🧩-plugin-development

[amber condor]

interesting

i'll take a look into it later

because i will also be adding a character counter

instead of displaying how many characters are left, it would display how many characters have been used

[hushed smelt]

hmm what is the name for when you click an image?

[trail ginkgo]

some tips: we have a built-in escape for matching minified variable names \i, they can change between builds and even have symbols/be more than one char, so it makes patches more reliable

same with the find

also small nitpick but the space after var will always be there so the ? isnt necessary ^^

[hasty stag]

thanks for the tips

{
    find: ".CHARACTER_COUNT_OVER_LIMIT",
    replacement: {
        match: /var \i,\i=\i\.type,\i=\i\.textValue,\i=\i\.maxCharacterCount,\i=\i\.className/,
        replace: "return null;$&"
    }
},

made it a bit better

[trail ginkgo]

still should prob replace the es with \is but yee

also, u dont need to put the whole match in a capture group, you can just use $& to output the whole match

[hasty stag]

oh nice

[cedar olive]

also avoid unecessary spaces

best would be "return null;$&"

[green vessel]

human tersing

[hasty stag]

edited now

[hasty stag]

@amber condor do you intend on allowing large message edits too?

[cedar olive]

not doable

[lavish sable]

Why, what's blocking that?

[hasty stag]

by that i mean the edit will be the first 2k chars in the message then send a message with the remaining

[lavish sable]

Yea that should work

[cedar olive]

you are gonna end up sending a message with others in the way

so like

first part of message

[hasty stag]

if edits will not be supported they need to have the blocking modal on them still

i tried it with his plugin and it will just ignore the edit

with no feedback

[lavish sable]

https://github.com/mwittrien/BetterDiscordAddons/blob/2aa5956d7ed349bc407b66d95acbeba526328745/Plugins/SplitLargeMessages/SplitLargeMessages.plugin.js Can you translate a bd plugin manually or is that not possible?

[hasty stag]

its possible pierre has pretty much done it

[cedar olive]

second part

[cedar olive]

ahhhh I see

[next stone]

oh boy I sure am the only one in the Devs list with a full name

either way, lgtm

[opal fern]

Lmfao

[amber condor]

i was afraid of that but i forgot

[dull magnet]

I was gonna do that but I realised that doesn't really help us in any way, it's much better to just have the types directly in our mod

makes it easier to maintain

[snow jetty]

Sounds reasonable

[next stone]

you can monorepo it

that's what I plan to do with my mod

[dull magnet]

we just have them in src/webpack/common/types

[next stone]

monorepo it and autopublish to npm

[dull magnet]

my favourite method

[amber mantle]

me when the static method returns a trans

[pure temple]

that should be in rust

[chrome folio]

q, does someone have a dash-to-space plugin or theme, too lazy to do myself

[dull magnet]

me

#🎨-css-snippets message

[chrome folio]

thanks

[hushed smelt]

hey is there any tutorial to do a basic plugin for someone who is starting?

[dull magnet]

not yet

look at other plugins for how to

[hushed smelt]

oh okey thanks

[viral roost]

there are some docs on github but they arent the most useful or the most up to date
https://github.com/Vendicated/Vencord/blob/main/docs/2_PLUGINS.md

yeah probably best to look at other plugins

[hushed smelt]

yeah i've read that but it isn't helping at all xD

[amber condor]

i found a fix for the edit thing

of SplitLargeMessages

are function arguments e by default?

[viral roost]

there are some exceptions

[amber condor]

is it safe to assume so?

[viral roost]

its not safe to assume its e

well 99% of the time it is

but not always

[hasty stag]

whats the need for it? do you just need to match or do you need to replace code with it

if you need to use it in replacement and you know for certain where it will be you could add a word capture group and use $1 to get it maybe theres better ways though

[amber condor]

i need to read from it

[hasty stag]

you could try what i said

[amber condor]

i will

i recently saw someone post something about lookbacks

is it possible to search for the most recent function \w+\((\w+)\) after the original match ?

[faint siren]

// $1: argument name (currently "e")
// $2: the rest inbetween the argument name and my actual match
// $3: my actual match
match: /(?<=function \w+\()(\w)(.*)(\(0,\w\.jsx\))(?=(.(?!\3))+?canDM)/,
// paste my fancy custom button above the message field
replace: "$1$2$self.patchPopout($1),$3",

@amber condor

[viral roost]

also you can use \i instead of \w to match minified vars

[faint siren]

some custom feature for vencord?

[viral roost]

yes

[faint siren]

ah bruh

[amber condor]

thats complicated 🥴

.*\Kfunction\s\w+\((\w+)\)(?=.*;if\((\w+\.length>\w+)\)({\w+&&null)) i somehow came up with this but its very inneficient

[faint siren]

yea, didn't know there's \i till now

[amber condor]

it doesnt even work

[dull magnet]

this is really bad

it will match the first function in the entire module then match up to the thing

[faint siren]

the whole module is this function

[tame nexus]

Does someone know how to create a store with react states? I know how to create a store and do it like this.

const store = new Promise<IDBObjectStore>(res => DataStore.createStore(PluginInfo.PLUGIN_NAME, STORE_NAME)("readwrite", res));

But what's the best way to manage the states with that? Sorry if I'm missing something obvious but I only have experience with https://github.com/pmndrs/zustand and know that Discord uses flux store. Or do I need to create a context and manage all my states manually?

[amber condor]

like lets say we have stringcookies1 cookies2 banana cookies3
and we first want to search for banana, then how do we get the most recently occured cookies\d before banana?

[faint siren]

well the thing i sent basically is the solution to your problem, but i'm too dumb to explain it

try matching cookie first, then use negative lookahead and then match banana

[woven lion]

do you want cookies2 or banana

i assume the former

[dull magnet]

what are u trying to do?

[woven lion]

if so, then this is what you want cookies\d(?= banana) for that specific example

[dull magnet]

/banana(?<=cookies\d).+?banana/

[amber condor]

cookies2

[woven lion]

then this, i guess

[tame nexus]

I'm trying to store my plugin settings. But I don't use definePluginSettings because I need to display them somewhere else and differently. And I never used definePluginSettings so I don't even know how it really works.

[woven lion]

looking for cookies with banana after it

[amber condor]

yeah but there wont be just spaces, there will be a to of random characters

in reality it would be cookies1 {var t;} cookies2 {somethingsomething = ()} banana cookies3

[amber condor]

this actually does not mutch anything for me

[woven lion]

then you can swap it out for whatever you want

im doing it based on your example lol

regex is an awfully annoying system

[dull magnet]

im in a game rn

[amber mantle]

balls gaming

[shrewd yacht]

remenber guys, pee is stored in the balls

[amber condor]

.*\K(cookies\d)(?=.*banana) this worksgreat but i assume it would be hella slow

[dull magnet]

"cookies1 cookies2 banana cookies3".match(/banana(?<=cookies(\d).+?banana)/)

first group will be 2

[dull magnet]

what kind of settings are u trying to store

[amber condor]

lemme try a diff site ig

[dull magnet]

switch to ecmascript

youre on php

[amber condor]

bruh why is there a diff in regex between languages 😭

[woven lion]

because implementation detail

@hushed bloom uses regex but uses a rust implementation which doesnt have these lookaheads for example

[tame nexus]

Something like this

type Settings = {
    width: number,
    height: number,
    fps: number,
    codec: string,
    resolutionEnabled: boolean,
    fpsEnabled: boolean,
    codecEnabled: boolean;
    ...
};

[snow jetty]

you can access and set settings from your plugin i believe

but i don't know if you can have custom settings that don't show in the plugin settings, if that's what you're trying to do

[dull magnet]

ah this is for ur plugin

you can just manually use settings

or use datastore

u dont need to make a store

just DataStore.get() and DataStore.set()

[tame nexus]

Hm I'm not sure if that's what I meant. But if I would set and get my settings like this and lets say I'm displaying them somewhere in a react component would they be rerendered if a setting changes?

[dull magnet]

no youd be responsible urself of keeping track

[snow jetty]

also ven when could i expect to know if the plugin i ported might get merged?

[dull magnet]

you can make a simple react hook to do it

i can help in a bit

[amber condor]

thanks i got it working

/;if\(\w+\.length>\w+\){\w+&&null(?<=function\s\w+\((\w+)\).+?;if\((\w+\.length>\w+)\)({\w+&&null))/

[woven lion]

dont use \w+ for identifiers btw

use \i

[amber condor]

im doing that right now

match: /;if\(\i\.length>\i\){\i&&null(?<=function\s\i\((\i)\).+?;if\((\i\.length>\i)\)({\i&&null))/, :)

[snow jetty]

omg this regex looks awful

i hate regex

[amber condor]

// match: /;if\((\i\.length>\i)\)({\i&&null)/, this was my original one

[snow jetty]

https://imgs.xkcd.com/comics/perl_problems.png

[woven lion]

faster than monkeypatches though

so

[snow jetty]

i remember trying to mod krunker.io by patching similar to how vencord does it

turns out it was possible a few months before i tried and it was how some hacks worked

but they improved their anticheat and i have no clue how they validate the game source

[amber condor]

now i just gotta patch this out

[snow jetty]

can't be that hard

surely

[amber condor]

yeah

wel maybe some proplem thos

[tame nexus]

That would be great :).

[dull magnet]

function useDataStore<T>(key: string): [T | undefined, (newValue: T) => Promise<void>, boolean];
function useDataStore<T>(key: string, defaultValue: T): [T, (newValue: T) => Promise<void>, boolean];
function useDataStore(key: string, defaultValue?: unknown) {
    const [state, setState] = useState({
        value: defaultValue,
        loading: true
    });

    useEffect(() => {
        let alive = true;
        DataStore.get(key).then(v => alive && setState({
            value: v ?? defaultValue,
            loading: false
        }));

        return () => void (alive = false);
    }, [key, defaultValue]);

    return [
        state.value,
        async (newValue: unknown) => {
            await DataStore.set(key, newValue);
            setState({
                value: newValue,
                loading: false
            });
        },
        state.loading
    ];
}

[soft onyx]

i hate keywordAlterts

[dull magnet]

usage ```ts
const [value, setValue, isLoading] = useDataStore<Settings>("YourPluginName-settings");

// might be undefined until loaded, if you supply a default value (2nd param), it won't be
value?.codecEnabled;
setValue({
codecEnabled: false,
...
})

[tame nexus]

Wow that's great and exactly what I needed! Thank you that helped a lot 🙂

[dull magnet]

i havent tested it btw

[tame nexus]

Looks good to me I'm gonna test it now

[eternal lion]

Can't u just use CSS to display: none it

[amber condor]

yes but i actually want to replace the number instead

so it can show how many characters are inside the message

[eternal lion]

Oh I see

[amber condor]

now just gotta find a reliable regex to reference the variable holding the count

[eternal lion]

For the color you probs dont want that to be red, you can maybe use CSS with !important

[amber condor]

i actually wanted to keep it red tbh

it shows you when you if the message is gonna be split

[woven lion]

id recommend a different color like blue or something

red implies its not going to send

[dull magnet]

i think the red is from discord

[woven lion]

it is ye

cause that'd be the overcount

[amber condor]

hmmm

well i'll see what i can do

because it still has to be red when its and edit message

[dull magnet]

[style*="color:red"] {
  color: brown !important;
}

so easy

[shrewd yacht]

can I change the flex-flow of the child of a column flex-flow?

[amber condor]

i guess it works

[amber condor]

what do you guys think?

the green line is what i've changed during patching

originally it was just children:I

[woven lion]

e is reused but n and o still refer to the variables at the top

if you really want to check, you could throw the entire function body into vscode, reformat it, and rename the variables using F2

then you can see which variables map to what

@amber condor

[amber condor]

i know

but theres a chance the characters will collide right?

because function(...) arent always function(e)

[woven lion]

[amber condor]

so in some case it might be function(n)

[woven lion]

that's why you use capture groups in your regex

then you can capture the identifier and use it later

[amber condor]

yeah i know

[woven lion]

rolecoloreverywhere makes extensive use of this

[amber condor]

thats what i currently use

but what if the two highlighted names are the same?

[woven lion]

they never will be

the build tool guarantees that

[amber condor]

you sure?

[woven lion]

yes

because if that second e was an n you've just broken the functionality of the app

they're two distinctly different values

so the build tool guarantees that it wont overwrite variable names if they have meaning

[amber condor]

not if the e is replaced everywhere with n (which might happen when discord "smallifies" their code right?)

[woven lion]

wha?

the minification step is done by the build tool

so its guaranteed that the variable names wont clash like that

the only reason why its reusing e here is beacuse e is not used anywhere else in that function

and you shouldnt need to either, its props are extracted already at the top to separate variables

[amber condor]

what i meant is: is there a chance discord will ship their code with these two highlighted names being the same? because the variables in the green line are added by my patch, not discord itself

[woven lion]

well we have no idea

swc could do anything it wants

but most likely no

[amber condor]

before my patch, only I was in the function

[woven lion]

okay, odds are your patch will remain working

[amber condor]

then iguess we just gotta live with that

[woven lion]

i severely doubt swc will decide to start using n or o

but even then

as long as your patch is written correctly

it should be resilient against it

[amber condor]

match: /;if\(\i\.length>\i\){\i&&null(?<=function\s\i\((\i)\).+?;if\((\i\.length>\i)\)({\i&&null))/,
replace: ";if(($2)&&($1.type.analyticsName!=\"normal\"))$3"

wait wrong one

match: /,{children:\w+}\)\)}}\)]}\)(?<=function\s\w+\((\w+)\){var \w+,(\w+)=\w+\.type,(\w+)=\w+.textValue[.\s\S]+?,{children:(\w+)}\)\)}}\)]}\))/,
replace: ",{children:$2.analyticsName===\"normal\"?$3.length:$4}))}})]})"

$1 should equal to e, but that's overwritten so thats why im not using it (even tho its still being captured

[woven lion]

you shouldnt use it regardless

because as i mentioned before they're already being extracted

you should use the extracted props directly when possible

[dull magnet]

that patch is strange

seems overly complex

[woven lion]

i dont quite understand it myself

i was thinking it was going to be something like this

https://github.com/Vendicated/Vencord/blob/main/src/plugins/roleColorEverywhere.tsx#L66-L69

[proud parrot]

**roleColorEverywhere.tsx: **Lines 66-69

{
    match: /function \i\((\i)\).{5,20}id.{5,20}guildId.{5,10}channelId.{100,150}hidePersonalInformation.{5,50}jsx.{5,20},{/,
    replace: "$&color:$self.getUserColor($1.id,{guildId:$1?.guildId}),"
}

[woven lion]

actually thats a good point

[woven lion]

this patch is very hyperspecific and will probably break very quickly when discord releases a new build that changes that code

[amber condor]

it had to be specific because there are alot of ,{children:\i}, even with the almost random looking curly, normal and square brackets behind them

,{children:(\w+)}\)\)}}\)] this is the shortest i can make it for the to-be-replaced part, because otherwise it will match to multiple things

[dull magnet]

what are u trying to replace

show screenshot and say what u wanna add

[amber condor]

so basically

when you type close to 2000 characters it will show how many characters you got left, if you go above 2000 characters it will show a negative number

but i want to replace that with just the amount of characters in the textbox

which is pretty easy

the I on the green line is the amount of characters thats left to use

which is pretty easy to accomplish

but it should only do this when its a normal text box (not text box when editing a message)

so i came up with this:

[cedar olive]

I'm almost sure vencord streaming quality bypass actually causes crashes

still testing with it disabled

[amber condor]

@dull magnet forgot to ping u

[hasty stag]

@amber condor i got this showing

[covert nimbus]

4

[hasty stag]

its the message length

[covert nimbus]

4

[hasty stag]

4

{
    find: ".CHARACTER_COUNT_OVER_LIMIT",
    replacement: {
        match: /(function \i\((\i)\){)(var \i,\i=\i\.type,\i=\i\.textValue,\i=\i\.maxCharacterCount,\i=\i\.className)/,
        replace: "$1return ($self.renderText($2));$3"
    }
},

this is the patch i made

renderText: e => <TextComponent length={e.textValue?.length} />
``` then i did this

function TextComponent({ length }: { length: number; }) {
    return (
        <div className="msg-content-length-text">
            {length}
        </div>
    );
}
``` and this is to show the number

[tame nexus]

Do you know if there is a way to create a global store using the function you sent me? When I use the useDataStore function it only creates a store in the react component where I call it. That means I could have two stores which are independent of each other. I could easily solve this issue by simply creating the store in a react context, but there is no way for me to wrap the app component with a provider. I need to wrap the app component because I need to use the store in components I don't really have access to. I'm just wondering how other people have solved this problem because it's really unlikely I'm the only one using some sort of a store.

[hasty stag]

im pretty sure you can do a global store using a class extending Flux.Store

[dull magnet]

you could just use normal settings api

if ur data is small like this

and useSettings() from anywhere

[shadow jacinth]

#suggestions

no suggestions wqtf

anyways my suggestion add profiles for rich presence then ill uninstall bd

[hasty stag]

is there a way to access the guild icon component

[green vessel]

u can use github discussions for plugin suggestions https://github.com/Vendicated/Vencord/discussions

[shadow jacinth]

oke

ty

[amber condor]

i already got it working too

[amber condor]

.

[hasty stag]

oh didnt see that

[amber mantle]

when the trolling

[covert nimbus]

that means it's no longer binding!!!!!!!!! /j

[amber condor]

real

[amber condor]

if you want to make edits

i made a branch on my forked repo

i know how i can patch the character count better

but ive spent too much time on Vencord the last week

i really gotta start working on school stuff because exams in 2 weeks and have to do a assignment before next week, but i hadnt started yet 💀

[covert nimbus]

procrastination

[tame nexus]

Is there a way to replace an object in the Settings Proxy object? If I do

getPluginSettings().currentProfile = profile

It just doesn't replace anything. The profile object conains informations about the current profile like (width, height, ...). The only solution iv'e found is

Object.assign(getPluginSettings().currentProfile, profile);

but this isn't really a solution because I need to replace everything and not merge it.

[dull magnet]

idk what getPluginSettings is

[tame nexus]

Oh sorry forgot to add the function. Here it is

export function getPluginSettings(): PluginSettings {
    return Settings.plugins[PluginInfo.PLUGIN_NAME] as PluginSettings;
}

PLUGIN_NAME is the name of my plugin.

[dull magnet]

dont use it like this

[dull magnet]

oh, this should work fine

it doesnt?

[tame nexus]

Wait I will check something

Okay no. If I change the code to

getPluginSettings().currentProfile = profile
console.log(getPluginSettings().currentProfile);

it still prints other the old profile. I also checked the profile that it sets and it's a different one so it's something else...

The weird thing is if I print out getPluginSettings() it has the changed profile but if I print out getPluginSettings().currentProfile I get the unchanged one. Maybe it has something to do with the Proxy? I don't really have knowledge about the Proxy object but maybe it has something to do with it.

Ah and I found out something else. If I set it the way I do it (getPluginSettings().currentProfile = profile) currentProfile becomes a normal javascript object and isn't a proxy object anymore.

[amber condor]

if i make a pr, will my branch still exist?

[dull magnet]

wha

not sure what u mean by that!

[amber condor]

like

atm i have Vencord forked

i made a branch called SplitLargeMessages

[dull magnet]

and

[amber condor]

if i make a pull request

and it gets accepted

does that mean my branche will get "deleted"

(i am github noob)

[dull magnet]

it won't but you should delete it afterwards

why do you ask?

if you want to push a second feature, make a second branch

[amber condor]

i was just curious because im github noob

its only 1 file, i would be able to just copy it back anyway

[dull magnet]

why would you need to add it back

if it's merged into main then you don't need the feature branch anymore

[amber condor]

true

but if i wanted to make a edit

i'd have to recreate the branch right

[tame nexus]

Okay I found a solution and don't know if it's a bug so it would be really helpful if you could tell me. If I change this

// Wraps the passed settings object in a Proxy to nicely handle change listeners and default values
function makeProxy(settings: any, root = settings, path = ""): Settings {
    return proxyCache[path] ??= new Proxy(settings, {
     ...

to this

// Wraps the passed settings object in a Proxy to nicely handle change listeners and default values
function makeProxy(settings: any, root = settings, path = ""): Settings {
    return proxyCache[path] = new Proxy(settings, {
     ...

it works because I changed the ´??=´ operator to =. Does that mean it's a solution to my problem or is it some sort of bypass of a mistake I made?

[dull magnet]

ah I see

hmm that shouldn't happen though

ah no I get it

[dull magnet]

add ```js
delete proxyCache[path];

and it should work

[tame nexus]

What is a set trap? Do you mean the set function? Something like this?

 set(target, p: string, v) {
            // avoid unnecessary updates to React Components and other listeners
            if (target[p] === v) return true;

            target[p] = v;
            delete proxyCache[path];

            // Call any listeners that are listening to a setting of this path
            const setPath = `${path}${path && "."}${p}`;
            for (const subscription of subscriptions) {
                if (!subscription._path || subscription._path === setPath) {
                    subscription(v, setPath);
                }
            }
            // And don't forget to persist the settings!
            VencordNative.ipc.invoke(IpcEvents.SET_SETTINGS, JSON.stringify(root, null, 4));
            return true;
}

[dull magnet]

yes

[tame nexus]

Okay I tried that it didn't work for some reason

[dull magnet]

ah

move it one line down

and change path to setPath

[tame nexus]

Ah yes now it works

[dull magnet]

the reason that cache exists is because otherwise it would create and discard hundreds of instances

which is just not necessary

every time you access a setting it would

settings.     // create proxy
plugins.      // create proxy and discard the first
YourPlugin.   // create proxy and discard the previous one
SomeSetting   // discard proxy

[tame nexus]

Oh that makes sense

[dull magnet]

wait

does android not use a momospace font

for codeblocks

I can't manage to align the //

[tame nexus]

I don't know hahaha

[dull magnet]

ah now

I got it

[tame nexus]

But is this whole thing some sort of oversight or something like that or only a fix for my plugin?

[dull magnet]

oversight

[tame nexus]

should I do a pr?

[dull magnet]

the reason this happens is because every time you get a value, it proxies it if it's an object and caches that proxy (for above mentioned reasons), then reuses that proxy for future use

if you overwrite the object later, it still has the cached proxy that still proxies the old object

So the settings actually do update correctly but the proxy returns the old value

simply deleting the cached proxy fixes it like u saw

[dull magnet]

yes please

[tame nexus]

Okay gonna do 🙂

[dull magnet]

git commit messages should be present tense :P

[tame nexus]

Oh sorry I don't have that much experience with github and even their naming conventions. And my english isn't the best as well hahaha.

[dull magnet]

that's fine

[dull magnet]

well it's slightly subjective, however most projects use present tense and even the git guidelines tell you to use present tense

https://git.kernel.org/pub/scm/git/git.git/tree/Documentation/SubmittingPatches?h=v2.36.1#n181

[tame nexus]

That's good to know! Learned something new again thank you 🙂

[snow jetty]

@dull magnet sorry for the ping, i just wanted to check with you if the edits i'm making correspond to the edits you requested, here's a diff https://img.remty.xyz/Rx7L65aW.png

[dull magnet]

ye

does it work?

[snow jetty]

yup

[dull magnet]

also u should probably add { noop: true } as 2nd param to the ErrorBoundary

[snow jetty]

compiles and runs just like before

dam

[dull magnet]

ever seen one of the red error blocks?

when opening settings

[snow jetty]

i think i have

yes

[dull magnet]

that's the default fallback of our ErrorBoundary

[snow jetty]

i see

[dull magnet]

passing { noop: true } makes it render nothing instead of the error box

it would be quite ugly to have a red error box randomly in settings so noop is probably better :P

[snow jetty]

right, makes more sense here

[dull magnet]

anyway every time you see one of those it means the ErrorBoundary just saved u from a crash

thats why u should always use them if possible

[snow jetty]

also what's the policy on trailing commas, cuz i saw i accidentally left some you removed

[dull magnet]

doesnt really matter, should just be consistent

[snow jetty]

okay thanks

also sorry for doing a PR from the main branch i forgot when making my commit

@dull magnet fuck i broke it

don't merge kek

wrapping in the errorboundary broke it somehow

[dull magnet]

huh?

[snow jetty]

yeah i need your help here

[dull magnet]

how is it broken?

[hasty stag]

the way i decided to do it was this but im not sure if you would want to

[snow jetty]

my colors are undefined

wait

hmm

[amber condor]

we could make it a option

but it'd be pretty performance heavy i assume

[hasty stag]

i dont really think

[amber condor]

having to calculate how many chunks there are for every keystroke

[hasty stag]

its just math.floor(messagelength / maxcharcount)

[amber condor]

not if you have SplitPlace.NewLine

[hasty stag]

lol i forgot about that

[amber condor]

idk if you read my code

[snow jetty]

i think i'm just stupid don't worry ven

[hasty stag]

i did but i only use the single char one

[dull magnet]

simply debounce it :P

[amber condor]

wat is debouncing

[snow jetty]

okay i'm just stupid

[amber condor]

fair enough

[dull magnet]

https://github.com/Vendicated/Vencord/blob/main/src/utils/debounce.ts#L19-L32

[proud parrot]

**debounce.ts: **Lines 19-32

/**
 * Returns a new function that will call the wrapped function
 * after the specified delay. If the function is called again
 * within the delay, the timer will be reset.
 * @param func The function to wrap
 * @param delay The delay in milliseconds
 */
export function debounce<T extends Function>(func: T, delay = 300): T {
    let timeout: NodeJS.Timeout;
    return function (...args: any[]) {
        clearTimeout(timeout);
        timeout = setTimeout(() => { func(...args); }, delay);
    } as any;
}

[snow jetty]

i wasn't calling my function with my colors object

[dull magnet]

essentially, every time you call it it will clear previous schedules and schedule the callback in delay milliseconds

[snow jetty]

the only place where typing doesn't happen and i didn't think to check guuuh

[dull magnet]

so if you type 10 characters it will keep cancelling the previous timers

then once you stop typing for 300ms, it will run the callback

:P

[snow jetty]

okay fix pushed ven

[amber condor]

hmm i guess its doable then

[hasty stag]

got an idea

actually no wouldnt work because the function only runs on send

[amber condor]

what was your idea?

[hasty stag]

make a global chunks variable to only use the function as little as i can

[amber condor]

that wouldnt really work no

[dull magnet]

wtf

trying to do some tauri shenanigans

[covert nimbus]

I'm curious, what's a usecase for this

[dull magnet]

ayo

[hasty stag]

debouncing wouldnt work since it returns a promise and i need to return a jsx element

[dull magnet]

what

just use an effect

[next stone]

rewrite armcord

[amber condor]

time to recreate discord in rust

something ive wanted for a long time

@hasty stag do you have your code opensource?

[hasty stag]

i can send it but its mostly yours

i added this to patches

        {
            find: ".CHARACTER_COUNT_OVER_LIMIT",
            replacement: {
                match: /(function \i\((\i)\){)(var \i,\i=\i\.type,\i=\i\.textValue,\i=\i\.maxCharacterCount,\i=\i\.className)/,
                replace: "$1return ($self.renderText($2));$3"
            }
        },

    renderText: e => <TextComponent data={e.textValue} />
``` and this

export function TextComponent({ data }: { data: string; }) {
    if (data.length <= 0) return null;
    const [chunks, setChunks] = useState<string[]>([]);

    useEffect(() => {
        debounce(() => {
            setChunks(intoChunks(data, getMaxMessageLength(UserStore.getCurrentUser().premiumType ?? PremiumType.None), splitterSettings.store.splitPlace));
        }, 300);

        return () => { };
    }, [data]);

    return (
        <div className="msg-content-length-text">
            {`${data.length}: ${chunks.length} ${chunks.length > 1 ? "Chunks" : "Chunk"}`}
        </div>
    );
}
``` this doesnt work for showing chunks yet i still need to figure out the useeffect

[cedar olive]

yooo is that good?

like does it work normally or misses functionality electron has

[trail ginkgo]

its probably not much different, just that you'd be writing ur 'preload' logic in rust

i wonder how good of an idea that'd be

[dull magnet]

it seems terrible

[dawn elm]

https://cdn.discordapp.com/emojis/1024751730237374484.gif?size=48&quality=lossless

[steel oriole]

i wonder how do i replicate BANger using js

?

[unreal mason]

the active now page seems to hide when the window width is under 1201 pixels, which is annoying since my discord monitor has a width of 1050pixels. does someone know how to lower this number with vencord?

[snow jetty]

If you're comfortable with installing userplugins I'm willing to have a look

[unreal mason]

i dont have a problem with that 🤷‍♂️

although ive never done it in vencord, i only switched recently

[snow jetty]

it requires compiling vencord yourself

updates are still automatic tho

[unreal mason]

oh. that makes it slightly more complicated, although id give it a try

is there a way to keep my current configuration after selfcompiling?

[snow jetty]

yeah it's not too hard

[trail ginkgo]

yea ur config isnt stored in the same place

[snow jetty]

is it-

[trail ginkgo]

ur config isn't stored next to ur installation no

[snow jetty]

it's always in roaming right

[trail ginkgo]

yea

[snow jetty]

but you can have the repo cloned elsewhere?

[trail ginkgo]

although i think the installer might put the installation next to ur config by default

[trail ginkgo]

yea

[snow jetty]

also am i right about updates still working the same

i never actually checked i believe

[trail ginkgo]

idk i dont update

[snow jetty]

lol

[trail ginkgo]

i think the updater still works

just uses a different method

[chrome elbow]

do you want to make it doesnt hide or something

[class*="nowPlayingColumn-"] {
  display: block;
}

[unreal mason]

yeaah

[chrome elbow]

put that to custom css

[snow jetty]

welp i didn't think to check css

[unreal mason]

ahh

that was fast

thank you

thats perfect

and without any compiling xd

[snow jetty]

yeah lol

[chrome elbow]

@unreal mason change it to what i edited, so it wont break when discord change these random letters, i forgor

[unreal mason]

ah right, makes sense

why does dc do the random letters tho?

[trail ginkgo]

it has to do with their css bundler, sometimes they use the same class names in different components, so the random letters specify which component it belongs to

so they dont overlap

[unreal mason]

alright

[dull magnet]

it works completely different under the hood, but it's the same flow for the user so you don't notice

[steel oriole]

is there a way to replicate banger plugin with js?

[dull magnet]

what

[snow jetty]

yeah that's what i thought

[dull magnet]

we have two updater backends, one that shells out to git and node and one that uses the github api

[snow jetty]

sounds like it might've been a pain to make

[dull magnet]

nah fairly simple

the git one we already had initially, it was the only way to install when vencord was new

then I just wrote an equivalent http updater

https://github.com/Vendicated/Vencord/blob/main/src/ipcMain/updater/index.ts

[proud parrot]

**index.ts: **

/*
 * Vencord, a modification for Discord's desktop app
 * Copyright (c) 2022 Vendicated and contributors
 *
 * This program is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program.  If not, see <https://www.gnu.org/licenses/>.
*/

import(IS_STANDALONE ? "./http" : "./git");

... (-10 lines left)

[dull magnet]

-10 lines left

[snow jetty]

lol

[trail ginkgo]

this file is 11 lines

or 10

[amber condor]

looks interesting

[shrewd yacht]

how can I make discord use the browser font?

[dull magnet]

css

[shrewd yacht]

what css?

[dull magnet]

#🎨-css-snippets message

just skip the Google fonts part

[shrewd yacht]

my problem is that I don't know what font to set it to so it's the browser font

If I say it I'll get banned ||it's fira code||

idk if it's serif or not

[trail ginkgo]

--font-primary: sans-serif; or --font-primary: monospace; will probably work

i think those are the browser provided font names

[shrewd yacht]

Yeah it worked thx

=> ligatures yay!

|----|-||----|>---|

[shrewd yacht]

How can I use css to modify "Avoid pronouns, use my name" to "Name" (The first takes too much place)

[dull magnet]

no

[shrewd yacht]

😦

no it doesn't take too much space or no it's not possible?

[dull magnet]

not possible with css

unless the plugin is edited to add some attribute

like data-pronouns

[shrewd yacht]

can I PR this?

ig I can hide it until hovering the message

with css

[shrewd yacht]

I don't use the plugin, I use the web extension which has a .pronoundb-pronouns class

[shrewd yacht]

knowing that, how can I do that ⬆?

there is also .vc-pronoundb-compact for the plugin

[dull magnet]

why would you post a css snippet that doesn't work with Vencord

[shrewd yacht]

That's a way to see it
I see it as working with the web ext
If you want I can delete it

[dull magnet]

yes

[shrewd yacht]

ok

[dull magnet]

make it compatible with Vencord

otherwise posting it here is kinda pointless

🔫

[shrewd yacht]

can I make a PR for the plugin?

[dull magnet]

yes

you could add the data attributes and show on hover

[shrewd yacht]

I think I'll need some help but I'm all for it

shall we use the same class as the extension?

(.pronoundb-pronouns)

[dull magnet]

no

[shrewd yacht]

I'll use the one you proposed then

[dull magnet]

if you use the same class as the extension then you must also guarantee our component is similar to theirs

aka in the same location, same styles, same htlm element

otherwise it's worthless

no compatibility is better than broken compatibility imo

[shrewd yacht]

I can do that

do you want it?

so @dull magnet ?

[shrewd yacht]

well I gotta go eat I'll do it after (usually takes 1h because of my sister)

[amber condor]

has anyone figured out how to create new windows?

[trail ginkgo]

like a popout window?

[amber condor]

yeah

[trail ginkgo]

that would have to be on the electron side

what are u thinking of using it for

[amber condor]

idk yet

i just wanted to get it working for fun

[dull magnet]

the quick css Editor does it

[amber condor]

oh yeah ofcourse

[dull magnet]

I wonder if we could fix quick css on Firefox by making it a popup of the extension instead

I think that's a thing

[torpid agate]

question difference between scripting and coding which one is better

[vague echo]

scripting is automating something
coding is well, the practice of writing code

[torpid agate]

ik the difference

but i meant which one will have more uses

i’ve heard a lot bout js so i was gonna start learning it then i realized i would only know scripting so i was seeing if i should learn one before the other

[opal fern]

coding lmfao

[torpid agate]

any language recommendations

[vague echo]

js and python are fine choices

[opal fern]

depends on usecase

[torpid agate]

idk that yet

just need one to learn

i’ve heard good things about rust

[vague echo]

(do not)

[torpid agate]

why not

[vague echo]

you're gonna have a hard time

[torpid agate]

gonna do it then

[amber condor]

i'd start with assembly

start with python

its easiest and doesnt have weird quirks like javascript

[dull magnet]

I think a lot of people call writing js scripting

dum

python is garbage

js is probably the best starting language

[amber condor]

ok but try explaining === vs == to a complete beginner

they'll just get confused and use both at random

[dull magnet]

nothing to explain, just use ===

[amber condor]

lets not get started on javascript types

[dull magnet]

gives you an easy step up to typescript, a great base for learning other C-like languages, and it's so versatile, once you know js you can basically make anything thanks to frameworks like electron

[dull magnet]

your argument is basically "obscure language feature and quirks you're told to just avoid by all guides make the entire language bad"

[amber condor]

javascript is just weird

if you want to start with javascript you should rather start with typescript

[dull magnet]

no lol

@cedar olive epic idea

In nitrospoof we could https://cdn.discordapp.com/emojis/1024751730237374484.gif?size=48&quality=lossless&name=emojiName

then check for that name query parameter and use it as the emoji name instead of FakeNitroEmoji

[amber condor]

the thing is, those quirks and features can happen on accident and cause massive confusion

because sometimes they cause errors and sometimes they dont

honestly i would only reccomend python or c(++)

[trail ginkgo]

that happens in like every language

you just learn the language over time

[dull magnet]

LOL

js is bad because making mistakes can cause massive confusion
I recommend C

[amber condor]

a mistake can happen as quickly as taking a inputing and doing math

if user input = "4" and the code did input + 2 you'd get "42"

[snow jetty]

I tried C
I gave up

[dull magnet]

how are u gonna recommend C where forgetting to end a string with \0 or accidently indexing an invalid array index can cause undefined behaviour and memory corruption that's impossible to debug for a newbie

[amber condor]

that kind of automatic type conversion is just bs

[dull magnet]

and at the same time say js is bad because you can make mistakes

[snow jetty]

What about rust then lol

[amber condor]

doing rust good is too hard

but tbf they can .copy() everything

albeit inneficient

[snow jetty]

I never tried rust

[woven lion]

you shouldnt .clone()* everything

that's what taking references is for

[amber condor]

i only reccomended C because you were talking about getting into C languages

[amber condor]

we're talking about beginners

[woven lion]

the beginner programming book gives an example of why you shouldnt use .clone() on everything

same with rbe

[amber condor]

C wasnt a good example for safety i gotta admit

[amber condor]

i avoid .clone() at all costs, but who cares if its someon's first file of code?

anyway

"2" * 2 -> 4
"2" + 2 -> "22"

[grim hare]

teaching poor practice first thing is never a good idea

[amber condor]

enforcing perferct code from the start is not a good idea either

[dull magnet]

.

[grim hare]

there's a difference between enforcing perfect code and teaching something that you shouldn't be doing

[dull magnet]

your argument is really stupid sorry

[amber condor]

so is yours

i never get how people like javascript

it sucks ass

[dull magnet]

ok

that's your own preference

[amber condor]

well i wouldnt reccomend rust to a beginner anyway

too hard

[dull magnet]

you're trying to convince beginners of your own opinion instead of trying to be objective

Javascript is bad in my opinion so you shouldn't use it

[amber condor]

javascript is bad because of things x, y, z, etc...

you said javascript is good for beginners because it has alot of capabilities, but so does python...?

[amber condor]

its your preference for liking javascript too

so in your case it would be "Javascript is good in my opinion so you should use it"

[dull magnet]

no

Javascript is objectively one of the best languages if not the best language to pick up as a beginner for a multitude of reasons

high popularity, best odds of getting a job
many great resources online
versatility
absolutely necessary to make web stuff anyway
ease of use
high readability
clike which makes it similar to many other languages (java, C#, etc) and gives you a great base to learn those later
easy step up to typescript later on

and that's just the beginning, there's way more good arguments

the only real argument for learning python first is that it's easy. But it doesn't offer you a good base for other languages, it has crappy performance, cannot be used for as many things, it's mostly good for scripts

[grim hare]

I always wonder how many people that bash on JS have actually used it seriously before, and if they have used it seriously I wonder how bad of code they must have written to consistently run into the "js bad" issues they always complain about

[snow jetty]

same reason why that's what we learned at school here lol

[dull magnet]

yeah, every time someone tries to call js shit it's them going on about things like ==, the fact that adding different data types yields weird results (duh??) and other quirky behaviour, but every up to date resource you can find online explicitly tells you to avoid things like this

there are many actually good arguments against js, but you can find good arguments against any language

[woven lion]

its like me saying

dont use assembly

it doesnt have operators

and control flow is obfuscated

[dull magnet]

Also I think criticising "2" * 2 = 4 is also really funny coming from a python user, you're trying to unironically tell me "2" * 2 = "22" is any less confusing?

[woven lion]

print(len([0x0for x in {1,2,1}]))

valid python

[amber condor]

python is also populair. Depending on the field JS has biggest odds, but so do C# and python in other fields. This shouldnt be reason to learn a specific langauge.
alot of languages have great resources.
python also has huge versatility
not true, you can make a website with python-only (not sure if i'd reccomend that, but if you want to you can)
ease of use? python is the easiest to use
most well-made code is highly readable
going from python to java, the only thing that confused me was using this without this being a function parameter, but that didnt take me long either
your 2nd langauge is always easy, the only thing that's usually the challenge is setting up environments etc, because youre not used to them. learning new syntax is easy, the important thing about your first language is learning programming concepts

[amber condor]

"its easy" shouldnt that be most important for a beginner? (its why scratch exists for young people... its easy?) performance doesnt matter, you can easily create 2d games in them and its fast enough with the abundant amount of c-librarys. python can be used for a fricking lot, desktop-dev, web-dev, creating games...

[amber condor]

this is easy for experencied to say because they know waht to avoid while coding, a beginner does not

[amber condor]

a beginner who just started to learn how to use variables shouldnt be bothered by the fact they have to check their code they didnt accidently use == instead of ===. Ofcourse the person eventuaklly finds out, but == doesnt lead to errors (right?), just weird behaviour.
its odd anyhow that === is common practice for is-equal rather than ==

[grim hare]

do you know the difference between the two

[amber condor]

yes i agree that something like .repeat(n) wouldve been better. But python errors while trying "2" + 2, in javascript it doesnt. my argument is that javascript has a very odd type conversion. Also, that python example doesnt actually do type conversion there. Every type in python has methods that describe how operators function on them, in this case it would be str.__mul__(self, other: int): return "".join(self for _ in range(other), ofcourse written differently because its written in C rather than in python

[amber condor]

one is loose the other is strict

iirc

[amber condor]

so what? as if an beginner would stumble upon how to write hexadecimal (btw for who doesnt know, it actually says 0x0f or x in {1,2,1}, its not an for loop)

[woven lion]

im using your own complaint against you

that was not clear to me the first time i saw that example

and i have been doing python for years

[amber condor]

there are alot of things that arent clear when you first come across them

[trail ginkgo]

yes in every language

[amber condor]

but in this case you wouldnt come across it very fast luckily

you dont just accidently write a hexadecimal while also accidently writing a list comprehension

[trail ginkgo]

assuming we're talking about python __<__3.11

how is "2" * 2 = "22" not confusing for a beginner

[amber condor]

how is "2" * 2 = 4 but "2" + 2 = "22" less confusing than "2" * 2 = "22" for a beginner?

[woven lion]

multiplying strings never makes sense

[trail ginkgo]

the first makes sense because people know what multiplication is

[woven lion]

there's a reason why we have things like string.repeat()

[trail ginkgo]

the latter 2 are initially confusing because you have to learn casting / that * specifically means repeat

[grim hare]

list comprehensions are also a pain to read for beginners, and sufficiently complex ones are a pain regardless of skill level

[amber condor]

"2" * 2, i want string "2" two times?

"2" * 2 = 4, alright so math just works on string i see
"2" + 2 = "22", huh math doesnt work?

[amber condor]

i agree that wouldve been better

[trail ginkgo]

a beginner isnt going to be thinking about types as rigidly as you are

[amber condor]

but python's "2" * 2 is a one-off, javascript does automatic type conversion everywhere which is way worse

[trail ginkgo]

thats kinda the whole thing they have to learn

[amber condor]

that doesnt mean theyre bad, just badly written, just like there are enough people that write poor javascript code

[woven lion]

* 2 works because in js it implicitly casts the string to a number since thats the only logical solution, but + 2 onto a string is a completely different meaning and that, in fact, works in almost* every language as standard

[amber condor]

they shouldnt be bombarded with all the different automatic type conversions from the start

[amber condor]

so py > js?

[woven lion]

/run

"2" + 2

[amber condor]

are you saying "2" + 2 works in almost every language?

[worldly oxide]

@woven lion I only received cs(6.12.0) error output

Cannot open assembly 'out': No such file or directory.

[trail ginkgo]

as a beginner, strict types might feel more bombarding depending on the person, as it requires a lot more keeping track of things

[woven lion]

oops

/run

System.Console.WriteLine("2" + 2);

[worldly oxide]

Here is your cs(6.12.0) output @woven lion

22

[woven lion]

yes

i do

[amber condor]

in what languages lol

[woven lion]

that's c#

lets see..

[amber condor]

what? having to keep track of all the possible type conversions is way more work lol

[trail ginkgo]

you dont have to as a beginner

[woven lion]

/run

pub fn main() {
    println!("2" + 2);
}

[worldly oxide]

@woven lion I received rust(1.65.0) compile errors

error: expected `​,`​, found `​+`​
 --> file0.code:2:18
  |
2 |     println!("2" + 2);
  |                  ^ expected `​,`​

error: aborting due to previous error

chmod: cannot access 'binary': No such file or directory
/piston/packages/rust/1.65.0/run: line 4: ./binary: No such file or directory

[woven lion]

ah right yes that doesnt work in rust

rust is too strict for that

[trail ginkgo]

people are a lot more neuroplastic in general than ur implying

[amber condor]

/run

pub fn main() {
  println!("{}", "2"+2);
}```

[worldly oxide]

@amber condor I received rs(1.65.0) compile errors

error[E0369]: cannot add `​{integer}`​ to `​&str`​
 --> file0.code:2:21
  |
2 |   println!("{}", "2"+2);
  |                  ---^- {integer}
  |                  |
  |                  &str

error: aborting due to previous error

For more information about this error, try `​rustc --explain E0369`​.
chmod: cannot access 'binary': No such file or directory
/piston/packages/rust/1.65.0/run: line 4: ./binary: No such file or directory

[trail ginkgo]

the point where a developer is annoyed because they feel like they have to keep track of types

is the point where you switch to TS

which is not a beginner stance

[woven lion]

/run

public class A {
    public static void main(String[] args) {
        System.out.writeln("2" + 2);
    }
}

wait

is it println in java

[trail ginkgo]

well u can see it there anyways

oh

[woven lion]

/run

public class A {
    public static void main(String[] args) {
        System.out.println("2" + 2);
    }
}

[worldly oxide]

Here is your java(15.0.2) output @woven lion

22

[woven lion]

yes

okay

/run

fun main() {
    println("2" + 2)
}

[worldly oxide]

Here is your kt(1.4.31) output @woven lion

22

[woven lion]

anyway i could get a massive list of these and just keep going

point is yes

"2"+2 does work in loads of major languages

exactly how js does it

[amber condor]

yea but java and c# arent beginner languages lol

[dull magnet]

??????

yes they are lol

[woven lion]

that's a different argument

[amber condor]

they require the user to know what types are from the get go

[woven lion]

and java and c# are taught at a very young age here nowadays

[amber condor]

ffs im just gonna quit it

[dull magnet]

many universities use Java as introduction language

[woven lion]

and

[dull magnet]

[woven lion]

the argument was

https://discord.coffee/9Cf316f.png

[dull magnet]

mine does lol

[woven lion]

to which i presented a couple of major languages that do it like this

[dull magnet]

anyway adding stuff to strings getting auto casted is nice and useful

I use it a lot in many languages

[woven lion]

with the subtext that most major languages do it like this

because it just.. makes sense

[dull magnet]

it just implicitly calls toString() in most languages

[woven lion]

[amber condor]

it doesnt matter anyway, no offense, but every web dev server always fanboys javascript. Most other prgramming servers would disagree.

yes i love TS but damn JS sucks for beginners

[dull magnet]

"https://google.com/search?" + new URLSearchParams({ q: "banana milkshake" })

[woven lion]

i dont like javascript

i like typescript because its not completely insane

like yes there's still the funnies

[amber condor]

it doesnt lol, type conversion shouldnt be automatic, especially if the beginner user doesnt know what types are

[woven lion]

but typescript yells at you

[dull magnet]

(it doesn't, you're still just going on about ur personal opinion)

it doesn't matter much honestly lol

[woven lion]

im gonna be honest you probably use autotype conversion without even thinking about it

[dull magnet]

it's almost like the majority of languages have implicit type conversions because it's incredibly useful

[woven lion]

even rust has implicit types

in some cases

[dull magnet]

Integer x = 1;
``` vs ```java
Integer x = Integer.valueOf(1);```

[amber condor]

yeah between String and &str

[woven lion]

rust doesnt have implicit casts for primitives

[dull magnet]

rust has a ton of implicit casts with into()

[woven lion]

you have to call into() for that

[amber condor]

yeah

[woven lion]

String != &str

[amber condor]

into() is okay

[amber condor]

i know

[trail ginkgo]

its that rust does implicit derefs to match a signature

sometimes implicit referencing

[amber condor]

its kinda funny because i acctually dont like Java either 💀

[dull magnet]

x = 42
y = 0.2

print(x + y)

even python has implicit type conversion

[grim hare]

the funny thing about this statement is that removing automatic type conversions will require the user to learn what types are

[dull magnet]

knowledge of types is also not even such a hard concept to understand

[grim hare]

indeed

[trail ginkgo]

most knowledge is fuzzy as a beginner

[dull magnet]

if anything, it makes it less confusing to the person because you can't accidently pass wrong data to functions

[trail ginkgo]

so even if a beginner is basing their understanding of types from like typeof

then thats still fine

[grim hare]

anyways
how about that [object Object] amirite

[trail ginkgo]

i prefer `${new class {}}`

wait i forgot what makes it just [object ]

`${{ [Symbol.toStringTag]: "" }}`

[amber condor]

knowing what a integer, a float and a string are is different than also having to know which types convert to which when you use them in certain ways

[amber condor]

i think thats the only one tho, but atleast this one makes sense, theyre both numbers

[dull magnet]

wait til you find out strings are also just numbers

[amber condor]

oh yes

imagine if "a"*2 resulted in 194 💀

[trail ginkgo]

true "2" + 2 simply multiplies "2" by 256 first

[amber condor]

someone should write a language where thats the csse tbh

[grim hare]

(it's called c)

(i did it in #🤖-bot-commands but here it is here for channel linear history)

/run ```c
#include <stdio.h>
int main () {
printf("%i", 'a' * 2);
}

[worldly oxide]

Here is your c(10.2.0) output @grim hare

194```

[shrewd yacht]

is there a way to make compact mode work a bit like cozy mode where when someone post two messages in a short amount of time instead of saying the username twice it "merges" them?

[dull magnet]

tbf that's a char which is just a number type

and doing math on chars is really useful

[grim hare]

true

if I used a char array I'd segfault probably though

you love pointers

[dull magnet]

int parseDigit(char digit) {
    return digit - '0';
}

parseDigit('9') // 9 (as int)

[amber condor]

imagine how peaceful and calm this chat wouldve been if this guy didnt ask for beginner language reccomendations 💀

[dull magnet]

it is peaceful and calm

[amber condor]

i was jus jokin

[dull magnet]

'a' * 'a'

🔥

[amber condor]

me when mommy didnt get choccy milk

[trail ginkgo]

a squared

[dull magnet]

/run ```c
#include <math.h>
#include <stdio.h>

int main() {
printf("%c", sqrt('a'));
}```

[worldly oxide]

Here is your c(10.2.0) output @dull magnet

h```

[dull magnet]

h

[trail ginkgo]

LOL

why is h the sqrt of a

that doesnt make SENSE

i would imagine sqrt('a') < 'a'

[viral roost]

why is the square root bigger than the original

exactly

[dull magnet]

Pierre: Js is confusing, use C instead
C: h

no idea why it's h LOL

[trail ginkgo]

[dull magnet]

I didn't know it would be h

I just did it for the funny to see what it would be

[viral roost]

c is fine* until you remember pointers exist

https://cdn.discordapp.com/emojis/859795873915207721.gif?quality=lossless&size=48

[trail ginkgo]

0[0]

(does not fail)

[dull magnet]

probably cause the sqrt of 'a' isn't a whole number

[dull magnet]

tbf it makes sense, it's just indexing into a pointer that is null (the pointer itself is null)

[trail ginkgo]

i think it would fail

[dull magnet]

it will just either segfault or work fine on weird systems

[trail ginkgo]

i was trolling about syntax that u have to learn to make sense

[grim hare]

/run ```c
#include <stdio.h>
int main () {
printf("%lu", (unsigned long) "h" * 2);
}

[worldly oxide]

Here is your c(10.2.0) output @grim hare

8405000```

[grim hare]

I love

c is so wacky it's great

[dull magnet]

/run ```c
#include <math.h>
#include <stdio.h>

int main() {
printf("%c", sqrt('a'));
}```

[worldly oxide]

Here is your c(10.2.0) output @dull magnet

(```

[dull magnet]

/run ```c
#include <math.h>
#include <stdio.h>

int main() {
printf("%c", sqrt('a'));
}```

[worldly oxide]

Here is your c(10.2.0) output @dull magnet

�```

[dull magnet]

/run ```c
#include <math.h>
#include <stdio.h>

int main() {
printf("%c", sqrt('a'));
}```

[worldly oxide]

Here is your c(10.2.0) output @dull magnet

�```

[dull magnet]

?????

random number generator

but how

[trail ginkgo]

.LMFAO

i think htis is buffer overflow

[dull magnet]

how would it be

[trail ginkgo]

i dont know but it looks like it

[dull magnet]

oh because it returns a double

maybe

[amber condor]

stay mad, i already took my words back abt C

[dull magnet]

/run ```c
#include <math.h>
#include <stdio.h>

int main() {
printf("%c", (char) sqrt('a'));
}```

[worldly oxide]

Here is your c(10.2.0) output @dull magnet

	```

[dull magnet]

hmm

/run ```c
#include <math.h>
#include <stdio.h>

int main() {
printf("%c", (char) (sqrt('a') + 50));
}```

[worldly oxide]

Here is your c(10.2.0) output @dull magnet

;```

[dull magnet]

/run ```c
#include <math.h>
#include <stdio.h>

int main() {
printf("%c", (char) (sqrt('a') + 50));
}```

[worldly oxide]

Here is your c(10.2.0) output @dull magnet

;```

[dull magnet]

yeah okay it was cause it returns a double

but that's so strange lmao

[trail ginkgo]

i need to know why

[grim hare]

/run ```c
#include <stdio.h>
int main () {
char * hi = "hewwo owo";
printf("%s\n", hi + 6);

int numbers[5] = { 0, 1, 2, 3, 4 };
printf("%i", 4[numbers]);
}

[worldly oxide]

Here is your c(10.2.0) output @grim hare

owo
4```

[grim hare]

pointers are amazing

[amber condor]

i need a translator

[trail ginkgo]

"its joke"

@dull magnet

/run ```c
#include <math.h>
#include <stdio.h>

int main() {
printf("%i", sqrt('a'));
}

[worldly oxide]

Here is your c(10.2.0) output @trail ginkgo

-1103781416```

[trail ginkgo]

/run ```c
#include <math.h>
#include <stdio.h>

int main() {
printf("%i", sqrt('a'));
}

[worldly oxide]

Here is your c(10.2.0) output @trail ginkgo

-821694712```

[trail ginkgo]

literally random number generator

[dull magnet]

ye

[trail ginkgo]

WHY

[amber condor]

run it 5 million times and check if its evenly distributed

[dull magnet]

it might be because the function takes a double but we pass a char, so it just uses our char + 7 random bytes or smth

but that'd be weird too

/run ```c
#include <math.h>
#include <stdio.h>

int main() {
printf("%f", sqrt('a'));
}

[worldly oxide]

Here is your c(10.2.0) output @dull magnet

9.848858```

[dull magnet]

no it works correctly

[grim hare]

if sqrt takes a double that's definitely what's happening

[steel oriole]

/run

#include <math.h>
#include <stdio.h>

int main() {
    printf("%i", sqrt('a'));
}

[worldly oxide]

@steel oriole

Error

Unsupported language: None
Request a new language

[dull magnet]

just giving printf the wrong format specifier

[grim hare]

o that makes sense too

[trail ginkgo]

yea i just wonder what its doing

[steel oriole]

/run

#include <math.h>
#include <stdio.h>

int main() {
    printf("%i", sqrt('a'));
}

[worldly oxide]

Here is your cpp(10.2.0) output @steel oriole

324471560```

[amber condor]

maybe the next bytes in memory?

[dull magnet]

Well you're telling it to print an int and passing a double

which doesn't make a lot of sense

[trail ginkgo]

i can only imagine it truncates tho

a double is more bits

[steel oriole]

why do you even have that bot
cant people make it say stuff that breaks rules?

[dull magnet]

yes so it just ignores presumably the last 4 bytes

[dull magnet]

So?

[oblique lark]

shouldn't it error if you give it the wrong format specifier

[dull magnet]

then we ban that person

[dull magnet]

this is C......

[trail ginkgo]

yea but truncation is deterministic

[dull magnet]

we don't know errors in C

[trail ginkgo]

not in v10

i think it does in a later verison or something

[dull magnet]

wrong format specifier is usually a compiler earning

but not error

[trail ginkgo]

true

[dull magnet]

how do you pass compiler flags again

/run

[worldly oxide]

I can execute code right here in Discord! (click here for instructions)

Update: Discord changed their client to prevent sending messages
that are preceeded by a slash (/)
To run code you can use "./run" or " /run" until further notice

Here are my supported languages:
awk, bash, basic, basic.net, befunge93, bqn, brachylog, brainfuck, c, c++, cjam, clojure, cobol, coffeescript, cow, crystal, csharp, csharp.net, d, dart, dash, dragon, elixir, emacs, emojicode, erlang, file, forte, forth, fortran, freebasic, fsharp.net, fsi, go, golfscript, groovy, haskell, husk, iverilog, japt, java, javascript, jelly, julia, kotlin, lisp, llvm_ir, lolcode, lua, matl, nasm, nasm64, nim, ocaml, octave, osabie, paradoc, pascal, perl, php, ponylang, powershell, prolog, pure, pyth, python, python2, racket, raku, retina, rockstar, rscript, ruby, rust, scala, smalltalk, sqlite3, swift, typescript, vlang, vyxal, yeethon, zig

You can run code like this:
./run <language>
command line parameters (optional) - 1 per line
```
your code
```
standard input (optional)

Provided by the Engineer Man Discord Server - visit:
• https://emkc.org/run to get it in your own server
• https://discord.gg/engineerman for more info

[steel oriole]

is there a way to replicate BANger using js

[trail ginkgo]

im reading linux docs

[dull magnet]

/run c
-Wall
-Werror

#include <math.h>
#include <stdio.h>

int main() {
    printf("%f", sqrt('a'));
}

[worldly oxide]

Here is your c(10.2.0) output @dull magnet

9.848858```

[dull magnet]

/run c
-Wall
-Werror

#include <math.h>
#include <stdio.h>

int main() {
    printf("%i", sqrt('a'));
}

[worldly oxide]

Here is your c(10.2.0) output @dull magnet

1927604840```

[dull magnet]

huh

no error

Well anyway any sane ide and compiler will warn you

[steel oriole]

is there a way to replicate the effects of BANger using js

[opal fern]

what the fuck does pnpm want from me

[dull magnet]

remove the import

[dull magnet]

you already asked that like 3 times and it still doesn't make any sense

[grim hare]

i'm guessing "command line parameters" means passed to your program, not the compiler

[steel oriole]

how is there not any sense

[dull magnet]

/run ```bash
cat > main.c << 'banana'
#include <math.h>
#include <stdio.h>

int main() {
printf("%d", sqrt('a'));
}
banana

gcc -Wall -Werror main.c

[worldly oxide]

@dull magnet I only received bash(5.1.0) error output

main.c: In function 'main':
main.c:5:14: error: format '%d' expects argument of type 'int', but argument 2 has type 'double' [-Werror=format=]
     printf("%d", sqrt('a'));
             ~^   ~~~~~~~~~
             %f
cc1: all warnings being treated as errors

[dull magnet]

there we go

[dull magnet]

it's already in js so I have no idea what you're asking

replicate the effects how

[steel oriole]

really?

i cant find the code

[dull magnet]

https://github.com/Vendicated/Vencord/blob/main/src/plugins/banger.ts

[proud parrot]

**banger.ts: **

/*
 * Vencord, a modification for Discord's desktop app
 * Copyright (c) 2022 Vendicated and contributors
 *
 * This program is free software: you can redistribute it and/or modify
 * it under the terms of the GNU General Public License as published by
 * the Free Software Foundation, either version 3 of the License, or
 * (at your option) any later version.
 *
 * This program is distributed in the hope that it will be useful,
 * but WITHOUT ANY WARRANTY; without even the implied warranty of
 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
 * GNU General Public License for more details.
 *
 * You should have received a copy of the GNU General Public License
 * along with this program.  If not, see <https://www.gnu.org/licenses/>.
*/

import { Devs } from "@utils/constants";
import definePlugin, { OptionType } from "@utils/types";

export default definePlugin({
    name: "BANger",
    description: "Replaces the GIF in the ban dialogue with a custom one.",
    authors: [Devs.Xinto, Devs.Glitch],
    patches: [
        {
            find: "BAN_CONFIRM_TITLE.",
            replacement: {
                match: /src:\w\(\d+\)/g,
... (14 lines left)

[steel oriole]

that is typescript

[dull magnet]

now what's typescript

[opal fern]

glorified javascript

[dull magnet]

a superset of Javascript

[opal fern]

same thing

[dull magnet]

Also nothing in that file uses typescript features so u can literally just rename it to .js and it'll work

[steel oriole]

i just got 13 errors screaming

[dull magnet]

what are you even doing

why do you need it in js

[steel oriole]

i plan to use it without client mods

jk

port it to bd

because yes

[opal fern]

Bro

Vote ban

[steel oriole]

does vencord have custom plugins support? no

[opal fern]

we dont provide support for betterdiscord

[trail ginkgo]

vote ban

[opal fern]

vote ban

[dull magnet]

what if we make a vencord desktop app

with electron

[opal fern]

i mean

no

not electron

something like gluon

[dull magnet]

no

[opal fern]

why not

you want to torture vencord users?

is that it?

[dull magnet]

realise that maybe there's a good reason electron is so popular

nothing offers as good apis as it

[opal fern]

true enough

but i dont see the benefit over just patching into discord

[dull magnet]

way lower resource usage

[opal fern]

But i like krisp :(

[dull magnet]

for example, does gluon support intercepting requests to modify headers

does it support custom protocols

does it support frameless windows, preload scripts similar to Electron

etc

[opal fern]

But tbf having an app can possibly open doors to stuff like screenshare with audio on linux

[dull magnet]

the same is true for things like tauri

they all offer pretty limited apis

[opal fern]

true enough

[trail ginkgo]

voice engine will be real

although u dont really need a new

desktop app for that

[opal fern]

what the fuck is voice engine

[dull magnet]

oh yeah we need to work more on voice engine

[amber condor]

tauri

[dull magnet]

tauri is garbage

[trail ginkgo]

👎

[amber condor]

idk ive never tried it

[opal fern]

bad performance

use neutralinojs instead

[dull magnet]

.

[dull magnet]

why are you recommending something you've never tried