development

[quartz kindle]

like many other email clients

[marble juniper]

Is this a valid sql statement?

DELETE FROM guilds WHERE EXISTS (SELECT * FROM guilds WHERE id = $1)

$1 is obviously replaced with a real value

[quartz kindle]

yes, but kinda redundant

DELETE FROM guilds WHERE id = $1 would do the same thing

[marble juniper]

Ok

[split hazel]

:o

quite complicated but i can give it a try

give some use for my https://remmacs.org domain

other than scamming

anyways i want to actually fix my databases TCP

so theres no set message length

im thinking make a buffer, check the message length and if it doesnt match the actual length, store it in some variable and wait

of course setting the limit to something generous to not run out of memory

like 100mb maybe

mysql's packet limit is 1gb

but for incoming packets from the database there wont be a limit

but by default there will be a limit of rows returned

if someone decides to request every record and theres a million of them, i hope you have enough time and memory

well same applies to the database tbh

[pine nova]

add rickroll 🏃‍♂️

[split hazel]

that is just cringe

[pine nova]

fr

[split hazel]

topgg mfs dying because someone made a rickroll joke

[pine nova]

i mean this is also an option https://www.youtube.com/watch?v=4ebas1ZvyVI

[quartz kindle]

node uses a rope structure

[boreal iron]

how do actually change the prefered guild locale (for testing reasons)?

[quartz kindle]

imagine you're waiting for a 200 byte message, but you get in 3 tcp frames:

on first frame:
start = rope = {
  size: a
  ptr: aaa
  next: null
}

on second frame:
rope = rope.next = {
  size: b,
  ptr: bbb,
  next: null
}

on third frame:
rope = rope.next = {
  size: c,
  ptr: ccc,
  next: null
}

final structure for the start variable would be ```js
{
size: a
ptr: aaa
next: {
size: b,
ptr: bbb,
next: {
size: c,
ptr: ccc,
next: null
}
}
}

on each frame you follow the next chain and count the sizes, until the combined size surpasses the expected message size (provided in a header in the first frame)

if the size surpasses, then you follow the chain again to assemble the buffer pointers

and then emit the final combined buffer, and reassing the excess bytes back to the first element

this is basically what node.js does in its tcp stream

this way you dont need to preallocate a massive buffer to account for all possible message sizes

[split hazel]

ima just malloc a base buffer size(e.g. 1mb) then check if all data has arrived (check for null character at the end of the buffer size) if not maybe make a linked array and store the buffer pointers in it then at the end assemble them, or just concat them immediately (maybe using realloc)

the obvious problem is how big of a contiguous block can i allocate with malloc

apparently its compiler and OS dependent

[quartz kindle]

1mb should be more than fine

[split hazel]

yeah

[quartz kindle]

how many bytes are you using for the length header?

[split hazel]

for now its only 2 but i can extend that to 4

no biggie

or even 8

but i dont need that much

[quartz kindle]

idk if its possible in c++, but in node its possible to receive tcp frames of 0 size, or 1 size, or anything weird like that

[split hazel]

it gets a bit complicated

the receive function takes in a max size, if a packet exceeds that size the next receive call will give you the other part of the packet

but thats per packet, it doesnt take into account split packets

[wheat mesa]

Just make incredibly unsafe macros to outsmart the compiler 😉

[quartz kindle]

i would do something like this (pseudocode)

buffer = <1mb buffer here>
bytesWritten = 0
tcp.on("message", data => {
  if(bytesWritten + data.length > buffer size) { realloc/extend buffer }
  buffer.write(bytesWritten, data)
  bytesWritten += data.length
  if(bytesWritten < 4) { return } // return if buffer not big enough to get our header
  header = buffer.readUint32(0)
  if(bytesWritten < header) { return } // return if buffer not big enough to contain our data
  ourdata = buffer.slice(0, header); // copy data from buffer
  buffer.write(0, buffer.slice(header)) // write excess bytes to the beginning for next messages
  bytesWritten = bytesWritten - header; // write excess bytes to the beginning for next messages
  emit(ourdata)
})

[split hazel]

c++ has no events silly

(jk)

[quartz kindle]

this is not the most efficient solution, because it involves copying buffers

but its the easiest solution

if you want max performance, you need to implement a zero-copy solution

which is a bit more complicated

[split hazel]

i think i know how i can kind of achieve that

i can always extend the buffer with realloc

how it does that does not concern me but the important thing is it works

[quartz kindle]

it can be done with the rope method i mentioned before

[split hazel]

why cant tcp just be normal and let you decide if you want to split packets or not:(

[quartz kindle]

because it cant lul

tcp has to follow the hardware transmission units

[split hazel]

honestly it'd be easier to just use a library that does this for you

they would probably do a better job anyways

[quartz kindle]

lel

i did it on tiny-discord, well kinda because node did half of it for me

[split hazel]

involves a lot of copying right

at the end

the json parser i use has to have the entire string at hand

cant have it in parts sadly

otherwise that would make it so much easier

i might ask some c++ experts to see what they think

[quartz kindle]

the rope method doesnt require copying to a temporary buffer

it stores the pointers to the original buffers provided by the tcp

can you create a buffer composed by multiple pointers in cpp?

[split hazel]

well here you have to make your own buffer and point to it

[split hazel]

nope

unless they have some std function for it

but then again you cant make a buffer composed of the size of a packet

you have to set a strict size for it

so if you want a buffer to fit that packet size you either have to shrink it (with realloc) or copy

but then what realloc does to achieve that is implementation details

it may or may not copy

though that doesnt really matter considering you can place a null terminator at the end

so you can assemble the entire string at the end

[quartz kindle]

im reading some stuff on zero copy, it seems they use mmap for that

[split hazel]

huh?

[quartz kindle]

the rope method does not require copying the tcp packets, but it does do a copy at the end when the final message is assembled

basically when the entire message is available, a new buffer is created for that message, and the parts are written to it

[split hazel]

yeah i can do that with a linked list no biggie

allocate size of message, write message, push pointer to linked list then assemble the string at the end

and continue until everything has been read

when it reaches the size of the string, it saves the position in the buffer and the pointer currently being read, processes the current string and continues with the next

it is a lot of copying though

because we have the size of the entire packet in the header, you can allocate the exact amount of memory needed for it

mapping separate buffers to appear contiguous would be very good though

oh wait

i thought of a million dollar idea

lets split that real quick

• make the receive function take in 4 bytes (the header size) at first

• allocate a buffer of that exact size and store the pointer in a variable

• make the receive function take in the "remaining" size of the data with the pointer pointing to the next position in the buffer we allocated previously

• subtract the remaining size by the amount of bytes written, if it is zero, send off the buffer for processing, if not repeat previous step

• at last the last step, free!

@quartz kindle thoughts?

I AM A GENIUS

not a single copy in sight

[quartz kindle]

yeah

you just need to also handle excess bytes

[split hazel]

no need

we have the max size parameter

we only take in what we need

i will focus on detecting incorrectly or maliciously crafted packets later

[fickle arch]

tried to install better-sqlite3 by adding it to package.json, deleting the node_modules and it still gave me an error, any clue how to fix this ?
the regular way have the same error.

[vivid fulcrum]

what node version are you running

[fickle arch]

v18.1.0

[vivid fulcrum]

is node-gyp installed?

[fickle arch]

yes

[simple stump]

This is a bit vague, but how can I prevent an user from sending too many requests? I have a few situations I need to fix for a website I’m making using ExpressJS. One is creating accounts/logging in where the user data is stored in a SQLite database. However, the user can slow down the server if they create too many accounts at once, but I’m not sure how I can prevent this. Another is through the search bar, where I send HTTP requests to an API and to another website. I’m planning on only having the user be able to search 5 times per minute (as a start), but I don’t know how to track that if they don’t have an account (staying logged in is also a problem I’m having since I don’t know how to track that).

[fickle arch]

limit request per user ?

[split hazel]

honestly its quite a tricky topic

the easiest way is to do it via a users ip

but if an attacker was motivated enough they can perform a DDOS where IPs dont really matter

[solemn latch]

you're really going to want to have logging in & sessions sorted out before moving forward with ratelimits(for ratelimiting per user)

express also has a few middleware options for rate limits per ip.
For account creation you'd ideally have a captcha used along with ratelimits per ip..

[split hazel]

yeah a captcha

forgot about that

a captcha is a must for ratelimiting registrations

logins too

after that you can rate limit per account

@quartz kindle to make you feel better

for tracking the buffer position im using a pointer to the position rather than an index

because indexes use multiplication on the buffer to get to the destination

and multiplications are slow:)

[feral aspen]

Well... we're going to have fun.

[stable eagle]

what, so fast?

[feral aspen]

[split hazel]

mfs probably changed the language to java as a breaking change

[boreal iron]

Does anybody have a "smart" idea how to parse the JSON and replace the double quotes of the value direct_message (of the key errors) by ` to turn it into a template string

{
    "errors":
    {
        "direct_message": "**Sorry** ${somevar}",
    }
}

[radiant kraken]

they'll call it a 'minor' change though

[split hazel]

lmao

[earnest phoenix]

My Approach would be:

1. JSON.parse(obj)
2. Recursive function ,which loops through keys and creates a string of key value

[cinder patio]

What's the goal here though?

[boreal iron]

nvm need to do it differently anyways

[cinder patio]

so ${somevar} gets replaced with the variable's value?

you'll need to use eval

or Function

[boreal iron]

gonna take my php language parser as example

[boreal iron]

yeah, that's what I want to avoid

[cinder patio]

You can make up your own templating engine which is simpler to parse

and parse the strings

[boreal iron]

yeah I do

[cinder patio]

that's your safest option

[boreal iron]

im gonna replace {{somevar}} by the arguments I pass

[cinder patio]

yeah that works

I'd go with something like @somevar cause it's simpler

Easier to escape as well

[boreal iron]

errr nope, that could get in conflict which strings containing @

double brackets are actually good to go in JS and PHP

[cinder patio]

you can easily detect the escaping character before it though

@

\@

[boreal iron]

yeah ik, I just prefer {{..}}

to be consistent in my projects, you know

damn can't test things atm

Is that the correct syntax for optional chaining for dynamic keys?

object[key]?.[val]

[cinder patio]

yes

[boreal iron]

alr

[lament rock]

Optional chaining for methods is cursed

value?.(args)

[boreal iron]

wdym by is cursed?

[boreal iron]

well works like a charm

    language(localization, group, key, placeholder = false)
    {
        if(!localization || !group || !key) return;
        
        if(localization.includes("en-")) localization = "en";
        
        const language = (this.client.localizations.has(localization)) ? this.client.localizations.get(localization) : ((this.client.localizations.has(locale)) ? this.client.localizations.get(locale) : null);
        
        if(language)
        {
            if(language[group]?.[key])
            {
                let string = language[group][key];
                
                if(Array.isArray(placeholder))
                {
                    const match = string.match(/{{.+?}}/g);
                    
                    if(match)
                    {
                        for(let i = 0; i < match.length; i++)
                        {
                            if(placeholder[i]) string = (placeholder[i].length > 0) ? string.replace(match[i], placeholder[i].trim()) : string.replace(match[i], match[i].slice(2, -2));
                        }
                    }
                }
                
                if(string) return string.replaceAll("{{", "").replaceAll("}}", "");
            }
        }
        
        return this.utils.log("warn", `Missing language record for '${group}->${key}'`);
    }

could probably be compressed a little bit more but who cares

which allows me to call the method and pass my args where ever my vars are in the right scope

[sharp geyser]

what da fuq

Isn't this just doing what i18n does?

[boreal iron]

idk

I don't like libs when I can do stuff my own

[sharp geyser]

True, the code looks a bit weird to me since it is scrunched up :p

[boreal iron]

also I'm replacing placeholders in my language files by vars where I need 'em

[copper cradle]

Ah yes, reinventing the wheel. Classic js

[cinder patio]

that indentation is cursed

[boreal iron]

For example:

// lang json file
{
    "errors":
    {
        "direct_message": "test {{123}} {{afas}} df"
    }
}

const wow = 345;

app.language("en", "errors", "direct_message", [wow])

// test 345 afas df

[boreal iron]

how is that weird?

I'm just using TAB not spaces

[sharp geyser]

I would actually prefer that over i18n ngl

It is way easier to implement

[cinder patio]

I guess not the indentation, just the style... Opening curly brackets on new line 🤮

[sharp geyser]

I assume whatever is in the {{}} will be replaced?

[boreal iron]

yes

[sharp geyser]

what about that [wow] thing?

[boreal iron]

the array does only replace the first placeholder of {{ }} as it has only 1 item

[sharp geyser]

Oh right, so that is what actually gets placed in there and the other 3 are to tell it where to look

gotcha

[boreal iron]

When I pass more items like [wow, 456] the result would be test 345 456 df

I can even pass more items than placeholders exist

that wouldn't do anything

[sharp geyser]

What happens if you nest inside direct_message with another object?

This is just a general example of what happens if you ever do need to nest

[boreal iron]

I don't need that over here, but my PHP equivalent build can lookup one structure deeper

[sharp geyser]

I see

[boreal iron]

For this example I don't need nesting as I already have the key errors and the value direct_message aka. 2nd key

In PHP a much deeper nesting wouldn't make sense, too

because I'm using an INI structure as language file

for example:

[static]
time_years = "year{{s}}"
time_months = "month{{s}}"
time_weeks = "week{{s}}"
time_days = "day{{s}}"
time_hours = "hour{{s}}"
time_minutes = "minute{{s}}"
time_seconds = "second{{s}}"
time_few_seconds = "a few seconds"

[footer]
link_tos = "Terms of Service"
link_privacy = "Privacy Policy"

while in that case when passing an empty item in my replacement array time_years would just be year and not years

that's very simply but effient

and when not passing any item in my array, then year{{s}} would get years

if(string) return string.replaceAll("{{", "").replaceAll("}}", "");

which allows me to replace any part of the final language string by a different string or for example by a var I pass

So I'm just loading my (JSON) language files into a map and access it whenever I need and replace my placeholder if I have to

can't actually think of any easier way to do it

[boreal iron]

well it's all about preferences

[proven lantern]

i like this part

why not optional chain it twice?

[cinder patio]

He needs language again after that

[proven lantern]

language?.[group]?.[key]

[copper cradle]

zamn

why width 8 tabs tho

[proven lantern]

i use the i18next lib, but only have American english

[copper cradle]

then what's the point

[proven lantern]

pluralization stuff

[copper cradle]

you could make a simple template for that

[proven lantern]

i also plan to support more languages

[copper cradle]

then that's ok

[proven lantern]

maybe european english. i'll add random u's to words

legacy english should be supported

[cinder patio]

simish too

[proven lantern]

latin is next

[copper cradle]

also old himalayan

[proven lantern]

i just looked this up. i want 3 sims

sim cards! not the sims 3. oof

[boreal iron]

why would I need the first chain? as language already exists (and is an object) it doesn't matter if group exists or not in JavaScript

if not it will still be undefined

and the statement would fail

[proven lantern]

you could remove that first if statement with the optional chaining

[boreal iron]

ah, that's what u mean

[proven lantern]

i mean things and stuff

[boreal iron]

actually not a bad idea

[proven lantern]

it's a miracle

[boreal iron]

huh?

[proven lantern]

it's a miracle that i didnt have a bad idea

[boreal iron]

regarding our i8n lib

the main feature I need is to replace placeholders

[proven lantern]

eval random strings. what could go wrong?
eval("`${alert('hello')}`")

edge doesn't want to protect me, but chrome does

[split hazel]

the javascript version of a buffer overflow

[proven lantern]

you can use this if you want to make life easier ```js
const i18next = require('i18next');

const fallbackLng = "en-US"
const resources = {
[fallbackLng]: {
translation: require('./en-US/translation.json')
}
}

const initializeLocalization = async (locale) => {
await i18next.init({fallbackLng, resources});
return i18next.getFixedT(locale);
}
module.exports = initializeLocalization;

[boreal iron]

nah... don't wanna use libs

[sharp geyser]

How does supporting multiple translations even work in i18next

[boreal iron]

my selfmade thing actually works perfect and is damn effient

[proven lantern]

when initializing i pass the locale. i can add more keys to the resources object to support more languages

[sharp geyser]

Mmm

[proven lantern]

you live dangerously. if you returned that object on a new line it would fail

return
{
  color: ...
  description: ...
};```

[boreal iron]

https://media.discordapp.net/attachments/483252234579083268/932614697780408330/unknown.png

wut

returning it on a new line, huh?

[proven lantern]

function x () {
    return
    {
        hello: "world"
    }
}```

what do you think this code does?

[austere surge]

nothing

[boreal iron]

I'm not returning an object but a string

the language method can't return an object

[proven lantern]

i'm returning nothing with my code

[boreal iron]

well I wouldn't open that bracket in a new line

[sharp geyser]

This is when you make your own superset of js that supports line breaking

:^)

[proven lantern]

it's a bad habit in js

[boreal iron]

that's the issue when you don't need to put a ; to end your line

[proven lantern]

yeah, ; should be required

[boreal iron]

but I still don't get why I'm living dangerously

just because I generally open the brackets in a new line?

[proven lantern]

correct

let x = 
{
  live: "dangerously"
}
;```

[boreal iron]

I'm only doing it, when ever it doesn't fuck up the syntax, don't worry

[sharp geyser]

Making anonymous functions in js is a good reason why semi should be required

[boreal iron]

also I would never end an clause without a ;

even in JavaScript I add ; to any line

[proven lantern]

fixed

[boreal iron]

to prevent that sort of bullshit to happen

[proven lantern]

i put my ; on a new line too

[sharp geyser]

I do it as much as possible but I just sometimes forget and let prettier do it for me

[boreal iron]

even if it isn't required

well npp doesn't do shit for me

once it belongs to your routines, shit can't happen anymore

[proven lantern]

notepad++. i just figured out what npp is

[boreal iron]

also I would define that inline
let x = { live: "dangerously" };

[proven lantern]

very nice

[boreal iron]

The reason #development message is multi line, is the readability

[proven lantern]

everything should be a oneliner

[boreal iron]

nah

I prefer human readability

[proven lantern]

well it can be multiline, but only one assignment is allowed

[boreal iron]

I could get rid of hundrets of lines if I optimize one liners but I don't like to

as I said earlier

well it's all about preferences

so... just because I'm not as lazy as you to add ; I can live my "weird" style

[proven lantern]

scala has similar problems because of the lookahead + auto inserted ;

[wheat mesa]

tbh I think js should require semicolons like every standard c-like lang does

it saves a lot of headaches

to have strictly defined behavior

[proven lantern]

yep, i'd change the way i write js if they did that

actually i probably wouldn't

this looks betterjs return { hello: "world" }

[boreal iron]

this looks better

pfff

go on

[proven lantern]

well this looks best

return { hello: "world" }

then do ctrl+shift+l

[split hazel]

no

return { hello: "world" };

does

[boreal iron]

awww speedicus saves the day

[sharp geyser]

honestly

[earnest phoenix]

You're trying to make a request with the client such as fetching a user before the client is even logged in

[lament rock]

didn't supply a token it seems

or that

[split hazel]

for discord.js unfortunately yes

you might have some luck trying to login then logout

i had the same issue but forgot how i solved it

hmm

you can try https://npmjs.com/package/discordtools

oh last update 4 years ago

nevermind

and deprecated

[earnest phoenix]

According to the source, seems like you haven't even provided a token to begin with
https://github.com/discordjs/discord.js/blob/df64d3ea382c07e66bc7cc8877ee430206c31d63/packages/discord.js/src/util/Util.js#L230

[split hazel]

i think i know how you can do that

eris allows you to perform api calls without logging in

[boreal iron]

I mean if you dont wanna use simple IPC for example to pass the vote data to your logged in client then why don't you init a new client without any intents and cache?

even if a second connection doesn't really make sense tho

[earnest phoenix]

I don't think you need to be logged in, seems like they're just missing the token completely

[boreal iron]

ahh nvm then

[split hazel]

@earnest phoenix do you happen to know if node has a TCP interface similar to C? i need to receive a max amount of bytes, each time being variable and the event API nodejs has just fires the most bytes it can at you
I had to make an ugly solution with a main buffer being allocated with a set size and when bytes arrive, slice the buffer and copy over the contents to a certain position in the buffer

[boreal iron]

would have been strange if a login would be required just to send an request to the api

but djs is strange as we know

[split hazel]

funnily enough in this case its easier to work with TCP in C than nodejs

if you didnt know the C way of receiving TCP bytes is by
recv(socket_id, destination_buffer_pointer, max_bytes)

[sharp geyser]

I don't think my lib requires to be logged in to send rest api requests either Tho if someone would like to fully test that go for it :^)

[split hazel]

blocking if that isnt obvious

[earnest phoenix]

The built-in node:net module provides such an interface, but may not be as low-level as a C TCP API to work with
https://nodejs.org/api/net.html

[split hazel]

yeah receiving data still requires you to use events

hm

i guess I can implement this myself

but eh

i feel that'll impact performance even more

@earnest phoenix this is what i wrote to work around the limitation https://hatebin.com/vghkkfpmpc

because tcp is a stream API when you send a packet, at the receiving end it can be grouped with another packet or split into two

which makes the stream approach of it terrible to work with

[sharp geyser]

Should I make a Base Channel class and then have classes like TextChannel, DMChannel, etc extend it ? It seems like that would be the most logical as a dm channel will have different properties than a text channel or even a category channel

[proven lantern]

ctrl+shift+l added semicolons for me. 3 key strokes is easier than pressing ;

[split hazel]

you're the type of guy to write a program to rename 5 files

[proven lantern]

i use ken rename

[earnest phoenix]

Looks good to me, that's a pretty good implementation unless you've got some issues with it (?)

[split hazel]

yeah not really just wondering

mainly not happy about the part that you have to copy from another buffer to another when you can directly write to the empty portion of the buffer

but I get thats not really easily implementable because js is a memory safe language

[sharp geyser]

What would be the most logical way to attach the guild the channel belongs to right onto the channel glass? I was thinking of getting it from the cache using the guildid the channel receives but that doesn't make sense as the guild could also not be cached

I was also thinking that to make a channel you must pass the guild with the channel class constructor but that would be rather hard to do as well for events where you either create a channel or update it and then you'd have no access to the guild there either

[split hazel]

i think its a poor idea not to cache every guild

its not much overhead and saves you a lot of headaches

[sharp geyser]

I am already caching every guild

I am trying to figure out how to get the guild the channel belongs to

and attach it to the channel class

[split hazel]

just attach the guild class to the channel?

if every guild is cached that shouldnt be an issue

[sharp geyser]

Mmm, should I even bother caching a channel on channel create?

If I don't have to bother I very well could easily do that

[split hazel]

i would let the user decide

maybe let the user set a channel cache limit

if it reaches that limit it will start overwriting older channels

on every action that involves a channel you should cache the channel discord sends you

[sharp geyser]

Hmmm that seems like a good idea but how should I tell if the channel is an older channel? If it is later in the cache or should I attach a timestamp of when it was added?

[split hazel]

good question

if you dont want to do it through time and instead through order:

[sharp geyser]

Also, how should I update the channel in the cache if it already exists, as the properties of the channel could change

[split hazel]

you should remove the first element of the map (assuming you'll use that for storage) since they're ordered and just set the new channel

be warned anything that holds the channel that isnt the map and is long-lived will result in a memory leak when you try remove them

[sharp geyser]

Wdym by this?

I understand what a memory leak is but not what you mean

[split hazel]

lets say you have a map of channels which holds a class to cached channels

then you decide to cache messages which also hold the channel class

if you remove the channel from the map the channel wont be released from memory since the message you cached still exists and holds it

you'd also have to remove the message to release it from memory

just a warning

[sharp geyser]

So should I also delete message as well?

[split hazel]

too much overhead

that was just an example but you shouldnt worry about that

[sharp geyser]

Mmm, I see

[split hazel]

just make sure your message cache is low

[sharp geyser]

ye

[split hazel]

a few excess channels wont hurt

[sharp geyser]

I plan on setting the limit for a lot of things by default

and then let them decide if they wanna cache more or less

But then I still have the question of how I should retrieve the guild the channel belongs to to attach it to the channel class I tried retrieving the guild from cache with the guild id you get from the channel object but it just returns undefined even tho the guild does exist in the cache

You say just attach the guild to it, I can surely do that when I cache the channels of the guild in the guild create event but not in say channel create or update

[split hazel]

i believe the guild object comes with the event

[sharp geyser]

From what i've seen no

just the guild id

[split hazel]

ah probably to limit bandwidth

[sharp geyser]

{
  type: 0,
  topic: null,
  rate_limit_per_user: 0,
  position: 2,
  permission_overwrites: [],
  parent_id: '957867801119449110',
  nsfw: false,
  name: 'aha funny word',
  last_message_id: null,
  id: '972974453367652362',
  hashes: {
    version: 1,
    roles: { hash: 'kyCmkswUy88' },
    metadata: { hash: '7KBK+FxBI+w' },
    channels: { hash: 'oVyKHFAhsMU' }
  },
  guild_id: '957867801119449109',
  guild_hashes: {
    version: 1,
    roles: { hash: 'kyCmkswUy88' },
    metadata: { hash: '7KBK+FxBI+w' },
    channels: { hash: 'oVyKHFAhsMU' }
  },
  flags: 0
}

all you get back

[split hazel]

you must be doing something wrong when retrieving the guild form the cache

code?

[sharp geyser]

import { ChannelData, MessageCreateData, PermissionOverwrites } from '../../../util';
import { BaseClient } from '../../BaseClient';
import { ChannelType, RESTPostAPIGuildChannelJSONBody, RESTPostAPIGuildChannelResult } from 'discord-api-types/v9';
import { Message } from './Message';
import { Guild } from './Guild';

export class Channel {
    public client: BaseClient;
    public id: string;
    public type: ChannelType;
    public guildId: string;
    public position?: number | null;
    public permissionOverwrites?: Array<PermissionOverwrites> = [];
    public name: string;
    public topic: string;
    public parentId: string;
    public guild: Guild;
    public nsfw: boolean;
    constructor(client: BaseClient, data: ChannelData) {
        this.client = client;
        this.id = data.id;
        this.type = data.type;
        this.guildId = data.guild_id;
        this.position = data.position;
        this.permissionOverwrites = data.permission_overwrites;
        this.name = data.name;
        this.topic = data.topic;
        this.parentId = data.parent;
        this.nsfw = data.nsfw;
        this.guild = this.client.guilds.get(this.guildId)
    }

    public async send(data: MessageCreateData): Promise<Message> {
        return await this.client.rest.post(`/channels/${this.id}/messages`, data);
    }

    public async create(
        guildId: string,
        data: RESTPostAPIGuildChannelJSONBody,
    ): Promise<RESTPostAPIGuildChannelResult> {
        return await this.client.rest.post(`/guilds/${guildId}/channels`, data);
    }
}

[split hazel]

have you tried logging the guild id?

[sharp geyser]

yes I get back the guild id as expected

but this.client.guilds.get(this.guildId) is undefined

[split hazel]

have you tried fetching the guild with the ID hard coded?

are you sure its even cached? can you print your whole guild cache

[sharp geyser]

Printing the entire guild cache returns the base store with the guild in it, and fetching it with a hard coded id works fine and returns the guild

https://www.toptal.com/developers/hastebin/hicixiqido.typescript

response

[split hazel]

im stumped then you have to be doing something wrong though

hard to debug when you arent in front of the code haha

[sharp geyser]

I can push to my repo and you can test whenever you have time

if you want

[split hazel]

sure

i can have a look at it there

[earnest phoenix]

Same

[sharp geyser]

pushed

[split hazel]

where are you instantiating the channel?

[sharp geyser]

In the webhook shard, on guild create and channel create

I make a new channel to convert the data returned to my class's data so I can keep the methods with the data

[split hazel]

cant see it at first glance so i'll just clone and run it

[sharp geyser]

Alrighty ty in advance <3

[split hazel]

windows :~

wsl time

[sharp geyser]

lol

[split hazel]

oh dear i dont have a typescript compiler installed 💀

should be fine tho

[sharp geyser]

typescript is one of the deps for it so it should have it locally installed in the project

[split hazel]

what is the config layout?

its not included in the repo

[sharp geyser]

it is just a config file with an export of TOKEN

[split hazel]

always a good start

[sharp geyser]

Yea no idea what happened there

[lyric mountain]

windows: git is not a recognized cmdlet, go fuckin google how to solve this
linux: git is not recognized as a command, perhaps you could try installing with apt install git

[sharp geyser]

I thought you were being rude then realized you are mocking windows

:^)

[copper cradle]

That's not a windows thing eiter

That's a shell thing

I'm seriously considering installing wsl

[lyric mountain]

windows mocks itself

[split hazel]

i see the problem

im so kind i'll fix it for you

[sharp geyser]

lmao thank you :)

[split hazel]

@sharp geyser https://github.com/MistyyBoi/strife/pull/1 :D

[sharp geyser]

Also question would it not be better to just
this.guild = guild ?? this.client.guilds.get(data.guild_id);

[split hazel]

sure

my personal preference is ? :

[sharp geyser]

Either way works tbh

I do like the look of ?? tho

https://www.toptal.com/developers/hastebin/muvamufave.yaml Mmm, is it just me or does this look bloated?

[split hazel]

might be just you

imposter syndrome

[sharp geyser]

Mmm well doesn't really matter as long as it works :p

What should I work on next 🤔

[split hazel]

more useful connection errors

a few times my connections timed out and i got a basic "request failed"

[sharp geyser]

Mmmm

[sharp geyser]

So i was looking at discord.js lib and they do something rather confusing

They have a Channel class right which is just a basic channel not really a text, dm or category, etc. Then they have GuildChannel which extends that, then BaseGuildTextChannel which extends GuildChannel then TextChannel which extends BaseGuildTextChannel. What is the point of all this extension?

[lyric mountain]

inheritance

[sharp geyser]

Well yes but could they not have done it simpler?

[lyric mountain]

let's say all channels have id, name, permissions, category, etc

instead of repeating that same code in all channels, u just make a superclass with common properties

ur not really supposed to use the superclasses themselves

[sharp geyser]

Yea i know this much, but why the whole long winded extension

Why not just have a Super class then extend that super class for the things you need

Like
Channel (Super Class) -> GuildChannel -> TextChannel

[lyric mountain]

because not all guildchannels are textchannels

and not all channels are guild channels

[sharp geyser]

Yes, which is why GuildChannel is a class in its own where all guild related channels extend it

[lyric mountain]

ah, u mean there's a class between those two?

[sharp geyser]

Yes Channel is the super class, GuildChannel is the class that extends it and has basic properties a guild channel should have and TextChannel has the properties of what a guild channel should have + text channel properties

[lyric mountain]

that's how it's supposed to be

in jda it's that way

[sharp geyser]

Yes, but they add a fuck ton of what seems to be useless stuff

I am sure it has its uses but it seems to be overly done when it could be much simpler

[lyric mountain]

maybe the Base classes aren't classes but interfaces

[sharp geyser]

Looking at the difference between BaseGuildTextChannel and TextChannel seem to be nothing different despite a few things you could probably add to TextChannel in itself

[lyric mountain]

well then idk, base doesn't really make sense tbh

[sharp geyser]

They seem to apply a interface properties to it using applyToClass or smth

So it extends GuildChannel and implements TextBasedChannel

[lyric mountain]

in jda it's like ```
┌ PrivateChannel
Channel ┤
└ GuildChannel ┐
├ TextChannel
├ VoiceChannel
├ ...

[sharp geyser]

I don't really see the point in having a BaseGuildTextChannel class tho it seems like they took an unneeded extra step

[lyric mountain]

is that what u mean it should be?

[sharp geyser]

Yea, that is what I was planning to do

But I wondered why djs did it so weirdly

Wait what djs uses something from sapphire

:^)

[lyric mountain]

sapphire?

[sharp geyser]

the djs command framework

used to be klasa but the devs had falling out so the ones left made sapphire

[quartz kindle]

because ThreadChannel exists

so they needed a common class for text channels that are not text channels

[sharp geyser]

Mmm, I guess that makes sense yea

I still feel there is a better way to do it tho but 🤷‍♀️

they are the pros

[lyric mountain]

In jda it just extends textchannel

[quartz kindle]

also soon there will voice text channels

[lyric mountain]

What

Why?

[sharp geyser]

I mean tbh isn't a thread channel just a special text channel?

[dry imp]

yea dope

[quartz kindle]

discord is adding text to voice

like a stream thing

twitch etc

[dry imp]

its better than to comment about the stream on a irrelevant channel

[lyric mountain]

D.py devs going crazy for that update

[quartz kindle]

djs literally broke

when it started receiving message events in voice channels

[lyric mountain]

It already started?

[sharp geyser]

v10 I think is a part of that

[solemn latch]

Yeah

It's kind of an annoying feature for mods

[lyric mountain]

And they still didn't finish the join agreement part

The screen that appears when u join a server, it has many "To be added" features

[quartz kindle]

yes its been a few months now

i had to upate djsl a few times to fix it

[fringe pendant]

Quick question: is server count setting via the API broken right now? Just noticed my bot hasn't updated in a while but as far as I can tell the API calls are going through

[boreal iron]

takes a while to update in your browser since the page is cached

[fringe pendant]

It's like 50 servers behind though, so I'm guessing something else is wrong then. I'll look into it thanks

[boreal iron]

well which duration since the last update are we speaking about?

[fringe pendant]

Estimate would be about a month

[boreal iron]

ow

is the count in a different browser the same?

[lyric mountain]

Mine is working fine, did u see if ur not getting any http error?

[fringe pendant]

I'm gonna try to set it manually and see if it updates, and if it does then I guess the library I'm using might have an issue

[lyric mountain]

What lib is it?

[fringe pendant]

It's the java library, version 2.0.1

java-sdk

[boreal iron]

I remember a few reports recently about the java lib iirc

[lyric mountain]

I use the java lib

[boreal iron]

might be an issue with it

[fringe pendant]

I tried updating to 2.1.1 but for some reason I can't seem to get it off maven

doesn't exist

[lyric mountain]

But tbh, I'm quite sure I'll not be using it anymore on my reworked bot

Doing direct http requests is easy enough to remove the dependency

[fringe pendant]

yeah it updated via manual API call, so something broke in the bot seems like

[lyric mountain]

Also iirc, the java lib has some vulnerabilities (uses org.json) reported on maven central

[fringe pendant]

nice

alright guess I'll implement it myself as you suggested then, thanks for the help

[lyric mountain]

Btw, before you go, I recommend using apache http for doing the requests

And moshi to parse data into json

[fringe pendant]

Noted, although I literally just update server count and nothing else. So... not too much work being done

[lyric mountain]

org.json is a terrible lib (it was created as a proof-of-concept) to use anywhere

Mostly because it lacks any optimization

[fringe pendant]

I usually use OkHttp, not a fan of that either?

[lyric mountain]

That's for requests, not parsing data

[fringe pendant]

oh yeah sorry I misread

[lyric mountain]

It's fine too, I like apache's bcuz it's very flexible

And since it uses factory pattern, u can set default settings for every request

[fringe pendant]

Yeah true, I've used Apache before for other things but for whatever reason I started with OkH with this and never changed. Does enough for what I need

[lyric mountain]

Oh, btw, do u use log4j?

[sharp geyser]

I gained something but I lost another as well

[fringe pendant]

I don't use log4j no

[sharp geyser]

I managed to split the channels up according to their type so now the data is properly typed with the methods that are meant for that channel but the channels cache on the client is only storing one channel

[lyric mountain]

Oh ok, I was going to warn u not to use earlier versions

Due to infamous exploit

[fringe pendant]

yeah don't worry I'm aware of that lol

[copper cradle]

I'd maintain the go package but tbh I haven't seen anyone using go here

[wheat mesa]

So I find that I am once again going down the hole of writing a minecraft mod, does anyone know a good way to avoid global state? A lot of the features I need to support rely on settings provided by the user, and in previous mods I've done this with using a bundle of static variables, but I know it's a bad practice

[lament rock]

pretty sure the mod loader provides a way of r/w of a config file

would have to read your mod loader's docs

[wheat mesa]

docs??!! on forge??! impossible

I'm using 1.8.9 forge, there's basically no resources for it

Because they refuse to support anything older than like 1.12

[lament rock]

surprisingly, they actually made some docs for later versions

yeah then you might be shit out of luck unless you find a mod's source that does what you want to do

[copper cradle]

then don't write mods for 1.12

/j

[lament rock]

Unironically, forge needs more love in the performance departments like fabric gets

or one of the mod loaders just gives up on trying to update and have it be obsolete

or both

[sharp geyser]

@wheat mesa Make your own mod loader

:^)

[boreal iron]

Is there actually some sort of process environment vars in node which can be an array?

[lament rock]

env vars are all strings

[boreal iron]

Ok let's say I'm exporting this:

module.exports =
{
    name: "evil",
    description: "very evil shit",
    do(param1, param2, lang)
    {
        // do some evil shit
                lang.something...
    }
};

[lyric mountain]

Let's pretend u didn't use "do"

[boreal iron]

I'm getting my params in the do() method right from where I require the module and execute do()

let's say lang is another class or object I wanna use on the module's name

I don't actually wanna define lang as global var

And I can't import/require it from anywhere else

[lyric mountain]

Can't u just require it inside the scope?

[boreal iron]

hmm... I feel like the only option is to export lang as module itself

and import it over here

in my case lang is a method in class

[lyric mountain]

The entire concept concept of having to export stuff in js so it's discoverable in other files is weird to me

In java u just write the class and that's it

[boreal iron]

yeah I don't have that trouble in PHP, too

[lyric mountain]

If u don't want to import u just write the full path to the class

Js is the only lang I see this behavior

Maybe that's due to its browser origins

[boreal iron]

Importing something that is I'm already passing as argument is weird

[lyric mountain]

Oh wait what, I just noticed that's a param

That still requires importing?

[boreal iron]

well the example was stupid

lang actually is a class, let's call it app and lang is it's method

[lyric mountain]

That should've work, if it's passed as a param u shouldn't need to import the class

[boreal iron]

// some other file
class app
{
    language() ...
}

// some very other file
module.exports =
{
    name: "evil",
    description: "very evil shit",
    exec(param1, param2, app)
    {
        // do some evil shit
        app.language(...);
    }
};

wait

the class app get's exported

module.exports = app;

can't I actually export language as well?

with a second export like module.exports = app.language;

which I can then import in my very, very, very other module

[lyric mountain]

I mean, that's what I mean

U shouldn't need to import every single method in a class if ur passing an instance of it as an argument

[boreal iron]

YES I'M PASSING THE ENTIRE INSTANCE

BUT

I'm passing it to exec(param1, param2, app)

[lyric mountain]

Yes?

[boreal iron]

I wanna use it ouside that scope already

[lyric mountain]

U mean outside of exec?

[boreal iron]

module.exports =
{
    name: app.language(...),
    description: "very evil shit",
    exec(param1, param2, app)
    {
        // do some evil shit
        app.language(...);
    }
};

[lyric mountain]

Ah, that

[boreal iron]

yeah that's the issue

[lyric mountain]

Well, that's something I don't think any lang can achieve

Forward reference in this case

[boreal iron]

that's what I mean, I need to export that method separately

which I can then import in this module

to NOT import the entire class (again)

[lyric mountain]

U could, tho, make it a class instead of a (whatever that structure is called in js)

Then instantiate the class with an instance of app

And use it in constructor to build ur command

(assuming that's a command)

[boreal iron]

err nope, just was a neat example I could copy real quick

hmm js classes are weird

[lyric mountain]

Like new SomeExample(app)

Then use app however u want inside that class

[boreal iron]

importing the class isn't the issue tho

[lyric mountain]

I mean, in ur example it'd not work since you'll be getting a static method instead of a reference's

[boreal iron]

but I don't want to initiate it another time in my module (since the entire class is already available as argument)

[lyric mountain]

2 different contexts

[boreal iron]

yeah

[lyric mountain]

You either use a class so u can store the reference in a broader scope or don't forward-reference it

[boreal iron]

that's why I asked

how I can do that in JS

[lyric mountain]

U can also set a placeholder in name

[boreal iron]

huh

you're actually right, lol

[lyric mountain]

Like, idk if those fields are all final, but can't u do something like this.name = app.language(...)?

If anything, replace that value where u import it

[boreal iron]

i could check if the module name is a string or an object, when it's an object I could pass my language key and value

[lyric mountain]

Ye

[boreal iron]

module.exports =
{
    name: { group: "something", key: "something else" },
    description: "very evil shit",
    exec(param1, param2, app)
    {
        // do some evil shit
        app.language(...);
    }
};

[lyric mountain]

For my commands I set their names like cmd/some_command, then when printing the name I grab that entry from the locale files

[boreal iron]

damn I haven't thought about that actually

actually a good idea tho

never thought I would appreciate tips from a Java user, but nice one

[lyric mountain]

[boreal iron]

there you go, good girl, take a cookie 🍪

man... localization stuff is expensive

[lyric mountain]

How are u doing it?

[quartz kindle]

i gave up on localizing timezones in my api

[boreal iron]

lol

NOW YOU FUCKING ARRIVE

[quartz kindle]

the user can localize them themselves if they want, fuck it

[boreal iron]

lmao

[quartz kindle]

theres zero consistency

[boreal iron]

just do an autocomplete list of timezones he can pick one of

[lyric mountain]

/localize inmyhouse please supply a valid timezone
/localize im_my_house please supply a valid timezone
/localize here please supply a valid timezone

Ah fuck it this bot is brokenz

[quartz kindle]

lmao

[boreal iron]

well... by having a default locale (en) and return the user/guild locale where ever I have access to that property

pretty simple

[lyric mountain]

No, i mean, are u using files + cache?

[boreal iron]

i'm importing the lang file, yeah

[lyric mountain]

Oh, no cache then

[boreal iron]

so it's in the cache until I reload it

or should I say memory

[lyric mountain]

Like, in my schema I get from the cache, if it's missing it loads the entry from the target locale file

So the cache auto-loads itseld basically, and keeps the entries cached for 30 mins before discarding it

[boreal iron]

I'm not letting the user pick a language etc.

[lyric mountain]

This way I don't have the whole file loaded, only the requested lines

[boreal iron]

I only wanna support the description_locales

[lyric mountain]

Still, don't load the entire file

What format are u using?

Don't say json

[boreal iron]

technically not, no but why should I care?

the import process happens once when the client being initiated

[lyric mountain]

Wasted memory

[boreal iron]

we're just speaking about a few bytes of ram

or megabytes

wtf who cares about RAM in 2022

[lyric mountain]

Definitely megabytes

[boreal iron]

I can fucking assign up to 198GB to the VM if I want to

but yeah I know what u mean

not going for microptimizing here

a few mb of ram as base load doesn't matter for me, tbh

[lyric mountain]

Remember fixing i18n-related issues gets exponentially harder the longer a project is

[boreal iron]

that's why I'm using my own simple system I used to show earlier this day

which is also capable of replacing placeholder in my language string if I wanna replace something with a var for example

[lyric mountain]

I faced the i18n demon eye-to-eye, which is part of the reason I'm rewriting my entire bot

Every single visible string is going to be localized

Ngl it's easier when starting again, but my file already surpassed a hundred lines

[quartz kindle]

im never gonna localize my help command, fuck that

[boreal iron]

yeah I'm currently rewriting my bot, too, making anything more dynamicly to be able to reload basically the entire app while keeping alive the gateway conection

[quartz kindle]

it would be like writing a book

[boreal iron]

Yeah I made the accident to translate the entire API docs

starting at line 170 to 271

doesn't sound much, but is 2s scroll time on 4K

[lyric mountain]

https://github.com/OtagamerZ/ShiroJBot/blob/rewrite/src/main/resources/locale/lang_en.properties

My current locale file

And I didn't even start the most verbose commands

[boreal iron]

I really dislike JS hasn't a native INI parser like php does

I prefer the INI format

over JSON

so much

[lyric mountain]

I like properties

No nesting, each line is a key-value

And u can stream the file to load the entry instead of loading the entire file

[boreal iron]

#development message

basically like this one

[lyric mountain]

Ye, we use that format at my work

[boreal iron]

while categories (as key) are sexy to work with

since you have to write same key, for example rarity/ in any line

I can just use the category

[lyric mountain]

I also use properties bcuz intellij has a i18n manager that shows the data side by side (edit all locales at the same time)

[boreal iron]

the ini file which will be parsed as array/object and I can easily access them by the key (category) and val (names)

if JS would only have a native INI parser

[lyric mountain]

The editor

[boreal iron]

danke schon

I don't trust those google translations

[lyric mountain]

Now you know how to say thank u in german

[boreal iron]

panzerfaust

now you know how to say good bye

ok enough of that

we were talking for an hour and I didn't make any progress

I feel like Tim

[boreal iron]

Ok fuck localization

The API isn’t ready for it yet

Once a command is registered/edited with a description localization for example, it does loose when I restart the app

Fetching the command from the API and boom localization is null again

[sharp geyser]

This whole localization gives me a wonderful idea

[boreal iron]

But the command registration promise does return the new and correct application command object

With the localization

But it’s not accepted by the API for some reason also doesn’t throw an error

My gosh

Fuck Discord

Ok I can now understand Tim

[sharp geyser]

Wait what are you talking bout at this point

[boreal iron]

Let’s say you register an application command at the API

You can now add a localization for the command name and description

While I don’t care about the command name, the description might be worth to translate

As well as any command option descriptions

[sharp geyser]

Yea

[boreal iron]

So when registering that command or editing it, a guild command in my test, I can add the description_localization property to the command object which is an object to

{ locale: string }

{ "en-US": "us description" }

The API will respond successfully with new application command object I’m logging

But nothings happens

When I restart the app and fetch the guild commands the description localization is null again

[sharp geyser]

Mmmm, maybe localization isn't fully out yet?

[boreal iron]

And even without a restart and clearing the cache no translation is happening

[boreal iron]

Idk I thought it is

Nvm

#discord-news message

It isn’t

[sharp geyser]

Yea, the api for it is, but client wise I don't think it is

[boreal iron]

Looks like it requires a special discord app built

[boreal iron]

Well if so the API wouldn’t respond with success but doesn’t save my localization

Looks like it’s not ready yet

[sharp geyser]

Yea, well hopefully soon it will be :p

[boreal iron]

Regarding the time they posted it and I can’t even register the localization strings I highly doubt

Enough bot rehauling for me right now

Back to web development

The api isn’t ready for anything they introduced lately

As well as the missing features

Meh makes me sad

[fickle arch]

I still don't know how to solve this, already do all the troubleshooting on the github page, and I also have node-gyp either.

[fallen shard]

danke schon

[split hazel]

why does node gyp always have issues

it is singlehandedly the most unreliable thing in node

[stiff lynx]

async function cercaCanale(id, client) {
  let canale;
  let contatoreRicerca = 1;
  do {
    canale = await client.channels.cache.get(id);
    console.log(`Canale cercato ${contatoreRicerca} volt*`);
    contatoreRicerca = contatoreRicerca + 1;
  } while (!canale);
  return canale;
}

Is this good or not?

[cinder patio]

no

First off, you're only getting the value from the cache

you're not fetching it

[stiff lynx]

the second problem is if the bot doesnt have the permession to see that channel

and it will loop forever

[cinder patio]

Second, fetching a channel multiple times makes no sense - you'll always get the same result, unless like you said suddenly a person gives you permission to the channel.

But that should be detected in an event

[stiff lynx]

I did this bcs I get a lot of errors bcs the bot doesnt find a channel

[cinder patio]

Third, it's API abuse

This will only make it worse

[stiff lynx]

I want to say that this is only a test, In my hosted bot I have

    let canale = await client.channels.cache.get(id)
    return canale
}```

lol, still in cache

but without cicles

I had also fetch

but still errors

async function cercaCanale(client, id) {
    let canale = await client.channels.fetch(id)
    return canale
}

This should be better, right?

[cinder patio]

yes

[stiff lynx]

with fetch is 100% sure to get the channel?

(with the right permissions)

[cinder patio]

yes

if it's the right ID and you have permissions to see it

[stiff lynx]

after this I have a question about bot verification, I have only the guild intent for a welcomer.
Is it enough to get the verification or It will be denied?

[pale vessel]

if you don't have permissions to see it you could fetch all channels of that specific guild and find it by id

[earnest phoenix]

@pale vessel your status is great haha

[stiff lynx]

How can I fetch for a guild?

[eternal osprey]

 for(const key in roledatabase){
      m = newMember.guild.roles.cache.find(r => r.id === key);
      newMember.roles.add(m)
      console.log(m)
     }```always returns missing permissions

While i am pretty sure the role hierarchy is set, alongside admin perms of the bot.

[earnest phoenix]

You're bots role is above the key role correct?

[dry imp]

@harsh nova scam

[lyric mountain]

...do-while

[stiff lynx]

💀

I'm trying to add a role in another server for an user that uses a command, but I'm doing somethin' wrong and I don't get what lol

const NPW = await client.guilds.fetch("972430091927818280");
const utente = await NPW.members.fetch(interaction.user.id);

const aggiuntaRuolo = await NPW.utente.roles.add(ruolo);

[lyric mountain]

users cant have roles

ah nvm, ur fetching it

why dont u just get interaction.member?

[stiff lynx]

works for different server?
The command is being used in server X, and I want to add a role in the server Y

ik that is something strange, but is what I need 😮‍💨

[lyric mountain]

nope, for that u do need to fetch the member

what's the error ur getting?

[stiff lynx]

const aggiuntaRuolo = await NPW.utente.roles.add(ruolo);
Cannot read properties of undefined (reading 'roles')```

[lyric mountain]

either the member isn't cached or they arent in the server

[stiff lynx]

It's very strange, bcs the console.log of utente and the role are correct

I have the data of the Y server, and not X

nvm, I've solved it 'NPW.' was not needed

[subtle fiber]

is js better than python?

and why

[dry imp]

they are better for different purposes

[boreal iron]

True

Best py purpose I know is uninstalling it

[hazy vigil]

Best py purpose is to make a bot to make fun of js

[dry imp]

worst js purpose is installing it

[hazy vigil]

True

[dry imp]

worst php purpose i suppose

[boreal iron]

lol compared to shit like JS or py php is far ahead

[dry imp]

cope with it fake

[boreal iron]

Quiet py user

[cinder patio]

lmao are you saying php is better than JS

That's objectively false 😦

[hazy vigil]

Make a discord bot in Holy C

[hybrid cargo]

I just made a bot with CSS

[dry imp]

guys remember rules, no lang war

[boreal iron]

Ofc that’s the only reason we’re here

[hazy vigil]

Why?

[boreal iron]

CSS very powerful…

[dry imp]

best prog lang imo

along with html

[boreal iron]

Aye

[subtle fiber]

bro

[lyric mountain]

https://accodeing.com/blog/2015/css3-proven-to-be-turing-complete

[subtle fiber]

html is not lang

[carmine summit]

havent you heard of ✨brainfuck✨

[dry imp]

it is a lang

[cinder patio]

bro really? Why

[boreal iron]

lol

[dry imp]

my fav lag

[boreal iron]

It even has the L for language in its name

[subtle fiber]

👍

[hazy vigil]

Make a brainfuck discord bot and i will be give you all my money xD

[lyric mountain]

I mean, u could just write a bot as normal and use the js > brainfuck converter to make the sourcecode become brainfuck

[boreal iron]

We’re moving straight of the topic that py sucks

Concentration please

[carmine summit]

well that doesnt change the fact that it has a maximum amount of characters you can use which is 30000

[lyric mountain]

for what?

[carmine summit]

i mean you could easily reach it by a few lines of code in js

[lyric mountain]

?

like, code doesn't have size limit, it's just that after a certain amount of lines most text editors will refuse to open the file

[boreal iron]

Windows notepad, best editor

[lyric mountain]

30k characters is easily reachable in normal code

[lyric mountain]

actually the only editor that'll open such big file is nano

[hazy vigil]

I will try to reach it in my bot

[lyric mountain]

no like, really, it's just 3k lines of 10 characters

or 300 lines with 100 chars

[boreal iron]

Tbh npp is really good too

In opening large files

[carmine summit]

what is this then?

[cinder patio]

I wonder how many people will be able to run my 1TB javascript project

[boreal iron]

[lyric mountain]

no idea, but no code enforces a max limit

[quartz kindle]

wat

[lyric mountain]

heck I constantly see people who surpass 30k chars all the time here

by putting everything inside index

[hazy vigil]

Well if it's about disk space, i can handle that, i have 13,75TB disk space

[carmine summit]

who uses multiple files in 2022 anyways

[boreal iron]

True

50k line in one file is what the kids like to see nowadays

Especially on mobile

[lyric mountain]

and with no indent

[boreal iron]

lol yes

[quartz kindle]

remember that line counter utility thing

that also says how much your code is worth

(kinda stupidly exaggerated tho)

[lyric mountain]

my code was worth some tens of millions

[quartz kindle]

lmao

[carmine summit]

wait, loops and functions are a thing?

[lyric mountain]

?

[fathom sonnet]

Bruuuh... I use sticky notes

[lyric mountain]

https://www.quora.com/Is-there-a-way-to-use-Microsoft-Word-as-a-code-editor

[quartz kindle]

nah, code in powerpoint

[cinder patio]

nah, code with powerpoint

[lyric mountain]

https://www.andrew.cmu.edu/user/twildenh/PowerPointTM/Paper.pdf

[subtle fiber]

how to get bot developer role

[lyric mountain]

develop a bot ||and have it accepted on top.gg||

[subtle fiber]

so im close

i already waiting a week

[lyric mountain]

can bitfields store negative values?

ah nvm, I can just use 255 and offset by 127

[split hazel]

thats cursed dont say that ever again

[tribal crow]

Hello hello! I need some help! For a long time I've tried to sort and rank different objects in my database, but i cannot figure out a way to do so. I've managed to sort it from highest to lowest, but how to i get the rank of each item? Any help is appreciated :)

[quartz kindle]

any kind of ranking requires some sort of criteria, so you want to rank them according to what?

[tribal crow]

ok so

lets say i have 5 of these:

{
user: string
money: number
}

then i want to rank them based on the ammount of money they have

[cinder patio]

Use sort

[tribal crow]

I have

[cinder patio]

assuming the objects are in an array

[tribal crow]

Yeah

[cinder patio]

Well then show your code

[tribal crow]

var times = UserProfiles.length;

var Sorted = UserProfiles.sort((a, b) => b - a);

const rank = UserProfiles.map(v => Sorted.indexOf(v) + 1);

let UserProfiles = await UserSave.find({ guildID: interaction.guild.id });

[cinder patio]

And what do you want do to? You're already sorting them

[tribal crow]

yeah,

but lets say i have an array with the numbers

[10, 5, 100, 29, 52]

I want to sort then from high to low

and then give the highest number the rank #1

the second highest #2 etc

[cinder patio]

So you want to sort them in an descending order

[tribal crow]

Yes

[cinder patio]

Then you do a - b instead of b - a in your sort function

[tribal crow]

alr

[cinder patio]

Also this could be simplified, by a lot:

UserProfiles.sort((a, b) => a - b).map(v => v + 1);

[tribal crow]

so by logging rank, i get [ 1, 2 ]

[cinder patio]

Pretty sure UserProfiles.map(v => Sorted.indexOf(v) + 1); is wrong because v is the value, not the index

[quartz kindle]

the rank is literally the index of the item in the array after sorting

[tribal crow]

and i get that how?

[quartz kindle]

who is the first rank? sorted[0]
who is the second rank? sorted[1]
what rank is user ID? sorted.findIndex(x => x.id === ID) + 1

[pale vessel]

and indexOf takes a value?

[tribal crow]

yeah, i understand this, but the point is that i dont know how many items there will be

depends on how many users there are

[quartz kindle]

it doesnt matter?

the ranks dont change

[tribal crow]

yeah, but im using a forEach() to add a field to an embed for every user

[quartz kindle]

still doesnt matter

[tribal crow]

-.-

i know the ranks don't change, but when adding fields i do it like this

embed.addField(`#${rank} <@${element.userID}>`, `\`❂ ${element.bank + element.wallet}\``)

right?

so i dont know what rank is added

[cinder patio]

Show more code

show your entire command

you're overcomplicating this

[tribal crow]

i am

my english sucks

https://sourceb.in/cfRTv8Wn8z

[quartz kindle]

Sorted = UserProfiles.sort((a, b) => a.money - b.money);

Sorted[0]

[wheat mesa]

Also, camelCase please

[cinder patio]

const sorted = UserProfiles.sort((a, b) => a.money - b.money);
for (let i=0; i < sorted.length; i++) {
   const user = sorted[i];
   embed.addField(`#${i + 1} <@${user.userID}>`, `\`❂ ${user.bank + user.wallet}\``)
}

Simple as that, sort the profiles, then loop through each one. The index of the profile is the rank, so you just add + 1 to it so it starts from 1

Also don't put await in front every function call

[tribal crow]

may i ask why? never understood what it does, just been told to use it

[cinder patio]

https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Promise It's used to await async code - you don't need to put it in front of calls to sync code

[tribal crow]

alr

[tribal crow]

ok this works, but is it possible to limit how many times it loops?

ideally 10 times

[quartz kindle]

for (let i=0; i < 10; i++) {

[tribal crow]

alr

but since the amount of users is lower than than 10 rn, i get an error because user.userID is undefined 8 times

[quartz kindle]

for (let i=0; i < (sorted.length > 10 ? 10 : sorted.length); i++) {

[tribal crow]

lovely

thank you so much

[tribal crow]

ok

so now i need to find the rank of one specific user...

how would i do that?

[cinder patio]

sorted.find(u => u.userID === id)

[tribal crow]

that gives me the object of that user

[cinder patio]

findIndex

[tribal crow]

how do i use it? Tried multiple things, none work

[cinder patio]

sorted.findIndex(u => u.userID === id)

[split hazel]

just found an infinite freelance developer customer glitch

lie about having a masters degree in cs

tbf hobbyist developers know way more than degree/master students

because you're actually motivated