help-development

[flat leaf]

you can set block data

but things like chests still won't have, for instance, items in the block inventory

and containers also use NBT to lock

[noble lantern]

ohh nbt for tile entities

[flat leaf]

yeah sorry that's what I meant

[noble lantern]

thought you meant like stone

was like huh lmao

You should just be able to get the NBT block on generation no?

[flat leaf]

you can set NBT data through NMS or even worldedit's API but you're not supposed to directly set blocks through the world during generation

what I currently tried was telling a scheduler to set the block's NBT a tick later and it technically works but it haults the whole server until the chunk is generated

which is really bad

[noble lantern]

idea

keep a referance to those locations

generate al lour chunks, and when a player goes nearby you generate chest contents

this is how vanilla mc does it

Go in spectator, try and open a chest far away

"Loot not generated" will show up

[flat leaf]

well chests have loottables, they generated once you open them

but afaik the loot tables are also stored in NBT

thing is these chunks only really generated when the player is near

[noble lantern]

Indeed but as far as this goes there not a lot of ther ways to handle it

[flat leaf]

i guess I could wait till they're really near but I still feel like there'd be a way to slow down the server if you're using, say, an elytra and travelling really fast

[noble lantern]

Dont exactly see an issue doing it that way unless its hindering performance on your generate code

I mean i would assume vanilla mc applied the nbt loot tables on chest generate

Theres a command to give you a placeable dungeon chest, and i know you directly apply NB directly when giving that item

Wouldnt see why it would be different for chunk generation, set nbtdata on generate

[zenith gate]

lol tell me about it, I wasn't thinking, I forgot I could totatly use a runnable inside the interact of the item.

[noble lantern]

in fact i think a plugin i worked on in the past did it like this

one sec

lemme look

Yep it did that but its oooold code

back when the chunk populator method was populate()

but its still same principle, on generation of the block apply the data or fill container

[flat leaf]

issue is it does

every time a chunk generates where NBT needs to be set and I run the code to set it, the server stalls for a good 5-10 seconds

I think the reason is that when running it on the main thread, I have to get an instance of the chunk and that means it has to wait until the chunk generates since worldgen is on another thread

[lime moat]

Hello, I have this code in a command: https://paste.gg/p/anonymous/f0f4320a9a884daba10b9309059faee1

Then I have this code as a fill: https://paste.gg/p/anonymous/37824ea968ec45698c93194b709d44bf

What should I do at line 37 to make it work?

I'm trying to make all the logs spawn at the same chance... this system previously used HashMap

A lot of stuff is probably wrong, so any guidance would help so, so much!

[noble lantern]

What event are you using by chance

[flat leaf]

I can't even manually set the block inventories either if I wanted to just handle containers

I'm not using an event, this is a chunkgenerator

in generateSurface()

meaning I have to set block types with the ChunkData object it gives me instead of directly going through the world

[noble lantern]

hmm

[zenith gate]

this should work in onEnabled right?

Player player = (Player) Bukkit.getOnlinePlayers();
        new BukkitRunnable() {
            @Override
            public void run() {
                Player.Spigot spigot = player.spigot();
                spigot.sendMessage(ChatMessageType.ACTION_BAR,
                        new TextComponent("❤" + ChatColor.RED + "" + ChatColor.BOLD + Math.round(player.getHealth()) + " " +
                                ChatColor.AQUA + "\uD83E\uDDEA" + ChatColor.BOLD + Mana.getPlayerMana(player)));

            }


        }.runTaskTimer(Wired.getPlugin(), 0,5);

[noble lantern]

you may have better luck using the chunk populate event (Might not be called that)

Since those give you direct access to Block

[noble lantern]

What does it do at line 37?

[noble lantern]

sadly yes

oh wait thats a timer

nvm

yeah itll work

[lime moat]

Well, using HashMap entrySet did this: https://fwoostyhub.com/🎅🏒📻🎉

but now I'm just using a list so I'm dumbfounded

[noble lantern]

ores is a list?

[flat leaf]

you mean use a blockpopulator?

[noble lantern]

Yes, one of those events

It should be really lightweight all your doing is setting block data

[flat leaf]

oh you mean literally hook onto the event

[noble lantern]

yee

[flat leaf]

hmmm

[noble lantern]

it gives you direct access to block

[lime moat]

This is the HashMap system which works fine: java int x1 = -114, y1 = 62, z1 = -55; Location loc1 = new Location(Bukkit.getWorld("world"), x1, y1, z1); int x2 = -102, y2 = 52, z2 = -67; Location loc2 = new Location(Bukkit.getWorld("world"), x2, y2, z2); Map<BlockData, Double> ores = new HashMap<>(); ores.put(Bukkit.createBlockData(Material.COAL_BLOCK), 0.15); // 15% ores.put(Bukkit.createBlockData(Material.IRON_ORE), 0.15); // 15% ores.put(Bukkit.createBlockData(Material.GOLD_ORE), 0.05); // 5% ores.put(Bukkit.createBlockData(Material.REDSTONE_ORE), 0.04); // 5% ores.put(Bukkit.createBlockData(Material.DIAMOND_ORE), 0.03); // 3% Cuboid.fillCuboid(loc1, loc2, Bukkit.createBlockData(Material.STONE), ores);

[noble lantern]

and your rly not making a world gen

[flat leaf]

that seems kinda gross but that may work

[noble lantern]

indeed it is slightly gross

[lime moat]

I just need to make basically the same system to work with a normal list so they have all the same chance of spawning

[noble lantern]

But its fine since your only doing one thing

if you were generating blocks in that event like a mountain

then oof

[flat leaf]

yeah no I can just save in a list the blocks that need their NBT saved and handle it in the event later

[noble lantern]

so you basically are removing chances

[flat leaf]

that should theoretically work

[lime moat]

Yes, exactly

I have really no clue what else to do and I've been guided on this journey :P

[noble lantern]

for (Something something : thatList) {

}

you can just set it in there directly

unless thats not exactly what your asking

i assume your just asking how to iterate through it

[lime moat]

It may be, but by chance could you give me an example of how I would use it in my instance? I'm so lost with this :P

[noble lantern]

I mean its relatively simple

you change ores to a List of BlockData

From there you change your put(BlockData, float) method to add(Bukkit.createBlockData(. . .));

And then you iterate over it like so

for (BlockData blockData : ores) {

}

[lime moat]

I currently have this with my fill: java if (woods != null && woods.size() > 0) { // picking random ore BlockData blockData = woods.get(random.nextInt(woods.size())); int i = 0; for (BlockData blockdata : woods) { if (random.nextDouble() <= entry.getValue()) { loc1.getWorld().getBlockAt(x, y, z).setBlockData(entry.getKey()); } // guess failed. Use base. else { loc1.getWorld().getBlockAt(x, y, z).setBlockData(base); } // break loop, we've done our work. break; } }

                        int x1 = -89, y1 = 62, z1 = -81;
                        Location loc1 = new Location(Bukkit.getWorld("world"), x1, y1, z1);
                        int x2 = -77, y2 = 52, z2 = -93;
                        Location loc2 = new Location(Bukkit.getWorld("world"), x2, y2, z2);
                        List<BlockData> woods = new ArrayList<>();
                        woods.add(Bukkit.createBlockData(Material.SPRUCE_LOG)); // 12.5%
                        woods.add(Bukkit.createBlockData(Material.BIRCH_LOG)); // 12.5%
                        woods.add(Bukkit.createBlockData(Material.JUNGLE_LOG)); // 12.5%
                        woods.add(Bukkit.createBlockData(Material.ACACIA_LOG)); // 12.5%
                        woods.add(Bukkit.createBlockData(Material.DARK_OAK_LOG)); // 12.5%
                        woods.add(Bukkit.createBlockData(Material.CRIMSON_STEM)); // 12.5%
                        woods.add(Bukkit.createBlockData(Material.WARPED_STEM)); // 12.5%
                        Cuboid.woodfill(loc1, loc2, Bukkit.createBlockData(Material.OAK_LOG), woods);``` and this in the command

Now, what would I do in these two lines? 😅 java if (random.nextDouble() <= entry.getValue()) { loc1.getWorld().getBlockAt(x, y, z).setBlockData(entry.getKey());

[noble lantern]

oooops

sec

for (int x = 0; x <= (ores.size() - 1); x++) {
ores.get(x);
}

you can use x as the iteration count/index getting for the List

all lists start at 0 (I assume you know that, and hence why you -1 from size())

if you dont you get a IOB error (indexoutofbounds)

[lime moat]

I replace this from the previous for loop you gave?

[noble lantern]

yep!

also

make sure you do a isEmpty() check before-hand if the values ever bcome non-hardocded

not that it will rly matter but its decent to catch that

[lime moat]

https://paste.gg/p/anonymous/5dcd8face1b640d3887a28387b2c55c9 I current have this, does it look kinda right?

Line 37 is what I changed

[noble lantern]

yes but your still using entry.getKey() which is prolly screaming red in your ide

How did you manage to write all this without knowing how to do this lmao

[lime moat]

Heh, lots and lots of help n googling, so now I need to remove java if (random.nextDouble() <= woods.getValue()) { loc1.getWorld().getBlockAt(x, y, z).setBlockData(entry.getKey()); and throw ores.get(x); in the place?

[noble lantern]

yep just remove the if, keep the body of the if and replace with List#get

Also is there any reason for you using block data to set blocks?

I feel like Material may suit you better unless you do plan on setting block data later on, then disreguard that part

[lime moat]

Nope, besides that's what I've used in the past :P

I'll try it on the server now! :D

[shadow zinc]

How can I power a piston with code?

[quaint mantle]

((Piston) block).setExtended(true);

[shadow zinc]

no way its that easy

[quaint mantle]

Ok

[shadow zinc]

you're talking garbage

[quaint mantle]

Check again

[lime moat]

I have this now: https://paste.gg/p/anonymous/98ac9b03acc54b08a16b88da23c408e0 and it replaced the whole mine with Spruce :P

[shadow zinc]

its not going to be any different?

[quaint mantle]

.

[shadow zinc]

yeah that doesnt work

[quaint mantle]

How so

[shadow zinc]

it updates the texture

it doesnt actually extend or push any blocks

[noble lantern]

Did you set the new block data

[shadow zinc]

yes

otherwise its texture wouldn't update?

[quaint mantle]

He means did you update the state

[noble lantern]

not necesarily

[lime moat]

I believe we have to incorporate BlockData blockData = woods.get(random.nextInt(woods.size())); somehow to get a random block.

[noble lantern]

final Piston piston = (Piston) block.getBlockData();
piston.setExtended(true);
block.setBlockData(piston);

[shadow zinc]

                            //For piston
                            if (data instanceof org.bukkit.block.data.type.Piston piston) {
                                //extend piston
                                if (cachedData.cachedRedstoneActivity.get(block) > 0) {
                                    piston.setExtended(true);
                                    block.setBlockData(piston);
                                    continue;
                                }
                                continue;
                            }```

[noble lantern]

yeah like that, your code you sent above didnt have it

[shadow zinc]

that was an example showing setpowered doesnt exist

[quaint mantle]

I dont think that will do anything

I dont think it would return a clone of getBlockData

[noble lantern]

he didnt say setPowered

setExtended

[noble lantern]

Piston inherits BlockData

[quaint mantle]

Doesnt change anything

@shadow zinc try this https://hub.spigotmc.org/javadocs/spigot/org/bukkit/block/BlockState.html#update(boolean,boolean) after making it extended

[quaint mantle]

Wtf

[shadow zinc]

you tell me

if that worked

[noble lantern]

tf

[shadow zinc]

i mean it might push the block

let me check

[quaint mantle]

Shitty client

[shadow zinc]

sure

[noble lantern]

yeah has to be client issue

that is no way possible on vanilla

[quaint mantle]

Stupid fabric

[noble lantern]

unless your textures are corrupted

[shadow zinc]

doesnt push the block

[noble lantern]

oh thats fabric?

[shadow zinc]

thats not client right?

[noble lantern]

Try using the block state update for players in radius

[quaint mantle]

shouldnt even have to

[noble lantern]

i didnt mean to say that

[quaint mantle]

but maybe his client is doing some sort of optimization

[noble lantern]

idk why idid

i mean to say just block state in world lol

however

he may need to do that too

just for that client

[flat leaf]

nah I don't think it's the client

I'm pretty sure setExtended just sets an NBT value

that's my best guess

[noble lantern]

yep

thats right

just checked

[flat leaf]

and that makes the block look proper when the block for the piston stem gets created

[noble lantern]

https://hub.spigotmc.org/javadocs/bukkit/org/bukkit/block/data/type/Piston.html#setExtended(boolean)

[quaint mantle]

Shit game

[flat leaf]

the actual logic for creating the extension block and moving it is not there

an API request could maybe be made but idk it just kinda is what it is

[shadow zinc]

bruh am I going to have to remove the block and set it one forwards

[flat leaf]

speaking of NBT though I really wish there was an API in bukkit to set NBT

[noble lantern]

got it!!!

Cast Piston

to Powerable

[flat leaf]

I would personally take a look at how minecraft does it

[noble lantern]

https://hub.spigotmc.org/javadocs/bukkit/org/bukkit/block/data/Powerable.html

there you go 🙂

[flat leaf]

yeah that probably works

[shadow zinc]

lol no

its not a powerable

[noble lantern]

or the block data to Powerable

try that instead

[shadow zinc]

i mean i got an instance check

it doesnt fly

[flat leaf]

I get why there's not, making the API consistent and shit between versions

[noble lantern]

Might as well try it

[flat leaf]

but man things like this make Bukkit in general pain to use sometimes

[shadow zinc]

okay

for you

[noble lantern]

https://www.spigotmc.org/threads/make-piston-move.417551/#post-3689731

If not then you need to do this

thats the most recent thread i can find

from 2020

so hopefully it still works

[flat leaf]

^^ keeping in mind though you don't have to suffer with obfuscated names anymore

[noble lantern]

as long as your using maven at least

[flat leaf]

so if you are using mojang mappings (which you really should be, they're great) it will not be called .a

[noble lantern]

spigot dont care for gradle users :((

[flat leaf]

f

why not just use maven

[noble lantern]

no one has converted special source plugin from maven to gradle and it suuucks

cause me like gradle

[flat leaf]

understandable

well yeah that's unfortunate

having mappings is great

[shadow zinc]

everyones favourite error

java.lang.RuntimeException: java.lang.ClassCastException: class org.bukkit.craftbukkit.v1_19_R1.block.impl.CraftPiston cannot be cast to class org.bukkit.block.data.AnaloguePowerable (org.bukkit.craftbukkit.v1_19_R1.block.impl.CraftPiston and org.bukkit.block.data.AnaloguePowerable are in unnamed module of loader java.net.URLClassLoader @5c29bfd)

[noble lantern]

yeah anytime i wanna do mappings i typically convert my project back to maven lmao

[shadow zinc]

oof

wrong one

[noble lantern]

@shadow zinc read article i sent

[shadow zinc]

ik

[flat leaf]

fun too to see what the community guessed right with their mappings and what is completely different

like tile entities actually being called block entities by mojang is very jarring

but makes more sense of course

[noble lantern]

i dont get why mojang just doesnt publicize theyre API

are they scared cause its so shit?

[shadow zinc]

yeah same thing

i'll read the article

[flat leaf]

wdym API

[shadow zinc]

even tho I have read it

[noble lantern]

IE the API mojang themselves use to make the game

[noble lantern]

i doubt that

If you did you would of told us you tried NMS

[shadow zinc]

oh yeah, I've tried NMS

[flat leaf]

I mean as far as I know there's no API, just the source code

which we basically have access to now except not really

Minecraft is probably the closest thing to actually being open source right now but not quite

we have names for everything, just not the actual original .java files

[noble lantern]

well they would basically have theyre source code as the API,

But instead they decide to obfuscate and make nms users cry

[noble lantern]

What exactly did you try for NMS?

thats gonna be your only option now i beileve

[flat leaf]

obfuscating in this case was really just not saving names of stuff when compiling

[shadow zinc]

idk some weird article I found, problem is all of these things are for no remapped

so a or b meaning nothing to me

[flat leaf]

the logic is still all the same, we still have to decompile everything

[noble lantern]

well look at the args

those are update methods

[flat leaf]

it's just that now when we do decompile minecraft we can get the names of everything back

[noble lantern]

Yeah i dont get why they dont just make an API jar

that mirrors methods from the server jar

it would make alllll this server shit so much easier

and the mirror jar is just empty code methods

and set scope to provided

[flat leaf]

I mean mojang doesn't really have an API like bukkit, there's no reason to

they very frequently change the structure of lots of parts of Minecraft

[shadow zinc]

This really means nothing to me, its outdated

[flat leaf]

which is a good thing, Minecraft has far exceeded its original planned scope and was designed by a swedish dude in his off time originally

[noble lantern]

I would assume they have a basic event system and NBT handler though

and those would likely not change very often

cause its like core minecraft for those

[flat leaf]

they have NBT handling shit in NMS

[noble lantern]

@shadow zinc

[flat leaf]

and they don't necessarily need an event system but they do have sorts of events systems in minecraft's code

the reason we don't use any of that shit though is because it's not meant to be used as an API as it's just the game's code

so they change stuff around often

[noble lantern]

i knowww but i hate building all the shit myself 😦

[flat leaf]

the point of Bukkit is to have one uniform API that wraps around minecraft and deals with what changes in the background without the developer having to know or really care

[noble lantern]

mojang could just like

write good code

so they dont need rewrites every week

:))

[shadow zinc]

Sorry, I don't know what you are referring to

[noble lantern]

You need to get a NMS BlockPiston

and call its update methods

[flat leaf]

all I will say is I do not envy Mojang developing a voxel game in Java

Java would be my absolute last choice for something so memory intensive

[noble lantern]

it might be named differently than BlockPiston in remapped tho

[shadow zinc]

of ffs why isnt it remapped

[noble lantern]

If handled right its fine

[shadow zinc]

a b c die

[noble lantern]

but most devs leave mem leaks all over the place

especially in games

[flat leaf]

sure but the core structure of java is lacking for memory intensive applications

[noble lantern]

its 10x harder to try and avoid mem leaks

[flat leaf]

like afaik you can't allocate complex structures to the stack... like what

if I'm wrong on that correct me but I could never fucking find out how

[noble lantern]

explain complex structures?

Wym by that? Like large complicated objects?

[flat leaf]

basically being able to put objects on the stack

[shadow zinc]

why isn't it remapped?

[flat leaf]

C# has structs which are like objects but immutable

[shadow zinc]

is this something I did?

[flat leaf]

and you can lay out a struct using other value types like primitives and other custom structs

[noble lantern]

Did you follow the remapping guide?

[noble lantern]

https://stackoverflow.com/questions/10481528/in-java-which-objects-are-put-on-the-stack-and-which-on-the-heap

I found this

[flat leaf]

and they are automatically put on the stack instead of having to go on the heap and be managed and eventually garbage collected

[shadow zinc]

im using gradle

so I had to do some stuff

im using paper weight too

[noble lantern]

yeah youll need maven for remapped

theres prolly a way in gradle

but im not that good

[shadow zinc]

I mean most of my stuff is remapped

never had problems until now

[flat leaf]

so basically what I'm getting is: you cannot directly create something on the stack, but the JVM will try to analyze how you create objects and put them on the stack if it's safe to do so

[noble lantern]

Yeah in a sense

[flat leaf]

I far prefer C#'s way of handling this still but that's at least better than not being able to put anything other than primitives on the stack at all

[noble lantern]

Yeah im sure long running things are added to the stack more than likely

like socket instances and theyre values likely get shoved to stack while short lived things like threads and what not run through heap

Just my best editmate tho

[flat leaf]

I just hate how much is delegated to be part of the magic internals of the JVM instead of up to the developer to control

especially since how the JVM behaves is not necessarily always guaranteed depending on what JVM you're using

[noble lantern]

Seems like the comunity has a few different answers for this ngl

i wonder if its possible to create your own JVM

[flat leaf]

sure it is

[noble lantern]

and make java stack based instead of heap based

prolly a lot of work

too much

ide rather write a mc server jar from scratch

[flat leaf]

probably a waste of time but sure you can, all the JVM does is interpret the byte code and tell the computer how to run it

[noble lantern]

yep, imma google that now i wonder if its been done

[flat leaf]

you wouldn't want to put everything on the stack though, there's a heap for a reason

but the stack should be delegated for smaller objects that aren't necessarily supposed to change

no need to create a thousand small objects a second that only get used immediately just for them to all clog the garbage collector

[noble lantern]

Hmm so make global scope finals, statics, and long running objects on the stack

https://stackoverflow.com/questions/731143/how-to-create-custom-jvm

Eww

im boutta make my own programming language

[flat leaf]

actually usually you want long running objects to go on the heap

the stack only has so much space

it's usually short lived stuff you want to go on the stack

[noble lantern]

so stack wouldnt work for large complex voxels then no?

cause typically those voxels could be long lived

eg loaded chunks

[flat leaf]

usually you put stuff on the stack while, say, doing a ton of calculations in a loop

if you're trying to store data in some sort of structure while iterating over every block in a chunk you don't need to make an object

that structure you put on the stack is automatically freed as soon as the current iteration ends

instead of having to go through the GC

[noble lantern]

ahh i see was just about to say doesnt the gc replicate that haha

TBF the GC has gotten rly good over the years

compared to when mc first came out

[flat leaf]

could also be useful when, say, sending information to the GPU in a certain way

yeah no modern java versions are much better at it

[noble lantern]

i actually dont think MC interacts with gpu much

[flat leaf]

well it does, minecraft just doesn't do many intensive graphical calculations

because it's a simple block game

it would still be stupid to render on the CPU

[noble lantern]

Honestly cpu bound games are weird

Rust is cpu bound

[flat leaf]

it's moreso that Minecraft just spends way more time per frame on actual calculation instead of graphical calculation

[noble lantern]

and thats a full fledged AAA 3d game

[flat leaf]

basically your cpu is doing more work comparatively to your GPU

[noble lantern]

multithreading when

: ' |

someone should shoot notch for not making the code async ngl

[flat leaf]

multithreading is used very frequently throughout Minecraft's code, I think vanilla even does world gen on another thread

but you can't make everything on another thread

[noble lantern]

yeah it does but theres still a bunch of stuff handled on main

IIRC entities and stuff is all on main

[flat leaf]

sounds like a fun time to deal with race conditions and making the code look horrible because you now have to synchronize every thread when passing data between threads

multithreading also has diminishing returns to a point, CPUs only have so many cores/threads

[noble lantern]

well you would just do processing async

and then call back to main thread

[noble lantern]

thread pools

been playing with that lately

rly fun

[flat leaf]

the issue is that when running something on another thread there is zero guarantee as to what the main thread is doing

so to exchange any sort of data you have to basically pause both threads at a point that is appropriate and exchange information like a hand shake

[buoyant viper]

when the ticker is 5 secodns ahead of the rendering thread

😎

[noble lantern]

public static int playerBalance;

CompletableFuture.runAsync(() -> playerBalance++);

ez

[flat leaf]

now access that from two threads at once

[noble lantern]

redis could possible work maybe as a middle man

[flat leaf]

I don't even think it's necessarily guaranteed what actually happens when you just access an int that is being written to currently

[buoyant viper]

TIL the search bar in the creative inventory in 1.8 only fits 15 chars, so i cant search for items such as "heavy_weighted_pressure_plate" :(

[noble lantern]

latest running thread typically sets the value

last*

[flat leaf]

in java you use an AtomicInt

in C# you use a lock

or C# may have types like the Atomic types, not sure

[buoyant viper]

in java u have the oh so wonderful synchronized keyword

(i have never used that in my life)

[noble lantern]

i have

its pretty useful

Object#wait

and Object#notify make use of it

you can make makeshift callback functions with all those methods

useful for multithreading too but i stoped using it

too annoying

but basically synchronized (object) just locks the object to the main thread when calling and allows you to execute things on main inside the block

[flat leaf]

ah so synchronized is basically like the lock keyword in C#

[noble lantern]

yep!

in fact

the argument name in javadocs is literally named lock

haha

[flat leaf]

and can be used pretty much exactly like it

[buoyant viper]

yes

[flat leaf]

makes sense

[noble lantern]

Would be useful if you handle data management on main thread

[flat leaf]

I am a little more familiar with C#'s threading systems over java's as usually I don't have to do much fucky shit in minecraft plugins

[noble lantern]

i actually need to make a async storage access system and thats gonna suck ngl

[buoyant viper]

for as different as Java and C# are, theyre basically the same language when it comes to writing code

[flat leaf]

yeah they're pretty similar

[noble lantern]

im debating on not caching from db and always pulling from mongodb

or redis

[buoyant viper]

except ones PascalCase and ones camelCase

[flat leaf]

I am far far more familiar with the internal workings of C# over java though

as I have just made more projects with it

like audio shit pretty much guarantees you will be doing some threading

but audio programming is fun

[buoyant viper]

usually need to thread audio in java too iirc

[flat leaf]

well yeah you need to thread audio in every language, you simply have to feed to much data to do audio along with everything else

[buoyant viper]

true

lol

i realized that after i sent it

[flat leaf]

audio is fucking crazy

like imagine if we had to generated 44.1 thousand frames a second for a game

obviously there is far more calculation to be done in a game but still

[buoyant viper]

a small price to pay for high FPS gameplay

[flat leaf]

lmfao

I feel like we couldn't even see a refresh rate past like 5000hz

[noble lantern]

omg wtf

https://www.javamex.com/tutorials/synchronization_volatile.shtml

bro

[flat leaf]

or like 1000hz

[noble lantern]

ez

thats so easy to share between wtf

[buoyant viper]

cant even count++

smh

[noble lantern]

smh

imagine

useless as shit

now i gotta make a whole different workaround for my multithreaded for loop counters

Wait so could you use volatile to say save player data accross threads?

https://docs.oracle.com/javase/6/docs/api/java/util/concurrent/BlockingQueue.html

Actually this seems more viable

[ivory sleet]

Yo

What are ya doing

[noble lantern]

Oh thats right i pinged you lmao

[ivory sleet]

Ye

[noble lantern]

i think the guys gone now

[ivory sleet]

I was sleeping then

A nice

[noble lantern]

but whats the name of the protocol thingy minecraft uses?

Like the actual data type transferred

i remember you telling me one time

[ivory sleet]

TCP?

[noble lantern]

Not TCP, i was talking about how i handle my socket data in JSON, and you mentioned something else that MC uses

someone else was doing the same thing

[ivory sleet]

NBT and SNBT?

Oh wait

[shadow zinc]

anyone got an idea how to remap gradle

[noble lantern]

I feel like it was named something else other than those hmm

it was a weird word

[ivory sleet]

I mean they just write their data in raw bytes if thats what you mean

[noble lantern]

Been asking myself for a while

[noble lantern]

Isnt there a special name for it?

[shadow zinc]

all I want is to power a piston ffs

but no

[zenith gate]

My recipe is being shown in the recipe book, But I cant actually craft the item, the result shows the correct as well as all of the items, what could be causing something like this? I get no errors either.

[shadow zinc]

it just wont happen

[ivory sleet]

Hmm no not really as far as I know

I mean you have their abstraction on top of netty

With a buncch of utility methods for encoding and decoding data types to raw bytes

[ivory sleet]

I know paper has a remapper plugin

But that’d add the paper api to your classpath

Which you might not want

[shadow zinc]

why not?

[ivory sleet]

Because paper isnt spigot

[shadow zinc]

so?

[noble lantern]

read through all your messages of the past 3 days and couldnt find it weird

[shadow zinc]

does it break anything?

[noble lantern]

also conclure i learned the hard ware not calling join() with runAsync + work stealing

[noble lantern]

Yes

If you use paper methods on a spigot server youll get errors

[shadow zinc]

why?

bruh

okay

so

im stuffed?

[ivory sleet]

Idk if you meant this @noble lantern #help-development message

[noble lantern]

nah it was like one word

it was a rly weird word too

maybe someone else said it hmm

[noble lantern]

Uhmm

you could use spigot remapping website to display version specific method names for mojang

Theres a website that does it

[ivory sleet]

How are you saving?

[lost matrix]

Baboonery

[shadow zinc]

can I get the link to that?

[ivory sleet]

screamingscandals

or sth

[lost matrix]

https://nms.screamingsandals.org/

[ivory sleet]

ah sandals

[noble lantern]

Im gonna have a class on main thread that has just a basic data handler class with calls to MongoDB

As well as a hashmap ofc to cash PlayerData POJO classes

I think the queue thing likely would work better for me maybe

[noble lantern]

such a fancy name lmao

[lost matrix]

screamingscandals would sound a bit dubious

[noble lantern]

i was even thinking in my head i bet smile knows the site

[ivory sleet]

But anyway burchard you only want to yse volatile when absolutely necessary since any volatile write ensures a flush of local registries

[noble lantern]

so definatly dont use it in player data objects

got it

[ivory sleet]

Well

[lost matrix]

volatile prevents caching on the cpu layer caches.

[ivory sleet]

^

[noble lantern]

Hold one let me write out my data handler rq and come back with my question

gimme like 5 mins

[lost matrix]

It also does not ensure atomicity

[ivory sleet]

You could if you wanna be very optimized use a varhandle or atomic field updater

[noble lantern]

just gonna psueodo code

[lost matrix]

let me double check on that...

[ivory sleet]

It has methods to do so

But on its own it doesnt

Well

What the fuck am I saying

I mean AtomicReference and AtomicFieldUpdater for instance

Not volatile itself

But volatile is a crucial part if you want to actually perform atomicity correctly

Or well memory fences as well with var handles work just as good

[lost matrix]

Ok so it doesnt. It just makes sure that all threads have the same representation and nobody uses a locally cached version

[ivory sleet]

But you need the memory ordering effect guarantee to avoid indeterminacy

[noble lantern]

okay im back

?paste

[undone axle]

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

[noble lantern]

This class would be on the main thread, except for the Optional<PlayerData> which will be async get from MongoDB: https://paste.md-5.net/gozuwitela.cs

Don't roast me its just pseudo code lmao

And the POJO player data class would be like this: https://paste.md-5.net/akaxaligut.cpp

And when a processing thread wants player data they would get the instance of PlayerDataFactory and call get(uuid) (Like said pseudo code don't roast me for the current design pls)

my current method of handling it would just be calling synchroniced on the PlayerDataFactory instance and handling data get/insert there

[ivory sleet]

Wait

[shadow zinc]

ffs should I just use maven?

[noble lantern]

but for some reason it feels wrong

[shadow zinc]

I want to die ngl

[noble lantern]

yes

[shadow zinc]

gradle aint worth it

[noble lantern]

unless smile knows how to do special source for gradle

[ivory sleet]

PlayerDataFactory is used across multiple threads for both reads and writes?

[noble lantern]

PlayerDataFactory is intialized on the main thread

[shadow zinc]

hmm

[noble lantern]

and runs on main

[shadow zinc]

I should wait then

Until smile gets on

[ivory sleet]

Yes but?

[noble lantern]

but mongodb requests (First data load) are on a work stealing thread

[ivory sleet]

Is it used over more threads than that?

[noble lantern]

Well it would be used by threads from the work stealing pool

The threads would get a static instance (Or DI instance) of that class and call get()

sec

[ivory sleet]

Cause usually

You have some sort of repository class

Which is basically a map, sometimes responsible for instantiating the underlying objects as well

So factory + map

And that class might often very well be used across multiple threads

[shadow zinc]

Could you send him a message so when he is available he can talk to me?

if thats alright, I don't want to ping anyone

[noble lantern]

It would get used like this in a sense

my code is more abstracted than that

runAsync is a CompletableFuture wrapper

[ivory sleet]

Okay

But do you ever write to the PlayerDataFactory

From other threads

[noble lantern]

only change variables inside PlayerData value retrieved from get, so i think yes?

No actual mongodb writing

just cache write

like playerMoney += 500; or some shit like that

It writes to mongodb every 5 mins on a diff future for cached PlayerData in the map

and onStop

[ivory sleet]

So how do you put the player DTOs in the map in the first place?

[noble lantern]

When the player logs in, validated with oauth2, and sends theyre JWT with UUID it will be loaded on first login

ie onConnect

[ivory sleet]

Isnt that on another thread?

[noble lantern]

Indeed it would be

so i would call loadData(uuid) onOpen

[ivory sleet]

But yeah I dont understand why you need that blocking queue as well?

[crude heron]

Hey does anyone know how to add effects on Monsters?

[noble lantern]

well i assumed it was needed for data to be safe accross threads

[ivory sleet]

No

Let me try to explain

In Java

Theres a memory model

[noble lantern]

like what if on thread does playerData.money += 50

i dont think it gets updated without the volatile keyword

[ivory sleet]

Normally we have whats called plain read and write

[noble lantern]

kk

[noble lantern]

What type of effects

[ivory sleet]

Which is normals reads and writes to variables

But when we do multi threaded stuff we need certain guarantees

[crude heron]

speed and resistance type

[noble lantern]

Look into Attribute

https://hub.spigotmc.org/javadocs/spigot/org/bukkit/entity/Zombie.html

Certain mobs like this one implement Attributeable

https://hub.spigotmc.org/javadocs/spigot/org/bukkit/attribute/Attributable.html#getAttribute(org.bukkit.attribute.Attribute)

From there you can call say zombie.getAttribute(Attribute.GENERIC_ARMOR);

[ivory sleet]

//one thread
var i = 3;
var j = 1;

//other thread
var a = i;
var b = j;

what are the possible values after both threads have executed?

[noble lantern]

And use these methods to edit it: https://hub.spigotmc.org/javadocs/spigot/org/bukkit/attribute/AttributeInstance.html

[ivory sleet]

Assuming they run concurrently

[shadow zinc]

Are there any generators for maven?

like minecrell?

[noble lantern]

hmmm

thread 2 would generate an exception wouldnt it no

unless it runs after thread 1?

[shadow zinc]

Can I use maven and gradle simultaneously?

[noble lantern]

The minecraft intellij plugin does it

[frosty tinsel]

Yes

I think so

[crude heron]

Ok thanks

[shadow zinc]

ooo

[ivory sleet]

burchard assume that it would become 0

Now guess x)

Hint there are 4 outcomes

[noble lantern]

wtf

im having a hard time thinking of one LMAO

if theyre both ran at the same time at least

[ivory sleet]

a = 0
b = 0

a = 3
b = 0

a = 0
b = 1

a = 3
b = 1

All four outcomes are possible

[noble lantern]

so basically no consistency between threads

[ivory sleet]

Yes

Because

var i = 3
var j = 1

can be re ordered to

var j = 1
var i = 3

And same guess for a and b

That wont alter the behaviour of the program as you can see (assuming its a single thread)

[noble lantern]

Yeah but it would lead to values you dont really want in a sense

[ivory sleet]

Yep

[noble lantern]

Like i dont want one thread spending a players money, while another thread also does it

[shadow zinc]

so build is gradle, package is maven?

[ivory sleet]

We are coming to that part soon burchard

[noble lantern]

okay okay x)

[ivory sleet]

So basically we use volatile as by the means of preventing this type of reordering

[noble lantern]

Honestly have no idea how youd run the 2 in unison

Like you can have the 2 in your IDE but i dont think theylle work in unison like your expecting them to

If you want remapped maven is likely easiest way to go about it

[shadow zinc]

so I need to create a plugin.yml now?

[noble lantern]

no just a pom.xml

[shadow zinc]

but there are no generators?

[noble lantern]

convert your gradle to pom (Striaghtforward process) and then add special source

Its pretty easy

gradle imports are

groupId:artifact:version

[noble lantern]

dw im stilll listening

just multitasking

[ivory sleet]

when we prevent re ordering we get somewhat of a determinacy we can safely rely on

[shadow zinc]

how tf do you convert gradle to pom?

[ivory sleet]

This is crucial when you want to read and write concurrently

Now assume we do have some ordering constraints

Then we can apply whats called atomicity

[noble lantern]

Manually I can give you an empty pom tho

[shadow zinc]

i got my pom

its from another project of mine

[noble lantern]

Yeah i saw a lot about AtomicIntegers and stuff

[ivory sleet]

Yes

[noble lantern]

DIdnt fully understand them though

[noble lantern]

So youd just take your gradle dependencies and format them to mavens xml format

[ivory sleet]

So you know how incrementing a variable is multiple operations right?

[shadow zinc]

yeah I just need to figure out how to shadow my kyori stuff

[noble lantern]

IE how it pushes the values to heap right, in that case yes

[ivory sleet]

like you gotta load the variable, perform the incrementation and then put the result back into the variable

[noble lantern]

yep

[ivory sleet]

necessarily just the thread local staxk

[noble lantern]

use maven shade plugin

[ivory sleet]

Stack *

But ye

[noble lantern]

here ill get you it

[ivory sleet]

Anywa

[noble lantern]

?paste

[undone axle]

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

[noble lantern]

https://paste.md-5.net/pezasurewu.xml

use scope compile for shading

your properties tags needs a java.version variable

[ivory sleet]

When variables can guarantee visibility we can perform a compare and swap when we want to perform operations that base themselves on old values

basically when we have performed the incrementation you check whether or not the old value was changed

[noble lantern]

Ah so thats what atomics do?

Just constantly checking for new values?

[ivory sleet]

Yes in principle

If the value was changed

Then we perform the incrementation again

And so on

If it wasnt

Then put the result into the variable

For this to work across multiple threads we need volatility

Since we want up to date and accurate reads and writes

AtomicInteger, AtomicReference provide methods that allow you to do this

updateAndGet, getAndUpdate

for AtomicInteger you have incrementAndGet, getAndIncrement

etc

So you dont havr to implement that shit

And for maps

You have ConcurrentHashMap

[noble lantern]

I dont think ill be doing any Incrementing on values

mainly division, multiplution, some addtion but not incrementals and substractions

[ivory sleet]

Yeah

[noble lantern]

Always wondered wtf that was for

always thought it was for something else

[ivory sleet]

Lol well it performs atomicity on a map

[noble lantern]

Would the atomicity also carry over to the objects inside said map?

[ivory sleet]

assuming you use for instance putIfAbsent, computeIfAbsent, compute etc

[noble lantern]

Yeah i always use putIfAbsent on my regular maps

[ivory sleet]

Ugh wym

[noble lantern]

lol sorry xD

[ivory sleet]

it guarantees atomicity on the map, not the variables of the values in the map

[noble lantern]

yep thats what i was asking :p

[ivory sleet]

That you need to fix yourself Im afraid of

Ye

[noble lantern]

oh fr

[ivory sleet]

But in simple terms, you go woth AtomicReference etc

and provide some convenience methods with your DTO maybe

void add(int delta){
moneyReference.getAndUpdate(old -> old+delta)
}

[noble lantern]

would create a wrapper method called setAndUpdate around a ConcurrentHashMap work?

Ie where you get object from hashmap, edit that object and then update/re-insert it into the map?

I see under the hood, this is similar to what things like AtomicInteger does

[noble lantern]

ohh

[ivory sleet]

yeah

[noble lantern]

So my variables inside player data w-

oh that way would work?

[ivory sleet]

Yes

You can even setup a unit test to prove that it works

[noble lantern]

and i wouldnt need any transforms to my actual pojo class? I can still just use raw BigInteger?

and other keywords instead of atomics

[ivory sleet]

You’d have to use AtomicReference then

But in principle same story

[noble lantern]

I wonder how gson parses those 🤔

if possible

[ivory sleet]

For the map

You probably dont need to do anything special

Like

If you just putIfAbsent or use put

[noble lantern]

i see, let me write something out rq

i think i understand

[ivory sleet]

To populate DTOs that are yet not contained

But remember

You only need atomicity when multiple threads reads and writes to a variable

[noble lantern]

Yeah theyres gonna be lots

Every request is handled async

and players can send a bunch of diffeerent requests that would alter theyre player data significantly

and some things take heavy math calculations

so sometimes a request could hang for a few seconds

[ivory sleet]

What I recommend you to do is sth like

[noble lantern]

and if a player sends a request like spend money before that one finishes it would cause an issue

That was mainly my main concern and main thought worry when kaing the async socket system

[ivory sleet]

onSomeThread(()->{
result = heavyCalculation() //if you are not dependent on old values
dto.add(result);
}

rather than dto.update(old -> heavyCalculation()+old);

But the latter might be necessary if u r dependent on old values

But yeah

You also need to add some checks

Then

To make sure the money cant go negative when performing an update

[noble lantern]

i would prefer to depend on the most latest updated value of that number

[noble lantern]

Yeah ill have those checks in my core logic

sec stil writing this out

[ivory sleet]

No burchard

Sometimes you’re not dependent

Like lets say all you do is adding a number

Then you dont care

But if you wanna square the old value and take the second derivative of the old value or sth weird then you need to perform an explicit update

Anywah theres much more to it if you wanna get real, but this should be enough for you to create a model that is scalable enough

Id tho… recommend unit testing

Especially the multi threaded stuff

[shadow zinc]

Can I get some help shadowing minimessages?

[ivory sleet]

To make sure you got it right

[shadow zinc]

their docs don't supply any examples

[granite burrow]

hey I got a question is there a way to show the crafting recipe in a GUI without it being clickable? I tried to disable the clicks that a player does when they are in the GUI I made, however all this does it stop them from clicking in the crafting inventory I made.

They are still able to take out any items that are in the crafting gird by double clicking with a similar item, and can add to that item by right clicking the item onto the similar item in the GUI

[noble lantern]

think im mostly done

[ivory sleet]

(:

[noble lantern]

https://paste.md-5.net/ukexepuxej.java - new player data handler my socket library handles those methods async, pretend that class isnt nested inside this class i was just lazy to make a new file

POJO Data class: https://paste.md-5.net/berepumemo.cpp

the subclass there would just be shown how its used

ik im prolly doing a lot of uncesary gets im not using

but its just psuedo code mainly

ill abstractify it once i get it work

Ill have to learn how to unit test lmfao

[ivory sleet]

Well ye

Uh

[noble lantern]

Got any code?

[ivory sleet]

Use updateAndGet

just using get and then set isnt exactly the same

[noble lantern]

oh its a functional interface for that method

hell yeah

[ivory sleet]

Ye

Same goes for the lap

Map

[granite burrow]

?paste

[undone axle]

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

[noble lantern]

I dont need ConcurrentHashMap tho do i

since im pretty much extensively checking?

[granite burrow]

sorry bout that, forgot to turn off the reply

[noble lantern]

or use it just to be safe

all good

[ivory sleet]

well

You could just use map.put(uuid,data)

[noble lantern]

oh yeah

forgot that overwrites

[ivory sleet]

In this case 🙂

[granite burrow]

Its just a basic check if the inventory is correct and if it is cancel the event.
https://paste.md-5.net/kuwojijixi.java

How I'm going to fix it though is just add a attribute to the item stack to make it different from any other possible item.

[ivory sleet]

But overall atomicity is very scenario specific

[noble lantern]

now the main issue is just seeing how these parse against gson

Idk this does seems pretty good and looks like itll work

it will be easy to test once its stup

[ivory sleet]

I have written a unit test that showcases how you might wanna test it

https://github.com/Conclure/Clonomy/blob/master/module-shared/src/test/java/DataTransferBaseTest.java

[noble lantern]

So you only want the actual 4 or 9 crafting slots to be non clickable?

[granite burrow]

the 9 crafting slots and the outcome to be not clickable, or editable in any way.

[noble lantern]

oh wow

that lots complex asf, ill definatly look into it haha

idk why unit tests always scary to me

[noble lantern]

I forgot theyre exact slot counts but

[ivory sleet]

Yeah

Gonna fetch some breakfast cya

[granite burrow]

its 0 - 9 with 0 being the outcome for the crafting table

[noble lantern]

so can you just if clickSlot == those cancel?

[noble lantern]

Cya! And thanks a bunch for explaining so well for me

prolly would of never figured it out and i prolly would of went to like redis or something 💀

[granite burrow]

the code I sent works for stopping the items from getting left clicked or right clicked out, however, if you shift right click a item into the crafting gird it will remove the item and if you double click the item it will give you all of that item.

[noble lantern]

for shift clicking and canceling destination slot someone asked that earlier

youll need packets to get destination slot of shift clicks sadly, and then cancel upon those

As for double clicking

thats a thing?

i know shift double click is

[granite burrow]

yeah just clicking fast inside of a gui should stack all similar items up to their max stacking point

[noble lantern]

hmm

i wonder if an event for that exists

oh also you wanna cancel inventory drag event btw

cause i bet dragging breaks your gui too

[granite burrow]

I solved that issue by just adding in extra attributes to the items that are in the fake crafting menu I make. but the shift clicking items from your inventory will still make them disappear :/

[granite burrow]

suprisingly no, just the shift clicking item removal, and the items being able to be pull out with double clicking

[noble lantern]

Odd question

but after you cancel double clicks

do you cann player#updateInventory?

call****

[granite burrow]

I do not, do I need to?

[noble lantern]

"do you can" lmao

sometimes you might need to

bukkit api sucks for that

try and see if it fixes it

lmk if it doesnt

[granite burrow]

alright let me remove my fix and test that out, that may fix the shfit clicking too, who knows 😄

[noble lantern]

shift clicking sadly needs the weird work around

unless your shifting clicking directly on the disallowed slot

idk why theres no getTargetSlot method

[granite burrow]

didn't seem to do anything

[noble lantern]

one more thing

Bukkit.getSchedular().runTask(() -> p.updateInventroy());

sorry

forgot to mention run it one tick later

[granite burrow]

the work around only works for double clicking not shift clicking :/ unless your talking about another weird work around and not adding custom data to the item in the gird to make no items stack to it

[noble lantern]

well shift clicking and getting destination slot needs NMS/Packets

sec lemme find the article

https://www.spigotmc.org/threads/shift-click-item-slot-destination.428224/#post-3748300

hot link ^

[granite burrow]

running it later doesn't seem to do anything about it either, guess im stuck with the added data. ill look into the shfit clicking thing hope to fix that.

[noble lantern]

sadness, was worth a shot

[granite burrow]

Its weird that when shift clicking the item just disappears though.

[noble lantern]

sometimes that method can fix some weird behavious with inventories

Its likely just because its a custom inventory

so the functionalities fucked a little

[granite burrow]

ah yeah prob

[granite burrow]

I found a non NMS work around, in my inventory click event I have this code:

        if (e.getClickedInventory().getType() == InventoryType.PLAYER) {
            if (e.getAction() == InventoryAction.MOVE_TO_OTHER_INVENTORY) {
                e.setCancelled(true);
            }
            return;
        }

basically what it does it check if the player shift clicks an item and if they do, it stops them, if they don't then they don't care about that inventory. However, I do wish that spigot had a InventoryAction.Double_Click since it doesn't seem to exist anywhere

EDIT:
🤦‍♂️ I think its called "COLLECT_TO_CURSOR"

[shadow zinc]

I need help, why am I getting this for maven ```[ERROR] Failed to parse plugin descriptor for net.kyori:adventure-api:4.11.0 (/home/user
/.m2/repository/net/kyori/adventure-api/4.11.0/adventure-api-4.11.0.jar): No plugin descriptor found at META-INF/maven/plugin.xml -> [Help 1]

this is my shadow part

                <groupId>net.kyori</groupId>
                <artifactId>adventure-api</artifactId>
                <version>4.11.0</version>
                <executions>
                    <execution>
                        <phase>package</phase>
                        <goals>
                            <goal>shade</goal>
                        </goals>
                    </execution>
                </executions>
            </plugin>```

[noble lantern]

why is it wrapped in plugin tags

isnt kyori a dependency

[lost matrix]

Since when is there an adventure maven plugin??

[shadow zinc]

oh fuck

look

maven

it aint it for me

[lost matrix]

I see

[shadow zinc]

oh

and

actually

special sauce

for gradle

you know a way?

[lost matrix]

 <dependency>
    <groupId>net.kyori</groupId>
    <artifactId>adventure-api</artifactId>
    <version>4.11.0</version>
 </dependency>

[lost matrix]

Nope thats a maven plugin. When using gradle ill always go for the userdev plugin from paper.

[shadow zinc]

I was using it

but it didnt remap everything

[lost matrix]

It does for everyone else... I'd bet on user error.

[shadow zinc]

yeah maybe, but fixing isnt really an option now

[granite burrow]

So that fixes the two issues with, shift clicking and double clicking.

However, I also ran into an issue where you can right click the item into the slot if you right click the slot enough times it will bypass the event being cancelled

Video of the event running even though its clearly cancelled

[shadow zinc]

apparently maven is better

so

[lost matrix]

Both maven and gradle have their caveats. They are both equally painful to use.

[shadow zinc]

well its going to be more of pain for me if I use gradle tbh

can I do this?

[lost matrix]

For recipe editing i always use a dropper or hopper inventory.
Ill let the user do whatever they want and simply read the matrix when the inventory closes.

[granite burrow]

its not supposed to be editable, its just another way for people to view the recipe other than in the recipe book. Because sometimes it glitches out

[lost matrix]

Then simply cancel the InventoryClickEvent when they have a view open.
All of them

[shadow zinc]

its got gradle all over it

its not happy that I deleted the build.kts

[granite burrow]

Thats what I have right now https://paste.md-5.net/eqepawotul.cs

[lost matrix]

Dont use the title as an identity...

[noble lantern]

oh oops

i totally missed that when looking over that

[granite burrow]

Is it not good practice? whats a better way to determine its the right view?

[lost matrix]

This can be done in way fewer lines. All you should need:

  @EventHandler
  public void onClick(InventoryClickEvent event) {
    if(viewManager.isRecipeView(event.getInventory())) {
      event.setCancelled(true);
    }
  }

[lost matrix]

Use a Set<Inventory>

Create a RecipeViewManager that contains a Set<Inventory>
When you open a view for a player you simply add it to the manager.
Then on the close event you remove it from the manager.

RecipeViewManager is a singleton like every other manager class

[lost matrix]

Write if you want an example

[ashen quest]

https://youtu.be/AVVQAU3lWZo?t=341
How can I code this thing (@5:41)

It makes it look like you are about to fall into the void but its just a trap

[shadow zinc]

come on, stop it

[ashen quest]

why

[lost matrix]

Thats just a bug

[ashen quest]

Ooh

well is it usable now

also is it from the API or do I need nms?

[lost matrix]

What exactly are you referring to?

[granite burrow]

no, I got that to work. However, it still bypasses the cancel event if you right click fast enough

[ashen quest]

Making it so player thinks he is about to fall into the void but its just a hole in the ground

[lost matrix]

Doesnt happen. What version are you on?

[ashen quest]

the way spigot fakes its ores for anti xray

[granite burrow]

1.18.1

[lost matrix]

Inventories are fully server authoritative. If you properly cancel the event then nothing can happen.

[lost matrix]

Send Chunk packets

[granite burrow]

I wonder why its happening then, cuz I can check if the event is cancelled after but the item still appears

[ashen quest]

can you elaborate

I need nms right

[lost matrix]

Neve had that happen after 1.12
Its a fake ItemStack or your code is not solid.

[lost matrix]

ProtocolLib will do

[shadow zinc]

Burchard you lied to me

Maven doesnt generate plugin.ymls

why would it?

sfsdfdsgdrf

[ashen quest]

so I send a chunk packet, thats all?

[noble lantern]

i never said maven did

i said the intellij plugin did

[shadow zinc]

you insinuated it

regardless

[noble lantern]

i didnt?

even then

why is plugin.yml a big deal

that shouldnt change when switching to mappings

[shadow zinc]

I never had a plugin.yml

it was generated by minecrell

do I need to make it manually now?

[noble lantern]

never heard of minecrell

yes all plugins need a plugin.yml

[shadow zinc]

.......

[noble lantern]

this is why we dont use libraries like that

[shadow zinc]

are there any generators?

[noble lantern]

That way we know how to make a barebone plugin

[shadow zinc]

I've done it before for my neosurvival

[noble lantern]

https://www.spigotmc.org/wiki/plugin-yml/

You can view how to make one here

it goes in /resources

[shadow zinc]

I feel like a teenage girl that just got cut off her allowance

[noble lantern]

(Considering your projects setup like that)

lmao

[ashen quest]

yes, theres a intelliJ plugin that will set you up with a basic plugin structure

[noble lantern]

Hes meaning just plugin.yml generator lol

[ashen quest]

oh

[noble lantern]

I suggested that plugin to him earlier

[granite burrow]

your right, im stupid, my code had old testing code from when I was trying to fix the other issues I had.

if (e.getCursor() == null) {
            return;
        }
        if (e.getCurrentItem() == null) {
            return;
        }

this broke it which makes sense, Its all fixed now

[ashen quest]

why do u need a gen for that

[shadow zinc]

because

[noble lantern]

i love when debugging code manages to keep itself in the codebase

happens alot when i go to bed after debugging a bunch of shiz and i start up the next day forgetting any leftover debug shiz

[shadow zinc]

I'm going to decompile my old jar and grab the plugin.yml from that

ik

ik

[noble lantern]

there you go

big brain

[shadow zinc]

ikr

so smart

[granite burrow]

worst feeling, I have spent countless hours doing that

[lost matrix]

I once wrote something for an answer here into one of my resources and pushed it. Only noticed it a month later.

[noble lantern]

oh god

i do that a lot

ill write out some code in a random project to show to someone

and i forget its there lmfao

in fact there code in one of my projects doing that rn

forgot which one tho

enjoy smiles

[lost matrix]

[shadow zinc]

He actually does a good job at explaining tbh

[noble lantern]

indeed but mans comes out with some shit i could only dream of sometimes

[shadow zinc]

Lol

looks like me too

anyways

why cant I find my plugin.yml in the decompiled classes?

[noble lantern]

should be right when you unzip it

[noble lantern]

if the plugin was built correctly i will look like this

[shadow zinc]

just a bad decompiler?

[noble lantern]

just use winrar

or any unzipping aplication

jar files are zip files

[shadow zinc]

bruh

you are a wizard

[noble lantern]

im literally the dumbest person in this channel, but thank you

[shadow zinc]

-100 self confidence

[shadow zinc]

How do i do this again version: {{ neoPerformance.version }}

for version:

again, I've always had my allowance, the real world is a bitch

[noble lantern]

wth is neoPerformance version

can i get some context

[shadow zinc]

I remember you could put some tag in the plugin.yml to grab the version

neoperformance is just the name of my crappy plugin

[noble lantern]

ahh

uhm

${version}

but thats maven

nvm u on maven

make sure you use the plugins i sent earlier

or that placeholder wont work iirc

[shadow zinc]

is that minecraft development?

[noble lantern]

no its called filtering

just as long as thats true for your resources dir it should parse fine

which it should be cause i sent you that entire chunk earlier

[shadow zinc]

ah

yep

thanks

[noble lantern]

just - the binary-resources dir bc that for my own plugin

[shadow zinc]

It's saying my plugin.yml is invalid

[noble lantern]

send the plugin.yml you have

[shadow zinc]

name: NeoPerformance
version: '${version}'
main: com.neomechanical.neoperformance.NeoPerformance
load: STARTUP
authors:
  - NeoDevs
prefix: NeoPerformance
commands:
  neoperformance:
    description: Show server performance
    aliases:
      - np
      - performance
    permission: neoperformance.admin
    usage: /neoperformance
api-version: 1.13
default-permission: op```

[noble lantern]

When you unzip the jar

where is your plugin.yml

[shadow zinc]

root

[onyx shale]

have a feeling this isnt ur project right?