>>110139176
today I gotta fix some bugs in my assembler. Then I might be able to make a cute demo with it. Sadly I have no idea about muh animations, never did much of that

>>110139176
I'm building the replacement and killer for Plex. Did you know plex and like 10 other companies hold patents for subtitle to video matching? It makes me have to make the code retarded for a simple feature

Gosh Terry had some wild hair

function pointer lookup tables will fuck you in the ass (no lube), when you try to write them in risc-v assembly.
Holy fuck.

>>110139176
What a handsome lad.

>>110139243
I just have directories of videos on my computer, but you do you.

File: ididit.jpg (1.69 MB, 3753x2268)

1.69 MB JPG

>>110140155 >>110140296

>>110139398
I fucking did it!!!
Now I can make real programs

File: VID_20261023_103204-vp9.webm (3.73 MB, 1728x972)

3.73 MB WEBM

>>110140296

>>110139595
More blog posting to keep the thread alive.
I made a little demo! Wow, much demo. Such demo.

neopix_anim:
  addi sp, sp, -32
  sw ra, 0(sp)
  sw s0, 4(sp)
  sw s1, 8(sp)
  sw s2, 12(sp)
  sw s3, 16(sp)
  sw s4, 20(sp)

  mv s2, a0 // iterations
  li s3, 0x050000 // red

  neopix_sweep:
    li s0, 0 // x=0

  neopix_col_loop:
    // erase prev col
/*
    mv a0, s2
    li a1, 0
    call neopix_draw_col
*/
    // draw new col
    mv a0, s0
    mv a1, s3
    call neopix_draw_col
    call delay
    // mv s2, s0 // save prev col
    addi s0, s0, 1
    li t0, 16
    blt s0, t0, neopix_col_loop
    srli s3, s3, 8
    bnez s3, neopix_sweep
    // TODO add N loops?
    li s3, 0x050000 // back to red
    addi s2, s2, -1
    bnez s2, neopix_sweep
    // j neopix_sweep

  call delay // small delay before return
  lw s4, 20(sp)
  lw s3, 16(sp)
  lw s2, 12(sp)
  lw s1, 8(sp)
  lw s0, 4(sp)
  lw ra, 0(sp)
  addi sp, sp, 32
  ret

neopix_draw_col:
  addi sp, sp, -32
  sw ra, 0(sp)
  sw s0, 4(sp)
  sw s1, 8(sp)
  sw s2, 12(sp)
  sw s3, 16(sp)
  sw s4, 20(sp)

  mv s1, a0 // s1 = x
  mv s2, a1 // color
  li s4, 0 // y
  neopix_dc_loop:
    andi t1, s4, 1
    xori t1, t1, -1 // neg
    andi t1, t1, 15
    xor t1, t1, s1
    slli t0, s4, 4
    add a0, t0, t1
    mv a1, s2
    call neopix_set
    addi s4, s4, 1
    li t0, 16
    blt s4, t0, neopix_dc_loop

  lw s4, 20(sp)
  lw s3, 16(sp)
  lw s2, 12(sp)
  lw s1, 8(sp)
  lw s0, 4(sp)
  lw ra, 0(sp)
  addi sp, sp, 32
  ret

delay:
  li t0, 100000
  delay_loop:
    addi t0, t0, -1
    bnez t0, delay_loop
  ret

File: Screenshot_2026-10-23_17-23-27.png (251 KB, 1318x1030)

251 KB PNG

>>110140333 >>110141998

>>110129104
Here's a bit more detailed view of what is going on for you. You can get this information from running readelf.
The default entry point for the executable when you don't specify it is "_start", which you can see is in a different position than "main".
Using objdump you can get the compiled assembly out, and you can see it doesn't immediately jump to main, it does some stack shit, loads the address of main and then jumps off to libc_start_main.
When you compile a program, there's more than just your code going on in it, specially when targeting a system like Linux or Windows.

>>110139595
>>110140155
Cool, but what am I looking at? You connected to some LED device over UART or telnet. Are you programming on some microcontroller controlling the LEDs?

>>110140296
I have built a risc-v assembler for rv32i in holyC.
And now I am programming a risc-v SoC running on an FPGA with the neorv32 CPU.
Uploading the binary via a custom bootloader (in C tho) and xmodem and then jumping to the binary location.

>>110140278
Yeah that's true. There's hidden shit with _start you get from libc. I'm not an expert.

As far as the segfault goes, looking at the write function arguments, the first argument is file descriptor. Is 1 a valid file descriptor? I think you need to fopen first and then pass that.

I don't get what you mean by it doesn't crash in C but it crashes in assembly? Do you mean throw a compilation error? The program doesn't run differently.

>>110140302
That is cool. I appreciate the lunacy.

>>110140333
>Is 1 a valid file descriptor?
Yes. 0 = stdin, 1 = stdout, 2 = stderr.
Even if you passed an invalid file descriptor to write, you wouldn't segfault, you'd just get EBADF back.
>fopen
That's the libc I/O functions, he's using the lower level functions.

>>110129104
To add on to what the other person said about _start, I'm pretty sure it's because libc was not initialized properly. Compiling your code and running it, it segfaults for me, with the debugger being present or not, not making a difference.
I don't know the exact details of what glibc is doing internally, but maybe some kind of syscall table needing to be initialized. Maybe a different/older version of glibc doesn't fail in this case, but basically, if you're going to use libc in your program at all (99.999% of the time), you really need to not override the entry point and let it do its thing in _start.

File: scrn_20261023_114611.png (6 KB, 640x480)

6 KB PNG

>>110142010 >>110142045

rv32im fully supported now. Only missing Zicsr extension support to be fully usable

>>110140448
>>110140333
>>110140278
I love compiler diving.

Had a funky weird __malloc error when I tried to manually setup my LSP (for C++) in Neovim from scratch. Compilers are eldritch monstrosities from the depths.

>>110141260
Goddang Terry, an easier colorscheme wouldn't have hurt you know :S

File: cuckbow.webm (3.86 MB, 972x1096)

3.86 MB WEBM

>>110142055

>>110141260
not exactly how I wanted it to turn out, but I am happy.
I think that's enough assembly for today. My brain is melting holy shit.
Now I know this thing kinda works. I should improve the assembler now more and maybe build a compiler on top

>>110142045
>maybe build a compiler on top

You mean write one from scratch?

>>110142010
it's not that bad really.
For standard holyC files he has at least highlighting. I still have to figure out where this is implemented and see if I can extend it for risc-v assembly
inb4
>but muh light so bad xDDDDD
fuck you then

>>110142055
Yes.
One that emits risc-v assembly that my assembler can use

>>110142055
The other path would be:
Write an OS with that assembler, with fat12 drivers and some kinda line editor and then bootstrap from there somehow.
But doing the bootstrapping all in assembly sounds so frustrating holy shit.
I don't even know how the ps2 emulator guy did not go insane

>>110142061
Yes, flashbangs are bad for your eyes, that's not esoteric knowledge in this day and age.

>>110142351
That's why you should sit in a room with AT LEAST 300 lx

>>110142559
What specifically? I've been tempted to learn it recently

time to get back into the temple. 8 hours are not enough.
with some linkin park now
https://www.youtube.com/watch?v=jZSPAp8kCl4

How do I muster the energy to program after college classes? I always feel tired even though I go to bed 9 hours before waking up. It's so hard to do additional work at home when you finish at 5pm and get home at 6pm.

>>110143547
just be happy and enjoy life

>cheap external blu-ray drive keeps failing in the middle of long writes
>can't figure out cause
>jury rigging a longer auxiliary USB-to-barrel jack power cable because the one that came with it was dogshit and notice voltage is sometimes 3 instead of 5
>external blu-ray drive requires absolutely zero voltage dips
>turns out modern high current USB-A power supplies fuck around with the voltage a lot
>use USB-C and a dummy PD connector I have to force it to be 5v
>even explicit USB-PD based chargers fuck with the voltage when they do this sporadic polling thing to check if something is plugged in even if they know it's already there because USB-PD definitely requires some level of charger-device communication
i frakensteined an even more horrific auxiliary power cable which just has capacitors all along it for smoothing and it worked

i was just starting to write the type definitions for some sort of more robust resumable disk image format that acted like a middle ground between the way regular disk images have to be written kind of all at once and the extremely slow packet based thing
still might do something with libburn though, the intended way many of the linux tools work together for the backups I want to do is clunky and doesn't really fit the problem
not to mention linux UDF support is half assed and ISO images don't support some parts of the data like certain filenames

File: 8a4a24699831aed2f2e6570f4(...).png (278 KB, 694x900)

278 KB PNG

>>110145101

>write my own tagging application
>want it to be server based so I can use it from multiple devices
>originally it also acts as an image server as well as a tagging application but I decide this is already a solved problem so switch to minio so now front end gets images from a different service instead of my backend
>everything working perfectly
>try to run it on my network
>have to do some annoying hosts file edits to get around browser cross origin rules and make it work in docker
>yeah but what if I setup my own dns server instead
>do that instead
>it works fine
>try to use it on android
>android just straight up fucking ignores custom dns on firefox even with all the applicable settings disabled like DOH and other settings on android

Holy fuck why is this such a fucking pain in the fucking ass shit fuck I just want this to fucking work on all my devices and I have to fight against fucking everything just to do that.

>>110139176
>yfw you realize your code is niggerlicious rather than divine intellect

File: me.jpg (24 KB, 316x916)

24 KB JPG

>>110147445

>>110139176

public static void main(String[] args) {
        learn2Code();
    }

    public static void learn2Code() {
        char[] string = { 73, 102, 32, 121, 111, 117, 32, 100, 111, 110, 39, 116, 32, 114, 101, 112, 108, 121, 32, 116,
                111, 32, 116, 104, 105, 115, 32, 112, 111, 115, 116, 32, 121, 111, 117, 114, 32, 109, 111, 116, 104,
                101, 114, 32, 119, 105, 108, 108, 32, 100, 105, 101, 32, 105, 110, 32, 104, 101, 114, 32, 115, 108, 101,
                101, 112, 32, 116, 111, 110, 105, 103, 104, 116, 33 };
        System.out.println(string);
    }

Today Ive been playing around with beautifulsoup, its really hard

Thank God for Rust.

>>110144879
>exposes all the troons

>>110145041
Just like C exposes all the Indians.

>>110145065
This India narrative you keep trying to push about C is very weak.

>>110141998
>Compilers are eldritch monstrosities from the depths.
Oh, you've met MSVC then?

>>110144366
>have to do some annoying hosts file edits to get around browser cross origin rules and make it work in docker
Normally you patch that shit together using nginx to rewrite/rev-proxy things to look like they all come from the same host.

>>110145065
Not all. Just better than 99%

>>110145067
It's a fact not a narrative.
Why does C sometimes seem especially popular among Indian programmers?
There is also a cultural and educational feedback loop:
Colleges teach C as a foundational language.
Students encounter C in exams and programming exercises.
Interview-preparation materials reinforce C and C++ fundamentals.
Graduates carry those habits into the workplace and into teaching the next generation.
This can make C unusually visible in Indian programming tutorials, online discussions, and beginner courses.

C is the lingua franka of India.

>>110145116
Damn, you seem to know a lot about this, Rahul.

>>110145227
>projecting jeet

File: IMG_9334.jpg (72 KB, 702x828)

72 KB JPG

>>110145227
Also you're brown. HAHA

>>110145116
C is still one of the most common languages in use.

does coding ever get any easier

>>110145267
Indians have a large population.

>>110145116
C established itself as the dominant language 20 years before India knew what a computer was. You're falsely connecting the two.

>>110145278
when you work in a familiar domain, sure.
also if you find the actual act of programming hard, you should re-think what you are actually trying to do. programming concepts are mostly quite easy when you do it right (dont overabstract, dont abstract too early, keep it simple)

>>110145290
The world is a bigger place than just India.

>>110145323
>falsely connecting
>Oy vey, stop noticing things, you Whites.

>>110145359
You don't need to know that. Just go back to india.

Why not just use Rust or C++ instead of being indian?

what the fuck is going on with clang's C++ 23 and C++ 26 feature implementation
even for the in dev versions it's been mostly stagnant for like a year at least if not more and any updates have been extremely small feature tweaks
it's at the point the stable versions of gcc have completely blown past it

File: screenshot_2026-10-24_08:02:41.png (205 KB, 1504x1316)

205 KB PNG

>>110147113

Work progresses porting Serpent to tkinter.

By /g/ee's logic is PHP better than Rust since PHP is more vulnerable?

>>110145797
Are you a faggot? Because you're acting like a faggot.

>>110145810
No, I'm straight. I'm acting like a straight. I challenge all indian narratives on /g/eet.

>>110145463
What’s a better language than C if you value simplicity and compilation speed over rust’s overbearing safety or C++’s pointless bloat and complexity?

I know the answer. It would probably be odin. I wonder if I switched to odin if this random schizo on the internet would stop calling me indian. Hmmm, nope. I think I will continue to write C.

>>110145861
You will never be white, brown jeet.

>>110145886
>say the line, bart

>>110145893
Memes won't stop me from calling you brown, brownboy.

>>110145912
you are the new malloc schizo

>>110145837
>I'm acting like a straight.
Sure, buddy.

>>110145922
>>110145939
No, but you are the Rust schizo.

>>110145953
I have said zero words about Rust.

Rust did nothing wrong. It amended C's mortal sins.

>>110145974
Rust did nothing wrong.

>68 replies in the thread
>zero negative posts about Rust

>>110145991
Ok, but I'm still gonna use C.

>>110145589
MSVC have just barely started 26

>>110146004
Enjoy your segfaults. Meanwhile I won't be making any.

>>110146024
I'm not going to complain here about it.

>>110146030
It's a fucking anonymous forum. You can complain about anything especially C.

>>110146024
>Meanwhile I won't be making anything.

>>110146038
Still not going to complain here about it.

>>110146049
>just bottle it up and shill what's oppressing you
Can't tell if masochism or Stockholm syndrome.

>>110146024
segfaults are so trivial to fix they aren’t worth mentioning. maybe they are if you’re a nocoder, but for anyone else they are the lowest hanging fruit of debugging

>>110146075
Nocoders don't make segfaults, you fucking idiot.

>>110146070
complaining is feminine and a negative personality trait

>>110146007
yea but I expect that from them, given the VC release schedule
plus they've been hinting at a nebulously upcoming major ABI break (vc-next or something) for a long while which will technically be the first one since XP

i guess other projects under the llvm umbrella are taking up dev time or something
i know there's plans for some kind of a C++ specific IR or something

>>110146093
nocoders don’t make anything, which is why they would be inexperienced enough to complain about segfaulting in the first place

>>110146102
>plus they've been hinting at a nebulously upcoming major ABI break
that got cancelled

>>110146098
Yet all /g/eet does is complain about Rust.
>>110146103
Segfaults don't explode but they're annoying and worth complaining about. Imagine bottling it in so Cindia's image isn't tarnished. Pure masochism.

>>110146117
>Yet all /g/eet does is complain about Rust.
86 replies and zero negative posts about Rust.

>>110146113
oh shit really
damn
they were hyping it up last i looked into it (which was admittedly years ago)
guess the bugs that require breaking the ABI to fix are just going to stay there

>>110146070
Like any other error that occurs when coding, if I get a segfault I just fix it.
It's not something that ruins my day or stresses me out. If you're gonna let every little thing that goes wrong get to you then programming, no matter the language, will be endlessly frustrating.

>>110146117
nobody here is complaining about rust. they just don’t have a use for its rigid paradigm and safety. rust just gets in the way of prototyping so I really don’t care about it, I don’t really think about it. I don’t see someone posting rust code and go “HEY HE CANT JUST DO THAT. TRANNY TRANNY TRANNY”. you on the other hand seem to be completely obsessed with C and anyone that uses it that you literally shitpost the thread to death for weeks on end at the mere mention of it. that’s why you are the thread’s newest resident schizo. you have autism

>but they're annoying and worth complaining about
they really aren’t. 99% of segfaults can be fixed immediately from just a single glance at the call stack. it’s almost always either a null dereference, a range bug or an uninitialized memory problem. all of these cases are absolutely trivial to fix and easy to find. the one single case where segfaults become iffy is if you have a race condition somewhere. I don’t worry about segfaults, I don’t think about them and when they happen I am happy that they are a segfault instead of a more subtle bug

>>110146227
If you actually used C you would hate it too. Everything you're saying is just a college tier understanding of C. Yes, segfaults are better than some hidden bug but they still suck and suck is suck.

>>110146227
also, I forgot to mention, in the case where it is a race condition, I would rather it be a segfault than the subtle bug case because the segfault is just the symptom in this case, not the cause and the symptom is screaming at you to fix it

>>110146281
I use C everyday dude. I don’t know what to say at this point other than you are having some kind of skill issue. this shit is not and has never been a big deal

Nocders love C because of its leet haxor rep. C gurus who are stuck using the language because there's no good alternative despise C.

>>110146308
>skill issue
You're just a skiddy. You don't use C.

>>110146308
There is nothing that screams leet haxor nocoder than unironically saying skill issue. You're just a brown jeet. You have no skills and don't understand the pain of using C. Some dude I know who has an A in modern physics bitches about C++ every day. C's way of doing things is hacky and assbackwards and it's nearly unparsable to compilers.

>>110146318
>you’re brown and a skiddy and a liar because you’re not running into skill issues I am having
I think the college tier understanding comment is just you projecting. I don’t think you have a whole lot of experience actually making anything. strange that something as abstract as a programming language can live so rent free in someone’s head

>>110146351
>I don’t think you have a whole lot of experience actually making anything.
>t. Dickpreet who is 20yo and has 25 years of experience on his resume.

At least no one here hates assembly.

>>110146340
I am calling skill issue because there is genuinely nothing else to call it. you are making segfaults to be some huge insurmountable problem when they just aren’t.

>C's way of doing things is hacky and assbackwards and it's nearly unparsable to compilers.
some things about C are certainly hacky and dated, but in terms of compiler complexity and the amount of translation work necessary to turn C into machine code compared to more complicated modern languages, nothing could be further from the truth. a great strength of C is just how predictable the assembly will be, assuming you don’t have optimization flags on, so I don’t know what fucktarded argument you’re even trying to make at this point.

>>110146363
>I don’t think you have a whole lot of experience actually making anything
you talk a whole lot about college and shit you don’t understand so it’s a safe assumption, anti-C schizoposter

>>110146395
You brownoid jeets only understand attacking reputations because you have such poor ones.

>>110146403
no, that’s what you’re doing; calling everyone you disagree with brown or indian because you literally don’t have anything of substance to say

>>110146388
Parsing C is notoriously difficult because its grammar is context-sensitive and inherently ambiguous, breaking the clean separation between a compiler's lexer (which turns text into tokens) and its parser (which builds the syntax tree).
Unlike modern languages designed to be easily processed with a single token of lookahead (like Go's func or Python's def), C was written in an era before formal grammars were widely adopted for programming languages, prioritizing memory-constrained, single-pass compilation over theoretical cleanliness.

>>110146413
It's a brown indian language that belongs in india.

>>110146418
that is true but it all just amounts to one backpatch you have to do between the lexer and parser. typedef was a mistake, for sure. if you’re not a compiler writer I’m not sure why it even matters though

>>110146426
you aren’t beating the allegations

>>110146448
>guilty until proven innocent
Brown jeet thinking. Absolutely non-white thought processes.

>>110146467
everyone is free to read the entire thread you just shit up and come to their own conclusions

>>110146418
C is slow as molasses on an 8-bit microcomputer. there was a time when C was considered the slow, bloated compiled language believe it or not. C was meant for big, expensive 16-bit minicomputers. BASIC, Pascal and Forth all compiled faster on a typical home computer. Forth is about as low-level as you can go and still be considered a high-level language.

>>110146483
Meanwhile in reality people with jobs already decided to replace C with Rust in Android because C sucks and is insecure.

>>110146497
people with jobs don’t program anymore retard

>>110146495
I like old languages like Forth and Lisp.

>>110146507
old languages are lots of fun. programming in forth feels like solving a puzzle. i do part of my coding with pencil and paper like i am solving a math problem.

>>110146519
I had an experience like that in Brainfuck.

>>110146495
i really want to love forth, like conceptually it ticks a ton of my boxe
but i find it so hard to get used to
maybe i just need to properly dedicate myself to it

>>110146418
This waas LLM generated.

>>110146690
>attacks the messenger instead of the content
Brown minds know no bitter end to their sinful ways.

why do rustfags act like security and safety are the end-all-be-all
they literally only matter for applications handling remote unsanitized input and even then the exposed attack surface is only scoped to the the code that is responsible for sanitizing that input
while that is a legitimate usecase for rust
at the end of the day there is no point trying to babyproof the entire world to protect retards from themselves
not only is it a fools errand, as retards will always retard, (shit like command and SQL injection still happen in the current year), but it's completely out of scope
safety and security shouldn't always trump things like performance or flexibility

like at some point one of the projects I want to experiment with is hostless gpu driven jit'd gpu code by seeing if I can stream instructions to a RWE page of VRAM
that's a huge security vulnerability
does it matter at all (in a hypothetical example where doing this was portable)? not unless the application is networked
and even then, if it is, and then if i incorporate it into some sort of user-driven content pipeline and I slap a disclaimer not to connect to untrusted servers then a path to code execution becomes a feature and it becomes the user's responsibility to not download malware

the inevitable conclusion of the attitude rustfags have will result in the end of free computing
shit like only being able to run code locally on extremely thin clients if it's signed and executing in a secure enclave while the vast majority of everything else is server side

>>110146340
there's nothing hacky or assbackwards about writing C code. it's pretty straightforward in fact. the pitfalls to avoid are well-known and there are multiple ways to go around them. you do those things (use stdint types, pass lengths with your data, group allocations) and run with sanitizers and your analyzer of choice. big deal

File: Screenshot from 2026-10-2(...).png (253 KB, 1139x939)

253 KB PNG

>>110146824

>>110146690
oh my god, you're right.

looking back, that was the most technically correct of all of the anti-C schizo's posts. it showed at least some domain knowledge of a subject, was actually neutral in tone and unlike every other post he made, I could find little to argue against. he couldn't even use his brain so he generated an LLM response. imagine calling other people indian. what a fucking loser lmao

>>110146788
Pain in ass. I want to write and run without setting up a complex toolchain.
>>110146789
>neutral in tone
You are such an obvious brown jeet. You even popped in the thread when India's day starts.

>>110146824
you should delete your posts in shame, I can feel the second hand embarrasment

>>110146702
I didn't say it was wrong. I just said it was LLM generated.

>>110146831
The only thing embarrassing is how much ub C has and how hard to parse it is.

>>110146850
that's nice, did an LLM tell you that?

I mostly agree with Terry about smart people admiring simplicity, but I don't think that's entirely possible with massive systems

File: IMG_3391.jpg (343 KB, 1170x1857)

343 KB JPG

>>110147178

>>110146853
>indian is also a codetrans snailcat

re

>>110146871
It's about doing what you can. Everything should be made as simple as possible, but not simpler.

>>110146871
The point is complexity reduction. The simplest way to solve a complex problem is still maximally simple for that problem. If you are comparing the simple solution to a complex problem against solutions to different problems, you're just comparing apples to oranges.

I hate c

>>110146547
>>110146574
brainfuck is a fair comparison. i might see if i can make c work or, failing that, use BASIC or Pascal. forth is kind of a nightmare. i think Power C would at least allow me to code my own level and graphics editors if not slower turn-based games or text adventures.

forth is maybe 2-5x times faster than compiled BASIC or C but it is basically unreadable. i could just cross compile but that sucks some of the fun out of it.

>>110147055
Dubs. Very nice.

>>110147055
try holyC

File: screenshot_2026-10-24_12:17:04.png (40 KB, 885x284)

40 KB PNG

>>110145592

>>110146880
Tell it to make the case for C

>>110147178
I'm not indian though.

>>110147108
why do you always type it with lowercase h, it's HolyC

1. Why do C programmers never get lost?
Because they always know where their pointers are.
Until they don't.
2. What's a C programmer's favorite pickup line?
"Hey girl, what's your address? I'd like to dereference you."
3. Why did the C programmer break up with his girlfriend?
She had too many references to her ex, and he couldn't handle the dangling pointers.
4. A C programmer walks into a bar.
He orders a beer.
He orders 0 beers.
He orders 999999999999999999999 beers.
He orders a segmentation fault.
He orders a beer.
The bartender asks, "What's wrong with you?"
The programmer says, "I don't know. I forgot to check the return value."
5. What's the difference between C and C++?
C gives you a gun and tells you to point it at your foot.
C++ gives you a gun, a foot, and 47 different ways to accidentally shoot yourself.
6. Why don't C programmers use garbage collectors?
Because they are the garbage collectors.
And they're still leaking memory.
7. What's a C programmer's favorite magic trick?
Making memory disappear without a trace.
It's called free().
Sometimes the memory doesn't disappear. Sometimes it disappears twice.
8. Why was the C program so confident?
Because it had no type safety to hold it back.

>>110147202
>i only want the LLM to generate things that are supportive of my position
It shows.

>>110147274
I'm not indian and I won't take part in indian activities like supporting C.

>>110147258
I had a dream that I was a muffler last night.

I woke up exhausted!

>>110147283
That's what I said.

File: 1789549491872584.png (225 KB, 480x360)

225 KB PNG

>>110144613
AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAHHHHHHHHHH

vectorized my entire string library with sse and avx and was surprised with how trivial it was. it was like 1 day of work

I don't know why I thought it was gonna be hard

i love c

>>110147506
try HolyC (with uppercase H)

beautifulsoup is really hard and im not making any progress

>>110147699
What's hard about it?

>>110147588
what did terry add to it

>>110147804
just hit F1 in TempleOS????

>>110147713
the words used, regex, scraping sites riddled with JS, ive been stuck for 2 days trying to make sense of it

gave up on internet scraping, going back to data manipulation

Started documenting on how I got from the latest TempleOS release towards getting it to work on my hardware.
Fuck, documentation is so boring. But it's good to get myself a clean distro again. My working copy is all messed up. So started from qemu again fixing step by step to make a release (TM) soon (TM)

>>110148654
what is THAT?

File: 1735466709663584.jpg (139 KB, 500x572)

139 KB JPG

>>110145084
How did you guess?

>what is THAT?
damn I fell for it

>>110145116
Indians are mostly on JeetScript.

>>110145861
Hasn't modern C also gotten some bloat features of its own? Haven't really been keeping tabs on its evolution.

>>110148580
Documentation is always boring until you need it.

>>110148954
I don't know man, I don't know much about C23, honestly. I don't think I'm ever moving beyond C11, and that's already too modern for a lot of people that like to dig their heels in at C99.

I just like having anonymous structs and unions as a language feature instead of a compiler extension. And I like the C11 memory model with atomics as a standard, again, instead of a compiler extension, and I like some of the tools becoming standardized instead of compiler.macro hacks like alignof/alignas. For those reasons I tend to gravitate towards C11, but I can't really think of anything more I'd want. With macro hacks you can get decent containers. I don't know how I'd feel about a new version of C that provided generics, for example. My first instinct would be not to trust it. C is soft of already solved and if you want more you should probably just pick a different language.

>>110148961
The problem is now that, I did let the AI write some ATA code to be able to boot on hardware.
Somehow I want to understand it, but that'd mean diving deep into the docs, but I don't really have the feels for it.
So I guess I'll just accept the AI output as working and archive the changes

>>110148954
>>110149068
The C standard committee (typically) is good at not adding too much useless bloat shit.
Using a later C standard isn't going to cause some major paradigm change in your program, at least compared to taking the C99-pill and abusing the shit out of designated initializers.

Things in C23 I consider interesting:
- <stdbit.h>
Get access to those hardware bitwise instructions in a standard way without having to try finagle a bunch of bitwise operators.
- Binary literals
0b0101
- typeof(), type inference with 'auto'
Helps with some macro stuff. C's type system isn't a clusterfuck like C++'s, so it's not expected that people will just replace all of their shit with this.
- 'nullptr' constant and nullptr_t
Seems like a stupid C++ism at first, but having read their justification, it makes sense.
NULL being able to defined as 0, (void *)0, or whatever the fuck the implementation wants causes issues with _Generic and varargs.

>>110149305
anyone on a language committee is there with the intention of designing language features. That is the exact way that C++ became a mess so I remain inherently distructful

>>110149334
They publish their meeting minutes, and I casually keep up with them.
They definite get why people like/use C, and aren't just trying to turn the language into C++. They shoot down tonnes of stuff.
But yes they aren't there to let C stagnate and do actually add stuff.

>went full retard with C++ meta-programming
>tiniest change to code results in over 1min build
>on 5950x
it saves me tens of thousands of lines of boilerplate code, but it's still brutal to work with

>>110149224
>I did let the AI write some ATA code to be able to boot on hardware.
The ATA code already boots on hardware though...

C is for indians. They stole it and they can keep it.

sepples is some gigantic bullshit sometimes, especially when you need something that isn't abstracted away like cross-platform process launching like in Python:
>import subprocess
>subprocess.run(["nig", "nog"])
and a variety of methods to control it and capture its output, even async during execution. too often it feels like I'm goddamn dastardly and muttley trying to stop the pigeon when programming in C++.

C with libdispatch and Clang blocks is cool

dispatch_async(queue, ^{
    printf("Hello from a libdispatch block using Clang!\n");
});

>>110149954
didn't on mine.
It got stuck on writes. Reads worked

>>110150238
this has very little to do with c++ being bad. you just chose the wrong tool if you are looking for ready-made abstractions for everything

Rust never cut off its penis operator => that means it's not trans.

File: lock_1763040182289.jpg (106 KB, 1088x1062)

106 KB JPG

LETS FUCKING LOCK IN BROS

File: python_pepe.png (237 KB, 700x1007)

237 KB PNG

>>110153797 >>110153869 >>110153873 >>110154844

I use Python.
I'm a programming Chud.

>>110153777
Checking divine trips for Python and Pepe.

>>110153777
Chad more like lit.

>>110153777
pls do not step on snek

>>110150168
>C is for indians. They stole it and they can keep it.
I'm not a big fan of C, but you're just fucking annoying. Don't you have anything else to talk about? Written any code lately? Got any cool projects you're working on? Just anything other than complaining about C and calling people indian

File: IMG_3402.jpg (99 KB, 1024x865)

99 KB JPG

>>110154070
I tend to annoy brown people.

>>110149750
Did you use models? I still have to get around to testing with those.

Also what kind of meta-programming did you do? Reflection and code gen?

>>110154268
Sorry, meant MODULES.

>>110150168
90% of jeets can't even write code that compiles, they're definitely not flocking to C. Or any other compiled language for that matter.

>>110154290
C is super easy to make compile. The only reason to not use Rust is you have some skill issue with the borrow checker and can't compile Rust.

>>110150238
>cross-platform process launching
Do you have the slightest idea just how FUCKING AWFUL THAT SHIT IS IN WINDOWS?! All because those DUMB MOTHERFUCKERS put argument parsing the standard library, and then let there be alternate implementations of that?! (POSIX process launching is fiddly, but mostly a solved problem.)
The simple cases are easy to abstract. But start dealing with complex edge cases, and it all gets bastard hard. Oh well, at least the screaming clusterfuck that was pre-NT windows is now excised (and hopefully buried face down in an unmarked grave in the Mojave). That shit was truly insane.

ive practiced with pandas, tinydb, pydantic, path, and argparse, my favorite python modules

Tonight is the night I start work on pinball in Tabby.

File: 1792638497049545.png (30 KB, 400x190)

30 KB PNG

>>110154986 >>110154993 >>110159248

Big things are happening.

>>110147826
regex sucks nuts. even once you learn it well enough to parse stuff well its always slower than just manual ass iterating through the string. websites are fucked now anyways though. every page is an AI generated single page app 50mb cluster fuck. my only advice is if you really want to do it use something like headless chromium to have all the javascript run before trying to parse it.... but once your that deep you can pretty much just wright javascript to parse out the exact shit you want using querySelectorAll("#blah").innerText or whatever instead of crappy regex or html libs.

File: 6.gif (1.16 MB, 720x576)

1.16 MB GIF

>>110154585
cool good luck. pinball is such a funny problem. there's some fine line there between real physics and fun arcade mechanics that I've never balanced successfully. try atari pinball for an example of exactly what NOT to do.

>>110153777
based pytoddler

>>110154627
anyone who posts fake stupid benchmarks where their language is 30x faster than the fastest languages on the planet they’re either being dishonest or stupid so I’m automatically not interested in this shitlang right out of the gate

File: reddit faggot.png (572 KB, 927x1020)

572 KB PNG

>>110154627

File: f2p2_rubish.png (31 KB, 822x1872)

31 KB PNG

>>110155054

"f2p2.rubish"

File: rubish2_rar.png (339 KB, 2262x2256)

339 KB PNG

>>110155047
"rubish2.rar"

File: IMG_3440.jpg (391 KB, 1200x1200)

391 KB JPG

>>110155606 >>110156462

What is Rust competing against again?

>>110154986
genuinely how i felt about his self-promo spamming everywhere too

dude vibed a language that can do something better and most real things worse, alert the press

>add namespaces
>add defer
There, C is perfected.
>tfw I pass shit by copy in C++ and the destructor runs and the resources get 10x freed

File: Screenshot 2026-10-24 at (...).png (49 KB, 1884x1390)

49 KB PNG

I successfully made progress in pinball. I got three balls bouncing around. Image unfortunately isn't animated.

>>110155369
So make C Go and it's perfect. Glad you're starting to see the light.

>>110155516
>added gc
>added generics
bloat

>>110155541
Lol, if you like extra work C is the language for you.

File: 1702269040940573.gif (2.11 MB, 200x150)

2.11 MB GIF

>>110155633

>>110155145
>c++ not complicated

>>110155606
Better to have the option to be simple and use a subset of the language than be too simple to get anything done.

I despise C but dang it's a good compiler target. It somehow compiles faster than Assembly assembles. I see no reason to use LLVM when C exists. I still hate manually writing in languages without generics like C though. Don't get me wrong.

>>110155633
To remove the complexity and problems of C++, you'd have to start everything from scratch in order to bake in your core assumptions and flush out the bullshit anyway. It's a complete wash. Switching from C to C++, the meaningful features you get out of it are like, overloading and uhhh, templates but the implementation of templates in C++ is so poor that they're hardly worth using in the first time. The cost on the other hand, is compilation speed. It's not even worth it imo

>C++26 vote failed
lmao

>>110155657
Better to have generics than not at all. But yeah other languages don't seem to have slow generics.

>>110155673
You can have generics in C with some simple macro tricks in conjunction with embedding type information into unions. I greatly prefer this to C++ templates. Even with errors reported through layers of macros, it's still somehow easier to read than the fucking nonsensical errors that templates tend to generate.

>>110155689
Yeah, and then you get obscenely long function names because the type gets included in it. The macros themselves also look ugly unlike C++ templates. It's better to solve a problem with a dedicated solution than with a hack.

>>110155712
No, I know exactly what you're talking about and I tried metaprogramming with that kind of convention and it does suck, even if you generate _Generic wrappers. You don't have to include the type name like that with the method I'm describing.

I can show you if you give a shit.

>>110155748
Sure, show me.

I do something like this in my codebase. I didn't check if it compiles but you'll get the idea.

typedef union generic_container generic_container;

#define container(T, ...) \
union __VA_ARGS__ { \
  struct { \
    T* Data; \
    /* ... */ \
  }; \
  generic_container* GenericMetadata; \
  T* TypeMetadata; \
}

typedef container(void, generic_container) generic_container;


#define typeof __typeof__
#define Type_Check(ResultPtr, ComparisonPtr) ((1) ? (ResultPtr) : (ComparisonPtr))
#define Container_Do_Something(Container, Element) \
  Container_Do_Something_Impl((typeof ((Container)->GenericMetadata)) (Container), \
                              Type_Check((Element), (Container)->TypeMetadata), \
                              sizeof (*((Container)->TypeMetadata)))


void* Container_Do_Something_Impl(generic_container* Container, void* Element, void* ElementSize) {
  // ...
}

typedef struct {
  int Value;
} foo;


int main(void) {
  container(foo) FooContainer = {0};
  Container_Do_Something(&FooContainer, &(foo){.Value = 123});
  return (0);
}

If you really need memcpy inlines, you unfortunately have to wrap the implementation in a do while macro instead. For large array copies or something, it's fine. I think __typeof__ is supported by all the big 3 compilers and it's also standard to C23 IIRC.

>>110155795
ElementSize should be size_t, oops

I remember when I made my own preprocessor and had fun with it. I could implement a compile time factorial in it.

>>110155795
for
>>110155752

Yeah pinball. Or currently ball bounce. Later I'm going to change the velocities and positions to floats so I can have more fine tuned control.

>>110155975
>floats
>more fine tuned
>>>>>>>>implying

>>110155979
Only one way to find out.

>>110155997
he's retarded, you definitely should be using floats

>>110155997
IEEE don't want you to know this but you can just use fixed precision (i.e. integers) for almost any realistic problem

I wrote a webapp that plays vocal exercises for me. I'm getting better at singing / karaoke.

I can adjust the tempo easily, which is nice.

>>110155652
> I still hate manually writing in languages without generics like C though. Don't get me wrong.
why? Go was actually completely fine without generics. we went from strings.Sort() to slices.Sort() in the standard lib. ive never felt the need to give a function the ability to process like 20 different data types or whatever.

>>110154297
>C is super easy to make compile.

Not to jeets it's not.

>>110155145
Normal people, but mostly biological females.

>>110155657
Debugging templates in C++ can be true hell.

>>110155665
I just wish they'd make modules less confusing.

File: ezgif-77298ae08028066a.gif (2.69 MB, 592x640)

2.69 MB GIF

>>110158832

Have any of you ever had a C or C++ job? The only people I knew who wrote C professionally were when I worked for a bank and they had an old as fuck banking system they maintained. The only person I knew who was doing Rust professionally was working on crypto shit. 99% of my career has been web shit. Even internal tools at the companies I worked for were intranet websites like sharepoint or whatever.

>>110154687
its really bad and usually not worth it, the best case scenario is catching the backend api json, but that never happens
ive been learing regex a little but I dont really see the point or any difference in speed for that matter, im still stuck with scraping that is so slow its not more useful than just visiting the site

>>110156737
I wrote my own string library and while the API itself took some trial and error to really get right, once I nailed it in, I find it more convenient and easier to write than regex even though it requires more code than a compressed regex expression. It's also probably multiple magnitudes faster.

https://aardappel.github.io/goose/
mallocschizo made a language

>>110157086
>vibecoded
inefficient compilers already made him crazy, what do you think inefficient LLM code would to him?

>>110157182
It compiles to C without malloc, it's ebin.
Will probably bitch about muh register spilling or other gay shit tho.

>>110157238
The mental overhead of using arenas is already trivial and completely solves memory management. I don't see the point in this shitlang.

>>110157248
>C without UB without malloc with automatic mmap and with closures and generics and global returns from nested functions
Sweet deal.

>>110157387
>transpiles to C
this is a toy, I’d rather just use C

File: scrn_20261025_094109.png (3 KB, 640x480)

3 KB PNG

I added icons to my menu.

Not sure if I am too dumb, but it seems like you can't align them without spaces, such that they are fast navigable with keyboard. So they are only for clicking which I rarely do. But they look cute. I am a designer now

>>110156487
I hope they get rid of profiles and contracts and all the dumb safety shit

Hmmm, if my calculations are correct, my SIMD pixel renderer is about 80% faster than my previous scalar renderer for gbemu.
I think there are still improvements to make, though.

File: music.jpg (435 KB, 4000x4000)

435 KB JPG

>>110158728

man, building an assembler is one thing. But making it usable fucks with my head.
I could just have fun with Rust instead of trying to build a fucking assembler in HolyC

>>110154268
no modules. I am using meta-programming to bind my C++ classes and functions to a scripting language engine (Duktape).

>>110158696
assemblers are straightforward and pretty easy. the way intel encodes instructions is annoying but the problem itself is simple. they’re way way easier than compilers

>>110158728
I am doing risc-v.
But yes, if you look from the outside it's very easy and it really is.
But then you start thinking about directives that are needed, handling multiple files and namespacing files, adding all the instructions, it's all just very daunting work.
Tho maybe the joy of the accomplishment might be good

>>110158745
in short: actually making it usable, so that you can build other systems with it

>>110154986
>>110155297
Mad John Carmack has never complimented your programming language?

Learn Erlang. Erlang is the best programming language.

File: tohru shrug.jpg (185 KB, 498x409)

185 KB JPG

>>110156525
Jobs are for wagies, and are only for wagies until employers figure out how to replace them. Nothing good comes from being a wagie. Your entire life reduces to a loop where people who are already wealthier than god look for creative ways to exploit your labor as hard as they can, while minimizing your salary, so they can become even wealthier. All while expecting you to be grateful for the "opportunity". Avoid employment and focus on doing the Math/CS research you like. Capitalism is just feudalism with extra steps to make it less obvious to dumb people how exploitative the system is and how much it takes from you, and how little it gives back.

>Cosmos
>Goose
Vibegod victory

>>110154627
THE
GOOSE
IS
LOOSE

>>110156456
Jeets use c in their intro to programming classes and their interviews. They can't compile Rust.

>>110156462
Rust is competing against idiots with skill issues who can't get past the borrow checker.

I accepted the future and now I'm working on my own agentic coding harness.
Each time I press the button it burns 4 cents.
I love the future.

File: IMG_0949.jpg (116 KB, 1200x1000)

116 KB JPG

Rust is gonna kick your ass. It's coming to get ya.

File: stallman.jpg (96 KB, 333x365)

96 KB JPG

>>110159248
loose is my goose

Rust has big penis programming aura.

back in the day, if you were programming on an 8-bit computer, a C compiler would crash if your source file had more than like 15 unique identifiers in it. you had to modularize the crap out of your code to make it compile and link.

>>110162893
nah, it's definitely a bottom surgery aura

>>110163088
From you.

File: file.png (1.11 MB, 1879x2239)

1.11 MB PNG

>>110139176
User profile view. The top right image is part of the custom style as a background image, it just isn't too large nor correctly resized. Overall, it works

https://github.com/casey/just
Most overengineered project award

>>110163088
c literally has a gaping hole in it.

>>110163181
kek I had that book

>>110158785
>Mad John Carmack has never complimented your programming language?
John Carmack complimented Goose?

File: 1785096424863149.png (454 KB, 1311x1525)

454 KB PNG

>>110163768

>>110163660

>>110163675
>I made a whole programming language that does what competent programmers were already doing except with no granular control
wow! amazing!

>>110163047
cool xD

>>110163768
Why are you mad though?

>>110163793
I don’t like dishonest people.

>>110163047
Totally a good language. Nothing sounds jeet about this type of language at all.

>>110163866
Why are you so obsessed with India? Are you capable of thoughts that don't involve India? Is this a bot?

>>110163889
>the shitlang c is taught first in indian schools
>the shitlang c is used for interviews in india
>"NOOOOOOOOOO THERE'S NO INDIAN CONNECTION YOURE JUST OBSESSED!!!"
Copius maximus.

>>110163889
>muh c
it woudlnt surprise me it actually bait to make non-maid /dpt/ a thing again


i saw the previous one and i was gonna move over but i got b&

i was gonna say "i finally figured out the innards of v0 of my stat engine, ima code now"
i didnt code since then but also bc life happened

maids did post here, though
and the smart ones too i think
dpt might be healing and indiaposter is an important part thereof. i think

File: this-is-my-bait.jpg (67 KB, 500x491)

67 KB JPG

>>110163969

>>110163939
>c is taught in every western school though
you sound mad about teh concept of "lowest barrier to entry, highest skill ceiling"

>>110163959
No, I'm just mad c is such a shitlang and glad Rust fixed it. I learned how to use real macros in Rust and they're so nice.

File: Screenshot from 2026-10-2(...).png (1.2 MB, 3183x1020)

1.2 MB PNG

>>110164014

>>110163969
you ont get what c is about
c is when you know what you want
and dont want to deal with no nannyshit

yeah, ok os-shit is disputable
yeah, ok, webshit is disputable

scientific computation though?
game engines?
simulation?
HPC in general- yeah nah bruv
tf do i care about a fukken nanny if im to tell her to stfu right from the get go?

yes, its the same pos code im posting since 2 years
its my business card
also yes, its tw:wh3 in the background
thats bc i crave violence, momentarily

back to the point:
i want zeroes to turn into ones
very fastly
i dont fucking care about 1Mloc webshit frameworks