help-development

[buoyant viper]

does parseDouble parse the literal "Infinity" as Double.POSITIVE_INFINITY ?

[worldly ingot]

I'm missing out on a lot of context I'm sure, but what's wrong with this?

public double parseDouble(String input, double defaultValue) {
    try {
        return Double.parseDoublue(input);
    } catch (NumberFormatException e) {
        return defaultValue;
    }
}

[lilac dagger]

yeah, this is what apache commons does

[worldly ingot]

Other than the glaringly obvious typo

[thorn isle]

hey i found the regex recommended by the javadocs

 final String Digits     = "(\\p{Digit}+)";
 final String HexDigits  = "(\\p{XDigit}+)";
 // an exponent is 'e' or 'E' followed by an optionally
 // signed decimal integer.
 final String Exp        = "[eE][+-]?"+Digits;
 final String fpRegex    =
     ("[\\x00-\\x20]*"+  // Optional leading "whitespace"
      "[+-]?(" + // Optional sign character
      "NaN|" +           // "NaN" string
      "Infinity|" +      // "Infinity" string

      // A decimal floating-point string representing a finite positive
      // number without a leading sign has at most five basic pieces:
      // Digits . Digits ExponentPart FloatTypeSuffix
      //
      // Since this method allows integer-only strings as input
      // in addition to strings of floating-point literals, the
      // two sub-patterns below are simplifications of the grammar
      // productions from section 3.10.2 of
      // The Java Language Specification.

      // Digits ._opt Digits_opt ExponentPart_opt FloatTypeSuffix_opt
      "((("+Digits+"(\\.)?("+Digits+"?)("+Exp+")?)|"+

      // . Digits ExponentPart_opt FloatTypeSuffix_opt
      "(\\.("+Digits+")("+Exp+")?)|"+

      // Hexadecimal strings
      "((" +
       // 0[xX] HexDigits ._opt BinaryExponent FloatTypeSuffix_opt
       "(0[xX]" + HexDigits + "(\\.)?)|" +

       // 0[xX] HexDigits_opt . HexDigits BinaryExponent FloatTypeSuffix_opt
       "(0[xX]" + HexDigits + "?(\\.)" + HexDigits + ")" +

       ")[pP][+-]?" + Digits + "))" +
      "[fFdD]?))" +
      "[\\x00-\\x20]*");// Optional trailing "whitespace"
 if (Pattern.matches(fpRegex, myString))
     Double.valueOf(myString); // Will not throw NumberFormatException
 else {
     // Perform suitable alternative action
 }

it's only like 50 lines which is primarily comments

[buoyant viper]

the slight off-chance that my default value is the same as a user input by pure coincedence

but my NaN idea is basically that but with defaultValue == NaN

[worldly ingot]

I mean surely you've got some sort of bounds, no?

What's your input?

[buoyant viper]

a number

xx.yy

[worldly ingot]

Obviously

What's the purpose?

[buoyant viper]

money

currency

[thorn isle]

iirc the context is an eco transaction

[ivory sleet]

i think guavas is even longer 💀

[worldly ingot]

Does negative currency make any sense in that case?

[buoyant viper]

no, but i early return in transaction methods if n is negative anyway

[thorn isle]

so yeah uh after taking this into account, you maybe do want to do primitive double and check for NaN/infinity instead

[buoyant viper]

so negatives are handled later on

[worldly ingot]

So then you're fine!

Default to 0!

Or -1, whatever

[buoyant viper]

fuck ur so right

[thorn isle]

because unless you or someone else validates the input, the player could enter the string NaN which is going to give you a NaN

which, again, if you do like receiver.deposit() on it and deposit just does balance += amount, it'll nuke the player's balance

[buoyant viper]

truth nuke

(literally)

[thorn isle]

i think vault probably guards against nans and infinities

or well i don't know if vault is in a position to guard anything, it's just a service registry

[buoyant viper]

i made the naive decision to write the eco implementation myself and at this point im in too deep to turn back

sunk cost fallacy

[sullen marlin]

Just make your economy communist

[buoyant viper]

TRUE!

everyones balance is getting nuked

[buoyant viper]

iirc its up to the implementation

i looked at Essentials once i dont remember much

but im pretty sure Vault is all interfaces

[ivory sleet]

iirc they got specs abt not using negative numbers

but thats about it

[thorn isle]

anyhow yeah, try catch and if it throws, return nan

then explicitly check Double.isFinite

[buoyant viper]

what if i add the finite check to the parser lowk

[thorn isle]

this will nicely represent "not a number" without boxing, and will also guard against someone trying to /pay me infinity

but since you will then be dealing with nan's, you have to take very good care to always check that it is finite

[buoyant viper]

doing the finite check outside of the parse and if its false i kill the player

[thorn isle]

so if you're planning to expose this to "outside callers", like other classes, maybe dependencies, other plugin modules, i'd still go with null in place of nan

[buoyant viper]

perhaps

[slate mortar]

is this my fault or some 26 jank

doesnt show my plugin at least

[buoyant viper]

thats been mo jank for years

[slate mortar]

oh fun

never seen that myself

[buoyant viper]

yeah its like relatively common actually

[slate mortar]

aight glad, since i'm currently messing with join & quit stuff, so i wouldnt have been surprised

[buoyant viper]

ye plugin themselves provide values rather than the economy plugin (all eco does is safeguard n < 0)

so its per-plugin to make sure they dont nuke the balance

the good part is im in charge of all the plugins that will use it (since its for My server)

the bad part is Im in charge

[thorn isle]

writing your own eco backend isn't a bad thing

i've had to do the same because all the plugins out there seem to either be terrible, bloated, feature creepy dumpster fires like TheNewEconomy (recommended by towny) or otherwise nice and simple, but don't work with towny

so i ended up writing one that specifically doesn't fuck shit over when used by towny, and is just a plain and simple uuid -> double store

[buoyant viper]

does Towny not hook Vault?

[thorn isle]

it does, but it has had trouble with uuid's ever since they were introduced

[buoyant viper]

oof

[thorn isle]

and since it needs to register non-player accounts for towns, shit gets whack and the uuid resolution differs from backend to backend

many backends like essentials eco do weird evil uuid version checks that prepend npc_ to accounts or something, and that makes it not work with towny

[buoyant viper]

i do like mine bc it has a pseudo-bank system where players have their on-hand/wallet balances, but can also store and access balances in an infinite amount of "bank" accounts

i just have no way of making more than one bank rn... so its just one Central bank

im like the federal reserve fr

User

• Wallet

Bank

• AccountHolder
• • Owner (User reference)
• • Collection<Account>

eventually ill think of a way to manage multiple banks but

maybe just like

BankRegistry

• • Collection<Bank>

[buoyant viper]

its a fun learning experience for sure

ive never touched SQL even with a 39.5 foot pole until now

never had a system that benefitted enough from having a storable database to constitute learning SQL

had to spend yesterday evening fixing my queries bc my initial database structure didnt work properly to let me use UPSERT in SQLite

and initially i had 3 separate queries (UPDATE on SELECT else INSERT)

[wet breach]

you will have fun with this

[slender elbow]

remember CRUD
Celect
inseRt
Update
Delete

[young knoll]

I thought it was
Chatgpt
gRok
claUde
Deepseek

[topaz cape]

yo how do you detect if the damager is anchor block explosion?

EntityDamageByBlockEvent#getDamager returns null cuz block alr gone

but it still calls the event

[slender elbow]

would probably need to chain some logic with the BlockExplodeEvent

[topaz cape]

Ah

[topaz cape]

this does not make sense but.. BlockExplodeEvent is called AFTER EntityDamageByBlockEvent??

[thorn isle]

i don't know if there is any sensible implied order for them

[topaz cape]

BlockExplodeEvent is cancellable

so it should be called first no?

[thorn isle]

yeah, it will prevent the blocks from exploding if cancelled

it doesn't cancel the explosion itself

[topaz cape]

exactly

huh

thats dumb

[thorn isle]

well, not really

suppose for example that you want to prevent some blocks from exploding

but still want it to damage mobs

perhaps you have an indestructible custom block that you don't want to be explodable

[topaz cape]

blockList#clear?

[thorn isle]

cancelling the event shouldn't make the explosion not deal damage to nearby entities

[topaz cape]

blockList is a mutable list

[thorn isle]

sure

[topaz cape]

so yes, its stupid

[thorn isle]

not really

[topaz cape]

yes it is

@-@

[wet breach]

Entitydamagebyblock event is not inherently tied to blocks exploding. There is other blocks that can damage. The reason blockexplode event comes after is because otherwise you would not be able to cancel the damage to entities as blockexplode event does not have entity list.

@topaz cape

[slate mortar]

what is the commonly used way to align multiiline-text properly, since apparently tabs aren't a thing, and spaces just don't work at all, i assume due to different character lengths

[thorn isle]

there's a MapFont class or something that you can use to calculate the rendered widths of text in the minecraft default font

using this and spaces and bold you can manually align things

[slate mortar]

the problem is that apparently with only spaces, it is impossible to align

sec

[thorn isle]

there's a magic trick

making a block of text bold makes it 1 pixel wider

so you can make your spaces bold to get pixel-perfect adjustments

[slate mortar]

this for instance.

when i move the "<name>" from the "enter" line one to the right, it does not seem to work either

[thorn isle]

yes

you need to make some of the spaces bold

[slate mortar]

oh the spaces

[thorn isle]

different characters have different widths, and you have to adjust the width of some of your spaces to account for it

a cleaner way of doing this is to use a custom font with spaces of widths 1, 2, and 3 pixels in addition to the default 4 of the regular space

but messing with bold also works, it's just more clunky

[slate mortar]

but that would involve a custom resourcepack i guess

which is always ass

[thorn isle]

well, it's fairly simple to set up, but it does mean that people without the pack will be seeing random squares in place of some spaces

[slate mortar]

yea i know how to do custom fonts, but custom rp's itself are smt i always avoid, since people do not like those, for a reason

[thorn isle]

myeah that is fair

[slate mortar]

welp i guess ima see about that custom formatting. thanks for the help

[thorn isle]

spaces it is then, but it will be fucked if anyone has unicode font enabled or uses a custom font, since then the character widths will be different

[slate mortar]

ugh

yk what, i'm just going to use comic sans

or just use no spaces at all the same mojang does.

[thorn isle]

dots can get you sorta close

that's what worldguard has been using for its /wg flags list for like 20 years

it's not pixel perfect, but each one is like 2 pixels i think, so you'll be 1 pixel off at worst

[slate mortar]

meh that should be enough, kinda same as the existing "help" command

[young knoll]

Make a custom monospaced font :p

[thorn isle]

unicode has a monospaced block and it does render ingame, but it looks kind of weird

[frosty bolt]

I'm facing a weird issue where getPersistentDataContainer() highlighted in red (Cannot resolve symbol) in IntelliJ IDEA. However, my pom.xml seems correct and I'm targeting Spigot 1.20.1.

[echo basalt]

zero width spacing or something

[echo basalt]

Does it compile fine? intellij is weird sometimes

[chrome beacon]

Make sure Intellij is up to date and that you have a modern JDK set for the project

[frosty bolt]

Does it compile fine? intellij is weird sometimes

[echo basalt]

crazy

[frosty bolt]

Oh, sorry, I was just trying to insert a picture.

[frosty bolt]

The program won't compile because of this error and I've already tried clearing the cache.

[thorn isle]

do you maybe a have a bukkit 1.8 as a transient dependency from worlguard or something

[frosty bolt]

No, I don't think there's anything like that, but I have an api vault in my pom.

[thorn isle]

that could do it, vault is ancient and does depend on bukkit

print your dependency tree

if you use intellij, you can see it in the maven sidebar; expand "dependencies" and look for bukkit/spigot

[frosty bolt]

<repositories>
<repository>
<id>jitpack.io</id>
<url>https://jitpack.io</url>
</repository>
<repository>
<id>spigotmc-repo</id>
<url>https://hub.spigotmc.org/nexus/content/repositories/snapshots/</url>
</repository>
</repositories>

<dependencies>
    <dependency>
        <groupId>com.github.MilkBowl</groupId>
        <artifactId>VaultAPI</artifactId>
        <version>1.7</version>
        <scope>provided</scope>
    </dependency>

    <dependency>
        <groupId>org.spigotmc</groupId>
        <artifactId>spigot-api</artifactId>
        <version>1.20.1-R0.1-SNAPSHOT</version>
        <scope>provided</scope>
    </dependency>
</dependencies>

</project>

@thorn isle Indeed, as soon as I removed the vault dependency from the pom, everything worked.

@echo basalt Sorry if I offended you, I just wanted to send a pic.

[thorn isle]

you can exclude it

shove

            <exclusions>
                <exclusion>
                    <groupId>org.bukkit</groupId>
                    <artifactId>bukkit</artifactId>
                </exclusion>
            </exclusions>

within the vault dependency tag

[frosty bolt]

thank you

[grim hound]

anyone knows why I can't "Go to declaration" (like being transferred to where the method/variable/class resides) in IntelliJ 2026.1 with the middle mouse button?

worked in IJ 2024

but it doesn't support gradle 9.4 so had to transfer

[thorn isle]

My case about some stupid gradle plugin requiring gradle 999.99 for no discernable reason in point

It's in the settings somewhere, but good luck finding it

My personal recommendation is to give up and just run gradle through the command line instead of the ide

26.1 is fucking horrible

Or even better, switch away from gradle altogether

I can run every maven build under the sky from IJ 2019.1

Since I maintain an intellij plugin and target everything from 19.1 to 26.1, I frequently see and use each version, its mych more apparent to me how it has evolved (and I have about 100gb of gradle garbage and IJ platform shit in my build cache, but thats besides the point)

But switching the versions to runIde and test the plugin on each is like watching a timelapse of Rome turning into ruins

Every time it gets worse and worse

The buttons get bigger and bigger, the ui gets rounder and rounder, all the tool tips and in-ui button texts disappear and get replaced by big fat icons that I have to guess what they do

More and more ai buttons and "start free trial" shit appears

[grim hound]

yeah fuckin hate that

will just keep the 2024.2 version Ig

[grim hound]

I imported the settings, and it's fixed a little bit

but still terrible navigation

I don't think it'd be easy to configurate that

[mortal hare]

why java doesnt support compact source files outside main class?

it would be so useful to have as an utils alternative

[slender elbow]

i feel like a broken record

[thorn isle]

"compact source file"

gb2 python

[mortal hare]

currently i use lombok's

@UtilityClass
public class UsernameValidator {
}

for this

but it would be quite nice to drop class level too

[thorn isle]

what's the issue again?

[sullen marlin]

Costs an extra AI token or two

[buoyant viper]

true

[ivory sleet]

my condolences

[green mauve]

is there an exploit when player can close inventory without sending a packet (or not close it and send a packet) and this way duplicate items?
ive heard of it and im using invenotryCloseEvent to remove the inventory from a set

[slender elbow]

to dupe items? unless it's a poorly coded plugin and the server has no anticheat, no

[echo basalt]

They can use invutils to run commands while a menu is open

Delaying the close packet on top means they get to open a menu before closing another

For example, if you have a gui that saves when closed and renders its items when open it'd result in a dupe

because you can take your item, reopen the menu and get it again before the old version saves

[drowsy helm]

Speed check doesn't account for anything like ice, speed effect etc

[drowsy helm]

uuuh

[pliant topaz]

?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/

[quaint basin]

Does Component#translatable not distinguish which plugin is running?

[chrome beacon]

??

That's just a translatable component it has nothing to do with plugins really

They're a vanilla thing

so no it doesn't care what plugins are running

[quaint basin]

lang.item-bought-successfullyI have this key in 2 plugins and one message is being overwritten

I thought that two different plugins could use the same key

[chrome beacon]

I mean yeah if you have two plugins trying to define the same key that's not going to work

[quaint basin]

ok thanks

One more thing: if a world unloads, are all chunks at the time of unloading called in the ChunkUnloadEvent?

[chrome beacon]

Try it and see

[quaint basin]

My plugin is behaving strangely, I imagine that's not

And what about the EntityRemoveFromWorldEvent? Do you have any idea if the EntityRemoveFromWorldEvent is called when the entity is in a world that is unloaded?

[chrome beacon]

Try it and see

[quaint basin]

https://pastes.dev/YJyI1isQfS

This error appeared when shutting down the server.

[chrome beacon]

?whereami

[undone axle]

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

[quaint basin]

[15:56:06] [Netty Epoll IO #1/ERROR]: Thread Netty Epoll IO #1 failed main thread check: Cannot perform command async
java.lang.Throwable
at org.spigotmc.AsyncCatcher.catchOp(AsyncCatcher.java:9) ~[aspaper-1.21.11.jar:1.21.11-17059-496f794]

I see org.spigotmc

[chrome beacon]

You're not running Spigot so it doesn't matter

[quaint basin]

it makes sense

Block block = world.getBlockAt(cx + dx, cy - 1, cz + dz);
Material type = block.getType();

I'm doing this in EntityRemoveFromWorldEvent and I'm getting this warning: Cannot update ticket level while unloading chunks or updating entity manager

What can I do?

[chrome beacon]

I mean that's a Paper thing

you should be asking them about it

[quaint basin]

So I'm screwed lol

[chrome beacon]

#help-server

This channel is for development

[vestal carbon]

thx

sorry

[rotund ravine]

Nah

[umbral ridge]

thats pretty cool https://github.com/pingcap/tidb

[dry hazel]

looks like cockroachdb but mysql instead of the postgres wire protocol

[thorn isle]

raft
nice

and written in go

looks very impressive, though i don't really need something like this

[umbral ridge]

yeah

cool stuff!

[pliant topaz]

Despite the fact this is not spigot, reading the trace, its about a command being called async. What excactly are you doing in your onDisable?

[thorn isle]

i don't see any plugin classes in the stack trace, and it looks like it's just a command packet being received and processed normally

does it happen every time? if it was just an one-off thing, maybe someone sent a command right as the server was about to shut down, and that isn't handled correctly by the server internals?

it does seem to try to pass it to the BlockingEventLoop which i think should schedule it to run on the MT, but it's a common pattern for schedulers to short circuit on shutdown

[grim hound]

Does RIGHT_CLICK_AIR in PlayerInteractEvent not work anymore? (1.21.11)

[quaint basin]

nothing from i know

[quaint basin]

From what I saw, it only happened that one time

[chrome beacon]

Are you holding an item?

[grim hound]

no, when not holding

[chrome beacon]

Then no the client doesn't send the interact packet

It's been this way for a long time

[thorn isle]

you need either a block at the crosshair or an item in your hand for an interact packet to be sent

the client doesn't expect anything to happen if it has nothing to interact with, so it just doesn't send an interact packet

[chrome beacon]

Or just constantly have an interaction entity infront of the player

[thorn isle]

yes a block or an entity at the crosshair to be more specific

[chrome beacon]

I mean yeah any entity will work

[thorn isle]

however i'd recommend against doing it with the interaction entity as that will take priority over targeting actual blocks at the crosshair

[chrome beacon]

that's also true

[thorn isle]

so you'll have to move it around as the player looks and moves around, which then in turn is latency sensitive

someone has suggested adding a fully transparent texture for an item and then filling all hotbar slots with that item, but i think that will make the player's hand disappear

so the best approach is i think just switch your input button to left click for whatever you're trying to do

[grim hound]

ah okay

tested, and ye, doesn't send

easiest to check would probably be making a large hitbox ride the player at all times and just reproducing the interactions with the api

not sure about mining blocks

will just use items then

[sly topaz]

it is fine if you just put it far enough that it is in entity interaction range but not in block interaction range

[grim hound]

it's different?

the two

[sly topaz]

by default block interaction range is 4.5 and entity interaction range is 3

those values are increased by 2 in creative mode, but I doubt anyone cares about creative

[thorn isle]

still not ideal as this will make the targeted block outline/highlight not appear on the client

[thorn isle]

unfortunately that is backwards in regard to what you need

you could increase it with an attribute and then manually check the range on entity interact packets, and drop any that exceed the normal range, but i suspect that might cause issues with anticheats; it's worth a shot however

[slender elbow]

nowadays you can make it so the rendered item is the hand

can't remember if it was a component or a resource pack thing on the item model definition

[thorn isle]

i know you could do that with core shaders, but i guess they added some special model that just renders the actual player hand?

[green mauve]

is it possible to get 1 tick of redstone? normal repeater's smallest interval is 2 ticks which is = 1 redstone tick apparently
so for example play a noteblock 2 times in 2 ticks

[quaint basin]

Is the EntitiesLoadEvent called on a tick after the associated chunk has been loaded? And what about ChunkUnloadEvent? Are the entities unloaded on the same tick that the chunk is unloaded?

[thorn isle]

not strictly related to repeaters, and i'm not sure but i'd guess, that blocks like noteblocks perform their behavior only when their redstone power goes from zero to nonzero

[quaint basin]

Block block = world.getBlockAt(cx + dx, cy - 1, cz + dz);
Material type = block.getType();

I'm doing this in EntityRemoveFromWorldEvent and I'm getting this warning: Cannot update ticket level while unloading chunks or updating entity manager

[thorn isle]

so it'd take 2 ticks either way, one tick to power it, another tick to depower it

[thorn isle]

undefined; the two are not coupled anymore

[quaint basin]

wdym coupled

[thorn isle]

they are not the same and have no well defined relation

and yes, it is ass, and it makes everything more difficult

entities could load several ticks after the chunk is loaded

not sure how unloading works

[quaint basin]

Yeah, I need to know about unloading

[quaint basin]

and why?

[thorn isle]

either way your error comes from the chunk being unloaded, or being in the process of being unloaded, at the time the code runs

[thorn isle]

because that's how it works, ask mojang

[umbral ridge]

hi hd luna

[thorn isle]

they separated, i.e. decoupled, i.e. made it not coupled anymore, between entity and chunk loading

entities and chunks are stored in completely separate files

[quaint basin]

And can entities load after the chunk unloads?

[thorn isle]

i don't know

[pliant topaz]

it is. some components like noteblocks dont operate on redstone ticks. With some specific timings (maybe 2 out of sync observer clocks, which produce one tick pulses) it is possible, atleast in theory. Try googling or searching on yt, someone most likely already has done something like this

[quaint basin]

Well, I was thinking of using ChunkUnloadEvent and iterating through all the entities. I need to get world blocks when the entity is unloaded. But I don't know if the entities are guaranteed to be unloaded on the same tick as the corresponding chunk is unloaded

[pliant topaz]

hi :)

[thorn isle]

they aren't

i think they could be unloaded already by that point

[quaint basin]

I don't know if this is a problem

[quaint basin]

oh

then yea this is a problem

[green mauve]

can i print current tick to console on event handler

?

[thorn isle]

i think you might be able to deal with it through nms, i suspect this is bukkit being dumb

bukkit only considers the highest, entity ticking chunk ticket level to be "loaded"

[quaint basin]

But then I wouldn't have that error I told you about

If the entity is unloaded before the chunk is unloaded, then I shouldn't get that error about getting a world block if the chunk is still loaded

[chrome beacon]

You really shouldn't have gotten yourself banned from the Paper discord

[thorn isle]

yeah but then you have no entity to call that getBlock logic on

[chrome beacon]

They heavily change how chunk loading works

[thorn isle]

i'm fairly sure the chunk is still loaded at the border ticket level, meaning the block data is still there

[quaint basin]

this is a impossible task

[chrome beacon]

not to mention the events you're using are Paper specific

[thorn isle]

so you should be able to get the chunk with nms and read the block with nms

[chrome beacon]

Impossible to be nice and follow the rules?

[thorn isle]

i got banned for trying to help people with 1.8

two thirds or more of the people here are paper users who either dislike the community or have been banned from it

[chrome beacon]

Yeah I know a lot of you have been

[quaint basin]

You get banned for being good; you don't get banned if you're bad

[umbral ridge]

https://giphy.com/gifs/BbJdwrOsM7nTa

[thorn isle]

i'm fairly sure he also got banned for asking for help with 1.8

[umbral ridge]

guys i need help with 1.0 beta how do i add plugins

[quaint basin]

no

It was because I said "lol" in response to a message and people thought I was mocking them. Something like that

[thorn isle]

either way, try nms

[quaint basin]

And then they muted me for a month. I spoke to the admin and they told me that next time it's a ban, and I told them to ban me

And so I got banned

[pliant topaz]

I'm not aware of another way in spigot, but maybe you could try the Bukkit Scheduler and reapeatingly schedule a sync repeating task every tick? Whats the xy for powering the noteblock every tick? I assumed it was only a general redstone question at first, but since you're asking about doing it with a plugin, what's your plan?

[green mauve]

just noteblock songs

[quaint basin]

What I find strange is that getting the block is possible, but getting the block type isn't

[green mauve]

xy problem is probably not the case

[thorn isle]

getting a block does nothing

[quaint basin]

yea i see now

[thorn isle]

it just creates a new CraftBlock pointing at those coordinates

same with chunks

[quaint basin]

It simply creates an object with x, y, z, etc

yea

[thorn isle]

only when you try using them to read the state does it e.g. trigger a chunk load

[pliant topaz]

If you're doing it through the plugin anyway, you could also go the route with Player/World#playSound

[green mauve]

yeah i know

but im thinking of noteblocks right now

[pliant topaz]

Then probably either the scheduler or some special redstone circuit, sadly my knowledge there isnt too great so i cant help you with that :/

[green mauve]

yeah thanks anyway you helped

[quaint basin]

I don't know how I can obtain the block with the nms in a different way than Bukkit

[thorn isle]

cast the world to craftworld and get the handle and get the nms chunk and get the nms block

[chrome beacon]

So what are you even doing with the block?

[quaint basin]

just getting the type

[chrome beacon]

for what purpose?

[quaint basin]

https://imgur.com/a/LeEFdt7 To save in memory the number of blocks the minion has, in order to know the minion's efficiency. In this case, there are 3 grass blocks, so the efficiency is not 100%, consequently it will generate less stone

[thorn isle]

here's a thought

[chrome beacon]

Do you not already know the efficiency on removal??

[thorn isle]

do that on chunk load

it's not like the blocks are going to change while the chunk is unloaded

[quaint basin]

https://imgur.com/a/kDFiT70 like this?

[thorn isle]

yeah that's worth a try

[quaint basin]

Good observation, lol

I can't do it in chunk load yea

[thorn isle]

i mean with chunk borders and pistons, adjacent chunks might change between the entity unloading and loading if it's on the chunk border, but i don't think it should be an issue

[quaint basin]

https://imgur.com/a/3TILeKB I need to retrieve the data from memory before loading the minion

I could do this with timestamps/time differences, but yeah, that's a big refactoring

[quaint basin]

But it seems to me that Bukkit does the same thing I'm doing

 public Material getType() {
        return this.world.getBlockState(this.position).getBukkitMaterial();
    }

[thorn isle]

possibly

bukkit and nms chunk load levels aren't 1:1 however

so it's still worth a try

[quaint basin]

The difference is that Bukkit uses the world to obtain the state block, while I use the chunk

getChunk will load the chunk if it is unloaded

[thorn isle]

i'm guessing it is loaded, but not at the entity ticking level

which bukkit considers "unloaded"

either way, try it and see

if you get the error, then it is actually unloaded

[quaint basin]

I haven't had that error for a while now

[thorn isle]

if it doesn't happen consistently, what i'd do is just schedule the block get to happen a tick later if the chunk isn't loaded when the code runs

this would be bad if it does happen consistently, as loading the chunk would make it unload again and it'd loop indefinitely

but if it's a 1 in 100 thing, it's fine

[quaint basin]

But the chunk won't be loaded in principle for a tick later

Or are you telling it to load the chunk one tick later?

[thorn isle]

it won't, but the block get call will load it

it also tries loading it here, but the problem is the code is called from nms code that's already doing chunk loading/unloading status shit

and you can't load chunks under that code

loading the chunk a tick later under the scheduler will work as normal

[quaint basin]

https://imgur.com/a/ljuvw11 Perhaps this if statement will resolve the infinite loop you're talking about

I set this to true in EntitiesLoadEvent

With this if statement, I have a guarantee that the getBlock call will be made for a chunk that is not loaded

[thorn isle]

don't use bukkitrunnables for stateless behavior

those are meant for self-cancelling tasks and tasks with state

[quaint basin]

Actually, the chunk might have been loaded, but the entity hasn't been loaded yet, but I don't think that's a problem

[thorn isle]

just use a regular runnable and pass it to the scheduler

[quaint basin]

wdym

[thorn isle]

your bukkit runnable

[quaint basin]

yes but why

[thorn isle]

it's redundant and not as readable

[quaint basin]

Do you advise doing this instead?

 Bukkit.getScheduler().runTask(BukkitMinionPlugin.INSTANCE, () -> {

            if (!bukkitMinion.isLoadedChunk()) {
                bukkitMinion.getMinionState().onMinionUnload(cx, cy, cz, radius, world);
            }

        });

[thorn isle]

yes

[buoyant viper]

https://tenor.com/view/m-gif-7556977615911675392

[quaint basin]

you are using this yet?

they have a dupe

If you drop an item and restart the server, your item will be on the ground and in your inventory

[lilac dagger]

SEVERE: FlatLaf: Using JProgressBar.setIndeterminate(false) not on AWT thread may throw NPE in FlatProgressBarUI.paint(). Use SwingUtilities.invokeLater().
java.lang.IllegalStateException
    at com.formdev.flatlaf.ui.FlatProgressBarUI.lambda$installListeners$0(FlatProgressBarUI.java:127)
    at java.desktop/java.beans.PropertyChangeSupport.fire(PropertyChangeSupport.java:343)
    at java.desktop/java.beans.PropertyChangeSupport.firePropertyChange(PropertyChangeSupport.java:335)
    at java.desktop/java.beans.PropertyChangeSupport.firePropertyChange(PropertyChangeSupport.java:268)
    at java.desktop/java.beans.PropertyChangeSupport.firePropertyChange(PropertyChangeSupport.java:310)
    at java.desktop/java.awt.Component.firePropertyChange(Component.java:8690)
    at java.desktop/javax.swing.JComponent.firePropertyChange(JComponent.java:4598)
    at java.desktop/javax.swing.JProgressBar.setIndeterminate(JProgressBar.java:916)
    at org.spigotmc.gui.panels.general.GeneralPanel.lambda$buildButtonActionPerformed$1(GeneralPanel.java:197)
    at java.base/java.util.concurrent.CompletableFuture.uniWhenComplete(CompletableFuture.java:884)
    at java.base/java.util.concurrent.CompletableFuture$UniWhenComplete.tryFire(CompletableFuture.java:862)
    at java.base/java.util.concurrent.CompletableFuture.postComplete(CompletableFuture.java:531)
    at java.base/java.util.concurrent.CompletableFuture$AsyncSupply.run(CompletableFuture.java:1794)
    at java.base/java.util.concurrent.CompletableFuture$AsyncSupply.exec(CompletableFuture.java:1781)
    at java.base/java.util.concurrent.ForkJoinTask.doExec(ForkJoinTask.java:511)
    at java.base/java.util.concurrent.ForkJoinPool$WorkQueue.topLevelExec(ForkJoinPool.java:1450)
    at java.base/java.util.concurrent.ForkJoinPool.runWorker(ForkJoinPool.java:2019)
    at java.base/java.util.concurrent.ForkJoinWorkerThread.run(ForkJoinWorkerThread.java:187)

am i the only one getting this error

it happens right at finish

[thorn isle]

judging by the one line in the stack trace that seems to be vaguely related to spigot, i assume you're talking about buildtools?

[young knoll]

?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/

[slow skiff]

is there a way not to send teams to client until they see a player of that team without manually sending packets?
it looks like when a player joins they instantly get told of all teams and their members

[thorn isle]

not only that, but the entire scoreboard, with tags and scores and all

this is sort of "doing it manually with packets"-adjacent, but you could maintain a per-player scoreboard with player::setScoreboard, and assign only nearby people to teams based on the real scoreboard

[lilac dagger]

Ah yes

Forgot to mention

Something with the laf theme

[kind hatch]

Leaf theme?

[buoyant viper]

im gonna assume maybe the icon to use the theme is a leaf lol

[young knoll]

I thought we had a sun and a moon

[kind hatch]

See, that's what I thought

[buoyant viper]

maybe they autocorrected LaF

idk

[mortal vortex]

yeah i think he meant LAF

[slow skiff]

sleeping mannequin disappears when player is >19 blocks away ignoring tracking range, why? hard coded client side?

[umbral ridge]

is it possible to modify for how long the action bar message stays visible after showing it

or is that kind of hardcoded

@vagrant stratus

[mortal vortex]

hardcoded i think

[slow skiff]

but you can extend by just sending it multiple times and sending empty to clear or shorten it

[mortal vortex]

yeah but it will disappear abruptly and not fade

(i think the actionbar fades, cant remember)

[slow skiff]

yes but its possible to calculate when it should fade and send the last one some time before that

[lilac dagger]

Yes laf

[slow skiff]

is it possible to make mannequin walk normally without packets or teleporting
can goals be added to them?

[young knoll]

Doesn’t look like it

They don’t extend Mob

[echo river]

You could create a fake player instead of a mannequin if u are using a plugin

[umbral ridge]

Jishuna implements Mob

[chrome beacon]

This still involves packets

[thorn isle]

My personal recommendation is libsdisguises

[echo river]

Oh yeah

[thorn isle]

they need to ban these gooner guilds

what is this shit

[echo river]

Ppl are too horny 😔

[fickle spindle]

there is any good library for config handling becuase i don't like the fuct that as right now all my config is messed up when i push a new update

[thorn isle]

are you the guy to whom i explained how to config.set and save for like half an hour a few days back

[civic mica]

Does anyone need plugin development ... xD

[fickle spindle]

maybe... hahaah

[hybrid spoke]

there's jshepherd

[rocky locust]

How can I protect my plugin against leakers? I've already tested ProGuard and Skidfuscator but apparently it's not that useful

[lilac dagger]

Proguard is the best you can do afaik

There's sone better paid obfuscators

[worldly ingot]

I just wouldn't bother. Anyone with enough time and dedication will deobfuscate it

Or they'll copy/paste your obfuscated code. It's not like it doesn't work just because it's obfuscated

Your code isn't high-value enough to need to be obfuscated

[rocky locust]

I'm about to open a paid plugin store and with my own licensing, you know, I just want to protect it as best I can, isn't there any alternative to make it as difficult as possible?

[chrome beacon]

The plugin will be leaked regardless of what you do

You're just making it a fun challenge to crack

[hybrid spoke]

The best protection against leakers are updates

• obfuscating won't stop people from sharing the jar. it just makes it more fun to read the code

[slow skiff]

the best protection is to not run code on the client but stream it from a license server and apply at runtime

but i dont recommend it, just make good content, good updates, and dont worry about cracks

[chrome beacon]

I mean even that isn't too hard to bypass

• now you run in to the classic problem of the plugin being better and more reliable if you pirate it

Actively making the experience worse for paying users is not the way to go

[clever kraken]

piracy doesnt need active prevention as the ceo of steam said "piracy is almost always a convenience problem"
make your store easy to use and everyone who is willing to pay will pay, everyone who isn't willing to pay just wont use your software even if there's not a pirate

[chrome beacon]

^

https://www.youtube.com/watch?v=EQweFurRz4g give it a watch :)

[fickle spindle]

how can i check if the player have played the server before? i used to use PLAY_ONE_MINUTE but it isn't working properly now...

[chrome beacon]

OfflinePlayer#hasPlayedBefore

[fickle spindle]

and there is a way to reset this value?

[chrome beacon]

delete the player data

[mortal hare]

is there any decent gui client like dbeaver which doesnt have clunky java ui

specifically searching something for redis

[onyx fjord]

vkcommander

[remote swallow]

Datagrip probably does, free for non commercial

[buoyant viper]

    public CompletableFuture<Response> sendAsync(Request request) {
        return CompletableFuture.supplyAsync(() -> {
            try {
                return send(request);
            } catch (IOException exc) {
                throw new CompletionException(exc);
            }
        }, threadExecutor);
    }``` does this look correct for passing an exception to the CompletableFuture?

which is then used like ```java
sendAsync(args).whenComplete((response, throwable) -> {
if (throwable != null) {
/* handle err */
return;
}

/* do stuff */

});```

[ivory sleet]

yea thats fine

u can also handle the error on its own with exceptionally

[lost matrix]

Im not sure i would throw a CompletionException.
Probably a RuntimeException, since the CF will wrap it in a CompletionException on default.

[buoyant viper]

i tried using exceptionally and it had an issue with return types

[ivory sleet]

it will unwrap (wont be rewrapped) CE so its fine

[buoyant viper]

i havent used CFs much tho so i probably just had the wrong method used before it

[lost matrix]

Exceptionally does not pass through the result, so it will be a CF<Void> with exceptionally

[buoyant viper]

would i need to like return null and then in a .thenAccept check if null?

[slender elbow]

god I hate CF's handling of CompletionException it's absolute ass

in some cases it wraps exceptions in it, in other cases it doesn't, but it never unwraps them

[lost matrix]

There are like 10 methods that let you handle exceptions in some way.

[buoyant viper]

my alternative idea is sendAsync takes a Request + Consumer<Response> and Consumer<Throwable> lol

[slender elbow]

so you always end up having to run a stupid conditional unwrap method somewhere to get the actual exception you want

[lost matrix]

Dont mix functional forwarding with CompletableFutures. Your approach is fine.

[slender elbow]

exceptionally is more so a recovery function, the signature is CompletableFuture<T> exceptionally(Function<Throwable, T>)

[ivory sleet]

oh wait it doesnt?

edit: well im psure it wont rewrap CE at least, unwrap was a bad choice of word

[slender elbow]

so yes it does pass the result downstream

[buoyant viper]

i would ditch the CF if i did 2 callback consumers instead :P

something like java public void send(Request request, Consumer<Response> onSuccess, Consumer<Throwable> onThrow) { threadExecutor.execute(() -> { try { onSuccess.accept(send(request)); } catch (IOException ioe) { onThrow.accept(ioe); } }); }

[ivory sleet]

i like it

though, if its always an IOException u're catching, why not Consumer<IOException> ^^

[buoyant viper]

idk ¯_(ツ)_/¯

i could change the type ig

[sly topaz]

this function looks rather pointless

why not just use Java's HttpClient

[buoyant viper]

because im on java 8 😔

[ivory sleet]

my condolences

[random grove]

I've been working on a custom inventory GUI system and ran into what I think is a gap in the InventoryClickEvent API.

When MOVE_TO_OTHER_INVENTORY fires (shift-click), the event tells you the source slot but not the destination. The destination is computed at the NMS level before the event fires, so plugins have no reasonable way to access it without diving into NMS — which kind of defeats the point of having a Bukkit API.

This makes it really difficult to write a slot-based inventory library that correctly handles shift-clicks without either blanket-cancelling everything or reimplementing item routing logic that ultimately doesn't match what NMS actually does.

Would it make sense to expose the destination slots on the event? Something like event.getDestinationSlots(). Since the information already exists at that point, it feels like a natural addition.

Has anyone else run into this, or is blanket-cancelling really the accepted approach?

[sly topaz]

what people do is just duplicate the internal logic with the API

and yes, people have ran into this before but it is niche enough of a problem that nobody has bothered to fix the API around it

[random grove]

alr

[mortal vortex]

alr

[umbral ridge]

alr

[pliant topaz]

alr

[echo basalt]

alr

I did make an nms based slot thing but even then I often find myself emulating slots because everyone and their mom wants to have display items that represent "empty slots"

[potent crescent]

alr

[onyx fjord]

i hate InventoryClickEvent i hate it

did you know

[autumn zinc]

Does anyone know how to use ClientboundSetEntityMotionPacket ?

[lilac dagger]

you give it an id and a vector @autumn zinc

[autumn zinc]

I'm trying to use it, but without results

one min i'll send code

[lilac dagger]

show code

((CraftPlayer) player).getHandle().connection.send(new ClientboundSetEntityMotionPacket(id, new Vec3(x, y, z)));

this is exactly how i'd send it

[autumn zinc]

        Display.ItemDisplay entity = new Display.ItemDisplay(
                EntityType.ITEM_DISPLAY, ((CraftWorld) player.getWorld()).getHandle());
        entity.setItemStack(CraftItemStack.asNMSCopy(new ItemStack(Material.STONE_SWORD)));
        Location location = player.getLocation();
        entity.setPos(location.getX(), location.getY(), location.getZ());

        ServerGamePacketListenerImpl connection = ((CraftPlayer) player).getHandle().connection;
        connection.send(new ClientboundAddEntityPacket(entity));
        connection.send(new ClientboundSetEntityDataPacket(
                entity.getId(), entity.getEntityData().getNonDefaultValues()));

        Bukkit.getScheduler().runTaskLater(
                plugin, () -> connection.send(new ClientboundSetEntityMotionPacket(
                        entity.getId(), new Vec3(0, 1, 0))), 20L);

[autumn zinc]

The entity spawns successfully, but without velocity in a second

[lilac dagger]

i don't think item displays tick for this to work

try it on a villager

[autumn zinc]

i also tried with default zombie

[autumn zinc]

ok

[autumn zinc]

        Villager entity = new Villager(
                EntityType.VILLAGER, ((CraftWorld) player.getWorld()).getHandle());
        //entity.setItemStack(CraftItemStack.asNMSCopy(new ItemStack(Material.STONE_SWORD)));
        Location location = player.getLocation();
        entity.setPos(location.getX(), location.getY(), location.getZ());

        ServerGamePacketListenerImpl connection = ((CraftPlayer) player).getHandle().connection;
        connection.send(new ClientboundAddEntityPacket(entity));
        connection.send(new ClientboundSetEntityDataPacket(
                entity.getId(), entity.getEntityData().getNonDefaultValues()));

        Bukkit.getScheduler().runTaskLater(
                plugin, () -> connection.send(new ClientboundSetEntityMotionPacket(
                        entity.getId(), new Vec3(0, 1, 0))), 20L);

same thing

[lilac dagger]

looks correct

but is there a reason why you'd do it with nms?

[autumn zinc]

do you mean i should use protocollib?

[lilac dagger]

no no

do you need per player entities?

[autumn zinc]

yes, it is necessary

[young knoll]

I mean, the api can technically do that too

[autumn zinc]

mm, how?

[young knoll]

Entity#setVisibleByDefault
Player#showEntity

[lilac dagger]

interesting

so this doesn't add it to the world?

or no

it just hides it

pretty cool

my guess is you need to send the packet in small chunks over a period of time

[autumn zinc]

i know, but it is very important to implement my system using packets
without creating a real entity or something else

[autumn zinc]

yes

[autumn zinc]

interesing suggestion, i'll try

        Bukkit.getScheduler().runTaskTimer(
                plugin, () -> connection.send(new ClientboundSetEntityMotionPacket(
                        entity.getId(), new Vec3(0, 0.2, 0))), 20L, 3L);

Like this? if yes it doesn't work ...

[lilac dagger]

yes

try it

[autumn zinc]

it doesn't work either

i tried to intercept set motion packet when a default zombie was moving, and i got this data

[15:20:25] [Server thread/INFO]: X: -372 Y: -627 Z: -329
[15:20:25] [Server thread/INFO]: X: -364 Y: -627 Z: -339
[15:20:25] [Server thread/INFO]: X: -357 Y: -627 Z: -347
[15:20:25] [Server thread/INFO]: X: -356 Y: -627 Z: -348
[15:20:26] [Server thread/INFO]: X: -355 Y: -627 Z: -350
[15:20:26] [Server thread/INFO]: X: -192 Y: -627 Z: -192
[15:20:26] [Server thread/INFO]: X: -105 Y: -627 Z: -104
[15:20:26] [Server thread/INFO]: X: -105 Y: -627 Z: -104

        String stringBuilder = "X: " +
                packet.getXa() +
                " Y: " +
                packet.getYa() +
                " Z: " +
                packet.getZa();

        Bukkit.getScheduler().runTask(plugin,
                () -> Bukkit.getLogger().log(Level.INFO, stringBuilder));
``` used this code

but i know that these values are already converted

[lilac dagger]

try new ClientboundMoveEntityPacket.Pos(

[autumn zinc]

there are no sub classes in ClientboundMoveEntityPacket

oh

[lilac dagger]

it's 26.1

[autumn zinc]

yes

[lilac dagger]

not sure how it's on lower

[autumn zinc]

there are i'm sorry

[autumn zinc]

i tried it, but the entity doesn't move smoothly, but abruptly

[lilac dagger]

one second, let me try

[autumn zinc]

how is it going?

[lilac dagger]

just got the server going

let me see

it doesn't spawn at all @autumn zinc

might be something related to 26.1 tho

[autumn zinc]

i try it on 1.20.1

can you?

[lilac dagger]

i can't rn

[autumn zinc]

do you thing it is because of vesion?

[lilac dagger]

might be something new in 26.1

[autumn zinc]

are you sure that your way of creating the entity is okay?

[lilac dagger]

yes

server entity is required in newer versions

give me some more time to check

ah nvm

i got it

i had to move the position above server entity

[autumn zinc]

holy 26.1

[lilac dagger]

they don't move indeed

might be something else as well

server entity seems to hold the vector data

[autumn zinc]

set motion packet constructor

    public ClientboundSetEntityMotionPacket(Entity var0) {
        this(var0.getId(), var0.getDeltaMovement());
    }

    public ClientboundSetEntityMotionPacket(int var0, Vec3 var1) {
        this.id = var0;
        double var2 = 3.9;
        double var4 = Mth.clamp(var1.x, -3.9, 3.9);
        double var6 = Mth.clamp(var1.y, -3.9, 3.9);
        double var8 = Mth.clamp(var1.z, -3.9, 3.9);
        this.xa = (int)(var4 * (double)8000.0F);
        this.ya = (int)(var6 * (double)8000.0F);
        this.za = (int)(var8 * (double)8000.0F);
    }

maybe it will help you? 🙂

[lilac dagger]

it's not related to the move packet

[autumn zinc]

?

[lilac dagger]

the packet itself is a vector and entity id

[autumn zinc]

we are trying to move the entity using the move or motion packet?

[lilac dagger]

those are just logic for transfering

yes

but the constructor does work to prepare it to send it to the player/players

hmm

[lilac dagger]

movement is really complicated for some reason

i think there's multiple packets you need to send first

why wouldn't a move packet not move an entity

@autumn zinc really sorry, i don't think i have enough time to debug this

                 
  if (this.entity.needsSync || this.trackDelta || this.entity instanceof LivingEntity && ((LivingEntity)this.entity).isFallFlying()) {
                        Vec3 vec31 = this.entity.getDeltaMovement();
                        double d0 = vec31.distanceToSqr(this.lastSentMovement);
                        if (d0 > 1.0E-7 || d0 > (double)0.0F && vec31.lengthSqr() == (double)0.0F) {
                            this.lastSentMovement = vec31;
                            Entity entity2 = this.entity;
                            if (entity2 instanceof AbstractHurtingProjectile) {
                                AbstractHurtingProjectile abstracthurtingprojectile = (AbstractHurtingProjectile)entity2;
                                this.synchronizer.sendToTrackingPlayers(new ClientboundBundlePacket(List.of(new ClientboundSetEntityMotionPacket(this.entity.getId(), this.lastSentMovement), new ClientboundProjectilePowerPacket(abstracthurtingprojectile.getId(), abstracthurtingprojectile.accelerationPower))));
                            } else {
                                this.synchronizer.sendToTrackingPlayers(new ClientboundSetEntityMotionPacket(this.entity.getId(), this.lastSentMovement));
                            }
                        }
                    }```

i found this

but it doesn't tell me much

[autumn zinc]

When i use move packet it works, but motion - no

[autumn zinc]

thx so much, i'll check it

[lilac dagger]

all this does it checks if it's higher than a value and if it's a projectile it sends 2 packets, otherwise just one

[autumn zinc]

in this code sends only motion packet to move the entity, without move packet, i thought these two packets must be used together

[lilac dagger]

well, this is what i found

i'll try to look for 20 more minutes

[lilac dagger]

let's try seeing which packets get sent during movement

it has to be more complex than just this packet

[autumn zinc]

I understood, that smooth movement can be achieved without motion packet - you just have to send move.pos packet every tick

like calculate offsets using this code

short x = (short) (entity.getX() * 4096D - previous.getX() * 4096D);
short y = (short) (entity.getY() * 4096D - previous.getY() * 4096D);
short z = (short) (entity.getZ() * 4096D - previous.getZ() * 4096D);
new ClientboundMoveEntityPacket.Pos(entity.getId(), x, y, z, false);

previous it is a last location of the entity

well if you want to move an entity 5 blocks along X+ in 1 sec (20 ticks), you have to calculate offset per tick:

double distance = 5; // 5 blocks
double time = 20; // 1sec
double offsetPerTick = distance / time;

and then every tick send pos packet with adding an offset per tick

[lilac dagger]

delta time for the win

but i wonder why this packet doesn't work

it should do pretty much the same thing

[chrome beacon]

?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/

[lilac dagger]

so add entity takes a motion

doesn't work

entity set motion doesn't work

i can't pick anything during packet sniffing related to this

i assume the pos is when the villager tries to move while on the ground

and it's also not every tick

[thorn isle]

last i checked the protocol, entities do send pos/posrot every tick, regardless of whether they actually changed position or not

not sure when they send motion packets, i think that might be up to the entity property data sync whatever the fuck logic

[lilac dagger]

hmm

let me try one thing

send a motion packet and then a pos

it still doesn't feel as fast to be per tick in my testing

i am actually stubborn

i wanna figure out what mojang did exactly

my guess is they use the set entity motion to smooth out in between position changes

[thorn isle]

that's handled through regular interpolation

motion is for clientside predictions for the next tick

[lilac dagger]

that might be it yeah

well, that's sad

[thorn isle]

what were you trying to do?

[lilac dagger]

trying to help iwust above

he went with position update and delta time

i hoped there was a way to cut that packet

[chrome beacon]

I already told you where to post

?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/

[buoyant viper]

i lowkey got that shit on tho https://github.com/winnpixie/http4j

[autumn zinc]

i wanted to move entities smoothly, and i achieved this buut it works perfect only with living entities such as zombies or villagers, when i try to do it with item displays their movement is still jerky

[buoyant viper]

the funny part is originally i started the project as just the Server portion, and on a newer Java version (like 17 i think), but eventually downgraded to java 8 after i started building the Client code, since it wouldve been redundant to have mine + the built-in client

i should split them really but u never know when ur gonna need a HTTP Server on java 8..

however, not having access to features like record does negatively impact my health

[lilac dagger]

what do you use the http server for?

😄

[buoyant viper]

local site testing (before i learned about jwebserver lol)

[slender elbow]

how would it have been redundant?

[lilac dagger]

cool :d

[buoyant viper]

and there have been a few times ive embedded the server code into other programs for callbacks with api tokens like Spotify

[buoyant viper]

redundant in the sense that i wouldnt necessarily need it

id just use the built-in http client over my own

[slender elbow]

why downgrade to java 8 to begin with tho?

strange choice

[buoyant viper]

wanted to write client, wouldnt make sense if i stayed on newer java that already had a client framework

[slender elbow]

so you downgraded for the sole purpose of writing a client? and you wrote a client because you decided to write for java 8?

[buoyant viper]

kinda yeah

i do still have some other projects on 8 that could benefit from having a nicer api instead of rawdogging HttpURLConnection so it wasnt a complete waste

and thus have switched to my client framework, instead of rawdogged HUC

i could split the projects, me no wanna

regardless of the java version though... i need to learn how to properly read the full channel rather than relying on hoping everything fits in my initial buffer

so some server functionality is half-baked bc i dont bother to read past N(bufsize) bytes (reading content body in a request)

[slender elbow]

neat

[buoyant viper]

a non-issue when i was on java.io, but NIO has ever so slight differences bc of the Buffers that i lost some correctness when i initially moved over

actually

i have an idea now

i think i can just allocate a second (larger) buffer, copy the remaining from the first, and then read from the channel again

dealing with variable-length input when u have to use fixed-length apis sux idk how C devs do ts every day

[lilac dagger]

literally the one thing keeping me away from learning nio sockets

the buffer is so weird to use

netty is non blocking and it has a nice byte buffer to work with

(old io)blocking sockets too

[buoyant viper]

it's my own fault really, that i didn't think to redesign more when i decided to switch to nio. this is the consequence of my actions

[lilac dagger]

well, you did good tho

i have to get to learn nio sockets as well

even if i don't like working with them as much

probably once i figure them out i'll like them

[buoyant viper]

its a nice api honestly

[cunning flare]

jag dog och skule hemta mina saker och blev banad i 30 dagar för grov stöld??

[lilac dagger]

ja ja

seriously tho, keep it english please

[buoyant viper]

its not vastly different from using regular .io at the end of the day (at least for me), and being used to nettys bytebuf will probably give u an edge

[cunning flare]

men får man inte hemta sina saker utan att bli banad??

[chrome beacon]

Du är nog i fel discord :p

[slender elbow]

i mean the jdk http client will give you the raw response as a List<ByteBuffer> (or, well, you pass a Flow.Subscriber<List<ByteBuffer>>), you can do the same really

or if you want to keep it in a contiguous buffer you'd need to realloc and copy

[buoyant viper]

im talking for the server portion rn

[slender elbow]

sure

[buoyant viper]

the client is still using .io actually lol

[slender elbow]

but it's all the same

regardless of whether it's the client or the server, you have a SocketChannel, and you need to read data into a ByteBuffer, or a collection of ByteBuffers

[buoyant viper]

yes

[slender elbow]

i'm just saying that's what the jdk http client chooses to do

not a strategy only to be employed in such setting

[buoyant viper]

i think i can get away with 2 buffers if i give ample leeway for the http head

i will still enforce a max size for the content body (i think this boils down to arrays in java not? being able to exceed 2^31 - 9 in length)

[lilac dagger]

i wonder how the header and body are placed from a client

[buoyant viper]

http makes heavy use of newlines as a delimiter

\r\n specifically i believe

[lilac dagger]

so header here\r\nbody here?

[buoyant viper]

the split between head and body is when u have 2 consecutive newlines

[lilac dagger]

oh i see

interesting

[buoyant viper]

HEAD
Key: Value

Body```

[lilac dagger]

so each key and value come with a \r\n?

and when it comes to body twice?

[slender elbow]

now do chunked encoding!

[buoyant viper]

fukckkkk that

one request or nothing big dog

[lilac dagger]

what's chunked encoding?

[slender elbow]

it's still one request!

[buoyant viper]

yeah but its split up :(

[slender elbow]

:)

[lilac dagger]

oh so like multiple requests in one

insteresting

how does that work?

[buoyant viper]

thats above of my pay-grade

[lilac dagger]

doesn't sound impossible tho

[slender elbow]

instead of passing the content-length header, the body is split up in chunks that are length-prefixed (plus crlf)

[lilac dagger]

oh

[buoyant viper]

HTTP 1.1 feature iirc right

[lilac dagger]

you got this winpixie

[slender elbow]

in the case where you don't or can't know the full body length ahead of time

[buoyant viper]

and then HTTP 2.0 builds on that with multiple requests in one socket i believe

[slender elbow]

so you choose to stream it instead

[buoyant viper]

i dont even have properly reading one Full request 💔

my saving grace if the client detects i dont support chunking i think i can avoid it

i will need to stop reporting as HTTP/1.1 though, pretty sure

maybe 0.9 or 1.0

clients are pretty smart at adapting their function to the server

sadly, i am not smart at adapting to the client

[lilac dagger]

wait, does this work for anything that's not application/json?

[buoyant viper]

chunking can apply to anything i believe

u can reconstruct any content type so long as u have the data u need

[lilac dagger]

i assume application/json is just a header for the processor

nicee

i might make a http client

[buoyant viper]

Content-Type just tells the client/server... the type of the content being sent

[lilac dagger]

ye, i kinda figured it as i said

[buoyant viper]

ye

u could omit the field, but ur not gonna have a nice time, like, anywhere

[lilac dagger]

tbh, the http protocol is not as complicated as i thought

i mean sure there's chunking and such

but with enough tinkering you can get it to work

[buoyant viper]

just wait until u deal with HTTP/2 and HTTP/3

[lilac dagger]

lol

[buoyant viper]

jk those r probably "easy", im just dumb lol

[lilac dagger]

nah, you just need to get into hyper focus and add those features

[buoyant viper]

just spend like a day reading the RFC lowkey

im not sure what HTTP/3 looks like since its not on TCP anymore

it uses QUIC (UDP)

[lilac dagger]

oh udp

[buoyant viper]

requests might look relatively the same, but chunking i imagine is a must-have bc of the lossy nature of UDP

[lilac dagger]

ye

well, if i have time i'll give it a try tbh

[buoyant viper]

or well not lossy, but... limited single transmission size

(but also a little lossy)

[buoyant viper]

its makes for a fun afternoon project :3

i still have no SSL support tho 😿

[lilac dagger]

that's for securing http no?

[buoyant viper]

yes

[lilac dagger]

you can try it tbh

[buoyant viper]

my server can only serve HTTP rn, not HTTPS (unless u cheat and hide it behind a secured proxy) (cough cough cloudflare) (or on-site nginx reverse proxy cough cough)

[lilac dagger]

well, did you look into the implementation?

[buoyant viper]

nah

my server implementation is mostly based on reading RFCs addressed for HTTP and observations from what my own browser sends

not so much how others chose to impmement

[lilac dagger]

spring boot has a nice client impl

[buoyant viper]

waow

i should try and add a function (similar to newer javas HttpClient) that can apply a content mutator for u to use when building the request

rn i have a method in the Response class thats "good enough" that looks like java <T> T getBodyAs(Function<byte[], T> mutator)

[lilac dagger]

pretty cool

[buoyant viper]

javas client has BodyHandlers that u use when sending the request i think

<T> HttpResponse<T> send(HttpRequest request, HttpResponse.BodyHandler<T> handler)```

[lilac dagger]

i might at first try to use java to do this

and see how it goes

[robust helm]

how can i get the itemstack when throwing an trident?
im using ProjectileLaunchEvent with getShooter().inventory.itemInMainHand but that doesn't seem to work

probably because the trident is already in the air when the event is called and no longer an item

ngl i didnt try adding a debug message

but maybe someone knows from the top of their head

[buoyant viper]

?jd-s

[undone axle]

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

[buoyant viper]

cast projectile to Trident i believe @robust helm

[chrome beacon]

getItem on the Trident

[robust helm]

thanks that should work :)

[buoyant viper]

oh im so big brained as hell

[slender elbow]

what if I send way too many headers to fit in your initial buffer

[buoyant viper]

SocketChannel channel = getChannel();
ByteBuffer initialBuffer = ByteBuffer.allocate(8192); // 8KB buffer for header

channel.read(initialBuffer);
initialBuffer.flip();

// do processing with head buffer
doStuff(initialBuffer);

// read content body
int contentLength = getContentLength();
int remaining = contentLength - initialBuffer.remaining();
ByteBuffer contentBuffer = ByteBuffer.allocate(contentLength);
contentBuffer.put(initialBuffer);

while (remaining > 0) {
  remaining -= channel.read(contentBuffer);
}```

[buoyant viper]

i Blow up

[sly topaz]

is the project you're making an http client

because otherwise I don't understand why you're spending time with this lol

[buoyant viper]

this is for http server

[sly topaz]

I see

[buoyant viper]

heaven forbid i want to toy around D:

i like reinventing the wheel

even if there are significantly more squares

[sly topaz]

I just thought it'd have some relation to minecraft since well, this is spigot in the end

[buoyant viper]

it can

ie. embedded web server for a control panel

[sly topaz]

I do it often as well, when I want to learn

[buoyant viper]

(even if there are much better frameworks for such goals)

[slender elbow]

oh okay

good luck?

[buoyant viper]

if ur sending a header over 8KB i implore u to reevaluate ur life choices

thatss so many bytes

(cookies can easily become an issue)

there is a way to tell the client they sent a header or content body too large to be processed so i could just do that

i think im calculating my remaining bytes wrong actually

and i think it does have to do with the buffer potentially filling up without completing

so contentBuffer might accidentally fill up with data from headers

[buoyant viper]

ok so ive added a safeguard against that for now

[buoyant viper]

i will no longer blow up and leak memory into the content buffer, but i also just wont process the content body LOL

ill need to rethink how i read headers, or just tell the client "hey, this is 2 large 4 me"

ill need to rewrite my readLine function though to let me know when i lose the buffer midway through a line read

[mortal vortex]

inuendo for cumming btw

[mighty wind]

        <dependency>
            <groupId>me.libraryaddict.disguises</groupId>
            <artifactId>libsdisguises</artifactId>
            <version>11.0.0</version>
            <scope>provided</scope>
        </dependency>

just switched pcs

loading up an old project

anyone know why this might be really angry

and red

[lilac dagger]

did you change the repo?

<repository>
<id>libraryaddict-public</id>
<name>libraryaddict repo</name>
<url>https://mvn.lib.co.nz/public</url>
</repository>
@mighty wind

<dependency>
<groupId>me.libraryaddict.disguises</groupId>
<artifactId>libsdisguises</artifactId>
<version>11.0.16</version>
<scope>provided</scope>
</dependency>

i use this version

but there might a newer one

[mighty wind]

this is from like a year ago

so idk

[mighty wind]

will a newer version break like

1.21.8

nvm it works now i think

[lilac dagger]

yay 😄

just reload maven so it pulls the dependency from the repo

[thorn isle]

libraryaddict can't into maven so he keeps changing the artifact coordinates every 3 days

[lilac dagger]

i just put my stuff on jitpack

but libsdisguise had to change the repo since it was hosted on md5's repository

[thorn isle]

and then on mythic's repository

and then changed the artifact id and org 5 times

and even then 80% of the time the deployed artifact doesn't match what you'll get from spigot/ci

there's a good chance he's changed some internals in a breaking way and just pushed it under the same version/snapshot without a corresponding spigot update, or vice versa

i've entirely given up trying to depend on the artifact from the repo, i just build the plugin myself and deploy it on the server and to my maven repo

[half gust]

does anybody know how, in a datapack, how to replace the lava in the nether with blue ice? I was able to replace all the lava pools, but not the lava generated ontop of basalt deltas.

[chrome beacon]

Try asking in the world gen discord

https://discord.gg/BuBGds9

[slender elbow]

saw the invite layout and thought it was another hot women camgirl egirl nude 18+ sexy spam

[vagrant stratus]

ngl i almost jumped to banning them lmfao

[idle quarry]

hey, does anyone knows why I can't compile my plugin.

I updated from 1.21.11 to 26.1.2 and now I can't compile it (maven)

First error was java 25 since I had j21 but now the error is just can't find any code

[young knoll]

?notworking

[undone axle]

"Does not working" is a useless statement. Please describe what exactly is not working, what you expect it to do, and what actually happens. If you get any console errors, also ?paste the entire stacktrace.

[idle quarry]

The error literally says can't find [piece of code]

https://cdn.discordapp.com/attachments/720329859175415879/1493658717340700702/image.png?ex=69dfc595&is=69de7415&hm=ccbf5fbf78c39ed2eac6fe75b002b3a9588cbfc79d9b667f1a7736cc1d0375c0&

i suspect something to do with Lombok but getters are working fine in the ide

[young knoll]

Does your ide show an error at that line

[idle quarry]

nope

same version of the code can be compiled with 1.21.11 and java21

I was trying to update it for latest version but ....

[chrome beacon]

Try not setting Java 8 as your Java target

[idle quarry]

I tried in pom 25

same error

[sullen marlin]

You need to update Lombok too

[idle quarry]

i did to the latest version

Latest is .44 (?)

[sullen marlin]

Oh you also need to add it as an annotation processor

https://hub.spigotmc.org/stash/projects/SPIGOT/repos/buildtools/commits/a0bbb6b111068fe89bd11730c8e66f9c1706598a#pom.xml

See maven compiler plugin

[idle quarry]

Thanks, i'll take a look soon

[idle quarry]

?paste

[undone axle]

Looking to paste something? Try a code block or one of the following websites:

• https://pastes.dev/
• https://sourceb.in/
• https://mclo.gs/ (best for server logs)

[idle quarry]

Still doesn't work, here is my pom
https://pastes.dev/jJntF3D4Y8

[chrome beacon]

Still missing the annoation processor?

[idle quarry]

okay it is fixed thanks everyone

also anyone can explain why is that needed for newer versions

i feel like every new version there is going to be something new that will fuck up my compiling process

does gradle have the same issues?

[fickle spindle]

which is the configuration library most used in plugins / public and private, because i'm starting to have the problem that it get all organised wrongly (i know i did a similar question in the past but today is the day i need the final answer)

[chrome beacon]

If you think that's a pain the you really will not like Gradle

They break stuff all the time

Anyways yes that would have happened in Gradle too

[young knoll]

The solution is to not use Lombok

[pastel swan]

What’s the first farm you build every new world?

[idle quarry]

I hate adding like 50 getters setters and things like that

😭

[worldly ingot]

I mean your IDE can auto generate them in one click :p

The solution REAL developers don't want you to know about: Make all your fields public! 🤫

[idle quarry]

haha, I'm so used to private fields

[worldly ingot]

As you should. I'm memeing

Encapsulation is important lol

[idle quarry]

yeah maybe not that important in plugins but in other aspect is 100%

[fickle spindle]

Bump (still looking for a solution / good library suggestion)

[idle quarry]

checkout packetevents

[lilac dagger]

all i type is get or set and the autocompletion already shows it

[worldly ingot]

I just type the field name, but yeah, whatever floats your boat

[worldly ingot]

Takes a split second to generate them, or you can use an actual dedicated menu for it to generate all of them at once for all your fields

[lilac dagger]

oh

this is better

thx, i'll type the field name instead

[remote swallow]

If your on intelllij iirc default generate keybjnd is alt+insert and you can generate getters settings equals hash code override etc

[lilac dagger]

it generates it

but not where i want it

oh nvm

it generates at cursor position

perfect

@remote swallow thx sir

this is the way

[remote swallow]

[slim wigeon]

As you know, I came from SA-MP with player ids like the yellow text. So what I asking here, is there a api like Vault that assigns player ids and I get that from the api into my plugins I create?

[mortal vortex]

SA-MP
?

[slim wigeon]

San Andreas MultiPlayer

[sullen marlin]

What do you mean player id

[mortal vortex]

Who the fuck knew that, who are you addressing with "as you know"

[sullen marlin]

Groups?

[mortal vortex]

Are you sure the thing in the square brackets is an ID and not like... idk, a level system?

player ids like the yellow text
That yellow text looks like a level system, seeing as YOU are on 0, and someone with "OG++" is 201.

[slim wigeon]

I forgot these ranks, I not trying to adv but I just thinking I could get a player id system like on my SA-MP server. If that is not possible, that is ok because my bots on the IRC server is programmed to accept player names or IDs

[mortal vortex]

Why do you need an API for 10 lines of code?

[slim wigeon]

That is possible but last I heard, the data don't transfer between plugins. Yes, there is this api like Vault that made this possible but I don't know what they did

[mortal vortex]

i have literally never heard of this

[mortal vortex]

Make ur own API then.

It's not like you're losing integration because, given how I have literally no idea of an API like this existing, its likely that not many plugins are hooking into whatever this API might be.

You might as well make the API urself.

[slim wigeon]

I looking at my source code and found this: private boolean setupShop() { if(getServer().getPluginManager().getPlugin("MGN_ServerShop") == null) { return false; } this.shop = (com.mrnategeek.serverShop.Main) Bukkit.getPluginManager().getPlugin("MGN_ServerShop"); return true; }That is one way which I going to use. Don't mind me, my mind. Its been a min since I worked on this. Don't mind me, sorry

[mortal vortex]

I'm not sure what you're trying to say

[quaint basin]

https://pastes.dev/w2LD9C2fP0 How can I guarantee that PlayerJoinEvent will be called if AsyncPlayerPreLogin is called? Perhaps this isn't possible, and I need to remove the map key from the cache if the player doesn't log in within x seconds (where x = timeout seconds)?

I imagine that if a plugin uses event#disallow, then the playerjoinevent will not be called

[slender elbow]

join event is not called if async pre login doesn't go through

so you already have that guarantee

join event is already after the login process when the player entity is being spawned and the player data broadcast to other players

if the login events are disallowed, then none of that can happen as there is no player to spawn

[mortal hare]

Whoa

nice glitch

[quaint basin]

But what if a plugin executes event#disallow on AsyncPlayerPreLoginEvent? Will PlayerJoinEvent still be called?

[sly topaz]

if you disallow the player gets kicked

so they don't join the server

this kind of thing is easy to test though, you can just give it a try

[lilac dagger]

If you disallow it won't go further

Player join event is happening at a later stage

[dark arrow]

hello I had a question
like I know all note block states and mushroom block state textures can be altered in texture pack , what else can I alter to get more textures that I can cutomize

[chrome beacon]

I believe there are some note block states you can use

[thorn isle]

waterlogged double slabs are unobtainable in survival and can be safely replaced (but they will act as water source blocks)

the age:0 sapling states appear identical to the age:1 saplings and can be safely replaced, although it's more complicated as you want a protocol level filter that replaces natural age:0 saplings with age:1 saplings

the nice part about saplings is that they're not solid and so won't occlude any adjacent block faces

[random grove]

If I schedule an async task, and it's runnable is in the middle of running, and the server shuts down while the runnable is running, will the server wait for the runnable to complete?

[thorn isle]

kind of, but not in any proper or controlled way

after the server is otherwise completely shut down, it will wait an extra 10 seconds for bukkit async scheduler threads to complete

they will not be interrupted and bukkit does nothing to signal to them that the server is shutting down; you can't rely in InterruptedException and have to manually check on a loop whether to exit the thread

if the threads don't complete in that 10 extra seconds, they just get killed and the jvm terminates

this is one of my major pet peeves, making long-running and blocking async tasks much more annoying to write correctly, as bukkit doesn't use the first class language feature specifically intended to notify threads to clean up and terminate that's been in the language for like 20 years

[random grove]

Alright, thanks. I will just have to make something on top of it then.

[thorn isle]

my personal recommendation is either using your own executor/scheduler and terminating it in onDisable, as this will fire interrupts correctly

or manually propagate interrupts to your bukkit async task threads, but that's kind of ass

[thorn isle]

e.g. suppose you block on say a http request that'll time out in 60 seconds, normally you'd just do like client.send(whatever).join(), but this is "incorrect" as the thread will have no clue that the server is stopping if it does, and just continue blocking until it gets killed, with no chance to do cleanup and possibly leaving things in a corrupted state

in bukkit what you have to do instead is

long started = System.currentTimeMillis();
Response response = null;
while (System.currentTimeMillis() < (started + 60_000) && !Bukkit.isShuttingDown()) {
   try { response = future.orTimeout(5_000, TimeUnit.MILLISECONDS).join(); }
   catch (TimeoutException ex) { continue; }
}

which is as you can see SO much easier than just try catch interruptexception

i asked about this at paper once, and it still makes me mad

"we don't want to do interrupts because they're hard to get right and confuse plugin developers, and then you'd have to check for and catch interrupts everywhere in async task code"

YEAH that's kind of the point of the exception; if you do something that blocks the thread, it's your main (and sometimes only) mechanism to pull back

what, are we intended to try catch ignore it in async task code instead and ad-hoc it in 200 different ways with manual boilerplate?

[young knoll]

Leaves with different distance values, tripwires in the active state

[sly topaz]

does this happen with Folia's AsyncScheduler too

I vaguely remember that one just cancelling tasks when terminated, don't remember if it awaits or not (though I imagine it should? How do you deal with classloaders closing before the task is finished otherwise)

[buoyant viper]

i think im gonna need to make that registry

(i dont like the current way i load my central bank)

what would be a good way to manage them in database? im thinking a table that has bank names, and generated tables for each bank

CREATE TABLE IF NOT EXISTS banks (name VARCHAR(255) PRIMARY KEY)

CREATE TABLE IF NOT EXISTS TABLE bank_bankname (owner VARCHAR(36), account VARCHAR(255), balance BIGINT, UNIQUE(owner, account))```

[grave abyss]

is there a way to decompile a minecraft client class when working on a plugin? (ping on answer pls)

[young knoll]

You’d have to add the client as a dependency

[grave abyss]

oh okay

[young knoll]

Or just do it externally

[mortal vortex]

blud did not ping on answer

[worldly ingot]

Like you just want to look at client source code?

If so, https://mcsrc.dev

[mortal vortex]

oh shit dis new?

[worldly ingot]

Relatively. Fabric people

[young knoll]

Wait that’s really nice

[ivory sleet]

is there a way to switch between vineflower and cfr?

[young knoll]

It even works on mobile for when I get those brain blasts in bed at 3am

[mortal vortex]

bruh istg ur always on mobile

[buoyant viper]

Real

[slender elbow]

no, but you can plop the client jar into slicer.run for that

[paper crest]

Hey, when I switch to Java 25 Lombok is not working. I have the latest version.
<dependency> <groupId>org.projectlombok</groupId> <artifactId>lombok</artifactId> <version>1.18.44</version> <scope>provided</scope> </dependency>
On Java 21 works fine.

[dry hazel]

did you add it as an annotation processor?

maven doesn't scan the classpath for annotation processors for projects targeting JDK 23+ anymore

https://projectlombok.org/setup/maven

[dry hazel]

you can also press a button on this site to get minecraft with mappings loaded in there automatically https://minecraft.katana-project.org/

[quaint basin]

https://pastes.dev/5d8kHeL59a Do I have a guarantee that the value of serverId on line 97 will have the value initialized on line 78 (the value read from the redisExecutor thread) if I never redefine it anywhere else except on line 78?

[hazy parrot]

yes

[quaint basin]

Who guarantees that?

the JVM model?

[hazy parrot]

it is going to read static variable every time it is executed

so yeah, if you are not modifying it elsewhere, it is guaranteed

[paper crest]

I added this and is still same

is not abstract and does not override abstract method

[quaint basin]

Whether it's static or not, it doesn't change anything