#❗-1-18-experimental
1 messages · Page 2 of 1
https://minecraft.fandom.com/wiki/Java_Edition_1.18
1.18, the release of Caves & Cliffs: Part II, is an upcoming major update to Java Edition set to release on November 30, 2021. ... It will completely overhaul ...
ez

:D
is 1.18 more optimized than 1.17? (in terms of resource usage, and yes i know there's more to process now but overall..)
No
so optimization is still garbage, oof
'what's an optimization?' - mojang, probably
they actually did quite a few things
you guys still hoping that mojang will ever optimize their game
they are
Is Kenny more optimized than Kenny?
At least 3x as optimized
Nice!
Mojang, optimising Minecraft? 🤣 🤣 🤣 🤣 🤣 🤣
No, Kenny optimizing Kenny.
kenny 🥺
Ah I did not read the full conversation, just saw mojang and optimization so I had to make a joke.
guess what i'm having more issues with paperweight because of course i am
admitting defeat and that paperweight will never work cross version because noone wants it to, updating 1.17.1 to data version 3 clearly is beyond the limitations of possibility
so i'm doing the one thing i didn't want to do which is write my 1.18 nms code in a whole other gradle project because why would i have a nice workflow when i can bend to fit with paperweight's brokenness
Cool
but now it doesn't even want to work at all
Could not resolve io.papermc.paper:paper-api:1.18-rc3-R0.1-SNAPSHOT.
fym
if i can import the dev bundle i can import the rest of this shit
so why won't you work gradle
im losing my mind
Because the deployed paperweight is not ready yet. Build it yourself and publish it to maven local
hi
ILLIMAR KANGUR
"How come all of this experimental stuff that's labelled experimental isn't production ready"? -half the people in here
im just going to give up and hope that it's fixed so that i can release 1.18 support at a reasonable time
It is fixed, the deployed builds are just cached.
ah
Tbf, it was never broken in the first place, just the deployed builds were
still have to do annoying multiple projects because 1.17.1 is data version 2 and gradle doesn't like having different plugin versions between submodules
which is an enormous pain in the dick
So is supporting old versions.
Worst case composite build
patience™️
im hoping
first time paper updates to 1.18 with new tooling & first time in quite a long time where paper releases pre-release versions
expect it to be broken one way or another
i'm fine with things being broken
that's obviously to be expected
i don't like how the refusal to have v3 data for 1.17.1 means that everyone who touches nms at all has to resort to having extra projects (not submodules, whole other projects) because gradle can't handle having different versions of the same plugin
if 1.17.1 just had v3 data then everything would be perfect
but it doesn't so everything's a shitty workaround with a horrible workflow
Then there's no need to complain on a daily basis, we are aware that things are not working as expected yet, that's normal for this kind of stuff.
it'll be fixed sooner or later
as long as it gets fixed at some point that's fine but from what i can tell there's no plans to fix it
You just have to chill for more than 2 seconds
i think the paperweight authors are well aware of the suboptimal situation
This kind of stuff is why people hate doing pre-releases.
Hurr durr you're saying that not everything works 100% from its beginning?? How dare you!
Instead of seeing it as a boon because you can start updating and check/test things, people are griping that things aren't done yet or are broken. I mean - come on.
Noone nowhere refused it. In fact, there are plans on providing a solution after other, more pressing matters are resolved.
You can subscribe to the issue to be notified when it gets done.
https://github.com/PaperMC/paperweight/issues/116
you're right i should mention that issue
No
Mojang should show stuff as soon as possible, to gather as much community feedback as possible
If that means that stuff doesn't make it or get late or whatever, that's just how game dev works and you gotta deal with it
I wish every game would ask players feedback
Stuff moves around releases all the time
The new world gen was planned for summer for example
Nothing is done until it's delivered
It goes to show how awful devs are at estimating time.
That's how it works
Either delaying or pushing broken shit
Sometimes you just discover you need a new approach etc
Not at all
Its not even devs
Game designers need time to iterate
You set a goal, so that you don't go iterate forever, but if you are close to that goal and not happy yet, you move the goal
It's impossible to predict
don't game designers recruit predictors? how so? (sarcasm))
just finish it 
I have a question about the max-block-height setting for the anti xray:
If you are above the block limit and then you enable the xray, will you see blocks bellow you (I mean non-fake ones) or ONLY above you ?
Idk if it's clear ^^"
It only mangles blocks below that height.
Player location doesn't matter, block location does.
If I understand your question correctly :)
Yeah, thank you
When I try to connect to the server, it says ''Please use the 1.18 RC-3 client'' even tho I am on that client already
Anyone experienced something similiar?
reproduce without plugins
Still occurs
reproduce without proxy
Though it is linked under bungee
that's your issue then
That's not possible, yet?
dunno if the latest waterfall supports 1.18 yet
I did update viaversion and waterfall for RC-3
Checking the commits of waterfall, it seems to be updated for that
has anyone tested the world blending when updating paper?
antixray config generation is not working on last commit? (nevermind, found the issue)
org.gradle.api.internal.tasks.TaskDependencyResolveException: Could not determine the dependencies of task ':applyPatches'.
(only gradle classes in the stack, removed them)
Caused by: org.gradle.api.internal.tasks.DefaultTaskContainer$TaskCreationException: Could not create task ':applyServerPatches'.
(only gradle classes in the stack again, removed them)
Caused by: java.lang.NoSuchMethodError: 'void io.papermc.paperweight.patcher.tasks.SimpleApplyGitPatches.doNotTrackState(java.lang.String)'
at io.papermc.paperweight.patcher.PaperweightPatcher$createPatchTask$patchTask$1.invoke(PaperweightPatcher.kt:262)
at io.papermc.paperweight.patcher.PaperweightPatcher$createPatchTask$patchTask$1.invoke(PaperweightPatcher.kt:52)
at io.papermc.paperweight.patcher.PaperweightPatcher$inlined$sam$i$org_gradle_api_Action$0.execute(TaskContainerExtensions.kt)
at org.gradle.api.internal.DefaultMutationGuard$2.execute(DefaultMutationGuard.java:44)
at org.gradle.api.internal.DefaultMutationGuard$2.execute(DefaultMutationGuard.java:44)
at org.gradle.configuration.internal.DefaultUserCodeApplicationContext$CurrentApplication$1.execute(DefaultUserCodeApplicationContext.java:112)
at org.gradle.api.internal.DefaultCollectionCallbackActionDecorator$BuildOperationEmittingAction$1.run(DefaultCollectionCallbackActionDecorator.java:95)
at org.gradle.internal.operations.DefaultBuildOperationRunner$1.execute(DefaultBuildOperationRunner.java:29)
at org.gradle.internal.operations.DefaultBuildOperationRunner$1.execute(DefaultBuildOperationRunner.java:26)
at org.gradle.internal.operations.DefaultBuildOperationRunner$3.execute(DefaultBuildOperationRunner.java:75)
at org.gradle.internal.operations.DefaultBuildOperationRunner$3.execute(DefaultBuildOperationRunner.java:68)
at org.gradle.internal.operations.DefaultBuildOperationRunner.execute(DefaultBuildOperationRunner.java:153)
at org.gradle.internal.operations.DefaultBuildOperationRunner.execute(DefaultBuildOperationRunner.java:68)
at org.gradle.internal.operations.DefaultBuildOperationRunner.run(DefaultBuildOperationRunner.java:56)
at org.gradle.internal.operations.DefaultBuildOperationExecutor.lambda$run$1(DefaultBuildOperationExecutor.java:74)
at org.gradle.internal.operations.UnmanagedBuildOperationWrapper.runWithUnmanagedSupport(UnmanagedBuildOperationWrapper.java:45)
at org.gradle.internal.operations.DefaultBuildOperationExecutor.run(DefaultBuildOperationExecutor.java:74)
at org.gradle.api.internal.DefaultCollectionCallbackActionDecorator$BuildOperationEmittingAction.execute(DefaultCollectionCallbackActionDecorator.java:92)
at org.gradle.internal.ImmutableActionSet$SetWithManyActions.execute(ImmutableActionSet.java:329)
at org.gradle.api.internal.DefaultDomainObjectCollection.doAdd(DefaultDomainObjectCollection.java:264)
at org.gradle.api.internal.DefaultNamedDomainObjectCollection.doAdd(DefaultNamedDomainObjectCollection.java:113)
at org.gradle.api.internal.DefaultDomainObjectCollection.add(DefaultDomainObjectCollection.java:258)
at org.gradle.api.internal.DefaultNamedDomainObjectCollection$AbstractDomainObjectCreatingProvider.tryCreate(DefaultNamedDomainObjectCollection.java:944)
... 137 more
I get this when trying to build a fork of Paper fork, did I miss some update?
actually stacktrace is much longer but I don't think its removed part would be useful
.paste all of it
the entire stacktrace: https://paste.ubuntu.com/p/hDzTRQhHdp/
Does your fork uses the same paper weight version as upstream at that commit?
I fail with that every time. 😄
I know you propably have a TON of these messages arround, so I am sorry for that, but any idea how long could it take for this version to stabilize?
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.
6.9 years to be more precise
6.94 years*
*6.9420 years
rc4 out 
Yup, looks like they just "fixed coal distribution"
Another lol dang
I am not asking for a precise release
Whatever that means
I thought they only wanted to release one release candidate
just wanted to know if I should expect it soon or later
There's no such thing, see the bot's message above.
Rephrasing your message doesn't change the answer
okay, just asking. I know its not easy, its just that I have a christmas event on my private server soon (13 days) and I want to know what to expect
anyways, thanks for at least botting me off XD
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.
I can read you know
Well you can expect that it will come out when the update is finished 😄
this is not a question
π-years
good one
don't be rude, they got an answer
anyways, its time for me to leave this server, just wanted to check what should I expect as I said.
no eta is best eta. take your time papermc team 
Best eta is community
Asking for release when full 1.18 isn’t even out yet. 😂🤦♂️
And the amount of activity in general
I loved last release, how the bot was counting how many times did people asked about eta... :-D
is 1.18 going to be any more different than the snapshots?
It'll be pretty much the same as rc4
Might be a bug fix or two
But the release candidates are pretty much the same thing as the official release
ohhh ok
wtf since when is there a rc4 😆
Since today
wtf that was quick
Tommorow 1.18 pogg
What sources do you have?
the release date has literally been announced a week ago
mojang invalid source what youtuber said this
Xisuma
why are they still releasing rc3? rc4 is already out
?
?
?
release candidate 4 is already out
Okay?
ok?
ok
Reading comprehension is hard
Spigot isn't on rc4 yet
Do you know how any of this work? @young tangle or do you spend some time to read #❗-1-18 ._>
rc4 was released like two hours ago
sorry im dumb
i hate how whenever you ask a question here you just get attacked
Even then
its fucking annoying*
it was a genuine question tho
not really though if the person doesnt even bother to read the information provided
good question gets good answer here
i thought rc4 was released yesterday and im just outdated
I do agree that #paper-dev does this a lot but when people spam for ETAs and updates literally 400+ times every single update, you lose any patience you had for those kinds of questions.
Especially when the person is asking hours after the vanilla update
Not to fault them. Misunderstanding about when it released
It happens a lot where ppl get a rude response. Not everyone has the time to read every single message in this discord. If you can't deal with answering a question, don't answer 🤷 If you are taking the time to write a response, might as well just give them the answer.
It will save them time and honestly stops a bit of the negativity
🧱
Seems like an in depth discussion that isn't really at home in #❗-1-18-experimental 🙂
ngl minecraft patch notes should have an "edited at ____" for ppl to know when it was released
especially since they use the same post
New topic, folks ;)
^
i wonder what exactly they changed because i didnt notice a thing about coal yesterday lol
How is the performance currently in 1.18-rc builds?
deepslate coal was generating too frequently i believe
decent ish
it's missing a few optimization patches
decent, except for worldgen
(which isn't really paper's fault)
well worldgen is gonna be a bit slower coz world is literally double the size as before 
you're still out of luck
even a 5950x is getting 60 chunks/s
8673068 chunks generated, overworld size: 95G
That sounds perfectly fine for 1.18 emperimental builds 😅. Thanks for the info!
I swear if mojang doesn't do an optimization update after a single update had to be split into 3 parts...
they optimized chunks and world gen/loading in 1.18 though
I think 1.18 was the optimization update
With all the new features lol
I love one line updates!!
yeah they sure did
https://sirkillian.gg/i/444j63brkujmqgza.png
. Oh god
they could still do more like that feature where they were going to turn large flat surfaces into a single face
vanilla 1.18 performs much better than 1.17 for me
pre-gen will still be mandatory
wtf did you do
walk.
what hardware
infra2 ovh, dont know specs from the top of my head
it was 2 players flying in creative into unloaded chunks
its mostly fine but we noticed it does these weird tps drops
actually im not sure about the old infra-2
not sure how to replicate the drops yet but we assumed it was a specific structure that gens
Is paper stable enough rn to not worry about world corruptions? Or should I switch to fabric for a bit
(the 1.18 experimental releases)
You should definitely wait after the initial release of 1.18 to worry about stability.
Doesn't matter if you use fabric or paper
1.18 itself isn't even out period you shouldn't be running it on any platform in production
I meant like Day 1 lol, either Nov 30th or Dec 1st
Guess I might just do vanilla until it's stable 'enough'
We don't know if we will release paper day one, or even be able to
We also don't know if vanilla is stable enough
There hasn't been a release without a quick .1 in a while
1.18.1 Changelog:
• removed bug that would delete system32
If you care about your world, wait two weeks, that's a good guess
That unironically happened with the dungeons launcher lol
this happened yesterday on RC3: https://sirkillian.gg/i/n9lptt5ivoswz8xs.png
This also happened:
https://sirkillian.gg/i/vkw63957zzio85d1.jpg
surely thats not intentional?
ask Mojang to fix those
Oh god
they wont fix it
those are micro biomes
they said they cant fix it but they will try to reduce them
it will be odd if paper try to alter the world generation
I am sure 99% of player will not want that
That's on the other hand is cool.
that said, the structure looks kinda cool, but sometimes you get like a random tall jungle tree in the middle of a desert / snow biome for example xd
The biome Minecraft is missing is volcano tbh
the microbiome thing is super ugly imo but its something we will have to deal with
I played modpack before and a volcano would look cool 
The biome minecraft is missing is performance 
Just waiting on the Rory biome 😔
how is it missing if it was never there to begin with?
True
I mean
You could've tell that to Fall Guy developer
what also bothers me is how they didnt change the dungeon structure... it looks super ugly in the new deepslate caves lol
That INCLUDES file sizes looks at ARK
Anyways I kinda wanted to do a 1.18 Christmas advent thing so I guess I'll just bite the bullet and run 1.18 without the .1 update while making 10 backups a minute
People still play among us as well sadly

Nah I meant vanilla dw (for my own server with like 2 people)
its like project winter but with shitter graphic when I first saw amongus
👍
... seed, cords, and build number?
no longer have the coords
would anyone happen to know where the experimental server jars can be downloaded?
Found another biome like that the other day, it was literally one chunk
it was the saddest ice biome ever
is it going to be a struggle for 3970X running 1.18 rc 3 ?
no
how resource intensive is running 1.18 server compared 1.17 server ? im worried that my puny 32GB ram handling 150 ish player won't be enough
its not that much more demanding
not from what ive seen at least. Then again, im kind of an idiot sometimes
because i saw the world size is HUGE compared to 1.17
also, pls tell me that 32gb was a joke
thats just the storage required for a world, that sure has gone up
true somewhat
you'd have to compare tho
sections above won't increas
caves below tho maybe
i can check rn
the region files dont look bigger in size, but there is probably more of them on a fully generated world
nope its not a joke, since ECC memory is hella expensive here in my country, trying to order 16 more gb from ali express but it need at least 3 weeks before it arrive
You do not need more than 10G to host a server
For majority of server 10G will be enough
^
the biggest one i have on rc3 is 10.1MB
Seems very much alike 1.17 tho
so im going to assume there is just more of them
i could maybe pregen a 1000x1000 later and compare
would be very useful
the heck
maybe ill do a 5k world lol
try 15K
Why stop there.
Try 30Mx30M /s
running around in spectator mode with 10 of my friends running together, booting up using linux VM, allocated 4 core of 3970x, quite smooth~
i was just about to say that
the fix for this is somewhere down that thread
beta chunky for 1.18 https://ci.codemc.io/view/Author/job/pop4959/job/Chunky/
performance seems good, phew
thank you 🙏
im doing a 5k world while i eat dinner
testing me too, seems to be fast 🥺
compared to 1.17, around 20 - 30 ish percent heavier
and watch out for bandwith usage when alot of player running around 😂
why would it be worse
chunks are taller but meh
well taller chunks maybe means slower loading times
how big is that ? 1k x 1k ?
bruh
owh ok
mechanical on a vps?
yeah is a cheap vps that i use for telegram or discord bots
what
how much per month
i'll try on my dedicated server with ryzen 3600 and nvme ssd
am i reading that right? 1.18 is 30% heavier than 1.17?
3.99 euros
its really not
ur just throwing arbitrary numbers around
ah, good
Chunk generation is like 40% of what it was in 1.17 but it's also doing at least twice as much work so...
dunno in normal use case but i am with my friends spreading around in spectator mode with 10 people, so our cpu usage spike quite heavily~
ah! [laughs in quad core server]
reasonably priced i think~
well its not like what i've been anticipated... XD because back then when i saw the pewview im scared that i'll run out of horsepower since the world is twice as big now..
Hi, I would like to build against the mojang mapped server jar for development and be able to automatically build my plugin using Github actions. Therefore I don't really want to have to build the jar every time.
I know a mojang mapped jar is currently available for download, do you plan to keep supporting this (if possible with the licencing and such)?
I think you want paperweight userdev
Ok, I will take look at that, thanks
ofc, its an outstanding cpu
why ubuntu (just curious)
also everywhere i look it says its a 2.7ghz cpu, how did you overclock a laptop to 3.8 ghz?
Go with Debian then
Strike out the to maximize performance and you are good to go
i7 7500 should be enough for 10 - 15 players, use Ubuntu :>
be careful with debian, you'll definitely get very good performance with it but it has issues with stuff like laptop gpus and wifi cards (please don't run a server on wifi)
There is bar null difference between using debian or ubuntu but a personal preference
Linux in general is gonna perform better than windows, etc
Distro is just a choice and choice is yours
oh sorry, its a 2 core 4 threads, unless you're using tons of plugins it will able to perfom just fine~
But, practically every distro is about the exact same with a bit of different tooling choices; apt vs yum, netplan vs ifup, etc
5000x5000 full pre-gen comparison
Paper 1.18-rc3
Time: 27:46
/world size: 1.1GB
Paper 1.17.1
Time: 10:07
/world size: 785.0MB
omg almost 3x
does paper 1.18-rc3 have starlight?
but that size is really nice~
ill test this again when the final builds start to come out
yes
Did 1.18 actually add any new blocks, or is it all just 1.17 blocks?
1.18 was mostly just finishing off what was started in 1.17
It's all 1.17 blocks
nothing new was really added afaik, see the changelogs
agree, also less wasted resource~
what is this error? Exception: java.lang.NoClassDeffounderror thrown from the uncaughtexeptionhandle
er in thread "Server thread"
That gives no info, we need the whole log
yay I don't have to go though adding blocks to the default config file
cmd are crap to copy from 😦
There's a latest.log file in your logs folder, hopefully
if it's that crap you can't get the actual basic info from, we can't help you
@hollow notch Best we can tell is that something is referring to a class that does not exist.
That's it.
so now I force it to copy , is rely hard to get it to copy in win 7
B:\servertesting\paper-1.18.0>java -Xms1024M -Xmx3048M -jar Paperclip-1.18-rc3-R
0.1-EXPERIMENTAL-reobf-3.jar -o true nogui
Starting org.bukkit.craftbukkit.Main
System Info: Java 17 (OpenJDK 64-Bit Server VM 17.0.1+12-LTS) Host: Windows 7 6.
1 (amd64)
Loading libraries, please wait...
[20:19:11 INFO]: Environment: authHost='https://authserver.mojang.com', accounts
Host='https://api.mojang.com', sessionHost='https://sessionserver.mojang.com', s
ervicesHost='https://api.minecraftservices.com', name='PROD'
[20:19:11 INFO]: Reloading ResourceManager: Default, bukkit
[20:19:12 INFO]: Loaded 7 recipes
[20:19:13 INFO]: Starting minecraft server version 1.18 Release Candidate 3
[20:19:13 INFO]: Loading properties
[20:19:13 INFO]: This server is running Paper version git-Paper-"0fe0613" (MC: 1
.18 Release Candidate 3) (Implementing API version 1.18-rc3-R0.1-SNAPSHOT) (Git:
0fe0613 on dev/1.18)
[20:19:13 INFO]: Server Ping Player Sample Count: 12
[20:19:13 INFO]: Using 4 threads for Netty based IO
[20:19:13 INFO]: Default game type: SURVIVAL
[20:19:13 INFO]: Generating keypair
[20:19:13 INFO]: Starting Minecraft server on *:25565
[20:19:13 INFO]: Using default channel type
[20:19:13 INFO]: Paper: Using Java 11 compression from Velocity.
[20:19:13 INFO]: Paper: Using Java cipher from Velocity.
>
Exception: java.lang.NoClassDefFoundError thrown from the UncaughtExceptionHandl
er in thread "Server thread"
This still tells us nothing
if that's al there is, we can't help
Shall I try fresh install instead?
Although you really should not being using Windows 7.
win 10 are no good and win 11 is even worse
Solution: Linux.
I´m not want my computer restart
wow some plugins even run right away without problem, love it ❤️
Just means they are not using parts of the bukkit api that have changed or been borked.
Sad thing not all things exist on linux and no wine are not a fix for the issue ether (it has it´s own issues to).
for running a server at least you must run some kind of VM or rent a vps or installing the server on separate machine using linux~
don't tell the kid to install linux just to get rid of a weird paper error, they probably have their reasons to use an unsafe old os
or just stubborn idk
adobe, all games i have. and winscp are some things some not exist or likely good.
bruh
oof
nevermind what i said, then
omg
don't forget to add "his main windows 7 pc" for the shock-factor when others read the chat
yea, no luck for the rest of my 50 ish plugins...
Sure it can
depends on the cpu
How do you think singleplayer works
I will ofc or force to move from win 7 when i upgrade my computer, so can not stay on win 7 forever XD
And no I only run testserver on my computer not any production server (It can´t handle that, it has to little ram).
os doesnt matter
besides, if its just to play with a handful of friends, it will be fine
yeah it will be foineee
no I have i5 2500k 980 gpu and 16gb ram and several ssd and hdd lol
you'll be fine
for a local lan multiplayer 8 ish player its fine~
it will not be foinneee*
I wish I had 16G of ram
Actually, it does
not much tho
IO perf is much better on Linux
I can say 16 is to little next time i will have at least 32gb ram 🙂
ext4/brtfs/zfs/etc is also faster than ntfs/fat32/etc
if you use more than 10gb of ram you are doing something wrong unless u have a very specific usecase
especially on System Drive Performance
say that to my computer some complaining and shot down my programs.
only idea run on 3gb two minecraft 1.17 clients and web browser and you are done.
thats not just discord
reminds me of irc
just take everything with a grain of salt and look at actual data to make up decisions
Looking forward to using 1.7.10 on a 1.18 server
I can only run 1.18 server if i close down most of the running programs (i use approximately 14GB to 15GB) with one 1.18 client
and one 1.18 server
if i has idea and web browser running
is it possible to boot up creative server using RC 3 ?
then how exactly does it use 14-15GB if server is only taking 3
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.
so their client uses 12 gb of ram?
one error i get when not have ram left
imagine you allocated 12 to client xd
it use around 3gb every client.
if you not limit it
but 3+3 != 14-15 lol
but how are you using 15 gb of ram with minecraft??
if anyone has free time help me to stress my server out, running 1.18 rc3 address : minecraft.kiryuukaito.live creative mode enabled!
go here and click on memory https://sirkillian.gg/i/6ip92pvs2fr8ga6v.png
awww i love this screen, such a beautiful ui. you're doing something wrong if your minecraft takes up that much ram though
Just use MineFlayer or something
i wonder if they're secretly running 420 chrome tabs

no firefox
@hollow notch do this
you were right
steam, gitub, fnaticop, autodesk, battle dot net, wargaming, atom, intellij, a bittorrent client...
that's like five chrome instances
...?
What a chad
I not run win 10
click the first tab
what language is this
click "processer" @hollow notch
then click "minne"
guys pls they uploaded a screenshot already
that's where i got this from
oh
That's W7, you're not gonna find PC info in the task manager lol
I have say all the time I dont have that in windows 7 🙂
@hollow notch Can you press the windows key + R and put in msinfo32 and press enter. Screenshot the window that pops up. That'll tell us your system specs.
is there a reason why you still use win7?
Me too and no task I has running use so much ether.
I not like force restart on windows 10
they already shared that they have 16gb ram, i7 2500k, a few ssds, etc
Ah wasn't sure what y'all were trying to get lmfao
he said he allocated 3GB to client, 3Gb to server and yet its running out of memory while he has 16GB ram
so now we wanna know where the 10GB is coming from lol
He's got like 40 other things taking up stuff
the PERSON said that his 1.18 client + server uses 15 gb of ram but we were all surprised at this claim so we asked for a task manager screenshot where we discovered they have like 69 background processes (steam, battle, etc)
Yeah you're trying to run a server on a PC running other stuff lmfao
Either buy and install more RAM or close stuff out
AVSLUTA! AVSLUTA!
cat 
Was way I say i will upgrade to 32Gb next time 🙂
literally just close intellij and atom instead of buying more stuff
buy a 240G ssd for swap 👍
I still need buy new stuff, this computer have start to get some issues (the build in network card are dead, some sata are not in best shape to day ) and 32Gb are not soo expensive 🙂
So when I need upgrade, i do it probably 🙂
i'd say its more of an user error...
yeah I know I can´t buy to day. Because the current situation.
But I run testserver to test plugin and i need have idea open and two clients to test out my own code and some games use lots of ram too (they load lots of data to ram).
I know for example factorio load the map to ram and have you big map loaded. I can say it easy use 8GB.
ah so you're coding 1.18 stuff while playing other games, everything makes a lot more sense now
minecraft cilents x2 lol
I not often run a second game beside(I not have ram to do that ether)
Depends
i5 cpu
10G is the highest youll realistically ever need for any server
the limiter these days is single-core CPU performance.
Me who gets away with 1GB of ram on my testing server...
highest
By the way, I found out that the default RAM allocation in Minecraft is far from enough for 1.18. I doubled the RAM usage, my frames doubled along.
if you pay for 30 you should use all 30
No...
I bought a machine lol
Sometimes allocating more ram then you need causes issues.
Just allocate what you think you need, and then add a few extra GB just to be safe.
yeah but if you have 30 sitting around it won't hurt to use them all
split it across 2 servers would be more efficient
I'll probably go for 16GB
Reasonable
It can. You are removing yourself from using it for multiple servers, or letting your OS use it for cache.
well yeah using 32 on a system that has 32 would be a problem, but given aikar has flags for 12+, 12's fine.
Meanwhile my testing server is crying about it's 1gb limit
There’s too little ram, and too much ram, you have to hit the sweet spot for your server specifically
No one setting or ram allocation will work in every server, it all depends on your needs
hi, i play with 12 friends on my server on survival with some farm... I have choice between a 8GO Ram debian system or a 16Go Ram System. Which is better for us?
eh no as much as you can up to 12 should be fine
The one with the better CPU single core performance
CPU's the more important part.
Do not use 1.18 for anything other then testing/development.
At 12 players, RAM isn't a big concern
it's a public cloud with amd epyc
you will be fine
they spelled epic wrong
Like what clock speed? What model? Etc
Does paperweight userdev already work with 1.18?
It should
getting some errors then when using 1.3.0-SNAPSHOT, https://paste.fvdh.dev/uyulilejex.lua
Clock speed isn't a good way of comparing CPUs
not the best CPU, but it'll work. I assume this is shared hosting / a VPS?
no a public cloud from infomaniak
overkill much
it's not what i have XD
i have choice form instance between 1 to 32 cores
Cloud usually means vps or shared hosting
yeah. it's like a vps i have ... but i can contrôle everything
Dumb question; is rc3 stable enough if I wanted to just run a small survival? Or should I just wait?
So not shared hosting, either dedicated or vps
Wait ... How much does it cost?
Yes and no, it’s kinda stable but there might be changes that could break things when it fully releases, if you have to run to in production backup everything, but you should probably wait til it’s fully released
Cool thanks
between 12€/month with 4cores and 8Go to 25€/month for 8cores and 16Go ram
There's not enough data to tell you, and there'll likely never be, as most ppl will wait for stable releases before putting it in prod
^
i think you can create your own configuration
It’s called experimental for ᔑ reason
you have the price here
I CANT CLICK ANYTHING ON THE INTERNET WITHOUT THE COOKIES PROMPT SHOWING UP
oh yeah this thing ... internet is a kitchen
Click to accept my cookies and see my message: ||lol 🍪 ||
Best server hosting for a small server of like 12-15 people?
I'm using https://chrome.google.com/webstore/detail/consent-o-matic/mdjildafknihdffpkfmmpnpoiajfjnjd?hl=en it's epic
already config or vps ?
That's not really the right channel for that
oh sorry
no problem
stupid fucking school google account arghhh
[20:55:00 INFO]: [Chunky] Task running for world. Processed: 9590058 chunks (98.08%), ETA: 0:47:40, Rate: 65.7 cps, Current: 197, 1548
almost, there..
i'm using ublock origin and its still not enough
I just use the i don't care about cookies extention, seems to work thus far
I use bloom.host. Great for the price
You have school google accounts jeff man
Damn I didn't know that was a thing
Nifty, the moar ya kno
@still sparrow pfp
Hah, people using teams in my hs str8 up using cropped hentai as pfps. Pretty sure the teachers caught on but i'm sure they aren't paid nuff2care
Those aren't the defaults iirc
Or atleast haven't been since lyk 12'
i dont think so
Isn't it some randum shape?
Like four squares and a square in the middle or sum shit
can you guys continue this in general not the channel made for 1.18 experimental build discussion
oh lol i didnt realise this is where we were
Fine.
Okay thanks. I heard about shockbyte. Is shockbyte good?
No, it's bad
Shockbyte is bad?
Rip
I live in Canada so it’s kinda weird finding a server host with servers in Canada.
Shockbyte looked promising.
I mean canada has OVH so if you're ok with quebec there should be heaps of hosts
Should I even try to use shockbyte?
I used to use server.pro
But shockbyte looks much better than server.pro
extravm, bloomhost and dedicatedmc is what's been recommended in here
^^
Depending on what side of Canada you're on you can also go with a server in Buffalo, NY if you can find one, it's close to the border
#paper-help tho
i've tried most hosting services, heavynode is the best accross the board
for inexperienced server owners
BisectHosting has servers in Montreal, Buffalo, and Oregon.
Montreal and Oregon are located at OVH, Buffalo at Colocrossing.
Just don't choose a host that doesnt tell you what cpu you get, thats the worst thing
The internet has a few dozen peering points around the world
just go for the general reason, no need to give a shit too much about where exactly it is, it rarely matters
Yeah, but when you get that 4 msec ping...
This version is quite ram heavy I've noticed
Not paper specifically, just 1.18 and spigot in general
I'm hoping paper and spigot can at least soften the blow
chunks are bigger now, there's twice the world
lot more disk space too
Seems to be better than early versions of 1.17.
Besides the fact the mojang is still releasing RC versions the day before release lol
What is your guess for averaged increased usage of ram
I mean for those who "pregen" to try to get ahead, their coal distrubation is fucked
that's why you want to wait til final final version
Not that I'm worried I run 20gb on a server that has about 30 players first day of a new world, then dwindles down to 10-12 on at once after a few days
Anyone who does this prior to release for production server deserves it lol
In this release I assume it will get a lot of early patches also after release
Once the masses rip it apart
That's what I've been telling everyone 😄
no one ever listens
i never preloaded my worlds lol
though this is kind of like the same issue that happened in 1.17.0, and then they fixed in 1.17.1
set a border and let players load it all in, that's what i did 
This update was too ambitious on the terrain too late into the update to finish it properly. And my god basically only Henrik did most of it lol
for small server that would be fine @lusty flicker
paper handles chunk generation nicely but every little bit adds up if your server is big
it usually won't lag if you do that but chunks tend to load in very slow for most servers with a lot of players in new chunks
so pregen is still reommended say around your spawn point 5k~10k even if you dont plan to have a borader
yea i noticed if you "hard-load" chunks the server just dies
I flew around in paper 1.18 for about 30 mins last night. For me it ran beautifully vs vanilla (doesn't take much but still). But that was just me solo
fabric client mod 
most survival servers have something like /rtp, now imagine 50 people rtp-ing every 5 seconds
from 100 fps to 200 fps
Particle.Barrier removed in 1.18?
yall think an smp is safe to start now?
with the wrong random tp plugin you'd be screwed XD
but if you use something good like essentials its fine
1.18 literally isn't released yet
i wrote my own to store 20 static locations just for 1.18
I meant the API, sorry,
and any legit survivals have huge worlds impossible to pregen all of it lol
I meant to the person asking if it was good to use yet
i've used it to check how broken it is but that's about it
yeah a good rtp will recheck a few locations so there isnt any performance drop when requested.
am I the only one who thinks 1.18 cave entrances are too large?
Imagine with 10+ people flying with elytra into new chunks at once XD
Is there a 1.18 change log or api javadoc?
Well, I don't know if it's just me/the client, but chunks load more lazily in this version
essentials uses an async location search queue so it isn't very heavy (which is basically what you just said)
I have many complaints about the world gen but most are too busy drooling over it to notice
JRoy rewrite 
I've published the javadoc jars, but, no javadoc site yet cos that depends on the actual thing
1.18 actually running really good
changelog otherwise, see the commit logs from upstream mostly
it looks like someone went over with a worldedit smoothing brush
Github?
our API is generally stable outside of stuff coming out/in during updates
Stop mentioning
there's not very much flat terrain in 1.18 tbh, is what I've noticed
javadoc jars are on maven
^^^^
it's a shame cause you get these mountainous "plains" biomes
It bothers me
yeah I wasn't expecting my 1.17 plugin to work flawlessly right away
The perfectly circular holes get me the most, and the parts of the world that look like it was blown up with 500 tnt
i still use a 1.16 plugin... tested in rc-3 paper and works flawlessly
Look inside the caves with night vision, it is disgustingly ugly lmao
people will get used to it sooner or later
i hope terralith will make 1.18 worlds a bit better
the shader is also very good
Agreed
Lets make a plugin which puts blocks in a world!
What do I name it,, errr...
I know, a part of the eye!
I get 125+fps with sildurs extreme with Iris
can't wait for tomorrow's 1.18 release... -60 players
My server won't have it until 60 plugins work lmao
we only have like 40
a lot of big plugins didn't work on rc-3 like essentials/deluxemenus
but that will probably change
Yeah I am not doing crap without essentials lol
I'm tempted to upgrade all the open source plugins to 1.18 myself just to release quicker
My game is mostly vanilla (backend plugins) so we don't use fake currencies etc
i want to get rid of NMS plugins asap
i only have essentials, plan, dynmap and bluemap
Can't wait to see many server update on day one.
Free bug reporting and a bug reduced experience in a week. ❤️
I stopped using maps. It kills the survival experience of exploring and finding peoples bases etc
What I'm seeing is a big delay when doing a RTP while the server generates the blocks below y=0. --forceUpgrade doesn't generate those blocks, so it happens after the RTP.
But I do a vanilla style, obviously that isn't for a lot of people lol
force upgrade just loads the chunks and runs them through DFU
Like, it does transformations but there is a good chunk which occurs during placement
or, well, when it's actually loaded as the post upgrade logic isn't ran through --forceUpgrade
and while it's doing it you say "DFU" to yourself over and over , as in "don't f*ck up" lmao
you can use map without update tiles and no player tracking
Data Fucker Upper
It's Done fuck up 
that happens with any random teleport to new chunks, if you want to optimise it --forceUpgrade or pregen (otherwise just wait for it to generate/upgrade as normal)
lol
i never recommend --forceUpgrade lol
Imagine pregenerating when your world is 1 million border or more
I think that there was a way to force it to load the chunks even if they're already gen'd?
This is with a pre-generated, and upgraded world.
generally you pregen a small "hard" radius and let people gen past a few k
cat, I've used it to proc DFU in minor updates even lol (probably overkill, but just to get rid of stupid maplike errors etc)
pregen that far would be a world killer
@violet wing selling custom worldgen pre-set when (?) partner with some artist from MC that's a million buck idea!
NASA level hard drive space needed lmao
did anyone try running a mc server on M1 chip?
But yeah I usually pregen 5k in all directions
Honestly the DFU recode has led it to be PRETTY fast.... tbh it's not rly needed to force update as much.
One thing I do find the modding side does better is custome dimensions and their custom structures
Kinda amazing that one person upkeeps an entire recode of the DFU x)
pretty lacking on the bukkit side.
kinda like optifine
Yea, that's the issue with worldgen
vanillas worldgen has moved substantially since that API has been made; here and there the odd things are added but it's mostly been stagnant
One of the many APIs I'd be happy for a replacement for

Mojang's implementations of these are, well, terrible
They constantly break them too
I'm just happy paper and the forks exists and do their thing. Even when I play singleplayer I use it because it's so smooth vs the garbage choppiness in vanilla
In vanilla 1.28 mining tree logs was intermittently freezing me up lmao
1.18
I really wish they'd redesign their bug tracker: it's so hard to find issues and know if you should submit or if you are wasting your time with a duplicate (likely)
It's jira
mojang ain't gonna make their own bug tracker from scratch
Exactly lol
and, well, nobody has a good bug tracker for laymen
they need github as a bug tracker
githubs issue tracker sucks ass
That's not even any better
Jira 
You would think they would want an 8 year old friendly version with the amount of people who play to weed out the "I bought the new Star Wars pack and Yoda is missing"
it's better than nothing but their issue tracker is missing much stuff
The issue is that nobody has a good bug tracking system "designed for the masses"
Jira is the industry leader.
Because the masses usually don't have a high enough IQ to file a proper report
For internal use yes it's amazing
most companies which do basically have a support team whom fill issues from the forums, and then there is 0 tracking for individual users bar hoping that one of the mods will go look at the filed issue and provide an update from the info there
I wonder how a mix of the two systems would work. A forum where users can report issues, and an issue tracker that only trusted users from said forum can open bugs on.
On the behalf of end users
and can people please get over the "it's off by one pixel" thing. Jesus who cares lmao
It doesn't, basically
Go see the bug section on the forums of any large software vender, pretty much
They're usually a gigantic disaster
What I mean is that what if there was a team that opened issues when the support post actually uncovers a bug, instead of misconfiguration or user stupidity
Like, whenever I hit an issue with vivaldi I often go see their help/support area and basically you'll often see some post from some time ago where it's like, "I've created a ticket and this is tracked with XXYYZZZ tracker number", theres 0 ways to subscribe to the issue, 0 ways to see where the issue even is in the scope of things, etc, etc
In this day and age trackers should be able to say "I see you've submitted blah blah, please see the following tickets to see if your issue fits any currently open tickets" based on key words
Like autonomously?
It's definitely doable, but probably not cost effective for anyone who can't afford to run something like GPT-J locally
I can't defend costs for anything being too extravagant when Moneysoft is involved lol
That's true
not every company wants to create their own issue tracker from scratch
That too
Best selling game of all time, still selling. You had a single person doing most of the massive terrain recoding
Why lol
some ticket systems attempt to link you towards existing stuff but it's generally useless when most tickets are filed by end users
because as your team size goes up, so does the overheads of managing said team
Something no manager anywhere has ever said.
That says more about the manager than anything lol
Managers think a baby can be delivered in one month if you assign nine women to the job.
Issue is more particular to code development than mangement in general
Lol. I would say most managers, MOST, from most companies are hard working and pretty good. It's the leadership outside of operations who are detached from reality
you can't really have two people working on the exact same code efficently in a manner which isn't them just sitting at the same desk
Like, unless you wanna be consistently doing git merge every few hours, etc
I'm sure tho there are many minor things they could take off the main persons shoulders. Like "I'm moving on, these peaks generate too high and flat, play around with the numbers until it looks right" lol
Most managers are promoted to the level of their incompetence.
Well, yea, but nobody else can touch those json files while you're toying with them
You easily end up in a case with coding in proxity around systems that you end up introducing so much headaches just tryna tag team the system together that it's easier to just say, sod it, you do your stuff, I'll go work on something else or twiddle thumbs unless you want me to come over there and peek
Yeah but it's doable with proper management. I doubt GTA has one coder at a time lol
I look at credits for games, and I'm amazed at the number of people working on them.
I worked for a big company, and we never had more than 10 software engineers on one project.
I mean, it boils down to how you split up the work
Like, you'll often have a team dedicated towards working on the actual game engine, and then you'll have another team working on the games wider scope of stuff itself, and then you'll have teams working on one aspect, another on another, etc
It all boils down to how much the code is seperated out and limiting the work you do
if you submit a 20k line diff, you'll probs be told to fuck off unless it's something which was massively co-ordinated beforehand
yeah lol
Grand Theft Auto V was developed by more than 1,000 people across multiple Rockstar Games studios, Rockstar North president Leslie Benzies told Develop in a new interview.Oct 7, 2013
Wow lmao
1k loc diff is a fuck of reason.
I doubt that is all programmers but phew
Minecraft be like nope I can't use it my "GOTO LINE 6789364797" will break
lmao
I had one manager who spent two years trying to fire me. She outright lied on my performance review.
Yeah there are bad ones no doubt. But I've worked for more good ones than bad myself
Then there was the manager who lied most of the time. But you knew he was lying and he treated us well, so it was all ok.
Some things they are told to lie about from the next level up tbh
Then there was the manager who became VP who would make sexist/degrading comments about women, but called several people into his office to complain about email messages they were sending to each other (lewd jokes).
what in the world did i come in to
That VP retired years ago, but he's been charged with 6 felony counts for fraud in Florida.
lmao damn
He was actually a pretty good boss.
We have been intwining leadership/minecraft/programming/asshole manager conversation all in one
lol
He was very accessible, and if you had a problem, he'd get it solved.
Unless you were a woman
F lol
There was a rat in one of the Navy labs we were working at, so he stomped on it (the rat) with his shoe.
I mean, it's part of why HOAs are a good idea at first
The sad truth is is that most people who want to get into places of leadership are generally narcisistic fuckheads
Like I said earlier, the worst part of it starts at the level of leadership directly detached from direct operations of whatever field it is. VP, director, etc
With HOAs, some of them take the job just to get their pet projects done.
The guy I worked for who was charged with 6 felonies - It was for forging ballots for his condominium association election!
The Peter principle is a concept in management developed by Laurence J. Peter, which observes that people in a hierarchy tend to rise to "a level of respective incompetence": employees are promoted based on their success in previous jobs until they reach a level at which they are no longer competent, as skills in one job do not necessarily trans...
The company I work for (cable provider) has a call center in South Africa. They were in civil unrest and closed the center, and an upper leadership called my VP and told him to call them and "get that fucking call center open"
Ok I'll stop hijacking the channel now lol
Has anyone done any back of a napkin benchmarks on chunk generation performance vs 1.17?
My gut just says "it's slower" but I'm curious how much slower.
It's about 40% the 1.17 rate
I saw "20%"said here earlier, but obviously needs to be taken with a huge grain of salt.
Leaf's quick testing showed 90 cps vs 230 cps on 1.17
Welp, I'd more trust what Amaranth said.
When he first looked at it he got 30 cps and it tanked the main thread so... at least it's not that bad 😄
Yes.
Not sure whether he tested on Paper or Fabric?
Fabric will get an idea of what we can expect once Paper applies all patches, in comparison to previous Paper.
iirc Paper is already 3x faster than Fabric
Not that "Fabric" means anything, really
Comparing 1.17 Fabric to 1.18 Fabric should get rough idea of 1.17 Paper to 1.18 Paper.
Well I think anyone who doesn't expect a massive performance drop on any platform needs to wake up lol
I hope paper has a stable version soon :D
They are doing awesome
I will def be up and running on my survival server far sooner than I thought. Even if I don't know exactly when yet lol
Any areas they are concerned with that play testing could help?
just navigating the world last night for 30 mins, it was running great. Chunks were doing quite well (with just me)
Altho I got shorted some coal 😉
oof.
so uh we arent getting paper 1.18 until like next week or?
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.
is the early versions of paper 1.18 for the snapshots or the full 1.18
Please actually read the posts in the channel
hint: The fill version is literally not released yet.
It's only half full at the moment
There most be a cat laying on your keyboard, nice typo 
I'm like, half dead
arent we all
ive been running 1.18 on my raspberry pi (before you say anything it’s the 8gb one and 1.17 runs flawlessly) there’s still some obvious performance issues but there hasn’t been anything that’s caught my eye, i’m gonna do some testing later
1.18 is slower than 1.17 and the CPU is the weak point on the pi, not the RAM.
yeah ik
ranned rc 3 on arm with 8gb ram, ran fine for a small player server
but it’s still pretty good tbh
yeah it’s a small player server
i did some optimizations to the spigot.yml and such so that saw some improvements
Cat when are you not half dead
Schrodinger's cat
https://builds.enginehub.org/job/worldedit?branch=snapshot-testing/1.18
if you're going to run a snapshot server jar, why not a snapshot plugin jar?
Still shouldn't run it in prod, but...
i mean, that one doesn't build lmao
blame paper :P
need to build a local devbundle
5984 built and is working in fabric at least
1.18 paper
fabric
🤦♂️
yea yea, but they use one jar
no they don't
jesus i;m blind
and even if they did the fact that the fabric part of the jar runs on 1.18 doesn't mean the hypothetical paper part would
(in fact, fabric and forge are bundled in the -mod jar, so those builds actually work on both fabric 1.18 and forge 1.17)
doesnt mean it doesnt either unless someone tested it and it said no, but people in this channel are likely good enough to expec thtings to maybe work and want to dive in early
bruh
mfw
Very confused rn
kyad, please read the last commit on the site you just linked
yup, got it
understood wiz, my bad, i'm tired
mhm
Wait, why would an offline build work but not an online build
it wont, i didnt read enough, sorryfor getting your hopes up
Need a sad paper chan emote
cuz paper's published devbundle is broken
so you need to build the devbundle manually/locally
I did push a new bundle a few hours ago
That just gives me more questions
Is the published dev bundle different from a local one?
I literally updated the published dev bundle earlier
So the issues should no longer be there
what the shit is io.papermc.paper:paper-api:task '.
there's a space in ur artifactname!!!!!
wat

Which class is that in?
not sure why you think that's a class?
So is there only 1 small error with the dev bundle? And everything else should be okay
¯_(ツ)_/¯
Which lecture is that in?
Stuff is still being tested and some stuff still isn't updated/ported over. So - no. TBD. No eta.
okay thx
that's the final exam
kudos to the team and contributors for your hard work recently
dont forget to take some time for yourselves too. take care :D
lord Kenny
This is what Kenny does instead. He is not the same.
I've noticed in the RC3 server jar my TPS plugin doesnt work, anyone lese have this issue?
ensure your plugin is updated (very likely not coz 1.18 is literally not release yet)
make sure that you're using the correct jar
I just clicked the link for the download
spigot also dropped some method names, so, plugins using internals will break
there was 2 jars
I used paperclip
make sure you got the one that was named reobf
do you think it would be better to just run the mojang server jar until paper and all other plugins are updated?
and I can just add them slowly?
its better to not run anything until paper is stable (coz even mojang jar can still have bugs)
can you use plugins with the mojang jar? no right
nope.



