help-development

[slender elbow]

what is that implying?

[sly topaz]

actually nevermind, I didn't think that one thorough lol

[rough drift]

can you generate chunks async using World#getChunkAt(int, int, boolean) in another thread

[sly topaz]

nope

[ancient vapor]

https://www.spigotmc.org/threads/scaling-and-rotating-a-display-entity-between-two-points.674095/
Could someone give me the complete code to achieve this?

[kindred sentinel]

Which event is fired on wither break block?

[blazing ocean]

say I have a string of an interface name, what would be the best way to check whether it's following the """convention""" of prefixing it with an I? Can't just check whether it starts with an I since it could be called something starting with I

(xy: writing a linter rule)

[slender elbow]

check that a) starts with an I, b) the following letter is uppercase?

[blazing ocean]

that is smart yeah

did not think of that

[slender elbow]

but afaik I interfaces are not convention in JVM world lol

[blazing ocean]

that's why I said """convention"""

[slender elbow]

that's a .NET thing for the most part

[blazing ocean]

I've seen people use them a lot and hate it

so, writing a linter rule

[sly topaz]

BlockExplodeEvent probs

either that or EntityChangeBlockEvent

[kindred sentinel]

Yeah prob it's entity change block event, ty

[blazing ocean]

[slender elbow]

i feel you 🫶

[mortal hare]

weirdo is back 😄

lets say im trying to implement IContainer and IMutableContainer. Is it better to implement IMutableContainer and then downcast it to IContainer when returning IContainer from a method as a return type?

or is it better to extend IContainer implementation and implement IMutableContainer methods in a separate extended class, returning appropriate implementation in the use-case?

returningIMutableContainer and then downcasting it to IContainer sounds like less of a hassle in terms of maintaining it, but im not sure if there's any cons of it

[mortal hare]

here there's bunch of I prefixed interfaces for you 😄

[remote swallow]

IAmSteve

[robust helm]

what do i do if my gradle just breaks every few hours after changing something in the build.gradle.kts?

i could use maven, but with kotlin it kinda sucks

like the whole build file gets read and it complains with different errors, then randomly the next day it works again without any changes

[blazing ocean]

cry

[robust helm]

thought so

[manic delta]

just restart intellij

[robust helm]

thats what i did for 2 hours yesterday

• deleting all caches

[manic delta]

or click this button

[robust helm]

also did

[manic delta]

how many ram do u have

[blazing ocean]

that's really not related

[rough ibex]

??

[manic delta]

really?

[rough ibex]

Why would that matter

[manic delta]

That happened to me when I had a 2 RAM PC always

[rough ibex]

What do the errors look like

[blazing ocean]

"2 ram pc"

[rough ibex]

lol 2 rams

[manic delta]

Now I have my laptop and it doesn't give me any problems

[rough ibex]

haha i have 32 rams

[manic delta]

me 16

[rough ibex]

(I am making fun of your usage of "ram")

[robust helm]

Cannot access script base class 'org.gradle.kotlin.dsl.KotlinBuildScript'. Check your module classpath for missing or conflicting dependencies
this was the last one, but i had others too

and i reinstalled a lot

[limber apex]

Hello I'm looking for a person who knows minecraft to make a software related to a launcher (I pay of course) dm

[blazing ocean]

what version are you on

[manic delta]

maybe his intellij is just dumb

[robust helm]

2024.3.4

[blazing ocean]

I also have a 2 ram pc (I have two sticks of ram)

IJ is not related to gradle

what gradle version are you on

[echo tangle]

is the only way to change someone's nametag (the name above their head) hiding their original one and summoning a textdisplay to follow the user

[blazing ocean]

yes

[rough ibex]

https://youtrack.jetbrains.com/issue/KTIJ-893 ?

[limber apex]

1.9.4

[blazing ocean]

not you

[limber apex]

oh okay

[manic delta]

1.21.1

[blazing ocean]

also not you

[rough ibex]

?services

[undone axle]

If you wish to request or offer development/art/building/administration services, please do so at https://www.spigotmc.org/forums/services-recruitment-v2.54/

[rough ibex]

lmao

[manic delta]

why not me

[rough ibex]

because not you

[manic delta]

damn

[limber apex]

yes ?

oh okay

thanks

[robust helm]

flipping deleting caches again worked now??

i might be going insane

[rough ibex]

Happens

[robust helm]

but i 99% sure that tomorrow morning ill have the same problem

[manic delta]

probably

[robust helm]

and now my dependency is still not loaded wth

[blazing ocean]

again, what gradle version are you using

[robust helm]

oh didnt see your msg. 8.8

[blazing ocean]

that's ancient

update to 8.13

[robust helm]

just change the gradle-wrapper.properties file right?

[blazing ocean]

./gradlew wrapper --gradle-version 8.13
./gradlew wrapper

[robust helm]

oh ty

[robust helm]

it flipping crashes after invalidating caches

like after restarting

[blazing ocean]

okay lemme make something sure real quick

run ./gradlew tasks and lmk if it errors or prints the task list

[robust helm]

prints the tasks list

[blazing ocean]

so it's an IJ issue

[robust helm]

maybe i should switch ide

[blazing ocean]

embrace neovim

[robust helm]

cuz these issues arent a rarity

[blazing ocean]

embrace not having a working LSP

[robust helm]

nvim has lsp

[blazing ocean]

embrace there being zero good kotlin LSPs

[robust helm]

Fuck

[blazing ocean]

you can try fleet but it just uses IJ as a backend

[robust helm]

hmm

scala instead of kotlin?

[echo tangle]

how would i make the textdisplay follow the user? manually teleporting is easier going to be too slow or too laggy

[blazing ocean]

add it as a passenger

[blazing ocean]

nah

[echo tangle]

oh, thank you

[robust helm]

or i could js give up java with all its crap

[robust helm]

its back...

[rough ibex]

youre blaming java for kotlin?

[robust helm]

blaming the java world

[rough ibex]

does java gradle not work either

[robust helm]

oh i meant to say give up everything with jvm

[robust helm]

it works better but also not perfect

well anyways. thx everyone

ill probably wait for some updates, god or learn java so i won't have to deal with kotlin/intellij

[manic delta]

damn

[hybrid spoke]

give up everything

[robust helm]

yea maybe i should move to asia

and farm rice or smth

[mortal hare]

is there any data structure in java which is like linkedlist, but retains its index after some element in the middle of list gets removed.

here's my take on naive implementation of such data structure

class PackedArrayList<T> implements List<T> {
  private final Deque<Integer> removedIds = new ArrayDeque<>();
  private final ArrayList<T> list;

  public PackedList(ArrayList<T> list) {
    this.list = list;
    
  }

  public void remove(int index) {
    if (this.list.get(index) != null) {
      this.list.set(index, null);
      removedIds.push(index);
    }
  }

  public void add(T element) {
    if (!removedIds.isEmpty()) {
      this.list.set(removeIds.pop(), removedIds);
      return;
    }
    this.list.add(element);
  }
  // im not gonna implement whole List<> in Discord.. duh..
}

[thorn isle]

linked int -> object hash map

that's not really a list at that point however

[mortal hare]

what i want is to reuse ids from the elements which were removed

effectively reusing space that removed object took

without resizing arrays and for previous objects to retain that id

[thorn isle]

might as well use an array list rather than a linked list at that point if the goal is to avoid shifting elements on removal

a common approach to this is to shift the tail element into the hole caused by the removal; but this of course means that the id of that element changes and needs to be updated in the associated data structures

generally in java land a list is expected to become smaller (size() becomes smaller) on successful remove, and the last index of a list is expected to be list.size()-1

so what you are initially describing isn't really a list as understood in this ecosystem and probably shouldn't implement List

you are free to create an impl like that though of course, just, don't expect it to work with anything that takes a (mutable) List

[mortal hare]

for example:

0: Object
1: null
2: Object

PackedList#add() places element in index: 1
PackedList#remove(2) set the index: 2 to null and tracks removed id for use inside add() method

[mortal hare]

but you can count virtual size of the list just by tracking add() and remove() methods

[thorn isle]

this would be achieved by set(1, null)

which is different from remove; remove is expected to either fail (with an exception) or change the size of the list

and if the size of the list changes, then the last index in the list should also change

set simply replaces the element at the index with another element, in this case null

[mortal hare]

what's stopping you to count length of the list by increasing and decreasing size field on add() remove() after if guards

basically virtual "size"

[thorn isle]

the issue is that you would end up with a list of size 2 but the last element would be at index 2, which is larger than 2-1

i.e. to get the last element in the list, you would have to do list.get(2)

but list.get(2) is expected to explode, because 2 >= list.size()

so your size would have to remain 3

[mortal hare]

https://tenor.com/view/cat-annoyed-eyeing-cat-meme-cat-funny-gif-14867130448325066479

[thorn isle]

which then means list.remove(1) didn't work as expected; the list didn't become any smaller

tldr use list.set(index, null)

this won't remove anything from the list but will substitute the element in that index with another

[mortal hare]

isnt my example doing just that?

i guess .get() wouldnt work in remove()

[thorn isle]

well yeah but what that is doing is set, not remove

so you can't implement remove to do that

that's breaking the list contract

so like i said at first, you can do that, but don't implement List

[echo tangle]

the textdisplay mounts directly above my head, how do i push it up to match it with my nametag

[blazing ocean]

add a translation

[thorn isle]

add a translation offset to the transformation

will look kind of funny if it also billboards vertically however

since it'll pivot around the player's head rather than itself

[manic delta]

the old method was placing a named armorstand right?

[thorn isle]

yeah, and the old method of moving it up/down was to mount a bunch of them in a pile

[mortal hare]

there's no constraints on javadoc that would mention that List<>.remove() implementation should effectively remove element from the backing collection/array. it mentions that remove() should affect size(), but you can calculate that by adding or decreasing size private field inside PackedList

[thorn isle]

you can do whatever you want with it but don't expect it to work when plugged into anything that expects a mutable list

do you really need it to implement List specifically?

[echo tangle]

yeah this does appear to be an issue lmao

[thorn isle]

a more appropriate data type would be an int -> object map

which can still be implemented with e.g. an array

alternatively you can use a regular old array list and just call set(idx, null) instead of remove(idx)

[mortal hare]

you can implement PacketList#size() by list.size() - removeIds.size()

[thorn isle]

and take this example

the index of the last element is 2, yes?

and list.size() - removeIds.size() returns 2, yes?

[mortal hare]

yes

[thorn isle]

so PackedList::size returns 2

but the last index is 2

this is illegal

that is, to get that last Object, you would have to call get(2)

which according to the specification explodes because that exceeds the list's size

you can make it not explode, but everyone will expect it to explode

e.g. compilers/linters will emit warnings about "index out of bounds exception"

    /**
     * Removes the element at the specified position in this list (optional
     * operation).  Shifts any subsequent elements to the left (subtracts one
     * from their indices).  Returns the element that was removed from the
     * list.
     *
     * @param index the index of the element to be removed
     * @return the element previously at the specified position
     * @throws UnsupportedOperationException if the {@code remove} operation
     *         is not supported by this list
     * @throws IndexOutOfBoundsException if the index is out of range
     *         ({@code index < 0 || index >= size()})
     */
    E remove(int index);

also, that the subsequent elements shift down on (successful) remove is specified in the contract

[mortal hare]

ok

yeah, this sucks

but its ok

[thorn isle]

just don't use a List

because it isn't one

[mortal hare]

well it is, just a nonstandard one

you can probably get away with get() by doing linear search by skipping null values

but this javadoc comment ends this debate 😄

[thorn isle]

it can still be e.g. a SequencedCollection or Iterable or whatever interface you actually might need that does describe what it is, but List isn't one of them

[thorn isle]

well... that would also fix the issue you have with the davadoc comment

because then you can get the last object with get(1) as the null is skipped

and hence the subsequent elements' indices were indeed shifted down by 1; 2 -> 1

index doesn't have to correspond to where the element physically is in the data structure, it's purely an abstract ordering thing

for example in an ArrayDeque, the start and end (0, size()-1) continuously shift in the underlying array

[mortal hare]

Yea

[thorn isle]

removing the first element in the deque doesn't actually physically shift the subsequent elements down; it just moves the starting index up

and the end index can have a lower physical address than the start index because it "rolls over" from the end of the array back to the start

[mortal hare]

But isnt this getting back to square one because the whole point this was to have nonshifting indices lol

[abstract totem]

You are running the latest version```

I just updated and am still facing the same issue

I have found a "fix" involving NMS but it is quite messy

[kind hatch]

?whereami

[undone axle]

https://cdn.discordapp.com/attachments/864180468680360027/944415049462714469/whereami.mp4

[lime stump]

Can I make an invisible entity have a glowing outline?
Like something like this

entity.setVisible(false)
entity.setGlowing(true)```

will this entity be see-through but be visible by it's glowing outline?
Because I'm trying this on an item-frame and it does not have a glowing outline. Although I could be doing this wrong, because for a while, it was the item-frame that was invisible, but a glowing outline would be applied to the item displayed inside the item-frame.

[abstract totem]

i genuienly didint think there would be any differences that major, regardless im trying with a spigot jar now

[thorn isle]

what's the goal here exactly? are you working on like a palette or something?

[blazing ocean]

?tas I'd say

[undone axle]

https://tryitands.ee/

[kind hatch]

You'll have to apply the invisibility effect to the entity.
It'll only work if the entity supports potion effects.

[lime stump]

I've been trying, dumbass.

I wanna know if I'm doing something wrong or if the item frame being set to invisible negates the glowing effect

[kind hatch]

That used to be the case with most things, but nowadays there are major differences in how paper does things compared to spigot.
Compatibility is the main reason why things just kinda work on both setups, but with paper's fairly recent hard fork announcement, there are things that just won't work.
Paper is becoming it's own standalone project.

[blazing ocean]

then mention that you've tried it and it doesn't work

no need to be rude about it

[mortal hare]

curiosity

[blazing ocean]

and also, crossposting in spigot and paper at the same time is crazy

[kind hatch]

https://hub.spigotmc.org/javadocs/bukkit/org/bukkit/inventory/ItemFlag.html
Whoa, lots of new flags are being added.
Noice

[blazing ocean]

tad bit jank

[abstract totem]

wow, yeah i was totally unaware thanks for filling me in, il update you once i get the spigot jar

[young knoll]

That’s a lot of flags

[echo tangle]

would the only way to fix textdisplays with translations rotating around the origin (for nametags) to spawn a clientside textdisplay for each player and rotate it based on said player's pitch?

[thorn isle]

i would just disable vertical billboarding so it only faces the camera horizontally

it's not exactly perfect i suppose but it'll be a lot easier than the alternatives

[sly topaz]

Doing billboarding manually when it is a feature of display entities is a waste and doesn’t work as well when latency comes into play

[echo tangle]

so then attempting to make nametags billboard vertically is pointless?

[sly topaz]

I mean, isn’t it able to do that by default?

[echo tangle]

to make the text display be in the same position as a regular nametag, you need to translate it up, which then makes it rotate around the origin so if you are looking at it from above it's off to the side

[sly topaz]

Ah I see what you mean now, in that case, your only option is to do billboarding manually however I wouldn’t recommend it. Just disabling vertical billboarding sounds like a good compromise to me but I don’t know your setup so I wouldn’t be able to say how necessary it is

[echo tangle]

yeah i see, it is very laggy. so there is essentially no way to set the player's nametag and have it rotate without lag?

[short pilot]

wsg

[young knoll]

Can you use another entity to offset it instead

Maybe an interaction entity

[echo tangle]

i was thinking this, but i don't know what entity i would use to offset it

[earnest girder]

what do I use instead of the deprecated AnvilInventory? I need to set the cost text and stuff

[chrome beacon]

The javadoc tells you what to use

If you're using Intellij you can get them in your IDE by opening the maven/gradle tab and click the little download icon and choosing sources and documentation

Alternatively use the web version

?jd-s

[undone axle]

https://hub.spigotmc.org/javadocs/spigot

[echo tangle]

also, i can't find a way to make said entity invisible without making it's passenger invisible, and i don't think there are any rideable invisible entities

[young knoll]

I mean interaction entities are invisible

So are display entities with nothing set to display

[echo tangle]

you can't use addpassenger() on those

[eternal oxide]

You should be able to. ALL Entities support passengers

[echo tangle]

you're right, but i don't think any of the invisible entities are able to be scaled so that the text display could be higher

[eternal oxide]

all display types can be translated

[thorn isle]

the translation doesn't affect where the passenger goes

you're stuck with an invisible scaled armorstand or something

...or make it not billboard vertically

[echo tangle]

^

[thorn isle]

that's the easiest solution with the least maintenance baggage

[young knoll]

Does the size of an interaction entity affect the passenger position?

[thorn isle]

what's an interaction entity?

[echo tangle]

no

you can use air as a projectile, but it still wouldn't be scalable

[thorn isle]

last i asked this someone told me they're invisible armor stands

[echo tangle]

don't they work in the same way where size doesn't affect?

[thorn isle]

with them you have 3 possible sizes you can use

large, small, and marker

marker is very small

iirc you need half a dozen to make up a row of text worth of vertical distance

putting them together with muh combinatorics will get you somewhere close enough

alternatively, setting teleport interpolation to 1 tick and manually moving the position of the text displays every tick to match the player's should look correct for other players; then use the hideEntity api to hide it from the player themselves

alternatively, use horizontal billboarding only

these are your 3 options basically

[young knoll]

They are for detecting interactions

And can have a custom width and height

[thorn isle]

what's the entity type?

EntityType.INTERACTION huh so there is a specific entity type for them, last i was told they're invisible armor stands

[echo tangle]

how small are they? if i stack a couple small armor stands, the text display is pushed up, but with markers it doesn't move up at all despite how many i put down

[kind hatch]

That was the old method.
The interaction entity was added specifically for this.

[thorn isle]

yes but i asked like 3 days ago or something and was told it's an armor stand under the hood

[young knoll]

Markers can’t be used with passengers

Because they don’t get sent to clients

[kind hatch]

I don't know if that's even true.

[young knoll]

Or do you mean armor stands with the marker flag

[echo tangle]

armor stands with the marker flag

[thorn isle]

it most likely wasn't true

[young knoll]

They have a 0 size bounding box

So they don’t add any offset

[thorn isle]

i don't think they added interaction entities in the past 3 days

[echo tangle]

oh, bruh

[kind hatch]

Ofc not. They were added at the same time they introduced display entities

[thorn isle]

i vaguely remember someone saying that a marker armor stand will still lift up its passenger by some small amount

it's possible they were wrong i suppose, they definitely have zero volume hitboxes

other nonsense has been attempted over the years including invisible silverfish, dropped air items, experience orbs, projectiles, what have you

[echo tangle]

it still looks a bit janky, but i could just do this or the air projectile

[thorn isle]

i'm not sure if it works with a text display directly since they can't have velocity

try mounting it on an invisible armor stand or other physics capable entity

the reason why it looks janky is because the player and the display physics aren't quite in sync, presumably because one has velocity and the other doesn't

[echo tangle]

can i use attributes to scale the invisible armor stand up and down?

[thorn isle]

sort of yes but that won't affect where the passenger sits i don't think

[echo tangle]

it doesnt, i tried

[thorn isle]

yeah you've got to put it on an armor stand

[echo tangle]

oh, small armor stands are too tall

[thorn isle]

and then apply velocity/teleports to the armor stand

i'm not saying to mount the armorstand on the player

the armorstand exists purely so the client can tick velocity + physics on it so the text display that's mounted on it doesn't desync with the player

[echo tangle]

ohh, you mean mount the text display on the armor stand to sync. so i would just set the velocity of the armor to 0, and teleport it every tick to the player?

[thorn isle]

you'll want to copy the player's velocity to the armor stand rather than setting it to 0, but yes that's the idea

then every tick teleport it to some offset from the player (to get the height right)

e.g. offset by +1.9 on the y axis or whatever magic number looks right

[echo tangle]

oh, alright, i'll try that thank you!

[wide viper]

is there any way to remove components from items like consumable component from food?

[eternal night]

no, you'll need NMS || or paper api ||

[wide viper]

oh unfortunate, thank you

[blazing ocean]

or WHAT api!?

[eternal night]

idk what you are referring to

[young knoll]

https://tenor.com/view/sniper-pubg-loading-sniper-reload-gif-4868344854109813711

[echo tangle]

i can't manage to teleport the armorstand, i don't think you can teleport entities with passengers, or atleast i think

[earnest girder]

InventoryOpenEvent:java AnvilView view = (AnvilView) event.getView(); this is causing this error:java.lang.ClassCastException: class org.bukkit.craftbukkit.v1_21_R3.inventory.CraftContainer$1 cannot be cast to class org.bukkit.inventory.view.AnvilView (org.bukkit.craftbukkit.v1_21_R3.inventory.CraftContainer$1 and org.bukkit.inventory.view.AnvilView are in unnamed module of loader java.net.URLClassLoader @77a567e1)What am I doing wrong?

[kind hatch]

Not currently. I think there is still an open PR about it.

You can dismount the entity, teleport the main entity to where ever you need it, then remount the passenger.

[chrome beacon]

Are you sure that's an anvil inventory being opened

[earnest girder]

yeah. I'm getting the same error in my InventoryClickEvent when I click in the anvil in game toojava Bukkit.createInventory(null, InventoryType.ANVIL, ChatColor.RED.toString() + ChatColor.BOLD + "Enter Name");

is event.getView() not the right way?

[echo basalt]

any OG aware of why this same exact code causes the horizontal slices to be "inverted" on 1.8?

Same exact code, happens consistently across all images and restarts

Facing the same exact orientation and this happens at the same coordinate

I've also gone ahead and checked, the coordinate direction system is the same, yaw 0 is when facing south and that's +Z

Might have to do with what direction it faces but even then it's a bit weird

because this makes sense for 1.9+ where item frames can look up/down but not on 1.8? not sure, will mess around with

[slender elbow]

You've reached the far lands

[young knoll]

Ah yeah that’d do it

[echo basalt]

hm

how is itemframe rotation set on 1.8?

because passing in that short or whatever on the spawn packet seems to do absolutely nothing

and the image could actally work if we rotated every item frame to face south instead of north

please tell me it's not yaw/pitch based

ah shit I need to look at obfuscated 1.8 code

been years since I've done this I feel old

[slender elbow]

my condolences

[echo basalt]

ah shit CraftEntity#getEntity is ass

@alpine urchin I suspect this is a packetevents bug

the spawn entity packet's straight up ignoring the data

which it does use in later versions.. so why not the older one

might need to set a velocity

[fierce salmon]

I am making a plugin for my server that times a parkour course and lists the top 10 times on a leaderboard. Would it be faster to store the times in a yml file within the plugin folder or should I use a MySQL database to store the times?

[earnest girder]

yml would be fine

[young knoll]

I mean, you can use both async which eliminates most of the concern

So it doesn’t really matter

[fierce salmon]

MySQL would be more work but it would look cooler imo

I might just do MySQL

[ivory sleet]

The nice thing w mysql is that u can sql to query top 10

[fierce salmon]

true

[worldly ingot]

Issue with YAML is that you have to load the whole configuration into memory

Then reading and writing becomes slower and slower the more entries you have

[fierce salmon]

Right

[worldly ingot]

Bare minimum you should be using SQLite, MySQL would just be a nice optional thing if you want a remote database

[fierce salmon]

after 1000 entries it'll be pretty slow im guessing

[worldly ingot]

Depends on your system but either way it has to parse through 1,000 lines and validate it as YAML, :p

A proper database is just better suited for lots of repetitive structured data

[kind coral]

I have a question regarding the premium resources, for the past year i have been developing a plugin mostly aimed at prison servers, which is in really high demand, my only problem is with Spigot, they don't directly allow Obfuscating your code, (because i want to publish my resource on multiple marketplaces), so i was left with the decision of publishing first on other platforms and then some time later on spigot, but with an older version,

this is because i really don't want my plugin pirated on the first week and dropping my sales, considering i have worked on it a ton.

is there any other way? like getting an authorization to use some special obfuscation tools? I am planning on using Zelix Klassmaster,

TL;DR
I want to obfuscate my code, spigot wont allow it, i publish my resource on other platforms (which allow obfuscation) and delay as long as i can updates on spigotmc (or i get authorization to use obfuscation tools)

[echo basalt]

still waiting for the day there's a proper db abstraction made for sql in java

something like jetbrains exposed

[worldly ingot]

I mean there's jOOQ if you really want

I don't know how abstract it is to different DB engines though

[echo basalt]

exposed works for different db engines

I should make a java port some day

but it's such a monumental task

eh jooq looks like a java exposed thing yeah

[sullen marlin]

Rule 4.1.6.1.1.1

[wet breach]

obfuscation has never prevent pirating

MC is a prime example of that

MC is even to this day is obfuscated and despite that custom server implementations have still come about regardless

and consistently updated

[left jay]

ok i have an odd error going on
so i have this method in a class in a plugin that creates a key value pair in an items meta, and a method that checks for the value of that pair.

in another plugin that depends on the first one, i have a method that does something based on the value of that key-value pair. however when i try to do this via the method from the first plugin, this plugin doesn't regonize that the key-value pair exists at all in the items pdc, but the original plugin does.

im wondering if theres anyway item metadata could leak in between plugins or if pdc changes can only be regonized by the original plugin.

[remote swallow]

Is it the same key with the same namespace

[sly topaz]

kinda late but you're supposed to be using MenuType#create method nowadays, not that one

and that returns a view

[wet breach]

PDC is plugin specific in their k,v pairs. In order to access the k,v pair of another plugin for pdc you need to obtain the plugin's instance in regards to the pdc. So all you need to do is get the plugin instance of the one you depend on and then using that instance get its pdc

[remote swallow]

Could also just use fromString

[worldly ingot]

Or, more abstractly, have a method that doesn't require a key at all, just a utility method that hides the key as an implementation detail

[remote swallow]

That too

[wet breach]

I like Choco's suggestion better, but Epic's method should work too

[echo basalt]

if I'm doing a core plugin and let's say a secondary plugin I just use the core plugin for all the pdc

(I have pdc wrappers that just do that for me)

But yeah your issue is that while the string keys are the same, their namespaces are different

if you don't care just use minecraft's namespace for everything

[gaunt pollen]

Good resource for learning nms code?

:D

[echo basalt]

uh

Just run mappings on it and explore whatever

I wrote a guide forever ago but it was really superficial

Bit hard to teach someone on how to navigate a codebase

[earnest girder]

Thanks

[drowsy helm]

Nms isnt really something you can learn explicitly from a guide. Set a goal to achieve and then what nms can do to solve it

[gaunt pollen]

Ic

Very new to nms so I'm pretty much clueless

[echo basalt]

I'm in bed rn thinking about how I'd make an entire video series around different nms concepts

[drowsy helm]

Are you trying to achieve anything specific?

[gaunt pollen]

Hm... Right now I'm relearning all the prerequisites so that I can start speedrunning my plugin.
But a few years back, I remember I was working on a player tag plugin where I wanted to set the player tag blue colored for only the red team and vice versa. I think my only option was playing with nms code back then, but I don't quite remember the specifics. Sorry, it's a bit vague.

[drowsy helm]

Nah makes sense. You’re sort of in luck as we have moj mappings now which makes nms a cake walk

It’s sort of just a matter of finding and sending the correct packet

And sometimes intercepting a packet with a listener, but for that I would recommend a library like PacketEvents

[gaunt pollen]

During my break from bukkit, I had to deal with datagram packets while I was working on a game server in java 🤢
So maybe I won't fully be clueless while dealing with nms this time

[gaunt pollen]

Sure thing

I'm a bit confused on where to start

Do you have any starting point recommendation?

For now I'd say particles

[drowsy helm]

Sec I’ll get some links

https://mappings.dev

[gaunt pollen]

oh right nms is veery version dependent

[drowsy helm]

https://minecraft.wiki/w/Minecraft_Wiki:Protocol_documentation

[drowsy helm]

With Moj Mappings it doesnt vary much

The wiki might be a lil outdated for 1.21.5 but should be mostly same. I usually just Ctrl-f keywords im looking for and can usually find the packet I want

The problem with the wiki is that the names they give for packets usually dont match with the spigot names

[earnest girder]

how should I use MenuType to create an anvil view? MenuType#create doesnt seem to exist

[minor otter]

In order to bypass a shield (blocking) I check if the player is blocking on entitydamagentity then player.damage, however this retriggers the event and that can lead to some issues so instead I can change the player health and play a sound but knockback/armor and damage reduction is lost unless I simulate it. Is there a more clean way to do this?

I could also add a cooldown to the shield to prevent the event from constantly re triggering but that defeats the purpose of the system im trying to make

[worldly ice]

i believe you can do something with damage causes to remove the blocking from a shield

damage modifiers*

iirc it's deprecated, but there's no other alternative besides what you said

[worldly ice]

using https://hub.spigotmc.org/javadocs/bukkit/org/bukkit/event/entity/EntityDamageEvent.html#setDamage(org.bukkit.event.entity.EntityDamageEvent.DamageModifier,double)

if not you could flag the player before you apply the damage, and in the event handler just do an early return if the damaged player is flagged

[cursive kite]

What sound are people using for double jump I cant figure it out

its like a launching sound

[minor otter]

I'm upset I didn't think of this before posting that's definitely an easy way of fixing my problem

Thanks

[minor otter]

Have you tried entity.ender_dragon.flap

[cursive kite]

I thought it was another

There is one that sounds a bit more like something launching

[kind coral]

i mean it would make the work totally harder, since i have a license system and i don't want to load my classes over HTTPS lul

[warm mica]

Your plugin has to work offline without the reliance on any license otherwise you can't sell it on spigot's platform

[kind coral]

yeah i know, thats why i would delay updates on spigot

since its the full resource and obfuscation doesn't do the job 100%

[thorn isle]

nothing will, even if you load them in over https from a remote server owned by you and they never touch the user's disk, someone is just going to dump them from memory, put them in a jar, and upload them on some black market site

[chrome beacon]

Only way to get around it is to host the servers yourself

And not let users choose what to install

[thorn isle]

all it really achieves is making it higher-hanging fruit (more money for the person cracking it; paradoxically more demand for it to be cracked) and making your end users's lives worse for relying on your remote server and licensing and not getting legible stack traces etc.

[chrome beacon]

Some crackers might even see it as a fun challenge

just cracking it to see if they could

[thorn isle]

tldr bake in a license number or something and make it difficult to identify & remove, and if you see it being leaked, pursue legal action with the buyer that license is associated with

crackers generally are too dumb to notice and/or won't give a shit about the baked in license number

[chrome beacon]

license can be replaced

but yeah as long as it's unique ig it's a good enough identifier

[thorn isle]

they can, but that assumes the cracker knows its there and knows how and cares to

[gaunt vortex]

hi i am making anticheat and i use ProtocolLibrary. i dont know where is some documentation on packets and how to read data from the packets.

[thorn isle]

https://minecraft.wiki/w/Java_Edition_protocol

https://dev.bukkit.org/projects/protocollib/pages/tutorial

[gaunt vortex]

PacketType.Play.Client.GROUND i want to get data from this packet

[thorn isle]

no clue what the actual name of that packet is

what data does it have?

[chrome beacon]

If this is your first time working with packets I do not recommend making an anticheat

especially so if you need us to spoon you how to read that

[gaunt vortex]

its wierd. it was always true and was sended only sometimes

[gaunt vortex]

i worked with packets before. i made hack client in java and bot using socket in python

[chrome beacon]

I was mainly talking about spigot serverside packets

[gaunt vortex]

i think i got it working. i only needed to use POSITION_LOOK packet type

[nocturne mirage]

I have an texture pack error someone who can help i will pay!!!

[chrome beacon]

First off don't DM random people. Include your error in the request for help and use #help-server

[smoky anchor]

I'll take 5€

• texturepacks are no longer a thing, it's a resource pack
• do what Olivo said

[nocturne mirage]

Im sorry guys

[wet breach]

just like what is done with MC, all it takes is to just replace the classes and methods involved with your so called licensing. And it isn't really that much effort to do it. Also you wouldn't even need to deobfuscate to do this either. Again, we have minecraft which contains hundreds of classes and it stopped no one from deobfuscating it and making implementations, why do you think your small plugin is any different?

[left jay]

hey i dont understand what either of these suggestions mean
i only kinda understand how fromString() works and dont see how it works cross plugins
and i dont understand what choco means by hiding the key

[rough drift]

final var mapStack = new ItemStack(Material.MAP);
final var mapMeta = (MapMeta) mapStack.getItemMeta();
assert mapMeta != null;
mapMeta.setMapView(view);
mapStack.setItemMeta(mapMeta);

itemFrame.setItem(mapStack);

aused by: java.lang.ClassCastException: class org.bukkit.craftbukkit.v1_21_R4.inventory.CraftMetaItem cannot be cast to class org.bukkit.inventory.meta.MapMeta

Am I forgetting how to make map items?

ignore

got the wrong material

[wet breach]

seems like you got some more researching and learning to do then 🙂

good news what you are wanting to do is possible

[chrome beacon]

FILLED_MAP

[rough drift]

yeah realized .6 seconds after posting

[chrome beacon]

Happens every time 💀

[rough drift]

works

there is a teeny tiny magenta pixel

some more magenta

[glossy laurel]

Not necessarily spigot related, but I want to create a single thread which would handle the code of a method that gets called. How would I approach this?

[blazing ocean]

new Thread(() -> {
  // ...
}).start();

?

[thorn isle]

most nebulous question challenge (impossible)

[warm mica]

Or use a single-threaded executionservice, if the task is short lived and frequently called

[slender elbow]

Executors.newSingleThreadExecutor().execute(() -> {
  // ...
});

like this, yes?

Clueless

[eternal night]

[slender elbow]

thanx bb

[sly topaz]

MenuType.ANVIL.create(player, "title")

note that this just creates an inventory view, you still have to open it yourself with HumanEntity#openInventory(InventoryView)

that way you can take an inventory view and place items before opening it though, which is nice

[glossy laurel]

How do I tell the thread to handle the method when it gets called tho

[sly topaz]

what are you trying to do

because what you're describing seems like a solution to a leading problem rather than the actual thing you're trying to solve

[glossy laurel]

Put stuff on a worker thread

[sly topaz]

but why, what do you need to put in a "worker" thread

[glossy laurel]

Not to use main threads resources

[sly topaz]

are you being dense on purpose, what are you not trying to run in the main thread, a calculation? a database query?

[slender elbow]

mine crypto, ofc

[glossy laurel]

^

Jk, lines to put on scoreboard

[sly topaz]

putting lines in a scoreboard can't be done asynchronously, at least not with the API

[glossy laurel]

Yeah im calculating them

Async

[blazing ocean]

then use the async scheduler

[sly topaz]

ah so you mean the strings that go into them are calculated asynchronously

[glossy laurel]

Yea

[sly topaz]

you could use a CompletableFuture#supplyAsync then do thenAcceptAsync with BukkitScheduler#runTask as Executor

[sly topaz]

I was contemplating whether to whereami

[glossy laurel]

What does thenacceptasync do??

[blazing ocean]

BukkitScheduler#runTaskAsynchronously

[sly topaz]

public static final Executor MAIN_THREAD_EXECUTOR = task -> Bukkit.getScheduler().runTask(YourPlugin.instance(), task);
CompletableFuture.supplyAsync(() -> {
  /* Calculate scoreboard lines */
}).thenAcceptAsync(lines -> {
  /* set scoreboard lines */
}, MAIN_THREAD_EXECUTOR);

[glossy laurel]

Bukkitscheduler#runtask as executor??? What??? You can do that??

[glossy laurel]

Okay so it thinks that it accepts async

But its secretly the main thread?

Xd

[sly topaz]

the issue with that is that going back to the main thread schedules another task and it's kinda annoying

[blazing ocean]

heh I just use coroutines

[slender elbow]

yet another day of the word "async" being misinterpreted by people in the spigot ecosystem

[sly topaz]

I think I am having a dejavu about this topic lol

[sly topaz]

in this case thenAcceptAsync is refering to itself being asynchronous from the previous step in the future, not necessarily from the main thread, supplyAsync is asynchronous to the main thread and so the reverse is the same for thenAcceptAsync

[slender elbow]

"but async means not on the main thread, huh?"

[thorn isle]

isn't there a BukkitScheduler::getMainThreadExecutor(Plugin) specifically for this?

[wet breach]

this is probably due to the fact in regards to spigot, async does use another thread in the implementation

[glossy laurel]

😔

[wet breach]

but you are correct nonetheless about it

[sly topaz]

paper method

[slender elbow]

papir

[blazing ocean]

papièr

[jagged thicket]

paper sex

[worthy yarrow]

https://paste.md-5.net/decajulisu.cs

Nothing serious here just curious, is this correct in applying metadata to an entity?

I also figure in this example I could have used a FixedMetadataValue but other than that not sure if it really matters

[echo basalt]

pdc when

[sly topaz]

allegedly, metadata causes memory leaks so you shouldn't use it

[worthy yarrow]

I literally just saw entities have pdc smh

[sly topaz]

it is a broad statement which doesn't really hold value to all scenarios but it certainly can happen

[sly topaz]

PDC is permanent tho, I'd just use a hash map

[worthy yarrow]

I mean pdc imo is easier to manage, although I would assume the pdc is cleared when the entity dies no?

[thorn isle]

i doubt it's cleared but the entity reference definitely becomes unreachable

so unless you're referencing it somewhere yourself it's as good as cleared

[sly topaz]

it is most likely cleared though

[slate siren]

hey guys how can I open a sign in front of the player for input?

[thorn isle]

i think the mob just gets marked as invalid and dead and little else is done on its data

[sly topaz]

destructors get called when an entity dies, I don't assume they don't clear nbt on purpose

[wet breach]

there is background tasks that eventually clear it out

[sly topaz]

place a sign somewhere then use Player#openSign

[worthy yarrow]

It just feels weird mapping entities and holding them from spawn -> death

[sly topaz]

I don't really see what's wrong about it

most plugins before PDC just used hashmaps which they dumped into yaml or whatever

[thorn isle]

i can't find it being cleared anywhere

the reference on CraftEntity isn't a weakref either

[sly topaz]

let me pull up the spigot source, I could be wrong since my memory is quite fuzzy about these details lol

[worthy yarrow]

I suppose my eventual use case would turn into a sort of mob stacking system, so it's not really even an issue I'd only be mapping a single entity anyway... I think

A single entity in the sense of one per stack as it were

[thorn isle]

i'm fairly sure the entity data stays available indefinitely (as long as it isn't gc'd, i.e. you're referencing it) but you can't do certain things like teleport it or use its pathfinder

[sly topaz]

PDC sounds fine for that, as long as you don't need mob stacking data offline for whatever reason

[thorn isle]

but like all of this is much besides the point; you probably shouldn't be using the entity pdc if you need to access the data beyond the entity's lifecycle

[worthy yarrow]

Yeah fair

[sly topaz]

it's often hard to gauge that as when you expand whatever you are making, you do end up needing that data offline lol

be it for migrations, or other unrelated things

[thorn isle]

yeah, this is why i kind of dislike the pdc

[blazing ocean]

hi kat

[sly topaz]

and Spigot doesn't offer any access to DFU so it isn't like you can just make a datafixer when it comes to data migration

[thorn isle]

there is also no good way to clear out data when your plugin is "uninstalled", which becomes an issue for players where inactive pdc entries pile up over time

basically every single time i've used the pdc on players especially i've had to migrate to something better later on

it's good on blocks though

[worthy yarrow]

There's no way to directly clear all pdc entries is there

[sly topaz]

yeah, players are good example of a place where PDC is just the lazy solution, a database always end up being the final form for whatever data you're holding

[sly topaz]

nope

[worthy yarrow]

Hi rad

[blazing ocean]

of one container? sure

of all? no

[worthy yarrow]

I would want to clear them all

so rip to that

[manic delta]

javier chambea

[sly topaz]

I'm surprised there's no clear method on the PDC tbh, you have to loop through keys and remove each one manually

[blazing ocean]

yea

[sly topaz]

leave me and my unemployment alone smh

[thorn isle]

i suspect it's kind of intentional, since you really can't account for what other plugins have stored on it or know whether it should be deleted or not

[manic delta]

agarra la pala

[worthy yarrow]

I really don't think I'll ever use the data offline, I think my only goal would be to stack mobs and even then I don't really wanna deal with recreating stacks on restarts and what not

[thorn isle]

i could maybe see a clear(Plugin) method that clears all the key-values where the key namespace belongs to the given plugin

[sly topaz]

that sounds reasonable, free PR also

[blazing ocean]

doesn't that like involve like, looping over all entities and tile entities that exist in all worlds

[thorn isle]

it'd just be for the single pdc

[blazing ocean]

oh right

[manic delta]

what if pdc is the friend we made on the way

[thorn isle]

i don't think there is any sane way of clearing them all in all the world

[blazing ocean]

yeah exactly

[thorn isle]

except maybe to associate a timestamp for each pdc entry and then store "deletion rules" somewhere in the server internals forever on clear, that get iterated through on pdc load and delete matching entries

[sly topaz]

Bukkit#getOfflinePlayers 🤑

[thorn isle]

i.e. entries with that key and a timestamp older than when the clear was issued

[blazing ocean]

now do that for all tile entities

[thorn isle]

yeah this'd be for tiles and entities that you can't realistically iterate over in one go

i mean you can but you really don't want to and god knows how long it'd take for large maps

[sly topaz]

at that point it's just better to manually parse the region files

[blazing ocean]

proceeds to block the main thread for a minute

[thorn isle]

the downside then is that now you have an indefinitely piling up list of deletion rules that all have to be checked each time any tile/entity loads

so it isn't really feasible either

[wet breach]

ideally you would do it either when a chunk loads or unloads

I have never been a fan of saving data directly into the world files though

[sly topaz]

yeah, it just kinda sucks it has to be that way, since it often becomes too late when you realize you could've done a better job at keeping track of the data

[wet breach]

well it could easily be changed where it just kept separate with a way pointing where that data goes

in this manner clearing out old plugin data in the world is a matter of clearing a directory

[thorn isle]

sounds roughly like what we already do but with a tighter integration into the api

wouldn't mind having that to be honest

[sly topaz]

it would be nice but I doubt Spigot would favor such a thing, also taking into account how much of a mediocre storage format region files are anyway

but one can dream

[thorn isle]

it could be abstracted quite far to not be limited to region files i suppose

[sly topaz]

inb4 PDC just ends up being backed by a sqlite database with a table per plugin

[thorn isle]

well yeah that's how it'd end up

e.g. imagine if PersistentDataType also had a, say, Z load(UUID) and a void save(UUID, Z) and the default impl for them would be to dump the data in the nbt

but they could be overridden to, say, query a database

they would then have to be called before the chunk containing the pdc holder is actually added to the world, so as to not block the main thread

[sly topaz]

it's a nice thought experiment though I believe it'd take a bit too many NMS patches for it to be considered a worthy undertaking

[thorn isle]

myeah

iirc paper recently-ish had a pr for async chunk preload event that can be used for this, without having to mess with pdc types

the idea is that you block the event and do your db queries async, and once it's all done, the chunk gets popped into the world with your data in it

[sly topaz]

I do wonder if that'd work at all on spigot given how differently the platforms handle chunk loading lol

[thorn isle]

i think its somewhat async in the mojang server already but i'm not sure

[slender elbow]

on vanilla & spigot chunk loading is async but the server stops ticking until the chunk is loaded anyway

[thorn isle]

spottedleaf has been tearing everything apart with massive diffs rewriting chunks and light and everything else so i can't really tell how it works anymore

[sly topaz]

I mean, it's just moonrise made into a patch in the end so ultimately you'd have a better picture by looking at the mod

it's still hard to grasp if you don't already understand vanilla's chunk loading system and it isn't like anyone has made much effort into explaining it

I do hope it doesn't end like Aikar's changes if anything lol

[thorn isle]

the patch is borderline illegible because there are no imports and everything is ca.spottedleaf.moonrise.patches.chunk_system.level.chunk.ChunkData.someNmsMethod$someMoonRiseOverloadWithReallyLongName

[blazing ocean]

you forgot $moonriseNmsMethodName0_moonrise

[sly topaz]

yeah, every time I find myself digging paper source, I just stop whenever I see any relation to moonrise as the decompiled code is just shit to read

but it is decompiled code so it is what it is

[thorn isle]

it isn't even decompiled, it's the actual patched & mojmapped source

they just avoid imports because of diff conflicts

so everything is always fully qualified

ironically if it were decompiled the decompiler would import it and use the short class name

so it might be more legible after compilation+decompilation

which is fucking 🤡 but here we are

[jagged thicket]

moonrise will be so fkn hard to maintain if it was not mixin first

[remote swallow]

you do realise moonrise was used in paper before it was called moonrise

[jagged thicket]

well i don't

its not good to know the history too much

[blazing ocean]

what

[slender elbow]

it even had a history before it was added to Paper!

[eternal night]

What

[jagged thicket]

According my understanding of moonrise, it is currently a fabric mod which is then ported into paper

#555462289851940864 message

[eternal night]

Yes but like, updating it in paper would not be more difficult than mixins

[blazing ocean]

if anything mixins are harder / more annoying to update

[remote swallow]

"Moonrise" was in tunity first then moved to paper then moved to a mod

[echo basalt]

moonrise my beloved

lets us do async entity lookups without blowing up the server

really useful for packet manipulation

[thorn isle]

and block lookups

[echo basalt]

is there seriously no good method of detecting when an item is "put in a chest"

I have to do all this BS emulation

[sly topaz]

what do you mean

are the inventory events not enough? What do you need to do emulation for?

[echo basalt]

I need to emulate how the chest behaves to figure out the before and after state

events are enough but it's a pain in the ass to do this

https://paste.md-5.net/luxocaraqe.cs

[sly topaz]

there is no before and after state in a chest though lol, let me check the code ig

[thorn isle]

i don't know what you're doing but if it's a gui, i've found implementing them through nms to be much easier, since that gives you access to oop style Menu and Slot objects, and the server internals will actually ask your Slot's whether something can be taken from/placed in them, rather than you having to emulate the 100 different click types yourself

[echo basalt]

it's literally just a quest where you need to put an item in a chest

[sly topaz]

I don't get why it is doing specific checks

[weak wasp]

That's only checking if players put something in a chest. What about hoppers and minecarts as well?

[thorn isle]

the trivial way of doing it would be to scan the inventory before and after each interaction, but then there exists an edge case where two players could insert things in the same tick and you couldn't tell which one inserted what

[sly topaz]

just listening to the inventory click event and checking whether the item they clicked ended up in the chest view should be enough

[echo basalt]

if you look I don't have a check for container type

but the quest is literally "put your ball in this chest"

[thorn isle]

shift clicks and such make that nontrivial

[echo basalt]

provide a context object?

saying like

[thorn isle]

since whether something ends up somewhere is completely up to the click type and inventory state

[echo basalt]

basically a diff of what changed

[sly topaz]

too bad InventoryMoveItemEvent doesn't quite work for this concept

[echo basalt]

item went from slot A to slots B, C and E

[thorn isle]

reflect into the container and wrap each Slot with a wrapper that reports canPlace(Player, ItemStack) to you so you know who placed what in which slot 🤡

[echo basalt]

hella work I don't get paid enough to bother

[thorn isle]

i really wish we had a proper menu/slot api in bukkit

the inventory api is okay but trying to do things like these is profoundly ass

[sly topaz]

just loop the inventory a tick after and check if the clicked item is there

[echo basalt]

I once tried to do a plugin where you could uh

[sly topaz]

it should work well enough™️

[thorn isle]

what if two players put items in it the same tick?

[echo basalt]

put whatever item you want on a beacon

because you can do it with events and enough emulation

[sly topaz]

they wouldn't put the exact same item

[thorn isle]

i guess you can cancel all inventory events from the other player during the same tick 🤡

[sly topaz]

even if they did, you could just check for that as well

[thorn isle]

but how do you know which one placed which item?

[sly topaz]

any information from the event can be passed to the task afterwards, you'd just need to check for it if it is necessary

[echo basalt]

https://paste.md-5.net/avikahidus.java

this is NOT accurate

[sly topaz]

and the event would have information of who clicked what

[thorn isle]

atp you might as well emulate it

since what the information on the event means depends entirely on the click type and inventory state

[echo basalt]

reactive inventories

[thorn isle]

suppose for example one player doing COLLECT_TO_CURSOR and the other one putting that same item type in the chest; you can't tell just based on the clicked item or cursor item who placed vs who took the item

[sly topaz]

@EventHandler(priority = EventPriority.MONITOR, ignoreCancelled = true)
public void onInventoryClick(InventoryClickEvent event) {
    var containerInv = event.getClickedInventory();
    var player = (Player) event.getWhoClicked();
    // momentarily put a PDC on the item to know who placed it or something
    Bukkit.getScheduler().runTask(plugin, () -
        for (var item : containerInv.getContents()) {
            if (item != null && filters.test(item) /* check that pdc here as well*/) {
                setProgress(player, getProgress(player) + 1);
                break;
            }
        }
    });
}

naive but it'd work

[thorn isle]

but now that i mention it, you could mark each container view as "locked" for a specific player after the first interaction with it, and prevent others from interacting with it in the same tick

in this way all of the changes in the inventory can be accurately attributed to the correct player, because there can only be one player

well, minus hoppers and shit

[sly topaz]

I think it's severely overcomplicating this lol

it is just a quest, there is no need for it to be extremely precise and efficient

[echo basalt]

lol turns out this isn't a chest but instead a custom menu from another plugin

[young knoll]

I mean that’s just a chest menu without the chest

[sly topaz]

that makes it easier tho, since most plugins make a view per player instead of sharing one

[young knoll]

An orphan chest menu if you will

[floral pond]

Hello, what is the best way to store an Item Stack with his metadata (colors of a banner, writings of a book...) in a json (or DB) ? Should they all be saved with their own logic ?

[sly topaz]

serializing the item to bytes

[young knoll]

The api way is to use BukkitObjectInput/OutputStream to convert it to bytes

[sly topaz]

and a db

[young knoll]

Tho with NMS you can serialize the NBT directly for a smaller byte array

And also add DFU support

[floral pond]

ok i will try serializing it first

seems like what i need

Thanks !

[sly topaz]

public byte[] serializeItem(ItemStack item) {
  Preconditions.checkArgument(item != null, "The given ItemStack must not be null");
  var nmsItem = CraftItemStack.asNMSCopy(item);
  var compoundTag = (CompoundTag) nmsItem.save(((CraftServer) Bukkit.getServer()).getHandle().getServer().registryAccess());
  return writeCompound(compoundTag);
}

public ItemStack deserializeItem(byte[] bytes) {
  Preconditions.checkArgument(bytes != null, "The given bytes must not be null");
  var compoundTag = readCompound(bytes, References.ITEM_STACK);
  return CraftItemStack.asCraftMirror(net.minecraft.world.item.ItemStack.parse(((CraftServer) Bukkit.getServer()).getHandle().getServer().registryAccess(), compoundTag));
}

private CompoundTag readCompound(byte[] bytes, DSL.TypeReference reference) {
  try (var bais = new ByteArrayInputStream(bytes);
      var ds = new DataInputStream(bais)) {
    // should this stay unlimited heap??
    var compound = NbtIo.readCompressed(ds, NbtAccounter.unlimitedHeap());
    var dataFixer = ((CraftServer) Bukkit.getServer()).getHandle().getServer().fixerUpper;
    var result = dataFixer.update(reference, new Dynamic<>(NbtOps.INSTANCE, compound), NbtUtils.getDataVersion(compound, -1), Bukkit.getUnsafe().getDataVersion());
    return (CompoundTag) result.getValue();
  } catch (IOException e) {
    throw new RuntimeException(e);
  }
}

public static byte[] writeCompound(CompoundTag base) {
  NbtUtils.addCurrentDataVersion(base);
  try (var bais = new ByteArrayOutputStream();
      var ds = new DataOutputStream(bais)) {
    NbtIo.writeCompressed(base, ds);
    return bais.toByteArray();
  } catch (IOException e) {
    throw new RuntimeException(e);
  }
}

this is a direct copy of Miles PR where I just translated stuff to mojmap, unsure if it needs any changes

oh addCurrentDataVersion adds the current data version to the compound tag, TIL

[young knoll]

Yeah that’s pretty much all you need

Is NbtUtils a Mojang class?

[sly topaz]

these are the imports:

import com.mojang.datafixers.DSL;
import com.mojang.datafixers.DataFixer;

import java.io.ByteArrayInputStream;
import java.io.ByteArrayOutputStream;
import java.io.DataInputStream;
import java.io.DataOutputStream;
import java.io.IOException;

import net.minecraft.nbt.NbtUtils;
import net.minecraft.nbt.NbtIo;
import net.minecraft.nbt.NbtAccounter;
import net.minecraft.nbt.CompoundTag;

can't put them on the same message since I don't have nitro

[young knoll]

Huh it is

TIL there are helper methods for the data version

[sly topaz]

it is GameprofileSerializer in spigot mappings

[young knoll]

I was handling that manually

[sly topaz]

could probably remove some of the imports since I used var in most places

[young knoll]

Var

Banned

[sly topaz]

I use it a ton so that people use it more

it's even more useful in these cases where you don't want to import things

[thorn isle]

i dislike excessive use of var because it makes it difficult to infer what everything is

try (var bais = new ByteArrayOutputStream(); is fine; the type is clear

var compound = NbtIo.readCompressed(ds, NbtAccounter.unlimitedHeap()); is edging the line

[sly topaz]

the good thing about var is that it was specifically limited with that in mind

[thorn isle]

var dataFixer = ((CraftServer) Bukkit.getServer()).getHandle().getServer().fixerUpper is absolutely haram

[sly topaz]

it is mostly because you don't need to know the type to infer what it returns here, if you know anything about the code in question anyway

[slender elbow]

ByteArrayOutputStream byteArrayOutputStream = new ByteArrayOutputStream() is peak java and code that everyone should write

[thorn isle]

it's not worth saving the 15 or however many characters the actual class name would be

[sly topaz]

java people were very worried var would hurt readability, but I've found that it rarely does so, even less when taking into account the majority of the time you're reading your code in an IDE which will tell you the type in a inlay hint anyway

[thorn isle]

var excels at hiding hideous and redundant generics, e.g. an enhanced for loop over a map's entry set

hardly ever do i see it warranted for types without type parameters

[thorn isle]

with the sole exception of assignment from constructor like this

[sly topaz]

it is worth it if you consider consistency

[thorn isle]

being consistent in being vague isn't a benefit

[sly topaz]

if you don't use var just because it makes the type a teeny bit unclear (which, your variable name should help with anyway) when reading the code outside an IDE, then you probably shouldn't use it at all at that point

I don't believe that to be the case

[young knoll]

They should have just made it v to truly save characters

[slender elbow]

var is no different than explicitly typed lambda parameters, and yet you very rarely see that

[thorn isle]

i type my lambda parameters

[slender elbow]

:doubt:

[thorn isle]

i also put nullability annotations on them

[sly topaz]

now that's just crazy work

even if the type isn't already inferred within the same line of code, reading the rest of the code should make that clear enough

if it doesn't, then read more code

only case I do agree that var shouldn't be used is when you lose the benefit of type inference in the type definition, making it redundant as now you have to be specific on the method call/instantiation

i.e. var list = new ArrayList<Integer>();vs List<Integer> list = new ArrayList<>();, in this case I prefer the latter

[slender elbow]

I mean, you're just moving Integer from one place to the other

it's not like you are hiding any information

[sly topaz]

it isn't about information at this point, just that I feel it is redundant at that point, but either choice is fine honestly

if anything one can do var for sake of consistency

[thorn isle]

doing anything all the time just for the sake of consistency is missing the point imo

for the record i was joking about the lambda nullability annotations, that is 🤡 as a baseline, though i do use them when they're actually warranted; case in point

[sly topaz]

I mean, it is really down to whether you like the feature or not, because the readability argument was extensively discussed in the mailing list when it came to accepting this feature, and it did end up getting accepted nonetheless so I don't personally consider it to be as much of a problem as people make it to be, rather people just find it strange for Java

I like the feature so I use it, is my bottom line

[thorn isle]

i liked it until i had to read an old piece of code on github without ide support

[sly topaz]

I used to do that a lot with paper/spigot code, I learned it is just easier to have the IDE open at all times with the code lol

I always end up being frustrated by github's navigation, even if it is far better than bitbucket's with its symbol recognition feature

[eternal oxide]

Or you could just not use var and everythign is instantly understandable at a glance

[thorn isle]

github's navigation is and always has been ass

and in the latest update they completely fucked ctrl-f; the browser search highlight now appears in a completely disjointed place from the actual search match

[sly topaz]

I have yet to see an example where it isn't understandable at a glance while using var, or rather an example where it matters. In most cases you don't really care about the type of a variable in the middle of a method, just what they do with it

[slender elbow]

some if its navigation has gotten better but it is certainly not an IDE replacement

[sly topaz]

even more so for codebases like spigot/paper where you have to navigate through patches

used to be way worse with paper before hardfork

[thorn isle]

i don't have any example to give you off the top of my head but it happens every time someone here uses var in their code that i need to look up javadocs for

rather than entering the type name in javadoc search it becomes a search to the n'th order

where the first thing i can search is class X which has method Y whose return type I have to check to get its javadocs to find method Z to get its return type to get its javadocs to find method W to ... to figure out what the fuck var entity actually is and why entity.remove() doesn't do what the guy expects it to do

having encyclopedic and exhaustive knowledge of the entire bukkit codebase does speed this up quite a bit but that's quite a tall order for someone to have

[sly topaz]

it is definitely more annoying when the method name itself is unclear and the variable name isn't indicative of anything, but that is only really true in sloppy code, so I don't suffer with it as much

[sly topaz]

I honestly find it surprising just how many people we have around with that kind of knowledge lol

in comparison to other communities anyway, at least wasn't my experience with other languages

I do wonder if some of it can be attributed to OOP

[thorn isle]

i don't think method names being somewhat vague is really sloppy code, half of the information on what a method is expected to do is its association with the class it's declared on

like sure rem() is sloppy code, but remove() certainly isn't, nor is say close(); but these words can have a completely different meaning depending on what is being "removed" or "closed"

this is why i disagree with dynamically typed languages in principle, just because you get something that has some method with some name doesn't mean it will (or can) do what you expect it to do

[slender elbow]

while var does not aid badly written code, not using it doesn't exactly save it either

bad code is bad

[thorn isle]

for sure

[eternal oxide]

In a world where 70%+ of code is bad, encouraging the use of var is probably a bad idea.

It would be OK if we could remotely slap wrists when bad code is written

[sly topaz]

by that logic dynamically typed languages shouldn't exist and we can't have any type of inference since it makes things unclear

[slender elbow]

i read brits and giggled

[sly topaz]

bad code? Blame the British

[eternal oxide]

oh typo 🙂

wrists

[sly topaz]

that was an unfortunate typo lmfao

[thorn isle]

a rule of thumb i use is that if you can infer the type without having to look more than a few lines up or down, var is fine, e.g. the following is good:

for (var e : this.someMap.entrySet()) {
    String k = e.getKey();
    UUID v = e.getValue();
    //do something
}

And the following is also fine:

Map<String,UUID> map = ...;
for (var e : map.entrySet()) {
    var k = e.getKey();
    var v = e.getValue();
    //do something
}

But the following is absolutely haram:

for (var e : this.someMap.entrySet()) {
    var k = e.getKey();
    var v = e.getValue();
    //do something
}

because the type declaration for k and v is probably 500 lines up where the map field is declared

[young knoll]

@remote swallow is responsible for all the bad code

I knew it

[slender elbow]

"dynamically typed languages shouldn't exist" is certainly a statement, they can have an entirely different grammar mechanism where the lack of typing does not hinder it

[thorn isle]

in the same vein

this.someMap.forEach((k,v) -> ...)

is also absolutely haram, and I would do the 🤡 and type the lambda parameters with:

this.someMap.forEach((String k, UUID v) -> ...)

[eternal oxide]

yep

[sly topaz]

I disagree

[eternal oxide]

I just don't use var as I would not trust myself to make unreadable code 5 years down the road

[sly topaz]

in reality instead of calling it k, v, you should call it text, uuid (or whatever text alludes to, i.e. name) and it is immediately clear

[eternal night]

Real peeps know, var is the ultimate abstraction technique

It is the perfect interface replacement

[remote swallow]

Heh

[thorn isle]

text and uuid will work for this simple example but for less known types that will quickly fall flat

[eternal oxide]

Sorry Epic it's my fault I exposed you

[sly topaz]

I really don't see the case where it won't, if you don't then you are being sloppy

[thorn isle]

no, name is not equivalent to type

[sly topaz]

since if the method is long enough, type is also lost in the sauce

[sly topaz]

not the same, but indicative

[thorn isle]

multiple variables of the same type can represent different things and be named differently

e.g. suppose a player -> player map where the key represents a sender and the value represents a receiver

[remote swallow]

Damn

Defamation on my name

[thorn isle]

would you prefer (playerSender, playerReceiver) -> ... or (Player sender, Player receiver) -> ...?

[sly topaz]

I would prefer (sender, receiver) ->, given it is used in a context where the types of sender and receiver are clear as it'd be by something that takes a Player, Player lambda

[thorn isle]

it's in no way guaranteed that they are players

when I see receiver I probably think of Audience

when i see sender I probably think of CommandSender

so unless those type declarations are less than a few lines away I'd declare them here

again on the principle that name doesn't and can't indicate what exactly something is; be it method or variable

[sly topaz]

if it isn't clear by type, it is clear by name, if it isn't clear by name, it is clear by the method call, if it isn't clear by the method call, then maybe the class name, if it isn't clear by the class name then the surrounding code. Context is important and if we start to analyze these situations in a void, we can reason ourselves into thinking it is unclear however in any practical examples it's very unlikely to be the case

[thorn isle]

type it and it will always be clear

no need for ifs and buts and thens and wherefore maybe if by the method name i can roughly guesstimate what this is

[sly topaz]

you have yet to bring a non-hypothetical situation where it is unclear though

[thorn isle]

i can resend the code block to you next time someone asks for help with code and it has vars in it

[sly topaz]

I stand by the fact the feature was accepted at all, and we're really just repeating what the java devs discussed in the mailing list lol

[thorn isle]

well yes, but the consensus the mailing list ended up at isn't "use it everywhere all the time"

it was accepted because it is conditionally & situationally acceptable

[sly topaz]

well, that'd be hard to judge given I know anything related to Spigot API and a good chunk of the server as my palm, but if it is anything outside that I'll try to be unbiased

[thorn isle]

it's not categorically acceptable nor categorically unacceptable

so we have it and we can use it where it does the job

preferences vary of course but personally i absolutely detest having to guesstimate things about my code based on soft associations like "hm this is called that maybe it does this"

[sly topaz]

I can understand that, however I dislike even more how it generates very stiff code style rules which you'd only see in bs enterprise code

[thorn isle]

if i don't have a class definition with a javadoc in front of me it might as well be single-letter variable names

[ivory sleet]

Doesn’t that depend on the scope in which the lambda is declared in?

for example auxiliary index variables like i, j and k don’t need to be much longer

for one they live a relative small scope

when it comes to lambdas, its ofc nice to have shorter names if code won’t suffer its readability too much, sometimes when currying or use some other functional pattern in java (especially if u get very functional) the verbosity is sometimes quite necessary

but yea sometimes u can also just make it point free and avoid the hassle in the first place

[ivory sleet]

tbf I think its a bit unfair to call it bs- I mean java enterprise has its good reasons for why things become so verbose, you may not like those reasons, but they still persist valid

[echo basalt]

I personally stick to single letter variables and method names

if proguard does it so do I

because I'm pro

[sly topaz]

I don't believe it is all that valid honestly

so much of enterprise code is just bs abstractions made by a junior dev who didn't understand what they were doing when creating that one FactoryBeanFactoryProvider

[ivory sleet]

Not at all

you have to understand enterprise java can consist of multiple modules, and millions of lines of code

you have tight deadlines because one u may work with some dev practice like test driven development

which means u'll be sitting writing tests, not actual prod code

secondly

you have deadlines

sprint deadlines, in lets say agile if u do some scrum or kanban shit

additionally, u have to code things in such a way that if business requirements (or tech requirements change) that change should be relatively easy to incorporate

devs get stressed, devs arent all super pals with one another on the team

50% of the devs might not even be motivated to work their life enthusiasm on the project

in all of this, explicit, super clear, verbosity, with a rigid set of rules and conventions - it does help

[sly topaz]

I mean, yeah I was just bringing up an example of where the code rules don't necessarily bring out any benefit other than making the code itself more verbose, but if we go into specfiics, there's of course many reasons for that to be the case. All in all, losely defined code style rules often end up being of more utility than an entire 12-page document detailing things some product manager considered after consulting the scrum master

[ivory sleet]

Yea no- Im just saying that its not entirely fair to call it bs, or say its invalid when these practices, styles and conventions stem from so much failure and an effort, a reason to address such sources of failure

[dawn flower]

how do i change the light level of all blocks (basically full bright)

[quaint mantle]

https://www.spigotmc.org/threads/set-light-level-of-light-block-1-17-x.519687/

[rough ibex]

That is for a light block.

[mighty wind]

anyone know if its possible to fake trigger frogs tongue animation in NMS

tried frog.setTongueTarget(targetPlayer); and frog.tongueAnimationState.start(0); to no avail but i think it might not be that simple

[sly topaz]

have you tried setting the pose to USING_TONGUE

[mighty wind]

i'll try when i'm back on. honestly i haven't completed the game since beta 1.8 so i don't really know how things work lol

[sly topaz]

to be fair, nobody knows how frogs work, nobody has done anything with them yet

[mighty wind]

would suck if i need to make a slime spawn

[sly topaz]

they use the new brain system for AI which only makes things more complicated since nobody knows how the brain system works

[mighty wind]

yeah i tried to dig into it

are we able to access it at all or change it

i mean you can use reflection but like...

[sly topaz]

as far as acessing it goes you just call getBrain on the handle then you may set a memory which should trigger the action in question, though that's already possible via the API

see LivingEntity#setMemory

[mighty wind]

will do i didn't know this was a thing

i'm guessing there's some way to just mock the memory to make it want to eat lol

[sly topaz]

just setting the memory and then the tongue target may just work honestly

on a high level, brains are composed of sensors, which are responsible for changing memories, memories which should be self-explanatory and behaviors which make use of these memories to do their actions. The behaviors are coupled to activities which are just a set of behaviors that the entity might do

you can read this fabric example https://wiki.fabricmc.net/tutorial:villager_activities which more or less gives one some idea on how it works (Task is yarn lingo for Behavior in mojmap)

[pseudo hazel]

the behaviors can also change memory like a target task that remembers the "target" memory

[mighty wind]

i need to just see how memories work i guess

[pseudo hazel]

memories are just values shared between all the tasks and brain of an entity

[mighty wind]

i don't have my laptop on me to test rn

do you have an example of how you'd change this

or mock memory

[pseudo hazel]

wdym, change existing memory values?

sadly i dont know the spigot api for this

[mighty wind]

oh its like one object each mob has

with a bunch of flags

ah nah i mean the NMS

unless u cant discuss that here

[pseudo hazel]

the brain has memory, which consists of variables with values of varying types

like an allay stores the liked player for example

which is used be several tasks like following the player

that follow task has a starting condition of needed the liked player memory slot to be filled for example

like each task has a pre requisite

[sly topaz]

you can set the memory using the API I sent above

[pseudo hazel]

a find target task has the prerequisite of a target memory slot being empty

[sly topaz]

for the shoot tongue behavior, you'd set the ATTACK_TARGET memory key

[pseudo hazel]

but iirc brains are mostly serverside are they not? it would be hard to fake a tongue animation without the frog actually targeting that entity

[sly topaz]

well it seems like the MemoryKey class is missing a bunch of memory keys so you'd use frog.setMemory(MemoryKey.getByKey(NamespacedKey.minecraft("attack_target"))) for it

[sly topaz]

brains are completely server-side, yes

[mighty wind]

theres attacktarget in the frog nms iirc

as well

tongue target?

that one ^

[pseudo hazel]

p sure its just the attack target

[sly topaz]

by setting the attack target you indirectly set the tongue target as that is set & controlled by the ShootTongue behavior

[pseudo hazel]

usually entities attack by swinging their arms, the frog attacks by using its tongue on the target

[mighty wind]

oh ok

what if i remove all their goals and pathfinders

i assume it will work

otherwise i guess i can write my own attack one if its necessary

would probs need to know which one it is to see

[sly topaz]

super(ImmutableMap.of(MemoryModuleType.WALK_TARGET, MemoryStatus.VALUE_ABSENT, MemoryModuleType.LOOK_TARGET, MemoryStatus.REGISTERED, MemoryModuleType.ATTACK_TARGET, MemoryStatus.VALUE_PRESENT, MemoryModuleType.IS_PANICKING, MemoryStatus.VALUE_ABSENT), 100)
ShootTongue has all of these activity requirements, the walk target has to be absent, look target registered, attack target present and is panicking absent

[pseudo hazel]

well sounds like you just want to change the way it targets entities

there is probably a task responsible for setting the attack target

[sly topaz]

entities with brains do not use goals at all

[mighty wind]

oh wow ok

[pseudo hazel]

they have tasks

which is pretty much a goal but it works with the brain system

[sly topaz]

yeah but I also mean that they're inherently incompatible

[pseudo hazel]

yes

[sly topaz]

entities which use behaviors do not interact with goals and viceversa, which makes things annoying for anyone used to goals but it is eh

[mighty wind]

i see

i will have to learn

[pseudo hazel]

oh well, I like the brain stuff

[sly topaz]

I mean, you don't have to do anything besides just setting the memory keys to these values and then triggering the activity by calling customServerAIStep on the entity or whatever the method is called

there's probably more forceful ways to trigger a behavior however that works fine in my experience

[charred lagoon]

Hey! I’m working on a Lifesteal server for up to 300 players and need someone who’s great with plugins and configuration.

Server type: Lifesteal.

Tasks: Plugin setup, advance configuration

Budget: $100–$150

If you’ve got experience with similar servers and want to help bring this to life, shoot me a DM with your past work or questions!

[thorn isle]

🤡

[drowsy helm]

?services

also $100-150 is like a few hours of work.

?services

[undone axle]

If you wish to request or offer development/art/building/administration services, please do so at https://www.spigotmc.org/forums/services-recruitment-v2.54/

[tranquil pecan]

teach me how to

[blazing ocean]

?

[drowsy helm]

how to what lol

[sullen marlin]

how to

[blazing ocean]

md how do I do the

[tranquil pecan]

💀

[tranquil pecan]

I mean how $100-150 is few hours of work

[drowsy helm]

assuming the person is charging minimum wage (let's say $20)

thats 5-7.5 hours

[quaint mantle]

Highly doubt that they're actually gonna pay $100 for this

[young knoll]

Up to 300 players

[drowsy helm]

I always giggle when i see people planning a server and expecting insane numbers like that

[young knoll]

Well they said up to

0 is also up to 300

[blazing ocean]

-281 active players

[dawn flower]

how much they paying for advertising

[plucky skiff]

Why does BuildTools remapped not actually install for 1.16.5?

[slender elbow]

i believe remapped wasn't a thing until later down the line

1.17 or 1.18

[young knoll]

Yeah 1.17 was the first one iirc

[plucky skiff]

The Mojang Mappings were released on 1.14.4

https://x.com/Dinnerbone/status/1169257039597645824

[slender elbow]

yes

[plucky skiff]

https://mappings.dev/1.14.4/net/minecraft/server/MinecraftServer.html

[slender elbow]

but spigot didn't use them until 1.17

[plucky skiff]

Ah, that's quite annoying.

Alright makes sense, thanks for the help

[dawn flower]

how do i make a hologram system (display entities) where the holograms (or display entities) are respawned every server restart without creating a billion display entities, i could also use packets if it's better

[young knoll]

You could just have them never despawn

Or just store the location, remove any existing holograms at that location and then spawn them again

[slender elbow]

or you could setPersistent(false) them and respawn them when the respective chunk loads

[dawn flower]

this is actually genius

[young knoll]

Yeah that's also an option

[slender elbow]

A genius one at it too

That is quite the bar

[manic delta]

Yeah using PlayerChunkLoad / Unload

[mental frigate]

hi guys, I've got this code

net.minecraft.world.item.ItemStack nmsItem = CraftItemStack.asNMSCopy(item);

net.minecraft.network.chat.Component displayName = nmsItem.getDisplayName();

System.out.println(displayName);```
which returns me something like that
```translation{key='chat.square_brackets', args=[empty[style={italic}, siblings=[empty[siblings=[literal{Netherite Chestplate}[style={color=#FF00FF,!bold,!italic,!underlined,!strikethrough,!obfuscated}]]]]]]}[style={color=white,hoverEvent=TypedHoverEvent[action=<action show_item>, value=net.minecraft.network.chat.HoverEvent$ItemStackInfo@58531c7e]}]```
and im trying to access the siblings, however `Component#getSiblings` returns an empty array, how do I get to this part `[literal{Netherite Chestplate}[style={color=#FF00FF,!bold,!italic,!underlined,!strikethrough,!obfuscated}]]`

[slender elbow]

just use getHoverName() instead of getDisplayName()

[mental frigate]

oh thats a thing

[mental frigate]

it worked, thank you!

[mighty wind]

can you make parrots sit on ppls shoulder

server and client side

[slender elbow]

Player#setShoulderEntityLeft/Right

[mighty wind]

says deprecated

i read its bc it doesnt work

[slender elbow]

where did you read that?

the javadoc itself?

the javadoc doesn't say it doesn't work

it says to use with care, and the client will only render parrot entities

https://hub.spigotmc.org/javadocs/spigot/org/bukkit/entity/HumanEntity.html#setShoulderEntityLeft(org.bukkit.entity.Entity)

[mighty wind]

oh so other people cant see it?

[echo basalt]

gonna put a zombie on the shoulder

[mighty wind]

why tf cant u server side a parrot on sum1s shoulder

can u do it with NMS?

[echo basalt]

you can but the game will only render it if it's a parrot

bruh

[slender elbow]

have you even tried calling this method?

[mighty wind]

oh yeah i have

it doesnt work rn im gonna see if i messed anything up

[mighty wind]

ohhh

[slender elbow]

you can only put a parrot

basically

you asked "can you make parrots sit on ppls shoulder", the answer is yes, use that method

any other entity does not work, but you didn't ask about other entities

[mighty wind]

um also does anyone know how to stop parrots from hopping off shoulders if a player is hit

apart from a scheduler perma setting it to ride

[eternal oxide]

try disabling its AI

[mighty wind]

is that just setAI(false)

i thought ai was like for armour stands or smth if u want them to be invis or smth

didnt work if so 😭

oh serverplayer has this.removeEntitiesOnShoulder();

i dont think u can override this? unless anyone knows how

[slender elbow]

not really

[mighty wind]

ggwp