[a / b / c / d / e / f / g / gif / h / hr / k / m / o / p / r / s / t / u / v / vg / vm / vmg / vr / vrpg / vst / w / wg] [i / ic] [r9k / s4s / vip] [cm / hm / lgbt / y] [3 / aco / adv / an / bant / biz / cgl / ck / co / diy / fa / fit / gd / hc / his / int / jp / lit / mlp / mu / n / news / out / po / pol / pw / qst / sci / soc / sp / tg / toy / trv / tv / vp / vt / wsg / wsr / x / xs] [Settings] [Search] [Mobile] [Home]
Board
Settings Mobile Home
/g/ - Technology

Name
Options
Comment
Verification
4chan Pass users can bypass this verification. [Learn More] [Login]
File
  • Please read the Rules and FAQ before posting.
  • You may highlight syntax and preserve whitespace by using [code] tags.

08/21/20New boards added: /vrpg/, /vmg/, /vst/ and /vm/
05/04/17New trial board added: /bant/ - International/Random
10/04/16New board for 4chan Pass users: /vip/ - Very Important Posts
[Hide] [Show All]


[Advertise on 4chan]


What are you working on, /g/?
Previous: >>108264010
>>
>more H1B please
>>
opencode wrote me a minecraft clone kek, its dumb but werks
>>
>>108304883
ok it's not even bad. Tried it out. 8/10
Had to remove 1 star because took me a bit to find the smart optimizer. First I was only selecting the imageboard preset, wondered why the output was too big and searched for a smart optimizer button, but then looked into the presets again.
I'll prolly use this every now and then. Thx
>>
>>108307807
np! glad you like it. newer build will be coming out soon so keep an eye out on the github. newer build adding some tweaks and chapter extraction support to crop & trim tab
>>
>>108307184
Building some stuff out in Figma make. Replicating the UI in a .NET app I made for internal enterprise tools.
>>
>>108307184
Reminds me of Terry explaining how MIT niggers have their minds crunched by the CIA. He was trying to justify why his operating system ran all applications in ring 0.
>>
>>108307922
alright. will be pushing the new build to github in about 10 minutes.

new build has tweaks to improve the optimizer for better accuracy to hit the target file size. now uses 3-point quadratic instead of 2-point.

double x1 = (double) crf_lo;
double x2 = (double) crf_mid;
double x3 = (double) crf_hi;
double y1 = Math.log (size_lo_kb);
double y2 = Math.log (size_mid_kb);
double y3 = Math.log (size_hi_kb);

double det = (x2 - x1) * (x3*x3 - x1*x1) - (x3 - x1) * (x2*x2 - x1*x1);

double qa, qb, qc; // quadratic coefficients

if (Math.fabs (det) < 1e-12) {
if (size_lo_kb <= size_hi_kb || Math.fabs (size_lo_kb - size_hi_kb) < 1e-6) {
warning ("Two-point fallback also degenerate — using midpoint CRF");
qa = y1;
qb = -0.1;
qc = 0.0;
} else {
double B_fallback = Math.pow (size_hi_kb / size_lo_kb, 1.0 / (crf_hi - crf_lo));
double A_fallback = size_lo_kb / Math.pow (B_fallback, crf_lo);
qa = Math.log (A_fallback);
qb = Math.log (B_fallback);
qc = 0.0;
}
} else {
double dy2 = y2 - y1;
double dy3 = y3 - y1;
double dx2 = x2 - x1;
double dx3 = x3 - x1;
double dx2sq = x2*x2 - x1*x1;
double dx3sq = x3*x3 - x1*x1;

qc = (dy2 * dx3 - dy3 * dx2) / (dx2sq * dx3 - dx3sq * dx2);
qb = (dy2 - qc * dx2sq) / dx2;
qa = y1 - qb * x1 - qc * x1 * x1;
}
>>
File: Screenshot-Chapter Split.png (531 KB, 2337x2154)
531 KB
531 KB PNG
>>108308094
its been pushed out. smart optimizer has also been integrated into the crop & trim tab for vp9 and x264 re-encoding. there is a toggle to enable / disable it. only shown if re-encoding is enabled. obviously, doesn't work for copy.
in terms of ux for the codec tabs, i'll probably later on move it as a simple toggle instead of hidden in the quality profile combo box.
>>
>>108308132
>>
File: 1772589967654614.png (140 KB, 2267x1233)
140 KB
140 KB PNG
I'm growing more powerful
>>
File: 1408656759085.jpg (63 KB, 500x408)
63 KB
63 KB JPG
>was a full-stack dev
>hate AI hype to death
>wanna develop security or privacy things and earn money doing that
Any recommendations? Languages, firms...
>>
>>108307970
https://i.4cdn.org/wsg/1772784769349455.webm
>>
>>108308178
>nothing happening, 6 times
>>
>>108308215
>christcukistry
>>
>>108308358
that's why your kode is trash
>>
>>108308420
>christcuck having deluded visions again
>>
File: 1762532195275070.png (226 KB, 1080x638)
226 KB
226 KB PNG
>chatgpt: nooo i can't rewrite this header file for you because moral reasons (lol my creators just signed a contract with the us military btw lets ignore that tho)
>claude: here you go king
>>
>>108308577
not programming
>>
>>108308600
compiles on my machine
>>
>>108308577
>>chatgpt: nooo i can't rewrite this header file for you because moral reasons (lol my creators just signed a contract with the us military btw lets ignore that tho)
so much this
>>
>>108307184
Retard thinks people spend their time designing parts that do nothing.
>>
>>108308472
20:3 And the LORD said, Like as my servant Isaiah hath walked naked
and barefoot three years for a sign and wonder upon Egypt and upon
Ethiopia; 20:4 So shall the king of Assyria lead away the Egyptians
prisoners, and the Ethiopians captives, young and old, naked and
barefoot, even with their buttocks uncovered, to the shame of Egypt.
>>
>>108308776
It's more along the lines of, "why optimize the whippletree on that horse-drawn carraige when you could be inventing the automobile instead."
>>
>>108307807
Why is this man dancing like an autist.
>because he's autistic
Yes that much is obvious, but why is he trying to dance?
>>
>>108308981
Because he's based
>>
>>108309015
based on what?
>>
>>108308981
Because he has a soul as opposed to you if you're asking that question
>>
File: FishBowlOfTheBetta.jpg (95 KB, 640x480)
95 KB
95 KB JPG
>>108308981
>this man
>She wasn't allowed to post here when Terry died
Get back to twitter, this here is a boomer forum
>>
> 4.
> GPT-5.4 (openai.com)
> 886 points by mudkipdev 19 hours ago | hide | 699 comments

AI should fucking kill itself
>>
>>108309691
https://news.ycombinator.com/item?id=47265045
https://openai.com/index/introducing-gpt-5-4/
>>
>>108307184
Jon Blow said SICP will make me gay. Is that true? Is SICP really that bad?
>>
>>108308973
too many parts. just ride the horse.
>>
File: 1772186757759350.png (63 KB, 267x157)
63 KB
63 KB PNG
>>108310046
>>
>>108310152
made me spit on my screen
>>
hot take: floats should not exist
all math should be done as ratios of arbitrarily long integers. not sure of the proper way to handle irrational constants yet, though
>>
>>108310269
Floats have their use cases, but more often than not people could just get by with integer math.
>looking at you, Factorio, and your retarded fluid nonsense
>>
>>108310269
Floats have their uses. If nothing else they're convenient prototyping when you care more about getting something that essentially works without need for the best performance or delving into the dark arts of numerical computation.
But you're right in that you probably should be using fixed point computations when and if you can. It's not like float really handles irrational numbers either. Only technically proper way to handle irrationals, that I can think of, is doing symbolic algebra.
>>
thoughts on QML?
>>
I have circles and triangles. Now I only need to fill them. A little more fun today and tomorrow I'll try to get the drawing done with go's image.Image then I can make cool games, no homo
Pizza first tho
>>
>>108310490
Nice.
>>
>>108310269
Then just don't use them
type Decimal is delta 10.0 ** (-2) digits 3; -- [-9.99, 9.99]
>>
>>108310152
link? watched before but can't seem to find.
>>
>>108310269
3 / 2 == 1 when 3 and 2 are ints
I guess we should get rid of ints because they do math wrong
>>
File: 1537999538766.jpg (283 KB, 1125x1476)
283 KB
283 KB JPG
if someone asks you what do you do for a living do you say you are a "programmer" or "software engineer"?

code ninja rockstar wizards need not apply
>>
>>108311167
That's not ints doing the maths wrong, that's you using them wrong.
>>
>>108311227
Maybe you're using floats wrong.
>>
>>108311179
I think of myself as a programmer but I say software engineer. Mostly because software engineer is what it says on my degree and that was the title of every programmer at the place I worked at.
Or that's what I used to say. I'm unemployed now though so who gives a fuck about what I think about anything.
>>
>>108311254
If by using floats wrong you mean using floats when I should be doing fixed point arithmetic, then yes I am. Because I'm lazy and most of the time it doesn't really matter.
>>
>>108311259
I call you a code monkey
>>
>>108311179
>vague understanding of the field they work in
That's most people, innit?
>>
>>108311446
That works too. Basically the only thing I was ever somewhat good at.
>>
>>108308132
>>
>>108311454
The overwhelming vast majority of people, yes.
>>
File: sad-alone.png (107 KB, 498x498)
107 KB
107 KB PNG
this week I achieved somthing that I'm pround of or let's say made the most important progress for this project,
but since I'm the only dev in the company and the only guy that would understand isn't coming back for at least a month, I got nobody to celebrate with.
>feels loney, man

anyway, here's to this junior dev's first time building a project from the ground up and make it production ready
>>
>>108311648
good job anon
>>
>>108311685
thanks anon
>>
Go is so cool.
>>
>>108311648
Good job, anon. Next they they will expect that and a little more of you. If you don't deliver, you'll be fired :^)
>>
>>108311648
Once you get experienced, being solo dev in a project is the best.
>>
>>108308577
>i can't rewrite this header file for you because moral reasons
What can be immoral about rewriting header files?
>>
>>108312003
>being solo dev in a project is the best
It gets lonely. Much better with a buddy to share ideas with.
>>
>>108311648
cheers
>>
>>108311648
Here's to you!
>>
File: 1744600638808295.png (47 KB, 417x364)
47 KB
47 KB PNG
HOLY SHIT YAAAAAAYYYYY!!!!!!!
>>
>>108310269
Irrationals can be defined by a function that generates the to arbitrary precision, semi-symbolic arithmetic basically
>>
>>108310344
>them dark arts of numerical computation
where can you learn about them?
>>
>>108314015
I took a university course as part of my maths major a long time ago. Bare passed it by the skin of my teeth.
>>
>>108313972
Here's to you!
>>
>>108314079
Was there techniques to accurately simulate differential equations kind of thing?
>>
>>108314122
Not explicitly as far as I can remember, but a lot of it was probably applicable to that problem. I did take a separate course in numerical differential equations. But I have to confess I was never very good at higher level mathematics.
>>
>>108310269
>all math should be done as ratios of arbitrarily long integers
what about a ratio of a fixed integer and 2^a fixed exponent
>>
>>108311648
Congrats anon. You did an amazing job. Very proud of you
>>
Is it better to start small with a project and then keep adding onto it later? I have this issue where I have an idea, but end up adding a bunch of features to the point of never doing it.
>>
>>108308981
how dare you. terry died for our sins. his dances lead us not into temptation but delivered us from evil. may we all code with power and glory. amen.
>>
File: firstgraphics.webm (1.72 MB, 1088x521)
1.72 MB
1.72 MB WEBM
After two weeks I'm starting to get graphics out.
>>
>>108314788
>Is it better to start small with a project and then keep adding onto it later?
Usually, yes.
>>
>>108307184
>What are you working on, /g/?
I am finally learning Ada, because of HN. Thanks HN.

with Ada.Text_IO; use Ada.Text_IO;
with Increment;

procedure Subprog is
A : Integer;
begin
A := 0;

Put_Line("Initial A = " & Integer'Image(A));
for I in 0 .. 24 loop
A := Increment(A);

if (I mod 5 = 0) and (I mod 3 = 0) then
Put_Line("FizzBuzz");
elsif (I mod 5 = 0) then
Put_Line("Fizz");
elsif (I mod 3 = 0) then
Put_Line("Buzz");
else
Put_Line("A = " & Integer'Image(A));
end if;
end loop;

end Subprog;



function Increment (I: Integer) return Integer is
begin
return I + 1;
end Increment;


Ready to build a compiler now.
>>
>>108315617
Based. You doing it mostly from scratch or muh agentic koding?
>>
>>108315922
Internet docs and hand crafted assembly in nvim.
>>
>>108315966
Cool. The only emulator I ever did was chip8.
I did write quite a few CPUs in VHDL, but getting some existing roms to work is kinda cool.
I don't care which language tho. Even doing it with AI is fine, it's just a different level of effort needed. I like the hard path most of the time, too
>>
>>108311179
software engineer (unemployable NEET who programs 8 hours a day for the fun of it)
>>
>>108311179
Usually I say software developer (even though i am an engineer with honors and twice best graduate, but i am a humble man (woman))
>>
Ok lets go to mommy eat some breakfast and clean the garden. See you later frens!
>>
>>108315919
Based.
You can use
A'Image
with Ada2012 btw.
>>
>>108314788
It's called feature creep and there's a bazillion articles and videos about it.
>>
>>108316671
Usually feature creep tends to creep up during the development cycle, doesn't it?

>>108314788
Decided for yourself. Is it better to have a working minimum viable implementation, or no implementation at all.
I'd say start as small as you possibly can and build on that. Just start, and get something working.
>>
can the hash algorithm in a hashmap just be anything? like I could literally just make an array X, take an array of numbers, add 2 to each of those numbers and then put them at that index in X?
>>
>>108317060
>can the hash algorithm in a hashmap just be anything?
yes
better hash = better performance
>>
>>108317060
>>108317075
e.g. return 0 for every key is a valid hash, but its going to fuck up your hashmap performance
>>
>>108317075
>>108317086
worth noting this is actually an attack vector (HashDoS) - if an attacker knows your hash function they can craft inputs that all land in the same bucket, turning your O(1) lookups into O(n) and DoSing your server. thats why python and rust randomize the hash seed on each run so you cant predict collisions from outside
>>
>>108317060
You know they actually spend a lot of time researching hashing, you would know this if you were educated.
>>
>>108316652
Thx. Yeah ill learn the basics first and then read about ada 2022.
Since I've been doing vhdl for quite some years now it feels comfy af. So should be quite easy to pick it up, hopefully
>>
File: umigalaxy.mp4 (3.36 MB, 1600x900)
3.36 MB
3.36 MB MP4
>>108307184
>What are you working on, /g/?
I got the 1.0 version of my nextjs/nodejs media tracker website running.
>>
>install appindicator on gnome
>tktray still doesn't work
bros I hate gnome so fucking much
>>
File: 1762704766835633.jpg (65 KB, 1080x854)
65 KB
65 KB JPG
>>108307255
>>108308577
>>
>check on the "implementing DDD" book aka the red one
>takes like 150 pages just to more or less show bounded contexts
>most of it was filler and ramblings
>sounds like my grandpa talking about an old coworker, where you maybe can use 5% of what was said
>the very limited code snippets are "I learned how to code from youtube videos 2 months ago" tier dumb
Jesus christ. I don't know how it became popular. Maybe it gets better later on but man, it's hard to read.
>>
>>108315919
>learning Ada, because of HN.
link?
>>
File: penguin.jpg (22 KB, 396x375)
22 KB
22 KB JPG
>implementing DDD
>>
>>108317230
>thats why python and rust
perl has been doing it since 2003

https://github.com/Perl/perl5/issues/6544
https://perldoc.perl.org/perl581delta
>>
File: 1772922310210.jpg (155 KB, 749x441)
155 KB
155 KB JPG
>>108318712
I have an interest in architecture and patterns, I just didn't know the resources for such things were absolutely terrible.
>>
>>108315919
in OCaml this is just
module Int = struct
include Int
let image (n:int): string = begin
if (n < 0) then
Int.to_string(n)
else
" " ^ Int.to_string(n)
(* end if *)
end
end
module Increment = struct
let increment (i:int): int = begin
i + 1
end
end
module Printf = Printf;; open Printf
open Increment

module Subprog = struct
let a = ref 0;;
begin
print_endline("Initial A = " ^ Int.image(!a));
for i=0 to 24 do
a := increment(!a);
if (i mod 5 = 0) && (i mod 3 = 0) then
print_endline("FizzBuzz")
else if (i mod 5 = 0) then
print_endline("Fizz")
else if (i mod 3 = 0) then
print_endline("Buzz")
else
print_endline("A = " ^ Int.image(!a))
(* end if *)
done
end
end

this is totally wrong as an implementation of FizzBuzz though. There's a spec, you know.
>>
more like implementing bbc into my wife's pussy
>>
Damn, Rob Pike got really angry at AI.
>>
File: rob-pike.jpg (192 KB, 1179x1487)
192 KB
192 KB JPG
>>108319937
this?
>>
>>108319976
Yeah. Yeah.
>>
Is TensorFlow still what people use for ml?
>>
I've known how to plan and write good programs for a very long time now but I don't think I've ever actually done it right even once. It's always a mess that gets cleaned up later, never elegant until maybe the end.
>>
File: 1623637525045 rage.jpg (16 KB, 225x224)
16 KB
16 KB JPG
>>108307184
>Be me
>Look something up on Google
>AI gives me a recommendation
if (Boolean.TRUE.equals(x.getValue()){

I think I'm going to quit computer science and go into aerospace or automotive engineering. I can't take this anymore.
>>
>>108320701
almost took the bait
>>
>>108320701
in Standard ML this is just
structure Boolean = struct
datatype t = TRUE | FALSE
end
infix equals getValue
fun a equals b = a = b
fun a getValue () = a

fun example x =
if (Boolean.TRUE equals x getValue ()) then
print("it's true!\n")
else
print("it's false.\n")

val () = example(Boolean.TRUE)
val () = example(Boolean.FALSE)

just a few less dots
>>
>>108320812
>structure Boolean = struct
> datatype t = TRUE | FALSE
>end
>Boolean.TRUE
>Boolean.FALSE
I don't get it. Why are you not required to write Boolean.t.TRUE and Boolean.t.FALSE?
>>
People say null-terminated strings are bad but what's the alternative?
>>
>>108309039
based
>>
>>108321498
length-terminated strings
>>
>>108321596
So when you can't know the size of the input, you just assign a really large buffer size?
>>
I like Julia.
>>
>>108321606
No. Typically you allocate a small buffer of arbitrary size, say 16 or 64 bytes (or anything you want, allocate more if you think if you know it's going to be big). You append the bytes to the end of the buffer and if it's full you reallocate it (double the capacity and call realloc(), or allocate manually, memcpy() into the new buffer and free the old buffer), and then continue.

If the data is really large, you can always split the data into separate chunks a large but fixed size, say 4 or 16M, and make each chunk points to the next and to the previous chunk, or have an array of pointer to chunks.

Welcome to C where data structures can be whatever you want and be allocated however you want.
>>
>>108319118
>module Printf = Printf
what does this do?
>>
>>108319976
the audacity to not even write the thank you letter yourself is what triggered him I'm sure
>>
>>108321606
"Depends".
>>
>>108322198
Yep. Why bother reading it if there is no effort put in writing it?
>>
File: 1765392109795042.png (1.01 MB, 1080x1580)
1.01 MB
1.01 MB PNG
What will be your new job after coding agents automate swe?
>>
>>108307184
I just started to learn Kotlin and how to build an app in Andoid Studio with Jetpack Compose and am now building a little app for my mom that will help her calculate stitches and rows for her knitting projects.
Does anyone here have experience with Android development and can give me some pointers?
>>
>>108322411
Lisp macro expert
>>
>>108322411
Debugging autogenerated vibecode?
>>
>>108321606
This is no different from null-terminated strings.
>>
>>108307184
Zamn bro that's so deep, people do what they're told to because it's their job and if they don't do it and tell you your ideas are fucking shit you fire them. Really makes you think, that's why Melon is a billionaire and you're not!
>>
>>108322763
do you have 140+iq? can you work 12 hours everyday for minimum wage?
>>
>>108320701
>>AI gives me a recommendation
>I'm going to quit computer science and go into aerospace or automotive engineering
God help us all. I'm never flying on an airplane again.
>>
>>108320696
>I know how to
>I've never done it right
You don't know how to.
That said, all projects are like that. You build small parts, make sure they work, put things together, solve unforeseen problem, clean up, build more parts...
The better you are at planning and coding, the smoother that process goes, but it's never a perfectly smooth ride.
>>
>>108322781
No. I guess I'll just starve to death.
>>
>>108322806
A lede like that is enough discouragement to not read the remainder.
>>
>>108322816
That's fine. Filtering information is a good skill to have.
>>
>>108322411
Swashbuckler.
>>
>>108322774
How many rockets have you launched?
>>
>>108322813
Way ahead of you, pal.
>>
File: RauSOx8M_400x400.jpg (23 KB, 400x400)
23 KB
23 KB JPG
>>108322411
fixing all the mess agents created in corporate software. Planning on quadrupling my income every half year.
>>
File: 1766382278819789.png (15 KB, 501x110)
15 KB
15 KB PNG
am I dumb or is calculus broken? if you plug in x=1 to 3x^2 you get 3 and if you plug in x=2 you get 12
12 + 3 = 15, so your average slops is 15/2 = 7.5

but if you try to calculate the same thing using x=1 and delta x = 1 you get
3(1^2) + 3(1)(1) + 1^2 = 3 + 3 + 1 = 7

7.5 != 7
>>
>>108323086
the delta is supposed to be infinitesimal
what happens when you use a smaller delta
>>
>>108323086
That looks like the derivative of X^3, so 7 would in fact be the average slops between X=1 and X=2. But then you tried to average the derivatives at X=1 and X=2 which clearly isn't going to give you the average slope of Y=X^3, it's going to give you the average slope of Y=3X^2 which is obviously an entirely different graph.
>>
I can't remember ever having used calculus for programming.
>>
>>108323313
karlie kloss...
>>
>>108323319
What have you used calculus for in programming?
>>
>>108323350
https://en.wikipedia.org/wiki/Newton%27s_method
>>
how do I get a job in coding
>no degree
>4 yoe (but it was at some shit startup)
I could pass the interviews but I don't get any !!!!! :(
>>
>>108323367
physics engine?
>>
how the fuck do people find ideas for projects? it seems, like everything i might want already exists and it's very disheartening. i never even feel the need to mod games, since i can just play a different game. was i just born as cattle that can only consume? everywhere, people write, that i should learn programming by fixing problems that i have in daily life, but i have none.
>>
>>108322787
Never said I asked the AI for this, never said this is what I was trying to do.
Why is everyone thinking everyone else is acting in bad faith these days? Were you guys hit too often as kids?
>>108320758
What bait?
>>108320812
Tragic
>>
>>108307184

Holy shit that guy is a fucking moron, jesus fucking christ.
>>
>>108323999
trips of truth
>>
>>108323796
I started making better versions of already existing projects.
>>
>>108322411
keep dreaming
>>
>>108323796
>how the fuck do people find ideas for projects? it seems, like everything i might want already exists and it's very disheartening. i never even feel the need to mod games, since i can just play a different game. was i just born as cattle that can only consume? everywhere, people write, that i should learn programming by fixing problems that i have in daily life, but i have none.
>>108324052
>I started making better versions of already existing projects.
ya I think this is the best advice, but even if you TRY explicitly to make something there's a 99% change you'll end up with something different. that's just how humans are. whenever people copy something they actually are copying the features of the thing they like the most. mac, windows, linux. gimp, krita, photoshop. vscode, notepad++, sublime edit. all apps are the same with slight difference because of peoples preferences.
>>
>>108321339
t is the name of a type, it's neither a value nor a module name. ML separates its sublanguage expressions more carefully than post-OO languages that just make everything an object
>>
>>108323999
I thought that was one of the rare few times where he made sense.
>>
>>108323796
>like everything i might want already exists
Build it anyway, to learn how it works and how to do it. How else are you going to grow your skills so that you can eventually contribute to a real project?
Also you need to fight enshittification by building your own things or maintaining your own fork.
>>
Hey guys, guys... hack the planet! :)
>>
In C there's no better way of comparing a bunch of strings than a chain of if/else if statements? Every other solution seems pointlessly inefficient
>>
>>108326284
comparing how?
>>
>>108326284
Explain how what you think the limitations are are specific to C.
>>
>>108326284
use a thing called a 'build system' to turn your list of strings into a static hash table.
>>
>>108317890
looks good
>>
>>108323389
are you US based?
>>
>>108326657
yes!!! :(
>>
File: 1756100286091758.jpg (140 KB, 892x1157)
140 KB
140 KB JPG
>>108326660
Use one of those video filters, but, instead of pretending to be a whore like the jeets do, you pretend to be a jeet.
>>
Ah, so the C++ OOP is an evolution of C structs, si?
>>
File: 1768048103157580.png (113 KB, 520x280)
113 KB
113 KB PNG
>>108327125
>>
>>108326634
Thanks man, I'm pretty sure this won't go anywhere because getting users is super hard nowadays, but it was a fun building experience and I learned a lot.
I set the website in a way where users can participate in the imageboard without having accounts, but it's still hard to get engagement.
>>
>>108327125
In the most overwhelmingly broad and unhelpful way to think about it, sure.
>>
>>108326284
If it's A LOT of strings and/or they're decently long, maybe it's worth precomputing some kind of key-value datastructure like a hashmap or trie, but most of the time, yeah you'll just have the if else chain.
>>
File: 1769392489402760.png (23 KB, 386x195)
23 KB
23 KB PNG
>>
>>108327125
>OOP is an evolution of structs
develop
>>
>>108326697
tell me about karen
why does she kujo
>>
Ok, I got animated stuff with font and tiles now.
Maybe time to write a lil game
>>
>>108329356
I can also run it on my desktop now and have key input and different scenes implemented
>>
File: 1738465456570502.png (1.1 MB, 1024x1024)
1.1 MB
1.1 MB PNG
>>108323796
Build me a simple Rentry viewer that runs offline and isn't a massive repo meant for publishing. I coded up a simple one but its a pos, bc i couldn't find one.
> why
I use rentry to publish FAQ for others as well as track my own projects, but want local backups in a readable format bc rentry will sometimes delete them for no reason, and runs weird ads... basically I don't trust them as good stewards of information.
>>
>>108323999
He's right tho. I see this over and over irl.
>>
>>108329356
The straightest Embedded C programmer.
>>
File: panel.png (1.48 MB, 2000x2588)
1.48 MB
1.48 MB PNG
>>108329515
it's Go though. I was too lazy for C. Go ist fast enough and makes me iterate faster
And VHDL. All the ledpanel data shovelling is done by an FPGA. Everything else is just Linux writing to a memory area
>>
>>108329425
Based scholar.
>>
So you guys like numbers and uhh and uuuuuuuuh
>>
>>108326284
there's fundamentally no other way to compare them without storing extra metadata
if you want to implement fat strings, you could trim off a lot of comparison time by first comparing the stored lengths. as other anons have pointed out, hashes also work, but it's only efficient if you have to compare the same string multiple times. but if you compute the hash once and then use it for future comparisons, that could save a lot of time
>>
>SIMD
>SPMD
>SSE4
>Branch predictors
>Prefetching
>AoS vs SoA
>Cache associativity
No thanks.
>>
>>108322774
>people do what they're told to because it's their job
Holy fucking sheep
> that's why Melon is a billionaire and you're not!
Unironically one of the reasons, yes. Just not the only one; and there's obv a limit to being a contrarian but it's powerful when directed the right ways.
>>
thought on using namespace std?
>>
>>108330626
criticisms of using it are a cargo cult
>>
>>108330626
no such thing in C
>>
Is there a protocol for time series data with good visualizers out there.
It'd be cool to just be able to shoot some data over a socket and get a nice plot.
>>
>>108331144
shove it into influxdb and make it available with grafana. Both are very easy installs and very flexible.
>>
>>108310490
Is this LOGO?
>>
I feel like a little baby boy.
I have absolutely no fucking clue about game dev.
I just now the basic theoretical game loop. Is game dev just a really big cluster fuck? It takes quite some effort to fit actual game logic into the classic process_input(), update(), render() pipeline, no??
Well anyways I can do something. Building good foundations here. Now just some very basic game logic and then attach SNES controller to the fpga to play it on the ledpanel
>>
>>108331474
What? What Logo? Lego? Legion? Yes, I am legion.
>>
>>108330626
just use a language with an actual module system like rust
>>
>>108331492
>Is game dev just a really big cluster fuck?
yes
>>
I think almost all the criticisms of C++ apply to the standard C++ library, and very many libraries (such as google, llvm, QT) implement their own replacements for the C++ standard library.
There is nothing actually wrong about the standard library, AI will understand the C++ standard library better than other libraries, and the "bloat" isn't actually that big, unless you are measuring based on if your code is under 64kb. But I can't say that C++ has added anything of great value recently.
>>
>>108331510
https://el.media.mit.edu/logo-foundation/what_is_logo/logo_programming.html
>>
>>108331719
>I think almost all the criticisms of C++ apply to the standard C++ library
It's not just stdlib.
>header files
>array decay
>awkward destructuring, tuples, tagged unions, async, etc
>no destructive moves
>no hygienic macros
>no module system
>no zero sized types, proper unsized types, bottom type and such
>no proper function types
>>
>>108331719
>AI
leave
>>
>>108331759
Programming languages just suck, huh
>>
Thoughts on Scala as a backend language?
>>
>>108332092
Just use Haskell instead.
>>
is it hard to learn generics? I'm a pylet who needs to learn a typed language.
>>
>>108331144
"protocols" have nothing to do with what you want
>>
>>108332092
dead language.
>>
>>108331144
CSV
>>
>>108332245
Why would they be? Generics are only a bit of an issue in C because it doesn't really have the concept for such a thing but otherwise they're easy to understand.
>>
>>108332092
if you already need the JVM and have an unreasonable allergy to Kotlin and Clojure ... then you should fix your JVM dependency or your allergies, because Scala's abhorrent.
watch https://www.youtube.com/watch?v=uiJycy6dFSQ . Specific outcomes have been patched out, but the mindset never changed.
>>
>>108332245
generics are no big deal in Nim, Rust, and really most languages with generics.
template hell is a problem that C++ completely brings upon itself, which is vastly improved by D which still has it enough that you can understand the issue.
Nim or D are probably the best languages for a pylet. Nim for more obvious reasons, D for being more into OOP slop.
>>
>>108311648
Noice, anon. Congratz
>>
>>108330626
I use it and I like it.
>>
File: 1759703487429466.png (57 KB, 240x240)
57 KB
57 KB PNG
>>108332245
I failed an interview partly because I couldn't implement generics in my language
I've almost never used them, 99% of the cases it's easier to use static types or interfaces
>>
this is where i found deepdream https://www.tensorflow.org/tutorials/generative/deepdream
is it cringe to use old models to make surreal visions??
>>
>>108332245
>is it hard to learn generics?
Depends how far you want to go down the rabbit hole, really. If you avoid trying to use covariant and contravariant types, they're fairly easy. There's disagreement over how exactly to implement things, but it's not too mad unless you're going full autist.
C++ templates are different; those are parameterized by more than types, and the C++ type system is demented anyway. C++ templates make it far too easy to dabble in sin, and tend to draw you into doing so. Horrible things, defended by idiots unaware of the range of alternatives.
>>
>>108328066
Yes, Jonny Blowie has gay lisp when he speaks.
>>
>>108328066
Elon Musk tier, pathetic
>>
>>108332245
yeah genetics are pretty hard
>>
>>108332633
>I failed an interview partly because I couldn't implement generics in my language
>I've almost never used them, 99% of the cases it's easier to use static types or interfaces
kek brutal. but also this is why I python my interviews. you just have all the fucking tools you could ever want in the stdlib AND you don't have to care about memory or types. just fucking algos baby
>>
>>108332633
>I've almost never used them, 99% of the cases it's easier to use static types or interfaces
To be fair, it's hard to see the right abstractions until you've repeated the code a few times and bad abstractions make things worse. Small projects don't benefit from them all that much.
>>
>>
>>108307184
>03/06/26(Fri)05:15:11
how come this thread lasts for 3 weeks? This forum is dead.
>>
Ok, it's not even THAT bad for my very first game.
Lets not focus on getting it smooth and all the bugs caught, but lets first get it onto the led matrix. After that I can have more fun with games
>>
>>108336140
Faster posts per minute = lower quality board
>>
>>108336936
It just works (tm)

Shoutout to the ffmpeg converter man. The tool works nice. Thx.
>>
>>108307184
Deduplicating archiver and compressor
So far, I have the deduplicating and compression parts done but I'm still figuring out the file format.
>>
>>108336140
anon thats not 3 weeks
>>
File: 1753589868875875.mp4 (349 KB, 1280x598)
349 KB
349 KB MP4
>>108331144 (Me)
Something like this but. Basically like a virtual oscilloscope. Can't really find anything.
An ImGUI plotting library might be a good foundation.
>>
>>108334917
>>108336936
>>108338618
what the fuck am I looking at? it looks cool, but what the fuck is it? you have a giant LED panel and you're playing some web game on it for some reason? are you translating a webgame into a LED panel light array to display? I'm super lost
>>
File: 1504583323024.png (992 KB, 1200x675)
992 KB
992 KB PNG
>>108338807
Made a minor change to my program that improved its speed by 10x
>>
im using advanced vibe coding TTPs to create an abstraction layer that sits between skids and technical proficiency.

kimi k2.5 -> crush -> bmad method

works really fucking well. spec driven development was the hottest thing on the block when I went to jail like 6 months ago so idk if there's anything newer that I should be doing, but what im doing now seems to be working just fine. so ill keep doing what im doing.
>>
>>108323796
https://github.com/DivergentAI/dreamGPT
>>
Fixed an if statement in my python code (type issue) and looked into scans for columns and rows in BQN
>>
>>108340560
>python code (type issue)
Haskell
>>
File: PXL_20250304_215044540~3.jpg (828 KB, 2161x1888)
828 KB
828 KB JPG
>>108340358
>what the fuck am I looking at? it looks cool, but what the fuck is it? you have a giant LED panel and you're playing some web game on it for some reason? are you translating a webgame into a LED panel light array to display? I'm super lost
I have built a giant led panel to show a clock and score on it.

The panels are driven by a Polarfire SoC which contains an FPGA and a risc-v quad core CPU running linux.
To control that shit i have a web interface and a bunch of esp32 remotes.
Now this all worked last year and now I need new stuff to show to my friends august this year. So I want to add games which we can play in the evening.
The web interface that shows the game is just for faster iteration so that I can test it on my desktop

The software (linux side) is all written in Go
>>
Freaking dead as hell today.
>>
File: 1742540970401401.png (208 KB, 393x402)
208 KB
208 KB PNG
STOP WITH THE FUCKING BIT MANIPULATION QUESTIONS
>>
File: tests.gif (1.85 MB, 720x640)
1.85 MB
1.85 MB GIF
I want to get off METATABLES WILD RIDE
>>
File: 1746725507541291.mp4 (1.05 MB, 480x852)
1.05 MB
1.05 MB MP4
best AI for designing (not coding) UIs?
>>
>>108343211
AI is hopeless at design because there's no test for good design. You're better off penciling some mockup in Krita and asking AI to translate that into some design.
>>
File: file.png (122 KB, 1261x895)
122 KB
122 KB PNG
I haven't worked on my job tracking application for like a week and a half because I couldn't decide what I wanted it to look like but I had a fucking breakthrough idea at the gym today and now I'm fully locked in and ready to get back to grinding
>>
>>108342951
Kinda looks like TIS-100
>>
>>108342951
oooh moving numbers
>>
>>108311106
https://youtu.be/RY0WclQU7x4
>>
>>108345404
Pfaha...
>>
>>108345491
i'm not gay
>>
dead forum. Admins please close this thread
>>
did /dpt/ die because of ai?
>>
>>108347602
no, because of discord and moot quitting his job
>>
bros I might need to rework old winform application.
Well the plan is to keep the functional part same so only the UI would be redone so there will be c# code.
The old code is a mess, I was thinking that I would move business logic from the UI parts to controller and redo the forms.
I could also change the UI library but I'm not really familiar with new windows stuff. What do the kids use today, avalonia? Is it any better than winform?
>>
>>108311461
how did you lose your job
>>
>>108311648
keep it up anonymous, or maybe take a break. Whatever floats your boat
>>
>>108322411
cunny coomer if my wife leaves me. Hag coomer if she doesn't
>>
>>108322442
yeah, give up

t. Go/Cpp/Rust dev that tried jeetdev for "fun" on a weekend
>>
>>108322411
Mathematics professor
>>
>>108347832
What the fuck is wrong with you retards? Why is everything always about jeets 24/7?
Kotlin is an easy to learn modern replacement for Java. And at least with android you're making a product that you can use yourself and isn't retardedly abstract. It's 1000x better than spending years learning cpp / rust just to build glorified tally counter systems for companies that no ones ever heard of.
>>
>>108348051
>It's 1000x better than spending years learning cpp / rust just to build glorified tally counter systems for companies that no ones ever heard of.
there is no better or worse desu. It just doesn't fucking matter what you build. The language doesn't matter either.
All languages are equal, but some languages are more equal than others
>>
>>108347602
/dpt/ died when the maidposters got banned.
>>
>>108348868
maidposters should be banned. Musk should escort them to mars
>>
>>108348910
>/g/ is so dead
>why doesn't /g/ make anything?
>maidposters arrive
>they post a lot
>they make lots of stuff
>everybody hates them
>they get banned
>/g/ is so dead
>why doesn't /g/ make anything?
Never improve.
>>
>>108348933
Good blogpost. Have you considered opening an account over at tumblr?
>>
>>108308094
What are you making? What is all that for.
>>
>>
>>108348051
> Saar, please stop this nonsensical abuse sir! Why always "jeet this, jeet that" on subreddit saar? We are working professionals only, sir!
> Saar, listen with full concentration! Kotlin is perfect for gorgeous looks on Android sir! Very modern replacement for Java which is old uncle language saar! You make application and you can use yourself for seeing gorgeous looks and maybe even getting show bobs pic sir!
> Why you want to do 10 years hard labour learning C++ and Rust like some stupid donkey saar? For what sir? FOR WHAT?
> Android with Kotlin is 1000x better saar! You build app today, tomorrow you have 1 lakh downloads saar! Girls are using your app for gorgeous looks and maybe sending you show bobs in DM saar! This is reality saar!
> Please saar, remove anger from body. Kotlin is easy to learn like drinking gange saar! Not like C++ where you spend 3 weeks understanding what is "pointer" and then you cry saar!
> Thank you sir, please visit again for more Android development discussion saar. Perfect for gorgeous looks, I guarantee saar!
>>
>>108348051
Low IQ racism mixed with low effort trolling.
>>
>>108348051
>Kotlin is an easy to learn modern replacement for Java
ironically less modern than actual Java
>>
>>108326284
There are tricks to do it extremely fast, faster than regular hash tables, but you don't want to do it manually so you call a tool in your build process to generate that code heavily coupled with the data.
>>
Function pointers in c inside structs are based and no one says they're not.
>>
>>108349467
type BigCock struct {
Size int
}

func (bbc *BigCock) CalculateSize() int {
return 5
}


checkmate C
>>
>>108349516
tardtyping
>>
>>108349516
In C, you could return that in a short.
>>
>>108321498
structs that have a char array, a length, and maybe a capacity (like how Go does)
>>
>>108349310
with indians it's specieism
>>
daily reminder that vibe koding makes you smooth brane
>>
File: The Nooticer.png (300 KB, 1080x823)
300 KB
300 KB PNG
>>108323796
>i should learn programming by fixing problems that i have in daily life, but i have none.
identifying problems that you have in daily life that can be fixed with programming or automation is a skill not unlike pic related
>>
>>108349904
Incorrect. Enhancing your workflow with AI makes you 100-1000x more productive. Face it. AI is the future. You are either enhancing and evolving yourself with AI right now to become Homo AI or you are a Left Behind Luddite, hopelessly stuck in irrelevancy.
>>
>>108348868
I miss the maidposters.
>>
>>108349912
wrong, im growing chickens and banging your wife. you'd notice if you ever went outside
>>
>>108347602
a couple of the vibe-coding generals hit their bump limits over the past couple of days but then the most recent one got jannied in the past couple of hours
they might have drawn off traffic from here, not sure
>>
>>108349914
I miss regdumper
>>
File: steven.png (334 KB, 1084x777)
334 KB
334 KB PNG
>>108307184
where was this sentiment when he asked people to print out their code at twitter?
>>
>>108349996
in a different script
>>
File: vibing.jpg (99 KB, 800x945)
99 KB
99 KB JPG
>>108349912
>>
>>108349912
I didn't say it makes you less productive, dummy.
Looks like your text comprehension was fucked by AI, too
>>
>>108349467
There's plenty of valid uses for them, but if you just cargo cult putting vtables inside of everything, only for everything to always be set to the same value, that's super retarded.
Not only is it's a lot more code, you basically kill any inline optimisations and are wasting memory.
>>
>>108347665
Ah crap, if there's nobody to help out this guy then I guess my C# questions will not be answered too. I better join them Microsoft forums.
>>
>>108322411
uoh hot lady
>>
>>108323350
Game development. But it’s approximations and linear algebra. Not true calculus.
>>
File: 1763275611502573.png (139 KB, 622x1026)
139 KB
139 KB PNG
https://leetcode.com/problems/maximize-spanning-tree-stability-with-upgrades/
>>
>>108331492
Yes. But you understand it’s a continuous loop until you tell it to end. Now consider “how do I make everything operate within each frame?” Then you’re well on your way. The hardest part is designing data structures and making them cache coherent for speed. Best to make things contiguous arrays of data rather than arrays of objects that contain data. Then you can go further to pack only the most important bits into cache lines so you can do many calculations in nano-seconds. Lots of culling is in order. No need to check collision of one thing against another when it’s obvious they aren’t near one another.
>>
>>108349516
in v this is just
type BigCock = struct {
size int
}

fn (bwc &BigCock) calculate_size() int {
return bwc.size
}

an important difference from go is that in go, Size is mutable and public, but in this v your privates are both private and immutable.
nobody wants to see your dick and your thinly veiled interest in dick pills is even more repulsive.
>>
Morse encoder decoder:
#include <unordered_map>
#include <string_view>
#define Sv std::string_view
static const auto [encodeMap, decodeMap] = []{
struct {
std::unordered_map<char, Sv> encode;
std::unordered_map<Sv, char> decode;
} result;
struct {Sv k; char v;} a[] = {
{".-",'a'},
{"-...",'b'},
{"-.-.",'c'},
{"-..",'d'},
{".",'e'},
{"..-.",'f'},
{"--.",'g'},
{"....",'h'},
{"..",'i'},
{".---",'j'},
{"-.-",'k'},
{".-..",'l'},
{"--",'m'},
{"-.",'n'},
{"---",'o'},
{".--.",'p'},
{"--.-",'q'},
{".-.",'r'},
{"...",'s'},
{"-",'t'},
{"..-",'u'},
{"...-",'v'},
{".--",'w'},
{"-..-",'x'},
{"-.--",'y'},
{"--..",'z'},
{".----",'1'},
{"..---",'2'},
{"...--",'3'},
{"....-",'4'},
{".....",'5'},
{"-....",'6'},
{"--...",'7'},
{"---..",'8'},
{"----.",'9'},
{"-----",'0'},
};
for (const auto& i : a)
result.encode[i.v] = i.k,
result.decode[i.k] = i.v;
return result;
}();

#include <string>
#define S std::string
static S encode(const Sv str) {
S o;
for (const auto c : str) {
const auto it = encodeMap.find(c | (c <= 'Z') << 5);
if (it == encodeMap.end()) {
if (c <= ' ' && !o.empty()) o += " /";
continue;
}
if (!o.empty()) o += ' ';
o += it->second;
}
return o;
}

static S decode(const Sv str) {
S o;
const auto end = str.end();
for (auto it = str.begin(), start = end;; ++it) {
if (it != end) {
if (*it == '.' || *it == '-') {
if (start == end) start = it;
continue;
}
if (*it == '/') o += ' ';
if (start == end) continue;
}
const auto mit = decodeMap.find({start, it});
if (mit != decodeMap.end()) o += mit->second;
if (it == end) break;
start = end;
}
return o;
}

#include <iostream>
int main() {
for (S s; std::getline(std::cin, s);
std::cout << s << '\n' <<
encode(s) << '\n' <<
decode(encode(s)) << '\n'
);
}
>>
>>108322411
Fix all the disasters that automated swe will create.
>>
>>108350855
Beep beep haha.
>>
>>108350939
# sudo apt install sox
for i in . . . - - - . . .; do play -n synth $( [[ $i == . ]] && echo 0.1 || echo 0.3 ) sine 800 >/dev/null 2>&1; done
>>
>>108350855
Cool! But I don't like those DEFINE! Work on your clarity!
>>
>>108350855
in d this is just
import std;

immutable morse = [
"-----", ".----", "..---", "...--", "....-", ".....", "-....", "--...",
"---..", "----.", ".-", "-...", "-.-.", "-..", ".", "..-.", "--.",
"....", "..", ".---", "-.-", ".-..", "--", "-.", "---", ".--.", "--.-",
".-.", "...", "-", "..-", "...-", ".--", "-..-", "-.--", "--.."
];
const int[string] decodeMap;
shared static this() {
decodeMap = morse.enumerate(1).map!(r => tuple(r[1], r[0])).assocArray;
}

auto encode(R)(R str) => str.map!(std.ascii.toLower)
.map!(c => c >= '0' && c <= '9' ? morse[c - '0'] : c >= 'a' && c <= 'z'
? morse[c - 'a' + 10] : c == ' ' ? "/" : " ");
auto decode(R)(R ms) => ms.map!(s => decodeMap.get(s, 0))
.map!(i => cast(char)(i == 0 ? ' ' : i <= 10 ? '0' + i - 1 : i <= 37 ? 'a' + i - 11 : assert(false)));

void main() {
foreach (line; stdin.byLine) {
auto m = line.byDchar.encode;
writeln(m.join(" "));
writeln(m.decode);
}
}

with different round-tripping of invalid input
>>
File: file.png (189 KB, 1642x1119)
189 KB
189 KB PNG
>>108350579
Fun problem
maybe one day I'll be able to remember how to implement union find off the top of my head



[Advertise on 4chan]

Delete Post: [File Only] Style:
[Disable Mobile View / Use Desktop Site]

[Enable Mobile View / Use Mobile Site]

All trademarks and copyrights on this page are owned by their respective parties. Images uploaded are the responsibility of the Poster. Comments are owned by the Poster.