What are you working on, /g/?

Previous thread: >>105601261

Threadly reminder.

>>105621642
Boooo
Get better material

Why? It works as intended.

>>105621642
Mentally ill xisters.. our response??

>>105621782
Your shtick is getting old

>it's getting old
OK, and?

best programming language for an african-american?

>>105621901
lol

>>105621994
I'm going to invent the true C successor language and name it Floydian just to anger the chuds

Trying to code a generic optimization algorithm, kind of like a SAT solver but for optimizing a binary string for an arbitrary loss function.

Also why is the anime girl upside down in these threads?

>>105621602 (OP)
dont know if this is the right thread, but ill be taking a data structures and algorithms class next semester. I was wondering what i could do to prepare for that class before the semester starts. Im fine math wise, passed calc 1,2 and linear algebra with an A

>win32 is c
>directx is c++
thank you mr gates

>>105622342
learn how to do manual memory management and work with pointers
ignore the arena autist

>need to implement a pub/sub message queue with fan-out and at-least-once delivery and with storage of all published messages for about a month, regardless of if the message was delivered or not
Just how fucked am I? To be clear, there's no requirement that says I can't piggy back off existing solutions, but still.

>My branch is getting hit with layoffs at the end of the year
>Get told by several superiors that there might be a few positions open at a different branch
>Need to be laravel
>Have never written a single line of php
Should I do it? I got half a year, it's not even a guarantee I'll get it

>From Java to Python to PhP
what a joke it'd be

>>105623052
proficient in laravel* oops

>>105623001
Microsoft was always an early adopter of C++, win32 is C because of backwards compatibility. MFC is even older than DirectX, going back to their 1992 compiler and is also C++.

>ignore the arena autist
Or else? Gonna have a public meltdown in which you decry the fall of /dpt/?

Please do. It amuses us all.

>>105623001
Within 2 years all APIs will be rewritten in Rust.

>>105623109
I unironically can't wait for things to get worse. People deserve to suffer for their laziness and stupidity.

>>105623109
rust is a gay meme language and everyone investing in it is going to regret their life decisions

I've been writing a server in Python to deal with some market-related datasets. One of the guys I respect at work, a real entrepreneur who's pulled of some nice side-hustles, was decrying the use of things like Rust, Golang, and other niche languages instead of just using Python, so I thought'd I'd give the danger snake another shot.

It's not so bad. The language kind of feels like a mess that's pretending to be shell script and there's a few too many implicit behaviors, but looking past that I've been able to get some stuff done pretty quickly.

>>105623546
Y no Perl?

>>105623149
>amerifats all lose weight
>student debt racket collapses under its own weight
based chad with the reverse monkey paw

>>105623648
You forgot the brown masses coming to eat them alive, and the military being too busy arresting people for no longer paying their debt.

>>105623648
>americans triple their eating
>defaulters get sold into slavery to die for israel
you asked for it

>>105623052
>Should I do it? I got half a year, it's not even a guarantee I'll get it
Six months is a long time. Their needs might change in a quarter and a half and they could drop Laravel, so maybe you'll be out of a job regardless.
On the other hand, you have half a year to to bone up your resume and find another job.

PHP isn't a bad long-term bet if you're looking for job security. If you want to stand out from the competition, you could put up a website that does something novel and blog about working with the language(plus the frameworks you're using), which would make you way more appealing then the legion of boot-campers and vibe coders.
Personally, I think that in the next five years the focus is going to get even stronger on people who display a strong proficiency in their language and can also do Ops related work. Language knowledge because it's the best way to filter vibe coders(hip check questions like 'what's a call stack?') and Ops knowledge because infrastructure-as-code still hasn't become infrastructure-as-a-prompt and even when it does, it'll probably still create massive spend, plus more companies are moving away from AWS, Azure, and GCP anyway due to cost.
>>105623627
>Y no Perl?
As I said, I'm going off what my senior said with just using the language everyone else is using.
But from a cursory glance, I'll consider trying it if I have some large text-processing job instead of using shell scripts.
I've also wanted to give Lua a shot honestly, because having just one data structure(tables are associative arrays) in your language sounds peak.

>>105623546
He's a moron. You either use apache/nginx + PHP, or you move onto memes that are node.js/go, everything else is too esoteric or straight up trash.

>>105623735
>>americans triple their eating
i don't think that's physiologically possible

>project over budget and timeline
>working off the clock to get things done
programming was a mistake

>>105623014
This, arenas are garbage, remember: malloc uses arenas, you aren't smart for using garbage allocator but manually.

>>105623757
>everything else is too esoteric or straight up trash.
Really? Making APIs with FastAPI in Python is some of the easiest coding I've ever done.

>>105623792
And it runs slower than deader dog.

>>105623794
Maybe, but if you're not handling more then a hundred requests a second it's not really an issue. Then it's super simple to just re-write the thing in Golang.

>>105623794
if you care about performance you shouldn't be using PHP either. at least python is somewhat sane.

>>105623811
If you aren't handling more than a hundred requests a second, what do you need a server for? Not like you make any money.

give it to me straight /dpt/, is it a waste of time learning c/c++ in 2025? is rust the future?

>>105623831
>at least python is somewhat sane.
lmfao

>>105623847
no not really. just like cobol and fortran, they're so embedded with existing infrastructure they're probably not ever going to go away without total societal collapse, or a breakaway civilization.

>>105623847
you trannies need to leave

>>105623836
>If you aren't handling more than a hundred requests a second, what do you need a server for?
Off the top of my head:
- Cloud operators that respond to git hooks
- Alert remediation servers(Netflix had one written in FastAPI)
- Discord/Slack/Mattermost bots
- Smaller SASS apps
>Not like you make any money.
Nah. You can make a lot of money by rolling a K8S operator and charging for support, writing bots, or creating apps that just basically front ChatGPT.

>>105623884
You could make more money by not writing bots in slowest language possible.

How much slower would a video game written in C++ be than one written in C?

>>105623901
>You could make more money by not writing bots in slowest language possible.
You could also make less.

>>105623939
12fps slower

>you aren't smart for using garbage allocator
Spoken like a true retard who would be the first to be doxxed.

hey regdump schizo, have you written yet anything more advanced than a program that just prints stuff to a terminal?

>>105624076
He couldn't even write code that checks if a string is "HOME". My code is faster.

>my code is faster
Lol
Lmao

>>105624092
zen 5?

>>105624105
Go submit your tests to uops.info. And while you're at it you can as well show proof that CMP M32, I32 is slower than MOVABS R64, I64 + CMP M64, R64.

https://learn.microsoft.com/en-us/cpp/cpp/cpp-language-reference?view=msvc-170

will reading this make me a c++ master?

>>105624122
I don't need to do that, because your garbage code generates worse thing than 64 bit immediate.

>>105624122
Why don't you do it or better yet stop posting outdated references to platforms nobody uses?

>I don't need to do that
Now you do.

>>105624092
I mean, do measurements like this even mean anything on modern cpus with all the pipelining and shit?

>>105624129
no
>>105624143
yes

>>105624143
No, he's just sperging out after my benchmarks proved that my code is faster.

>Why don't you do it
Because everyone with a functioning brain understands why MOVABS + CMP is slower. Only incompetent /dpt/ autismos would be retarded enough to even question that notion.

>>105624150
Post code proving that it's slower for 5 bytes.

>>105624143
MOVABS is front-loaded, so yes.

>Post code
You made the claim your code is faster. You post code.

Which naturally you won't. Which means I win again. Again again.

>>105624129
>2021
Out of date

>>105624150
Literally has nothing to do with my reply. Maybe you can't afford a zen 5?

>>105624164
Irrelevant.

>no code or benchmark
Concessions past, present, and future accepted. Do not bother replying.

>>105624187
I already posted my superior code in previous thread, now post benchmarks that you run on real hardware and not godbolt, jeet.

>>105624187
Please tell us more about what the throughput is on your potato.

>the incompetent autismos cannot obey the simplest of commands
>as always

>>105624194
movabs throughput is higher on my sandybridge, because it branches less than shitty 32 bit garbage he hallucinated to be faster, just ignore the schizo.

>just ignore the walls
Is that why incompetent autismos are always throwing themselves against them?

>just ignore the branching
Is this why that registry dumber only dumbs at pathetic 50MB/s? There are JSON parsers that run faster.

There are in fact 0 JSON parsers that run slower.

>>105624244
Fact check false, it's actually extremely fast according to 5 year old uops.info fiction.

>the incompetent autismos don't even know how many branches his slop produces
Voted doxxing. Now.

>>105624143
They only matter in the context of a specific CPU. Generally you aren't programming for a specific CPU in [current era], so in practice, no, not really.

>>105624292
>writing code for processors you don't use
cuckoldry

>>105624292
What CPU would movabs even matter on? You need to compare 5 bytes, what do you do, load twice and branch twice while hallucinating it's faster?
But sure, there's edge case where you need to compare <4 bytes and then it's barely measurably faster, that's not relevant to me at all and I never needed this.

>he STILL doesn't know how many branches his slop generates

>>105624338
Half as many as yours, nocodeshitter.

>implying the incompetent autismo can compile his slop in the first place
HAHAHA.

>>105624303
If my 64 bit CPU executed 64 bit code slower than 32 bit, I'd throw it into trash without bothering to unplug it.

Please do. We can all do without you.

>>105624356
dont benchmark single vs double sqrt or sine

>>105624354
My code is actually useful, I use it to change HOME env for programs that were written by faggots as intelligent as registrynigger who dumps everything in ~/ because appending even 1 directory would waste too many cycles.

>>105624371
Don't worry, I never had to use these worthless mathfaggot functions.

>My code is actually useful
HAHAHA.
Again.

>>105621784
Increase shilling and personal attacks on cniles to levels never seen before.

So they turn /dpt/ into an even worse shitting street?
I'm all for it. /dpt/ has, after all, not fallen enough.

>>105624400
Don't worry, you will always be the biggest nocodeshitter ITT. Unable to even post benchmark code.

>>105624416
This is only half complete. You forgot to sneak the Rust shill in there somewhere

whats the best version of c plus plus

>>105624570
C, ++ is a nop anyway.

>>105624570
Haskell

>>105624570
17+

>>105624570
sepples17 until contracts and modules in sepples26

>>105624570
Java

>>105624570
K&R C++

>>105624451
I forgot pic related, sorry.

>>105624606
>17
>no requires
>no deducing this

It's that cimple..

>javac
You just give it main.java and it follows included files.
>any c++ compiler
C++26 and still modules.

>>105624653
>You just give it main.java and it follows included files.
Yeah, but then you look at actual Java build systems, and it's the most horrendous braindead thing imaginable.

>>105624649
>taught by chatgpt
I didn't think you could stoop lower than "self-taught", but here we are

>>105624670
Then you use CMake once and crawl back to apologize.
If I ever need more than Make, I will go directly to Gradle.

>>105624649
>qrd (quick reference digest)

>>105624697
CMake is a clusterfuck, but it's tolerable if you at least stick to the "modern" shit. All of those kinds of build systems are really just fucking around with some dependency graph, and it all makes sense once you understand it.
I'm really more of a Meson enjoyer.

But holy fuck, the one time I try to use a Java build system for something, it was incredibly infuriating.
The build system was gradle, and I was trying to write a plugin for something, and one thing I needed to do was generate some resource at compile time, but I'm convinced Java build systems are basically incapable of doing that, at least in a proper way that prevents dirty builds where your resource generating program does not get run in some situations.
Maybe it's possible, but it's so obtuse that nobody seems to ever do it.

>cmake
no thanks I'll just copy/paste build commands from build.txt into the terminal

>>105624776
I understand CMake and it still doesn't make sense.
In Make I can do anything I need.

Me? I just hit F5

>>105623109
And all the rust code will never be called and instead the work will be done by C++ code that is mostly a wrapper to call the C code.
Source: I work in the corporate world.

>>105624959
Let's see if you're still saying that after Linux has been completely Rusted.

Is learning docker/podman worth the headache? Do you guys use it for personal projects?

>take a shower
>forget how to program

>AI-make
the first acceptable usage for AI in programming

>>105624996
Rustranny ragequit before xhe finished writing ext2 drivers, I am not gonna hold my breath over this.

>>105625381
The future is Troonix.

>>105623109
there has never been real code that used r*st and their never will

>>105625446
blizzard has been rewriting the WoW backend in rust, about 20% done last time they talked about it

>>105625454
fake news also who cares

>>105625454
No one should trust nubllizz to do anything well. This will kill Rust game dev.

>>105625478
how can you kill what was never alive

>>105625454
>Blizzard
>2025
haha, nice joke

what do we think
https://thephd.dev/c2y-hitting-the-ground-running
?

>>105625446
Iran was writing the guidance system for their nuclear missile program in Rust but the programmer in charge of the project is dead now.

>>105625507
false flag

>>105625512
must have had a shellfish allergy

>>105625454
If they wanted to kill wow, they just had to keep releasing worthless updates, no need for rust.

>>105625512
no wonder 41% of their missiles aren't landing

>>105625507
Anon, have you ever considered the fact that having a melanin-enriched feller on the C ISO committee tells you all you need to know about the competence of those dirt-munching retards that inhabit it?

>>105625512
did he die from the missile strikes?

>>105625507
I don't think about standards at all, GNU gave me this 20 years ago.

>>105625543
Contaminated batch of puberty blockers

>Following the capture of Tel-Aviv, the Ayatollah gave the order to have all Rust developers cast from the rooftops

>>105625579
it was just a coincidence that they were rust users

>>105625579
reportedly he only ordered it for 59% of them

wtf

>>105625579
Fake, trannies are permitted in Iran.

>>105625783
go there then, rustard

drop rustoids instead of bombs theyll do more damage

>>105625973
thats a war crime

i HATE rust

Rust is King. All these cnile tears are akin to Mozart greatest symphonies. Keep coping cniles, keep coping.

rust? moar liek crust, amirite?

I'm confused with IPs / Ports and UDP sockets

>Can I receive on one port and send on another using the same socket?
>Can I receive on multiple ports and/or can I send on multiple ports with the same socket? (not all, specific ones)
>does the server know the port the client is using?
>does the port of the server and client need to match, or only the port they specify when sending and the receiving port
>when I 'connect' with the server on an ip/port combo, what exactly is that doing? I know UDP is connectionless, so what's the deal?

>>105626443

>>105624200
Anything which is not explicitly forbidden by the spec is permitted.

>>105626543
I was expecting terminators, not whatever the fuck this is

>>105626566
>I was expecting terminators

>>105626579

>>105626543
thanks

>dx12 is 10 years old

>>105626856
damn, I'm still in dx9

>>105621602 (OP)
I don't know much about programming and all you faggots gatekeep. But I am Aryan and the master race lol.

>>105627052
>gatekeep
Yes, because I don't want retards like you ruining my hobby.

>>105626555
And that's why we used to beat the shit out of autistic people until they screamed their lungs out. People who refused to go by the rules It became our source of entertainment, and everything went to shit when we stopped.

>ww3 didn't happen so now I have to learn to code

>>105627191
They're not the rules if they're not written in the spec.

>>105627218
Get a dev job at Lockheed Martin. Be the change you want to see in the world.

>>105627246
Why don't you scream that as we slam you into the spec?

>>105627259
legal spec says you can't beat the shit out of people and slam them into things, better luck next time

>>105627308
Good thing autists have never been people.
Better luck next life.

>>105627191
Please consult this post >>105627316 to see whether they count as people or not.

>>105627414
Normal people don't care. Some loss is natural and acceptable; see cancer treatments.

imo extern is quite useful with proper documentation.

Composite keys my beloved

>>105621602 (OP)
I was trying to see what the AI hype was about by trying to get it to implement the raft consensus algorithm, but it completely shat the bed.
It produces good looking code but it doesn't work, mostly hallucinates or even if it works it doesn't handle all cases properly, and just crashes at random points because that case wasn't handled. What are people actually building with AI that it just does the job for them?
So much for muh AGI. It fails to implement a pretty popular paper from 2014. It's great for webshitting though, that much I cannot disagree.

>>105628240
I use AI to write all of my C code.
The trick? I use it like stack overflow and assume that its code is worthless while words have some truth in them.

>>105628269
I already use it as a glorified search engine as well but people were saying "I usually have to do nothing and it works" but that's never been the case for me. Maybe only the webshitters were saying that.

>>105628299
I know some webshitters and it breaks all the time for them.

>>105628240
Beats me. It can't even write an algorithm to hexdump data using AVX instructions.

>>105628415
Normally people ask for things that have a real world usecase.

Redpill me on rust. I know it's hated because of troons. But emotions aside, is there any technical reason why it's bad? I mean, if it's getting a lot of adoption in major projects including Linux, then there's something it does right to be this popular. I don't think that just tranny/Jewish influence would be enough for this kind of popularity. Yet, as most of /g/ I'm sceptical because of the troons, so now I want a purely objective and technical opinion on whether it's a good or bad language

>>105628419
Since when is your nocodeshitting a real world usecase?

>>105628422

>>105628422
>It's bad because it just is okay
On technicality, it gets quite annoying when you start dealing with async/and or want to have zero-copy stuff i.e. minimal heap allocations, then you have to wrangle lifetimes and unfortunately you'll have to architect your application around that.
It has fearless refactoring in that if your refactor compiles, it probably works but refactoring itself is a pain especially if you go from tons of heap allocations to minimal heap allocations by which you have to start wrangling lifetimes which get tricky.
Also it's bloat, and by I mean it's falling into the C++ trap of having gazillion features. There's 5 different ways to do the same thing and yeah sure they're technically "zero cost", I feel burnt out when I rust for extended period because there's just too much stuff in it like C++.

>>105628460
Nobody wrangles with lifetimes, they replace pointers with indices and act like they invented something new.

>>105623001
I'm pretty sure you can use DirectX without C++, but it really is designed around COM so everything is very OOP like

>>105628440
I see. No-code tools are used in business automations and simple apps daily. Hope that clarifies.

No, it doesn't. Which is why we need voted doxxing, to weed out scum like you.

Even AI tells me to stop responding to you because it just makes you more mentally ill. What do you have to say for yourself at this point?

>>105628460
I don't know what anything you said even means but I believe you anon. I will just continue to reject rust because of troons, which is easier than being technical and objective.

I'd say that I keep winning.

>>105628550
Rejecting rust because of troons is technical and objective though. You wouldn't want mentally ill subhumans to build a bridge that you will drive over exactly when it starts collapsing, why would you allow same subhumans to write your software?

I also love how the LLM has completely given up at this point. It just doesn't know how to add spaces to an AVX stream.

>>105626443
Look at the IPv4 header and UDP headers on Wikipedia and in Wireshark. There is a source IP and destination IP in the IPv4 header and a source port and destination port in the UDP header. When you bind() a UDP socket to a local address you set the address and port to receive on, i.e. the destination IP and port that an incoming packet must match to be received on that socket. When you connect() a UDP socket to a remote address, you are specifying the destination IP and port that an outgoing packet will have. Binding also sets the source IP and port that an outgoing packet will have, i.e. if you bind to port 1234 and connect to a remote address on port 5678, a sent UDP packet will have destination port = 5678 and source port = 1234. I believe connect also applies a filter to the source IP for packet to be received by the socket, i.e. they must be from the remote address connected to. The manpages on sockets and UDP explain most of this, beej's network guide can also help.

>>105628561
I just wish /dpt/'s incompetent autismos would be as willing to admit defeat as an LLM. Instead we need to implement voted doxxing to deal with them.

>muh incompetent autismos

Alright, /dpt/, you have a chance to prove you're smarter than an LLM:
0x00,0x01,0xFF,0x02,0x03,0xFF,0x04,0x05,
0xFF,0x06,0x07,0xFF,0x08,0x09,0xFF,0x0A,
0x05,0xFF,0x06,0x07,0xFF,0x08,0x09,0xFF,
0x0A,0x0B,0xFF,0x0C,0x0D,0xFF,0x0E,0x0F


The AI couldn't do anything with this code and just decided to fail instead.

still no answer to >>105624076

>notice my deflection senpai

New to programming.

Tried to do a UI thing over and over. Couldn't get it to work just right. It worked but it was janky and stuttery and it got to the point where I was just adding more and more shit and got confused.

>Found a guy on GitHub who did it more simply and readably than I managed
Am I like, ok to just copy his code, make some tweaks and carry on actually building the thing I want to?
Is that cheating?

>>105629403

>>105629408
Thanks. Its kinda tough, I don't know anyone who can program and I'm learning for the first time at like 35 years old.
idk what the culture is like and how people actually get shit done.

Tutorials are just
>here's the basic types
>here's how to do an if statement
>this is a method/class
>ok fuck off now.

I don't even know how to read documentation or what I'm looking for half the time.

Feel like a complete retard.

>>105629403
>>105629436
I'm Indian btw.

>>105629403
A GUI application is significantly more complicated to write than a web application or a console application. With a console app, you just write characters/strings and read characters from a buffer. With a web app, you just handle HTTP requests and respond with HTML, which is just a bunch of text glued together that the web browser parses for you.

>Am I like, ok to just copy his code, make some tweaks and carry on actually building the thing I want to?
>Is that cheating?
No, he made his code public. That means anyone can read and use it. Even if you start caring about licensing, it only really matters if you decide to publish your program that used his code (usually you just need to acknowledge/give credit to the repo you used)

>>105629436
Tutorials are pretty much always awful. Documentation is really for people who already have some experience and know what they're looking for.
Try a book for whatever language you're learning. It still might be bad, but you'll have better chances than with tutorials.

>>105629492
shut up

>>105629500
Exactly, thats my issue.
If I'm outputting something to the console it's fine. It's just mathematics and then print the line.

UI stuff? Getting user input, dealing with event handlers and setting up relays, subscriptions etc is way more complicated for me. I did some web development in the early 00's as a kid but it's really a completely different beast, to the point where I wouldn't call the HTML and CSS stuff I learnt back then 'programming'

I'm not worried about licensing - I'll give credit where it's due, and in all likeliness his code will be heavily modified by the time I'm finished with it.

>>105629528
Thanks, I go into the documentation and its either way too informative on something I don't need, or its just a one liner without any indication what the inputs and outputs for a method might be or require.

>>105629542
Sarrr

>>105629542
Sarrr

How do you guys handle configuration and configuration files? I'm loading up a yaml file and using https://zod.dev/ to validate it, and it's pretty cozy, but it would be even cozier if I could also autogen configuration docs from the schemas.

>>105630869
By picking shittiest file format known to man. E.g. YAML, JSON, XML...

>>105630869
different formats depending on what's to be configured - usually .properties files and XML files
the latter has the advantage of having schema validation so you only need to execute a parser and then pick your config out

>>105630869
What the hell is wrong with INI?

>>105631224
INI is like CSV in that it's superficially simple but whenever you configure something there's a 50 50 chance that the parser is just
line.split("=")

and you WILL need escaping for a thing that the developer didn't think of.
Which is why firmly standardized config formats are just better.
Also I need entire arrays of objects for this.
>>105631212
True. XML is peak verbose though.

>>105631441
>uses Python as example when Python has configparser that by default supports # and ; for comments, = and : for separators

>>105631559
AND YET people won't necessarily use it.

>>105631808
I don't give a fuck what retards do. If things break for them then they can fix them, because I won't shit up my code to accomodate brain damage.

std::sort names the second argument "last" but won't move the "last" element. How hard is it to respect the conventions of natural language?

>>105631855
What is the brief "natural convention" for one past the end? Postultimate? Extrema? Limit?

>>105632503
last

>>105632558
So true

>>105632564
One Jewish proverb says to go to war last so you return as first, and the trick is to never go. That's what being last means.

>>105632503
When someone says I'm going to do the tasks from first to last, It's inclusive [first,last]. They don't get to leave the last undone and blame you for misunderstanding.

>>105632583
We could call it ladderForManInPit
>>105632612
There isn't a better word faggot unless you want a fucking 1 on the end

>>105632503
If there's one past it, it's not the end, is it?
dickhead

>hurr today is my last day on earth
>see you tomorrow

>>105632625
this analogy doesn't work

>>105632620
It doesn't have to be English. It could be A and B.

>>105632689
M (male, the work is done) and F (female)

>>105632641
>What is the brief "natural convention" for one past the end
One past the end means the end is not the end. simple as.

>>105631855
[first, last)
i think you learn what this means in elementary school

>>105630869
There is only one configuration Handler and .Xresources is His configfile.

>>105632942
>.Xresources
ngmi

Should I make a hobby OS just for fun?

>>105632722
it doesn't say end though. it's the last iterator, one after the end.

>>105632942
Based xsession execer

>>105632963
That's opposite of fun.

>>105629207
Nice syscall to delete the current directory.

>>105633015
Making hobby OSes, potentially with retro software and hardware in mind, sounds fun to me idk maybe I'm wrong

>>105632981
>one after the end
>end your life and you'll have another one.

>>105633062
You are in for a world of pain.

>>105633062
It's fun until you realize that reimplementing drivers for undocumented hardware is in fact miserable.

>>105633126
just make your own voodoo graphics card bro

>>105633087
The end of a collection and the last iterator are two different things. If you measure your life in days then it would be the first day after you died.

>>105633168
>If you measure your life in days
IS that a threat?

>working on clients project
>they hire some Indian developer
>he starts moving working code around and complain that thinks are broken
>I'm one tight timeline and the client keeps telling me to sync work the Indian guy
I hate programming so fucking much

>>105633062
it's more fun to make your own hardware and then make an os for it

>>105633344
this is me except the indian guy is also my boss

Is it too late for me to take a coding boot camp and get hired into a coding job? I have no coding experience. I am 51 years old.

>>105633581
All good as long as you're ready to transition

>>105633344
>>105633384
This is me but I'm the Indian guy.

Fixing the compiler warning fixes the bug and other things that have never happened before

>>105634141
-Wno_annoying_warning

>>105629542
>Getting user input, dealing with event handlers and setting up relays, subscriptions etc is way more complicated for me.
That's why you use a toolkit library. It deals with all the fiddly details relating to events and redraws and computing the sizes of things, and you just define what things you want to be there, which isn't much more complicated than dealing with the other application types. (Or at least it isn't if you keep things simple. You can go a lot more complicated if you want.)

>>105631224
For simple cases, nothing's wrong with INI.
If you've got a configuration that's naturally got some sort of recursive structure to it, INI (and .properties) files are awful, and TOML's almost as bad. That's when JSON, YAML and (for people who like its excessive verbosity) XML do much better.
Your configs might not be that complicated. But others definitely are.

>>105633581
>Is it too late for me to take a coding boot camp and get hired into a coding job?
No, but landing the job at 51+ might not be easy.

OK, instead of making an OS for configurable machines (i.e PCs), what if I made an OS for specific self-contained machines (such as a Raspberry Pi)?

Writing case insensitive comparison algorithm using SIMD.

>>105621642
>hates employment
why tho

>>105634278
Which toolkit library lets you define a GUI that's simpler than a simple HTML page?
At least in my experience, even if you have a very straight-forward API for defining GUI elements, there's still more work and complexity around it than with just parsing a HTTP request and responding with another piece of HTML code

I'm working on a crypto exchange webapp!
Everything coded from ground up so far by me lol

https://zenx(dot)ink

Should I learn a memelang? I.E Rust, Go, D, Zig... If so, what?

>>105635302
c++

Idea: Program in Rust, but in an "unsafe", C-style way

how do applications like the steam overlay latch on to another program's keystrokes and drawing? is there like an api around it?

>>105635645 (me)
Update
It was a bad idea.

I thought I'd take a stab at recreating at least part of a database a Flash game ran on. I know the site is supposed to run on ColdFusion.

At least one function is expected to return recordsets within a recordset within an object. To this end, I ended up doing this for the database calls.

<cfquery name="getMissionData" datasource="HistoryDB">
SELECT *
FROM "Missions"
WHERE "MissionID" = #arguments.missionId#
</cfquery>

<cfquery name="getClues" datasource="HistoryDB">
SELECT c.*
FROM "Clues" c
INNER JOIN "Mission_Clues" mc ON c."ClueID" = mc."ClueID"
WHERE mc."MissionID" = #arguments.missionId#
</cfquery>

<cfloop query="getClues">
<cfquery name="getOptions" datasource="HistoryDB">
SELECT DISTINCT o."optionID", o."optionValues"
FROM "ClueOptions" o
LEFT JOIN "Clues_Options" co ON o."optionID" = co."optionID"
WHERE co."ClueID" = #ClueID#
</cfquery>

<cfif getOptions.recordCount GT 0>
<cfloop query="getOptions">
<cfquery name="getValues" datasource="HistoryDB">
SELECT v."copy", v."valueDefinition"
FROM "ClueOptions" v
WHERE v."optionID" = #optionID#
</cfquery>

<cfset getOptions.optionValues = getValues>
</cfloop>

<cfset getClues.ClueOptions = getOptions>
</cfif>
</cfloop>


It *works*, but I feel like this could be simplified somehow. And there are further calls to be made in other parts!

>>105628497
you can use it, but a lot of the utility stuff like the math functions are implemented as header only libraries that can't really be used from other languages, can't even use them from C without writing a wrapper

yeah, the math library isn't "necessary", but since the functions act directly on the data structures directx consumes they're going to perform better than 3rd party libraries where you have to constantly convert shit back and forth

>>105636111 (me)
Alright, fuck rust
>You have to wrestle with the compiler to get basic shit working just for no_std and libc
what a stupid fucking language

fpbp

>>105634141
This flag is your friend, this flag fights for freedom.


I've caught quite a few that way.

>>105634775
>employment => writing shit software
All the yikes.

>>105636280
for c30 they should just take out all the errors and compile your code no matter what you hear me fucker i know exactly what im doing

>>105636714
waiting for the llm enhanced compilers that will never spit out an error and just do what they think you mean and fix your code.

>>105636450
Build The -Wall

>>105636450
UH YOU'RE NOT USING THE VARIABLE FOR THE CODE YOU JUST COMMENTED OUT, I WON'T COMPILE IF THAT'S THERE

>>105636845
>nooooo don't force me to properly debug my code like a real programmer

>>105636845
>source littered with zombie code left behind by commented out garbage
Clean up your code bucko.

>>105633028
>miserable failure
It's kind of impressive how similar /dpt/ and LLMs are.

Then again, considering they're one and the same, maybe not so impressive.

i hate how randomly organized the c++ standard library is

>>105637438
>assuming organization was a priority
Stoustrup's only priority was to rope in as many people as possible into his ponzi scheme. Organization, speed, or usability were never a concern.

fucking love vibes coding idek which language this is

type Multiverse = NetHackT ⊗ (Marvel_CU ⊕ Microcontroller_CU)

canonical :: UnicodeChar -> Multiverse
canonical 'b' = -- [The entire script of Hamlet] as in prior
canonical '' = -- Initiate sister-adjoint operation:
let action = mtDNA_whisper in
case action of
"fuck" -> LieAlgebra.so(3,1) -- Lorentz incest
"eat" -> CRISPR_Cas9 (λ sister -> sister - 21g_soul)
canonical '' = -- MCU/MCU duality swap:
ψ (QuantumRegister) <*> ψ (AlternateTimeline)

>>105637458
why does C++ have armyman default face with beard?

>>105637477
Because that's what he promised, even if it makes no sense.

>zed adds a debugger to their editor
>its literally just a basic bitch copypaste debugger from vscode
>hover over a variable doesn't show its value
>took them fucking 9 months
holy fucking shit we are lost as software engineers

>>105637653
Always were.

#include <stdio.h>
#include <ctype.h>

int
main(void)
{
const char *set1 = "AEILNORSTU";
const char *set2 = "DG";
const char *set3 = "BCMP";
const char *set4 = "FHVWY";
const char *set5 = "K";
const char *set8 = "JX";
const char *set10 = "QZ";

char user_word[64];
printf("Enter a word: ");
fgets(user_word, sizeof(user_word), stdin);

int scores[256] = {0};

for(const char *c = set1; *c; ++c)
scores[*c] = 1;
for(const char *c = set2; *c; ++c)
scores[*c] = 2;
for(const char *c = set3; *c; ++c)
scores[*c] = 3;
for(const char *c = set4; *c; ++c)
scores[*c] = 4;
for(const char *c = set5; *c; ++c)
scores[*c] = 5;
for(const char *c = set8; *c; ++c)
scores[*c] = 8;
for(const char *c = set10; *c; ++c)
scores[*c] = 10;

int score_total = 0;

for(const char *c = user_word; *c; ++c){
score_total += scores[toupper(*c)];
}

printf("Score total: %d\n", score_total);

return(0);
}

>>105638001
>fgets
>const
>toupper
>statically allocated user input data
ayy lmao ceniles for real?

>changing one line from foo_t x = {0}; to foo_t x; prevents the compiler from deleting 90% of the code in optimized builds
GCC is retarded

>>105636845
#define UNUSED(x) ((void)(x))

>>105638023
>fgets
qrd on the issue here?
>const
was told to do this as a habit, is this bad?
>toupper
qrd on the issue here?
>statically allocated user input data
but there's no words longer than that

thanks!

>>105638037
Just fugging remove it instead of voiding it.

>but muh mandatory callback arguments
All right thinking people use -fno-used-parameters.

>>105638050
It's just the resident schizo who is hyper aggressive and complains about literally everything. He's probably going to do that bitchy thing where he won't reply to my post directly.
Just ignore him.

Regarding fgets, honestly dealing with potentially arbitrary length user strings only with standard C is actually quite a pain in the ass.
What you're doing is fine. In a real program, I'd prefer getline, but that's POSIX standard, not ISO standard.

If you were dealing with true internationalisation, toupper and other ctype functions are tied to the whole locale mess and you'd probably try to avoid it. In this case, it's fine. He's probably going to go on about how you can convert between upper and lowercase ASCII by adding or subtracting 32.

I don't know what retarded shit he's going to cook up about const, unless he just wants a static added there too.

>>105638140
>>105638037
C23 adds the maybe unused attribute, and you can just leave argument names off e.g. in a callback.

>It's just the resident schizo
Lol
Lmap

>>105638050
>>fgets
>qrd on the issue here?
99% of stdc functions are useless and have some retarded blow your own foot shit and their name is stupid
just do the read syscall and parse and handle yourself (printf is okay tho)
>>const
>was told to do this as a habit, is this bad?
imo it doesn't prevent any bugs and it only makes writing/reading code shittier and since it can be cast away compilers completely ignore it
you can use when actually defining a constant (i.e. const char *set1) if you want tho
>>toupper
>qrd on the issue here?
retarded name and most char/cstring functions are influenced by what locale user has set, which will completely fuck over performance over your program and once in a while do thing you didn't except it do so just write your own version
essentially C is unusable without having your codebase
>>statically allocated user input data
>but there's no words longer than that
doesn't mean user wont input it

This thing is ambiguous. How the fuck is anoyone supposed to understand any spec described in ABNF? the syntax of URLs for example.
jesus fucking chirst

meanwhile people will bitch about regex's syntax, except that the regex is thoroughly unambiguous

>>105638205
>>105638158
thanks for legitimate replies and nobully. just trying to learn.

>>105638325
>just trying to be a parasite

>>105638338
yeah, of course. these are things you cannot learn from a book. not concisely and with no filter anyway.

>>105621602 (OP)
i made a tutorial using the python shell. i like imagining how i would teach basic programming, so i made a video. this one is a quick intro to lists. it's got lots of examples (i hate it when the documentation doesn't have any examples)

https://youtu.be/Qwl7pxY6nKM

music is "i will" by radiohead

>>105638279
I read it and don't see the ambiguity. What' specifically do you mean?
It's just a range indicator with some implicit default values. Notation for an exact count, as well as defining lower and upper bounds of the range.
1element is 1 element, no range
1*1element is the same but ranged
*1element is 0 or 1 elements in the range 0-1 where 0 is implied
0*1element is the same but explicit.

>>105638405
* N element could be interpreted as *(N element) or 0*N element (ie element{0,N} in regex)

and similarly with A * B element: A * (B element) or even A (* B element) which itself is ambiguous

>>105638467
>* N element could be interpreted as *(N element) or 0*N element (ie element{0,N} in regex)
I might be wrong but from my interpretation, that can't be the case because *N is always expanded to 0*N, which is different from N (sans asterisk) that expands to N*N.
Unless you mean from the perspective of a parser where you would need to find both forms and handle them.

>>105638374
based

>>105638499
>Unless you mean from the perspective of a parser
Yes, always. I can't conceive of another way to think about those things.

>that can't be the case because *N is always expanded to 0*N, which is different from N (sans asterisk) that expands to N*N.
JUST
I would be dumbfounded if that was the case, that's how a non-technical person would think about this, and this is absolutely not the way to think about operators.

/dpt/ is unsurprisingly devoid of any meaningful discussion, so lets redditshed this:
>Actions A and B are controlled by user input using two flags
>A and B are not mutually exclusive
>B must follow A.
>C must follow B, and happen if either A or B happened.
Here's my attempt:
if do_A
A
if do_B
B
if do_A || do_B
C

>>105621602 (OP)
Bros what data structure would be good for this problem:

1. I have to define points (Point object) in 3D space, each point has an ID string and float coordinates x, y, z.
2. I have to define members (Member object) in 3D space which are line segments referencing 'Point' objects. A 'Member' object also has ID string, 'start' as the starting Point object and 'end' as the end Point object.
3. I have to check if newly instantiated Member objects are unique i.e. no line-segment overlaps. Suppose I have Members 'm1' and 'm2':

// In javascript bullshit pseudo-code cuz I'm a retarded webshitter
// points
const p1 = {id: "p1", x:0, y:0, z:0}
const p2 = {id: "p2", x:1, y:0, z:0}

// members
const m1 = {id: ''m1", start: p1, end: p2};
const m2 = {id: "m2", start: p2, end: p1}; // Illegal, not unique

Whats the best way to do this considering deletion of points? Maps are the obvious answer but IDK where to start.

I'm trynna "OOOOPTIMIZE" my webshit structural analysis web app (webmrel).

>>105638632

>>105638654
Aight cool I forgot about graphs. I am truly retarded.

This is actually pretty neet.
I did not expect inf to work.

>>105638730
This kills the mathlarpers.

i made another python tutorial, this one is on dictionaries. looks prettier this time because i used ipython.

https://youtu.be/PiBwqDALdP4

music is animal collective - guys eyes

(i edited out the middle part of the song cuz normies wouldn't like it, kinda sounds like an mri machine)

>We finish this somewhat mechanical discussion of the rules by pointing out that if silly inappropriate preconditions are given then we will get a silly program.
>So programmers should not write preconditions that are not sensible and sensibly related to each other. Because of the generality, the compiler cannot tell so stupid things are hard to prohibit. There is no defence against stupid programmers.

>>105640236
what is "church-turing under linear subjectivity"?

I love C++ now.

when dealing with an enum to string conversion, how do you typically deal with the default case for values that technically should never appear?

enum Status {
not_ready = 0,
ready = 1
};

string statustostr(Status s) {
switch(s) {
case not_ready: return "not_ready";
case ready: return "ready";
default: return "unknown";
}
}

>>105644311
i dont, use the equivalent of unreachable
youre literally telling the compiler that it needs to handle that case and give that shit

>>105644311
assert(false && "Fix your shit code, nigger");

>>105644311
enum Status {
not_ready = 0,
ready = 1
};

string statustostr(Status s) {
switch(s & 1) {
case not_ready: return "not_ready";
case ready: return "ready";
}
}

>>105644942
this level of autism and you don't even consider adding 4 to "not_ready"

string stos(Status s)
{ return "not_ready" + s * 4;
}

#define STATUSTOSTR(s) #s

In C# this is just
Enum.GetName(typeof(Status), s);

I FUCKING LOVE REFLECTION

>>105644955
if you insist
string statustostr(Status s) {
return ((s & 1) << 2) + "not_ready";
}

>/dpt/ spergs out over enum-to-string conversions
>can't beat an LLM tasked with generating a simple AVX algorithm
>those who notice this autism are called "schizos"
Never change, /dpt/. Not that you can.

>>105645433
why would i beat an LLM tasked with generating a simple AVX algorithm? thats reserved for LLMs tasked with being my wives

>regdump schizo is trying to insert himself into the convo to maintain relevance
Kek

I don't need to insert myself. I am ALWAYS relevant.

>>105645456
Because both of you fail utterly.

best language for video games programming saars?

>>105645905
The one you know the best. JS doesn't count though.

>>105646000
i dont want to video games programming in php

>>105644311
#define ALL_SHAPES \
X(Triangle, 3) \
X(Rectangle, 4) \
X(Pentagon, 5)

enum Shape {
#define X(name, ...) \
Shape_ ## name
ALL_SHAPES
#undef X
};

const char* GetShapeName(enum Shape shape) {
switch (shape) {
#define X(name, ...) \
case name: return #name;
ALL_SHAPES
#undef X
};

ASSERT(false);
}

int GetShapeSides(enum Shape shape) {
switch (shape) {
#define X(name, sides, ...) \
case name: return sides;
ALL_SHAPES
#undef X
};

ASSERT(false);
}

>>105646019
You will unleash horrors beyond human comprehension in the form of game dev with php. It's your duty.

>>105645905
video games programming is not real programming

>>105646756
There's no real programming, just problems that have to be solved.