development

[unreal mist]

im using bn.js

only issue is this keeps happening

[dusk linden]

transcend space and time, build a supercomputer to run 25 million 962 thousand 864 instances of bn.js at once, use all of those instances of bn.js to stack into a superclass called “eternalnumber.js”, max that out then you will have enough XP.

[unreal mist]

apparently i cant go bigger than 300k levels :/

[dusk linden]

why’s that

[unreal mist]

bc it just does this whenever i run the command

[dusk linden]

is it due to it taking too long to calculate?

you only have 3 seconds to acknowledge the interaction

[unreal mist]

could be

ill have to check

[dusk linden]

if it is u can defer the reply then edit it

that’s what makes it show thinking

iirc

[unreal mist]

i forgot it wasnt already deferring tbh

[trail sapphire]

how u guys make bots?

[deft wolf]

We code them

[radiant kraken]

[neon leaf]

weak numbers

[warped sierra]

I've finally seperated my buttons into categories 😭

there is currently 50 button interactions

[simple cradle]

GG!

Im to lazy too xd

[warped sierra]

i ended up rewriting my handlers for menus and modals too so they all get sorted

[warped sierra]

my latest update to my JsonDB util now has response times of as low as 0.03ms

caching is a beautiful thing

-# from my testing, thats 70x times faster than before

and i've improved batch operations to be up to 74x times faster

[neon leaf]

when just use pg

sorry to disappoint but its sometimes even faster than 0.03ms

[warped sierra]

-# dont ruin this for me

nah, this is just for small commission projects that will never have any actual use for a dedicated db

since i often take bot commissions for private single community bots, the need of an actual database is non existent

so in those cases i just use my json db

in a full production environment i obviously wouldnt use it because, personally not a fan of it

its just one of those boredom projects that i slowly improve whenever i feel like it

[queen needle]

Just use string.db

it's 10/10

[unreal mist]

part of the assignment is create an app in c++ using max of 100 lines or less

1 line instead of 3, reckon theyd be mad or proud 😎

[warped sierra]

A little bit of both

[unreal mist]

may get my progress tutors opinion

im curious how much respect hed lose for me over this

[warped sierra]

Last time I got told x lines or less for a js challenge, I obfuscated it so it was one unreadable line

[queen needle]

I love to do that

[unreal mist]

see i wouldnt do it if it was multiple lines, but for the ones that are single lines i would:

like this would stay as it is

    if (choice == "today") {
        time_t t = time(nullptr); /* Grabs the current time, in seconds */
        tm* now = localtime(&t); /* Converts the time in seconds so its readable, into yy/mm/dd /h/m/s */
        month = now->tm_mon + 1; /* Allows month to start at 1 instead of 0, for 1 - 12 format */
        year = now->tm_year + 1900; /* Calculates how many years it has been since 1900, for accuracy
    }

but this? 1 line should qualify

bool leapYearChecker(int year) { 
  return (year % 4 == 0 && year % 100 != 0) || (year % 400 == 0); 
} 

bc its still readable as one line in the function than multiple lines, when its in a single line

[warped sierra]

If its readable as one line then it should definitely be one line

[unreal mist]

exactly

[warped sierra]

But if appears confusing as one line then it shouldn't be a one liner

[queen needle]

if it can be one line it should be oneline

someone can learn the art of reading the horrible one liners

[warped sierra]

almost anything can be one line if you really try

[queen needle]

the best is the javascript function to check if a letter is uppercase

[warped sierra]

for my next trick, my 1k line file will become a one liner

[queen needle]

const $=_=>$>_

[warped sierra]

then I lose my jsdocs tho

[queen needle]

or the for loop to get the nth fibonnaci number

for(let i = 0, l = 1, c = l, a = l; i < n; (i++, c = l + a, a = l, l = c))

[quartz kindle]

the curse strikes back

[lyric mountain]

you can also use the formula so you dont have to iterate all the way to it

[queen needle]

wdym

[lyric mountain]

there's a formula for the nth fibo number

F(n) = (φⁿ - (1-φ)ⁿ) / √5

[queen needle]

ah

I didn't know that for some reason

[keen hazel]

Mobile Coders Be Like

[warped sierra]

looks like the classic chatgpt output

[unreal mist]

i feel proud, guess how many lines?

[warped sierra]

63

[unreal mist]

nope :P

[warped sierra]

2

[unreal mist]

definitely not

its more than 50 but less than 100

even if i made that single line function into 3 lines, its still less than 100 but more than 50

[warped sierra]

I guess 50 -> 100

[unreal mist]

its 85 lines if i kept that function in 1 line, 87 if not

[warped sierra]

oooo nice

im just setting up testing for my npm package rn

[unreal mist]

it checks for leap years, highlights the day you select if you pick your own date, and todays date if you select today

[warped sierra]

date mod 4 = 0

[queen needle]

can you try october of 1582 in the program

and for day do 4

[unreal mist]

it goes as far back as 1970 iirc but hang on

oh wait thats to calculate correct days etc

[queen needle]

thats wrong 🙁

[unreal mist]

It’s using a different calendar that’s why

[queen needle]

It was when we switched calendars

that's why we had to make the days up

[unreal mist]

yeah so my checker wont be 100% accurate for dates before 1970

but for a simple, max 100 lines to work with its pretty damn good

[queen needle]

i think so

[queen needle]

awesome!

[unreal mist]

think its a bit iffy for before 1970 but oh well

at least the last 125 years work

[unreal mist]

Only thing I haven’t actually added is a fallback check to make sure you can’t enter a year below 1970, month below Jan and a day below 28, 29, 30, 31 (depending on year and month)

[neon leaf]

🔥

[pearl trail]

what nvme is that

[neon leaf]

2x8tb Samsung pm9a3 series

in hardware raid 1

[unreal mist]

Is JavaScript interpreted or compiled?

[wheat mesa]

A mix of both

[unreal mist]

Oh, sources online say different ones and I put about one of them since I was confused

[wheat mesa]

Compilation to machine code happens at the time of interpreting the bytecode

It’s called JIT compilation

But it’s not compiled in the typical sense of “Source code -> Assembly/Machine code”

[unreal mist]

Give me like 5 mins to finish walking home and I’ll send what I’ve put bc I feel like it’s incorrect now 😭

[radiant kraken]

pretty small docstring

all this for a single method

[unreal mist]

whats the ///, i rarely see three of them together

[unreal mist]

im guessing im partially correct then

based on what youve said

[wheat mesa]

The stuff about threading is incorrect. JavaScript’s runtime is inherently single-threaded

The rest of it is mostly subjective. Saying that one is easier to read is mostly an opinion

[unreal mist]

from what i read online, it talked about js being multi threaded

[wheat mesa]

JS internally has multiple threads it manages, but your code runs all on a single thread

[unreal mist]

oh

[wheat mesa]

Worker processes are “kind of” multithreading, but that’s kind of a cheat. As far as I know, there is no real multithreading within a single javascript process

[neon leaf]

yes, js is concurrent

not parallel

[radiant kraken]

docstring comments in rust

otherwise // are ordinary comments

[unreal mist]

ohh

[neon leaf]

yes

I have 0 docs 🔥

and 80gb target folder

for a 20mb bin

[radiant kraken]

me when #general message

[pearl trail]

now only god knows what my code does on my project that haven't been worked on for months

[neon leaf]

to be fair it does take 4minutes to compile on a 9900x

without deps

[radiant kraken]

nah i'm sure you will soon!!!!

[pearl trail]

xDD

[radiant kraken]

i'm sure you left very informative comments

you're super organized

[pearl trail]

nah

comments are for weaks /j

[dusk linden]

highly relatable (i don’t need comments i can read guys) (this is a joke)

[radiant kraken]

don't you have TODOs?

or FIXMEs?

[pearl trail]

lol just todo yeah, but most probably i will refactor everything again, since i do that project once a year, and my skill increased during the gap

[radiant kraken]

really shows how awesome you are!

[pearl trail]

you tooo

[radiant kraken]

i could never be more proud of you

you're the best!

[unreal mist]

so essentially js can only use one thread at a time but can process multiple tasks at the same time as compared to python only doing one task at a time?

[neon leaf]

it can process them concurrently (so it feels like its doing multiple things at once)

because things like io isnt done on your code thread

so while thats waiting, your code can continue in other places

[unreal mist]

so im along the right lines with that now then

[neon leaf]

yes

[unreal mist]

ty

[neon leaf]

but im not sure if python has async too nowadays

but never used it so idk

[unreal mist]

im not sure, i do know it has multi thread capabilities, just locked behind GIL

[loud canopy]

ok fuck it

im hosting database and everything on arch laptop

and connect from main machine

like im tired bro

this pc has period

24/7

everyday a new issue

[wheat mesa]

No

Python and JS have similar event loop architectures

Async python essentially follows the same idea as async JS, it’s concurrent but not parallel

[unreal mist]

im so confused

[wheat mesa]

You are right in that JS can only use one thread at a time

But you are wrong in saying that Python is only doing one task at a time

Python and Javascript both have something called the event loop

The event loop allows them to handle things concurrently while waiting on blocking code to complete (usually I/O or network)

However it is all handled on one thread for both languages

Python and Javascript are what’s known as “concurrent”, but they are not “parallel” (these terms are often confused)

[unreal mist]

so this is wrong
Due to Python only being able to run one thread at a time during interpretation, it prevents several tasks occurring at once. Despite having multithread capabilities, Global Interpreter Lock prevents Python from this, which in turn causes scalability to be a much harder process

[wheat mesa]

I mean it’s right and wrong. Concurrency is almost like simulated parallelism

TECHNICALLY, they are only doing one thing at a time. However, both languages can switch to different tasks when waiting on other blocking tasks, which gives the “feeling” of multiple tasks happening at once

If that makes sense

[unreal mist]

so it becomes a similarity rather than a difference

[wheat mesa]

Yes

[queen needle]

This is probably a stupid question, but when languages are single threaded or only use one thread does that refer to like the actual CPU and only having one thread on that?

[dusk linden]

pancakes and waffles!!

[neon leaf]

simplified yes

in the real world the kernel of your os moves around stuff all the time

[frosty gale]

it does get really confusing i dont blame you, but in general in async everything is coordinated by a single thread so you can run things that execute quickly on that single thread but offload tasks that can take a while (aka blocking) onto different threads

so that single thread basically processes all requests/messages and as soon as it hits something that could take a while it pushes that off to another thread and works on something else it can execute quickly while that processes in the background

people call this single threaded and defend that stance with their lives but i disagree

[queen needle]

Wait so if it's pushing to another thread then how is it single threaded?

[neon leaf]

because those threads are only for io

not for actual code

[frosty gale]

yep, actual code runs on that single thread but anything that can take ages (aka a network request, disk read, etc) is pushed onto another thread which handles that

takes ages is oversimplified here but its good enough

[neon leaf]

in comparison it does take ages

[frosty gale]

so in theory it is single threaded but in reality its still multithreaded because holistically that process has multiple threads doing something, at least from the operating system/outsider standpoint

[unreal mist]

im still really confused, so it doesnt get locked by gil? 😭

[lyric mountain]

event thread isn't a separate thread from main

if it were then you'd not be able to lock your entire application by having an infinite loop inside a promise

[unreal mist]

huh

this is hurting my brain so much

[wheat mesa]

Threading is very complex and the terminology is confusing

[unreal mist]

i have 4 points to do that id noted down for this report, if i end up over the 1000 and can remove the threading part without ending up lower than 1000 words ill leave it out for now 😭

i want to talk about c++ and c# still for it

[frosty gale]

you can compare async to a company ig, ceo runs the company (lets just say so for this example), ceo wants X done at the company to do Y but he cant do it himself because it will take too long and block doing anything else, so he offloads it to an employee and continues with the next task, when that employee is done he lets the ceo know and then when the CEO has time he does Y

not sure what you mean by gil though

[unreal mist]

gil is global interpreter lock

everywhere online talks about python being limited bc of that

[wheat mesa]

That’s just the idea of not having multiple threads executing bytecode instructions at once

It’s another way of saying the interpreter is single threaded

[lyric mountain]

that'd be actual threading, async would be akin to doing a group project but you only have one pen

[neon leaf]

except if you are using rust tokio 🔥🔥🔥🚀🚀 which runs all code on multiple threads secretly

then you have 4 pens for example

but yes, the pen example is better

[vivid fulcrum]

modern c# nailed multithreading/concurrency

the task system is such a pleasure to work with and it ties really nicely into async syntax sugar. one of the best designs out there imo

[unreal mist]

wait

idk how to describe what i want to say hang on

so js runs the code, using things like settimeout() you can delay tasks briefly

thats what you mean right

the task is still done

just later?

[wheat mesa]

Agreed

[lyric mountain]

uhh...this Medium analogy is the simplest example I could find:

Imagine you’re at a restaurant:

• The Chef (Call Stack): The chef prepares one order at a time. If a dish takes a long time to cook, the chef moves it to a separate station (like the oven or timer) and starts working on the next order.
• The Waiter (Event Queue): The waiter keeps an eye on all pending tasks (like dishes in the oven) and brings them back to the chef once they’re ready.
• The Manager (Event Loop): The manager ensures the chef only works on one task at a time and keeps the workflow smooth.

https://medium.com/@ignatovich.dm/the-javascript-event-loop-explained-with-examples-d8f7ddf0861d

[unreal mist]

so the bit about the waiter is similar to how setTimeout() works then 😭

im trying my best to understand

[lyric mountain]

setTimeout would be an oven (or whatever appliance you choose)

[unreal mist]

but settimeout could have stuff inside to do though

but only when its ready

[lyric mountain]

there

whatever is ran in setTimeout is the dish

the chef wont touch that task until it's "ready to serve"

which would be the timer expiring

[unreal mist]

yeah which is what i was trying to say, it still serves the dish, just later

[lyric mountain]

yes

in a threaded scenario it'd be like having multiple cooks each doing their own stuff

[unreal mist]

this whole thing really shows how stupid i am

[lyric mountain]

nah it's pretty confusing until you get it, cuz for the programmer they all appear identical

the advantage is that it's lightweight and doesnt (cant) have concurrency issues, the disadvantage is that you cant really work in parallel and a busy promise can hang the whole thing

on the other hand threads can execute several tasks at the same time, but they're expensive to spawn and can run into concurrency issues if you're not careful

the cost is why some langs have both native and virtual threads, the latter working similar to js' event loop

[wheat mesa]

yeah threads are super confusing

[queen needle]

And also you understand most of these things when you know a language but maybe not to the same level, just through my use of JavaScript and how it's worked when I've used it I understand most of this in context but could not explain it

[loud canopy]

i love waking up just to see discord updates breaking my UI

didnt touch this stuff for months

"oh now the formatting doesnt fit"

[solar kayak]

How can I update the count of servers on my bot page?

[unreal mist]

i think my tutor may hate the way its written

[lyric mountain]

holy hell

[unreal mist]

This is v2, had to keep it 100 lines or under and with v1 and the validation checkers it was too squished and hit about 95 lines, so I broke it up, added more loops and it’s now 75

[lyric mountain]

please save the number of days to a variable so u dont do up to 31 array accesses 😭

[unreal mist]

its already in one

[lyric mountain]

the for condition

int days = daysInMonth[month - 1];

else you'll access that array every single iteration + 1

[unreal mist]

my for loop makes sure the day starts at 1 though doesnt it

[lyric mountain]

yes, but the final condition check (the one that exits the loop) will also access it

[unreal mist]

oh?

[lyric mountain]

it needs to check if it's false yk

[unreal mist]

so anywhere with this int daysInMonth[] = { 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 };

needs to be this? int days = daysInMonth[] = { 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 };

[lyric mountain]

no, like, your array is the correct way to approach it (tho month days vary but anyway)

just save the current month's length to a variable

so u can do for (int i = 1; i <= days; i++)

[unreal mist]

they do vary which is why for leap years i have it adding a 29th day, and i map correctly the number of days :P

[lyric mountain]

oh

[unreal mist]

i also have it calculate the correct day of the week that the year starts on

the validation checker only says its false if its outside the ranges, since im using while loops

probably a bit hard to read bc zoomed out

[wooden ember]

anyone know how my bot is offline but never crashed?

the js program is still running, no errors. But the bot is offline... bit of a confusing scenario.

[stark kestrel]

hard without any information, context/code

the thing i could see is that something blocked the websocket pings to be replied and then closed the session

depending on the library or your code, it may be considered to be still running without logs or errors

[wooden ember]

Had to go do something and I just came back. The issue seems to have fixed its self.... I guess we will never know since there are no log messages and I did nothing. The bot is pretty much exclusively on servers that are fully dead so I can be pretty sure that no one has interacted with it too.

[cedar crescent]

@shell tundra any issue with my bot??

Beast Named Bot

[shell tundra]

investigating a report

[cedar crescent]

What report??

Some users trying to remove my bot from topgg

[shell tundra]

we don't disclose that

[cedar crescent]

If you want any proof then ask me

I have

[cedar crescent]

I know who reported

But i have proofs

[prime cliff]

[clever tundra]

[warped sierra]

Pro tip: switching database midway through a project sucks.

[dusk linden]

😭

[warped sierra]

I have spent 2 hours rewriting models and just updating my interactions like buttons and menus

I haven't even got to updating commands

[lyric mountain]

you dont keep ur DDLs saved?

I keep them all inside my resources (tables aren't here since I use JPA) to recreate if needed

[dusk linden]

I just have a module for database functions so i can change them all in one place

[solemn latch]

👀 whats a good method of learning c++?

[dusk linden]

I would tell you if I knew c++ 😭

[solemn latch]

Yeah 🙁

O

I'm trying to get back into embeded programming

[dusk linden]

good luck with that!!

which languages do you know?

[solemn latch]

Javascript and a bit of python, not much D:

[dusk linden]

still good!!

[solemn latch]

I'll learn more over time ^-^

I havent had much reason to learn other languages

[dusk linden]

I mean pretty fair

I just learn whatever im using at that time, haven’t been bothered to learn C

[solemn latch]

I've thought about learning C before C++

idk if its worth it these days though

[dusk linden]

yeah me too, but I just feel like c has to be learned since its like fast lol

I already know too many other languages to not know C 😭

[neon leaf]

buy an arduino kit, do stupid shit, keep doing stupid shit, ✨ now you are decent at c++ and can continue

[solemn latch]

Oh I have a p4 😄

[dusk linden]

LOL

[neon leaf]

mhm, I personally prefer arduino boards for c++ fun (a lot easier to work with imo)

[dusk linden]

I just got one of them!

[neon leaf]

because with most arduinos, code size and efficiency is a real concern

so you need to optimize right from the start

[dusk linden]

helpful

[solemn latch]

I got excited when I saw the p4, so now I want to get back into it.

I heard the p4-m3 will be like $7~ on full prod

and the h264 encoder means you can stream video in high quality

should make great cat cameras for when I'm on vacation

[dusk linden]

yeah that sounds great

[unreal mist]

yall, what else should go on a github readme

[dusk linden]

https://tenor.com/view/jmk-gif-22808446

a tribute to beans

[queen needle]

Lua?? What for?

[unreal mist]

the game im making, and another side project i have

but if ive used/using/familiar with it, i added it

it shows what i know

[queen needle]

Ohh that's your profile

I thought that was toastuku GitHub and I was confused

[unreal mist]

nono

[dusk linden]

it looks nice though Noah, good job on it!

[unreal mist]

i made this on a test repo so when i post it as my profiles readme i wont have so many changes from testing

[queen needle]

Yeah way better than mine, Idk what to put on my GitHub readme so I just made a simple one and let it be

But it looks great!

[sour cairn]

dank

[crystal wigeon]

is there a way to restart only a specific sharD? like my bot died in some servers but its working in others

[lament rock]

Well. It depends on the lib. All that theoretically should matter is detecting when your WS connection dies

then reconnect

[loud canopy]

man i love looking at pokemon battle system formulas and

https://tenor.com/view/technology-copy-word-copy-paste-gif-13993009

[warm yarrow]

Does anybody know how to report? Server numbers to top.gg

[delicate zephyr]

create a ticket via the #staff-tickets channel

https://img.pyrocdn.com/DhVosgMQ.png

[sharp geyser]

not enough arrows

cant see what you are referring too

[grim inlet]

🤔

[unreal mist]

sudoku?

[pearl trail]

pls do cls every update.
sincerely, guy who used to have daily assignments to make cli apps for 2 months

[unreal mist]

I’ll be having it update properly when I get better with it, so that it isn’t sending over again 🙂

[stuck reef]

so cute

[clever tundra]

ooooh that looks nice

[stuck reef]

ikr ikr

got some spacing issues other than that ive almost finished my bot

[unreal mist]

Will update how it goes when I’m more familiar but I’ll probably do some at uni during the class I have programming in 😛

[warm yarrow]

I meant that as in an API request to give this server number to top.gg to display on your panel

[clever tundra]

Oh

@warm yarrow https://docs.top.gg/docs/API/v0/bot#bot-stats this???

[lyric mountain]

also use box drawing chars

oh oh and colors!

[unreal mist]

I plan to add colours too

I just need to learn more first 🥺

[restive ledge]

Hy

[unreal mist]

I do know how to add colour, I just want to get the rest figured out first

[lyric mountain]

Ye that's the easy part

[crystal wigeon]

im stuck using djs. how can i monitor my shards and restart em?

[wind bison]

one of the weakest documention seen so far for topggpy....

💀

[stark kestrel]

it's probably the easiest API to use because of its simplicity

[wind bison]

no example mentioned

[wind bison]

lavalink wrapper is much easier to setup than topggpy

discord.py is much easier than topgg.py

[deft wolf]

And it is developed by more people I guess

[wind bison]

they have more complex functions as well which is still easy to use

cus they have proper documention for each and every single letter with usage examples everywhere

[deft wolf]

And that's why they need examples. Topgg.py is so simple it doesn't need them

[wind bison]

its so simple for those who know py properly

not so begginer friendly

[deft wolf]

Discord bots are not beginner friendly in general

[wind bison]

nope

they are

thats the issue right everything around dc is begginer friendly due to proper documentation

when you dont provide one ..... it becomes tough to understand how their api works exactly

documentation was so weak infact ai was not able to do it

@deft wolf

💀

[deft wolf]

Ah yes, AI

[dusk linden]

just what all programmers should rely on

[wind bison]

creating a basic and advance examples might help for begginers just like lavalink and discord.py

[dusk linden]

feel like it’s the other way around “the documentation was so strong even AI can do it”.. I mean ai really isn’t that competent

humans can program better than ai

[wind bison]

why to do hard work when i can do smart work.... small things like creating a discord bot .... AI. can do that ... when it comes to office work... AI cant help

[dusk linden]

uhm whatever you say man

[wind bison]

work smart not hard

[dusk linden]

working “smart” to u is using ai to create bug-riddled code?

if you say so

[wind bison]

working smart according to me ....get the written stuff from AI ... have a look what AI has provided ...if all good ... deploy

write manually only when needed ... (work smart)

[dusk linden]

it is not smart work lol

it’s not even work

anyways can’t be bothered arguing about a brain dead topic

[wind bison]

better than writing manually for hours...complete time waste

0 productivity

get the work done within few mins is smart work

[dusk linden]

it’s not even related to what I said originally, I told you AI is NOT better than humans at coding, and the top.gg api is really not hard to understand.. if you aren’t competent enough to do so, then that’s on you. it isn’t related to whether using ai is a good idea.

[wind bison]

umm yes i implemented it .....had to write manually tho.....for small things like 1+1 i dont wanna write ...i ask AI to get it done for me .... i wasted my time on this just because they dont have good documentation

[dusk linden]

unrelated to what i said

not arguing about this

goodbye

[wind bison]

your words : if you aren’t competent enough to do so....
it is related mate ..i

[dusk linden]

!!

read pls

[wind bison]

i will ask ai to read

😂

[earnest phoenix]

Hi, I have a bot long time ago but i think it has been deactivated recently for some reason. How do I reactivate it again?

[clever tundra]

Define deactivated

[earnest phoenix]

it’s not usable anymore even though it says it’s online

wait no it’s offline

[stark kestrel]

do you really need examples for something that has the most basic usage? lmao

[unreal mist]

Oh yeah, if I’m not mistaken tho there’s only about 8 colours?

[clever tundra]

Then just restart it

[earnest phoenix]

but how? I think i lost my server to restart my bot

[clever tundra]

Then make a server, upload the code, and start it

[earnest phoenix]

i did but doesn’t work

[clever tundra]

Any errors or anything

[earnest phoenix]

no it just says offline but I have already uploaded my zip folder

[clever tundra]

Does the code actually start or is it just nothjng

[earnest phoenix]

It’s just nothing i think

[clever tundra]

Can the code run from inside a zip file because last I checked running files inside a zip isn't possible

[earnest phoenix]

wait actually I have archived the zip file

but still nothing

[clever tundra]

What language

[earnest phoenix]

python

[clever tundra]

So can the .py file be accessed? So not in the zip etc
And how are you trying to host this

[deft wolf]

Some "bot hosts" prefer this way. You upload zip file and they do the rest

[earnest phoenix]

no it’s not in the zip file and I use Kata Bump to host it. I used another server before but for some reason it had been closed down and i couldn’t use it anymore

[clever tundra]

So how are you launching it

[clever tundra]

Never seen that in my life

[earnest phoenix]

I use KataBump now

[clever tundra]

No lile

Luke

Like

How do you run the file

What are you doing to run it

What command

What button

How

Why

[earnest phoenix]

I have the Rerun.bat file?

[clever tundra]

How are you opening that file

How are you launching that file

[solemn latch]

this feels so sketchy 👀

is this real?

[clever tundra]

I don't know-?

Truman show aah

[solemn latch]

also doesnt this violate discord TOS to do this

[earnest phoenix]

i dont understand your question😅

[clever tundra]

Ok so

For your bot to work

You have to run the file

Katabump or whatever just keeps it online and running

You have to do smth to start it

What is that smth

[earnest phoenix]

Oh i created a server there and uploaded my folder. Then there’s a start button and I clicked it

[deft wolf]

What doesn't break Discord's ToS these days

[clever tundra]

How does it know what file to run

Like did you put the file in the settings?

So if it's in a folder it should be starting:
foldername/filename.fileextension

[earnest phoenix]

there’s an option for you to choose when you create a server

[clever tundra]

Was that choosing language or file

[earnest phoenix]

language

[clever tundra]

Ok so it's not that

Is there a server settings option

[earnest phoenix]

yeah

[clever tundra]

Click it

What do you see

[earnest phoenix]

SFTP Details, Debug info, Reinstall Server, Change server details

[clever tundra]

Can you send a screenshot

I would delete that from here if I was you

That has important info which can make people delete your files so I wouldn't keep that here

[dusk linden]

how to get hacked

tutorial

[dusk linden]

why would you ask for a screenshot then 😭

[clever tundra]

And none of that is particularly important, you basically need to try and find where you tell it what file to start, there should be somewhere you enter it but since I haven't used that platform before I'm not entirely sure then

[clever tundra]

I didn't know the stfp stuff would be just sitting there

Normally it's a click to reveal thing

[dusk linden]

if it wasn’t gonna be there then why would you need a ss

like, what were you expecting to gain if everything was converted

[clever tundra]

I was expecting somewhere where you enter the file path to start

But nevermind

[dusk linden]

ah alr

[earnest phoenix]

You mean this?

[clever tundra]

Yes

So if app.py is in a folder, it should be structured in a way where it is like:
foldername/app.py

[swift barn]

what happens when you start the server though ?

[clever tundra]

They said that it's just nothing

[earnest phoenix]

should this be in the host section?

[swift barn]

thats like not possible

[clever tundra]

I don't know man 😭

[clever tundra]

The what section?

[swift barn]

bro should try the python file locally see if it runs

[earnest phoenix]

like here?

[clever tundra]

If app.py is inside Fawkes, it should be:
Fawkes/app.py

[earnest phoenix]

@swift barn

[swift barn]

thats the mistame your meant to unarchive that folder

in the project root directory

[earnest phoenix]

i did

[swift barn]

so that app.py is here

[clever tundra]

It can run inside a folder

[swift barn]

did not

[swift barn]

no it cant

[clever tundra]

Bruh

Not inside a zip

[earnest phoenix]

wait so i should change my file name?

[clever tundra]

But a regular folder yes

[clever tundra]

No

[swift barn]

then change this to folder/app.py

[clever tundra]

Dw

[earnest phoenix]

because it isnt there?

[clever tundra]

(Fawkes/app.py)

[earnest phoenix]

like that

[clever tundra]

Yes

Now try and press start

[earnest phoenix]

still offline

[clever tundra]

What's inside the Fawkes folder?

[clever tundra]

Try and change it to Fawkes/Fawkes.py

[earnest phoenix]

ok

[clever tundra]

Ahhh

Alright so you want to copy the contents of your requirements.txt file, and structure it like:
word word word

And then paste that in the box to the left of the Fawkes/Fawkes.py and then run it

[clever tundra]

Ok so

Paste in this:

discord.py PyNaCl pandas dnspython

[earnest phoenix]

but idk where to find it or do i need to create a file?

[clever tundra]

The box on the left, copy my previous message and paste it in there

[pearl trail]

basically, it's trying to find /requirements.txt but it's not there

[earnest phoenix]

ohhh

[swift barn]

add discord
and all other python packages separateed by spaces

[earnest phoenix]

here?

[clever tundra]

Yes

Perfect

[earnest phoenix]

ok so it's online now. Let me see if it's actually working

wait so the server is online but the bot isn't

[clever tundra]

What does terminal show

[earnest phoenix]

should i kick the bot and reinvite?

[pearl trail]

sorry but pls read the error 🥀🥀it says missing intent

[clever tundra]

Delete that

Please delete that now

That has your bots token it can be used to give anyone access to your bot

[earnest phoenix]

oh sorry didn't know that

[clever tundra]

No worries, just trying to save you pain in the long term

[earnest phoenix]

thanks tho

so what should i do now

[clever tundra]

Go to developers.discord.com

Login and open your bots info

Go to the bot section, then enable the intents

Like this

The top one is supposed to be enabled, dont mins my crude drawing lol

[earnest phoenix]

ok and also i need to reset my token

[clever tundra]

If you need to reset it, then reset the token in the dev portal, then update your code to include it

[earnest phoenix]

ok will do it now

the bot is still dead

    asyncio.run(runner())
  File "/usr/local/lib/python3.12/asyncio/runners.py", line 195, in run
    return runner.run(main)
           ^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/asyncio/runners.py", line 118, in run
    return self._loop.run_until_complete(task)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/usr/local/lib/python3.12/asyncio/base_events.py", line 691, in run_until_complete
    return future.result()
           ^^^^^^^^^^^^^^^
  File "/home/container/.local/lib/python3.12/site-packages/discord/client.py", line 918, in runner
    await self.start(token, reconnect=reconnect)
  File "/home/container/.local/lib/python3.12/site-packages/discord/client.py", line 846, in start
    await self.login(token)
  File "/home/container/.local/lib/python3.12/site-packages/discord/client.py", line 675, in login
    data = await self.http.static_login(token)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/container/.local/lib/python3.12/site-packages/discord/http.py", line 843, in static_login
    raise LoginFailure('Improper token has been passed.') from exc
discord.errors.LoginFailure: Improper token has been passed.
container@katabump~ Server marked as offline...
[KataBump Daemon]: ---------- Detected server process in a crashed state! ----------
[KataBump Daemon]: Exit code: 1
[KataBump Daemon]: Out of memory: false
[KataBump Daemon]: Aborting automatic restart, last crash occurred less than 60 seconds ago.

how come there's still an error for the requirement file

[pearl trail]

your token is wrong

its nothing related to requirement file

[earnest phoenix]

should it be the long one?

i already changed it tho

it should be the one below this right?

[swift barn]

are you sure you changed the token? to the updated one? after you reset?

cos thats clear asf

[earnest phoenix]

yeah i mean i changed it after i posted that

[swift barn]

ok ok

[earnest phoenix]

what other servers do you use?

[swift barn]

me? i host mine on a vps

[earnest phoenix]

like which vps?

[solemn latch]

I like galaxygate

Novonode

Or hetzner

[earnest phoenix]

is it free

[solemn latch]

No

Nothing in the Dev world is free, eventually you'll be charged money. Sometimes with no warning.

[earnest phoenix]

this looks cheap

[solemn latch]

The galaxygate team is amazing

Good group of people ^-^

[earnest phoenix]

ok I might consider changing it after it's fixed

because it's still dead rn

not sure if it's my token because pretty sure i have already changed it

it's online now but my commands arent working

[lyric mountain]

Depends on the terminal, newer ones support full range

https://stackoverflow.com/questions/4842424/list-of-ansi-color-escape-sequences

[unreal mist]

oh im talking about this part

thats the color codes i use for the calendar, i havent used any other value for it

[frosty gale]

they got a bit iffy once i decided to cancel my server with them but other than that they were always great

woudl recommend overall

[solemn latch]

👀 did they? wild

[craggy pine]

I too use galaxy gate and have nothing but good experiences with them. The discord staff is quick to assist with questions if needed.

[clever tundra]

I just use a raspberry pi

[topaz grove]

How do i get past these rate limits?

bot1-1 | discord.errors.NotFound: 404 Not Found (error code: 10062): Unknown interaction
bot1-1 | WARNING:discord.webhook.async_:Webhook ID 1369357800144109599 is rate limited. Retrying in 3.00 seconds.

[swift barn]

batch ur request, and respect the rate limits and retry again

[topaz grove]

when the ratelimits happened it threw defer errors saying they needed to all be deferred

[swift barn]

yeahh so defer all responses

[topaz grove]

aw man

i gotta work through about 50k ish lines of code, basically just setting up reusable helpers to safe defer everything

i hit a rate limit out of nowhere, i dont think anyone was using my bot so what exactly couldve caused it is what im wondering

[wheat mesa]

50k LOC in a discord bot is wild ngl

[topaz grove]

oh yeah its a massive project man

dont wanna go into details to respect advertising rules

[topaz grove]

its prolly getting closer to 60k lol

[neon leaf]

I am rusting 🔋

[wheat mesa]

The enterprise level applications I’ve worked on are like 100-150k lines, with some fairly advanced functionality

50-60k better include sharding and respective IPC logic lol

[wheat mesa]

Hell yeah

Rust my beloved

Been learning about some multithreading stuff that I’ve been neglecting, channels are such a simple but cool tool

[neon leaf]

async lifetimes

I dont think I will ever posess the skills required to write proper async lifetimes

that requires 100 years of experience

[wheat mesa]

I can imagine

I’d love to learn more about rust for distributed systems and parallel computing but school has been forcing me to spend a lot of my time on useless theory 😭

Application/interview season rn too

[topaz grove]

of course man

just becuase of funding its sharded and clustered on a ovh vps, but as we scale ill move to a kubernetes cluster via public cloud on ovh, or invest in pyhsical servers

[wheat mesa]

Kubernetes my beloved

[topaz grove]

best of the best

ive been researching all week how massive bots scale and thats what i figured out

these guys are paying upwards of 10-25k a month on hosting

its insane

but they also make like 10x that a month

[wheat mesa]

Kubernetes is pretty much the “best” solution for massive scaling

[topaz grove]

yeah

ovh rate is 0.02/hour for b3 32 and that equates to like alot when youre using 3 nodes for a startup

so its 0.06 x 24 x 30

so like 130 a month

im not pulling the trigger on that until i can establish a decent userbase

[wheat mesa]

You don’t need to write for scale if you don’t have scale

It’s good to be prepared but don’t get ahead of yourself obviously

[topaz grove]

i am writing to scale though because i want this to scale

everything is already setup i have a full backend with redis, postgres, etc

[wheat mesa]

You can write for scale but don’t overengineer until you have a proven concept

Yeah that stuff is fine

[topaz grove]

as in like

super successful?

[wheat mesa]

But I meant like multiple nodes and k8s deployments, microservices, etc

[topaz grove]

yea

yeah im not doing any of that yet

[wheat mesa]

It also depends on the stuff you’re processing

[topaz grove]

i have a vps3 on ovh and i think it can reasonably handle a load of 25k servers given my shard / cluster setup

im not gonna push it though

id upgrade to vps5 before then doing k8 and kubernetes clustering

[wheat mesa]

Bots with real time voice/audio streaming features are significantly more computationally expensive to run than others

[topaz grove]

yeah im not doing that

[wheat mesa]

Whereas normal interaction based bots can be pretty lightweight, especially if you don’t need gateway

[topaz grove]

again, not trying to break advertising rules, but its exceptionally better in almost every way to dank memer

full fledged economy bot

so mainly interactions

[wheat mesa]

If it’s fully interactions, web servers can be pretty light

You’d be surprised how many requests you can handle

[topaz grove]

yea

i ran into a big bug though with rate limits

so ive been patching all of my code lightspeed lol

[topaz grove]

wtf thats a big bot

[neon leaf]

not a bot

[topaz grove]

what application?

[neon leaf]

game server management panel

[topaz grove]

i dont know a bit of rust or c++ i can only imagine the headaches

[neon leaf]

tbh ive had a ton of fun with this project

[topaz grove]

@wheat mesa did u mean me lol

[neon leaf]

there are always points where u need to think of an issue for 2 days

but in general its nice

[wheat mesa]

Yeah sorry driving lol

[topaz grove]

my bot is almost entirely discord.py (am using sql for schema and table creating) and i also am using html for my website

im not a big js believer for discord bots despite it having early access to components v2

[wheat mesa]

I will say that if you’re intending to build for scale, python probably isn’t going to be the best choice

Python is notoriously slow

If you’re mainly bound by I/O like network calls then it’s probably not a problem

But if you’re doing heavy CPU intensive work, Python is going to be pretty slow

[topaz grove]

Yea

idk

im using alot of db drivers like asyncpg pooling with redis

[wheat mesa]

Python definitely isn’t a bad choice for a bot but just won’t scale as well as some other languages

Which that’s fine, if you really need to, you can rewrite later

[topaz grove]

Yea

[wheat mesa]

There’s almost always going to be something you can optimize, doesn’t mean that it has to be perfect

[topaz grove]

it would take a while though wouldnt it

[wheat mesa]

Assuming you’re not familiar with some other languages, yeah

[topaz grove]

0 experience

[wheat mesa]

C# and Java are both pretty good in terms of a balance between complexity and performance

Rust and C++ is where you’ll see much better performance increases, but with the penalty of increased complexity and development time

[topaz grove]

Yea

Would you reckon that the top 0.01% of bots are using that, i tried to look everywhere online and could only find out that most are using js

[wheat mesa]

Scalability isn’t necessarily always about the language. It’s about how your architect your app

These languages are all capable of writing terribly slow apps if you are using them poorly

[topaz grove]

Thats true

[neon leaf]

loop {}

[wheat mesa]

The real gains will come from optimizing your architecture at scale, e.g. using a database that is close by (or ideally in the same data center) as your host, or being able to offload work to somewhere else if it doesn’t need to be done immediately, etc

[topaz grove]

Yea

[wheat mesa]

Async message processing queue systems are a fantastic example of offloading work to somewhere else. Producers and consumers are completely decoupled and work is only performed when the consumers can take a message

[topaz grove]

I cant believe how much engineering goes into this stuff

[topaz grove]

Yep

[wheat mesa]

Which is exactly why you shouldn’t fully focus on it yet. Prove your concept before designing for scale

There’s a billion ways to optimize, but having a super optimized app that nobody wants to use is also a pretty bad problem that can’t be solved by more engineering

[topaz grove]

Yep

i totally get that

im just working on getting this thing out there

and it is hard because everyone thinks that its either a nuke, a blockblaster phish, or another type of malware

[prime cliff]

Using OVH is the main issue overchsrged prices and bad support
You don't need $130 a month to run something at best medium scale for you

[topaz grove]

Hetzner?

I heard it’s way cheaper

[prime cliff]

Yea hetzner is what most people use i also have another host that i bought hybrid dedicated server for a lot less too

[topaz grove]

isnt it way less secure though? thats why i went with ovh

[lament rock]

What makes you think that?

[prime cliff]

It's only less secure if you do something dumb but other than that nope

Also docker can help with security too

[topaz grove]

idk

just from research

[lament rock]

Systems from cloud service providers in general are pretty secure assuming you don't leak your creds somehow

Or your software doesnt have vulnerabilities. Even if, can make certain things not public facing

[topaz grove]

Okay

ill look into it more

thanks for all the help

ill stay on my vps for now until im ready to scale it

[unreal mist]

Not me being asked if I knew how to make an app, like one day maybe (considering I’m in game dev) 😭

[warped sierra]

might be a dumb question, is there any method to check if a role is a bot role (like when you add Dyno and your server now has the Dyno role) within d.js

[spark flint]

yes

{
  icon: null,
  unicodeEmoji: null,
  id: '874846937747161120',
  name: 'Dyno Premium',
  color: 0,
  hoist: false,
  rawPosition: 22,
  permissions: '8798227229887',
  managed: true,
  mentionable: false,
  tags: { botId: '168274283414421504' },
  flags: 0,
  createdTimestamp: 1628650163829
}```

managed: true

[frosty gale]

they wouldn't be in business if that was the case

its all good having a host that is fast and cheap but if it isnt secure then its sort of pointless

[topaz grove]

for sure

Im definitely looking at scaling it when i reach more people

[autumn geyser]

[topaz grove]

im kinda stressing about downtime tho lol

still trying to figure that bit out

[wheat mesa]

Downtime isn’t a problem on majority of hosts. Unless you’re going with some unknown provider then you’re not gonna have an issue with downtime

[topaz grove]

no no i mean for like updates

i know its usually like 2 or 3 seconds but sometimes it can mess with the bot especially during timed events / commands

[wheat mesa]

Why would that be an issue? You can redeploy in a few seconds

[topaz grove]

Yea

i know

[wheat mesa]

Ah you mean in the middle of things

[topaz grove]

yes

have you seen how dank memer does bankrob?

how its like a countdown thing

similar story here - i just wouldnt know what to do if something like that is going on and the bot just up and restarts

and also some things that cant really be stored on the pg tables

[lament rock]

If you're using an interpreted language like JavaScript, there are ways to hot reload things that are bound to receive changes and just not touch your core architecture.

https://npmjs.com/package/heatsync

[topaz grove]

my bot is coded mainly in discord.py

[wheat mesa]

shameless plug here

I remember you wrote this after I complained about how dogshit the file cache was for hot reloads in es6

[lament rock]

It is not shameless at all. Some huge shit was added as well where you can update existing class instance methods and remember variable values across file reloads. I am genuinely proud of heatsync

[lament rock]

Take a look at "cog reloading"

[topaz grove]

ight

[earnest phoenix]

how do i downgrade a version in Python

[lament rock]

Uninstall the current version and then install the target version

[unreal mist]

If it’s a similar thing to node, should be able to just install target version and use the version manager to swap

[swift barn]

hit and miss with that

[unreal mist]

That’s why I said if it was similar, since I don’t really know how it works with python

[pearl trail]

either venv or system installation

[neon leaf]

so satisfying

[unreal mist]

Is that py?

[neon leaf]

its rust

[unreal mist]

Ohh

[vivid fulcrum]

why not use bitwise for permissions

[neon leaf]

because they need to be extendible

[vivid fulcrum]

i picked that design up from discord and so far applying it in my projects is smooth af

ah

[neon leaf]

and there are already more than 64

so itll just be a mess

[vivid fulcrum]

yeah I gather

[still rivet]

Hey. I have a Gacha Game bot.
I was wondering jf anyone here know how to automate rewards from voting for the bot?
Say the user votes the bot. It should automatically fetch and give some rewards.
I am really struggling to figure out how to get it to give the rewards automatically, without the players having to run the command 2 times to get the rewards.
If you have a tutorial or know how let me know please! 🙏🏿

[quartz kindle]

top.gg webhooks

https://docs.top.gg/docs/Resources/webhooks/

[frosty gale]

the amount of heap allocations coming from that struct makes me want to scream

[neon leaf]

good

theres a reason its arced

[frosty gale]

i wonder what allocator rust uses

hope its jemalloc or something since youd want to avoid heap fragmentation with lots of allocations and frees

[neon leaf]

i manually compile against jemalloc

default isnt

[quartz kindle]

iirc they moved away from jemalloc because they didnt want to be dependent on it

i believe they still recommend it, they just dont ship with it anymore

[neon leaf]

something like that yea

or it was just that they wanted to stick to glibc default allocator

sadly jemalloc cannot save me either 🙏

1GB memory usage after 1 month uptime, with 95% of it being unused

(I blame redis values of up to 50kb)

[patent bramble]

So I know there's an API endpoint to get command permissions in a guild, but is there one to get the overall integration permissions?

[dusk linden]

im getting the error:

AxiosError: Request failed with status code 500

when posting server count to top.gg. this used to work, i haven't changed any code:

const result = await axios.post(
    `https://top.gg/api/bots/${clientId}/stats`,
    { server_count: client.guilds.cache.size },
    { headers: { Authorization: postToken, "Content-Type": "application/json" } }
);

[patent bramble]

#topgg-api

[frosty gale]

why is jemalloc archived now

ah https://news.ycombinator.com/item?id=44264958

jasone

Jemalloc Postmortem

Hacker News · Jun 13, 2025, 1:37 AM

lol

[prime cliff]

I've just tested replit and holy fk is annoying how can anyone use this, forced AI features and using AI to specifically create a project instead of frameworks and menus.

You have to open a new tab with a large scroll dropdown menu with a bunch of annoying choices that have sub-options and sub-menus.

[quartz kindle]

jemaloc is dead? oh my

i remember one time when i added jemalloc to my bot and it reduced memory usage by almost 50% lol

[stark kestrel]

Was used in android before

Then quite some versions ago started using scudo

[timid lintel]

[frosty gale]

though i think facebook forked it and is going to maintain it now

which is confusing because i thought facebook maintained jemalloc anyways?

a lot of the maintainers of jemalloc were facebook

its a good allocator so companies and developers arent going to let it die anytime soon

[quartz kindle]

from what i understood, facebook has always used a version of jemalloc that is tailored specifically for their systems

the general purpose version they released was actually the modified one, not the original one

i guess they just got tired of maintaining a general purpose version that is much harder to maintain than the tailored one they have in-house

[frosty gale]

facebook so big mfs have to start maintaining their own allocators now because existing ones arent good enough

[quartz kindle]

they all do that

google has their own allocator too

all big companies have their own in-house nasa tech because anything slightly general purpose is not good enough

[eternal osprey]

why does my audit log fetch return nothing?

 const auditLogs = await oldState.guild.fetchAuditLogs({
      type: AuditLogEvent.MemberMove,
    });```
this is what i am using.

[stark kestrel]

are there actually logs

i know that audit logs just never really work

[eternal osprey]

yeah there are if i check the dc audit logs

they are just not being returned lol

always a collection of 0 items.

[teal anvil]

How to claim crypto currency reward after voting

[stark kestrel]

aaand left the server

[prime cliff]

Lol what kind of scam is the dude falling for

[quartz kindle]

do you have the permission?

Requires the VIEW_AUDIT_LOG permission

[prime cliff]

Shouldnt it error if the request fails 🤔

[slow elk]

Not on promises I don't think.. unless using a .then.catch type deal

[slow elk]

Discord.js has always kind of been broken when it comes to audit log requests. I tried it before and it wouldn't work most of the time

Some bots might be directly requesting the API and/or bypassing the cache, which would technically work but is more susceptible to rate limits

One way you could potentially do it is by using webhooks, but I don't know if it's more efficient to do so

[frosty gale]

i wonder how much of their engineering is actually overkill where they couldve used an existing solution without developing an entire new thing
or at least use an alternative which fits their requirements

there has to be at least a few instances

[quartz kindle]

i can imagine instances where it mght be overkill, but i can also imagine instances where they have no idea what they are doing

[neon leaf]

I hate myself

@lyric mountain I need some advice on how to avoid this 🙏
I can still change the api if needed, fairly early in development

[swift barn]

thought it was py at first , rust is clean

[lament rock]

Wow sql can get really ugly

[wheat mesa]

This is why I use ORMs

Implicit joins my beloved

[neon leaf]

all rust orms i tried sucked 🙏

[wheat mesa]

Yeah I’d imagine so

I’m spoiled from EF Core

[lyric mountain]

that's a lotsa left joins

[neon leaf]

that may not even be the worst part

1.3kb per server, and some of the struct is already boxed (total likely closer to 2kb)

[sharp geyser]

There is diesel

It’s somewhat manageable

[neon leaf]

that prob wont fix the general slowness issues tho

[sharp geyser]

Nothing beats ef core tho

[lyric mountain]

is the issue mostly query speed or payload size?

[sharp geyser]

Nah diesel is just "nicer" to use

[neon leaf]

query is slowing down a lot yes, but payload is also a growing concern

[sharp geyser]

It likely won’t fix any query speeds

[lyric mountain]

for speed I'll need the output of EXPLAIN

just add that before SELECT

it'll give the steps & metrics of the select

[stark kestrel]

The sad part about ORMs is that complex queries are messy

And sometimes not efficient, and people using an ORM likely never really learned SQL - and then good luck getting them to optimise a query by manually writing it

[lyric mountain]

and for payload, you can use a lazy loading approach

[stark kestrel]

I'd rather use something like sqlx

[lyric mountain]

as in, have some fields only be fetched when needed

[neon leaf]

 YB Batched Nested Loop Join  (cost=5.35..436825.31 rows=1000 width=13701)
   Join Filter: (nest_eggs_1.nest_uuid = nests.uuid)
   ->  Hash Left Join  (cost=5.35..1106.97 rows=1000 width=12573)
         Hash Cond: (servers_4.uuid = server_subusers.server_uuid)
         ->  YB Batched Nested Loop Join  (cost=0.00..1098.89 rows=1000 width=12509)
               Join Filter: (servers_4.egg_uuid = nest_eggs_1.uuid)
               ->  YB Batched Nested Loop Left Join  (cost=0.00..1000.55 rows=1000 width=11132)
                     Join Filter: (roles.uuid = users_1.role_uuid)
                     ->  YB Batched Nested Loop Join  (cost=0.00..902.22 rows=1000 width=10512)
                           Join Filter: (servers_4.owner_uuid = users_1.uuid)
                           ->  YB Batched Nested Loop Left Join  (cost=0.00..803.89 rows=1000 width=7636)
                                 Join Filter: (node_allocations.uuid = server_allocations.allocation_uuid)
                                 ->  YB Batched Nested Loop Left Join  (cost=0.00..690.00 rows=1000 width=7076)
                                       Join Filter: (server_allocations.uuid = servers_4.allocation_uuid)
                                       ->  YB Batched Nested Loop Left Join  (cost=0.00..591.67 rows=1000 width=7020)
                                             Join Filter: (backup_configurations.uuid = servers_4.backup_configuration_uu
id)
                                             ->  YB Batched Nested Loop Left Join  (cost=0.00..493.33 rows=1000 width=642
8)
                                                   Join Filter: (node_backup_configurations.uuid = nodes_1.backup_configu
ration_uuid)
                                                   ->  YB Batched Nested Loop Left Join  (cost=0.00..395.00 rows=1000 wid
th=5836)
                                                         Join Filter: (location_backup_configurations.uuid = locations.ba
ckup_configuration_uuid)
                                                         ->  YB Batched Nested Loop Join  (cost=0.00..296.67 rows=1000 wi
dth=5244)
                                                               Join Filter: (nodes_1.location_uuid = locations.uuid)
                                                               ->  YB Batched Nested Loop Join  (cost=0.00..198.33 rows=1
000 width=4672)

[lyric mountain]

talk about stairs

[neon leaf]

🙏

for anyone curious, the full query is

[lyric mountain]

ok so, that's a concerning amount of counts

are u using the uuid datatype or string?

[neon leaf]

datatype

[lyric mountain]

do all these uuids have indexes on them?

[neon leaf]

yes, they are all primary keys

[lyric mountain]

I mean on the FK side

servers.node_uuid for example

[neon leaf]

yes

[lyric mountain]

well, do you need the backup configurations everytime?

[neon leaf]

nope, only when an admin wants to edit a server or a user wants to create a backup

[lyric mountain]

better make them lazy then

[neon leaf]

I think this is the general problem, im loading this 2kb struct into ram for every server request, even if i only use 3 fields of it

[lyric mountain]

yeah that select is way too big of a polycule, it's better if you include only the essential fields and defer the rest for a when you actually need them

you can also implement graphql but...it'll require implementing graphql

[neon leaf]

I guess then only load all server fields + owner data + subuser data since those are basically always used? (owner & subuser for checking auth)

[lyric mountain]

yep

[neon leaf]

my frontend dev is suing me
hes suggesting I load "minimal" metadata on paginations (uuid, name) for the objects, and the full object when requesting the server directly

I guess that could work too? or would the joins still eat up performance

[lyric mountain]

that'd be lazy loading

{
  id: 1234,
  name: "abc",
  address: { loaded: false },
  parents: { loaded: false }
}

basically

[neon leaf]

no no, what I explained would be

{
  id: 1234,
  name: "abc",
  address: { id: 1, name: "hi", ...loading },
  parents: { id: 1, name: "www", ...loading }
}

[lyric mountain]

I mean, the idea would be to reduce the number of joins to a minimum

if ur getting any data from it then it's still a join, just with reduced payload size

[neon leaf]

yep

this would already reduce a few joins though

since I wouldnt bother with nested stuff anymore

[dense lotus]

Hi

[frosty gale]

pov you use a relational database

but yeah i think lazy joins is the simplest work around aside from just dealing with loading the entire thing, have an api route for parts of the resource you use together/need

if you need multiple you can also request each part in parallel

or.. you could sort of reimplement graphql where you send a query containing the parts you need in an array/list of booleans, then the server can conditionally add joins for them in one single query

usually im one to hate graphql and how complex it makes api calls, but with your requirements i think it would do wonders and fit in very well

[neon leaf]

ill probably try cooking something up with rust traits

but I have promised my frontend dev to not touch any code until monday (im fighting the urge=

[round ridge]

Hiii, any one know an ai that is free to use with no limits like cleverbot use to be?

[acoustic bough]

a bot or an api

[round ridge]

api

[vivid fulcrum]

download a model locally

¯_(ツ)_/¯

free and no limit does not exist unless you're running locally

[acoustic bough]

well the limit is your hardware

and cost is electricity

[swift barn]

https://openrouter.ai/models?max_price=0.1

[stark kestrel]

actually max_price=0

[acoustic bough]

https://openrouter.ai/models?order=pricing-low-to-high

[quartz kindle]

what does context and tokens mean here?

[neon leaf]

context window of model

how many token it can remember

[quartz kindle]

so higher = better?

what about the tokens thing on the right side? like 7B tokens

[acoustic bough]

higher = remembers more, but worse quality

[neon leaf]

how big the model is

lower = uses less resources, may be less accurate (also depends on quantization)

[quartz kindle]

worse quality?

[acoustic bough]

the more context you give it, the more it just makes up and pulls out if it's ass

[neon leaf]

i mean ye much context with small model = 💩

bigger models handle it fine

[acoustic bough]

until a certain point

[quartz kindle]

so deepseek 3.1 free is the best one?

[acoustic bough]

amount of tokens or context window != quality

and not every model is good with everything

one might excel at math but suck at text writing

one might be awesome for literature but suck at coding

https://openrouter.ai/rankings#categories

[eternal osprey]

chat how do i get the id of a custom emoji if i don't have nitro

that shit dissapears even after using \

[stark kestrel]

inspect element

[quartz kindle]

inspecto patronum

[lament rock]

You can also right click the emoji and copy the link and the link has the id

[short sun]

apparently inspect element works on the app too?

im sorry is discord trying to hire me

[deft wolf]

Dev Tools always worked in the app

[prime cliff]

Only in canary unless you specifically enable it in normal version

[deft wolf]

I don't remember activating it, but maybe I did it a long time ago and forgot. "Developer mode" is also something I think is enabled by default but then I remember that some people don't need it

[stark kestrel]

that's why i just said to use inspect element?

[short sun]

oh lol

[lament rock]

To enable inspect element in the mainline app, you have to edit the config value with some really funny key

[short sun]

??? i never edited anything i just clicked control shift i and it opened

[lament rock]

And you're using the default client? Not canary or ptb?

[stark kestrel]

to be honest i had the same

never added that weird entry in the config json file

yet have inspect element on the default client 🤷‍♂️

[lament rock]

Maybe they added it back recently

it definitely did not work a few months back when I reinstalled my Windows

[plush grove]

Hi, some friends and I are trying to create a bot, but none of us are good at technology, and we're really confused with the commands. Could you please help me? We don't know anything.

[stark kestrel]

Then start by learning a programming language

Bots are not for beginners that don't know anything as you say

[plush grove]

I created the bot, the problem are the comands

[stark kestrel]

We don't know anything.

As I said, learn a programming language

Which will tech you how to read docs aswell, which is needed for making bots

Creating the bot in the Discord dashboard is a few clicks, making it work needs you to code it

Hence, learn a programming language and how to code - then after some time you will know how to code a bot

[vivid fulcrum]

or use a pre-existing bot, chances are something already exists for your needs among the thousands of bots that do the same thing :^)

[earnest phoenix]

hello

[frosty gale]

or use the latest groundbreaking developments in LLMs to make your bot :)

llms are unable micro-optimize my c/c++ code so im not out of a job yet sadly

they will happily use a heap allocated vector in tight loops instead of reusing buffers

[quartz kindle]

when you try a "free" AI code editor

[deft wolf]

At least you have time to make some coffee

[acoustic bough]

vibe coding is the future

just gotta wait ten mins for your turn

[spark flint]

"its my turn soon"

https://tenor.com/view/buzz-lightyear-gif-21327280

[acoustic bough]

boss sorry I can't meet the deadline

[quartz kindle]

i hate tghis so much

but some things are just so annoying to do manually

its still annoying to check every single change made by the agent to make sure its not wwrong lol

[lament rock]

I think reading over some code done by the Ai is easier than writing it out

but you lose that understanding you would have if you wrote it

Typically, I just have Ai review my code and offer suggestions