#🪅-progaming
1 messages · Page 54 of 1
then i can do this when compiling
and use some macro to condense this down to a single line afterwards
because every variant does the same thing
in the case of your interpreter imagine you want to dispatch a function for each variant of your astnodes in your Evaluator
just do this and let res = self.compile_res_or_something(...)
*without using macros because rust-analyzer has a seizure
yeah ill do this after i implement the Codegen trait for every node

ill have a big stream of these things
and then condense them all down to 1 line
looking at this makes me want to make a compiled language
lmao
yeah you have the same issue actually
the issue with creating structs like this is that you cant pass them around
what happens when you dont want to destructure and just want to capture the struct and pass it along to another function
you cant
which is why for a while elle had a huge single file, each node had an implicit struct attached to it which you can only access through matches
because passing around every field to a function is ugly
got, netacad/cisco is fucking retarded https://www.netacad.com/content/i2cs/7.1/courses/content/m5/en-US/components.json
and this way with the trait is much nicer imo
function queryDeepSelectorAll(selector, rootNode=document.body, all=true) {
const arr = []
const traverser = node => {
if (!all && arr.length) return;
// 1. decline all nodes that are not elements
if(node.nodeType !== Node.ELEMENT_NODE) return
// 2. add the node to the array, if it matches the selector
if(node.matches(selector)) {
arr.push(node)
if (!all) return;
}
// 3. loop through the children
const children = node.children
if (children.length) {
for(const child of children) {
traverser(child)
}
}
// 4. check for shadow DOM, and loop through it's children
const shadowRoot = node.shadowRoot
if (shadowRoot) {
const shadowChildren = shadowRoot.children
for(const shadowChild of shadowChildren) {
traverser(shadowChild)
}
}
}
traverser(rootNode)
return all ? arr : ( arr.length ? arr[0] : null );
}
function queryDeepSelector(selector, rootNode=document.body) {
return queryDeepSelectorAll(selector, rootNode, false);
}
const res = await fetch('https://www.netacad.com/content/i2cs/7.1/courses/content/m6/en-US/components.json')
const json = await res.json()
for (const el of queryDeepSelectorAll('input', document.querySelector('[class*=\'coursePrimaryContainer\'] iframe')?.contentDocument?.body || document.body)) {
if(!el.id) continue
const [id, index] = el.id.match(/[^-]+/g)
const { _items } = json.find(item=>item._id === id)
if(_items[index]._shouldBeSelected) el.click()
}
add a domain seized information at the top
brotha its netacad's own domain
idk what that is
CISCO's certificate/course exam website
you mean you want a function that accepts specifically ExpressionKind::Unary?
cisco like cisco jabber?
no, cisco as in one of the biggest networking hardware manufacturer and distributor
is this on the site that gives you the cerfiticate
omw
its this link
which has 0 protection on it
not even a logged in check
is that a yes or a no
thats a yes
im not looking at a billion lines of raw html
lmao thats crazy
yeah
its an image
client-sided exam verification is wild
esp considering it gives u a certificate
💀
which then gives u cisco hardware discounts]
which itself is a scam but w/e
technically yes
i want a function which accepts the struct
so i can use its fields
with your approach literally the only way to do that is to pass each field to the function manually
this is way too ironic
havent found an easy way to get the course ID
but that script auto-fills all the exams for the one course
dis1
which is peak content
i mean its just awful security and design
you're finally splitting your compiler?
yes lol
will take a few days because there are so many nodes to refactor
so far i have this many
but this isnt nearly everything yet
because after i split the compilation of nodes i have to split the qbe things
but urs seems kinda verbose
idk tho
gn
tl;dr
enum Foo {
Bar { a: u32, b: u32 },
Baz { a: u64, c: u64 }
}
fn process(foo: Foo) {
match foo {
// Too much destructuring, sucks, process_* functions need more parameters as the enums are extended
Foo::Bar { a, b } => process_bar(a, b),
Foo::Baz { a, c } => process_baz(a, c)
}
}
// vs
enum Foo {
Bar(Bar),
Baz(Baz)
}
fn process(foo: Foo) {
match foo {
// Easier, delightful to the eye, grouped with the struct process func
// assuming each struct as a process function, ideally a trait
Foo::Bar(bar) => bar.process(),
Foo::Baz(baz) => baz.process(),
}
}
just plain better
cybersec certification lacking any form of security LOL
its pretty simple
ya irony is trully strong with this one
gn
if youre writing rust verbosity is your least concern
the language itself is verbose af
this is just objectively worse as the number of fields in the struct increases
yop
wtf wait
clicking on a message deletes it 😭
this is probably a vencord plugin
huskkkkkkkk
does anyone have a message logger to recover that code i deleted ugh
yeah it is
evil
click to delete is so evil
which segment
1:
struct Foo { x: i32 }
enum Node {
Foo(Foo)
}
// --------------------
match node {
Node::Foo(this) => this.compile(self, ctx)
}
// --------------------
impl Codegen<'_> for Foo {
fn compile(self, gen, ctx) { ... }
}
2:
enum Node {
Foo { x: i32 }
}
// --------------------
match node {
Node::Foo { x } => compile_foo(self, ctx, x)
}
// --------------------
fn compile_foo(gen, ctx, x: i32) { ... }
thank you !!!!
yeah that one is pretty broken
whoever decided click to delete is sane when you double click to edit
youre not invited to my birtjday party
@valid jetty are you like really really really good at rust
are you still trying to do that propability thing
no
i wanna see if a code can be optimized to hell and back to not use 2k% of cpu
its Sieve of Eratosthenes + exponentia + primes
Gn
anyways if anyone wants to optimize this code for n>2k it would be greatly appreciated
use dashu_int::UBig;
use rand::distributions::Uniform;
use rand::{thread_rng, Rng};
use std::time::Instant;
/// Sieve of Eratosthenes to generate prime numbers up to a given limit.
fn sieve(limit: usize) -> Vec<UBig> {
let mut is_prime = vec![true; limit];
is_prime[0] = false;
is_prime[1] = false;
for i in 2..limit {
if is_prime[i] {
let mut multiple = i * i;
while multiple < limit {
is_prime[multiple] = false;
multiple += i;
}
}
}
is_prime
.iter()
.enumerate()
.filter(|&(_, &prime)| prime)
.map(|(i, _)| UBig::from(i))
.collect()
}
/// Function to check if `x` is not divisible by any primes.
fn is_valid(x: &UBig, primes: &[&UBig]) -> bool {
primes.iter().all(|&p| x % p != UBig::ZERO)
}
/// Computes the percentage of valid numbers in range `2..=2^n`.
fn fun(n: usize, primes: &[UBig]) -> f64 {
let start_time = Instant::now();
let upper_bound = UBig::ONE << n;
//let space = &UBig::from(2u8)..=&upper_bound;
let primes = primes.iter().take(n as usize).collect::<Vec<_>>();
assert!(primes.len() == n);
let sample_size = 1000000usize;
let (count, percentage) = if upper_bound > UBig::from(sample_size) {
let between = Uniform::new(UBig::from(2u8), &upper_bound);
let mut rng = thread_rng();
let mut count = 0;
for _ in 0..sample_size {
let x = rng.sample(&between);
if is_valid(&x, &primes) {
count += 1;
}
}
let percentage = 100.0 * (count as f64) / (sample_size as f64);
(count, percentage)
} else {
let mut count = 0;
let mut x = UBig::from(2u8);
while x <= upper_bound {
if is_valid(&x, &primes) {
count += 1;
}
x += UBig::ONE;
}
let percentage = 100.0 * (count as f64) / (upper_bound.to_f64().unwrap() - 1.0);
(count, percentage)
};
println!("n: {}, count: {}, percentage: {:.2}%", n, count, percentage);
println!("Time elapsed: {:.2?}", start_time.elapsed());
percentage
}
fn main() {
let primes = sieve(1000000);
let xs: Vec<usize> = (1..1000).collect();
let ys: Vec<f64> = xs.iter().map(|&n| fun(n, &primes)).collect();
}
[package]
name = "msg"
version = "0.1.0"
edition = "2021"
[dependencies]
dashu-int = {version = "0.4.1", features = ["rand"]}
rand = "0.8"
^ the percentage decrease also is not too similar to the graph i've experimented so something might be off
is_valid can be optimized
use !primes.iter().any(|&p| x % p == UBig::ZERO)
shortcircuits
so if you have like
[0, 0, 0, 2, 0, 0, 0, 2]
its only gonna go up till the first 2 before it returns
as opposed to .all which will go all the way to the end
i dont think this is her goal
she wants to not use an enum at all
yes she is
what
i think she want a separate struct for each astnode
this confuses me a lot in rust ngl
its quite simple
once you start doing stuf you just start understanding it
people might say lifetimes are hard but they are really simple
you have to imagine a graph
a relationship graph of how a value lives
rust is kinda hard to code in like i spent an hour to make my function work due to constant type mismatches of errors 
in zig it’s possible to use !MyType as a return type which basically means the function can return an error but we don’t care what type it is
ig in rust you can achieve that with -> Result<MyType, Box<dyn std::error::Error>>?
zig would be great if they had rust-like (proc)macros
atleast they're adding async back now
thats how you do it if you dont care what type the error is, but thats kinda bad practice, you really wan explicit error types
also that dyn is confusing too like isn’t it basically -> Result<MyType, Box<T>> where T: std::erorr::Error?
erorr
sorry my ide is not that good i think
with dyn you dont have a concrete type, whereas with a genericc you have a concrete type
fn process_err_vec<T: std::error::Error>(errs: Vec<T>) { ... }
fn process_err_dyn_vec(errs: Vec<Box<dyn std::error::Error>>) { ... }
process_err_vec(vec![SomeError, SomeError]); // Works
process_err_vec(vec![SomeError, OtherError]); // Will NOT work
process_err_dyn_vec(vec![Box::new(SomeError), Box::new(OtherError)]); // Works
@nimble bone i made 5 of my classmates switch to bun
spreading the broken runtime
it would be better if they didn't document shit that doesnt work or exist
oh
i see i think
zig decided to not document ANYTHING 😭
documentation is the cowards way out
real developers love scouring through codebases
ok it has some but LOTS of stuff in std lacks documentation
zig has docs
not the best
but they do
well the language itself is rather incomplete
so whatever
let me get a specialization phd in compilers and language dev so i can read some obscure code base to print stuff to stdout
zig would be easily my new best lang if
- they loosen their shadowing restrictions
- add rust-like macros (esp proc)
- release async
wdym by first? i assume it’s something related to the fact that for example compiler will basically not let you compile the program if you haven’t used a return value of a function (why)?
their website uses no macros as a selling point
shadowing means using identifiers with the same name
yeah i just assumed it was some stupid ass restriction like this one:
you can’t just do MessageBoxA() you need to do _ = MessageBoxA()
fn foo(): void { ... }
pub fn main(): void {
const foo = 123; // doesnt work
}
still the computation is pretty big for n>2k
could there be a way to implement CUDA for better gpu usage for computation?
lmao this reminded me of this code sample i saw on x the everything app 😭😭😭 https://x.com/yutongwu111140/status/1905964811743048046
also idk i kinda get this one
shadowing is also nice to change the type of a variable without defining a new identifier
let foo = "hello"; // &'static str
let foo = foo.len(); // usize
fyi the colons are odd here
nah the compiler should be smart enough to tell the two apart
had no syntax hl so i fumbled
reminder i have written like just a couple hundred lines of zig
not the most
let me see if i can do this in kotlin
yeah zig really really hates shadowing
yeah like i don’t like this aspect of zig too where absolutely unnecessary restrictions are forced on compile level
“no i won’t compile your code because you named your variable wrong”
/run
fun deez() {}
fun main() {
val deez = "nuts"
}
@frosty obsidian I received kt(1.8.20) compile errors
file0.code.kt:4:9: warning: variable 'deez' is never used
val deez = "nuts"
^
compiles fine
wtf is val- oh it’s kotlin
you can use rayon here to add some parallelization, and there are also some other things you can do to optimize this code like not looping over multiple times insieve
zig would cry and weep at this
insane
i think if a language has to say nuh uh to me for doing something normal then its just not for me
zig users would go great lengths to defend this
thats for like C style macros
where its easy to mess up
rust macros are just the best, and even if you output invalid syntax it atleast tells you, unlike c, where depending on how silly the invalid syntax it you might end up with UB or not compiling
https://github.com/naasking/async.h this thing here uses macros
one slight mistake and now your entire control flow is fucked
contrary to this in rust (you'd use procmacros) you simply wouldnt

rust macros are so fucking hard to read
you dont need to read macros
just use them 
now i might say macros are not hard but i barely have a life and im mentally unstable who loves writing macros, so...
but when i try to understand how some app works this becomes a nightmare
#define KERNEL_LOG(...) \
kprintf("[%s::%s]: ", __FILE__, __func__); \
kprintf(__VA_ARGS__)
#define CHECK_STATUS(status, func_name, interrupt_way) \
switch (status) { \
case STATUS_OK: \
break; \
case STATUS_ERROR: \
KERNEL_LOG("%s - failed!\n", func_name); \
interrupt_way; \
case STATUS_NOT_IMPLEMENTED: \
KERNEL_LOG("%s - not implemented!\n", func_name); \
break; \
case STATUS_BROKEN: \
KERNEL_LOG("%s - broken!\n", func_name); \
break; \
}
i’m plausibly decent
oh hm
\s my beloved
it looks nice in 80 columns
rayon would slash the time in like 8 so it's not a huge difference for very big numbers
ig ill just wait until there is a supercomputer that's with like a 9070ti or something, extrapolating isn't really good because there isn't a general formula to go by and who knows maybe there's a huge dip somewhere in the graph
GUHHHHHHHHHHHHH
Bug Description: Detours allocates IMAGE_NUMBEROF_DIRECTORY_ENTRIES (i.e., 16) entries in the m_SectionHeaders array (https://github.com/microsoft/Detours/blob/master/src/image.cpp#L260). 16 is the...
threads, simd, etc etc you can throw everything possible in there to make it faster
or simply use a more efficient algorithm
idek what you're trying to achieve here
# kil
# these will hold things later I just want to make the errors go away
# search for a thing to eat
galaxydim = 40000000 # in LIGHT DAYS. the milky way is about 36500000 LD across but we "round up"
# if we want to return the array, for some reason
# this is probably horrible but i don't care
# --- CONFIG todo: AUTOMATE THIS *facepalm*
# it's not a star but who cares
# player physics (very simple)
running = True # lol
# all hail bep
# in the off chance i do a kilo of meth and decide to implement 3D
# declared global because i have to for some reason
# declare int out of paranoia
# all of this math is probably unnecessary but i can't be fucked to rewrite it
# make the universe clock tick, don't touch this
# horror
# what kind of thing are we?
# yes, changing the type so late is dangerous but it doesnt matter here
# handle the awful twelve hour format now
# begin worldgen functions
# these shouldn't be touched at all, even by me
# if all checks have succeeded...
break # leave this hell
if num < 0: # there's a warning here but there is no normal case in which it matters
comments from my code
the ratio of numbers not divisible by the first n primes in the interval [2, 2^n]
why light days and not ly/kpc
at least i get this right 🎀
im refusing criticism from someone who is dealing with this kind of math
shadowing.le: Lines 1-38
use std/prelude;
fn main() {
x := "foo";
$assert(x == "foo", nil);
x := 1;
$assert(x == 1, nil);
{
x = 2;
x := "a";
$assert(x == "a", nil);
{
x = "b";
$assert(x == "b", nil);
}
$assert(x == "b", nil);
}
$assert(x == 2, nil);
// Redeclare a new `y` using `:=`
y := 13;
y := y * 3;
$assert(y == 39, nil);
// Redeclare a new `z` using `=`
z := "foo";
z = z <> "bar";
z <>= "baz";
$assert(z == "foobarbaz", nil);
io::println("All `variable shadowing` tests have passed!".color("green").reset());
}
this is literally math for fun
what even is a kpc
kiloparsec
kiloparsec?
????
even if you want to use light days, use AU instead
im using light days because i like them
what. font is this
comic shanns mono
i knew it wasn't straight comic sans

yea but you should use SI
use meters
it stands for kilo parallax seconds
one parsec is 3.26ly
it’s a more convenient measurement especially when working with things very far away
i genuinely can't even tell if you're making shit up
no im serious
actually light is a wave, no speed
it might be april fools but i am 100% serious
yes 😭 but my point still stands
no rust 💔
you should rewrite in rust
lc.wa 1 kpc to ly
Because they have eight wheels and four people on them, and four plus eight is twelve, and there are twelve inches in a foot, and one foot is a ruler, and Queen Elizabeth was a ruler, and Queen Elizabeth was also a ship, and the ship sailed the seas, and in the seas are fish, and fish have fins, and the Finns fought the Russians, and the Russians are red, and fire trucks are always "russian" around
@dawn ledge if you rewrite what i have so far in rust, ill write the rest in rust 

i'm writing a game to explain how scale is terrifying
it just needs to have consistent and sensible game mechanics
did you implement F = G m1m2 / r^2
did you implement
comment goes hard
quit yapping greek
i can only understand just one variable here ℏ
what about implementing the fundamentals of an n body orbital system, accounted for gravity and eccentricity and the random supernova every 1 billion cosmic years, the Roche limit, lagrange points
the universe would be running at 1 fps if it was a computer today
i can
i've been fighting scope creep for this project very hard, don't tempt me
universe written in rust
i could probably write a sort of module for my game that makes the stars drift around and shit but im going for static stars for now
but stars do drift tho
in assembly
very aware of this
whenever will you implement yarkovsky o keefe radzievskii paddack effect
mathematicians are not real
the core of the game is psychological horror
i will give you more
psychological horror when a 6.76 sun mass star randomly goes supernova
@leaden crater write a program that can detect whether another program will infinitely loop or stop
i forgot femtoseconds were a unit for a nanoangstrom
i cant code a new turing machine
is this one of those seemingly simple problems but it's actually computationally impossible or something
yes
i'm in CS 1 rn so go easy on me
what's so hard about it?
im slowly coding up a solar system like in universe sandbox 2
oh boy several people are typing
anythhing that doesnt quit in 2nanoseconds is an infinite loop raaaaaaaaaaaaaaaaaah
/gen?
it was proven mathematically that it’s impossible, if you have a machine which contains itself
time complexity O(1) or halt everything
yea but it's fractal dimension (i am 0.0001% done)
Btw I tried to make a version in rust but I realized it would be a pain in the ass to get the cef headers working and link against libcef_dll_wrapper for the v8 api
epitome of fake programming when you use a config.cfg file instead of variables
config dot config
Pass it into itself

a programm that never halts doesn't exist because the machine on it will eventually die out
qed
assume no degradation of the semiconductor and infinite energy flow
what now :^)
let’s say you hypothetically create a machine which halts if the program loops infinitely, but loops infinitely if the program halts, call it A
if you create a new machine B which uses machine A, then pass that back into machine A, what’s gonna happen?
by definition machine A will halt if the machine never halts, but machine B passes the output from machine A into itself, so it does loop infinitely
therefore it’s not possible to predict whether a machine halts or not
assume the great attractor is near earth
assume the world is gonna end
NOW
i saw a physics problem that was like “calculate the freefall of this cow. assume g = 5, the earth is flat, the cow is a perfect sphere, there is no air resistance, there is no drag coefficient”
(yes g = 5)
it sure as heck wasn’t 9.81
avg physics question
assume the cow is a fucking sphere
well with the cow being a perfect sphere we are breaking the laws of farming
damn not even 10, straight up 5
assume g = pi^2
Ol’ Betsy’s laws
me when math
i went from trying decrypting usm files to accidentally finding primes using bash
- cows must be plump but do NOT make them perfect spheres you stupid mathematician idiot
- collect cow milk in a tub but do NOT find the perfect shape to safely hold the most volume that’s just stupid
- B = gmc (Bfood = gravitational field strength * my cow)
Ol’ Betsy’s laws of farming
have you tried printing using a pen and linear
how much CH4 can be in a vessel of 1 cow volume assuming 22.4 L /mol
coding chat
turns into farming chat
wonderful
turns into math nerds yap
wish i could use this this to check if people are real
hop off latex hop on typst
so much better honestly
if you use nvim i can give you some of my scripts to make the typst experience better
actually
$\frac{-b \pm \sqrt{b^2-4 \cdot a \cdot c }{2 \cdot a} $
(-b plus.minus sqrt(b^2 - 4ac))/(2a)
$ (-b plus.minus sqrt(b^2 - 4a c))/(2a) $
line 1: Missing } inserted.
line 1: $
forgot cdot
i used texit so it uses $
if you use slash commands i think
i use in msg box
how to use biscord
Authorize
X
No scopes were provided
Close
I'll try on pc later
Invalid scope
maybe my app is just too outdated
yes I'm on 203.10
:3
Plz like the video and subscribe for more!!!!
me when my code compiles
what are some cool nerd books i can get
that are not expensive as hell
i wanted to maybe get the c book but WHY IS IT $70???
i just want something to physically read
like idk at school
when bitches have fucking 180fov and see when i use my phone for 2 seconds
ordinary differential equations by vladimir arnold
i too want physical books but they're rather expensive for something i would barely read
but reading pdfs just sucks
i cant read unless its in physical form
but i also cant buy physical copies
so i dont read
print the pdf
unnecessary effor
read epub
same thing different flesh
finna print the rust book
print this book
rude book
german book
Error
The Printer does not accept copyrighted german material
@dawn ledge
NOP
That looks like UB to me

soon they will vibe code after falling in techbro rabbit hole
thats what they've been doing till now
It prints CoreLibrary
we have to do 3 group coding projects per year and all half of my classmates submitted was copy pasted chatgpt code
they’ve already fallen into web dev hole
this is worse
idk
my guy, we are using nextjs for the project AND YOU START WRITING PHP CODE?!!
WHY IS THE FUNCTIONCALL COMPILATION CODE
600 LINES
AND THAT DOESNT INCLUDE THE CODE TO MONOMORPHISE GENERIC FUNCTIONS
...
i think i have this
@valid jetty when elle wasm???
qbe has wasm
or at least theres a wasm fork
@valid jetty when jsx in elle
it does not
send
the list slowly drops as i move more away from compiler.rs lmao
ok finally function calls moved away
compiler.rs is down to 3491 loc
itll probably be about 1000 by the time i move all of the astnodes to different files
i recommend you add this to your settings.json :3
{
"features": {
"edit_prediction_provider": null
},
"collaboration_panel": {
"button": false
},
"chat_panel": {
"button": "never"
},
"assistant": {
"enabled": false,
"version": "1",
"button": false
},
"show_edit_predictions": false,
"edit_predictions": {
"enabled_in_assistant": false,
"disabled_globs": ["**"]
}
}
oh
do i have to restart zed for that
maybe, not sure
:3
@hoary sluice im quite literally going 1 step at a time
if i do too much i could break something and have no idea why
its easy when im working with thousands of lines
getting there
cant you just shorten that instead of duplicating the same statement
so bad
which is what i will do when im done with all the nodes
yeah its insane
ARM support is awful
but its VERY VERY VERY rapidly improving
which makes me
i wonder what will come first, bazite arm64 or steamos/proton arm64
well a mod of proton for arm64 already exists, so i guess steamOS
@royal nymph
just my luck zed stopped giving me the quick actions menu for imports

WHY
why is it gone
sorry
everything changed when it clicked that the f(x) stuff in my precalc is very much like the functions i have been writing in python
i already have the mental framework from doing progaming
@deep mulch do you know how to calculate an rlc filter
and how to derive a 2nd order digital FIR filter from its impulse response without using inverse laplace transform
like how to derive the formula
I don't remember the equation
my calc can run actual python lmao
i have the fx cg50
f(x) = := x =
but idk why it doesnt work
you don't really need it
Boo Casio bad
yea, but its annoying without it
I’ve tried it, the buttons are mushy :/
This however is something I miss on my preferred but very locked down HP Prime
refer to the following items
elif {} is literally else { if {} }
i take advantage of that https://github.com/acquitelol/ichigo/blob/mistress/src/lib/utf8.le#L11-L21
utf8.le: Lines 11-21
fn char::unicode_length(char self) {
if (self & 0b10000000) == 0 {
return 1;
} else { if (self & 0b11100000) == 0b11000000 {
return 2;
} else { if (self & 0b11110000) == 0b11100000 {
return 3;
} else { if (self & 0b11111000) == 0b11110000 {
return 4;
}}}}
}
an advantage of this is that you can see the number of branches by just counting the number of } at the end
yes, i know
but its annoying because you have to do #if defined(MY_MACRO)
#if defined(MY_MACRO)
#else
# if defined(MY_OTHER_MACRO)
# else
# endif
#endif
why binary and not decimal/hex
because the way you tell the length of a single unicode unit is the number of bits set of the first byte
if you’re not working on the bit level you lose the information of exact 01111000 bits set with hex or decimal
at this scale you can understand by just looking, in hex or decimal they just look like magic numbers
You could also do something with counting leading ones
i see
yeah
ive never worked with that
i mean that’s basically what the code does isn’t it
unless you mean counting with bitwise operations
like
let iota = 0;
for i in 0..4 {
if (c & (0b10000000 >> i)) == (0b10000000 >> i) iota++
}
oh
lol i see
@valid jetty i need ur help to derive the fir coefficients of a 2nd order digital filter using the impulse response of an RLC filter and either being able to explain inverse laplace in 1 sentence or not using it cause i have to do this on the whiteboard tomorrow
i have no idea what an fir coefficient or impulse response of an RLC filter is
but you know how to explain laplace in 1 sentence
isnt that a pokemon
If you transliterate it weirdly from the jp name, yes
split a function of mixed frequencies into the frequencies that make it up, or you could say it converts a function of time into a function of complex frequency
programming sucks
discord is programmed
that first one is fourier
discord sucks
how can there be a complex frequency
im picking up farming
yeah i got silly for a second then realized what you said
i know of laplace from e^At but idk how to describe it, it jsut work
isnt it e^-st
one part controls oscillation one part controls magnitude
complex numbers are just math packaged 2d vectors
but isnt magnitude on the y axis of the frequency domain
yea in complex frequency the y axis is the Im axis which in this case is rotation
if you show it on an argand diagram at least
in the frequency domain its x=frequency and y=magnitude, what is it in the laplace domain
Re=magnitude, Im=rotation/oscillation
this is in the complex s plane
rotation as in phase?
yea
so its phase vs mag instead of freq vs mag
mag is real axis
is real axis x or y
from what i understand a complex frequency is just a frequency with phase represented as the im part
s=σ+jω
real axis is x
how is the magnitude on x wtf
σ=Re=magnitude
ω=Im=oscillation
because this is different
why do u know this but not what an rlc filter is
i do maths not electricity
electricity 😭
There is electricity inside you this very moment
but you know what a frequency, magnitude and oscillation is which is an electronics concept
im grounded
i am inside you at this very moment
what
Go to your room.
@valid jetty youre gonna have to teach me a bunch of math next week
i have an exam next thursday and im gonna explode
There are about 10^9 galaxies in our universe.
Each galaxy has about 10^11 stars. Since about 0.3%
of the stars have masses larger than ten times the
mass of our Sun, it is estimated that 10^17 supernova
explosions have occurred throughout the history
of the universe. This means that, on average,
one supernova explosion occurs every second
somewhere in the universe
yes because physics is still applied maths
and laplace transform is quite a theoretical concept even though there are applied uses
like the rlc thing
its hard to figure out how it works cause our teacher is the only one who does it like that
tldr pls
here is the doc for conveniency
https://www.ipmu.jp/sites/default/files/webfm/pdfs/news11/E_FEATURE.pdf
that is not what tldr means
10⁹ galaxies × 10¹¹ stars/galaxy × 0.003 (>10M☉) ≈ 10¹⁷ supernovae ÷ universe history ≈ 1/sec
btw @valid jetty teacher said im not allowed to do fft library during the lab cause it doesnt involve any analog measurements
thanks
@placid cape
flowcharts have never made sense to me
pipe lolcat
like why would you make a diagram and make a formal system for what each node does
it means i have to focus more on remembering the right shape than brainstorming the algorithm
it’s so much easier in pseudocode for me
Everyone uses different conventions anyway
drawio
me neither
but i need more pages in the document
how do i color this tho
Based
im gonna color it if i have time
i have 15 pages of content (u need at least like 40), i found out today that they check if its written by ai even if you wrote it yourself and then fixed the phrasing and i have to print it TOMORROW
hows this then
yea which means you can’t even say there’s a benefit to them because “everyone can easily recognize the symbols and understand what it means instantly” because everyone does it differently anyway
ok what abt this
do they make sense to you
refer to the following image
I love
well.... good luck i guess
explain it in words not numbers
yes this is perfect
how would they know if you get ai to write it and fix the phrasing
plagiarism checkers and ai detectors
everyone knows those are never accurate
someone put the bible into one and it sait the text was 88% ai generated
how do you know it wasnt

a human written diploma thesis isnt gonna show 100% ai
mine shows like 70%, my partners shows 100%
either way the only way to really know is by human inspection
which they will do
ai has a specific tone in which it writes text that makes it obvious
ik
but if theres a paragraph written by copilot thats 1:1 from some website its plagiarism
it might not sound ai cause copilot copies it from there
grök deep think +" write with university niveau vocabulary"
or ai with research enabled
that doesnt prevent plagiarism
prevents ai detection
im not gonna give money to elon musk to have 6 hours of hatred towards ai flowing in my veins instead of just writing it myself which is super cozy with some lofi at 3 am while being sure that i wont be exploded for plagiarism
i can entirely avoid ai detection and plagiarism for ichigo because i only need to reference myself lol
are you writing it with ai
ai is fine I've used it to enhance vocabulary in works that would've been otherwise deprecated
what's ichigo
i love how you can just see when i take breaks based on the commit date
no
im gonna have a japanese and english version of the document available written in typst
there isnt really a way to cleanly get ai to write that i dont think
don't trust ai for wiring something important anyways
a language written in elle with purely japanese syntax
i make this many commits in 2 weeks
heres an example
関数 メイン()『
整数 値 = 乗算(
加算(
乗算(13、2)、
減算(23、10)
)、
乗算(5、2)
)。
整数 ミクの数 = 除算(値、剰余(100、90))。
プリント(「%d\n」、ミクの数)。
』
these are only really refactoring commits so its not that much effort
yea im not gonna write my bachelor with ai
and ill plan ahead more
now im stuck with about 30 hours to write it
god i wanna just go to uni already
including sleep and school and preparing for a more or less exam
i wanna be free of the curse of physics
physics are fine if its measurement theory
photoelectric effect????????
like filters
what
when are u graduating
june 2026
i turn 18 in june lol
but im gonna take a gap year anyway so i can get money to go abroad
not really if theres an exam on my birthday
but idk if there is yet
are you still planning on doing the whole japan thing
yes
why not do erasmus
what's elle
i saw another Japanese coding language before but i don't remember the name, pretty interesting
Erasmus+ is the EU's programme to support education, training, youth and sport in 👉 Europe 👈 .
another language i made, this time written in rust, but this one is way bigger its nearing 20k lines of pure rust code and 10k lines of elle code for the stdlib
20k..
pretty sure you can still do a year abroad in the uk
core
nadesiko?
i think? i don't remember the name
no but i think erasmus is targeted towards studying abroad in europe
japan is in asia lol
eu countries are picked more often but theyre not exclusive
oh hm
you can go to syria
about 160 countries total, most of them require proper arrangements
but japan is probably on the easy side
there's also wenyan and https://nas.sr/قلب/ for Arabic, esoteric languages are amazing
im probably gonna go to romania or poland
you know romanian???
yes 😭
why??
because my parents are romanian and exclusively speak it at home
don't go to romania, the country is in self deprecating module since last year
??????
what????
no way
my pronunciation is probably a little bad since i dont speak very often but i am fluent in it
de ce nu am stiut
puhbu is dumdum can't pass history exam
(nu put sa scriu accente da imagineaza ca este un sh accent pe rosie)
roshie
read literature (don't)
cred ca nu o sa-i place lu vee ca vorbim cu o alta limba si nu engleza sau germana
romanian is engleshized very much anyways
da foarte lol
omg am inteles asta
3 words in English 3 words in romanian and then the local variations
eu sunt poate B2-C1 in romana pentru ca este limba mea materna
nu stiu cuvinte lungi dar pot sa scriu cu confidenta
e limba "heritage" nu?
da
c2 automat in documente ptc materna
limba care parintii mei vorbesc
muttersprache
casa communicez cu ei trebuie sa stiu si eu romana lol
ei no vb jp/en?
ei stiu engleza acuma
dar tot vorbesc in romana acasa
jp este complet diferit, asta o invat singur
nu m as mira daca s ar uita la tv romanesc
wha
?
aha
@hoary sluice hey at least the language structure of romanian is VERY similar to english
i understand everything ur saying but i cant form coherent sentences myself
lmao fair
i didnt practice for a long time so i know the vocabulary but only subconciously
@valid jetty i thought you're in jp bc (almost) everything about your profile is jp
me when german
bist du deutsch
Nein
yeah ive been studying japanese for the good part of a year or so (i didnt keep track)
de ce nu vorbesti c2
i want to go there for uni in a few years
daca vorbest din casa
aber 8 jahren deutsch gelernt, b2 niveau aber es ist shit
icic
i wanted to go to niigata uni or smth last year but when i saw the prices i deleted it from my history and forget about it
they banned the c word
pot sa vorbesc in o conversatie dar nu prea pot sa citesc literatura foarte repede
o sa vada masina banana ca te abati de la lb eng sau germana
vom vorbi romana exclusiv pana vee observa (observa e correct?)
da
citeste operele pt bac
dar se zice mai bine "pana cand vee observa"
masina banana 😭
nu stiu ce e cand
in romanian "until when vee notices" and not "until vee notices"
dada, pana cand o sa observe, o sa, caci vb colloquial
cand este "when"
ah ok
i think ill do a lot better in moldova
nu stiu ce vb inseamna, eu nu prea scriu in romana aproape deloc
normal vorbesc duar in en/jp
talk english to young people and russian to old people
vb - > perscrutat "vorbi"
ah
with what materials
anki?
i made a huge thing of what i use and how to learn a while ago i think
a vs ă e random
i haven't really found any material that is good besides setting yt location to jp and watching the news
move to japan
cannot
since this list being created, i want to also add airlearn and teuida to this list
i have to finish my diploma thesis can we talk on wednesday 😭
i still have to finish this and next year + get a job that doesn't pay like $300 mo
similar to duolingo but better content
what country do u live in
fish land
airlearn has more cultural information and explains instead of just throwing you into a bunch of phrases to learn
ist free or?
fish land can be a lot of things
ist free but you can only do 5 lessons a day
look at Europa what country looks like a huge fish
why are you saying ist
repeating what they said lmoa
my keyboard is mixed
5 isn't a lot
is that a fish
I'll probably have to speedrun until December
it is because the lessons are quite long
ya
nobody calls romania fish country
and you can skip to a point further on if you want
i do
i ALWAYS thought it looked like a fish
and why schengen map and not Europe, brexit
what is ist
ist (de) = is (en)
rosie isnt german and you said you barely speak german
i said my keyboard is mixed
idk if i will ever be able to master a 4th language
everything i know i learned in my childhood
native russian, started german at 5 and english at ~8
何年か日本語勉強してるんだよね、大学で勉強したいから
whats zh
Chinese
do u speak all of those
but you said this before
im struggling to learn a third one lmao
yes but now in jp™️
en, de, ro yes
turkey no
zh, jp no but kanji is fuck so i have to use both
kanji is the only reason jp is hard other than the somewhat hard grammar structures
deutch und russian
tja, krieg
shiritori
born in kazakhstan moved to austria at 5 but im not 95 years old so its fine
imagine learning jp for shiritori lmao
letim v Kazakhstan or something
i forgot how to phrase it
is that a reference to something
i learned a bit in my free time but now i don't have a lot of free time so it's just shiritori
yes
that is not kazakhstan
oh
well tj isnt similar to kaz at all
i can understand mixing kyrgyzstan and kazakhstan
but the tajiks speak a persian dialect afaik



i don't speak any of those
i just know da and tak and that's all
tak is polish
no Ukrainian
i think i know more Turkish than japanese atp
from Turkish song sozleri
how did this channel change from programming to language debate
uhhh
language programming
language language
same language
@valid jetty can i add you btw
nu e o debata
i guessed
wrong
its dezbatere
btw does anyone here know how to decrypt usm files (I've tried online tools)
lc.g usm files
o dezbatere este mai mult ca si un "argument"
dar sunt aproape acelasi lucru
sure
e o synonym
a debate is not really an argument though
ig its more peaceful
synonym foloseste "un" -> "un synonym"
you can look at this https://github.com/ToaHartor/GI-cutscenes
A command line program playing with the cutscenes files (USM) from Genshin Impact. - ToaHartor/GI-cutscenes
i just use whichever article russian would use
lmao
argument -> argumentare (scris)
dezbate - > dezbatere (vocal)
I've tried that
it didn't work on many newer ones
past v3.4 or so
XDDDD
i mean for all we know, it was
we dont know what technology they had at the time it was written
@leaden crater do you have the decrypt keys
guhhh why are all these usm projects in C#
Even i don't remember how I found out she knows Romanian lmfao
:3
condoleanțele mele btw
i have so many secrets you will never know
for example what happened on 16/07/24
yea!!
i will find out when i adopt you
omg
you're getting adopted?!!
no i think hes kidding (i hope)
and allegedly torture you until you tell me
"""allegedly"""
baseless accusations
si tuuu
e, si mie dar ce putem sa facem
ok that sentence is too complicated for me
Ai grija de tine, tembelo
"I hope you two are doing well
take care of yourselvess"
omg ty
you're welcome 
sper ca -> i hope that
sunteti -> you (pl.) are
amandoi -> both
bine -> well
aveti grija -> take care
de voii -> of yourselves
macar nu ai zis ca sunt o rosie
roșie tembelă
rosie change your name to 🍅
ive expanded to other japanese genres too
https://www.youtube.com/watch?v=RyRfLSOewbU listen to this
『D/N/A』
25時、ナイトコードで。 × 鏡音リン
作詞・作曲:Azari https://x.com/xxxazari
イラスト:PiPi https://x.com/PiPinavigation
動画:omu https://x.com/omu929
公式サイト:https://pjsekai.sega.jp
公式X:https://twitter.com/pj_sekai
(C) SEGA
(C) Colorful Palette Inc.
(C) Crypton Future Media, INC....
I LOVE DNA
that's her friend nickname for me
such a good song
pj Sekai cover
yooop
so goated
SO GOOD
I like that one more
This is peak
but yea now im listening to a lot more different things which arent inherently pop
I want to pray.
you just need the decrypt key tho
pray
do u know this song
Provided to YouTube by Victor Entertainment, Inc.
隙 · yuri
隙
℗ Victor Entertainment
Released on: 2025-03-26
Composer: 原口沙輔
Lyricist: 原口沙輔
Auto-generated by YouTube.
were about to see
yoop
i dont know it lol
tell me if ya like it
idk why ゆり is even blocked, it literally means lily in japanese
cuz people kept sending gifs of lesbians kissing and stuff iirc
BRUH i thought it wasnt playing, i simply didnt have my earphones plugged in
lmfao
roșie I devolved to samsung
i like this
holy shut my Samsung keyboard automatically made rosie into roșie
what were you using before
apple level price, not quite apple level quality
is that really a devolution :clueless:
with the most delayed android 15 update
im still on an ios version from 2 years ago
it lags more and has shit battery
so i can jb
nooooo rosie an apple user
im gonna try and get a second hand ip13 or ip13pm on a jbable ios version
yeah but im jailbroken
roșie will move to android in 6 years
also I did pass it (I think, waiting for the results lmfao)
yatta
not on spotify 😭
stupid ellie

YouTubeify
I know this one xd
this is good
guh this aint even half of what ytmusic has
listen to this one for like the first 2 mins because the chorus is really good
is it finally time to switch
I will do so in a bit, I need to delete some apps from /product/priv-app and replace an overlay (kill me)
lmao okie
do we fw pgadmin4
its too late for me to switch
i have several playlists with 400 songs
yeah
or postgresql in general
I switched because I liked downloading songs with 3rd party apps xd
my spotify playlist is like 500+ songs
419, 408, 192, 103, and a bunch of smaller ones for genres i dont listen to as often
yep
unfreinded



aveți grija de voii
