development

[wheat mesa]

inline is a suggestion to the compiler

[radiant kraken]

inline inlines your code to whoever calls it

it's good for one-liner functions

that serves as an alias to call another function

[sharp geyser]

so if they were to call push_back normally it'd just inline the insert code into the file they called it in?

[radiant kraken]

mhm

[sharp geyser]

[wheat mesa]

It tells the compiler "hey, whoever is calling this function, instead of calling it, what you're really doing is insert(length, element), so just call that instead"

The compiler can choose to ignore you whenever it wants though

Inline is good for returning private variables that you want to have get-only access

[sharp geyser]

Okay so I need to

1. Resize the vector to be able to take in a new element
2. Insert said element

[wheat mesa]

class A {
private:
  int m_Member;
public
  // Caller doesn't have direct access to m_Member, so encapsulation is preserved
  // But no precious stack space is wasted by doing a function call
  inline int GetMember() const { return m_Member; }
}

[radiant kraken]

you can also add constexpr

to that code

[wheat mesa]

no you cannot

[radiant kraken]

yes you can

[wheat mesa]

m_Member is still memory allocated, the compiler cannot guess what it is ahead of time

let me test

[radiant kraken]

no, i'm referring to the function

not the property

[wheat mesa]

I know

[radiant kraken]

pub const fn get_thing(&self) -> Thing {
  self.thing
}
``` is legal in Rust so

[wheat mesa]

oh wait you can

why

that's weird

[radiant kraken]

why not

[sharp geyser]

grr see this is why I am not motivated, I am trying to think of how to approach this push_back method but I don't even know where to start

[radiant kraken]

this entire code is using const functions lol https://github.com/null8626/decancer/blob/main/core/src/codepoints.rs

[sharp geyser]

My first thought is to resize the array so it can take in a new element, which seems to be correct

but idk how to resize it by, initially I thought to take the current length and plus 1 it

[radiant kraken]

theoretically you can, though i recommend ```cpp
while (length < capacity) {
length *= 2;
}

[wheat mesa]

doubling the capacity of the array is generally an acceptable solution

[sharp geyser]

I already have a resize_to_fit function

[wheat mesa]

Can cause problems with large amounts of elements but for your case that's not a problem at all

[sharp geyser]

might as well use that

[wheat mesa]

Don't

[sharp geyser]

Why?

[radiant kraken]

wait if length is zero then set it to 1

[wheat mesa]

Because ideally you should only have a resize() function

The resize_to_fit function was made specifically for your previous insert method

Which is flawed because you shouldn't ever be "resizing to fit some index"

[sharp geyser]

Well wouldn't I essentially do the same thing?

I need to change the capacity and make a new array and move the elements to that array

[wheat mesa]

It's the same concept but you should get rid of that function altogether because you should not be resizing to an index like that

[sharp geyser]

oh well yea

i'd do it based off the length and capacity no?

[wheat mesa]

The capacity should be controlled/changed ONLY when you run out of space in your array; every element should be contiguous

Yes

Length and capacity only

[radiant kraken]

capacity: size of the array in memory
length: the amount of populated members in said array

[sharp geyser]

Okay so when should I technically call this new method?

[radiant kraken]

capacity must always be equal or greater than length

[wheat mesa]

Whenever length >= capacity

[sharp geyser]

Should I check to see if the array can take in a new member and if not then call it?

[sharp geyser]

Oh right duh I am so fucking dumb

So could I still use the skeleton of my resize_to_fit method but just remove its reliance on an index?

or would it be different code?

to me it seems it'd be roughly the same

but instead just make it check on length and capcity instead of capacity and index

[wheat mesa]

Roughly the same yes

Try to think about what it does though

If you double your capacity, you have to allocate a new block of memory. Then, you have to copy those elements over to the new allocated block of memory. Then you must delete the old memory, and set your array pointer to the new block of memory

[sharp geyser]

yea

[wheat mesa]

You can copy elements with a traditional for loop, or you can copy over elements with memcpy

[sharp geyser]

yea

[wheat mesa]

They work fundamentally the same and I'm sure the compiler would end up optimizing them to be the same exact code

[sharp geyser]

  void resize() {
    int newCapacity = capacity;
    while (newCapacity <= length) {
      newCapacity *= 2;
    }

    T* temp = new T[newCapacity];

    memcpy(temp, array, sizeof(T) * capacity);
    capacity = newCapacity;
    delete[] array;
    array = temp;
  }

this is what I currently have

is this remotely correct?

[wheat mesa]

Yes that's perfect

[sharp geyser]

then say I wanted to use it i'd do

void push_back(T element) {
  if(length >= capacity) {
    resize();
  }

  // Rest of code here
}

[wheat mesa]

Yup

Also just a quick note on the differences between delete and delete[]

delete is used for when the pointer only points to one object in memory

delete[] is used for when the pointer points to the first object in an array of objects that need to be deleted

[sharp geyser]

[wheat mesa]

The allocator tracks this data in the underlying implementation, but just make sure you use the right type of delete operator, otherwise you may cause a memory leak

[sharp geyser]

first object in an array of objects?

[wheat mesa]

Yes

[sharp geyser]

Not quite sure I follow

[wheat mesa]

If you allocate with new only, you must use delete, if you allocate with new Something[Size], you must use delete[]

[sharp geyser]

Well yea I follow that

I am not quite sure I follow your explanation

How is array in this case in an array of other objects?

[wheat mesa]

When you do ```cpp
int* a = new int[5];

a is just the memory address of the first integer in this array

[radiant kraken]

a is basically the memory address to the entire array

[wheat mesa]

Yes and no

[radiant kraken]

it just defaults to the first element

[wheat mesa]

a points to the first element

But the elements are laid out continuously in memory

| first | second | third | fourth | fifth
^^ a points here

[radiant kraken]

*a; // element 0
*(a + sizeof(int)); // element 1
*(a + (2 * sizeof(int)); // element 2
*(a + (3 * sizeof(int)); // element 3

[wheat mesa]

Then when you do a[1] what it's really doing is
| first | second | third | fourth | fifth
^^ next element

Array indexing is just pointer arithmetic behind the scenes

[radiant kraken]

or
a[0]; // element 0
a[1]; // element 1
a[2]; // element 2
a[3]; // element 3

same thing

[wheat mesa]

The compiler hides it from you because it's slightly counterintuitive and annoying to write

[radiant kraken]

pointer arithmetic ftw tho

rust unsafe enjoyer

like seriously all of my Rust projects have a github issue saying that i used too much unsafe @wheat mesa

[wheat mesa]

Ideally you should be avoiding unsafe altogether

I don’t get what you’d be using it for unless you’re doing something inherently unsafe like embedded dev or smth

[sharp geyser]

I mean I understand how it works

but your wording confused me

delete[] is used for when the pointer points to the first object in an array of objects that need to be deleted
this wording confuses me because to me it sounds like you are saying array is in an array in memory and its the first element

[radiant kraken]

nope my C self won't let it

[wheat mesa]

array is a pointer

[radiant kraken]

i also add SAFETY commands alongside unsafe calls tho

[wheat mesa]

It points to the first element

[sharp geyser]

okay but technically its not deleting just the first element no?

[radiant kraken]

nope

[wheat mesa]

You’re correct

[radiant kraken]

it deletes the entire array

[sharp geyser]

its deleting everything at that address which is the entire array

[wheat mesa]

It deletes all of them

No

[sharp geyser]

ok

[wheat mesa]

The address is not the entire array

[sharp geyser]

null said it was

[radiant kraken]

IT IS

if you dereference it or if you call pointer[0], you get the first element

[wheat mesa]

When you use delete[] it tells the allocator that the pointer you’re deleting points to an array, and it does the necessary work to delete it

[sharp geyser]

okay so basically it is pointing to an array just not the entire array

[wheat mesa]

Yes

[sharp geyser]

but delete[] knows to not just delete the first element but rather delete everything in the array the first element is in

[wheat mesa]

The allocator knows how much memory that the pointer owns and so when you call delete[] it releases that memory

[radiant kraken]

yes thats for the operating system to know

[sharp geyser]

I am understanding now

Hey I think I got it!

  void push_back(T element) { 
      if (length >= capacity) resize();
      array[length] = element;
      length++;
  }

[wheat mesa]

There ya go

[sharp geyser]

anything I could be missing?

[wheat mesa]

Not sure what that negative number is there though

[sharp geyser]

me neither

[wheat mesa]

Make sure your length starts with 0

[sharp geyser]

aha

that was the issue

I set it to 1

How would I grab a reference to a value in an array so I can delete it then return it? Cause if I just delete the value then the variable containing the value would be empty as well no?

[radiant kraken]

you dont

try this```cpp
if (index >= length) throw le_error();

T value = pointer[index];
std::memcpy(&pointer[index], &pointer[index + 1], (length - index + 1) * sizeof(T));
length--;

return value;

[sharp geyser]

Well you see I am not doing this based off an index

I am just removing the last element

but I thought it'd be nice to get what that last element was

[radiant kraken]

oh right

try this```cpp
if (length == 0) throw le_error();

T value = pointer[length - 1];
length--;

return value;

[sharp geyser]

wait what

that works?

[radiant kraken]

yuhh

[sharp geyser]

decreasing the length of the array destroys the last element?

[radiant kraken]

no

it doesnt decrease the size of the array in memory

it just marks the array as having less members

and it works

[sharp geyser]

Thats so weird

So its still technically the same size in memory

cool

[radiant kraken]

i guess you can decrease the array size but thats just adding another overhead

[sharp geyser]

no need

the array can stay the same size not like it really matters

[radiant kraken]

*the capacity stays the same

[sharp geyser]

yea

I know what you meant

dw

Doesn't matter if the capacity is the same or not tbh

I wonder why it removes the element tho

I mean I guess it kind of has to so that way it doesn't break

but like

[radiant kraken]

thats the point of pop_back()

[sharp geyser]

well yes null I know

I am wondering why decreasing the length just removes the last element.

[radiant kraken]

it doesnt

it only marks it

[sharp geyser]

[radiant kraken]

you can technically decrease the size with another new and delete[]

but it's adding unnecessary overhead imo

[sharp geyser]

Okay so like, I know its capacity stays the same null, but its length is no longer the same

[radiant kraken]

yes

[sharp geyser]

I think you are confusing what I am saying

[radiant kraken]

yes i am

im tired

lol i always confuse what you say and you always confuse what waffle says

[sharp geyser]

What I was saying was doing length-- removes the last element from the array but I know now its capacity still stays the same

[radiant kraken]

yes, it just decrements the length int

[sharp geyser]

I was wondering why doing such will cause c++ to remove the last element, though I have a theory its because now its length is less so it has to do something so that its in compliance to what the length should be, which is remove the last element

[radiant kraken]

nooooo

[sharp geyser]

[radiant kraken]

C++ doesn't do anything here

you just make it appear so

[sharp geyser]

okay well its still confusing

cause all I am doing is decrementing the length

so I don't see how that causes this behavior

[radiant kraken]

well your Vector methods depend on the length property

so it affects all of your methods

[sharp geyser]

C:\Dev\cpp-classes\Vector.h(45): error C2561: 'Vector<int>::pop_back': function must return a value
  C:\Dev\cpp-classes\Vector.h(44): note: see declaration of 'Vector<int>::pop_back'
  C:\Dev\cpp-classes\Vector.h(44): note: while compiling class template member function 'T Vector<T>::pop_back(void)'
          with
          [
              T=int
          ]
  C:\Dev\cpp-classes\main.cpp(12): note: see the first reference to 'Vector<int>::pop_back' in 'main'
  C:\Dev\cpp-classes\main.cpp(6): note: see reference to class template instantiation 'Vector<int>' being compiled
  ninja: build stopped: subcommand failed.

wtf is happening

  T pop_back() {
      if (length == 0) return;
      T back = array[length - 1];
      length--;
      return back;
  }

I am clearly returning something

[pale vessel]

what about the first line?

[radiant kraken]

me when throw

[sharp geyser]

me when I’m too tired to care anymore

[timber hatch]

anyone who is familiar with react or nextjs, i'm building my first react/next.js project and i'm encountering a weird bug.

i am trying to markdown and format timestamps.

while on initial input it works.

if you follow these steps you get undefined for non relative timestamps:

• enter a relative and non relative timestamp
• wait for relative to update
• tab out of and back into the page
• the non relative timestamp becomed invalid.

My code: https://pastebin.com/Yf2hBe0S

anyone who could help me solve it and educate me as to what was wrong would be greatly appreciated :)

[timber hatch]

^ here's a video to better demonstrate the issue and make up for my poor explanation skills

[digital swan]

how can i stop the middle text (a tag) going over into the text on the right?

overflow-ellipsis does nothing 😔 i hate css man

[pale vessel]

try gap-1 or so for the flex container

[surreal sage]

🫠

how do i like

make it cope

leave me alone

[digital swan]

already using gap-2, anything higher pushes the number out of the container

line-clamp cuts off the word but doesnt add an ellipsis

[sharp geyser]

@wheat mesa can you explain something for me? Null was too tired last night to explain properly

[wheat mesa]

gimme 15-20 mins then I can

[wheat mesa]

Aight what’s up

[copper cradle]

oh my god

it's not removing the element, the length is just decreasing

[sharp geyser]

Can you explain to me how this works?

  T pop_back() {
      if (length == 0) return;
      T back = array[length - 1];
      length--;
      return back;
  }

It's behaviour is that it does indeed seem to remove the last element in the array, but I am still not sure I am understanding what null said because I don't get how its happening when im just decrementing the length. Now I know its capacity stays the same but its length is no longer the same in memeory right?

[wheat mesa]

The memory layout doesn’t change

You’re just essentially marking the data as free space you can overwrite

[copper cradle]

yes

[wheat mesa]

That T is still in memory unless you write over it or you deallocate the array

[sharp geyser]

Okay so I am basically freeing that spot in memory which releases it?

[copper cradle]

no?

[wheat mesa]

no

The memory is still there

You're just managing the state differently

[copper cradle]

wait, imma just bring up my paint

[wheat mesa]

You're just saying "Okay so I still have this capacity, I'm just going to say there's only 4 elements instead of 5 elements in there now"

You're not deallocating anything

You're just essentially "marking" the space as unused so that when you call push_back again, something will overwrite at that spot

[sharp geyser]

hm

I think I get it now

[copper cradle]

@sharp geyser see, the element is still there, the len is just 3 now

[wheat mesa]

If you actually deallocated that T then the person getting the value from pop_back would get an invalid pointer

although actually you're copying here so that wouldn't happen

[sharp geyser]

So essentially there is still technically 5 items, its just one is an empty spot that still exists in memory, it just can be overwritten?

[copper cradle]

if you add something to the end of the vector array list whatever you call them these days, it'll just replace whatever was at the black circle

[wheat mesa]

it's not empty

it contains the value of T

until you overwrite it

[copper cradle]

let's say that accesing it now will be undefined behaviour

[wheat mesa]

It's like how when you delete something on a physical hard drive, it often isn't ACTUALLY deleted, just the space it took up is marked as "free" and can be overwritten

[sharp geyser]

hm

Okay so, I get that its marked as free, but there is no physical value there right?

[copper cradle]

there is something there

[wheat mesa]

the value is there

[sharp geyser]

Like say I set 3 to be at the back, then I pop it off, im not going to see 3 anymore

[copper cradle]

3 is still going to be there

[wheat mesa]

That's true but the 3 might still be in the memory

It actually will be in the memory if you haven't done anything else to it with your current implementation

[copper cradle]

it's going to stay there because that memory is allocated to the vector, and if you don't do anything to it then it's going to stay there

[wheat mesa]

When you print it, it only goes up to length elements though

[solemn latch]

👀 what lang are we talking about?

[wheat mesa]

That's why you don't see it

C++

[sharp geyser]

Ahhhh

Gotcha

So it still exists, but since my methods rely on the length it wont print it

[copper cradle]

don't access out of bounds data ^^

[wheat mesa]

Change your method to print to capacity and see what happens

[sharp geyser]

If I was to try and print beyond the length id still see it I assume

[wheat mesa]

You'll probably get whatever you last popped and then a bunch of garbage data past that

[copper cradle]

yeah, and then you'll get either a lot of garbage or an immediate segfault

[sharp geyser]

I need to rebuy clion

this visual studio shit is pissing me off

[copper cradle]

just use vim

[sharp geyser]

NOPE

[copper cradle]

if you're writing c++

[sharp geyser]

not using vim, I tried that and I hated it

[wheat mesa]

nvim is pretty nice it's just a big learning curve and I'm too lazy

[copper cradle]

fair enough

then you'll never get this

[sharp geyser]

Neovim requires a license?

[wheat mesa]

no lol

[copper cradle]

no

[sharp geyser]

https://youtu.be/c4OyfL5o7DU

[green kestrel]

I don't want to use neovim

[sharp geyser]

ye so it looks so annoying

ima just buy clion :D

[green kestrel]

it's too much work to type commands into my text editor, I just want single keypressses that do stuff.

I used vim for years

but using it for larger projects is just painful. and yeah people who say "but there are plugins for this and that"...

[wheat mesa]

Expensive

Just show school ID

[green kestrel]

I don't want to spend weeks tracking down plugins and tinkering with my editor

[sharp geyser]

I dont have one

Waffle you do know Im not in school

[green kestrel]

my editor is a tool I want to work out of the box,.not a toy

[sharp geyser]

[wheat mesa]

If you still have your high school one you can just photoshop the year

[sharp geyser]

lmao

I don't think I do

[wheat mesa]

They won’t question a thing

[sharp geyser]

I got rid of a lot of my hs stuff

[wheat mesa]

Or if you have your transcript

Photoshop the year

[green kestrel]

clion is ass. just saying. phpstorm though now that's nice

clion has weird ideas about vcpkg workflow

[sharp geyser]

Clion is good to me

[wheat mesa]

CLion is awesome

[wheat mesa]

Just don’t use windows for C++

Duh

[green kestrel]

agreed tbh

vscode remote dev to a Linux box.

[sharp geyser]

I might dual boot linux sometime

[wheat mesa]

CMake is so much easier to work with on Linux

[sharp geyser]

I wonder if I can just put linux on a spare ssd that I have and be able to boot from that

[copper cradle]

CMake is ass

[green kestrel]

iirc clion still tries to use vcpkg on Linux,.yeah it works there.... sorta. but why use it when libs are in apt or whatever?

[copper cradle]

yes you can do that

why couldn't you?

[sharp geyser]

cause idk anything about boot loaders

I didn't know if it all had to be on the same drive

[copper cradle]

that has nothing to do with boot loaders tho

[sharp geyser]

listen weeb you severely underestimate how stupid I am when it comes to operating systems.

[green kestrel]

cmake is ok. a lot of cmake based projects are NOT based, and are ass

what do you prefer as alternative?

[sharp geyser]

I know how to build a pc, I know how to use one

but idk a damn thing about the os itself or how it works

[wheat mesa]

Cargo

lul

[green kestrel]

are you one of these pure make masochists?

[copper cradle]

make

yes

I am one of those

[wheat mesa]

If C++ had a centralized dependency manager then everything would be a billion times easier

[green kestrel]

enjoy lol, make is portable as an anvil

[sharp geyser]

make one

[wheat mesa]

You can’t just make one

[sharp geyser]

[wheat mesa]

The language has been out for like 40 years

[green kestrel]

mainly because of the ton of non standard gnu make extensions

[wheat mesa]

vcpkg was Microsoft’s attempt at solving it, even that sucks compared to something like cargo

[green kestrel]

no don't encourage him we got like 50 "standard Package managers" already

[copper cradle]

yeah that's a no no

[green kestrel]

at least c++ doesn't have npm.

[wheat mesa]

The internet wasn’t a huge thing like it is now when C++ was first released, I don’t think many people thought about the notion of a “package manager”

[green kestrel]

npm is like a lighter on a fireman's uniform.

[sharp geyser]

pnpm is amazing /s

[green kestrel]

can't argue havent tried it

[wheat mesa]

The npm ecosystem is disgusting

Bloated as fuck

[sharp geyser]

its not

[wheat mesa]

js developers use 3 million deps for every tiny project

[sharp geyser]

pnpm sucks as well

[green kestrel]

npm ecosystem in one word: is_even.

[sharp geyser]

is_true

[green kestrel]

is_number

[wheat mesa]

The dependencies take up more space than the projects themselves take 99.99999% of the time

[sharp geyser]

is_odd

[wheat mesa]

is-odd is a funny one because it depends on is-even

[green kestrel]

is_odd. depends on is_number and is_even

[copper cradle]

ok pnpm just manages a global node_modules instead of having one per project, that actually sounds nice, although it just symlinks packages to the local node_modules

[sharp geyser]

waffle wish me luck ima try and dual boot

:D

[copper cradle]

honestly js itself was a mistake, and node made it worse

glhf misty

[sharp geyser]

Any articles or videos you guys recommend?

[wheat mesa]

js was literally created in a week, that's pretty impressive

[sharp geyser]

I dont wanna fuck my shit up

[wheat mesa]

don't

[green kestrel]

I'm still surprised sun never sued them for trademark infringement when they renamed it to JavaScript from livescript

[sharp geyser]

why not

[wheat mesa]

just use vs for right now

you're going to lose motivation by worrying more about your environment than your actual projects

[green kestrel]

JavaScript ruined any reputation java had left

by false association

[wheat mesa]

they're not even related!

[green kestrel]

I know! lol

[sharp geyser]

ima just use vs code

[wheat mesa]

dear god no

not on windows

just use visual studio

[copper cradle]

I told you

[wheat mesa]

don't overcomplicate it

[green kestrel]

the people that named it knew too they were just riding on it's coat tails

[wheat mesa]

misty you're not even to the point that you need dependencies yet

you shouldn't be worrying about your IDE like that

[green kestrel]

what are you making in c++?

[ionic schooner]

overall vim is much much more powerful than say vscode but takes a lot of learning and configuration

[sharp geyser]

but I hate visual studio!!!!!

[wheat mesa]

he's making some data structures for practice

[sharp geyser]

I can't even uncomment shit thats commented

I have to manually go through and remove the //

[wheat mesa]

maybe I'm a zoomer but I just like VS and CLion because you don't have to set a billion things up, it's plug and play

[green kestrel]

vim is an editor for tinkerers, the sort that run arch btw. I don't have time to play I have real work to do 😁

it is powerful if you spend tons of time playing with it's configs

[copper cradle]

there's these things called uh, plugins, have you tried them?

[sharp geyser]

Theres a plugin that lets you uncomment and comment things 💀

that should be a thing by default

[green kestrel]

this

there are those who produce stuff and those that just play with settings all day

[wheat mesa]

even if I had a super familiar vim setup, I wouldn't be able to think fast enough to make it worth it lol

[sharp geyser]

Also I have no fucking clue on how to access the settings menu

[wheat mesa]

yeah don't put me in the "those who produce stuff" category so quickly lol

I probably have one or two "flagship" projects that I've actually finished

[green kestrel]

when I go in vim I have about 10 or so things I do, including regex search and replace, delete to line, jump to line... not much else tbh

[copper cradle]

then there's those that produce stuff and on their free time tinker with the tools they love

[wheat mesa]

One being a java game engine, the other being a basic treewalk interpreter in rust

[frigid leaf]

exports.run = async (client, message, args) => {
if(!message.member.hasPermission('message.guild.ownerID')) {
return message.channel.send('❌ You need to own the server to load the backup in this server.');
} anyone know how I can make this code make it so only the author that created the backup can use the command?

[green kestrel]

ah yes rust what level are you

[wheat mesa]

Right now I'm working on a proper opengl engine with C++ though, it's revived my love for low level

[copper cradle]

use a code block please

I ain't reading that without one

[sharp geyser]

there is no level

[frigid leaf]

mkay

exports.run = async (client, message, args) => {
    if(!message.member.hasPermission('message.guild.ownerID')) {
        return message.channel.send(':x: You need to own the server to load the backup in this server.');
    }```

[green kestrel]

yeah there is there are two. newbs that play rust, the game and those who don't

[wheat mesa]

I wasted so much time playing video games when I was like 12-15/16

could've been a programming god

[sharp geyser]

Honestly same

I wasted so much time on gaming I never focused much on programming

[green kestrel]

I purposely bring out that emoji when people talk rust tbh, because it winds people up more than

[copper cradle]

when creating the backup, store the id of the person that created it in a database, and then when restoring the backup check if the id of the person running the command matches that of the person that created the backup, idk if that's what you're looking for, honestly that's what I got from your question lol @frigid leaf

[wheat mesa]

tbf though it was probably worth the memories made in the process

[wheat mesa]

:norust: is blasphemy

[sharp geyser]

I found out how to comment and uncomment things

why its CTRL + K + C I have no idea

[frigid leaf]

any idea how to do that i use discord-backup library so im not sure how to store it automatically

[green kestrel]

[wheat mesa]

last 3 are valid

first one is not :c

[green kestrel]

I don't have a nocpp, anyone got one?

[copper cradle]

that's on you then, I don't even know what the discord-backup library is, but are js devs now just creating discord bot features as packages 💀

[frigid leaf]

Lmao, true i guess i could just rewrite it so it just stores it in a postgresql db

[wheat mesa]

https://tenor.com/view/sepples-cpp-cplusplus-bjarne-stroustrup-gif-25570450

[pale vessel]

[wheat mesa]

(lambdas)

[green kestrel]

[green kestrel]

coroutines

[copper cradle]

yeah, but you can still use the lib you're using, just store the id when creating the backup and store something to identify the backup, and when restoring it just check if ids match and that's pretty much it

[wheat mesa]

oh god

I'm gonna have to learn coroutines for a game engine aren't I

either way I'm sure it'll be an interesting process

[green kestrel]

not likely

[wheat mesa]

next up on the list is figuring out texturing batching with layers

[green kestrel]

most game engines are stuck in 2001 with shitty entity component systems that are too much OOP

[wheat mesa]

oh

[green kestrel]

coroutines are c++20

[wheat mesa]

I was going to make an ECS

[green kestrel]

dpp has them

but we are forward thinking

[wheat mesa]

entity component systems are super powerful as long as you actually think about the data and not the structure of classes

[green kestrel]

ECS is standard way to do game engines but don't be tempted to OOP everything, that's like using only the hammer in your toolbox, c++ has many tools

[wheat mesa]

Rust definitely helps with orienting your brain to think about it data-wise

[green kestrel]

yeah think data

[wheat mesa]

composition over inheritance

[green kestrel]

ECS is supposed to be data driven not OOP

indeed

[wheat mesa]

vtables are performance killers

[green kestrel]

no

not in themselves

[wheat mesa]

vtable lookups are performance killers*

lots of inheritance -> lots of vtable lookups

[green kestrel]

with a good block allocator it's ok, and you won't notice the difference in performance unless you're calculating rocket trajectories for spacex

a good block allocator will keep related stuff nearer to each other so good for cache

[wheat mesa]

Yeah cache locality helps a lot

Inheritance being very bad for it because everything lives everywhere

[green kestrel]

also, use vectors where you can and std::array

[wheat mesa]

I planned to use pretty much all std::array

[green kestrel]

not unordered map or map, and never std::list

[wheat mesa]

alongside std::bitset

[green kestrel]

if speed is what you want

[wheat mesa]

yeah

[green kestrel]

it's easy to turn array or vector into a simple hash table

[wheat mesa]

linkedlist bad for caching and random access, map and unordered map incurring a cost for random access as well

[sharp geyser]

So for my clear method I simply set the internal array to a new array of type T, and set the length back to 0, is this fine?

[green kestrel]

unordered map is O(1) for access, access is fast but it murders cache

also unordered map doesn't free when items are removed,.it just grows forever

[wheat mesa]

hashing still has a cost

[green kestrel]

MORE BUCKETS!

it's not a noticeable cost

[wheat mesa]

Should be fine but just be aware that if the user stores pointers within your vector, it will not deallocate them

[green kestrel]

insertion costs 1000x more as it has allocations

[wheat mesa]

But that's not on you to figure out

That's for the caller to manage

[sharp geyser]

Oh okay

I was about to say

[green kestrel]

dpp uses tons of unordered maps

but dpp priority is memory use not speed

[wheat mesa]

Your vector is just responsible for managing the chunk of memory that the T* array points to

[green kestrel]

and we "rehash" them periodically, like a GC step, we allocate a completely new unordered map and copy the contained items to the new one then delete the old one

this keeps it's rampant memory usage in check

[wheat mesa]

that's actually quite a neat trick

[green kestrel]

at the expense of some CPU cycles and a lock

[wheat mesa]

is dpp multithreaded?

[green kestrel]

yes

[wheat mesa]

dayum

[green kestrel]

and thread safe

[wheat mesa]

I barely touch threading with a 10 foot pole in Java, much less C++

[green kestrel]

it's hairy

but c++17 adds shared mutex

shared mutex is awesome

[wheat mesa]

Java makes it relatively easy for doing things like not locking a UI thread by usage of thread pools, but making data structures thread safe and such is a different story

It's bittersweet that my time with Java has come to a temporary end

[green kestrel]

you can lock a shared mutex as a reader or writer.

a reader can claim the lock so long as there are no writers

a writer can only claim the lock if there are no other readers or writers

[wheat mesa]

Yeah mutexes are nice

Java has had them for a while

[sharp geyser]

Wth is begin() and end()

[green kestrel]

makes things like monitoring a queue really nice

[green kestrel]

std::begin?

or methods?

[wheat mesa]

begin() is a pointer to right before the first element of a container, end() is a pointer to after the last element of a container

[sharp geyser]

like vector has vector::begin

[wheat mesa]

they're iterators

[green kestrel]

begin is pointer TO first element

[wheat mesa]

oh is it?

thought it was right before

[green kestrel]

end is pointer to AFTER the last (not valid)

[wheat mesa]

thought you had to increment before dereferencing

[sharp geyser]

Is it worth trying to implement begin and end in my vector class

[wheat mesa]

Since I see the for (auto it = s.begin(); it != s.end(); ++it) pattern a lot

[wheat mesa]

maybe not right now, I'd start to think about insertion and deletion at particular array indexes

iterators can be a little tricky to think about

[green kestrel]

btw if you want to iterate a container don't use iterators

old way:

for (auto iter= container.begin(); iter != container.end(); ++container)

new way:

for (const auto & item : container)

❤️

new way is far nicer

and never EVER remove an item from a container you're iterating without taking special steps to avoid UB

idk if that's an issue in rust too

[wheat mesa]

I think it's usually just a bad idea to be modifying a container you're iterating over in the first place

Regardless of language

Unless you really know what you're doing

[green kestrel]

you can modify the things your container has in it

but adding or removing from it while iterating is recipe for pain

[wheat mesa]

Yeah but modifying the size of the container in any way is a bad idea

Yeah

[green kestrel]

adding is usually "non fatal" bug, it will iterate at least one item twice

removing is source of horrible heap or stack corruptions

[wheat mesa]

Java treats this as a complete catastrophe and throws if you attempt to do this whilst iterating within a foreach

ConcurrentModificationException

[green kestrel]

neither are wise but the first can go unnoticed in code for years

lol, that's not a bad idea

[wheat mesa]

iirc the reason is because the iterators are invalidated after you add or remove since the GC can shuffle memory around how it deems fit

[green kestrel]

in dpp you'll see this pattern a lot:

lock mutex
copy container
unlock mutex
long looped thing on copy of container
delete copy

[wheat mesa]

You can still do it in a traditional for loop but not while using iterators

[green kestrel]

copying the container being tons faster than iteration to do "x"

and when you lock some mutex you want it locked as least time as possible

[wheat mesa]

I'd imagine copying is pretty cheap in contiguous memory since memcpy is super cheap

[green kestrel]

yup

so things like processing lists of http requests

processing a queue can take literal minutes

if it's big or you're near rate lokitn

but copying the queue vector takes microseconds

[wheat mesa]

man I love low level

so much to think about and the more you know the faster your code gets

[green kestrel]

you should contribute to dpp lol I did everything from scratch

[wheat mesa]

modern C++ scares me

[green kestrel]

custom websocket client and http client

the socket stuff dives deep into C

lots of pointers and raw buffers

could have used libuv but where's the fun in that and I hate dependencies

[wheat mesa]

I'm trying to modernize my C++ knowledge by writing my engine with "Modern" c++

aka using more than "throw this pointer around, who cares"

maybe a unique_ptr or shared_ptr in there somewhere and a little bit of optional when I start writing the ECS

[green kestrel]

one hint id give is if you feel you're typing too much boiler plate you're probably doing it a bad way

and if you find yourself typing long type names over and over remember auto and "using"

[wheat mesa]

most of the boilerplate I'm writing is just trying to modularize opengl stuff

[green kestrel]

people who complain c++ is too verbose are usually trying to emulate java

[wheat mesa]

I wrote a lot of different classes (this is my first time using opengl) following videos from TheCherno's opengl series, just to realize that I can stuff all of it into a renderer class and it would be better anyways

[green kestrel]

if you want an example of how not to write c++ look at libmysqlclient++ (the official mysql c++ lib).. it's a horrific mess that tries to emulate jdbc

[frigid leaf]

anyone know any API that is useful for my discord bot, I want it to fetch automatic lyrics while the song plays

but like I want to stay on the cheap side or even like free side X)

[wheat mesa]

For example here's part of my renderer constructor lmao ```cpp
m_QuadBuffer = new Vertex[s_MaxVertexCount];
m_QuadBufferPointer = m_QuadBuffer;

// VertexBuffer
glGenBuffers(1, &m_VertexBuffer);
glBindBuffer(GL_ARRAY_BUFFER, m_VertexBuffer);
glBufferData(GL_ARRAY_BUFFER, s_MaxVertexCount * sizeof(Vertex), nullptr, GL_DYNAMIC_DRAW);

// VertexArray
glGenVertexArrays(1, &m_VertexArray);
glBindVertexArray(m_VertexArray);
glEnableVertexAttribArray(0);
glEnableVertexAttribArray(1);
glEnableVertexAttribArray(2);
glEnableVertexAttribArray(3);
glVertexAttribPointer(0, 3, GL_FLOAT, GL_FALSE, 10 * sizeof(float), (const void *) 0);
glVertexAttribPointer(1, 4, GL_FLOAT, GL_FALSE, 10 * sizeof(float), (const void *) (3 * sizeof(float)));
glVertexAttribPointer(2, 2, GL_FLOAT, GL_FALSE, 10 * sizeof(float), (const void *) (7 * sizeof(float)));
glVertexAttribPointer(3, 1, GL_FLOAT, GL_FALSE, 10 * sizeof(float), (const void *) (9 * sizeof(float)));

I could replace the VertexArray stuff with a nice little layout manager, but then I remembered that it's pointless because this engine is specifically for me

[green kestrel]

what are the vertexes for

[wheat mesa]

And I do not plan on needing more than a position, color, uv, and layer

[green kestrel]

shouldn't that be from a model file or something?

[wheat mesa]

I'm making a 2d engine

No need

[green kestrel]

ah is it your display plane?

[wheat mesa]

Everything will be rendered as a rectangle

And this is just for telling the GPU how the data looks in vram

[green kestrel]

id be tempted to display just one quad with a texture on it and update the texture lol

[wheat mesa]

position 0 is 3 floats that represent position, 1 is 4 floats that represent color, etc

[green kestrel]

but last time I did 2d, I did software rendering

[wheat mesa]

this is similar to what I'm doing

I wrote a textureatlas class that manages stuffing multiple textures into one and just sampling from different parts of it

(thanks to stb_rect_pack for doing the hard work)

The next thing on my todo list is to make it so I can access multiple texture units instead of just one

that way I can reduce draw calls even more

those images are programmatically packed into the black rectangle

[eternal osprey]

hey guys, i finally made some time to put my api online. So as stated before i have an api running on my vps and my own site on my other hosting platform. I currently do fetch requests to the api and eventually also redirect customers back to the site etc. How would i possibly connect my api to my site?

Like setting a subdomain

such that i can fetch my api using domain/subdomain.com

my site is also running on https

so i recently tried using fetch to my express api which would not work as https -> http was not allowed by cors

[green kestrel]

is the window with hello world imgui?

[wheat mesa]

yea

[green kestrel]

first make sure the API is accessible on port 443

point subdomain at it via dns

and use certbot and letsencrypt to get free ssl

after that you could use cloudflare to obscure the real IP

[eternal osprey]

i see obvs https has a 443 port.

[eternal osprey]

can i instead also create an a record pointing to my ip?

[green kestrel]

yes

[eternal osprey]

in the dns records of my site

i see

[green kestrel]

just untick the orange cloud thingy

in cloudflare

[eternal osprey]

idk it's handled by my hosting

i can't even access cloudfare i think

i can only add dns records manually but that's about it

[copper cradle]

what did I miss here

[sharp geyser]

@wheat mesa pinging you cause I know you know how to help but if anyone reads this and can help please do so

so for my insert method I can think of a problem i'd have to face.

What if they are trying to put something at an index that already has a value. My initial thought was to just shift everything after that index down one and then add in the element at that index but idk how i'd do that unless this could work

void insert(int index, T element){
  if(index >= length) return;
  if(array[index]){
    // If there is already something there
    for(int i = 0; i < index - 1; i++){
      // will likely need to change what i or my comparison is but I can't think right now lol
      array[i] = array[i + 1] // this seems wrong as well but again my brain no think properly
    }
    array[index] = element
  }
  length++;
}

There might be some other problems but im not too sure rn of any others i'd face.

[wheat mesa]

Your intuition of shifting is correct, but backwards. Shift everything to the right, not the left

Or maybe I misinterpreted and that’s exactly what you meant

[sharp geyser]

well

I want to shift everything after the index down one

so that way the element can take over that index

this is in case they are wanting to insert something at an index that is already taken

[sharp geyser]

hm, how exactly should I be shifting? I feel like I should be using the index in some way shape or form since that is what I am basing the shift off of

but all my testing gives me some odd behavior where negative numbers take over the spots

or it just fills most of the indexes with the last element

Here was my next thought but it doesn't produce an appropriate result:

  void insert(int index, T element) {
      if (index >= capacity) resize();

      if (index >= length) {
          length = index + 1;
      } else {
      length++;
      }

      for (int i = length - 1; i > index; --i) {
      array[i] = array[i - 1];
      }


      for (int i = 0; i < length; i++) {
      std::cout << "Here is the elements after inserting: "
                << "\n" << array[i] << std::endl;
      }

  }

I thought using the length of the array for i and then seeing if i is greater than the index would work. I was thinking of starting backwards and shifting the elements down until it reaches the index like that

[drowsy jewel]

Why not start from where you want to insert the index?

Something like ```c++
for (int i = index + 1; i <= length; i++) {
array[i + 1] = array[i]
}

length++
arr[index] = element

[quartz kindle]

that would delete the existing contents of i+1

[drowsy jewel]

Woops

My bad

Wait no

[drowsy jewel]

I think this shld be ok

Well basically you would be shifting elements to the right, I'm sleepy so I think my brain isn't functioning correctly. Once you shift the elements, place the element you want in the index then increment length

[crystal wigeon]

are there any game devs here?

or someone with good knowledge with rpgs

[sharp geyser]

I mean I am a beginner game dev

What question do you have, I might be able to help

[crystal wigeon]

Ah mmm im more like trying to figure out how to balance abilities

I have a discord rpg game

But it’s currently failing cause it’s not good enough

Need some game dev knowledge to fix and balance things

[sharp geyser]

Okay, what are you currently doing?

[crystal wigeon]

Would be great if you can help me out

[crystal wigeon]

Mmm do like the rpg game has many abilities

[sharp geyser]

Ofc

[crystal wigeon]

Some of them are useful and some are completely useless and dead

not viable

[sharp geyser]

a lot of rpg games have some kind of abilities that players can have.

[crystal wigeon]

You should maybe look at izzi bot XD you’ll get an idea of what’s happening

So the abilities are on cards that players can collect

[sharp geyser]

Okay

Do you feel the abilities are too overpowered?

[crystal wigeon]

not exactly but there are some abilities that players don’t use at all

Cause they useless or under powered

is what they’re saying

Feedback from players

[sharp geyser]

Hm, well. You can either find a way to rework the abilities so they are good, but not game breaking, or remove them and replace them with some other ones.

If you haven't already i'd also attach a rarity/drop chance to them if thats how your game is styled

You don't want people just getting these abilities for free, they have to work for em.

[crystal wigeon]

mm that’s the thing I need help with like talking to someone to rework them it would be great to talk to game dev as they would know better

Yeah

There’s rarity and stuff

[sharp geyser]

Well, its hard to tell you much without knowing your game's plot/theme.

I can't make any judgements

[crystal wigeon]

Yeah, which is why I suggest taking a look at izzi bot

That’s my bot

[sharp geyser]

Some games have abilities that make sense for them, while other abilities don't make sense for that plot/theme

[crystal wigeon]

If you want to help out that is

yeah

[sharp geyser]

I can take a look

[crystal wigeon]

great,

[crystal wigeon]

https://discord.gg/Wj4t9Y6ePb

ping me when you join here, i

i'll fill you in with the details

[sharp geyser]

Question, why does your bot require the ban and kick members perm?

[crystal wigeon]

you do have to play a bit at first tho, to understand the logic

[sharp geyser]

You dont advertise any mod commands

[crystal wigeon]

it doesnt you can remove it

[sharp geyser]

I'd recommend updating your invite link then

[crystal wigeon]

but it can kick / ban people

just not advertised as mod bots

[green kestrel]

ever wonder if the police use onetime passwords? /j

[sharp geyser]

What’s a one time password

[pale vessel]

a password that u remember one time and use for every website

[green kestrel]

[sharp geyser]

Oh I have plenty of those

[feral aspen]

https://cdn.hamoodihajjiri.com/9lr6F4CyVb
Same URL however different website? 👀

One leads to notion.so and other leads to notion.apps-pg.com (ad)?

Main question is, how does Google consider both websites same URL?

[ionic schooner]

Notice the top one says sponsored? Thats the difference. It is the same website but ones a paid advertisement

[feral aspen]

However, it's not the same website; the top one leads to notion.apps-pg.com, which is a scam website, while the bottom one leads to notion.so, which is the real website?

[ionic schooner]

Oh that’s weird asf

[feral aspen]

Oops, sorry. I just saw this (a month late to replying), it was a prototype that espouses Tolerance; since that was the theme of the competition.

[spark flint]

google is good at scam ads

[radiant kraken]

espouses tolerance?

[feral aspen]

Yeah, just a prototype of an app that promotes Tolerance.

[radiant kraken]

nice

[feral aspen]

Yeah, anyhow, the contract is now terminated; now that the program is finally over.

^ You had to learn swift, do assignments, pass Apple's exam, & graduate. We did all.

[neon leaf]

fuck this took a bit

[radiant kraken]

noice

[green kestrel]

what are you using for the browse thumbnails

ive just finished developing a resurrect feature for seven spells... the idea is that you can respawn near to your place of death once per hour, if you dont have a resurrect you have to wait or do what you always had to do before - start from the start

so ive basically added an optional non-hardcore option

of course premium users get to resurrect more often... 😉

and if you die and have no resurrect, it says this, basically "did you know premium users can res more often, if you were premium you'd be back in the land of the living rn"

😄

[neon leaf]

help tje ai logged into my google account

[green kestrel]

it knows what's best for you, don't resist this will all be over soon

[vivid fulcrum]

if you're still interested, https://www.techradar.com/pro/security/google-ads-are-being-hijacked-to-serve-up-dangerous-malware

[winged linden]

yo question

            ":white_check_mark:・ufo-verification",
            {
              type: "text",
              permissionOverwrites: [
                {
                  id: interaction.guild.id,
                  allow: [
                    "VIEW_CHANNEL",
                    "SEND_MESSAGES",
                    "READ_MESSAGE_HISTORY",
                  ],
                },
                {
                  id: interaction.guild.roles.everyone,
                  deny: ["SEND_MESSAGES"],
                },
                {
                  id: client.user.id,
                  allow: ["VIEW_CHANNEL", "SEND_MESSAGES"],
                },
              ],
            }
          );
          await channel.send({
            embeds: [embed],
            components: [row],
          });

          await interaction.editReply({
            content: "Setup Complete!",
            ephemeral: true,
          });```

what does this need perms other then Manage_Channels & Embed_Links?

its sometimes failing for some users not sure why

[winged linden]

i have a checker only for those 2

[sharp geyser]

Manage roles if you want them to be able to manage channel perms. Also if you set the need to have 2FA enabled in the server to do any mod actions then type person will need 2FA enabled to manage the channel

It’s a discord security feature

[copper cradle]

zamn

[lament rock]

could have base gifs and then tint and manipulate and combine the images depending on conditions

[low orbit]

i just resulted to using videos ngl

[lament rock]

webp if its an option

[sharp geyser]

so wtf is a bitset, is it just something that stores booleans as 1 bit?

[wheat mesa]

yes and no

Basically it treats integers as "sets" of bits

[sharp geyser]

it can only store 1's and 0's no?

[wheat mesa]

booleans are inherently inefficient because they use 1 byte of storage (sometimes more) but only NEED 1 bit to function

But using bitwise operations you can stuff multiple booleans into one integer

[sharp geyser]

hm

So how does it take something that normally takes 1 byte, and slims it down to 1 bit

does it just force its bit size? (if thats even possible)

[wheat mesa]

Instead of 8 booleans with 1 byte each, you can use a 1 byte number and treat each individual bit as a boolean

You can make this dynamic to handle more than that though, which is essentially what a bitset does

[sharp geyser]

huh

  std::bitset<10> myBitSet;

  myBitSet[2] = 1;

[wheat mesa]

Yes

binary is right to left

(Well, for this configuration it is)

[sharp geyser]

Ooooo

right

Is there a way I can tell how much something is taking up?

[wheat mesa]

sizeof(T)

[sharp geyser]

oh right

[ionic schooner]

C moment

[sharp geyser]

wait

you said a bitset stores them as 1 bit right?

[wheat mesa]

not necesarily

[sharp geyser]

hm

then im confused on the purpose of a bitset

[wheat mesa]

If you have multiple booleans it reduces the overall size when compared to a boolean array

There is no benefit if you just have one boolean

[sharp geyser]

hm

[wheat mesa]

    bool a[8];
    cout << "Size of a: " << sizeof(a) << "\n";

    std::bitset<8> set;
    cout << "Size of set: " << sizeof(set) << "\n";

Size of a: 8  
Size of set: 4

[sharp geyser]

hm

is this in bytes?

[wheat mesa]

Yes

[sharp geyser]

neat

Btw waffle

what is a good lib/framework to use to make desktop apps with c++

[wheat mesa]

Qt

But don't do that

You're going to kill your motivation

[sharp geyser]

dang I was hoping it'd be simple like electron

[wheat mesa]

not at all

[sharp geyser]

welp that squashes my project idea then

[radiant kraken]

lmao nope

you're not building desktop apps with html/css/js

@sharp geyser try GTK, GLFW, SFML, or Imgui

[wheat mesa]

I'd recommend learning more c++ first

[radiant kraken]

like which part of C++?

[wheat mesa]

memory management and stuff

[sharp geyser]

what do you recommend I do then

[wheat mesa]

graphics programming is very hard

[sharp geyser]

I have no idea where to start to even learn memory management

[radiant kraken]

memory management is ez

just new and delete

[sharp geyser]

makes sense

[wheat mesa]

ownership

with that being said, you should do projects that keep you interested in learning

SDL is relatively easy as a graphics framework, but just don't go into it expecting quick results

[radiant kraken]

this aint rust

[radiant kraken]

i mean that goes the same for every new project using big frameworks

[wheat mesa]

ownership is still a concept in C++

it just isn't enforced

[sharp geyser]

ya see idk what projects to do. I am just starting out learning C++ again so I am going with a fresh slate.

[radiant kraken]

i suggest attempting GUI ig

just dont go for the big frameworks first

start with the simpler ones

[sharp geyser]

https://www.geeksforgeeks.org/top-50-cpp-project-ideas-for-beginners-advanced/

Ima work through that list slowly. It should teach me what I want to know ;p

[green kestrel]

geeksforgeeks is an awful site

beware their "tutorials"

sometimes they're even worse than chatgpt

basically the article writers don't have any real experience, they just parrot what they know and there is no peer review of what they write

[sharp geyser]

it might be aweful but it gives me some ideas

Man you can't use floats in enums

[radiant kraken]

you can

[sharp geyser]

no?

[radiant kraken]

how?

[sharp geyser]

because enums in C++ can't accept floats or doubles or anything not integral

[radiant kraken]

ooooh

i thought you meant unions

[sharp geyser]

no

[radiant kraken]

why would you want floats in enums tho?

[sharp geyser]

I was making a grade enum and assigning values to the letter grade.

going based off a 4.0 grading scale with a max letter grade of A

[radiant kraken]

you can make the enum Grade { A, B, C, D, E, F } but process the floats yourself

like a helper function

[sharp geyser]

fair

also what is E

[radiant kraken]

idk

we dont use letter based grades here

[sharp geyser]

oh ok

I thought E was something for you

[radiant kraken]

only percentage-based grades

[sharp geyser]

[radiant kraken]

[sharp geyser]

Well you can still turn a percent into a letter

if you go based off typical grading standards that accept +/- with a max letter grade of A

95-100 A
90-94 A-
87-89 B+
83-86 B
79-82 B-
77-79 C+
73-76 C
69-72 C-
67-69 D+
63-66 D
59-62 D-
0-58 F

idk if these are the same everywhere

[radiant kraken]

damn

below 75 is a D here

C is 75 to 80s

[sharp geyser]

but out of all the grading systems i've seen in a lot of parts of the world this seems to be the closest to universal

Though I think Asia is more strict

[radiant kraken]

also there's no F, D is the lowest

below 75 is a fail

[sharp geyser]

Yup asia strict jeez

I mean I get it

America and UK are very lax imo

America Especially

[radiant kraken]

and yet americans complain about their education system 😩

[sharp geyser]

you could be as dumb as a box of rocks and still get your diploma

I complain that it is too ez

I passed high school with a 2.7gpa

I should not of passed

[radiant kraken]

someone's lurking

[sharp geyser]

but I did

probably flaze

he always does

[radiant kraken]

what's the minimum

[sharp geyser]

I didn't apply myself in high school and I suffered for it

[sharp geyser]

tbh idrk

I've seen people with a 2.0 pass

I think the only requirement is to too just good enough to get the credit

and thats it

once you get all your credits you are no longer their problem

[radiant kraken]

my high school average is a near 91% (not including last semester)

[sharp geyser]

they send you on your way with a diploma

I think my grad year though almost everyone had a 2.7 or higher

[radiant kraken]

dang

[sharp geyser]

I knew a number of kids who had a perfect 4.0

quite honestly I probably could of been one of those kids, I just slacked off majorly in freshman and sophmore year

I didn't start trying until junior year but by then my gpa already tanked

[radiant kraken]

which subjects are you good at?

[sharp geyser]

Science and History

[radiant kraken]

nice

I liked history but know nothing about science

failed almost all physics, biology, and chemistry exams

[sharp geyser]

I want to learn physics but I didnt take the chance to

[wheat mesa]

Eh you’re not missing out on much

Physics is cool but it’s not a necessary class to have

[low orbit]

What even is gpa

We don't have that in uk

Well England atleast

[grim aspen]

grade point average

[hidden gorge]

is every minute too fast?

[deft wolf]

It depends where this data comes from. If it's from the database/cache, it probably doesn't matter much

But every minute is still too fast in my opinion because no one will sit and check how many servers you have every minute. Even if you set it every 30 minutes, these values ​​would not differ much from each other

[radiant kraken]

1 minute will be potential API abuse if your bot grows larger

[hidden gorge]

it only sends to my server is that good?

[radiant kraken]

i mean, your bot will process and send more messages, no?

[solemn latch]

Realistically discords unlikely to care, you're unlikely to hit the global ratelimit because of this. It just means your bot in your server will have 1 less message edit per minute. Which is the only one you'll potentially hit as your bot edits other messages in the server.

To me its more of a question of why 1 minute?

Its very unlikely anyone(including you) will see much change by checking that channel. It could be hours, or a day and it wouldn't degrade the experience much(most wont notice)

[slate kestrel]

i mean regardless i'd probably check every minute and only update if it differs

[spark flint]

I don’t think you need live stats like that either lol

Highly doubt your bot will join a new server every minute, best to update every 15/30 minutes instead

[slate kestrel]

that's true, it's not necessary. Can just fetch it with a /stats command

[solemn latch]

Real analytics are way more fun anyway ^-^

[radiant kraken]

Or you can just make a website for it

[radiant kraken]

i think he was trying to display his bot's real-time stats in his support server

[slate kestrel]

Yeah i figured so

[wheat mesa]

Plus, why not just get the user count via guildMemberAdd/Remove and guildCreate events, store it in memory, and save that to a database every now and then

Sure it wouldn’t be accurate 100% of the time but it saves a lot of bandwidth overall

You’d only really have to fetch all of the information at once from the discord API if you have significant downtime

[hidden gorge]

Thing is it isn't using a database

It pulls from the bots cache

I only created it bc I got bored

[deft wolf]

Wouldn't it be better to do something actually useful?

[hidden gorge]

My ideas for my bot randomly come and go so I never really have something useful to do

[earnest phoenix]

Hello?

Is anyone there

[warm surge]

#general is probably best place talk

[earnest phoenix]

Im not here to talk

[warm surge]

oh

what’s up

[earnest phoenix]

I need help with the dbl thing

[spark flint]

Wrong dbl

[warm surge]

wrong swrver

@earnest phoenix https://discord.gg/ue7dMHdz

[earnest phoenix]

Thank you guys

[warm surge]

Your welcome

[pale vessel]

@bitter granite pertemuan sigma 🗿 https://youtu.be/rkgUtEuAHmc

u stopped playing right? u haven't logged in in almost 80 days

[bitter granite]

Havent for a while
But verry nice

Althou why #development

[pale vessel]

LOL

[bitter granite]

||Nevermind charater development ||

[pale vessel]

i forgot why im here ngl

because the usual people were here talking for some reason, so i thought im in general

[sacred python]

How can I create a subcommand with discord.py?

[pale vessel]

https://discordpy.readthedocs.io/en/stable/interactions/api.html#discord.app_commands.AppCommandGroup

set the type to https://discordpy.readthedocs.io/en/stable/interactions/api.html#discord.AppCommandOptionType.subcommand

an Argument with the type subcommand probably works too. not sure why they have both Argument and AppCommandGroup when u could set the type for both

[sacred python]

Can you give me an example please?

[past field]

question

i made a command that should have notified ppl with these roles

but it doesn’t actually notify them in the command.. is that normal?

[solemn latch]

Its been awhile since I've done bot development.
But I think bots are required to enable a mention when they send a message.

Lemme look into it real quick.

Also, how is it going mo? ^-^ getting comfortable with dev?

[solemn latch]

Can I see the send(...) code?

[past field]

It’s actually going pretty good! I’m learning so much from @deft wolf !! I’ve gotten about 22 slash commands so far that all function how I need/want them to

[past field]

yes 1 moment

[solemn latch]

awesome!

[solemn latch]

Intresting 👀

[past field]

it was embedded at first, but i thought that was why it wasn’t notifying anyone, so i reverted it to plain text and it still isn’t notifying those roles 😭

[solemn latch]

It doesnt feel like you've done anything wrong here. Have you done other commands that mention people and it works?

[past field]

let me try

it looks like it will notify users by their @mention, just not roles

my bot has admin/all perms toggled on

[solemn latch]

You might need to use allowed mentions.

https://discordjs.guide/popular-topics/faq.html#how-do-i-control-which-users-and-or-roles-are-mentioned-in-a-message

^ take a look, if you need help ping me

[past field]

ok checking it out now

thank you @solemn latch

[solemn latch]

npnp

[past field]

hm

so basically

i would replace ${notifyUsers} with allowedMentions

@solemn latch correct?

well the entire line of course

[solemn latch]

You shouldn't need to modify any of the message variable

You just need to edit this line await interaction.reply(message);

[past field]

ah ok i see

[solemn latch]

I've not used djs in so long, tbh I'm not entirely sure how this works.

[past field]

ok i’m going to try it

ok so i’m seeing that notifying tagged roles in slash commands are limited on discord

that might be why lol

[past field]

@solemn latch are you still awake? haha

[solemn latch]

I don't sleep ^-^

[past field]

lol ok so I have a question, and if it’s going to be too much to help with i completely understand

so everything that I’ve been doing to create my bot and run it has been done on my macbook pro

so i took the same folder, uploaded it to google drive, and downloaded it on my windows PC. I know the paths are going to be different, but I remember you helped me correct the path when I first started this project

would you be able to walk me through it again? i can’t remember what you had me do

or is it a lot more complex than that? @solemn latch

[solemn latch]

It can require modifying paths sometimes when going between windows and mac(or Linux).

Sadly I can't help with that tonight

[past field]

ok i gotcha, no worries! i had a feeling it was more complex than what i was thinking lol

[solemn latch]

Not something you need to worry about, but in the development world we have a program called docker.
It's the best worst thing ever, but it means when you write programs it can run on any system and act like they run on a specific OS.

Really neat for making sure whatever you're developing can be developed from any machine. Useful when working with multiple developers, or you want to be able to swap between your windows and mac machine.

[past field]

Ah ok I gotcha, I’ll have to look into Docker

[sharp geyser]

I’d look into using path.sep in some way shape or form

if this is nodejs

[past field]

yes it’s nodejs

[sharp geyser]

path.sep will give you the operating system specific file path separator

[past field]

ok i can look into it

[sharp geyser]

If you have any questions I can also help

[feral vine]

[sharp geyser]

Actually I lied, use path.join, path.sep gives you the platform specific separator, but its easier to just use join as it uses path.sep anyway.

[past field]

ok i will research it a bit

i am extremely new to this by the way

like just started learning all of this 2 days ago

[sharp geyser]

const path = path.join("whatever", "something") // /whatever/something unix or \whatever\something windows

All good

everyone starts somewhere

One place to never be afraid to ask for help in is top.gg

most of the people here are experienced and if not there are plenty of other people who are new just like you.

[pale vessel]

i've always used slash and they work just fine on both windows/linux

[past field]

well

[pale vessel]

i wonder where it wouldn't work

[sharp geyser]

they typically do work on windows, but sometimes it can cause issues

[past field]

it keeps saying that my bot token can not be used

but more importantly

[sharp geyser]

I've had cases where i'd use require statements and it'd not find the file/folder I was looking for because I was using slashes

[past field]

i do have a question 😅

[sharp geyser]

Go ahead and ask.

[past field]

So I’m coding my bot with javascript with slashcommandbuilder in visual studio code