#❗-1-18-experimental
1 messages · Page 19 of 1
not true, at less in where i lived....
ofc, for devs? That's literally nothing
They live in a data center 
if you want the capacity in which a proper dedi can handle, you're either getting cheap gigabit eithernet or not considering costs properly
That option only open if you own the physical server
wat
I wish it was standard to have 8g. Ram

build one just cost ~$450 and can use for 3 year+
(VPS cost $240 per month
🚌
Yeah big future plan
why tf would you ever spend 240$ a month on a vps
if you want to burn money you can just send it to me instead
What hosting lol
Keep off topic out of topic channels
are the experimental versions good enough for upgrading an existing world from 1.16 to 1.18?
This might already be common knowledge, but I figure I'd put it out there in case someone else runs into it. The newer java 17 required breaks backwards compatibility with some older docker versions due to additional security. Also, the openjdk:17-alpine docker image doesn't use the release version of java 17, which causes 1.18 to crash.
Wasn't addressed to you
oh
Yes, you can do so... but take backups first
And also schedule regular backups
And keep the server jar as up to date as possible
i have regular backups
how do i update a server tho
do i just replace the jar and start it?
First of all, test out the 1.18 jar locally on your PC with the plugins you want to use (to see if they work) then update your jar on the server and make sure that it's using Java17
(1.18 requires Java17)
thanks 👍
Has anyone else experienced never Ending Raids?
On my Server is a Raid that is still on Level one, even though 400+ Pillagers were killed already...
Nvm we were killing Outpost Pillagers...
my 8gb vps is
10 dollors
a month
btw without pregen my server is running alr for now
$10?!!
which provider? mostly I found almost > $30 for 8GB RAM
Hi! Simulated distance is ticking distance, right? And view-distance is non-ticking? And is it possible to test it in game?
okay, so this dispanser will work only close to me, right?
Also, make sure the value for View Distance either isn't set in spigot.yml or changed as well.
and what now? I put water in my dispenser and nothing happened
inside the veiw distance and withing ticking distance no change
outside the ticking distance no water flow
or get placed
nope, the water is IN my dispenser and that's it.
whut
it is basically no tick better distance
yeah
it will not get placed in
the no tick zone
I do not understand. I am close to this dispenser. I am clicking on it...
use redstone and try to active it from from outside the notick zone
isnt there something more simple?
maybe shoot arrows
Learn about Entity Activation Range 2.0 (Wake up Inactive): https://github.com/PaperMC/Paper/commit/e0ea2e0e147875b5b0eeab747d19fa00a5e19ef2
no
maybe armor-stands will dissapear when I will be in non-tick zone? I noticed that armor stands disapearing very fast when I am further away
maybe redstone-clock?
I will see redstone-clock from very long distance and I will see if it is working or not
Unfortunatelly the redstone clock is working veeeery away from me. So the "non-ticking" distance is a myth
Does starlight is implemented in paper?
Yes
someone know why this is happening?
https://github.com/PaperMC/Paper/issues/7054 known issue
time4vps
i got spawned inside a mountain or something even in two different seeds
will server performance improve when stable releases come out
If people die server dies too nice
So um changing to engine mode 1 led to a 50 -70% reduction in bandwidth usage compared to engine mode 2. If you guys were interested.
message.txt by @next owl: https://p.sulu.me/H2O9dJ
log
This is the 1.18 channel
sorry habbit lol
wait
i am on 1.18
that old log
sryy
leme send new
sorry for the ping, but do you perhaps also have a repo for that? xd
ye
i will do it later
Nope. 🛌
it didnt log it
Ah well, try to get a log.
howw
Next time it crashes?
ok
outside of the JVM being killed by the OS
it's pretty much impossible for it to die with 0 logs
log file, crash report, hs_err file, etc
^
Check to make sure your on the right server, because 1.18 still should be providing logs.
I mean, if you started your server after it crashed, then it will be in the last gzip file
Like, they're saved with the timestamp in them
or, well, the date
is it possible for the server to crash so badly that it doesn't release locks on the world folder?
Once when starting one it complained that it couldn't lock the world dir, and another instance was not currently running
If you’re absolutely sure, you can delete the “session.lock” file. I’ve had it happen to me before, but only one time in my multiple years of running servers.
hey folks I'm running paper 1.18-57 on my server and ive noticed that the transition from the overworld to nether and back is slow, to the point the client locks up and takes about 5-8 seconds to complete. i thought maybe it was a plugin i was running so i created a new world with just paper and the behavior seems to be the same. the hardware is home hosted and is plenty strong to run a server, was running a 1.17 world without issue. im curious if this is known behavior or if theres any remedy for it at this time?
Copied from other channel, Decrease the chunk send rate in the “chunk-loading” config section.
Problem is 1.18 chunk sending totally clogs up the network, so you need to slow it down a bit.
roger, i assume thats just in paper.yml somewhere then, ya?
yea
just found it
min-load-radius: 2
max-concurrent-sends: 2
autoconfig-send-distance: true
target-player-chunk-send-rate: 100.0
global-max-chunk-send-rate: -1.0
enable-frustum-priority: false
global-max-chunk-load-rate: 300.0
player-max-concurrent-loads: 4.0
global-max-concurrent-loads: 500.0```
this section i would assume?
the "target-player-chund-send-rate" is the one i should turn down?
oh thats actually really weird, that chunk-loading section was missing from my brand new 1.18-57 world for testing but exists in my actual server..not sure why that would be the case. i was going nuts trying to find those values to modify for testing
got it
that config won't currently do anything in 1.18, it hasn't been implemented yet
If your server just exits like that with no log, likely something external is killing it. How are you hosting the server?
i got the log
yes but it just ends, if it exited normally it would have logged more
ye
it crashed
yes, but for a lot of types of crashes more will be logged
this isn't it crashing, more than likely something external has killed it
it crases
ah I see, ok. How are you hosting it?
alright, so then i guess for the time being just live with it and eventually there will be some optimization options for the behavior? its not crashing the server of anything, more just a quality of life thing
yeah that will be added back in the future
roger, ty
*1.18.1 RELEASE CANDIDATE 2
MC-245010 - Sometimes certain chunks will never load
message.txt by @next owl: https://p.sulu.me/9XG09f
do you have a panel installed i.e. ptero, or are you just using screen/tmux/whatever to run it?
screen
something external is killing it, check the output of dmesg -T | grep -i 'killed process'
how do i check tha
im kinda new to this
just run that command, it will print an output
When is it going to be fully ready for 1.18?
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.
Oh
there aren't any serious bugs
no, what is blocking a proper release is that it is not finished, as in not everything from 1.17 has yet been implemented for 1.18
not that it is unusable and buggy for small servers
Out of curiosity, is there a place to view which features haven't been carried over yet?
this has most of the big ones ^
Thanks so much!
big 1.18-specific bugs left are https://github.com/PaperMC/Paper/issues/7054 and https://github.com/PaperMC/Paper/issues/7043
although the second is effectively just https://github.com/PaperMC/Paper/issues/6779
are 261 issues actually issues?
there are a lot that are feature requests/small things
and others are effectively just sitting there, never to actually be fixed/worked on such as a lot of aikar's tracking issues
might work to at some point move feature requests over to discussions, people like essx/discord's api docs have done that and it seems to work decently well for them
although idk
is this normal?
Pastebin.com is the number one paste tool since 2002. Pastebin is a website where you can store text online for a set period of time.
no
f
so the experimental build should be fine to use for my friends server so long as we make fairly frequent backups right?
it's never that bad
but yes backups are important
cool beans, ty
yes for friends its plenty enough
but for public not yet
awesome
?
Sorry
My cat jumped on my keyboard 😭
I was reading the channel
Didn't mean to send this
I'm getting errors in my logs what I've not seen before:
8.12 19:13:58 [Server] ERROR No key placement in MapLike[{config:{state_provider:{state:{Name:"minecraft:hay_block",Properties:{axis:"y"}},type:"minecraft:rotated_block_provider"}},type:"minecraft:block_pile"}]; No key feature in MapLike[{config:{state_provider:{state:{Name:"minecraft:hay_block",Properties:{axis:"y"}},type:"minecraft:rotated_block_provider"}},type:"minecraft:block_pile"}]
Is this paper related, or vanilla?
ok...cool. Thanks
Has anyone else had issues with sugar growing forever on the experimental builds?
What do you mean by growing forever? Took forever to grow or growing non stop until 256?
I'm not sure if its non stop, but higher than 3 tall.
There is a Config option in spigot.yml. Make sure no one touches it
Otherwise remove your plugin and see
cane-modifier in the spigot config appears to be default at 100. I'll check plugins.

How do I increase the view distance of players? setting view-distance or simulation-distance in the server config doesn't do anything
Hello, call me dummy but ... the client viewdistance now syncs with server?
so that I can see players over more chunks
yeah, client now matches distance of the server
Client can only get as much as you give them. Just ensure you don’t misconfig anything here
in a post on spigotmc somebody says I have to increase the entity-tracking-range-players in the spigot.yml, but the setting doesn't exist anymore. What's the setting in the 1.18?
This is actually nice
@proud sleet you can read this walk through
That’s pretty much all you ever need to change to keep server lag free
Paper has more options than spigot if you come from there 
It seems the patch to handle the new simulation distance is taking a while for paper. Is there a reason? Is it a wonky system or?
I’m getting this from the to do list that I saw posted here
the person who is aware of that system the most has other things to do
Gotcha
I want to add custom structure to 1.18 is there any good way to do it?
(custom dungeon)
I asked in #paper-dev but I am not sure if I will get helped here with 1.18.
moreso that you've waited all of 2 minutes
dev is the proper channel for dev related things
no-tick-view-distance:
Works in 1.18?
see
dont forget the range settings in spigot config!
Donating works better than praying
https://opencollective.com/papermc
😂
I'm only a teenager mate sorry
Does any1 at all have like any idea in any way shape or form how long it may take?
I don't need an exact time
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.
yes when you turn 13
possibly 😄
You can look at the update tracking issue if you want to see whats still missing for a full update: https://github.com/PaperMC/Paper/issues/6136#issuecomment-982297893
Thank you
Is everything ticked the done things?
yes, thats how a checklist usually works.
Just know that this isn't really an indication on how long it will take, just how much is done already
Do you guys have any idea about the current state of paper 1.18? Should it be save to move a 1.18 survival server tot paper from vanilla?
if you make regular backups it should be fine
Are there any known big issues at the moment?
https://github.com/PaperMC/Paper/issues?q=is%3Aopen+is%3Aissue+label%3A"version%3A+1.18" apart from these, not that we're aware of
My Server runs fine without any problems since #32
Ok thank you for the info
Can someone link me to the Paper bug tracker?
Still having this issue. Anyone have a solution?
think somebody had an idea of what it was but it's basically on their todo list
👍 good to hear
you're just somebody I used to know
looks like the server really doesnt like generating high value chunks
It works fine with low values but when i set it to something like 10.000.000 it shits itself
[23:38:19 FATAL]: Error executing task on Chunk source main thread executor for minecraft:overworld
net.minecraft.ReportedException: Exception generating new chunk
at net.minecraft.server.level.ChunkMap.lambda$scheduleChunkGeneration$34(ChunkMap.java:1264) ~[paper-1.18.jar:git-Paper-60]
at net.minecraft.util.thread.BlockableEventLoop.doRunTask(BlockableEventLoop.java:149) ~[?:?]
at net.minecraft.server.level.ServerChunkCache$MainThreadExecutor.doRunTask(ServerChunkCache.java:1248) ~[?:?]
at net.minecraft.util.thread.BlockableEventLoop.pollTask(BlockableEventLoop.java:122) ~[?:?]
at net.minecraft.server.level.ServerChunkCache$MainThreadExecutor.pollTask(ServerChunkCache.java:1260) ~[?:?]
at net.minecraft.server.level.ServerChunkCache.pollTask(ServerChunkCache.java:831) ~[?:?]
at net.minecraft.server.MinecraftServer.pollTaskInternal(MinecraftServer.java:1401) ~[paper-1.18.jar:git-Paper-60]
at net.minecraft.server.MinecraftServer.pollTask(MinecraftServer.java:1384) ~[paper-1.18.jar:git-Paper-60]
at net.minecraft.util.thread.BlockableEventLoop.managedBlock(BlockableEventLoop.java:132) ~[?:?]
at net.minecraft.server.MinecraftServer.waitUntilNextTick(MinecraftServer.java:1362) ~[paper-1.18.jar:git-Paper-60]
at net.minecraft.server.MinecraftServer.runServer(MinecraftServer.java:1268) ~[paper-1.18.jar:git-Paper-60]
at net.minecraft.server.MinecraftServer.lambda$spin$0(MinecraftServer.java:317) ~[paper-1.18.jar:git-Paper-60]
at java.lang.Thread.run(Thread.java:833) ~[?:?]
message.txt by @fierce ravine: https://p.sulu.me/jGKZG9
if it's a chunk pos
the wb command allows it though
vanilla worlborder is 30 mil x 30 mil
oh, chunkpos, nvm
well idk
they're locations though, not chunk pos
my plugin respects worldborder so any location it finds is within that bounds
you're sure teleporting there doesn't cause it too?
like, try using the same seed in SP and see what happens when you go there
lemme test
lol
when i join i'm at that location
after the server crashed from the chunk gen error
whats seed/pos
[23:43:37 INFO]: DarkEyeDragon[/127.0.0.1:61246] logged in with entity id 248 at ([world]36061.5, 63.0, -354918.5)
[23:45:53 INFO]: DarkEyeDragon[/127.0.0.1:61264] logged in with entity id 234 at ([world]171.0, 64.95357628820153, 966.0)
seed -6340746453277938574
36061.5, 63.0, -354918.5?
i cant really validate if those were the original coordinates i rtp'd to though
wait what
you spawn in random locations when you join now?
if you don't have a spawn set it picks one around the world spawn iirc
omg im stupid LOL i have a test plugin that teleports me randomly on join 
no error tping here
yeah disregard the ending up on the location on re-joining
works fine when teleporting regularly
i really cant tell if its my plugin or the server fucking up at this point
I nuked my .gradle just now and I'm still having the same error :\
I'll try that as well. I nuked my global .gradle 😅
just need to nuke paperweight caches basically, which are located in .gradle/caches
(of your project)
A new issue has shown up after nuking my .gradle. When editing in my project, everything imports fine, but when building all references to bukkit/spigot/paper related stuff results in a "cannot find symbol" error. For example:
error: cannot find symbol
NETHERITE("claim.power_type.5", Material.NETHERITE_BLOCK, 4); // 144x144, 72 radius
^
symbol: variable NETHERITE_BLOCK
location: class Material
is your java target 17
thats your toolchain
it's material
yes
their spigot/paper api is outdated
they are using userdev
it also does it for non-Material stuff
so that would happen if they add an api dep manually
error: cannot find symbol
PersistentDataContainer data = chunk.getPersistentDataContainer();
^
symbol: method getPersistentDataContainer()
location: variable chunk of type Chunk
wait what
we have a manual api dep and use paperweight tho?
heck, we literally have spigot-api, paper-api, and userdev deps
hey, i have the problem that the colour of the name in the chat of the people that has a team doesnt show, it shows as there werent in any team, this happens from today, when i updated to paper
er, you really shouldn't
at least not in your adapter project using paperweight
enable use-vanilla-world-scoreboard-name-coloring in paper.yml
the server added by paperweight depends on the api on it's own
but it's not transitive to other projects depending on the userdev project
and our adapter is an entirely separate module
yeah that makes sense
How stable is the experimental build atm?
Experimental
How experimental is the stable build
How built are the experimental stable's?
How
H
damn it beat me to it
how exablimerintal is the bastle bdiuld
... what?
(It’s a joke from above)
How stable is PaperHQ? 😛
have you ever tried to build a sandcastle on a moving car with dry sand
yes, actually
Well, it only gets more stable if people help test it.
Just make sure to take backups frequently, and you’ll be fine!
the trick is to use layers of paper towel or toilet paper interspersed between layers of sand as a lattice to help the dry sand keep it's shape
I mentioned paper, so it's on topic, right?? 😂
"yes"
Forget about sandcastles on moving cars and paper -- I'm just curious about @peak pulsar 's bio...
anyone know about when 1.18 stable will be released?
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.
But do your part, make sure ur testing and reporting and issues you encounter. 🙂
It helps everyone.
Owen started typing before NeoMan even finished
Yes, but let’s keep this 1.18 convo. x)
Sentences are a thing, and this is not the right channel for off-topic discussion.
it's because people think I'm a furry bc of the pfp
my pfp is the transfem fox girl Linux mascot that never caught on
instead we have a dumbass penguin
Kali dragon ftw
ew

How would you determine if it’s stable or not ?
It’s stable when paper officially says it’s stable
But, it can only become that way if people help test the software.
Do your part. 🙂
opin sours
Already ahead of you just wondering is all.
Been working on a testing server but
I’m sure you know the players of servers. Don’t understand what it’s like on the backend haha
run a separate instance on port 25566
Oh of course, no worries!
Demanding for updates and I’m sitting here telling them it’s not stable lol
That’s the plan
To limit system resources, set a world border that's pretty small. Then players can test out 1.18 and just know it's temporary
How would I set that up
world-border plugin and using /wb trim
would it be smart to update my server from vanilla 1.18 to paper 1.18 expirimental or should I just wait for a stable build?
I’d say go for it! Just please, make backups.
It helps to have all the help for testing updates. 🙂
How often should I make backups
Just note, if something breaks to report it.
Like, daily.
using rsnapshot or similar they don't take too much space
Ok. I was just wondering becuase I started said server last saturday and me and a bunch of other people have already made a bunch of progress. How often should I backup?
^ Whenever you build something worth saving
Ok thanks. I'll make sure to put an annoucement out to the players to make sure they are comfortable with it.
That’s a good idea, but do note that things seem to be pretty good currently.
No problem, just make sure to report any issues. And feel free to come here if you have any questions or concerns.
How would I setup rsnapshot?
a tool for backing up your data using rsync (if you want to get help, use https://lists.sourceforge.net/lists/listinfo/rsnapshot-discuss) - GitHub - rsnapshot/rsnapshot: a tool for backing up your ...
I personally just compress it to a .zip file and put .DATE-BACKUP at the end and that's that
Will do. Thanks!
See I've always wondered what a backup would do with chunks that are loaded in while the backuo is running
Would it corrupt region or chunk files?
I've been doing that since 2013 and never had a chunk corruption in my backup
As long as ur just reading it should be fine?
Okay sounds cool
There’s a chance some may be outdated unless you run /save-all before backing up
But otherwise it should be totally fine.
I'm sure I could put something like that in an auto restart plugin.
As a command to run or something
How often should I do a backup regardless of expermanrtal or not ? I won’t lye I have been lazy and haven’t set up a backup system yet
the server saves automatically every 5 mins by default
Daily
Ok
You could!
Because if you have a cron job that runs after it saves everything and restarts
Or you could run a script or something
Then it would have basically the most up to date stuff
You can write a .sh script and run it every X hours/days with a crontab
It saves before restarting right
It depends on how much work you're willing to lose. If you're willing to lose 6 hours of work, do a backup every 6 hours. Willing to lose more or less work, adjust the backup interval accordingly.
You can also write a script to automatically delete old ones after maybe 10 or so backups. I prefer backing up manually tho tbh
I need to setup the second drive in the container for backups
So backups are not done to the main ssd
I got another SSD thats offsite that I can use
I went to SSDs and haven't gone back
Loading lots of small files is so much better
Something like this, I think: find '.' -name "*.sql" -maxdepth 1 -ctime +10d -delete
I used that to delete old *.sql backups
I don’t have offsite backups either 
Honestly while waiting to move to 1.18 I might just move stuff to velocity and get tcp shield, then get a backup system in place.
1.18 is jank
I just backup to an external HDD. In case of a fire or while on vacation or something I can just unplug it and have all my backups safe
Like I got it on a test server
And it's just slow lol
And anyone trying to use chunky to load chunks is having memory leaks
And it's slower for clients on their end
Because more crap to load
But I'm happy that they finally have in game audio device switching
Slow ... well I’m gonna see how well it runs with 1gb of ram oh boy
that's mostly mojang, it has to load double the amount in one chunk than before 1.18
1gb ram jeez
My mate is gonna like that, headset dies and it breaks
the memory leak issue seems to be primarily with upgrading worlds
I got the main server running 4
Yes but literally upgrading worlds is what everyone wants to do
It's cool to get an extra 64 blocks
Especially for an SMP
Now a factions server it's fine if you reset that
the extra 64 blocks is partly why it loads slow on clients and why the chunks take longer to pre-gen
Ya but it's still got a memory leak converting
I wonder is this easy to reproduce? Someone should memory dump and analyze it
I haven't found this issue
The MOSS discord #chunky channel is really hopping with the memory leaks
looked over a heap dump as well, couldn't tell you where all the protochunks are coming from
I’m gonna be Pruening the world file prior to porting to 1.18
regardless if you don't run chunky over the whole world the issue won't present itself as badly
just upgrade as chunks are loaded, there is no need to force upgrade or run chunky over the world
Some dude there said he went to spigot and ran it then went back to paper
Ah dang
What about minecart chuck loaders?
Well dfu yea is much faster thanks to leaf ❤️
I world-border trim everything outside like 1500, then regenerate it with 1.18
I didn’t actually realize how many people pregenned chunks, ig it’s much more important now with the serious slow down in 1.18 generation tho
I just like to have everything pregenned because of dynmap
Definitely helps w/ performance for sure
I mean I guess I have okay specs for performance so I can go without
A 5900x and as much RAM per world as I want
Well generation should be done async so it’s not like you’d have heavy lag. It’s just resource intensive I’d say?
Unless I’m missing something here
I try to keep alloted ram to the minimum for garbage collection
Yeah just resource intensive that's all
Well the current issue is that chunky and world gen has memory leaks
On old converted worlds
But the issues are of because some people haven't reported issues
Well it would help to be able to have those reported… x)
I’m running 4770k and 16gb ram
Nice
That’s not just for minecraft so I wish I had all of the ram for mc
Well after covid started, I said screw it and I built a dedicated Minecraft server
does this happen to any of you
server online but discord srv not outputing and players cant join
restart fixes it
Run it literally for fun
Hmm well my 1.18 test server has discord srv working I believe, let me check
We had a modded server setup but it was lagging both the host and the server so we decided to build a dedicated rig. Some spare parts from my work at the time later and tangy was born
@next owl 1.18 DiscordSRV is working perfectly for me. The fact that players can't join means it may be something going on w/ the network connection to Minecraft on startup?
no no discord srv
not a problem
server not taking any input
or letting players in
restarting fixes everything
So you only have the problem while booting from a shutdown?
Oh lol sorry lol
So about the timeout thing, the only solution is to wait until 1.18.1 release ? 🥲
Prolly? I though paper had a patch for that, unless that’s handled client side
(I may be thinking of wrong thing)
no after like 40 mins
for server running
when everyone left
then no one could join
till i restarted server
I've reported it but there aren't much things to do since the crash logs are not useful to fix the issue
overlay which weekday it is
and each contributor's personal schedule (lots of uni finals this week)
there's a bunch of empty checkboxes on https://github.com/PaperMC/Paper/issues/6136 still
which plugin was that?
What are the current issues... oh i see that lol.
i guess my next question is can I just switch my spigot 1.18 to paper 1.18? I have been told plugins will still work but like.. is this true? do datapacks still work?
Yes they do.
is there anything like crazy wrong with 1.18 version of paper?
any reason not to switch?
That's up to you. Look at the issue tracker and if you are comfortable, go ahead and switch.
well like i dont know this kind of stuff super well. Like do you think it's playable for people on a server, espically versus 1.18 spigot
I mean at this point all 1.18 related things should be considered unstable (and should continue to be until at least a few weeks past 1.18.1 release), however there are no known major issues with Paper 1.18 yet
dope i will switch then..
just make sure you have regular backups
yessir
@silent compass I would say don't switch yet. No need to rush. You can also watcher their main TODO. https://github.com/PaperMC/Paper/issues/6136
meh i already did it. I could always switch back. I just heard paper > spigot
spigot is for ignorant people
@silent compass No switch backs I don't think
it's alright
For me the temptation is Fabric vs Paper
Well also @peak pulsar i havnt made a server in like 4 years. spigot was the way to go in 2016
Yeah you can switch back to spigot, but like why
Basically optimised for MMO or highly accurate for technical play
i tried fabric awful performance compared to paper and lack of features to
I personally have paper
do you guys know of any warden plugins/datapacks btw?
fabric vs paper is like apples vs hippopotamus
i mean a basic thing. like him walking and hitting. that was literally in the game already
And they work well enough in 1.17.1
i just wanna have the warden
i have a general quesiton. Do Datapacks effect servers?
apples definitely taste better
Datapacks will work on servers, yes. However, they almost always come with huge performance impacts, due to how datapacks work
that was my question. So like, how bad is it to have them?
Anything a datapack can do, a plugin can also do, so try to find a plugin alternative
are plugins just better?
correct
like lets say a silktouch datapack vs a silktouch plugin
(silktouch spawner basic thing)
not sure what that is, but the plugin will perform better
https://www.planetminecraft.com/data-pack/silk-touch-spawners-datapack-1-13/ vs https://www.spigotmc.org/resources/mineablespawners-1-8-1-18-silkspawners-alternative.59921/
i mean, it really depends on the datapack, but i'd say plugins are just easier to modify and more modular
and there's almost never a reason to use a datapack over a plugin
also plugins can have configs, permissions, etc etc
like to me this is such a small datapack idk why i would need ap lguin
like i dont need perms for this it's just a simple thing
is there any reason to use the datapack over the plugin?
personally: i dont wanna setup perms i have 0 need.
yeah looking in the datapack, it will perform like shit compared to a plugin
and also ALL THIS datapack does is make it so enchantments of silktouch mines the thing.
Really though? like compare the two
One player sleep datapack is better than plugin.
Change my mind
why do you need a datapack for that
Oh
just change the sleep percentage
Wrong message sorry
The datapack has a bunch of crap that has to run every tick due to datapack limitations, a plugin can do stuff on events i.e. when a spawner is broken, making it infinitely more lightweight
@modern lintel /gamerule playersSleepingPercentage 1
hmm
that is actually like useful info without any bias to me. So like hmm.
oh you've gotta be kidding me.... ~600 commands for every online player every tick https://f.u11.io/AmJMBP
Sadly I don't see a better way to do it, datapack limitations strike again
over-engineered
Oh, thats just awful
Nope it's not, it's just the same command 591 times with just a couple of values tweaked each time
Datapacks are extremely limited in functionality
Doing anything interesting is already a hack
that is why i chose this specific thing. Silk touch spawners seems like something that doesnt need al ot
for plugins, it's not. Probably like 30 lines of code
hmm
Except it requires reading NBT and conditionally executing a command to drop a specific item (also with specific NBT) at a specific location. And you must do all of this without if statements.
There is literally 0 brain power in the custom function stuff and pretty much anything "comlex" you do is basically just going to result into a complete utter horror show
so i have a question then. and sorry this is off-topic. In the permissions.yml can i just throw
"mineablespawners.mine"
in there like on it's own
many basic things completely devolve into a horror show of complexity fast because of how limited it is
use a permissions plugin.
LP ❤️
no i am sayin can i just do that
(Because, no, you can't, it doesn't work like that)
ah
oof i hate permission plugins dude it's just so much extra nonsense i dont wanna deal with
everything's extra nonsense to you it seems
LP's web editor makes it pretty painless
I mean
well no it's not @dark wigeon . If i want one permission done why am i downloading a whole plugin?
you are more than welcome to use permissions.yml, I just refuse to provide support for it as many people think that they're gonna avoid so many headaches using it and don't understand basic yaml or reading :L
because I garuntee you'll need more permissions
not like it affects performance anyways
the bukkit wiki has guides on how to do it, it is doable for basic stuff, but, otherwise, glhf
It's effectively impossible to run a server without a permissions plugin unless you have literally no plugins
well i dont know that. like to me it's "more files = more effort for the server"
I mean servers require effort so...
if it's literally one perm
You slap the plugin in the folder
start it up, run a single command to add it to the default group, and skip off like it's nothing
no duh lol. i mean the server itself is doing more effort for it. like that is how it seems to me like it will run bad.
Nope, LP does everything async so it won't affect performance whatsoever
but in most cases yeah you're right, you don't want to add hundreds of plugins just cause
hi, i use https://github.com/dhouck/anti-creeper-grief but its only working with my "standard" dimensions (overworld,nether,end) .. not with my custiom dimensions, even if its enabled in level.dat ... you think its caused by paper or should i check datapack?
i would appreciate it kindly if you told me how this works. Like what is this singular command to do this type of thing. My server isn't a plugin intensive thing I just only need like very very few of these. No pay to win or ranks or anything for me it's just default and op
/lp group default permission set <the permission> if using luckperms will add <the permission> to everyone
^ But I really suggest skimming the "getting started" guide next time you're on the pooper or something
LP is a pretty powerful plugin and, "I only need it for one perm" is often the start of 20 more and then hey I need somebody to mod...
(I need to poke turbo about making the links case insensitive again)
still wondering what that one command would have been
Manchas
...
Don't mention
unless you were reffering to lp.
ah
for perms.yml I refuse to offer support for that file as people generally aim to shoot themselves in the foot and I refuse to go down with them now
i guess? It's whatever. So with LP do i need to readd all the other permissions other plugins give me by default?
or just new ones
By default LP will respect plugin-assigned default permissions
alrighty
So lol
omg @rocky nest wanna know the difference between datapack and plugin for that one datapack?
half a gigabyte that datapack was using
of ram*
now i wonder if i can repalce this lol: https://www.planetminecraft.com/data-pack/amplified-nether-1-18/ though i dobut it's doing too much damage
Like wow, that was drastic change lol
I don't think worldgen datapacks are as bad as most others, but don't quote me on that
i wouldnt think so either since it would (hopefully) do it on generation
yeah poking around it appears to, I have almost no experience reading datapacks though
This one is cool: https://www.planetminecraft.com/data-pack/technical-enchant-for-1-17/ i wonder if it would poop the game.
I think the only reason the other one was so drastic is because i was running so many lines per second
that one almost for sure will cause issues as well
there's plenty of proper plugins for custom enchants
i like dis one though 😦 . Kidding, well i do. But yes this was a valuable conversation for me
Mojang offer 0 promises that you'll be able to upgrade your world when using a datapack, especially with how volatile they are, etc
that is quiet true
This take my atention, what is?
ignore them, they're a moron
has prob be asked a million timez already.. but what do u guys think, how stable is the release rn?
only 3-4 issues left on the github for 1.18, i think it's starting to be stable
lol
It’s still experimental. So it’s not as stable as a stable build.
Use at your own risk. Take backups. Report bugs.
I used the vanilla tweaks datapacks while upgrading, threw some errors at first but they seem to be fine now. completely anecdotal though
with 4 public bug left, i think we are past the exp build, it's more like a beta at this point
yeh yeh but experimental releases can be more or less stable as well
There are still patches that are not updated yet.
And those patches have some complex shit in them. So it very likely will get more unstable before a stable build is released.
sadge
can you explain a bit more? like paper patch or minecraft patch (like 1.18.1)
Paper is built by patching Minecraft.
There are a bunch of patches that do different stuff. Features, bug fixes, performance enhancements, etc. The patches are what makes Paper Paper.
When there’s an update the team goes through and updates all the patches for the new version. They aren’t all done being updated yet.
Once those are updated and the bugs ironed out then it’ll be more stable. But the patches that haven’t been updated yet are patches you want. They’re gonna make stuff better.
How many are left?
You can look at what is done and what not here: https://github.com/PaperMC/Paper/issues/6136
Damm they are already down to 6
Does that include Tunity Patches or is it just the OG Paper patches?
I think that the one they merged into paper yes, but i dont have confirmed this information
that's from papers patch list
Which, given that tuinitys patch set was made a part of papers...
Okay got it, I just didn't know if they were differentiated
Is it normal for Paper to Max out the "main" CPU-Core with only two players online? Its running on a i7-8550 at ~3.9GHz effectively
And I dont have any Plugins besides LuckPerms and a small one for moderation (I checked in timings - No time noticeable time spent on either)
CPU usage is a useless metric and unreliable
The vanilla server doesn't "sleep" between ticks too much, like, it sleeps for a few nanoseconds in a loop until the next tick starts
Yeah... but when a core is at 100%, there is no room for any faster execution
And I almost never reach 20TPS
Basically, no, it's not normal, but CPU load counters have been inaccurate for years so using just that as a factor means nothing
provide timings, etc
Huh... I might just ask the Players WTF they are doing first... Just wanted to make sure Im not chasing ghosts
them building 5 lag machines
Nah... Im pretty sure they are not, and the timings give no indication they do... But they might travel at max elytra speed for great distances or something...
Is Chunky prone to bugs on 1.18?
same here, at less start from 1.16.4... I forgot 1.8.x same or not. mc just use single core for almost everything
yea
it won't work for me
when somone dies
my server crashes most of the time
no other lag or anything
constant 20 tps
but only this problem
chunky literally just loads chunks
nothing more, nothing less
there is a known leak which only seems to impact people running chunky over existing worlds
yes
that is affecting me
which is nothing to do with chunky itself afaik
maybe is it because my server render is set to 18?
That could definitely be a key factor >.>
running it on a fresh world it starts to lag out after a while too...
don't mention
Ran chunky 2 days ago on a 50k x 50k worldborder. took almost a day, but didnt find any issues. Guess i was lucky.
started at a nice 12 cps then it drops to 3, time for a restart... /chunky pause restart /chunky continue
12???????
12 is much slower than the lows I've seen on 1.18, noting that the chunk gen slowed down significantly
pause and continue working either basically confirms my sus or says that your issue is 100% irrelevant to what I've been generally seeing
so, there is that
You did close your server for this right? you dont have people playing?
chunky just loads chunks
if it's what I think it is, not having players on can cause some fun issues due to how mc processes stuff
Could you elaborate Electroniccat?
my suspicion is that there is some process which is basically backing up
having players on allows for some areas of the code to be called more often which can aleviate that
it happens even on 0 players
-.-
That makes sense electroniccat, does not explain why chunky works a magnitude times faster on an empty server than with even a single player on standing around.
Im semi confused over "just loads chunks" though.
I thought it generated the chunks before a player loads them the first time, thus skipping the generation part and going straight to the loading part when a player discovers a chunk.
from an API perspective, that's just a load
chunky really just asks paper to load a chunk
the nature of loading, upgrading data, generating, is 100% hidden
with the generate flag set to true
having players on moving around, etc, causes stuff like the distance manager to be ticked more
Yea, so it requests a bunch of chunks, incase they have never been loaded before, it generates them, unloads them and moves on to the next bunch. right? or is my logic wrong here?
not having players on allows for more sleep time between ticks for it to do other stuff
The entire system is overly nuanced and complex
chunky does afaik not explicitly mark chunks to be unloaded, paper just does it after a while
So, question then.
Is pregenning with chunky or other filler plugins worthwhile?
I personally noticed lag dissapearing when many players were exploring pregenerated chunks compared to non pregenerated chunks
sryyy
mostly, no
Like, I recommend pregening spawn, but, for the aspects of most servers, spawn is the rea which is heavily loaded by people wandering around that area and then people tend to branch off
it's only when you hit sync loads that it 100% matters, otherwise the load of most users is generally negligable
No.
Literally only you is reporting of the server just randomly dying from shit like dying or sleeping
Hmm, makes sense. I guess the availability of a /wild or /rtp means more chunks get requested by more players faster due the lack of incremental exploration. It immediately dumps people into the wild, having to generate the whole area right away instead of gradually.
You most likely have not enough hardware power to run all the things your server is doing. Thus crashing, could be a myriad of issues.
yes, with lots of random access pregenerating is useful
hmm what specs you recommend also it didn't happen in 1.17.1
Yea thats usually when i noticed chunkgeneration being the laggiest. When streamers started spamming my SMP in their chat for viewers to join, and a dozen new peeps join in and smash the /randomtp command
This is a bad question. You should look at what you want to do with your server, and then find hardware that fits that.
7gb for 8 players?
yes
Well, unless you piled in a ton of plugins, you should never run out of ram
idk
[Thu, 9. Dec 2021 13:01:21 EET INFO] Plugins (15): AuthMe, CoreProtect, DiscordSRV, EasySetSpawn*, EpicGuard*, Essentials, EssentialsChat, InventoryRollbackPlus, LuckPerms, ProtectionStones, TAB, UltimateAutoRestart, Vault, WorldEdit, WorldGuard
this is all i have
authme 🤮

I've never used authme, thats for cracked places right?
Easysetspawn, wouldnt essentialsSpawn do the same thing?
Inventoryrollbackplus, doesnt CoreProtect do this?
Protectionstones???
UltimateAutoRestart?????????
i don't want sooo much essentials
and ir rollback for joins and deaths
and ps is for land claiming
So.. literally 2 things essentialsspawn does, probably more efficient too
Altho no direct experience, so not 100% sure, just doing a guess.
But start taking regular timings reports, log the TPS, keep track of what players are doing when TPS drops or the MSPT spikes
That way you can start to determine why your server is acting up
Only time i have abrupt crashes is when i get a bunch of dumbasses standing around at badly build goldfarms in the nether. Or someone watched that shithole of a Grian and recreated their boat bomb bullshit
also tpa also seems weird
"seems weird" is a very unspecific definition
it crashes the server too
That... actually makes little to no sense
especially without context
what is in the chunks the player TPA's from
what is in the chunks the player TPA's to?
The only time i've seen that happen is when people TPA'd away from Ilmango's server crashing portalbased goldfarm. Which i then promptly deleted from the world >.>
No point having simulation distance over 8 chunks. (128blocks)
Renderdistance over 10, eh, it looks nice, but with the fog etc it doesnt even work right. Unless you use mods to remove the fog.
ye i changed now sim to 6 and view to 10
before it was 8 and 18
simulation at 6 means you wont tick farms etc that are 128 blocks away to your NWSE, only above and below.
8 at the most, mobs etc despawn after 128 blocks unless you changed that. I intend to keep mine the same distance as i allow mobs to be active in.
Although to save load during primetime, probably limiting it to the same 5 chunks i set the render distance at.
No point ticking things people cant see.
Dont forget to inform your players of that change. Some might have to move AFK spots in farm areas to be able of covering all their farms in the area.
okk
Anyway, even though the current paper builds are relatively stable (ive been on #20 since they released that one), i do agree with paperMC's message of NOT using the experimental builds on LIVE servers.
Downgrade to Spigot 1.18 (even though it is far less stable as spigot always is compared to paper) if you really want to play with your friends online.
Upgrade to paper once they start releasing stable 1.18.1 builds.
Im personally most likely waiting on paper 1.18.1 before opening my servers.
Mojang dropped 1.18.1pre-1 monday i think? its already out on Bedrock, expecting it within 14 days for java.
dafuq?
So, that' not enough there as far as I feel
thats 7 players worth of chunks?
what?
or, am dumb
nvm me apparently
The dump I have is showing some other bug
if people ar ehaving issues with chunky draining them out of ram, I'd need a heap dump
Someone know when MC 1.18.1 will be available?
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.
Thanks!
You mean 13th. Sunday isn't a working day.
Well, it says this week, so then it'd be either today or tomorrow.
Not sure why the general text and the infobox mention different dates, but considering the 12th isn't a working day like you said, I'm gonna follow the infobox
There is no point of arguing the date. Its what the Wiki says. By the way this have nothing to do with Paper 1.18 Experimental. Lets keep it short.
I assume it's today. Since releasing on Fridays isn't the best to do. 😆
Yeah, today would be nice
according to the infobox, 1.18.1 released last week, so I think we can safely ignore that lmao
1.18.1 hasnt been released
obviously
yes very obvious
according to the infobox
I think we can safely ignore that lmao

oh ok
I'm hoping it comes out today or tomorrow, been waiting for 1.18.1 to update personally, and those fog fixes will be nice for sure
ur right
still i just wanna confirmation bias post 😦
truuuuuuuu
btw reducing sim and render distance stoped crashing ig
it its been 1.5 hours
fabric is just vannila with extra steps with mods
also data pack is lagier than a plugin
it is always on
but plugin only uses hardware when triggered
- data packs have to do a lot more to get the wanted results. Not very efficient
Has there been any reports of Paper 1.18 corrupting worlds? I literally need to have a 1.18 world generated without any issues and I want to know if that'll be fine on the latest Paper. (Eg have the generate random feature-seeds working, etc)
I mean, not tmk, but if any issue occurs you should report it and try to reproduce. As long as you make backups if anything happens you’ll be fine.
There have been reports of user faults by using force upgrade e.g.
DONT genrate without backup
i look pretty often at this channel and nobody has mentioned corruption with new worlds, generating a 1.18 world should be fine. just backup often
I'm literally generating a fresh 1.18 world.
you'll be fine, just keep backups
Just want to make sure everything will be fine since this world I'm generating is what I'll be using in production when 1.18.1 releases
maybe... if you want to be sure then pregenerate in vanilla+fabric then move to paper?
What about Spigot?
Not pregenerating right now; I need to scramble feature-seeds in paper.yml
sorry im not sure what you're asking, spigot generation is the same as paper's
Never mind.
Anyone has an idea? 😬
Use plugin not datapack @peak pulsar https://github.com/MC-Machinations/VanillaTweaks
Enable anti creeper module on this
Maybe its not working correctly cause i had those dimension import problems .. also stuff like rain is everywhere indoor and stuff like that
if you use a plugin
it will have options for world settings
and custom Worlds
Thank you for Plugin hint! ill check that 🙂
Wait you only want it to happen in certain world?
and why do you use data packs while on bukkit
Don’t creeper only spawn in overworld?

worldedit+worldguard are overkill for just stopping creeper explosions
Deny creeper griefing globally in the world you don’t want that
At thw moment is okay for me if its global setting for every world, its important for my original overworld and also old 1.17 imported overworld (now working as custom dimension)
i use for school
okey thank you very much
I want the explosion, but not the broken blocks
hmm then plugin
would be better
The data pack had perfect features for my case, but there is something wrong with some of my dimensions so i should look for plugins and chdck if there is still the same problen
this would be great as a WorldGuard flag
Thanks guys
who wanted the heapdump for the leak when using chunky to upgrade on existing 1.17.1?
I can't mp or tag electronicat
cat wanted
is it normal that my chunks in the latest paper version are gerneration really slow. especially in the nether.
no
Not had any issues at all personally
Yes, 1.18 chunks take way longer to gen
Or do you mean compared to previous previous 1.18 Paper versions?
do you mean "latest paper version" as in "build 63 is slower than build 62"?
1.18 in general has slower world gen
but if you mean specifically a paper build... that might be a bug
Is that also the reason for such high memory usage for 1.18?
eh its not that much higher than 1.17.1
Java will use what you told it to use. Versions don't magically """use""" more memory itself.
the world is twice as tall so yea theres going to be increased memory, but it shouldnt be that much compared to 1.17
Hmmm okay, might have to give paper a go, cause im trying to run a close to vanilla server, and it keeps using up all of my memory and going OOM
b r u h
i assumed you were using paper since you are in the channel where we discuss the experimental paper builds in the server where we talk about paper
Nah, just thought to ask if it was minecraft in general
if you need a vanilla server then you might be able to patch some performance with fabric and phosphor, lithium, carpet, krypton, starlight, hydrogen, etc
but paper is definitely going to perform better than the vanilla jar
lots of optimizations by smart people in there 😄
Have to give it all a go again, and enable the dupe patches, so the players can have the vanilla experience still
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.
?
it has been 6 minutes maybe now there is an eta
Bot was feeling lonely
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.
omfl 😂

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.
its been 12 mins maybe now theres an eta
You don't need to spam commands every x amount in topic channels.
its funny tho hehe
No, it clutters topics. If you want to spam bot commands for sake for wasting time, use #bot-spam
What percent done is 1.18
Timings?
@devout musk that java 17 build is out of date
that docker image for whatever reason ships the EA of it
there's a better docker image to use nowadays iirc, cant remember what its called though, if you search the server you'll probabl find it
I've seen most people use openjdk:17-slim
temurin best
still unresponsive...
if ptero, reduce Xmx
8gb too much?
I mean, big issue is that people hit the memory limit in ptero and that causes it to pause the process basically
