happening to me right now

[peak pulsar]

yes, but funny enough my BE players have no issues.

[crude orchid]

bro r u on tcpshield?

[peak pulsar]

no

[crude orchid]

my bedrock players just load infinitely, and then dc

only some players tho

if you use a vpn to go to seattle area it happens for example

its like routing or some shit like said above idk, only happens in some locations

[peak pulsar]

I am lowering view distance by 33%, so will be interesting to see if it makes an impact.

[crude orchid]

view distance is 1 on my test server

and i have the issue on there

so idk i dont think so

hopefully for you tho

[peak pulsar]

hmm, I think we might have different issues, since my BE players are fine

[wheat spoke]

I have one friend who can't join for what I assume is the view distance thing, but i have others that join fine

i don't know if he's using iris

er. sodium*

[native hornet]

I petition we take away leafs burritos until he fixes the patch

[cunning tiger]

is the lectern glitch fixed?

[dark wigeon]

yes

[quasi panther]

patch accepted, bug closed, build 77 should have it in it

[still citrus]

Hey I know about the ol asking for an eta thing but: are the builds for 1.18.1 in the #❗-1-18 chat considered stable?

[feral spire]

read the note on download page

[marble mural]

This is werid, when respawning it takes 5-7 seconds for my client to actually update. so those 5-6 seconds no mobs are loaded, none of my movement actually registers (client side it does but when finishedl oading it fixes)

[feral spire]

reduce view distance and simulation distance and see

[marble mural]

client or server?

[feral spire]

server

[marble mural]

current ones are at
view-distance: 14
simulation-distance: 10

what's recommended, without ruining the experience

its just a server for me and my friends (3-6 players max at once)
8gb ram 6gb allocated to game.

[feral spire]

I do not know what you have

but I all I know is the chunk limitor patch and respeact client view distance patch aint in the paper yet so if you set it too high, there is a chance client just dies

[marble mural]

ahh

[feral spire]

you may lower it and wait for it to be officially brought back.

[marble mural]

some of that is a bit confusing

alright

my main pc, should easily run fine, was fine before 1.18 paper

my server, isn't anything amazing. really just an i3, not sure what gen

[feral spire]

yes coz the patch abovementioned is in effect.

[marble mural]

nah still takes 4 seconds to be in affect.

the way im testing is using a freeze plugin.

i /kill myself

once I respawn i can run around for like 4 seconds before it teleports me back and freezes me

[quasi panther]

your server is running on an i3? 🤨

that's not "isn't anything amazing", that's "amazing in the wrong direction"

[marble mural]

it runs

only using 50% rn

and im running chunky

[quasi panther]

eh, whatever works ig

[marble mural]

but i know its not the cpu

as tps is stable

[ivory parrot]

How can I fix this error upon startup? Running 1.18.1 purpur https://pastecode.io/s/ynmw1rvw

[peak pulsar]

go ask for support in the purpur discord server

[quasi panther]

yeah, it's probably what eternity said: 1.18 paper hasn't optimized the clientside view distance or the chunk-send rate limiting yet, so the server sends the full view distance worth of chunks even if the client setting is much lower AND it sends all of them as fast as it can regardless of how well your client (and connection) can handle it

folks on worse PCs/connections just lag out because of it

[ivory parrot]

Doesn't seem to be a purpur related issue - seems to be Paper 1.18 related (world update? how could I force?)

[marble mural]

hm odd

[peak pulsar]

try on paper and report back

[quasi panther]

(@ iggy) I was seeing that error in vanilla 1.17

wait, no, spigot 1.17

still, not a Paper or Purpur problem

[feral spire]

hes not talking to you I think

its the person above him with nonpaper log. That will go to their support channel.

[ivory parrot]

Purpur said it was a vanilla chunk issue 👍

[peak pulsar]

ah k

[honest cape]

Anyone knows how to download java17 in wisp panel? (It isn't on the papermc tutorial)

[latent parrot]

wisp is merely a panel hosting companies may use

the java version and game that is run using it, is its own thing

[rugged rampart]

ask your host for help

[latent parrot]

^^ pretty much

[inner crypt]

Isn't wisp just ptero as a service?

So you just need to use the right egg

[peak pulsar]

Just noticed less updates for 1.18.1 recently

[languid lodge]

So?

[feral spire]

Update frequency or patch count does no have linear relationship on development process

[mortal token]

epic over math studiying moment

[timber ridge]

Hard this.

It's also December. End of year, holidays, etc. People get busy.

[peak pulsar]

Ye I know

Just was observation 🤷

[rugged rampart]

if you want more updates, PRs are always welcome :p

[peak pulsar]

No, I'm not fussing over this xD

Paper will be ready whenever its ready

[iron portal]

also as It was said already I think, the rest of the patches, wich aren't applied yet, can only be applied by one person who knows them the best, and that person is very busy.

[peak pulsar]

My coding skills aren't that amazing lol

[inner crypt]

Why did you send the message then?

[peak pulsar]

Was just observation

[peak pulsar]

Also this confirmed my thoughts

[lethal robin]

The main two things remaining to be patched just got patched anyway right (76 & 77)? Wasn't it just stuff like optimisation #7115 remaining?

[languid lodge]

We are still waiting for Leaf to awaken from his slumber

[peak pulsar]

Well, Furry convention is an important event

https://tenor.com/bbRjS.gif

[languid lodge]

not cool

Leaf is legend

[peak pulsar]

And a furry

[peak pulsar]

True though

[solemn blade]

ive lowered from 16 to 10, but i still fall through the world for a good few seconds after joining and theres still a lot of lag when respawning (cant interact with anything for first few seconds)

[versed frost]

He meant view-distance and not your render-distance.
Go modify the view-distance from 10 to 7 in your server.properties file

[solemn blade]

I know i modified the option in server.properties, 7s a bit low imo but i'll try it

[versed frost]

View distance of 10, means you are loading 21x21 chunks which is 441 chunks
With a view distance of 7 it would be 15x15 chunks which is 225 chunks
That is like half the load than what the server is normally on.

[solemn blade]

this is it on 7 when joining the server

[versed frost]

and it is worse when on view distance of 10?

[solemn blade]

ill do it on 10 now, but ive noticed the lag when respawning is little to none which is good

[versed frost]

Uhuh, it seems pretty okay to me.

[solemn blade]

ok yeah respawn lag is back when its set to 10. video is processing now bear with

[versed frost]

That would be cause the server needs some time to load everything 🙂

[solemn blade]

yeah i get that, ive just never experienced this before despite having the engine mode on 2 in 1.17. is there no option to like only apply the antixray to the first chunk around the player instead of the entire render distance?

[tropic hawk]

how does biome blending work ?

[pulsar dagger]

you mean on the client renderer?

it averages the color from neighbors in the configured radius

[tropic hawk]

anything i gotta setup server side

[pulsar dagger]

no

it's done client side

[languid lodge]

Logged on for 4 days straight at this point to farm moss and its working great

[dark nymph]

What is your server hardware?

could just be that not "up to date" or smn

[dark nymph]

I believe that there isnt if it was possible to somehow make it like that there would deffinitely be workarounds for the antixray

Mods and such exist which allows one to highlight ore deposits of desired kinds

[charred ermine]

a friend of mine had the same thing on some survival servers and it turned out to be their internet. I made it faster by disabling anti x-ray but afaik engine mode 1 works even faster; it just doesn’t hide ores next to caves and just caves in general

[solemn blade]

cant check rn but its an ancient laptop, before anyone makes assumptions i just want to say its had no problems over the past year and been much better than any free server host i could find

[solemn blade]

true, but i just want to take some precautions to discourage the temptation of just loading up a resource pack. i dont need it to be 100% fool proof

[solemn blade]

ill bear internet in mind, but engine mode 1 still displays many ore deposits for me and still provides an unfair advantage

[dark nymph]

believably better than any free host tbh. But yeah that might be why it is lagging, especially in 1.18

[solemn blade]

probably yeah, but it's definitely down to the extra stress from the xray engine

its fine without it

[dark nymph]

anyways I cant really imagine you running a "big" ish server playerbase on an old laptop, what is it for?

[solemn blade]

just an smp with friends, max 7-10 concurrent players

[dark nymph]

Deffinitely dont use anti xray in that case!

[solemn blade]

haha ive been doing this with them for over a year, the majority of them cant help themselves

[dark nymph]

If you want to check that no one has been xraying and you use plugins I suggest these 3: Coreprotect, protocollib and supervanish

[solemn blade]

ive already made a plugin to log where and when people mine valuable ores, but it isn't going to stop it happening in the first place

[dark nymph]

yeah it isnt going to stop thats true

ig you could just go vanish and spectate them and give them a punishment of sorts for xraying? Like a temporary ban for an hour or 2

[solemn blade]

they dont learn unfortunately

if they have the opportunity to do it, theyll do it

[dark nymph]

afaik there isnt a way to modify papermc's antixray but there probably is a plugin for that?

[solemn blade]

ive literally had to make a plugin to restrict some of them to only join on lunar client because they keep hacking

[solemn blade]

wdym modify?

[dark nymph]

Like change the rendering

[solemn blade]

im not aware of any plugins but ill have a quick look

[dark nymph]

https://www.spigotmc.org/resources/orebfuscator-anti-x-ray.82710/ this one, though its pretty resource heavy like all anti xray methods (in my experience)

[solemn blade]

yeah no that ones really bad imo

[dark nymph]

your server has like 20 ppl max in total right? Could you just like make rules that everyone has to agree to and if they dont obey they get the hammer?

[solemn blade]

i wish it was that simple lmao

theyll just do it anyway

[dark nymph]

To me it sounds just fair to get banned like for example a week or 2 if you pretty much broke the rules you made yourself?

or agreed to

[solemn blade]

yeah but then ive got to spend most of my time moderating them and arguing with them when they get banned

[dark nymph]

yeah that part of it sucks

[solemn blade]

plus the fact that if loads of them get banned the server will just die from lack of activity

so id rather just avoid the temptation in the first place

none of them have ill intentions, they just have no self control or awareness of where the line is

[fleet mesa]

cough

[half walrus]

Anti Xray configuration guide: https://gist.github.com/stonar96/ba18568bd91e5afd590e8038d14e245e

[fleet mesa]

never use orefuscator

[solemn blade]

yh i wasnt going to

[solemn blade]

doesnt have the config option im looking for

not much i can do at this point i think

[fleet mesa]

what are you looking for?

[solemn blade]

^

[fleet mesa]

why would you want to do that...

[solemn blade]

essentially i want a higher render distance on the server, but increasing it is causing too much lag with antixray

[fleet mesa]

hacked clients could remember where ores are

[dark nymph]

In terms of decreasing lag while still using antixray i think you just need more resources for the machine

[fleet mesa]

you might aswell just not use antixray at that point

[dark nymph]

it essentially does nothing if one wants to really xray

[solemn blade]

aight thanks

[long bane]

Can you elaborate on this? How do they see it in the first place?

[solemn blade]

if its within render distance, your client is aware of where blocks are and what they are

[long bane]

OH. Nevermind. I didn't read what the original request was

[dark nymph]

I believe hacked clients can pretty much see all the loaded chunks/blocks and a hacked client has access to that info and can just highlight those desired blocks like an ESP of sorts

[long bane]

Makes sense

I thought we were saying that abut engine mode 1 in general

[dark nymph]

oh ok

[peak pulsar]

Oh why not?

[fleet mesa]

The performance is waaay worse than paper's implementation.

[hidden cargo]

people still use orefuscator?

[dark nymph]

Some people probably do

but i dont think its mainstream in any sense

[vapid prawn]

It hides chests, the built in one doesn't

Hiding chests is also kind of expensive, on both client and server

[dark nymph]

:O

Depends on the server ig

For factions and such it might be good

[bold breach]

I ran it on a server with 10 gigs of ram and a 4 core cpu with around 40 consecutive players without issues

[dapper lotus]

ray tracing consistently for chests, etc, is expensive

doesn't matter the hardware, it's a costly thing to do, will it be fine on most servers? yea

are you gonna be pushing the mid 100s player marks with it? probs not 🤷‍♂️

[feral spire]

Every small extra work per player does add up

[mighty spade]

hey, that's a good idea, can you share it?

[feral spire]

Not a good idea coz the client can tell server they are vanilla which render the plugin useless

[mighty spade]

do the cheating clients implement such behavior though?

[flat harbor]

is one gigabyte too little for a 1.18 server

[vapid prawn]

That'll keep the idiots from cheating but if enough servers do it, your server gets more popular, or someone who knows how to program wants to cheat it's barely a speedbump

[dapper lotus]

Well, most of the hack clients just pretend they're vanilla, but, otherwise ^

[feral spire]

Some do

[dapper lotus]

It's literally a trivial bit of code to write to faux that shit

[feral spire]

It’s an arm race pretty much

[mighty spade]

true, but most skiddies don't into code

[vapid prawn]

Pretending to be vanilla is one thing, pretending to be lunar client and faking whatever extra data it lets you steal from your users is another

[mighty spade]

much less actually compiling anything

[feral spire]

Someone will make it for them if enough of people ask

[dapper lotus]

No, but the cheat clients are literally race to the top

[vapid prawn]

But I think I have heard of a cheat client that pretends to be a stock lunar

[mighty spade]

what about forcing certain players to use a custom client which forks lunar

and only changes the client string to something custom

[feral spire]

Ugh

[dapper lotus]

it's pointless cat and mouse

[mighty spade]

it's not pretty, but it would work

[dapper lotus]

You invest more effort into fucking around with this shit

scare off a large % of your player base

[vapid prawn]

It would work until one person who knows how to code wants to cheat on your server

[dapper lotus]

and basically make new player intake a "fuck that"

[mighty spade]

keep in mind, only "graylisted" players would be forced to use this custom client

[vapid prawn]

And if you don't make it custom enough a generic client will grow the ability to set a version string or whatever and then everyone can do it

[feral spire]

Put your effort into making a community that discourage cheats

[mighty spade]

players who have already been caught cheating

[forest bridge]

Custom clients won't really help, since you can just inject things directly into JVM.

[mighty spade]

that does assume a certain level of know-how though

[forest bridge]

Well, that's how it ended here in Russia tho, there are cheats that can be injected this way public available.

[vapid prawn]

Not as much as you'd think, this is a game to the people who write these tools

They love showing off their latest tech and eventually either they share it with everyone or someone copies it and shares it to make the first person/group less special

[mighty spade]

alright, ill never accept "give up" as an answer to preventing cheating though

[forest bridge]

Server anticheats exist precisely because of that.

Of course not everything can be detected

[dapper lotus]

Hey, you are welcome to engage and cat and mouse

We're not gonna stop you

[forest bridge]

But there are a lot of things that you can do still.

[feral spire]

Not give up. Just invest your time into making your community better than to prevent people cheating

[vapid prawn]

Down that path lies rootkits

[dapper lotus]

We're just saying that the specific way you're already tryna go about it has been defeated already

You, er

[mighty spade]

how?

[dapper lotus]

tryna rip a mix tape der amaranth?

[forest bridge]

The issue is that to be able to make server anticheat catch more things we need cooperation from mojang, so they add more data into protocol (like for example pressed keys on movement).

[vapid prawn]

Nah I disabled autoplay

[versed veldt]

is #68 deemed stable?

[feral spire]

It’s gonna be hard to explain but take my community for example. We originated from Dota and now just play whatever game server decide to “play together” and we gain member among the way @mighty spade most core members are here for 4 years or longer

[dapper lotus]

shit seems stable

idk what 68 is, but, if it's not the latest like, use latest

68

literally the first 1.18.1 build

[feral spire]

Once you have a community that everyone knows each other suddenly no one would cheat because it’s discouraged @mighty spade

[dapper lotus]

plz...

[vapid prawn]

Basically he means the community is small enough public shaming works

[mighty spade]

what about new players?

[vapid prawn]

New players don't happen often in a group like that

[dapper lotus]

new players are new players

if they're blatently cheating they're often reported fast or booted even by the shittest of anti cheats

[vapid prawn]

It's a friend group, not a bunch of people to buy your microtransactions

[dapper lotus]

I used NCP for detection just as like a "staff plz check", most actionable reports came from players

[vapid prawn]

The key is to have something like NCP only block the most egregious stuff while reporting the rest to staff

People will try a tiny cheat then if that works they'll escalate to see what they can get away with

You're logging the whole way as they make it more and more obvious what they're doing

[feral spire]

This is how it works generally. stuff like ore announcer will tell you instantly if someone is Xraying etc

[peak pulsar]

Had my first server hang up since switching to 1.18.1, a mod switched to spectator mode with only 2 players online and then this happened. https://pastebin.com/nbte5k14

[dapper lotus]

probs a metric fuckton of item entities somewhere

[feral spire]

the good o' dropping a bunch of fucking shit on the ground or do you have someone making a farm that has shit tone of waste on the ground?

[peak pulsar]

So, I don't have to bother you guys in the future, what in the log gives it away? I was not online at the time and the mod isn't giving me any clues.

[dapper lotus]

the fact that it was ItemEntity ticks in each of the snapshots

each of them with a different entity UUID

[wide talon]

Is there a stable release yet or is it experimental?

[ivory parrot]

Hi! So I have an error using the Purpur jar, however it's a vanilla bug and I was wondering if I could get support for it here?

I can swap over to Paper if required.

[vapid prawn]

If it's a vanilla bug I don't know what you expect us to do in chat 😛

[feral spire]

Mojang moment

[vapid prawn]

Reproduce with Paper, file an issue on github, reproduce with vanilla, file one on mojira

[ivory parrot]

Okay

[somber jewel]

I want to use paper 1.18.1, but is it stable? and sure?

[ivory parrot]

I was more wondering if you could diagnose it better

[feral spire]

You can use it. Make frequent backup @somber jewel

Switch to paper and see? @ivory parrot

[vapid prawn]

Yeah I guess if you're on paper you can try

[ivory parrot]

Gonna do that rn, but current error to see if it changes when I swap:
https://mclo.gs/USZMJcY

[somber jewel]

Do you know of any problem that generates when using paper 1.18.1 that damages the world or something known?

[vapid prawn]

Oh you can ignore those

Unless you're seeing world corruption those MapLike errors are just Mojang doing a poor job of writing their upgrade tooling

It's not an actual bug afaik, just spammy

I mean, the actual bug would probably be that it logs that, not that the thing it's looking for isn't found

[feral spire]

Not really.

[vapid prawn]

It should only happen once per chunk at most I think

And only the first time it's loaded

More likely you'll probably see it a few dozen times for your whole world

[dapper lotus]

We know the issue

we've complained to mojang about the issue

and they still keep introducing the issue...

it's data which isn't needed post world-gen iirc

they just don't have a mechanism for cleaning it or just not saving it

[vapid prawn]

Paper tried to silence it in 1.17 but apparently did so by trying to make the error go away instead of just shutting it up which then broke things?

[dapper lotus]

leaf actually upgraded the data or something, iirc

according to mojang that data can just be safely yeeted however which is why they kinda don't care and it keeps slipping

[somber jewel]

I use paper 1.17.1, and I want to update to paper 1.18.1, but I'm afraid that something will get corrupted.

[dapper lotus]

theres 50k+ servers running paper 1.18+

[feral spire]

Should be fine. Make a backup

[ivory parrot]

Paper errors
https://mclo.gs/VQj0epC

[ivory parrot]

Ohhh. So this is the thing then