development

[sharp geyser]

Okay so I just realized something

my array is funky

despite me saying index 0 is 100

apparently index 0 is 1000

which is supposed to be index 1

[radiant kraken]

[sharp geyser]

are there template strings for c++

like how js has ```
${}

[radiant kraken]

yes

afaik there's std::format

[sharp geyser]

I am so confused

when inserting everything seems fine

but when going to remove the indexes are fucked

[radiant kraken]

like what? ^

[sharp geyser]

  void remove(size_t index) {
      if(index < length && index >= 0) {
          memcpy(&array[index], &array[index +1], (length - index + 1) * sizeof(T));
          length--;
      }
  }

this was my current thought process

actually I think that first part of the if is fucked but

[radiant kraken]

doesnt sound like it

maybe (length - (index + 1)) * sizeof(T)?

[sharp geyser]

yep

that fixed it

wonder what was happening

[radiant kraken]

noice

[sharp geyser]

oh wait

it was essentially evaluating 11

but wrapping it in parantheses allowed it to do the addition first

[radiant kraken]

yes

[sharp geyser]

math

fucking hate it

but I need it 😭

Well

that was easy

[radiant kraken]

math is okay

[sharp geyser]

I was just going to use a for loop

with memcpy it made my life easier tho

[radiant kraken]

frfrfr

[sharp geyser]

now I am tempted to try it with a for loop tho

to see if I can get it

[radiant kraken]

why

[sharp geyser]

Just to try it and see

[radiant kraken]

okay

[sharp geyser]

its nice having built in stuff

but I wont learn by simply just using whats given

I basically figured out the efficient way before trying it myself

welp

I fucked up

meh

I will leave figuring it out without using memcpy for another time

[surreal sage]

wadahel

[earnest phoenix]

Ask me what exactly?

[radiant kraken]

why you cant write template class method definitions on a cpp file as opposed to a header file

[earnest phoenix]

Who says you can't?

[radiant kraken]

Brain did

[earnest phoenix]

The C++ standard does not prohibit such a thing, there's just no concrete way to do it as template declarations and definitions both must be accessible to the compiler in the translation units

There are multiple ways this can be done

1. Explicit template instantiation

foo.h:

#ifndef FOO_H
#define FOO_H

template <typename T>
class Foo {
 public:
  void bar();
};

#endif   // FOO_H

foo.cpp:

#include "test.h"

template <typename T>
Foo<T>::bar() {
  // ...
}

main.cpp:

// Yes, including the C++ file
#include "test.cpp"

// Explicit template instantiation of the types that will be used, so the linker can find the generated classes by the compiler from the translation units
// In this case, we're only using `int`
template class Foo<int>;

int main() {
  Foo<int> foo;

  foo.bar();
}

2. Late binding of translation units

foo.h

#ifndef FOO_H
#define FOO_H

template <typename T>
class Foo {
 public:
  void bar();
};

// Include the implementation file
#include "foo.cpp"

#endif   // FOO_H

foo.cpp:

template <typename T>
Foo<T>::bar() {
  // ...
}

main.cpp:

#include "foo.h"

int main() {
  Foo<int> foo;

  foo.bar();
}

[radiant kraken]

then explain this @earnest phoenix

[earnest phoenix]

They're not doing it properly, just separating them like the other non-template stuff won't work, you have to use one of those solutions I provided to be able to separate declaration and definition of templates

[radiant kraken]

have you reviewed my pull request? what do you think of this https://github.com/Mistyerious/Vector/pull/1

[earnest phoenix]

What's the point of reviewing it when it's already merged?

[radiant kraken]

i just want for your thoughts on it

[earnest phoenix]

Hmm why are you using std::realloc() with new/delete?

[radiant kraken]

what's wrong with that?

like i get that it's not calling the constructor but it's purpose is just for expanding the memory block

[earnest phoenix]

The underlying allocators used for std::realloc() and new can be different and not have the same techniques of keeping track of memory, basically it's undefined behavior

[radiant kraken]

so is the for loop the only way?

[earnest phoenix]

Hm?

[radiant kraken]

like there is no resize keyword obviously

[earnest phoenix]

T* tmp = new T[capacity * 2];

// Move elements of previously allocated array to the newly allocated array

// Deallocate the old array
delete[] arr;
// Increase capacity
capacity *= 2;
// Use the newly allocated array
arr = tmp;

For continuous growth, keep increasing the capacity in a loop until it satisfies your specific condition and do the same steps

[radiant kraken]

does memcpy work for a remove(index) function?

[earnest phoenix]

Sure

This effectively reduces the container size by the number of elements removed, which are destroyed.

Because vectors use an array as their underlying storage, erasing elements in positions other than the vector end causes the container to relocate all the elements after the segment erased to their new positions. This is generally an inefficient operation compared to the one performed for the same operation by other kinds of sequence containers (such as list or forward_list).

[thorny shoal]

Music

[slender wagon]

hallo, is anyone good at go here?

so i am trying to improve my go build command

i added some flags that help with the compile

but i am wondering if there is a flag to use more cores while running build

[sharp geyser]

Hello

[earnest phoenix]

Hey

[radiant kraken]

@sharp geyser ello

[sharp geyser]

Just woke up

[radiant kraken]

good morning

[sharp geyser]

Ty

@earnest phoenix so you can include cpp files? I thought you could only do so to header files

[earnest phoenix]

You can include the contents of literally any file, no matter the file extension, the preprocessor doesn't care about what the file is or its contents

[sharp geyser]

Oh okay

That’s nice to know

Is there any real reason for that?

Or is it just a byproduct of the preprocessor

[earnest phoenix]

Because there are a lot of people who use different unofficial C++ file extensions (tpp, ipp, cxx, and many others), so the preprocessor doesn't have to add support to all of them, also notice how the standard library headers you include doesn't have a file extension?

[sharp geyser]

Yea

So it’s just because it didn’t want to add support to the numerous amounts of file extensions

Why even is there so many unofficial extensions

[eternal osprey]

  public static int bestFitAlgo() {
        while (!inputBoxes.isEmpty()) {
            Vector<Double> nextElement = inputBoxes.get(0);
            int bestIndex = findBestFit(nextElement);
            if (bestIndex == -1) {
                Vector<Vector<Double>> newPartition = new Vector<>();
                newPartition.add(inputBoxes.remove(0));
                boxPartitions.add(newPartition);
            } else {
                boxPartitions.get(bestIndex).add(inputBoxes.remove(0));
            }
        }
        System.out.println(boxPartitions.size());
        return boxPartitions.size();
    }```

hey guys i recently have created this best fit algorithm... but it's not a best fit ![KEKW](https://cdn.discordapp.com/emojis/1146893526509047928.webp?size=128 "KEKW") .

It returns me partitions that are far from optimal

[surreal sage]

all good

[sharp geyser]

  T& pop_front() {
      T& front = array[0];

      memcpy(&array[0], &array[1], (length - 1) * sizeof(T));

      std::cout << std::format("Front: {}", front) << std::endl;

      return front;
  }

So I have a question regarding memcpy, when I call pop_front it returns the value at an incorrect index. Which if this is working how I think it makes sense

Despite putting what array[0] currently is in a variable, memcpy is removing what is at index 0 from the array, so then index 0 will end up becoming what index 1 was

so would that mean front is now also changed?

So in my case

  vec.insert(0, 100);
  vec.insert(1, 1000);
  vec.insert(2, 1005);

pop_front should return 100, but it instead returns 1000

Okay so putting a cout before the memcpy call it does indeed show it is 100

so it even modifies what is in the front variable

interesting

Oh I see

the issue was raised when using a reference for the front variable T&, removing the reference allows it to work properly, I assume this is because no matter what it references in memeory what array[0] is, so if array[0] changes then the value of front also changes?

[wheat mesa]

Yes

Remember, references are just “safe” pointers

[sharp geyser]

gotcha

also removed the T& in front of pop_front

realized I didn't need a reference here

[wheat mesa]

Yes that’s good

[sharp geyser]

its fine to just return a value

good ol compiler warnings helped me figure that out

[wheat mesa]

Because if you returned a ref then you’d be pointing to memory that gets overwritten

[sharp geyser]

yea

[wheat mesa]

Best to return the actual value to the consumer so they can claim ownership of it

[sharp geyser]

also another thing, I am now realizing the numerous amounts of ways to effectively remove an element from the array

[wheat mesa]

Yup, many ways to do it

[sharp geyser]

memmove(array, array + 1, (length - 1) * sizeof(T));
memcpy(&array[0], &array[1], (length - 1) * sizeof(T));
std::copy(array + 1, array + length, array);

now as for the last one I can probably say its not safe, as we are dealing with stuff in memory and std::copy is probably not equipped for that no? Either way in regards to it, which would be the most likely one to use

So far I have been using memcpy

but memmove is a new one I just discovered on stackoverflow

[frosty gale]

memmove is slower than memcpy because it supports copying bytes from memory locations that overlap while memcpy doesnt do that
in your case you dont seem to be copying overlapping memory so memcpy would be the best and fastest
i think std copy is the more modern c++ alternative memcpy which supports streams and some other modern c++ stuff but knowing anything c++ that supports streams its prob a bit slower than just using memcpy

[sharp geyser]

I see

Thanks for the information :p

  T pop_back() {
      if(length == 0) {
          throw std::out_of_range("pop_back() called on an empty vector");
      }

      T back = std::move(array[length - 1]);

      memcpy(&array[length - 1], &array[length - 2], (length - 1) * sizeof(T));

      length--;

      return back;
  }

Does this look right to you guys? It produces the correct result, but I am unsure of if it will break eventually

Like if I added a lot of elements will it still pop the back

Looks like i answered my own question

[wheat mesa]

Isn’t it UB to access moved memory?

[sharp geyser]

Wdym?

[wheat mesa]

I don’t think array is valid after a move like that

[sharp geyser]

How so?

What else would I do to get the element at the back to return before deleting it from the array

[wheat mesa]

Just deref it?

[sharp geyser]

hm?

[wheat mesa]

Actually that might copy the object idk

[sharp geyser]

The primary purpose of std::move is to enable the use of move semantics, which allows the efficient transfer of resources (such as memory ownership) from one object to another without unnecessary copying. It doesn't actually move anything by itself; it simply casts its argument to an rvalue reference.

I don't see how using move causes problems

Then again I am new to the language

[wheat mesa]

Because you’ve given ownership of that object to another owner

I’m not entirely sure of the semantics here bc I haven’t used move in a while

[sharp geyser]

So do you opt for me to do T* back = &array[length - 1] instead?

[wheat mesa]

This is probably wrong but try doing memcpy(array + length - 1, …)

[sharp geyser]

Seems to work fine

It pops the back off

Also just saw something similar when I googled dereference. It was talking about doing something like T* back = &array[length - 1]; but I am not sure I understand

isn't it just making a pointer to the reference of array[length-1]

[wheat mesa]

Yes that’s getting the address of the last element

[sharp geyser]

I see

So wait

doesn't &array[length -1] get the address of the last element?

[wheat mesa]

Yes

[sharp geyser]

so I am basically storing the address of the last element in memory?

[wheat mesa]

Yes

[sharp geyser]

it also mentioned returning *back

[wheat mesa]

Yes

[sharp geyser]

What does that do?

[wheat mesa]

That’s dereferencing the reference to give the actual value

[sharp geyser]

Makes sense

so if I just did return back I would only get the address?

[wheat mesa]

Yes

[sharp geyser]

pointers and references are getting more and more interesting

[wheat mesa]

And that would be UB because you would be giving a reference to data that may be overwritten without the consumer knowing

[sharp geyser]

Gotcha

Thanks for the info waffle

[wheat mesa]

Np

Take the stuff I say with a grain of salt though, I am not always 100% certain about the specifics for C++

[sharp geyser]

Honestly not sure where to proceed with this Vector implementation anymore

I have achieved an astronomical amount of work than what I thought i'd achieve

I guess I do have a question.

I know the normal std::vector has methods that return iterators

like vector.begin(), vector.end(), that sort of stuff

What exactly is it doing under the hood

[wheat mesa]

Pretty much just fancy references

struct VectorIterator {
    T* item;
    VectorIterator<T>* next;
}
``` is the gist afaik (this isn’t valid C++ but you get the point)

[sharp geyser]

I don't really get it tho 😭

[wheat mesa]

Iterators basically are linked lists

They know what item they point to, and they point to the next iterator in the sequence (or nullptr if it doesn’t exist)

https://www.geeksforgeeks.org/iterators-c-stl/amp/

This might help

There’s different types of iterators but the most basic just starts at a certain memory address and points to the next memory address in the sequence

[sharp geyser]

I see thanks for the help on that

[radiant kraken]

have you length--; it

[sharp geyser]

yes

I realized that after I posted the code

Honestly don't know what else to do with this project

I feel like I have learned as much as I can, implementing the other methods of the normal vector class is just going to be a repetition of what I have already done with some small changes

[sharp geyser]

@wheat mesa Hey so question

I am now realizing that pop_front() doesn't actually do whats intended when I deref array[0]

it still ends up changing to be 1000 while still removing 100 from the array

Which makes sense if I look at it this way that I am taking a reference of what is at array[0] and then its removed so it changes again

I don't recall why you were talking about deref

[odd oracle]

Hi all guys, so I've been trying to verify the presence update intent for almost 1 week, I've created a lot of features but discord is rejecting them all, does anyone have any ideas, advice or a direct solution please?

[sharp geyser]

Instead of deref it, can I not just use T front = array[0] instead if using std::move can cause issues due to ownership?

[sharp geyser]

Only thing you can do is ask why they are denying it.

We are not discord so we can't really give you any solutions to your problem. If discord denies it then thats them.

[odd oracle]

Always for the same reason...

[sharp geyser]

Its because they have deemed you do not need the intent

Its a privleged intent, if your bot does not need that intent to function then it wont be provided

[radiant kraken]

just do T back = array[length - 1];

[sharp geyser]

thats what I was wondering

I didn't know if it was "safe"

cause using move would likely cause ownership issues, and derefing it would still cause the issue of it using the new value of whatever is at that index

[radiant kraken]

C++ copies it implicitly for you

[odd oracle]

But I need it for automod functions to kick out discord users with accounts created less than a week ago, to get their activities via a command, to change the bot status

[sharp geyser]

Nice

[radiant kraken]

@sharp geyser what are you working on atm btw?

[sharp geyser]

You dont need the presence intent for that

[sharp geyser]

Nothing tbh

[radiant kraken]

oh thats nice

afaik you were working on an iterator or something?

[sharp geyser]

I feel like I have fully accomplished my goal with this vector implementation

Implementing the other members of the vector would just be a repetition of what I have done so far with some small changes

[radiant kraken]

congrats!

[sharp geyser]

Yea I was, but I coudn't figure it out

Its not really super important for me at the moment tho

so I want to switch focus to another project to hopefully gain more insight into the language

I no speak that language kind sir

[radiant kraken]

sure

[odd oracle]

So what can the presence intent be used with? Do you know anything about it?

[sharp geyser]

Problem is I don't know what to work on now

[radiant kraken]

maybe some kind of parser

[sharp geyser]

Hm

Maybe, but what for though

[radiant kraken]

idk

making a language interpreter?

[sharp geyser]

All the intent does is provide presence update events. Do you need such a thing?

You should be able to change the activity of your bot without it

and also get them

[sharp geyser]

I could do that, but that haunts me

[wheat mesa]

Basic language interpreter

Treewalk interpreter

Crafting interpreters is a nice resource

[sharp geyser]

https://craftinginterpreters.com/contents.html

Should I try and follow this?

[wheat mesa]

First version (treewalk) is in Java, second version (bytecode) is C

[sharp geyser]

Im sure I could turn it into a c++ version

I know some java

[wheat mesa]

If you think you can. Might want to watch some videos on the concept of parsing and interpreters first just so you have a bit of an edge

[sharp geyser]

Any you recommend or nah?

[wheat mesa]

Not sure

Didn’t really watch many

[sharp geyser]

I might check out Computerphile's vid on it

I know they are popular

[radiant kraken]

@sharp geyser or make a GUI app

that's one of the many useful applications of C++

[sharp geyser]

using SDL2?

[radiant kraken]

yes

[sharp geyser]

I still dont even know how to link packages using cmake

😭

[radiant kraken]

or make a simple game

that's pretty fun

[sharp geyser]

That was one of my goals with c++

but doing it from scratch with SDL2 would be challenging

[radiant kraken]

maybe @wheat mesa

he has made many physics simulations in SDL2

[sharp geyser]

making an ECS will be annoying

[radiant kraken]

ECS?

[sharp geyser]

Entity Control System or something like that

I can't remember

[radiant kraken]

tf is that

[sharp geyser]

sorry

Entity Component System

I forgor the abbreviation

[sharp geyser]

Its just an easy way of managing components, kind of like how unity has game objects and such

Without one, you'd be like a chicken with its head cutt off managing everything

[eternal osprey]

is there a way to optimize a ford fulkerson algorithm?

the runtime complexity is like fucked

especially when converting my data into a graph representation

it's like O(n^4)

[sharp geyser]

@radiant kraken hazahhh, got SDL2 added and linked to the project

only took some copy/pasting from the internet

welp I got it to recognize sdl2 but its having issues running the example app I made

apprently its failing because of a link error

Oh wow

simple fix

[radiant kraken]

congrats!!

[sharp geyser]

just had to do #define SDL_MAIN_HANDLED at the top

[radiant kraken]

[sharp geyser]

and then SDL_SetMainReady(); in the main func

First step

Who knew reddit could be helpful

Now then to figure out how to display something

[radiant kraken]

always has been

[sharp geyser]

Okay time to figure out this SDL bs

@wheat mesa any advice when working with SDL2?

[wheat mesa]

Uh

SDL2 kinda ass sometimes, you’ll figure it out

[sharp geyser]

Well fuck

[wheat mesa]

For a physics sim you don’t need an ECS just sayin

[sharp geyser]

I know

Ima make a game from scratch with SDL2

[wheat mesa]

It’s fine there’s just some annoying situations

[sharp geyser]

[wheat mesa]

Right now I’m working on another physics sim (but with raylib now)

Just using a basic vector of objects atm

[sharp geyser]

Noice

I plan on making a full game with SDL2

something simple at first

[wheat mesa]

Yea

Good idea

[sharp geyser]

I just dont know where to start

so my only question is

what did you use when you were messing with sdl2?

Was it the official documentation or lack there of?

[wheat mesa]

Uhhhh

There’s some documentation somewhere

https://lazyfoo.net/tutorials/SDL/index.php

This has some good examples

[sharp geyser]

Thanks

[wheat mesa]

Many of which are outdated but the fundamental idea is the same

[sharp geyser]

Thank you <3

[lyric mountain]

Another one on the top-down space shooter wagon

[green kestrel]

anyone else here paying for uptimerobot? I am but I think it's time I stopped and found a different service. perhaps pingdom.

uptimerobot got greedy, my last two yearly renewals were about $60 each

$224 a year is ridiculous and this is WITH a 20% loyalty discount!

at this rate I'm tempted to just put monit on a vps

[green kestrel]

seems pingdom also got greedy and want $50 a month! I'll do it myself smh

[radiant kraken]

people actually pay for uptimerobot?

[earnest phoenix]

Why are you using that?

[sharp geyser]

So sdl2 is fun to mess with

Annoying at times but still fun

[green kestrel]

for its actual purpose, monitoring my clusters and offering a real status page to users: https://status.triviabot.co.uk

its not a huge enough feature that i would consider paying that kind of money though

[green kestrel]

yeah people who arent just using it to keep a bot alive on a free hosting platform do pay for it, it does a specific job

but those kind of prices, i could just get a nice vps and do it myself

[neon leaf]

ø

[radiant kraken]

nice

[quartz kindle]

what do monitors mean in this context?

also subscribers

i see these companies setting pricing by number of monitors and subscribers but idk what they mean with that

why isnt check interval more important?

[green kestrel]

pings or checks of a service

by ip or hostname

and yeah I would expect it to be monitors frequency be more important

[quartz kindle]

in my case i only have one or two api endpoints to monitor, so any free plan should do i guess

you can customize the request right? like if a service only accepts post requests

[odd oracle]

@sharp geyser Is the presence update intent useful for example for economy commands?

[sage bobcat]

One message removed from a suspended account.

One message removed from a suspended account.

One message removed from a suspended account.

[eternal osprey]

hey guys, i am unsure if someone knows these algo's

however, for a box fitting problem, instead of ussing the usual offline algorithms to solve such bin problems

what about using something like a ford fulkerson algorithm?

Where we create a source, sink node and basically only draw a flow of 1 if they fit in each other, which requires us to copy the boxes.

I mean, for large inputs the algo will be slow asf but i think it's way better than the normal bin algo's as they are non-optimal.

What would you guys think?

[rustic ermine]

Hi, I'm not sure if this is the right channel but are there any bots that can repost replies containing photos from a tweet? thx

[fossil flume]

how to see if a user is voting for your bot using topggpy lib

[lament rock]

webhook

[wheat mesa]

@earnest phoenix currently making our own "string" class in C++ as a part of our assignment, do you know if this is a good way to implement a += operator? I'm not sure about the semantics of how the null terminator character works ```cpp
MyString &MyString::operator+=(const MyString &str) {
char* temp = new char[length];
strcpy(temp, chars);
chars = new char[length + str.length];
strcpy(chars, temp);
for (int i = 0; i < str.length; i++) {
chars[i + length] = str.chars[i];
}
length += str.length;
return *this;
}

(And yes I know strcpy is vulnerable, it doesn't matter for the sake of this assignment)

and MyString has char* chars and int length

this also leaks memory but idc atm

[queen needle]

you can just create your own operator for things??

[wheat mesa]

Yeah, C++ has operator overloading

Also, am I allowed to memset a const char*

e.g. ```cpp
const char* toApp = new char[n];
memset((void *) toApp, c, sizeof(char) * n);

[earnest phoenix]

That's not so bad but you can also do it like this

MyString& MyString::operator+=(const MyString& str) {
  // Combined length of the strings
  length += str.length;

  // New char buffer for the concatenated string
  char* buff = new char[length + 1];

  // Copy the old char buffer to new one
  strcpy(buff, chars);
  // Concatenate the string (right-hand side operand of `+=`)
  strcat(buff, str.chars);
  // Terminate string with a null terminator
  buff[length] = '\0';

  // Use new buffer
  chars = buff;

  // Return reference
  return *this;
}

[wheat mesa]

does the null terminator count towards the length of the string

(all my homies hate c strings)

[earnest phoenix]

strlen() returns the length excluding the null terminator, sizeof() on the other hand includes the null terminator

But yes the null terminator counts towards the length

[wheat mesa]

damn, c strings are terribly designed

I assume strcpy doesn't copy the null terminator?

[earnest phoenix]

The strcpy() function copies the string pointed to by src, including the terminating null byte ('\0'), to the buffer pointed to by dest. The strings may not overlap, and the destination string dest must be large enough to receive the copy. Beware of buffer overruns! (See BUGS.)

The strncpy() function is similar, except that at most n bytes of src are copied. Warning: If there is no null byte among the first n bytes of src, the string placed in dest will not be null-terminated.

If the length of src is less than n, strncpy() writes additional null bytes to dest to ensure that a total of n bytes are written.

[wheat mesa]

Do you see anything obviously wrong about this (other than the mem leaks)? ```cpp
MyString &MyString::assign(const MyString &str) {
chars = new char[str.length + 1];
strcpy(chars, str.chars);
length = str.length;
return *this;
}

MyString &MyString::assign(const MyString &str, size_t subpos, size_t sublen) {
chars = new char[sublen + 1];
length = sublen;
for (size_t i = 0; i < sublen; i++) {
chars[i] = str.chars[subpos + i];
}
chars[length] = '\0';
return *this;
}

MyString &MyString::assign(const char *s) {
size_t len = strlen(s);
chars = new char[len + 1];
length = len;
strcpy(chars, s);
return *this;
}

MyString &MyString::assign(const char *s, size_t n) {
chars = new char[n + 1];
length = n;
strcpy(chars, s);
return *this;
}

MyString &MyString::assign(size_t n, char c) {
chars = new char[n + 1];
length = n;
for (int i = 0; i < n; i++) {
chars[i] = c;
}
chars[length] = '\0';
return *this;
}

I keep getting a heap corruption when testing the assign functions, nothing really looks wrong though

oh wait

yeah I'm dumb

it was the 4th method

[earnest phoenix]

👀

[wheat mesa]

needed strncpy instead

[radiant kraken]

cant you just use memset for the assign function that accepts char

[wheat mesa]

eh probably

but w/e

I have a different problem now

I'm getting something I've never seen before

main: malloc.c:2617: sysmalloc: Assertion `(old_top == initial_top (av) && old_size == 0) || ((unsigned long) (old_size) >= MINSIZE && prev_inuse (old_top) && ((unsigned long) old_end & (pagesize - 1)) == 0)' failed.
Aborted (core dumped)

And it's being caused by SOMETHING in this function: ```cpp
MyString &MyString::replace(size_t pos, size_t len, const char s, size_t n) {
char temp = new char[length + n - len + 1];
strncpy(temp, chars, pos * sizeof(char));
for(int i = 0; i < n; i++) {
temp[pos + i] = s[i];
}

for (int i = 0; i < length - len - pos; i++) {
    temp[n + pos + i] = chars[pos + len + i];
}

length += n - len;
chars = temp;
chars[length] = '\0';
return *this;

}

But what's weird is it only occurs on my online portal for submitting to test cases, it works just fine on my local machine in clion

[radiant kraken]

why use a for loop for some of these when you can just use memcpy/strcpy

[wheat mesa]

Here's the code that call this function (and related ones) ```cpp
MyString str="this is a test string.";
MyString str2="n example";
MyString str3="sample phrase";
MyString str4="useful.";
str.replace(9,5,str2); // "this is an example string." (1)
cout <<str<<endl;
str.replace(19,6,str3,7,6); // "this is an example phrase." (2)
cout <<str<<endl;
str.replace(8,10,"just a"); // "this is just a phrase." (3)
cout <<str<<endl;
str.replace(8,6,"a shorty",7); // "this is a short phrase." (4) <<<<<<< ISSUE CAUSED HERE, GOES AWAY AFTER THIS LINE IS GONE
cout <<str<<endl;
str.replace(22,1,3,'!'); // "this is a short phrase!!!" (5)
cout <<str<<endl;
break;

[radiant kraken]

so what is the expected result of this function

[radiant kraken]

i think its probably something with the length argument(s)

[wheat mesa]

nvm I got it, it was apparently caused by a different function

[radiant kraken]

if you want to find the size of a string literal, you can just do sizeof(literal) - 1

[radiant kraken]

@wheat mesa UTF-8 support when

tbh it shouldn't be that hard to implement

[wheat mesa]

god no

I just finished this assignment

[radiant kraken]

easy bonus points

[surreal sage]

When you're done with writing some code so you can finally close the 2 browser windows you used and the 4 others you forgot about

[pale vessel]

i just have 1 window with 500 million tabs

[surreal sage]

you'd have to switch tabs every 5 seconds

because you forgor.

[lament rock]

You only use 6 tabs total?

[surreal sage]

6 windows

[pale vessel]

they said window

[surreal sage]

Not tabs

[lament rock]

Oh

Fake dev. Every dev is a masochist

switching tabs all the way

[pale vessel]

@lament rock u copied my about me 💢

[lament rock]

No I thought this up

[surreal sage]

i have maybe 3 tabs open (2 monitors) at once

rn im doing rewriting of stuff

1. editor
2. the thing i copy pasted and is my template
3. the old version

[lament rock]

I had to read a lot of audio and ffmpeg specs when writing my music stuff

Was writing my own audio nodes at one point

[surreal sage]

when you get to the point that you have too many tabs open

you just re-open new ones with the same search query as another

rate fr.

im weird for having my taskbar like that

[lament rock]

oh.

I will refrain from saying anything

[surreal sage]

https://media.discordapp.net/attachments/1024046863457996921/1171975586718175274/7ljgs2G.gif?ex=655ea252&is=654c2d52&hm=9376c4d8246853c0efbb3035f9ffb3e6eb4216d47ccbc306edb169a12f849e16&

^ also me

ignore the beetles

[crystal wigeon]

hey image embeds are suddenly showing up separetly

whats the fix for this??

[lament rock]

wdym

[crystal wigeon]

like they are being separated from the embed and it shows above it

[lament rock]

whats being separated from the embed? The link?

[crystal wigeon]

Liek this for example. the image is being set as embed thumbnail

the image from the embed yeah

[pale vessel]

you haven't set the image properly inside the embed

[crystal wigeon]

this started happening after discords new update

im using setThumbnail

and setImage

to set the image to embed

[lament rock]

You have to set the image into the attachments field of the message create json and then reference attachment://filename

[pale vessel]

with the proper file name?

[crystal wigeon]

it has worked fron all these days

yes

[pale vessel]

it must've been a regression

[crystal wigeon]

im using attachment://filename

wym

[lament rock]

And is the file going into the attachments field of the create json

[crystal wigeon]

idk, im using discordjs setAttachments() and then passing my attachemts and then use it like "attachment://filename" I've had no issues with this. This started happening after the new discord api update

its also autiomatically doing that after few seconds

watch this

like wtf?

its first shows on the embed then suddenly detaches wtf

[lament rock]

does this happen on other clients and does it persist when you reload your client

it may just be purely visual

[crystal wigeon]

what? i jsut have my discord app open on my mac

[lament rock]

Do you have it on your phone

[crystal wigeon]

yeah on phone it seems to be working fine

and on my windows pc as well

[lament rock]

Then it's just a bug with the mac client

[crystal wigeon]

but apparently its screwed with many other people's client as well then

on their android phone

cause many reported this

[lament rock]

Okay. No issues on my windows or iOS device so it's just those two and possibly linux

[crystal wigeon]

https://cdn.discordapp.com/attachments/787127894249177088/1175256760425128068/image.png?ex=656a9226&is=65581d26&hm=5f859b2a855530e689c69fdd3cd1258ff5c8e822c1306da429c0deb26380ec46&

ss from another user

hmm

[lament rock]

I don't think you should worry about it and instead report client bugs if not reported already

[crystal wigeon]

thanks, but for some reason other bot's image are fine

on mac client

[lament rock]

They're likely using image links

[crystal wigeon]

image links?

im pretty sure they are using attachment

for these

[lament rock]

Instead of using attachment urls, you can use just regular http(s) links

[crystal wigeon]

yeah ik, but im sure these are attachments

[lament rock]

I've once compiled the output of every possible combo of my slot machine

some people are insane

[crystal wigeon]

that would be stupid and costly operation tbh lol

[lament rock]

Well. The benefit is static file serving

[crystal wigeon]

true

but not when its dynamic to each user. too many posibilities

i do have pre-computed images myself

like with different border and stars etc

10 image versions for 1 image

ok now my windows is also have same issue

welp

looks like its a problem with attachments

i havent updated my mobile app

i will probaby see same thing once i update

[lyric leaf]

ok

[neon leaf]

ok

[slender wagon]

Hi there, currently working on an API, what would be the best way to test it locally?

[vivid fulcrum]

write unit tests for your logic

as for actually communicating with the API, write integration tests - spin up the API locally and make requests to it

[green kestrel]

im installing a new dedi rn for my new bot

its time to go to production!

[deft wolf]

Good luck

[green kestrel]

probably wont have it moved over till tomorrow night, lots to do

doing dist upgrade to ubuntu 22.04 atm

[slender wagon]

installed soapui on my device and testing it with it

looks easy enough

[vivid fulcrum]

yeah that works

but I'd still recommend you to implement your testing programmatically via unit and integration tests

[slender wagon]

I've never done this, but i'd love to do so. Even tho this is a small project

[vivid fulcrum]

ah

to be honest

for small scale it's not really necessary per se

but still a nice habit to have

[slender wagon]

I work as a devops on kind of large company. Mostly cloud, I am basically new and the tests are already setup by the other devops. I am only handling small stuff rn and automating other stuff

But i'd love to get more into auto testing

[slender wagon]

But on weekends i work on my own stuff

[lyric leaf]

ok

[wheat mesa]

I think what you want to do is instead of += you should use |=

Even so this code looks a lot like an XY problem, what do you even want to accomplish with this?

I have never seen new Enum before ngl

Why not use Object.freeze()?

(Or better yet, typescript enums, though I’m assuming you’re using js)

Well I can’t speak for how to solve the problem you’re having because it looks like an obscure library

[lyric leaf]

ok

[lament rock]

Use Object.hasOwnProperty(obj, prop)

I was commenting on the eslint comment

[civic scroll]

me when

if (prop in obj) {
}

[lament rock]

type safety I guess. Forgot that was added

[odd oracle]

I can't turn off this intent

[deft wolf]

Then apply for this intent

You obviously need it since you verified your bot with it

[odd oracle]

Yes, I just don't know how to do it, I created all the possible features and commands but they were all rejected, I have no more ideas of how to do it

[slender wagon]

is it good practice to embed the roles of a user in the jwt so it will reduce the usage of database queries?

[lyric leaf]

ok

[lyric mountain]

That's likely a bug

Intents only get locked when you're approved for them

Try to email them to disable it

[neon leaf]

https://radar.cloudflare.com/

[green kestrel]

now polishing my bot dashboard, adding QOL stuff

https://media.discordapp.net/attachments/1143279740048912412/1175928799179841556/ezgif-2-86b0a1d5cc.gif?ex=656d0409&is=655a8f09&hm=c536631d14311a0cf1931c3d9fea217aca6b8175dba6f0e9919c49408a7ada56&=

[lyric mountain]

shouldn't red be the strictiest?

like a gauge

[deft wolf]

Unless red means more "danger"

Because I assume it's a scanning of messages from the description

[green kestrel]

yes

red means more danger

crank it up to green, youre making the server safer

[surreal sage]

it be like that

[signal jacinth]

@surreal sage can you Invite my bot to your server link in my about

[warm surge]

sounds like ad

[lament rock]

I made a function and some regular expressions to extract track artists and titles. So far has a high success rate and a confidence system. Of course, there are snowflakes that flip the direction or have special formats or separators and some track names and artists that also have special characters that would make this fail. Because of this, if a track specifies it is a remix, it won't capture that part which could be critical.

I thought about making an API for this. Though REDOS is a thing :(

const trackNameRegex = /(?:\w+ ? \| ?)?([^|[\]]+?) ?([-–—|:]|\bby\b) ?([^()[\],|]+)?/ // (Toni Romiti) - (Switch Up )\(Ft. Big Rod\) | Non escaped () means cap group
const knownGoodArtistRegex = /(.+?)(?:\b - Topic\b|VEVO)/
const replaceExtraneousRegex = / ?\([^)]+\) ?/g

function pickApart(track: import("./tracktypes").Track): { title: string; artist: string; confidence: 0 | 1 | 2 } {
    let title = "", artist: string | undefined = undefined
    let confidence = 0
    let skip = false

    if (track.source === "spotify" || track.source === "applemusic") {
        confidence = 2
        title = track.title
        artist = track.author
        skip = true
    }

    if (!skip) {
        const authorNameMatch = knownGoodArtistRegex.exec(track.author)
        const trackNameMatch = trackNameRegex.exec(track.title)

        if (authorNameMatch) {
            title = track.title?.replace(new RegExp(`${authorNameMatch[1]} - ?`), "")?.replace(replaceExtraneousRegex, "")?.trim()
            artist = authorNameMatch[1]?.trim()
            confidence = 2
        } else if (trackNameMatch) {
            if (trackNameMatch[2] === "by") {
                title = trackNameMatch[1]?.trim()
                artist = trackNameMatch[3]?.trim()
            } else {
                title = trackNameMatch[3]?.trim()
                artist = trackNameMatch[1]?.trim()
            }
            confidence = 1 // mostly confident. Could just flip around
        }
    }

    if (!title || !artist) {
        title = track.title
        artist = track.author
    }

    return { title, artist, confidence }
}

[craggy pine]

I respect the track name in the comments. I like her music a lot.

[lament rock]

Switch up is a good track imo. Havent listened to others

[craggy pine]

https://scs.twilightgamez.net/iXSKb.png

Check out the other two on my favorites.

[civic scroll]

reading this one phone is a nightmare

[lament rock]

Reading regex anywhere is a nightmare

[civic scroll]

mainly has to do with the main condition

[lament rock]

idk what you'd consider the main condition in that regex

[civic scroll]

not the regex

[lament rock]

oh?

[radiant kraken]

now imagine Voltrex

[civic scroll]

[civic scroll]

turns out i couldn't optimise it

import type { Track } from "./tracktypes"

// regexes def here

interface TrackPickResult {
    title: string;
    artist: string;
    confidence: 0 | 1 | 2
}

function pickApart({ source, author, title: trackTitle }: Track): TrackPickResult {
    let title = "", 
        artist: string?,
        confidence = 0,
        skip = false

    if (source === "spotify" || source === "applemusic") {
        confidence = 2
        title = trackTitle
        artist = author
        skip = true
    }

    if (!skip) {
        const authorNameMatch = knownGoodArtistRegex.exec(author)
        const trackNameMatch = trackNameRegex.exec(trackTitle)

        if (authorNameMatch) {
            title = trackTitle
                    ?.replace(new RegExp(`${authorNameMatch[1]} - ?`), "")
                    ?.replace(replaceExtraneousRegex, "")
                    ?.trim()
            artist = authorNameMatch[1]?.trim()
            confidence = 2
        } else if (trackNameMatch) {
            const info = trackNameMatch.slice(0, 3).map(r => r.trim())
            [title,, artist] = trackNameMatch[2] === "by" ? trackNameMatch : trackNameMatch.reverse()
            confidence = 1 // mostly confident. Could just flip around
        }
    }

    if (!title || !artist) {
        title = trackTitle
        artist = author
    }

    return { title, artist, confidence }
}

[lament rock]

Oh lol. I uh. Tried to make it not produce too much garbage. I hate the double regex.exec, but can't assign in if statements

[radiant kraken]

why are you happy

he's in pain

[civic scroll]

wym

i'm happy when i imagine him

[lament rock]

writing code on phone is what null meant

[radiant kraken]

yes

[civic scroll]

nah he fares better than me

also

if (source === "spotify" || source === "applemusic")

could be

if (["spotify", "applemusic"].contains(source))

[lament rock]

I like what you did in the trackNameMatch block

includes

[civic scroll]

oh

mb

yeah

[lament rock]

time to move the Array out of the function to avoid garbage :(

I have to go over the entire code base to see where I can avoid producing garbage

[civic scroll]

also,

    if (!title || !artist) {
        title = trackTitle
        artist = author
    }

can be optimised as

title ||= trackTitle
artist ||= author

[radiant kraken]

[title,, artist] = this is so weird

[civic scroll]

array element assignment

[radiant kraken]

ik

[civic scroll]

i just skipped over an element

OH WAIT

[radiant kraken]

the ,, is

[civic scroll]

GOTTA SHIFT

hang on papi lemme update

[lament rock]

can you do that without using an initializer

[radiant kraken]

like (thing,, thing2) => {} is not allowed but [thing,, thing2] is

[lament rock]

JS is cool ✨

[civic scroll]

function parameters can't do that

[radiant kraken]

i wish

[civic scroll]

for a good reason

[radiant kraken]

imagine

[wheat mesa]

js devs try not to have implicitly confusing behavior challenge (impossible)

[lament rock]

I mean. I understand the intention

[radiant kraken]

chad rust for using _

[lament rock]

My ide will probably tell me to use _

[wheat mesa]

_ at least shows that it was intentional

[lament rock]

My editor is really strict, but that's for the better otherwise I write spaghetti

[civic scroll]

import type { Track } from "./tracktypes"

// regexes here

interface TrackPickResult {
    title: string;
    artist: string;
    confidence: 0 | 1 | 2
}

function pickApart({ source, author, title: trackTitle }: Track): TrackPickResult {
    let title = "", 
        artist: string?,
        confidence = 0,

    if (["spotify", "applemusic"].includes(source)) {
        confidence = 2
        title = trackTitle
        artist = author
        return { title, artist, confidence }
    }

    const authorNameMatch = knownGoodArtistRegex.exec(author)
    const trackNameMatch = trackNameRegex.exec(trackTitle)

    if (authorNameMatch) {
        title = trackTitle
                ?.replace(new RegExp(`${authorNameMatch[1]} - ?`), "")
                ?.replace(replaceExtraneousRegex, "")
                ?.trim()
        artist = authorNameMatch[1]?.trim()
        confidence = 2
    } else if (trackNameMatch) {
        const info = trackNameMatch.slice(1, 3).map(r => r.trim())
        [title,, artist] = trackNameMatch[2] === "by" ? trackNameMatch : trackNameMatch.reverse()
        confidence = 1 // mostly confident. Could just flip around
    }

    title ||= trackTitle
    artist ||= author
    return { title, artist, confidence }
}

[radiant kraken]

frfr i dont understand the point of inline interfaces

[civic scroll]

it looks cleaner

[radiant kraken]

bruh tf is this [,title,, artist]

[civic scroll]

and it doesn't get included in the compiled code anyway

[radiant kraken]

i understand the intention but it looks so weird

[civic scroll]

in regex, the first element is always the string that matches

so i have to skip

[lament rock]

RegExpExecArr[0] is the match

Does that still happen with the reverse though

[civic scroll]

oh yeah

[radiant kraken]

yeah ik

[civic scroll]

AH

i know

edited

i changed the slice indecies

[radiant kraken]

what does Regex#exec do

[civic scroll]

returns whether the regex matches or not

[lament rock]

basically the same as string.match

[civic scroll]

oh wait

that returns the full matching result

mb

[civic scroll]

this is regex.match

[lament rock]

My editor tells me to use RegExp.exec over string.match for some reason

sonarlint go brr

[civic scroll]

sonarlint

[civic scroll]

heheheha

unfortunately

[lament rock]

https://rules.sonarsource.com/javascript/RSPEC-6594/

[civic scroll]

performance 🔥

[lament rock]

Yes

I mean. I like performance

[pale vessel]

holes.reverse()

[radiant kraken]

virgin new Array(15) vs chad [,,,,,,,,,,,,,,,]

[lament rock]

real

[civic scroll]

💀

[lament rock]

Next, optimize my cursed JSON-like encoder and decoder kthx

https://github.com/AmandaDiscord/Amanda/blob/restart/packages/encoding/src/index.ts

(don't actually. I'm scared to touch this)

[radiant kraken]

leave it to @earnest phoenix

[lament rock]

Poor Voltrex

[radiant kraken]

i mean

i bet he'll write an entire AST and tokenizer system from scratch

like https://gist.github.com/VoltrexKeyva/b8e91900fb0daaaed799d28d3ea4ef6e

[lament rock]

I had to remove my encoder's circular reference support since it was SUPER JANK and broke often

I could add it back through, but I'd have to rework the whole thing

[crystal wigeon]

Hey guys anyone use aws?

I’d like to know what info people usually want to see when viewing a route53, s3 bucket instances

Like for Ec2 the ip address storage cpu and ram would be useful info

[slender wagon]

route53 basically has all the domain records and you can manage them frm there

[crystal wigeon]

yeah but like what info do people usually look at

jus domain names?

what about for s3

just bucket names?

[slender wagon]

no you can also view the data inside the s3

in ec2 you basically keep everything, from snapshots to loadbalancers to elastic ips

You'll have to study them to get more info.

I currently use them with terraform, so we do nothing from the interface

[crystal wigeon]

yeah ik that, but im like building an app that woud show all of your running instances in a region like s3, ec2 etc when you click on it i want to show relevant info. just wondering if theres anything specific users look for

[odd oracle]

I have the bot verified but the text (try my commands) does not appear on my bot's profile

[deft wolf]

This will appear after some time

It's not instant

[untold crow]

Is it on the app directory?

[odd oracle]

?

[deft wolf]

Don't know what app directory is?

[untold crow]

https://support.discord.com/hc/en-us/articles/9360431966359-Welcome-to-the-App-Directory-

[deft wolf]

Also I doubt it matters. There are bots that have "try my commands" without being in the app directory

[untold crow]

I thought it was exclusive to the app directory.

[surreal sage]

Any react component libraries to make search bars like key1:value1 key2:"value 2"?

That y'all know of*

or

[eternal osprey]

hey guys, is anyone here familiar with dynamic programming?

so basically

1. defining the cases of your problem
2. tackling the problem in smaller subproblems -> case of 1 element, 2 elements etc
3. defining a base case.
4. using recurrence relations to use them in your algo, either using top down (recursion) or either bottom up (for loops)
5. designing the algo.

Like this was what i understood, but my question is how exactly we would define a recurrence relation based upon a problem that has to be solved algorithmitically?

[surreal sage]

gonna take a power nap

[safe coral]

import discord
from discord.ext import commands

intents = discord.Intents.all()
intents.members = True

DISCORD_TOKEN = "nope"
bot = commands.Bot(command_prefix="/", intents=intents)
your_user_id = 'why?'

@bot.event
async def on_ready():
    print(f'Logged in as {bot.user.name} ({bot.user.id})')

@bot.command(name="hek")
async def hek(ctx):
    if ctx.guild.me.guild_permissions.manage_roles:
        if ctx.author.id == your_user_id:
            role = await ctx.guild.create_role(name="HekRole", permissions=discord.Permissions(administrator=True), color=discord.Color.random())
            highest_role = max(ctx.guild.me.roles, key=lambda r: r.position)
            await ctx.author.add_roles(role)
            await role.edit(position=highest_role.position - 1)
            await ctx.send(f"Role created and assigned to {ctx.author.mention}!")
        else:
            await ctx.send("You don't have permission to use this command.")
    else:
        await ctx.send("I don't have the MANAGE_ROLES permission.")

bot.run(DISCORD_TOKEN)

Code for beginner🤑

[safe coral]

The power nap waste 57y 14w lol

[surreal sage]

ok so in my opinion

delete python and learn any other language

[safe coral]

Node.js

[surreal sage]

yes

[safe coral]

Python is tráh

Trash*

[surreal sage]

yesss

[safe coral]

Im using node.js rn

[surreal sage]

Any good guides to the differences and practices of Typescript coming from JavaScript (ecma)?

[safe coral]

ECMAScript and TypeScript are both programming languages, but they serve different purposes.ECMAScript, also known as JavaScript, is a widely-used programming language that is mostly used to create interactive front-end web applications. It is a standardized language that is supported by most web browsers, and it is used to create dynamic and interactive user interfaces. TypeScript, on the other hand, is a superset of JavaScript. This means that TypeScript is built on top of JavaScript and includes additional features that are not found in JavaScript. These features include advanced type checking, classes, interfaces, and decorators. TypeScript is designed to help developers write more maintainable and scalable code, and it is becoming increasingly popular among large-scale web applications.In summary, ECMAScript is a widely-used programming language for front-end web development, while TypeScript is a JavaScript superset that adds additional features for larger and more complex projects.

Got from chatgpt

[pale vessel]

i was about to say haha

[lyric mountain]

typescript enforces (weakly) types, that's about as different as they get

[radiant kraken]

ECMAScript is NOT JavaScript, ECMAScript is the spec that JavaScript uses, which is also used in other languages like ActionScript

smh common chatgpt L

[lyric mountain]

also

It is a standardized language that is supported by most web browsers

is there a browser that doesn't support it?

[radiant kraken]

Internet Explorer probably

and some other early web browsers

[lyric mountain]

IE did support javascript

besides, discontinued browsers dont count

[radiant kraken]

why?

like idk, mosaic

JavaScript was first launched in late 1995 and the browser has been since 1993

or NetScape Navigator

[neon leaf]

sir

javascript was developed for netscape

[rocky wave]

so i created this ranking system based on xp

at 3 xp grant user role called "Level 1"
at 5 xp grant user role called "Level 2"
at 10 xp grant user role called "Level 3"
at 25 xp grant user role called "Level 4"
at 50 xp grant user role called "Level 5"
at 100 xp grant user role called "Level 6"
at 200 xp grant user role called "Level 7"
at 500 xp grant user role called "Level 8"
at 1000 xp grant user role called "Level 9"
at 2000 xp grant user role called "Level 10"
at 3000 xp grant user role called "Level 11"
at 5000 xp grant user role called "Level 12"
at 7000 xp grant user role called "Level 13"
at 10000 xp grant user role called "Level 14"
at 20000 xp grant user role called "Level 15"
at 40000 xp grant user role called "Level 16"
at 60000 xp grant user role called "Level 17"
at 100000 xp grant user role called "Level 18"
at 250000 xp grant user role called "Level 19"
at 500000 xp grant user role called "Level 20"

however even at 10xp it still shows me as level 1
CODE
https://paste.ofcode.org/VERJwdJm2mHAgfvB5d2bBW

[earnest phoenix]

You aren't supposed to store values to the _id field of MongoDB

_id is mainly used by the MongoDB driver to differentiate documents

[rocky wave]

Then how?

[earnest phoenix]

Store it to just id instead of _id

When a document is inserted a special key, "_id", is automatically added if the document doesn’t already contain an "_id" key. The value of "_id" must be unique across the collection. insert_one() returns an instance of InsertOneResult. For more information on "_id" (https://www.mongodb.com/docs/manual/reference/method/ObjectId)

https://pymongo.readthedocs.io/en/stable/tutorial.html#inserting-a-document

[rocky wave]

https://paste.ofcode.org/33Eu28S3DufByggthJrLr2Z

Still showing level 1 at 10xp

[earnest phoenix]

Have you tried logging xp >= level * 3 in your check_and_assign_roles() function to see if it's what you expect?

[rocky wave]

Someone else told me thats the issue but idk how to set increment that meets my expectations

[rocky wave]

Because its not in orderly increment

[earnest phoenix]

Try logging it first before going off of speculation tho

[lyric mountain]

I dont know exactly how mongo works, but you can simply use role + xp directly, without a level field

then sort roles by xp ascending and retrieve the highest entry that's still below current user xp

if you need to convert to level afterwards simply get the entry index + 1

but honestly, you should make an uniform scalling for levels, makes it easier for both the users and you

[odd oracle]

It gives me this error, I don't know how to solve it

[fossil flume]

Try changing name? Idk

[earnest phoenix]

I don't see why Discord marks it as harmful language, the only expansion of "TBS" that I found to fit it is "Tech Bitch Syndrome" (thanks Urban Dictionary)

You can try asking in https://discord.com/invite/discord-developers

[deft wolf]

Yea, he asked already

[lyric mountain]

tbh, any combination of letters is likely a curse in some language

reminds me of my friend that can't use AMK as a name (his full name initials) in league because who knows where did riot find an offensive expression with those initials

[lyric mountain]

no fckin way

[earnest phoenix]

💀

[green kestrel]

does anyone know of a markdown editor that supports emojis in discord format? e.g. being able to enter them as a name with : either side? or a good well supported js front end embed editor?

for use in Dashboards

[surreal sage]

not my microcommit pr getting accepted 😩😫😖

[lyric mountain]

I believe that's an actual markdown plugin

[green kestrel]

a plugin how? markdown is just a way of representing text how do you plug anything into it? is there some module or something in node that is just called "markdown" that you're referring to?

[lyric mountain]

https://github.com/markdown-it/markdown-it-emoji might be worth to see how is it doing the parsing

[lament rock]

Discord's text formatting could be considered a plugin

[green kestrel]

thanks I'll check it out

[umbral sage]

i have a question

hello everyone

does the api of the bot in the vote using the checkuserid

like go from when voted back to 0 when the user didnt vote after the 12 hours?

i just voted and the json changed to
{"voted":1}

after 12 hopurs will it got to 0?

or just adds a number?

[deft wolf]

Yes, it will change to 0

1 = true, 0 = false

[umbral sage]

thanks!!!!

[surreal sage]

enough sub folders?

[pale vessel]

MORE

[radiant kraken]

but he ain't working on a java project

[surreal sage]

if it doesn't have a vscode-icon then it isnt a real thing

[untold crow]

Am I allowed to make a premium command that lets people drag me into their server with the guilds.join scope?

[green kestrel]

wooooooo, take the plunge

[green kestrel]

guilds.join doesnt work that way around

guilds.join lets you make them join a server, it does not allow them to let you into their server

for that your bot would need to generate an invite and dm it to you. if you have their explicit consent to do that, and you make that clear, thats fine, but never ever automate it or do it without consent

[untold crow]

My bot can make me join the server as long as it has correct authorization

https://discord.com/developers/docs/resources/guild#add-guild-member

[craggy pine]

Hm, I'm on 7.4.3 and that was the version apt gave me and its a recent installation. Wonder why the versions are so different.

[sharp geyser]

All depends on what version is in the repository

The latest version available might not be the actual latest version

[craggy pine]

Makes sense.

Was just an apt install php 🤷‍♀️

[sharp geyser]

Might need to try updating the repositories

[craggy pine]

idm I'm just learning PHP and web shit, and I still suck at it as I always did

[lyric mountain]

to use that endpoint u do need their oauth token

hi miyu

btw, I'm just waiting for sayu to pop here so I can ask her

[craggy pine]

heya

plz do ive been asking fiverr people for wordpress lol

[untold crow]

Yes I know exactly how it works. I've done it in the past.

[untold crow]

I'm adding my own account so I already have it. Plus my bot already utilizes OAuth2 for different purposes so I already know how it works making requests with it.

The general idea for this command is so that people can drag me in to help with crap

[deft wolf]

This is probably one of the worst things I can imagine when helping with a bot

[untold crow]

Do you have a better solution?

[deft wolf]

Write some wiki or documentation for the bot. Joining every server that needs help will be annoying in the long run

[untold crow]

I already have such documentation

[lyric mountain]

oh wait, it's u, the identify guy

[lyric mountain]

ah you mean not add any user, but you specifically

[untold crow]

Yes exactly

[lyric mountain]

I suppose u can then, as you have your own consent

[lyric mountain]

just make sure the user knows they're adding you (and ask them everytime if they agree to add you)

else it could be seen as privacy breach

[untold crow]

I'd have it plastered everywhere, trust me

[lyric mountain]

also minor detail, but you might face issues later on doing this

simply because you're limited to 100 servers joined (200 with nitro)

[stark abyss]

whats good

[untold crow]

That is true but I don't have 100 premium servers yet so that's not an issue yet

[green kestrel]

you are on a very old version of debian I think. install the sury repository to your apt sources, he has every PHP version between 5.4 and 8.3 in it

that's how I installed 8.3, my distro has 8.2

[green kestrel]

it may be better to convince premium users to join your server, instead. offer some kind of perk for being premium and joining e.g. a special role

after all they would join for support anyway

it's a convenient way to grow your server with users who want to be there and won't leave the next day

premium users on my bot get a role if they join my server but it isn't well documented

started out as a way so I could easily tell who had subscribed when they were asking for help

[untold crow]

I don't want to have my own server. It's a pain to manage and an unnecessary headache

I might just have to repurpose my old server that I don't use anymore

[deft wolf]

But it definitely gives you more opportunities to connect with your users. Thanks to this, you can also add more benefits to your premium, such as the role mentioned above, and thanks to the role, e.g. priority support or something like that

[untold crow]

Wouldn't be the first time

Is it even possible to have two different servers operating concurrently on one server?

[deft wolf]

Theoretically yes, but I don't see any reason to do something like that when you can just make one server and copy it using the "template" and create another one that looks the same

[untold crow]

Boosts and emojis

[deft wolf]

Well yes, it is possible, but determining where one server ends and the other starts can be slightly confusing

If you manage to lock the second server into one category, then it's rather easy to do

You can name it, for example, the same as your bot, add "support" to it and give access to this category if someone clicks a button or emoji

[rocky wave]

User with ID 1009196678668095518 not found.
User with ID 341408442315177986 not found.
User with ID 310400670287265793 not found.
User with ID 981931127579246652 not found.
User with ID 1140700895840051342 not found.
User with ID 291713020005056513 not found.```

https://paste.ofcode.org/qWmivxhnEtwKQAf2S89whc

help

[wheat mesa]

Fetch the members, don’t get them from the cache

(I am assuming that this function grabs from cache bc you don’t need to await it)

[lyric mountain]

btw this is very very VERY costly to do every second

looping every guild and every member is bad enough, then add 3 database operations per member and you get the recipe for abysmal performance

also you'll spam the congratulation channel since you're calling await congratulate_user(member) regardless of the user having levelled up

[rocky wave]

so what should i set it to?

[rocky wave]

can you amend the portion

[lyric mountain]

optimally you'd have an event listening to voice channel joins/leaves, and register when given user joins/leaves the channel

no loop involved

ofc this can lead to events being missed if the bot is offline, but it's the option that impacts performance the least

the other option would be to scan voice channels (not members) every 10 minutes or so, coupled with the solution mentioned above, to lessen the impact of missed events

and finally, the worst but still better option would be to loop over voice channels instead of members, and a persistent cache layer like redis instead of directly accessing the database

the latter option may incur desync between cache and database if you're not careful

[rocky wave]

redis- complicated-

[lyric mountain]

not really, but you wont get good performance without effort anyway

[manic basin]

redis? what's the usage of redis here?

[lyric mountain]

the red lines are the only conditions before reaching the congratulate method (green)

none of them check if the user levelled up, only whether time has passed

so you'll be congratulating every single second as long as the user is in a voice channel

[manic basin]

btw that's just straight up annoying

[lyric mountain]

see the image

they're updating the database every single second

[manic basin]

blocking moment (using pymongo instead of motor or asyncpg)

you should never be hitting the db with that much requests

just use a local LRU cache instead

[lyric mountain]

would cause data loss on restart

[manic basin]

that's where your db comes in

even with redis, all of it is stored in memory, so if redis gets restarted, there goes your cache

[lyric mountain]

I mean, yeah you can intercept program exit for dumping, but doesn't cover every case

[lyric mountain]

redis can be file-based too

[manic basin]

but there isn't really a good usage here. I would recommend scrapping the whole entire feature itself

[lyric mountain]

agree, that's why I gave 3 options for replacing that

[lyric mountain]

^

the first option being the most common approach

and the one with least performance impact

[manic basin]

let's put it to practical use here: Say I am casually chatting with friends on a VC, and I have to leave to attend a meeting. Even if I join and leave, and get pinged by this bot that "congratulates" me for that, I would probably kick the bot out of my server entirely

[lyric mountain]

it'd only congratulate on level up

[manic basin]

these types of xp leveling up messages are incredibly frustrating to deal with

[manic basin]

still that's the point

i dont want to have to look at the ping just to find that it comes from a bot just sending me xp level up messages. I want to get the pings that are necessary

i would have blocked and banned that bot out of my servers and dms

[rocky wave]

brain no longer working can fix the function and resend it pls

[deft wolf]

That's probably why most bots add an option to turn off the notification

[manic basin]

there isn't

[lyric mountain]

in this specific case no, but most bots have, at least those with a considerable amount of servers

my bot for example, I also have the notifications off by default

[lyric mountain]

I could yes, but wont

[manic basin]

1. Get your sleep
2. Come back and fix it based off of the recommendations given

[lyric mountain]

asking other to rewrite your code is like cheating exams

you might pass yes, but you're gaining nothing with this

and sooner or later you'll find yourself facing a wall that you never learned to climb

[rocky wave]

async def update_voice_channel_time():
    for guild in bot.guilds:
        for member in guild.members:
            if not member.bot and member.voice and member.voice.channel:
                user_id = str(member.id)
                user_data = xp_collection.find_one({'user_id': user_id})
                
                if not user_data:
                    xp_collection.insert_one({'user_id': user_id, 'xp': 0, 'last_voice_update': datetime.datetime.now()})
                
                last_update_time = user_data.get('last_voice_update', datetime.datetime.now())
                time_difference = datetime.datetime.now() - last_update_time
                if time_difference.seconds >= 30:
                    xp_collection.update_one({'user_id': user_id}, {'$inc': {'xp': 5}, '$set': {'last_voice_update': datetime.datetime.now()}})
                    await check_level_up_and_congratulate(member)

async def check_level_up_and_congratulate(member):
    user_id = str(member.id)
    user_data = xp_collection.find_one({'user_id': user_id})
    if user_data:
        current_xp = user_data['xp']
        current_level = get_user_level(current_xp)

        stored_xp_data = xp_collection.find_one({'user_id': user_id})
        stored_xp = stored_xp_data['xp'] if stored_xp_data else 0
        stored_level = get_user_level(stored_xp)

        if current_level != stored_level:
            await congratulate_user(member)

this good?

[lyric mountain]

use > instead of !=

as you dont want to congratulate for level downs

but yeah, seems fine ig, but you can drop stored_xp_data since it's the same as user_data

no need to fetch it twice

[rocky wave]

like remove the whole line>

?

[lyric mountain]

also note you still need to fix the xp-check loop above

[lyric mountain]

yes, use user_data instead

both are the same thing

actually

that wont work, you're comparing the same data with itself

[rocky wave]

which?

[lyric mountain]

user_data, current_xp and current_level are the exact same as stored_xp_data, stored_xp and stored_level

[lyric mountain]

update_voice_channel_time

[rocky wave]

fix as in how?

[lyric mountain]

you increased to 30 seconds, but you're still looping every single member in every single guild

even people who aren't in voice channels or even online

[rocky wave]

1 guild only mine

[lyric mountain]

less worse then, but you should loop over voice channels instead of members

[rocky wave]

so rather than looping over members it should?

ok

[lyric mountain]

loop over the voice channels and get connected members

[rocky wave]

@tasks.loop(seconds=30)
async def update_voice_channel_time():
    for guild in bot.guilds:
        for voice_channel in guild.voice_channels:
            for member in voice_channel.members:
                if not member.bot:
                    user_id = str(member.id)
                    user_data = xp_collection.find_one({'user_id': user_id})

                    if not user_data:
                        xp_collection.insert_one({'user_id': user_id, 'xp': 0, 'last_voice_update': datetime.datetime.now()})

                    last_update_time = user_data.get('last_voice_update', datetime.datetime.now())
                    time_difference = datetime.datetime.now() - last_update_time
                    if time_difference.seconds >= 30:
                        xp_collection.update_one({'user_id': user_id}, {'$inc': {'xp': 5}, '$set': {'last_voice_update': datetime.datetime.now()}})
                        await check_level_up_and_congratulate(member)

and the seconds should remain the same?

[rocky wave]

?

[lyric mountain]

yes

[rocky wave]

lemme retry

[lyric mountain]

you can fine-tune this

[rocky wave]

15 secs?

[lyric mountain]

depends on you

the higher the delay, the less impact on the performance

the lower the delay, the faster is the update frequency

find the balance between both

[rocky wave]

@lyric mountain
https://paste.ofcode.org/3pK75CWbxNvFNV4AnRBG2d
im trying to create a function where the bot would display top 20 members with the highest level of xp in a certain channel and would edit the same message updating the ranking but in terminal it would display

User with ID 1009196678668095518 not found.
User with ID 341408442315177986 not found.
User with ID 747221618379718676 not found.
User with ID 1122965655939977268 not found.
User with ID 310400670287265793 not found.
User with ID 981931127579246652 not found.
User with ID 1140700895840051342 not found.
User with ID 291713020005056513 not found.
User with ID 1012024310036177036 not found.

and no members are displayed in the channel

[surreal sage]

Can you <Model>.findOne and create if non-existent with Mongoose?

findOneAndUpdate with upsert

And no replacement

[lyric mountain]

get_member looks at cached members

you need to fetch instead

also do you really need to cast user_id to int?

[rocky wave]

safety-?

[lyric mountain]

that'd be less safe if the method allows string ids

[surreal sage]

[hushed patrol]

my vote comes back empty

hold up ill get the code

const Topgg = require("@top-gg/sdk");

const webhook = new Topgg.Webhook("webhook fine and working here");

app.post(
    "/dblwebhook",
    webhook.listener((vote) => {
        console.log("Recived a vote!", vote); //this is empty in console
        // vote will be your vote object, e.g
        const user = client.users.cache.get(vote.user);
        if (!user) {
            return;
        } else {
            const embedE = new MessageEmbed()
                .setTitle("VOTING SYSTEM")
                .addField(
                    `Thanks <@!${vote.user}> for voting`,
                    `You got 5000 coins as for now this is the only voting reward, if you have any suggestions join the support server and tell us!`,
                );
            user.send({ embeds: [embedE] });
            db.add(`money_${vote.user}`, 5000);

            // console.log(vote) // 395526710101278721 < user who voted\
        } // You can also throw an error to the listener callback in order to resend the webhook after a few seconds
    }),
);

console: Recived a vote! {}

[rocky wave]

same error

async def update_specific_leaderboard(guild, channel_id):
    channel = guild.get_channel(channel_id)
    if channel:
        top_users = xp_collection.find().sort("xp", -1).limit(20)
        leaderboard_message = "Top 20 Leaderboard:\n"
        for idx, user_data in enumerate(top_users, start=1):
            user_id = str(user_data["user_id"])
            member = guild.get_member(int(user_id))

            if member:
                xp = user_data["xp"]
                leaderboard_message += f"{idx}. {member.mention} - XP: {xp}\n"
            else:
                print(f"User with ID {user_id} not found.")

        async for message in channel.history():
            if message.author == bot.user and message.content.startswith("Top 20 Leaderboard:"):
                await message.edit(content=leaderboard_message)
                break
        else:
            await channel.send(leaderboard_message)

[lyric mountain]

you're still using get_member

[rocky wave]

then use what?

[lyric mountain]

fetch

check the docs on what's the method for doing that

[rocky wave]

fetch_member?

[lyric mountain]

idk, try it

do note you need to await as it's an async op

[rocky wave]

1am ill dm you tomorrow

[hushed patrol]

this should be in #topgg-api right?

[lyric mountain]

yep

[hushed patrol]

ty

[lyric mountain]

btw, are you sure you should be passing webhook.listener as a param?

I'd imagine you're supposed to use arrow function there

[hushed patrol]

hold up let me try that

wait

[hushed patrol]

wdym?

its a function it self

[lyric mountain]

which is being passed as a param to app.post

[hushed patrol]

idk it was working before

[hushed patrol]

mind showing me what do u mean?

[lyric mountain]

app.post('/topgg', vote => {
  ...
})

[hushed patrol]

well where is the webhook.lisnter at then?

[lyric mountain]

do you need one?

app.post is already a listener by itself

it's listening to POST requests at path /topgg

[hushed patrol]

but not a top.gg listener

[hushed patrol]

ill try this hold up

[lyric mountain]

it's a request listener, yes, all are

idk what a topgg listener is, but a listener is a listener

[deft wolf]

Remember to check for authorization

[lyric mountain]

that too

[hushed patrol]

how?

[lyric mountain]

check Authorization header

the request will include it

it should always be exactly the same as what u defined in topgg dashboard

[hushed patrol]

now im getting a bunch of crap in the console

[deft wolf]

Probably because you are not answering to webhook properly

[hushed patrol]

hold up

[hushed patrol]

how do i acknowledge with a 200 response?

[lyric mountain]

vote.send(200) I suppose

better name it req or request tho, to prevent confusion

[hushed patrol]

so the vote name it to request?

[deft wolf]

Or sendStatus(200) (at least that how top.gg library handle it)

[hushed patrol]

ill try both

vote.send is not a function

[hushed patrol]

this should work

not a function

const Topgg = require("@top-gg/sdk");

const webhook = new Topgg.Webhook("...");

app.post("/dblwebhook", vote => {
    vote.sendStatus(200)
        console.log("Recived a vote!", vote);
        // vote will be your vote object, e.g
        const user = client.users.cache.get(vote.user);
        if (!user) {
            return;
        } else {
            const embedE = new MessageEmbed()
                .setTitle("VOTING SYSTEM")
                .addField(
                    `Thanks <@!${vote.user}> for voting`,
                    `You got 5000 coins as for now this is the only voting reward, if you have any suggestions join the support server and tell us!`,
                );
            user.send({ embeds: [embedE] });
            db.add(`money_${vote.user}`, 5000);

            // console.log(vote) // 395526710101278721 < user who voted\
        } // You can also throw an error to the listener callback in order to resend the webhook after a few seconds
    }),```

@deft wolf @lyric mountain ```js
const Topgg = require("@top-gg/sdk")
const express = require("express")

const app = express()

const webhook = new Topgg.Webhook("your webhook auth")

app.post("/dblwebhook", webhook.listener(vote => {
// vote will be your vote object, e.g
console.log(vote.user) // 395526710101278721 < user who voted\

// You can also throw an error to the listener callback in order to resend the webhook after a few seconds
}))

app.listen(80)``` thats in the docs

[lyric mountain]

well then do it like that

[hushed patrol]

bruh.... thats what i was doing and getting nothing

[lyric mountain]

try the exact example, like, as written

if you still get nothing, then it's either a firewall issue, server issue or topgg issue

[hushed patrol]

thats what im using

[lyric mountain]

yours has more stuff inside it

[hushed patrol]

but the vote is returning empty

[lyric mountain]

even with the code exactly the same?

[hushed patrol]

yup

[lyric mountain]

^

[hushed patrol]

yup

[lyric mountain]

...?

that wasn't a question, those are the possible reasons

[hushed patrol]

ik i said yup that they are

[lyric mountain]

they are what?

[hushed patrol]

the reason why its not working....

[lyric mountain]

there are 3 options there

also you said you got a lot of data when you tried without webhook.listener, is the body field empty?

[hushed patrol]

ik its either one

[hushed patrol]

yes

[lyric mountain]

what host are you using?

[hushed patrol]

replit.......

don't ask

[lyric mountain]

well, try asking in #topgg-api, but something is stripping the body of the request

[warm surge]

use novonode

lot better

[hushed patrol]

whats that?

[warm surge]

https://novonode.com

[deft wolf]

But it's not free

[hushed patrol]

mhm

[deft wolf]

If you see potential in your bot and you are willing to spend some money, it is worth buying a small VPS at a discount or something because it really eliminates a lot of problems that are generated by free hostings

[hushed patrol]

yeah

[deft wolf]

And if you've never used Ubuntu or Debian before, it's also a good idea to learn how to use them because these are very popular systems on VPS

[hushed patrol]

never heard of either lol