dev-general

[pastel imp]

I assume using display entities

[desert horizon]

Nope

[abstract osprey]

~~ it uses Nova~~

[desert horizon]

Bossbar and fonts

[desert horizon]

Also no

[abstract osprey]

what

wtf are you in the dc for then

😭

is it just a similar impl?

[desert horizon]

Yeah. It's my own implementation of a WAILA.

[abstract osprey]

ahh fair enough

[desert horizon]

Tho, I did use the textures of Nova here

But I had to edit the block icons... all 991,
Because of how MC creates fonts.

Since it only makes them as wide as the furthest non-transparent pixel of an image (from left to right) did I had to add a semi-transparent background to all the images for a consistent size.

[abstract osprey]

oh man

[desert horizon]

Good thing there was a batch option in Gimp

[ocean quartz]

What is Nova?

[pastel imp]

a plugin

[ocean quartz]

thanks..

[pastel imp]

https://modrinth.com/plugin/nova-framework

there ya go

[ocean quartz]

I see interesting

[desert horizon]

It's quite cool

Its own WAILA is really cool (Especially since it also supports textures of blocks from like ItemsAdder) and was the inspiration for doing this myself.

[desert horizon]

Haha.... Just found out some characters force a Right-to-left direction and it resulted in this xD

[abstract osprey]

no bedrock support though :(

that’s what I don’t like about nova

It makes sense though

Bedrock support would be hell

[desert horizon]

I personally don't like plugins trying to support BE if they are made for JE

[sweet cipher]

The class I am taking now used Dafny for the beginning part of the semester

The loop invariants are pretty much just look at the post condition

[prisma wave]

fireeee

[prisma wave]

sometimes, we had some more complex ones though which needed a bit of wrangling

[desert horizon]

Little teleport thing

[desert horizon]

To whoever in the past talked about my plugin sending too much text: At least it isn't Iris....

[pastel imp]

god that must be in debug mode

[desert horizon]

I hope so too

[desert horizon]

I'm really that lazy that I made this:

public class FunctionKeyMap<K, K2, V>{
    
    private final Map<K2, V> values = new HashMap<>();
    private final Function<K, K2> function;
    
    public FunctionKeyMap(Function<K, K2> function){
        this.function = function;
    }
    
    public void put(K key, V value){
        values.put(function.apply(key), value);
    }
    
    public V get(K key){
        return values.get(function.apply(key));
    }
    
    public Map<K2, V> map(){
        return values;
    }
}

So that I can do this:

FunctionKeyMap<Material, String, String> texturePaths = new FunctionKeyMap<>(material -> material.key().asString());

...and have a map with String, String from a Material and String...

(Probs could be made better too... idk)

[prisma wave]

why not just EnumMap?

[desert horizon]

Because I need a String key as I use that for identfication and don't wanted to do the Material#key().asString() 20+ times

[prisma wave]

for identification
bit vague but okay

[desert horizon]

I get the block's namespace:id key later and check for it in the map.

Tho, my current system is a mess and I need to find a better setup...

[desert horizon]

Spent some time making a section on my website for my ecloud expansions...
Kinda shows how the Formatter expansion is the most complex/large one.
https://andre601.ch/expansions/

[wintry plinth]

Would be cool to show the download count if there’s an api or something for it

[desert horizon]

I noticed an interesting behaviour with fonts that have been shifted using negative space widths

If the font has its shadow disabled is ir rendered behind content following

Like f.e. I disabled the shadow for the block icon and here it's gone.

But when I re-enable the shadow is it displaying properly

Really interesting behaviour

[desert horizon]

The ecloud API doesn't expose such a thing.

(Also didn't know that there was a v3 lmao)

[wintry plinth]

That’s such a shame..

[desert horizon]

We could probs suggest it. Tho unless it's tracked internally already would the stats probs not be accurate

[desert horizon]

Published the source and wouldn't mind feedback: https://codeberg.org/Andre601/IAWAILA

Command system will use cloud in the future, so I'll change it.

[wind patio]

why no spaces between () and {}

[ocean quartz]

White space price is too high nowadays gotta save up a bit

MD5's parenthesis style is hoarding all the spaces

[desert horizon]

Just my personal style

[wind patio]

i would suggest some improvements, but i cant look at it due to missing whitespaces

😔 🫰

[desert horizon]

mood

[distant sun]

It is tracked

https://api.extendedclip.com/usage/

And ig the data comes from here https://bstats.org/plugin/bukkit/PlaceholderAPI/438#expansions_used

[desert horizon]

wohoo! My expansion is on the first page

right above itemsadder (img is from itemsadder)

[desert horizon]

There we go... added some stats to the expansion pages (fetches from bstats)

Need to push it now

[wintry plinth]

Looks dope

[desert horizon]

Of course when I want to update the site are the actions runners causing issues... Have to wait I guess

[desert horizon]

Okay, it's now live

[desert horizon]

Since I started using the ItemsAdder API have I found various issues... Really hope LoneDev will fix them.

[pastel imp]

nexo on top

[desert horizon]

Meh...

I don't bother because A) Don't want to spend money and B) don't want to support something that is part of some stupid drama

[pastel imp]

calling it a stupid drama is plain stupid

lol

but oh welp

keep hoping for your fix then

[desert horizon]

I mean, I am in direct contact with Lone...

And he's always open to improve stuff

[pastel imp]

god claude sucks with svelte 5 xD

brilliant with 4, but 5.... gosh xD

[desert horizon]

Ah yes, Thanks IntelliJ...

Not sure what I can or should do here honestly...
Want to try out a library, but it only has a single Event class with a method. All other ones only extend it... And I don't want to do countless instanceof checks in a row here...

[brazen ether]

generics are a pain

[potent nest]

Basically, each instance of ? is a new one, incompatible with all other ones

That‘s not IntelliJ‘s fault

[brazen ether]

i think he meant intellij doesn't help handle the error

[desert horizon]

Yeah, and the "expected capture of ? but got capture of ?" doesn't help much there

[potent nest]

Yeah you have to understand Java for that I guess

[wind patio]

any recommended DI libs? Guice, dagger, spring?

[crude cloud]

yes

[wind patio]

😔 🫰

ima juice it

[crude cloud]

idk they all kinda just work, requiring scaffold and setup of different kinds to varying degrees

it also depends on how complex your project it, i don't often find myself needing a DI framework or in a situation where it would make things easier or something, but i like dagger personally; if i were to be doing something with spring i'd use spring di ofc, or if i'm making a velocity plugin i'll be using guice since it's built-in, but eh

just try whichever and see what you like i guess

[wind patio]

https://github.com/hakan-krgn/spigot-injection
this seemed interesting

[crude cloud]

omg it's john tesla

[wind patio]

currently trying out guice

[crude cloud]

the inventor of tesla

ah that tries to be more than just di but a whole spring framework

[wind patio]

kinda yeh

[desert horizon]

I have a hard time understanding the purpse of DI here... Tho I also never had any need to use it, so it's just lack of experience for me I would say.

[wind patio]

wdym by "here"?

or by here you mean you dont understand the purpose of it

[desert horizon]

ye

[wind patio]

well, in a typical rest api scenario (for ex. Spring app), you would have a controller, which handles the request/response, a service, which handles validation and data fetching from repository and a repository, which does the actual database queries

in this example controller has a service dependency, which is "injected" into the controller, service has a repository dependency.

a DI library would make your life easier in a way where you would not need to do the usual "new ObjectService()" and then pass it to the controller "new ObjectController(objectService)" (same for repository)

also for singletons, you dont need all the "instance = this / getInstance" boilerplate, since a DI library can usually handle that for you too via a single annotation

well that's just from my experience, someone can correct me if Im wrong

[prisma wave]

It depends what you’re trying to explain exactly

I think that’s a good explanation of dependency injection as a process but not necessarily why you’d want to use a DI library

DI libraries like guice handle the basic boilerplate but they also move even more away like binding interfaces to implementations, calling factory classes where necessary, and they encourage you to compartmentalise your code a lot more

[wind patio]

yeah that was a part of the explanation, but yeah, DI libs basically just reduce boilerplate

for DI lib ex. I don't need to manually instantiate and store VouchersRepository as a field/variable and I can just "inject" as a parameter anywhere I would need

||oops thats a service and not a repository, but the idea is the same||

[brittle leaf]

when wrapping text, does it look better to have the last word wrap if it goes over the length limit, or to allow that word and then on the next word wrap over?

cus im trying to figure that out

[prisma wave]

The answer is probably very subjective haha

How would you even do the second option though, without overflowing the view area?

[agile galleon]

^

[desert horizon]

I feel like it depends also in what context it's used

For example: Javadocs.
I want to keep a consistent width for the comment, but when a method or class link goes beyond that am I accepting this (or worst case put that entire thing on a new line, but that usually makes the other line look really short and cut off.)

[brittle leaf]

i probs shouldve said the context, this is for an item's lore

[desert horizon]

I feel like a hard limit is needed here, especially since if it gets too long it can become hard to read on certain GUI scales and screen resolutions

Kinda sad that Mojang doesn't have automatic wrapping of long text in lore... I mean the Figura Mod has it!

[distant sun]

Is possible to track user interaction and link clicks in a CSR app? I want to see how many clicks I get for each url from my site that sends the user to another website.

[brittle leaf]

just started messing about with nbt data on itemstacks again. interesting that they have made some hella changes to how that works, very interesting

i havent touched them since i was working on 1.19 and its just a few extra lines to modify it it seems

[brittle leaf]

only a little different. you call for CUSTOM_DATA rather then calling the tag directly from the nms itemstack object

[wind patio]

probably yeah, have you looked into GA4 or GTM?

or Posthog, Plausible, Mixpanel

or use backend proxy

[ocean quartz]

Posthog is pretty nice

[wind patio]

via posthog you can probably do something like:

document.addEventListener("click", function (event) {
  const link = event.target.closest("a");
  if (link && link.href.startsWith("http") && !link.href.includes(window.location.hostname)) {
    posthog.capture("outbound_click", { url: link.href });
  }
});

[agile galleon]

Umami is lightweight and great

[distant sun]

Ill probably use google analytics. Thanks.

[distant sun]

Wdym?

[crude cloud]

why not use PDC for custom data? https://docs.papermc.io/paper/dev/pdc

[wind patio]

well instead of

<a href="https://someurl.com">Blah</a>

you would use a backend

<a href="/redirect?to=https://someurl.com">Visit Example</a>

Where you would "capture" the click

[distant sun]

Yeah my question was more about where can I track the data

[brittle leaf]

i just like trying to do stuff outside of the api. if i was actually making a product then id use the api for it like pdc

[wind patio]

oh, well, original message said "is it possible" 👀

[distant sun]

Fair

[wintry plinth]

Vouch for Umami, I switched to self-hosting it. So awesome

[agile galleon]

And, additionally, blockers like uBlock don't block your own domain when you self host it

[abstract osprey]

When the hell did OfflinePlayer have PDC?

I swear to god it didn’t have that before :/

[crude cloud]

and you'd be correct

it didn't have that before

[abstract osprey]

Do you happen to know when they added OfflinePlayer to support PDC holders perchance

I last used PDC with the 1.21.1 API

[crude cloud]

https://github.com/PaperMC/Paper/pull/8117 whenever this was merged idk

1.21.1

[abstract osprey]

For fuck’s sake

I remember being like

‘Man, I wish I could access offline player PDC’ while being too lazy to check the OfflinePlayer methods

Maybe it was just the docs that didn’t reflect the change? I was using the api last year

[forest pecan]

I really wish I could use something other than citizens for NPC libraries

There’s npc-lib but even that is nothing but using citizens is annoying because I feel like I encounter a bug for every fresh build…

And it’s like the only NPC plugin out there

[wind patio]

time to create a new competitor - "Foreigners"

[abstract osprey]

what are you trying to do with them

are you using citizens traits

[desert horizon]

Well, there is ServerNPCs but that one is limited in features I would say

[remote goblet]

FancyNpcs is a banger

but thats like 1.21 + iirc

[ocean quartz]

Good you shouldn't be using anything lower than that

[desert horizon]

Using older servers is ewww

Same using spigot

[pastel imp]

I might hate vue more than react rn...

updating the whole frontend of hangar is gonna be SUCH a pain

[desert horizon]

good thing I don't touch either I guess

[pastel imp]

xd

[pastel imp]

HELP

whenever I click print screen button

Intellij opens?!

like tf

out of a sudden

[ocean quartz]

Sounds like a nice feature

[pastel imp]

not when I a ctually need that button

bro wtf

[wind patio]

win + shift + s?

[pastel imp]

oh wth that works for screenshots

but why is the print screen assigned to opening god damn intellij

[ocean quartz]

Did you setup some macros or something?

[pastel imp]

I did not

this suddenly happened

lmao

hella weird

[desert horizon]

I just read that HashMap isn't synchronized. What would that affect, if anything?
Because I want to ensure keys are not duplicated (I did encounter such an issue in the past funny enough, which shouldn't be possible?), so not sure if synchronization here is important or not...

[potent nest]

if you want to use a map with multiple threads, don't use HashMap

[desert horizon]

That isn't a concern of mine really.

[wind patio]

ConcurrentHashMap?

[wind patio]

well, you can get duplicate keys if multiple threads are putting values in it

[steel heart]

keys might be duplicated also if hashCode or/and equals are improperly implemented - u prob alr know that tho

[potent nest]

what is then?

[crude cloud]

a synchronized hashmap is only ever really helpful when you are dealing with multiple threads

if you are not dealing with multiple threads, then the fact it isn't synchronized does not matter to you

yeah what is the concern?

[desert horizon]

duplicate keys. But since that only happens on multiple threads, which I'm not having here should this not be an issue.

Although... the method is executed by Cloud's Paper command manager

And it's set to use async command handling, so there probs could be such an issue

[crude cloud]

well, do you need async command handling?

[abstract osprey]

what is the method in question

[desert horizon]

You mean the method used on the hashmap or the method it's used in?

[abstract osprey]

used on the HashMap

[desert horizon]

Does it hurt to have it?
I just use the async coordinator of Cloud itself

[desert horizon]

put

[crude cloud]

multi threading is not the kind of thing you throw to the wall and things just work

data structures and operations need to be specifically designed for it

otherwise completely unexpected effects might happen

anything from an outdated field read to a jvm crash, deadlocks, etc

so yeah it does hurt to have simply because the option is there

it makes things harder to reason about and work with, so you need to evaluate if running things across separate threads is an improvement to your system and how much of an improvement

[forest pecan]

Usually the “less time” benefit from parallelization doesn’t make a big difference for things like commands

In fact, there might be more overhead sometimes by just creating extra threads itself rather than just running them sequentially on the main thread

Also from a developer standpoint, you have to handle now that everything is going to be possibly called not on the main thread, and if you’re accessing the Bukkit API, well.. wouldn’t that be a bit annoying?

You would need something that’s intensive that would ideally be able to be mostly independent so you don’t have to worry about syncing information as much or anything like that

[desert horizon]

Well, looking at the docs, the asynchronous stuff is apparently the command suggestions stuff and uses Paper's own async API for it alongside Brigadier

(Or well, Brigadier is used first if available, followed by Paper's async API from what I can decypher here)

[crude cloud]

if you're explicitly using the async coordinator, everything related to cloud will run on its own thread pool

parsing, execution, suggestions, exception handling

not just brigadier suggestions

it is best to use the simple coordinator

unless you have a reason otherwise

[brazen ether]

anyone interested in hiring CodeBreaker_ShadowScript

bro went with a Pinterest hacking pfp 😭

[prisma wave]

i love hacking

[abstract osprey]

who stole the punctuation

[static zealot]

The hacker

[desert horizon]

Feel like that was a bot because they tagged me for a unrelated thing lol

[lavish notch]

There seems to be some sort of exploit on SpigotMC that allows users to change the currency at checkout when purchasing a premium plugin - someone bought ShopGUI+ for 20 czech crowns instead of 20 USD - has anyone heard about this?

[crude cloud]

isn't the transaction entirely handled by paypal or something like that?

[lavish notch]

I've just managed to replicate the exploit, this is nuts.

I'll wait for Spigot to hopefully fix it and then disclose exploit steps.

[crude cloud]

lol

gg

[static zealot]

You know you can trust me :))

[desert horizon]

so probs never

I have my honest doubts it will be fixed, unless they really work on the xenforo thing

....or you know.... remove paid plugins

[lavish notch]

MD_5 has already replied to my email believing there to not be an exploit and that he has messaged BRC about it. So I'll wait to hear from BRC about the details MD_5 has given him.

[crude cloud]

"yo there is this exploit you can do by doing this and that"

"no"

[half harness]

Btw who's brc?

[lavish notch]

ShopGUIPlus developer.

https://www.spigotmc.org/resources/shopgui-1-8-1-21.6515/

He's the one who initially discovered someone paid 20 CZK instead of USD for his plugin and still got added to the buyers' list.

[half harness]

Oh

[agile galleon]

Maybe it wasn't the best idea to publicly announce that there is an exploit

[cinder flare]

well didn't you hear, md5 said there's not bro

[agile galleon]

If that what I'm looking at is the exploit then it's very easy to find (i hope it isn't)

[agile galleon]

Ah yeah sorry 🙌🥀

Does it still work? I don't want to be poking around myself in these dangerous waters

[steel heart]

cant believe Id ever have to do this, but lets ball

[pastel imp]

Wait what

Who announced an exploif

And what

[brazen ether]

.

[potent nest]

joe

[pastel imp]

Bruh

[wind patio]

yeah, it really is

[agile galleon]

Then that's the stupidest fucking thing I have ever seen

I pray for md5 to have disabled automatic verification or something

[wind patio]

didn't really know what the bug was

took me like 10 seconds to find

[pastel imp]

let me try to find it xd

bruh

ye

[wind patio]

that's what no server-side validation gets you

[pastel imp]

like, lmao

this can be abused so easily

[abstract osprey]

I’m more shocked at how this has only been discovered now

[pastel imp]

I knew md_5 was stupid, but didn't know he was THIS stupid

like, it's just a matter of time until this spreads

[abstract osprey]

This isn’t even like.. a minor exploit

Which makes it even worse

[wind patio]

I think automatic verification is disabled, so not much use

[pastel imp]

yeah

this is going to affect many people

but tbf, in the bright side, maybe people will move out of spigot because of it

I am down for that

[abstract osprey]

don’t worry! md_5 says there’s not an exploit so we can all move on with our day :)

[pastel imp]

yeah exactly!

does that mean I can tell all my friends and their families about the exploit?

should probably ask md_5 that

oh wait

[abstract osprey]

it’s not an exploit!!!!

[pastel imp]

sorry

the non existent exploit

[abstract osprey]

smh

[pastel imp]

gonna try to see if I can contact anyone that has some type of influence/contact with md_5 to make him see things clearly

lol

tbh

[abstract osprey]

cough

[pastel imp]

I would let it burn

gladly would

[abstract osprey]

[pastel imp]

but this doesn't affect only the platform

it affects developers...

[abstract osprey]

it affects a lot of pissed off developers

[pastel imp]

(which tbf, their own fault they chose to use spigotmc)

wanna know something even funnier?

you can change both amount and currency LMAO

@lavish notch ^^

[wind patio]

yeh

[agile galleon]

Does it verify you automatically for the download of the resource? Or has he disabled it

[pastel imp]

didn't purchase anything

I don't want to abuse it...

so idk

[agile galleon]

Same same

[lavish notch]

@agile galleon @pastel imp So it turns out the user wasn't added to the buyer's list with the cheaper purchase.

They made a full payment aan hour later under a different PayPal email and was added then.

[agile galleon]

Ah great, false alarm then

[prisma wave]

Based

[wind patio]

based?? based on what

[desert horizon]

I'm kinda curious about it....

[desert horizon]

You would be surprised how many exploits don't get found out for a long time.

[agile galleon]

Honestly it's very easy to find and it doesn't seem like it is an actual exploit

It's just you sending the developer a donation, as you don't get the plugin according to Mackenzie

[desert horizon]

Would you be okay to share the details with me in DMs?
I would like to make some devs of premium plugins aware of this, so they know and can possibly take meassures.

[agile galleon]

Measures against what exactly?

[desert horizon]

🤷‍♂️

Just want to make them aware, so they aren't confused why some random user sends them like 1 cent in a random currency or smth

[pastel imp]

Wait so it lets you purchase yet it doesnt add to the buyer list if you modify currency or price?

[lavish notch]

It would seem so.

[pastel imp]

Okay that is better

[steel heart]

:,) whats ur opinion on prolog btw?

[crude cloud]

so good

prolog is really interesting

conceptually

but who would ever actually write that lmao

[forest pecan]

wtf 😭

import static me.clip.placeholderapi.libs.kyori.adventure.text.Component.empty;

intellij importing the wrong one

[steel heart]

Lol

[wind patio]

I have a weird problem with IntelliJ - my "test sources" for a module are not being recognized.
I tried marking them as "Test sources root" via the IDE, but it gets reverted after reimporting the gradle project.

tried various things, for ex. my root project build.gradle:

subprojects {
    idea {
        module {
            testSources = files("${project.projectDir}/src/test/java")
        }
    }

    sourceSets {
        test {
            java {
                srcDirs = ["${project.projectDir}/src/test/java"]
            }
        }
    }
}

and module build.gradle:

sourceSets {
    test {
        java {
            srcDirs = ['src/test/java']
        }
        resources {
            srcDirs = ['src/test/resources']
        }
    }
}

idea {
    module {
        testSources.from(sourceSets.test.java.srcDirs)
    }
}

any ideas what might be wrong? can't seem to figure it out at this point

[crude cloud]

I don't use the idea plugin and never specify any source set directory (unless it's custom stuff) and it works wonders

[wind patio]

I had the test dir before, then I removed it completely (somehow), and now I cant get it back lmao

[ocean quartz]

Just java or java-library plugin, don't touch sourceSets and by default both main and test work just fine

[crude cloud]

the key is to not touch anything

[wind patio]

true that

[wind patio]

alright, well, the issue was .idea/workspace.xml had the <module>.test in "unloaded modules" section

[agile galleon]

Well the paypal email is public so you could've done that all along

[half harness]

https://i.imgur.com/vc6iBvR.png
mmmm
Android Studio + Material Theme UI = beautiful

ok I imported the settings from intellij and it's much better now 🙃

[brazen ether]

i hate this game

one is blended and the other is occluded in thsi screenshot

the difference?

SPAWN ORDER

fucking SPAWN ORDER makes this difference

that's so stupid! this is garbage graphics programming!

[abstract osprey]

this is why I avoid anything to do with rendering or similar lmfao

Mojang makes this shit unbearable enough without touching that

[wind patio]

Using build.gradle my output jar was around 20mb, I migrated to build.gradle.kts and the jar dropped to ~13mb.
Is it normal or did I miss something when migrating?

[wind patio]

nvm I see org/sqlite is for some reason excluded or just straight up missing

[brazen ether]

lmfao

[brazen ether]

these should never make a difference

[wind patio]

yeah I dont understand why, there are no changes anything regarding sqlite

the one on the left is when I built after migrating to .kts

org package does not contain sqlite for some reason

I mean, I haven't added implementation() for sqlite, but for some reason build.gradle includes it, build.gradle.kts does not

[crude cloud]

i mean, i would need to see both scripts side by side

[wind patio]

I'll try to see if I can post it tomorrow, gtg sleep

but yeah, a very weird issue

also if I do include sqlite jdbc in build.gradle.kts, the output jar is ~28mb, which is higher than build.gradle output jar (reobfJar)

my only guess is that shadowJar's minimize option is more strict/aggressive in kotlin dsl and sqlite is being removed due to being brough transitively

I suppose I can try to debug minimize, see what it removes

[crude cloud]

that does not make any sense

a plugin can't know the dsl used for writing the script

also minimizing is cringe

[brazen ether]

you could just look at the file

just decompile the jar

you don't have to guess

uon even gotta decompile it

legit just unzip your jar

and you can see what's in there

[half harness]

I'm guessing that the issue lies in an mistake while migrating from groovy to kts

especially if it was your first time

not a shadow issue

[static zealot]

https://www.youtube.com/watch?v=pVI_smLgTY0

[potent nest]

so based

[wind patio]

yeah I did, checked and saw sqlite missing

[wind patio]

why is that

[forest pecan]

I finally got my own repo set up 💀

it took me like 3 hours lmao

trying to get nginx to work

[desert horizon]

What are you using for it?

[forest pecan]

reposilite

this was my first time

but isnt this url great https://repo.brandonli.me/#/

[desert horizon]

Ah, you mean repository like that

I assumed you meant a git repo service

[forest pecan]

oh nah

i wanted my own repo because it'd look nice

and also i cba publishing to central

[desert horizon]

I personally just use the repo host's packages thing, which thankfully allows anonymous access unlike GitHub

[forest pecan]

Yeah

[desert horizon]

Tho, there are some problems, like that it doesn't display the groupId on the page.
https://codeberg.org/Andre601/-/packages/maven/ch.andre601.asl-api-api/4.1.0

[distant sun]

Why are you making an api for the api of american sign language?

[desert horizon]

Because:

[brazen ether]

asl-api-api

I have a genius idea for what my next project is..

[pastel imp]

Let me guess

Asl-api-api-api?

[desert horizon]

Yeah. Just how Codeberg (or rather Forgejo) handles names for packages... <groupId>-<artifactId>

[desert horizon]

Huh..... Just made an update for the API and this time it was processed without any real issues.
The past times I always had a timeout with a "bad request" error.... Only thing I really changed was adding a <groupId>...</groupId>, to the individual pom.xml files of each module (which is redundant as they inherit it from parent)... So I guess forgejo doesn't like the pom not having a missing groupId, or the generated metadata file was messed up....

[distant sun]

Wasn't there a website that had all mc items / blocks and you could embed them on a website? I need something for a project.

I think I found one: https://assets.mcasset.cloud/1.21.5/assets/minecraft/textures/item/acacia_boat.png, but they might be a bit too small to display them on a website

[still portal]

https://minecraftallimages.jemsire.com/ these might work if the ones you’ve found don’t scale nicely

[distant sun]

cool, thanks!

[desert horizon]

https://minerender.org

(Probably)

[abstract osprey]

random question but, does anybody have some good resource for mc sounds?
making a plugin/gamemode and good sounds (pitch/vol/sound combinations) are really hard to get right

[surreal garnet]

Tbf what I’ve done on my project is play sounds I think may fit in-game, since you can mix with pitch and stuffs

[dry harbor]

That's what I do too lol

Like you'd be surprised how good pufferfish blow sounds are for picking up / dropping items 🤔

[spare falcon]

Maybe can use this site to help mess around with 'em, since you can select multiple and test/play them at the same time https://minecraftsounds.com/

[abstract osprey]

oooh that's a good shout

thanks y'all ^^

i'll go stalk 8 year old spigotmc forums too

[lavish notch]

Does anyone know how / if it's possible to dynamically unregister a command on startup?

Basically looking to have a command that can be disabled via a config property and I'm having difficulty unregistering it - so far managed to get it to produce the usage message defined in the plugin.yml but no further... looking for a unknown command output and thus another plugin will be able to take priority should it have a command with the same name registered.

[potent nest]

could you register them manually instead?

[static zealot]

We have dynamic commands for DeluxeMenus. The code for it is here: https://github.com/HelpChat/DeluxeMenus/blob/main/src/main/java/com/extendedclip/deluxemenus/menu/command/RegistrableMenuCommand.java

I had ran into the problem with commands not unregistering as well. The problem is I don't think you can completely unregister them. There is always some fragments of it remaining until a server restart. If I remember exactly, it's actually the bukkit:help command which has no way of unregistering a command. But other than that, it is unregistered from all other places.

[lavish notch]

I was hoping to avoid dynamically registering them as I wasn't too sure what impact it would have on their function.

[lavish notch]

I did consider referencing DeluxeMenus but I wasn't too sure if it was appropriate for my use case. 😄

[potent nest]

shouldn't change much. When using paper-plugin.yml, that's even the only way to register commands at all

[lavish notch]

Somehow that worked - mindlessly copied some of that.

[static zealot]

I hope so. I spent more time than I would like to admit on getting it to work somewhat good

[lavish notch]

😄

[desert horizon]

Something I noticed in the past with DM (And that afaik is still present) is that such commands may not execute, depending on how they are executed.

I remember that setting a command for a ItemsAdder Item didn't run it, if it was registered via DM, as I got the vanilla "incomplete or unknown command" response thing.

And afaik does IA execute the command using performCommand(String)

[static zealot]

And the command was registered? You had tab completion for it and all that?

[desert horizon]

From what I recall, no. It appeared red when typed, but entering it still executed it, if I did so via chat, not via the item.

[static zealot]

Well then the command wasn't actually registered. To do that you have to add register_command: true inside the menu and restart the server.

I think if the command isn't registered, it simply listens for the command process event or whatever it is called and executes the code then

[desert horizon]

Seems a bit weird

[static zealot]

what exactly?

[desert horizon]

that you would need to separately configure it to register the command to begin with.

[static zealot]

I guess? The option wasn't always there and even after it was added, it was quite buggy. We had a bug going on for years and it was only recently solved for example. /dm reload didn't properly unregister the aliases and didn't let go of the old menus so you would be able to open menus without changes being applied to them.

And not everyone wants tab completion for all commands. Sometimes they want the commands to be used by other plugins and such

[forest pecan]

Correct me if I’m wrong, but setting the order of a ByteBuffer using ByteBuffer#order never actually changes the contents, but only influences the read and write methods correct?

I’m supplying a byte[] from a native application that always provides the order in native endianness

Since I use ByteBuffer#wrap and ByteBuffer#array a lot, if I were to wrap my native endianness byte[], then set the byte order using ByteBuffer#order, and then call the array function back, it’s not like any of the values are actually converted. Since in the wrap function they assume that your data is already in big endian format and it only actually uses the order in the read and write methods

[wind patio]

yes

[forest pecan]

Ok I spent like an hour or two debugging this

lol

[wind patio]

just make sure that when you read from or write to the buffer, the ByteOrder matches that of the underlying data
the data itself is untouched unless you explicitly modify it

[forest pecan]

Yes exactly

[forest pecan]

Actually that makes entirely sense, because endianness only affects the bytes in the data type and not an array. An array of bytes in small and big will have the same corresponding bytes in order

[potent nest]

Cool people use VarHandles instead anyway

[crude cloud]

no they don't when the HH plugin is BROKEN

it deleted my intellij

[potent nest]

It‘s 100% bug free (besides some bugs)

[crude cloud]

it bricked my firmware

[potent nest]

Must be a Kotlin bug

[brazen ether]

#1367769665346207744 message "you won't be paid very well" at least they're honest!

[wind patio]

So our "team lead" insists (since we have an "agreement" (which I don't even recall) to prefer to use Stream API instead of for loops) that I should change this for loop (for a dynamic query building):

for (var data : maskedData) {
    var field = table.field(data.columnName(), String.class);
    var value = data.maskedValue();

    updateStep = updateStep.set(field, value);
}

into Stream API, which would look something like:

updateStep = maskingOperationResults.stream()
    .reduce(updateStep,
        (step, data) -> step.set(table.field(data.columnName(), String.class), data.maskedValue()),
        (step1, step2) -> step1); 

    public void update(/* ... */) {
        // ...
        var updateStep = dslContext.update(table)
                .set(table.field(DATE_MODIFIED_FIELD), Timestamp.valueOf(now))
                .set(table.field(USER_CREATED_FIELD), adaUserCreated);

        for (var data : maskingOperationResults) {
            var field = table.field(data.columnName(), String.class);
            var value = data.maskedValue();

            updateStep = updateStep.set(field, value);
        }

        // ...
    }

to which I told I don't see any advantages of using Stream API here and it just makes it less readable. thoughts?

[prisma wave]

i mean i would be in favour of the stream api in principle there since it's a textbook fold/reduce but reduce is kind of an annoying method i guess

[oblique heath]

i think you should prefer whatever is more readable

and as much as i love fp, in most complicated cases imperative programming is more readable

[distant sun]

Yeah I sometimes use for loops over stream api because I have more control over the flow and the code is more readable that way (imo).

[brazen ether]

LMFAO

#1367769665346207744 is a comedy club bro

[oblique heath]

"must be done today" 💀

[brazen ether]

it's hilarious

[prisma wave]

would argue this isnt actually very complicated but sure

[wind patio]

I mean, personally if it makes things easier and more readable - I'm all up for it, but in this case I would argue that the stream api is more confusing than just the plain old for loop

[oblique heath]

imo for reduce specifically, unless you can pointfree the perfect method it usually gets out of control quick

[wind patio]

I would understand if its a bad implementation, difficult to read or poor performance, but I am sure none of these apply in this case

generally I would just "fix" it and go on with my day, but it's not the first time I get "change this but it will not have any effect whatsoever I just want it to be this way" code review comments and it's getting annoying lol

probably an indicator to look for a new job

[prisma wave]

Only because most people don’t know what reduce means

[brazen ether]

ts really is art. beautiful

[prisma wave]

perhaps .set is a bad name for this kind of method because it implies a mutable update

[abstract osprey]

i'd love to interview one of these people and know their train of thought in this

[abstract osprey]

and this lol
$80-100 bucks for a sub-24h deadline for a complicated plugin

[woeful flame]

i completed that commission. honestly, it wasn’t anything huge they just wanted some basic mechanics modified, I'd say $80 isn’t too little for that, at least in my opinion

[crude cloud]

i love streams!

but j.u.s is meh

reactor!

[abstract osprey]

eh fair enough, although the deadline was a little ridiculous, glad you got round to it though

[forest pecan]

Streams are a pain to debug

In my opinion I would rather use the loop in this case

[brazen ether]

awesome, i'm surprised they found smth with that quick of a deadline

[cinder flare]

yeah i would definitely do the for loop there, i love folds in theory but in this scenario it's just calling methods on a builder over and over again

if you had to use streams, honestly i would probably prefer a .map().forEach(), so it's clear that you're mapping to a pair of columnName -> data, then setting that in the query builder or whatever

now if you could pass in a hashmap already populated with that data to the query builder, then you could just map().toHashMap() or whatever and pass it in and that would be pretty nice

[wind patio]

yeah its kind of a weird case

in a typical scenario I would go with a .map or something like that as well, but I just don't see the need for stream api here

anyways, my tech lead is annoying as fuck sometimes

always with this back and forth, we talk, we come to a conclusion that what I did is "good", but then he comes back and tells me to re do it lol

"SOLID", "OCP" blah blah blah

bro I don't care, it works, it does the job, its readable, the performance is fine, its not that difficult to maintain, just fuck off already

so yeah, see yall next monday

[onyx loom]

the managers just gotta keep you working for no reason

[crude cloud]

egh

hate people that follow those "principles" like they're bibles

even calling them principles makes me gag

guidelines :)

[wind patio]

god forbid you make a class that does TWO THINGS

oh no no no young buck that aint gonna fly

[forest pecan]

Does anyone know like some website that provides compiled dummy shared libraries across Linux platforms for like different arch

I'm using a Java library that uses OpenCV which loads a library functionality that's not even used, but only the name is used for linking if that makes sense

it uses libgtk but its not even needed for my usecase cause i have nothing to do with ui

and im just trying to load a fake dummy lib so it thinks that it exists

Actually wait Go cross compile is perfect for this so I don't have to get into docker mess

[desert horizon]

Currently thinking about adding support to my plugin for merging multiple images into a single one to use as the favicon, but I'm not too sure how I should implement this into my FaviconHandler...

To not break pre-existing configuration (And to also not having to deal with checking for what type the option is) would I probs go with a semi-colon separated String to determine the invdidual parts.

Difficult part is - as mentioned - the FaviconHandler. Would apreciate ideas for it:
https://codeberg.org/Andre601/AdvancedServerList/src/branch/master/core/src/main/java/ch/andre601/advancedserverlist/core/profiles/handlers/FaviconHandler.java

To clarify: The F generic is the Favicon instance the server/proxy is using. and the createFavicon(image) method simply uses whatever method the server/proxy provides to create a Favicon from a BufferedImage.

[desert horizon]

Made this stuff now.... Not too sure if that is any good honestly
https://paste.helpch.at/jihenilaxu.java

[agile galleon]

Cloudflare is blocking GitHub actions from accessing my Reposilite instance, any ideas on how to fix this?

[potent nest]

https://docs.github.com/en/actions/using-github-hosted-runners/using-github-hosted-runners/about-github-hosted-runners#ip-addresses

https://github.com/marketplace/actions/bypass-cloudflare-for-github-action

[agile galleon]

Currently using that, but it doesn't add the IP address to the WAF and then kills itself... I have recreated the API token twice and copied the correct Zone ID, I don't know what else to do since it doesn't say if the request was successful or not

Recreated it a third time but now custom building that and now it works, apparently the Edit Zone template isn't correct anymore

[pastel imp]

Yoo any volunteers to get tagged on a tweet? Appwrite doing a giveaway and I gotta tag 3 devs... thing is.. me not use twitter at all soo me have no one there 🥺

[desert horizon]

Eh.... idk

[pastel imp]

forget it

the winners would get a custom hoodie with a contribution graph

which mine would be 0

which seems unfair to others

[lavish notch]

Looking for opinions, I can't decide if this Java warning message is obvious/big enough for the below average (or average at this point) SpigotMC server owner... what do you guys think?

I'm going to pop it as an on-join message for server operators as well but I'm contemplating adding line breaks above and below or something similar to emphasis it more.

For context, I have a feature release pending (Folia Support to be specific) which requires Java 17 and the plugin is currently Java 8 based. I want to avoid the negative reviews and sentiment from idiots not be able to read basic error messages like 'Compiled with a newer Java version'.

*also ignore the typo in the warning message lol, just noticed as I posted this

[potent nest]

I assume you are only printing the warning when using <17?

[static zealot]

What warning message?

[lavish notch]

Yup.

[lavish notch]

[potent nest]

I mean I guess it can't be any clearer than that

[lavish notch]

For example, there are people who run servers with 100s of error/red messages in their console and just wouldn't read that somehow. Not sure how to better target those people... I guess you could call them edge-case people.

[potent nest]

if they complain because they're bad at managing their servers, there isn't much you can do

[ocean quartz]

If someone runs the server with hundreds of errors on startup then they deserve to not see the message

[lavish notch]

But they'll then leave a 1 star review because of their incompetence.

Beginning to wonder if this is too much text for an on-join warning and if I should include like a Java X detected, please update so the reader gets the idea that the message will go away when they update the server Java version past 17.

[oblique heath]

you can send a warning in chat to people with elevated perms

• add a config option to disable that ig

since someone will always complain

[pastel imp]

For a second thought it was stopping support for spigot but sad

[lavish notch]

BRC doesn't really have the time to pick up Paper's new features sadly. Even then, if we were to leave Spigot, we'd also likely leave 1.8 to maybe 1.12?

[lavish notch]

The point is to force people to update as well, we're increasing the minimum java version from 8 to 17 in July.

[desert horizon]

The sad thing is, people will still complain.

You cannot avoid this

[lavish notch]

I know...

[desert horizon]

Bruh, Barry being mean

[lavish notch]

Even Barry is after me. 😦

[desert horizon]

I f.e. have this, but I'm sure people will ignore it

[lavish notch]

It's just sad how stupid people can be.

[desert horizon]

And lazy

[lavish notch]

That too.

[desert horizon]

Not to long ago did someone ask for support while using v3.3.0-b1

Which is the latest version I have on Spigot

...where I have this big text at the top of the page

[lavish notch]

Ouch

[wind patio]

don't bother, those who read console/logs - will see it, those who don't - wont

at least you did your part

also personally I would log with WARN level, not INFO 🤓

[lavish notch]

It's not actually done via Logger - did it via ConsoleSender so I could ensure colour shows in all kinds of consoles.

[wind patio]

yeah personally it bothers me every time I see any red/yellow lines

[wind patio]

gimme like 2 hours and I am rewriting that shit to paper

[lavish notch]

😄

[cinder flare]

lol Paper 1.19+ is the way to go

that's still like 95% of servers and is a massive upgrade

[lavish notch]

This is SellGUI's bstats for reference.

SellGUI being an addon for ShopGUI+

[wind patio]

so from this chart 1.8 is lower than 0.5%, whats the point of supporting it

you are just holding yourself back lol

[cinder flare]

yeah I stand by my assessment lol

that's even better than i thought it would be tbh

[lavish notch]

I would happily jump to 1.19+, it's just ShopGUI+ is such an old plugin that many still use it on say 1.8.

[cinder flare]

then they can still use all your older versions lol

[lavish notch]

Bare in mind, the graph I sent is from an addon which not all SGP people use and also some people turn bstats off.

[cinder flare]

it's not like you're deleting it, just not actively updating it

[wind patio]

I don't think old server versions that you would drop support for would even need any "new" features, given there would be any

[cinder flare]

yeah exactly, if they want to use an 11 year old minecraft version, they can use a 1 year old addon version or whatever

[wind patio]

as Star said, they can continue using old versions

[lavish notch]

True true.

Well, I suppose I can pop dropping 1.16 and below on the table maybe.

[wind patio]

you deprecate old stuff and tell the users to move on and update or use unsupported versions

[lavish notch]

The negative reviews though.

[wind patio]

who cares lol

[lavish notch]

Me sadly.

[wind patio]

imo the plugin is popular enough, people can also have their own opinions without looking at the reviews

[lavish notch]

Fair enough.

[wind patio]

usually I just ignore version/update or outlier complaints since they don't bring any value

[pastel imp]

May I suggest leaving up till 1.20.x? 🙂

[pastel imp]

Why you only releasing in codeberg? Does hangar link to it too?

[lavish notch]

1.16-1.19 still kinda popular

[pastel imp]

Kinda does not mean it is relevant lol

Majority is 1.18+ at least

1.18 should be the lowest anyone should go

(Totally not because of minimessage being included on paper)

Tbf my opinion is biased af

Afterall, I am the one losing countless hours redesigning Hangar lol

So ofc I will be biased towards paper

And newer versions

[lavish notch]

🔥

[pastel imp]

confusion

[desert horizon]

If you would check the repo, you would see links to modrinth and hangar

Also, any of you good with python and pip?
I've reset my PC and now installed python again (Via their Python install manager) and now want my pip install s to be stored on the D: drive instead of C:
I changed the --target via PIP_TARGET environment variable, but now installed packages that add a command - in my case MkDocs - don't have the command installed and working... Not sure what else I can/should do.

[wind patio]

I remember I had to use python and pip back in the day

this shit was so annoying

[desert horizon]

Yeah, it can be

[pastel imp]

Well, on a bus and on my phone, not really ideal to check anything, no need to answer passive aggressively lol

[desert horizon]

All I really want rn is have packages be installed on D: instead of C: as the later is mainly for the OS itself

[wind patio]

iirc the issue is that the entrypoint is wrong

you could try using virtual environment

well, you can do this:

Install on another drive:
pip install --prefix=D:\python-packages mkdocs

Modify your PATH environment variable, add:
D:\python-packages\Scripts

Create a new environment variable PYTHONPATH, set the value to:
D:\python-packages\Lib\site-packages

[desert horizon]

uh....

I feel like we have some differences here?

Like where does the Scripts folder come from?

[wind patio]

when you install via pip install --prefix=D:\python-packages mkdocs it creates it

[desert horizon]

That's.... annoying

I probs just bite the sour apple and have packages on the default drive until I find a fix...

[wind patio]

well, if that makes any difference I have Pip v24.3.1 and Python 3.10.6

[wind patio]

why is that?

[desert horizon]

Well, mine is python 3.14

[wind patio]

did pip install --prefix=D:\python-packages mkdocs not create a D:\python-packages\Scripts dir for you?

[desert horizon]

I didn't bother with this as I don't see this as the right aproach here

Because having to do this for mkdocs would mean I have to do it for every other python project that adds commands

Tho, even resetting package stuff to the default does have the issue

So I start to believe this is a general python issue

[wind patio]

you would just install other packages via:
pip install --prefix=D:\python-packages <package-name>

the rest of the steps are one time set and forget

the difference here is I use --prefix instead of --target. Target is used to install packages into a flat, standalone dir and does not use the standard python directory layout

[desert horizon]

I think the primary issue was either the install manager or the 2.14 version, which apparently is beta

(Not sure why the default is to install a beta, but whatever)

[wind patio]

oh you said you are using 3.14, not 2.14

[desert horizon]

deinstalling all python stuff and then re-installing 3.13 via an installer seems to properly add pip as a standalone command

[desert horizon]

I mentioned that

Yeah, the mkdocs command works

So it was related to the Python installation or the way it was installed.

[cinder flare]

Tbh if you can afford to, using uv for Python package management has been a dream

It’s basically Cargo from Rust but for Python

Fixes all my issues with their horrible package management and stuff

[potent nest]

pixi >>>

(I'm not a python dev)

[cinder flare]

oh it's a conda thing?

uv is basically a replacement for normal pip/pipenv/virtualenv stuff, not conda related or anything

[potent nest]

I have no clue

I only know a few people who regularly contribute to pixi stuff

[cinder flare]

yeah i mean from what i saw it doesn't really seem that comparable lol

and uv is like at least 10x more popular, so i would recommend uv

[prisma wave]

does anyone remember the name of the development collaboration platform website that's kinda like jetbrains space

it had a like black and silver theme

i think either star or matt was talking about it ages ago but i cant find it anywhere

it's quite a new one, not a well established thing like Trello or Jira

nvm i found it lmao it's called Linear

was in my bookmarks this whole time my eyes just glossed over it

[abstract osprey]

oh

I didn’t read down far enough

[prisma wave]

classic

[abstract osprey]

thanks iOS for the autocorrect

[pastel imp]

Was about to say xd, I use it

[desert horizon]

Really need to find a good way of merging multiple BufferedImages in an async fashion (i.e. in a CompletableFuture.supplyAsync call).
The one I tried doesn't work, probably due to a logic failure on my end.

[forest pecan]

what do you mean by merge

like in a pipeline?

[desert horizon]

Combine them into a single image. Like treat each as a new layer to add

[forest pecan]

how would the combining work

[desert horizon]

My idea is, that the user can define multiple sources separated by a semicolon (To still support the current String option)
I would go through each part, obtain the BufferedImage for each and then combine them one by one.

The annoying part is, that I want to keep this async, so that it doesn't delay the Event this is called in too much... I already do that with the current setup for a single image where local ones are pre-made at enable and URL or Name/UUID ones are created asynchronously.

[prisma wave]

combine them how

[forest pecan]

^

[prisma wave]

overlaying them?

seems like it'd be easy enough to do a divide and conquer if so

[forest pecan]

putting them side by side?

[desert horizon]

I did say I treat them as layers right?
Like I don't think there is much guessing here

But yes, overlaying them

My current nightmare code: https://paste.helpch.at/ufebokabuf.java

[forest pecan]

I see like templates

[prisma wave]

The one I tried doesn't work
in what way

[desert horizon]

The image it has to fetch doesn't get added

Only the local one, which already exists

[prisma wave]

what does the debug output show

[desert horizon]

Like I tested with ${player uuid};vanillaplus.png which effectively would be converted to https://mc-heads.net/avatar/286f8d0c-b571-4720-b7ab-f2929cb38120/64;vanillaplus.png
And the output is just the vanillaplus.png image.

[prisma wave]

does it work with 2 local images?

[desert horizon]

https://paste.helpch.at/upuragezop.md

I separated the two ping event calls.
First one is with no cached image and second is with cached.

[prisma wave]

Layer 1 was null! Skipping...
alright well this is why

the web requests takes a while so it isn't completed in time

future.getNow(null); this is probably not what you should be doing

[forest pecan]

Yeah its not enough time

[prisma wave]

in fact composeFavicon should probably take a List<BufferedImage> instead

use CompleteableFuture.allOf and then you know when all the results are completed

[desert horizon]

Can't think of another aproach that wouldn't delay the event unessecarely

[prisma wave]

well yeah...

[forest pecan]

you can't really do anything about that...

web request in an event

[prisma wave]

if something takes 5 seconds it takes 5 seconds, asynchronous or synchronous

doing it async doesnt make it faster it just avoids lagging the server

[desert horizon]

Handling allOf is annoying tho, as it returns a CompletableFuture<Void>, so I need to handle the individual CF in a different way

[prisma wave]

yeah its a stupid method but you can then safely use .join() on all of them

instead of getNow

[desert horizon]

I did try something like that

[prisma wave]

and?

[desert horizon]

Well, it didn't seem to actually receive the stuff. Tho, I didn't use join() then, so that was probably the issue

But looking at this, I think I can throw a lot of the CF stuff away.
The only method used outside the class to make a Favicon is the favicon(String) method. So I can probably just make a CF there.

[prisma wave]

It does seem like you’re misunderstanding/misusing them a bit, yeah

[steel heart]

one day we'll have structured concurrency 🙏 and then we can say good bye to allOf() (not really but kinda :D)

[desert horizon]

Seems to work now

[cobalt marlin]

i mean virtual threads are a thing now

[potent nest]

but structured concurrency is still in preview

[forest pecan]

virtual threads are nice but thats different

[desert horizon]

Any info on that?

[potent nest]

They reworked some stuff, I didn’t followed it too closely so not sure what changed and why

But I guess they‘re getting there slowly

[desert horizon]

Looked at it a bit and eh....

[potent nest]

What eh

Like… what are your expectations

[desert horizon]

idk honestly

[steel heart]

Yea- they’re cool and in combination with gen zgc we can get insane performance, tho iirc virtual threads still have some issues w thread pinning like blocking inside a class initializer due to a native frame on the stack or resolving symbolic references (think there were some other small issues that cause unnecessary pinning)

[crude cloud]

yeah uh VTs don't really solve the need for CFs, lol, they're completely orthogonal

[abstract osprey]

iirc there are some allocation issues for virtual threads?
something along those lines

it was fixed in java 24 iirc?

but

virtual threads are amazing yes

[steel heart]

not allocation issues I think- but as said pinning issues- one major issue was pinning due to synchronized, but that’s been fixed there are still some small pinning issues tho

[desert horizon]

Would any of you have experience with OpenAPI and fetching its JSON in Java?
I have a JSON file for an API and want to extract its paths values to return a different, more simplified JSON like this:

{
  "paths": [
    {
      "url": "https://api.example.com/v1/some/path",
      "method": "GET"
    },
    {
      "url": "https://api.example.com/v1/some/other/path",
      "method": "POST",
      "request_body_schema": "https://docs.example.com/api/#post-schema"
    },
    {
      "url": "https://api.example.com/v1/yet/another/{path}",
      "method": "GET",
      "parameters": [
        {
          "name": "path",
          "location": "path",
          "required": true,
          "type": "string",
          "supported_values": [
            "value1",
            "value2"
          ]
        }
      ]
    }
  ]
}

...assuming the paths JSON of the openapi file is like this:

"/some/path": {
  "get": {
    "...": "Unrelated JSON"
  }
},
"/some/other/path": {
  "post": {
    "requestBody": {
      "content": {
        "application/json": {
          "schema": {
            "$ref": "#/components/schemas/post-schema"
          }
        }
      }
    },
    "...": "Other unrelated JSON"
  }
},
"/yet/another/{path}": {
  "get": {
    "parameters": {
      {
        "name": "path",
        "in": "path",
        "required": true,
        "schema": {
          "type": "string",
          "enum": [
            "value1",
            "value2"
          ]
        }
      }
    },
    "...": "Other unrelated JSON"
  }
}

I could probs use the parser they offer for Java to make POJO's, but at a first look does it seem a bit complicated in structure.

[cinder flare]

Yeah I mean parsing to a pojo then mapping to another pojo and serializing is as good as you’re gonna get probably

[potent nest]

nothing that couldn't be done with a simple jq tramsformation

[cinder flare]

lol when are they rewriting jq in Java, it already has j in the name!

[potent nest]

jq4j

[pastel imp]

Why not use openai's api java wrapper?

Like, as you said, it could make your life easier

Why are you choosing to go the "manual" route?

I mean, I guess complicated structure is fair

Although you are probably more prone to error by using json

[abstract osprey]

Holy crap, people didn’t like that #showcase message

[agile galleon]

Why is it produced like a YouTube short

I can't even see what's happening through the frantic zooms

[wind patio]

just gen z stuff

[abstract osprey]

it’s the most downvoted showcase I’ve seen so far lol

[ocean quartz]

Tbh the edit is terrible

[oblique heath]

feels like what i would expect some ai powered video editing program to come up with

[forest pecan]

i don't understand the edit

he just points to the italian brainrot at the end but i dont understand it

[pastel imp]

me neither

showcases something that has been done tons of times already

like it's something wow

with a very bad edit

and then shows some random text at the end

idk

lol

[lavish notch]

Does anyone know if it's against Spigot's rules to promote off-site payment methods on a plugin page? (Eg. Bank Transfer or Crypto)

This would be an option for individuals who don't have access to PayPal due to regional restrictions.

[pastel imp]

Ew crypto

[oblique heath]

ew banks

[desert horizon]

From what I know and saw is it allowed, if you get permission first

[lavish notch]

I'm not doing crypto, was just an example since I've seen it's semi-common.

[solar cradle]

Hi! Anyone here a CSS nerd?

I have managed to add a backdrop-filter of blur to a <dialog> element, but I cannot seem to figure out how to fade it in. One StackOverflow thread mentioned using opacity() but the below CSS just snaps in the blur (And it seems that Safari doesn't support the :open tag so that isn't great). Any help would be appreciated :)

dialog::backdrop {
    transition: backdrop-filter 2s;
    backdrop-filter: blur(4px) opacity(0);
}
dialog:open::backdrop {
    backdrop-filter: blur(4px) opacity(1);
}

Here's a minimum reproduction Codepen that I've set up, most of this is just copy paste from the mdn docs, my real code is quite different - although the CSS is identical.
https://codepen.io/Zax71/pen/MYwvLwd

[desert horizon]

Welp.... Tried to use the Swagger parser now and.... It complains about required not being an array?
For a schema component that doesn't have such a thing? Like.... it's weird

[oblique heath]

i have found those errors to be super unhelpful

probably just some json somewhere being wrong in some way that has very little to do with that error message

[desert horizon]

The JSON is valid, as I'm using it to generate content for an API page and it would otherwise raise errors there if it wouldn't parse to valid JSON.

[oblique heath]

when i say the json being wrong i mean it probably doesnt adhere to json schema, not that its not valid json

[desert horizon]

I found the issue

Apparently does required inside schema objects resemble a list/array of properties that are required to be included in the schema

So if I f.e. have

{
  "text": "hello world",
  "something": true
}

and I want to define that text is required, then I would have "required": ["text"] set in the JSON schema

[abstract osprey]

have you tried making a keyframes rule?

  @keyframes fadeInOpacity {
      0% {
          opacity: 0;
      }
      100% {
          opacity: 1;
      }
  }

then you could make a css class called whatever (e.g. fade-in)

.fade-in {
      opacity: 1;
      animation-name: fadeInOpacity;
      animation-iteration-count: 1;
      animation-timing-function: ease-in;
      animation-duration: 1s;
}```

and apply that class to the dialog element in HTML :P

[solar cradle]

Ooh, I have found a solution but it uses some new CSS features that aren’t in Firefox yet so I’ll give that a go as well

[abstract osprey]

Oh right sorry, didn’t realise that.

[solar cradle]

Dw, it was in a different server, you had no way to know xD

[brazen ether]

implementing a simple-shape rigidbody physics solver over the next week or so and quickly got this hunkered down; demonstration of angular velocity and why your phone always flips sideways when when you try to spin it

(basically just copied code from here: https://www.youtube.com/watch?v=4r_EvmPKOvY&ab_channel=blackedout01 but gonna see about re implementing this with quaternions because there are some things that seem to be easier with them than rotation matrices)

[brazen ether]

quaternion rotation. no more orthonormalization issues

• faster

https://math.unm.edu/~vageli/courses/Ma375/literature/rrr.pdf

awesome paper for learning more about quaternions in general too

explains a lot of the basic stuff well

[desert horizon]

Does one of you guys know an easy way to create a site that accepts a /api/webhooks/<id>/<token>/<bitmap> URL, takes the <bitmap> to determine the set of events to forward and based on the received info send a Discord Webhook with the <id> and <token>?

Kinda want to make a sort of webhook filter for Gitea/Forgejo, as their Discord webhook sends webhooks for all events, even ones you may not want (i.e. I enable releases and they send webhooks for release creation, edit and deletion when I only want the first)...

Kinda want to make it a thing others could use if they want to, hence the bitmask, as I think assigning a unique number to each event type could be the best aproach...

Main issue I see right now is get a setup to host on vercel AND to also determine what kind of event a received POST request is, as f.e. release contains the same event type header, but has a action in its JSON body afaik.

[desert horizon]

Got something basic that actually works....

Tho, now I need to find out all the individual event types a webhook can trigger AND also sent the apropriate JSON.... fun.

[brazen ether]

day #2 of working on this... rk4 integration method for super a stable system, raycast and physically accurate linear and angular momentum from impulse application

after this it's collision solving + generalizing to ellipsoids + optimization... and then it goes into Comet!

[brazen ether]

collision detection+ response

[agile galleon]

Is there any website to see packet definitions and how they have changed over time, I can't figure out for which versions I need to make a seperate definition...

[abstract osprey]

linkie for packets when

[agile galleon]

What's linkie?

Or is there possibly an NMS viewer for each version?

[brazen ether]

Minecraft protocol wiki history LMFAO

[abstract osprey]

something used for mappings

http://linkie.shedaniel.dev

absolute godsend of a website

lets you move your code between different mappings and/or versions

[agile galleon]

Is there a way to see the constructor of classes on that website?

[abstract osprey]

You used to be able to view the mapped class code but it just returns 500 errors now

[agile galleon]

yeah

[sweet cipher]

https://mappings.dev/

[pastel imp]

tbf this channel kinda indeed died a tiny bit after the GH linked roles thing

but at least convos here are actually relevant

[crude cloud]

bob and vagene

[brazen ether]

ngl I don't mind

#development fills the same role of shitposting

[pastel imp]

xD

[distant sun]

https://youtu.be/yup8gIXxWDU?si=9mBBYTyaLM1J7_EM

[brazen ether]

something seems off here 🤔 https://media.discordapp.net/attachments/955636370817314877/1388482203771605023/2025-06-28_04-27-49.mp4?ex=68612452&is=685fd2d2&hm=29800056913aba4285be4e707fa80112b804994e43d8889b3485d1abd7f4d3c2&

[brazen ether]

https://github.com/Kotlin/kotlin-lsp

standalone Kotlin lsp SOON

this a gem for me and my fellow neovim users 🙏

[brazen ether]

though right now its

quite bad

[desert horizon]

You gotta love stuff breaking without any known reason :/

[potent nest]

sounds like broken index

just do the repair ide stuff

[brazen ether]

LSP crashout

intellij has this thing where it can't function at all if it's in a project which is lacking a build system

even the semantic highlighting is worse than what treesitter gets you in neovim

so if you decompiled something or you're looking at open source with private build configurations, it becomes way less useful

[desert horizon]

Fixed it. Thanks

tho, this happens every single time I reload maven/resolve dependencies

So something else must be broken

I get the feeling that the MC dev plugin may be a possible cause here, as I just encountered an IDE error and everything hints at the dev plugin being the cause.

[crude cloud]

tbf you can highlight with different colours a package vs an interface vs an abstract class vs a non-abstract class vs a static method vs an instance method vs ...... on and on

so it kinda needs to have all classes indexed

i'm not saying that it's for the colour highlighting alone

but if you want anything more than a bitch basic "hmm yes this looks like a method and this looks like a keyword" syntax highlighting, you need that info

i have it set up so every thing is highlighted differently, and it's very nice and easy to read!

[brazen ether]

you're right, i have a proper semantic colorscheme setup for nvim and that makes the bigger difference

cursor trails in intellij btw!

[wind patio]

very annoying

personally, the less visual clutter & animations the better

[crude cloud]

dear christ

it looks like i'm playing neon white all over again

[desert horizon]

Way too distracting.

[brazen ether]

helps keep track of cursor position

[desert horizon]

Not for me tbh

[brazen ether]

feature a lot of terminals have that I enjoy and it was fun to implement

[brazen ether]

do you use ideavim

it not then yeah I can see why it wouldn't make much sense for you

[half harness]

https://i.imgur.com/NxD20cQ.png
btw what is this?
-# not a mac user

[cinder flare]

Window Manager with way too many desktops lol

[brazen ether]

mac has desktops you can use

basically like

diff workspaces

there's 16 max

[brazen ether]

no actually

default thijg

and this isn't even being displayed by a wm

I use ~10 currently

different one for each application

mostly

or just the most that would make sense to have on one screen

[cinder flare]

You’re telling me that’s not like yabai or something?

[brazen ether]

I use yabai but you don't need yabai to have that

diff desktops are vanilla thing

sketchybar for that display

[cinder flare]

Well yeah I mean same on windows I guess

Oh the bar isn’t just part of yabai?

[brazen ether]

ik aerospace has their own thing but yabai works so I have no reason to switch

[brazen ether]

nah they don't have that

[cinder flare]

Damn

[brazen ether]

I got cmus integration for the music display widget :>

[cinder flare]

Bro’s living the like arch Linux lifestyle on macOS lmao

[distant sun]

Why does the line number change .. like that?

[brazen ether]

relative line numbers

ideavim thing

like if I need to go to a line, I can glance at the relative like number and then type like 29j or 8k (j and k move cursor down and up, respectively)

tho you just have it by intuition after a while

[distant sun]

Ok ig

[wind patio]

mfs would rather remember 30 different shortcuts than use 3 mouse clicks

[cinder flare]

omg how big is your .ideavimrc

i recently added a bunch of shit to mine lol, still trying to get myself to adopt stuff slowly

i recently "discovered" % hops between bracket pairs, shit is so nice in like huge SQL functions or whatever to find what's matching what

cause i would just like click on it and scroll down to find the matching highlighted bracket, but that shit is hard to find lol

[cinder flare]

to be fair, it's a lot less about remembering once you get into the habit

and you don't take your fingers off of home row most of the time so it's super quick

[ocean quartz]

I have no opinions on vim because well use what you like, but I always find funny to see people using it like for example ThePrimeagen and pressing a bunch of keys just to have the cursor move a bit lol

[brazen ether]

this is it

set mini-ai
set functiontextobj
set multicursor

set scrolloff=10
set ignorecase
set smartcase

set incsearch

map Q gq

Plug 'machakann/vim-highlightedyank'

map <Space>r <Action>(ReformatCode)
map <Space>e <Action>(GotoNextError)
map <Space>E <Action>(GotoPreviousError)

map <Space>1 <Action>(GoToTab1)
map <Space>2 <Action>(GoToTab2)
map <Space>3 <Action>(GoToTab3)
map <Space>4 <Action>(GoToTab4)
map <Space>5 <Action>(GoToTab5)
map <Space>6 <Action>(GoToTab6)
map <Space>7 <Action>(GoToTab7)
map <Space>8 <Action>(GoToTab8)
map <Space>9 <Action>(GoToTab9)

map <Space>s <Action>(SplitHorizontally)
map <Space>v <Action>(SplitVertically)

set number relativenumber
set clipboard=unnamedplus

[brazen ether]

it's extremely easy to remember

literally muscle memory

[cinder flare]

yea idk i just think it's cool

and it's really nice to just move the cursor around really quick instead of like holding the arrow keys for 5 seconds or whatever

especially when it's smart and you can do better than just simple cursor moving, like the matching brackets thing i mentioned, or like editing basically any textobject (ci(, ciw, etc.)

[brazen ether]

i would say most useful things for me:

• space + number for going to tabs, but i eventually will replace this with my own intellij plugin so i can navigate around things just like how i do in neovim
• relative line numbers is the best way 90% to quickly go to some arbitrary line that you can see
• space e for gotonexterror is awesome
• macros are awesome and have been a huge timesave for me; learn to use them intuitively, that's how they get fast
• for navigating inside a line, searching with / + n is pretty great
• i often find myself replacing parts of words, so learn f, F, t, T ; it's really common to do some motion like ct<something>

[brazen ether]

it's legit just enjoyable

[brazen ether]

btw you can download mini-ai intellij plugin to get text objects around functions and classes

but it doesn't work 100% of the time at least in kotlin

i'll fork it and fix it eventually?

cause it's really cool in theory

i set up custom text objects in neovim for yaml stuff

[brazen ether]

use cib instead of ci(

easier to type

[cinder flare]

well usually i'm specifying which bracket i want to edit in a lot of nested ones

[brazen ether]

ah yeah makes sense

[cinder flare]

like () in a function surrounded with {}

[brazen ether]

B for { btw

[cinder flare]

oh wait actually? i have a bunch of the textobject ones but they kinda suck iirc

[brazen ether]

and also big thing @cinder flare try to get into habit of using your right shift more if you don't already

[cinder flare]

and yeah i use f,F,t,T all the time

oh I never use right shift lol

[brazen ether]

yeah 😭 i'ma fix them when i have the time

[brazen ether]

i've noticed that there's a lot of situations where it's really unergonomic to use left shift

like hurting my hands typa thing

[cinder flare]

https://paste.helpch.at/ohovecuyuj.xml

here's my .ideavimrc

quite a bit in there lol

[brazen ether]

also it's faster to use right + left at the same time instead of left

lemme c'

[cinder flare]

oh my right pinky is just not trained on that lol, that's gonna take a while to get used to

oh yo why do i have double c-w binds lol

[brazen ether]

ima steal some of these

my ideavim config is so sparse compared to nvim

[cinder flare]

yeah those are really nice, i usually use gd and gu a lot

also a really nice one for me has been <leader>ga to show Git Blame gutter annotations

[brazen ether]

yes i use gd and have been neglecting to do other ones

[cinder flare]

cause idk if there's normally a keybind but i had been right clicking, hovering over git, clicking on annotate

and that shit is slow as hell

cause i always wanna know which one of my coworkers wrote a fucked up function or something lol, so that's been a huge timesaver for me

and yeah i recently learned more about the jump list, so c-o and c-i, which has been sweet with like gd and stuff

[brazen ether]

i'm gonna try and clean up my intellij from some stuff

cause

i don't use vast majority of what's on screen

minecraft plugin dev is just not that intense

[cinder flare]

but yk what really pisses me off, gj and gk don't work in the non-modal commit message editor

and it's like a known issue that they only work in the main editor window and not in any other text box

[brazen ether]

i don't use intellij for git stuff atp, i use lazygit, but i don't even think it has vim impl for commit msgs

[cinder flare]

but i always write long git commit lines cause the formatting on those is weird, then i try to navigate up/down cause it wrapped it, but it doesn't work 😠

[brazen ether]

tho it def supports external editor so really easy to get nvim setup there if i wanted to

[cinder flare]

oh i have really come to love IJ git stuff working at my current job, it's so nice to see inline commit histories and diffs and shit, to review my changes before committing, stashing changes when i have to swap branches to work on a random ticket or something, it's been really nice

[brazen ether]

biggest thing missing from intellij for me is nvim-style undotrees

i hate local history

[cinder flare]

oh yeah i never really got into them, i just press u a lot lol

[brazen ether]

all that exists outside intellij ofc but yeh it's not the same

[cinder flare]

and i've slowly been getting myself to use buffers more, it's just so unergonomic to me to press "a or whatever all the time

[cinder flare]

well from what i understand the stash stuff is not like base git and that shit saves my life i stg

er

not stash, shelf

which saves like context and a bunch of crap, and you can have multiple changelists and it separates them nicely in-editor, idk i just have really grown to love it

[brazen ether]

i rewrote the rendering for an nvim plugin, and this is ofc an exaggeration but i find it genuinely useful in work when there's multiple branches for a file like this from complex undos, it also marks when you saved it so you can easily go back to saved states

[brazen ether]

stash is in base git

[cinder flare]

yea yea i meant shelf

[brazen ether]

that's just non-commit work saving right

[cinder flare]

yeah it's just like doing partial saving for when you want to swap branches or whatevs

[brazen ether]

is there a difference between git stash

ah doesn't seem like it

[cinder flare]

yea you can select certain files

[brazen ether]

but yes it's super useful

[cinder flare]

stash is every uncommitted change

shelf is you can pick and choose into multiple separate stashes really easily and stuff

[brazen ether]

i've been working on physics stuff and that can't exactly be broken up into atomic commits

yeah lazygit has allat

[cinder flare]

ah well maybe i should get into it lol

[brazen ether]

stash has been super useful

[cinder flare]

but tbh i just love IJ stuff

[brazen ether]

nah

if intelij works don't bother

[cinder flare]

recently discovered how to use the IJ profiler and SQL Query Plan analysis, holy shit it's been a gamechanger

[brazen ether]

i have 0 experience with any of that shit

[cinder flare]

i'm working on a huge commit right now to just optimize a million things that i never would've known without nice flamegraphs and stuff

yeah my only profiling experience before this was like Spark Minecraft reports lmao

[brazen ether]

does this work with mc dev 😭

that seems way more useful than spark

[cinder flare]

yeah theoretically it should, i mean it's probably just using the same Async Profiler that Spark uses

but it does show you like gutter annotations for long methods and stuff

[brazen ether]

i'ma check that out

physics sim is very performance sensitive

[cinder flare]

like that one sql query took a long fucking time lol

[brazen ether]

ah yeah that's definitely important

[brazen ether]

is that in CE

or ultimate only

[cinder flare]

probably Ultimate onyl?

but yeah it shows you the flamegraph super nice, you can click on stuff to navigate, sort it by CPU Time, Real Time, or Memory Used

super impressive shit i gotta say

[brazen ether]

$800 yearly

that's a lot of dabloons

[brazen ether]

(for ultimate + ai)

[brazen ether]

yeah does seem to be!

but awesome shit from them

it's nice when enterprise level software is actually enterprise level

[brazen ether]

oh nvm $270

[cinder flare]

yeah no it's like, superb

i always loved IJ, but since i've started my job i've really come to appreciate all it does and why it is truly the best JVM IDE experience you can get

[brazen ether]

oh shit they have free student shit

[cinder flare]

and yeah that shit is pretty cheap tbh, i'm probably gonna get just the yearly Jetbrains All Products plan once my student discount runs out

[brazen ether]

applied for that

[cinder flare]

yea it's amazing

[brazen ether]

maybe they'll accept it?

[cinder flare]

i've been using it for like 6 years lol

and then you get like a 30% discount when you graduate or whatever

it's like amazing

[brazen ether]

I'm a highschool student so I hope they accept that

[cinder flare]

oh yeah totally

i did that shit in middle school lmao

[brazen ether]

I should have been on this earlier wtf 😭

[brazen ether]

dawg I been using inferior software since 7th grade in this case

this is crazy

[cinder flare]

haha lol