Your demise.

>>108840292
Nothing, I'm too tired to game, too tired to code, but I'm also crazy bored so I guess I'm just gonna do chores so I can do whatever I want tomorrow.

The big issue I have now is that it is really difficult to interrogate the SPU2 to see why I am not getting sound on the hardware.
I am pretty sure I have the sound data making it to the IOP RAM, but getting SPU2 state is fairly limited.
Was the DMA from IOP RAM to SPU2 local ram successful?
Are registers set correctly?
At least I am fairly sure that registry access is working correctly because any time you try to read or write somewhere you shouldn't you get exceptions, and I can change and then read the current volume setting and so on.

>>108840563
... are you *sure* there is no good SDK for PS2 development?

>>108840576
I am sure there are no good assembly guides for doing everything in assembly, yes.

>>108840599
Sure, but have you actually *looked* at how games are doing sound on the SPU2?

>>108840601
I am referencing the ps2sdk, yes.

>>108840732
... was that SO hard to say?

File: 1778993596919734.png (169 KB, 1905x1040)

169 KB PNG

I've turned my hms program to a stream.
I thought about adding days, weeks and years but I think that would just be convoluted.

I decided I would do an implementation of Hoffman encoding from scratch. I didn't realize how of a task that actually was. I worked on it for a few hours and I'm not done with the encoding function yet, not to mention decoding.

>>108841252
I meant Huffman encoding. lmao

I think, probably, the DMA is not working on hardware.

>maidnigger continues hijacking threads and shoving his autistic obsession into the OP

Who's better - Anti-UNIX schizo or maid autist?

I'm so fucking done with youtube. Everything is that stupid fucking AI voice.
I'm going to fucking do it... I'm going to read the documentation.

>>108842254
mia mia miamia...mia mia mia mia miaaa

>>108842254
At least it's not an indian voice anymore

I saw a dream that I was using AI to write some documentation for me in real-time. It started going off rails and writing some weird nonsense in one place so I tried to fix it in the middle, but then it slowly started replacing everything with question marks.

>>108842520
Cool story ... you boner.

File: 1764969880139605.gif (525 KB, 498x416)

525 KB GIF

=========================
Summary of SPARK analysis
=========================

-------------------------------------------------------------------------------------------------------------------------
SPARK Analysis results        Total         Flow                                           Provers   Justified   Unproved
-------------------------------------------------------------------------------------------------------------------------
Data Dependencies                40           40                                                 .           .          .
Flow Dependencies                 3            3                                                 .           .          .
Initialization                  152          152                                                 .           .          .
Non-Aliasing                     15           15                                                 .           .          .
Run-time Checks                 897            .                 897 (CVC5 99%, Trivial 1%, Z3 1%)           .          .
Assertions                        3            .                                       3 (Trivial)           .          .
Functional Contracts            386            .    384 (CVC5 79%, Trivial 18%, Z3 3%, altergo 0%)           .          2
LSP Verification                  .            .                                                 .           .          .
Termination                     267          267                                                 .           .          .
Concurrency                       .            .                                                 .           .          .
-------------------------------------------------------------------------------------------------------------------------
Total                          1763    477 (27%)                                        1284 (73%)           .     2 (0%)

Finally got it passing

>>108842520
sounds incredibly realistic for a dream

There are people right now who are perfectly fine with losing the ability to work on their own code if they for some reason lose access to their paid AI subscription and these are the same people who used to call Adobe evil for charging people a subscription for being able to open their files. Why the hypocrisy?

Turns out there is a FIFO register in each SPU2 core I can write half words directly to that should put them in the local memory.
Simpler than getting the samples into IOP RAM and then setting up the DMA, and my plan is to just pack the SPU RAM with a bunch of samples that I can tie to different voices to replicate as many of the configurations of the Gameboy APU channels as I can, so I don't have to be dynamically generating and feeding in samples (most of the time).
But it's going to take me some time to rewrite this shit.

>>108842740
based koder

>>108842254
idk how people ever learned anything from youtube. super high level concepts maybe, but text has always been a far more digestible format as its easier to reference / cross reference other sections and reread something. also it actually goes into all the specific details rather than a quick overfview

File: huff.png (574 KB, 624x830)

574 KB PNG

>>108841252
>>108841263
>huffman

>>> print(freqs)
[(' ', 16), ('a', 1), ('b', 3), ('d', 1), ('e', 13), ('f', 2), ('g', 1), ('h', 4), ('i', 7), ('l', 4), ('m', 1), ('n', 4), ('o', 7), ('p', 1), ('r', 3), ('s', 2), ('t', 4), ('u', 2), ('v', 3), ('w', 4), ('y', 2)]
>>> print(encoded)
011111110111111010000100011011111011001011001110100110011111000111011010110110010011001011111011100100100011011111011001011001110100110011111000100111011010010000110100010111101110100100011011111011001011001110101100001100111110001110110101101001001011010101001001011010101001110011100100000111010100000100111011010011100011010101011110

its alphabetical order insertion

File: decoder_ring.jpg (50 KB, 1000x500)

50 KB JPG

>>108843931

>>108843931
>huffmaid encoding in 37 lines
The most based maidposter

>>108843931
Yeah, calculating the encoding wasn't too awful. I didn't remember the algorithm well though so I did a reasonable amount of reading about it, too. I'm doing all 256 possible bytes so I can handle binary files as well. I also re-implemented a heap and some other things along the way.
I'm still working on the function that actually reads a file and writes the Huffman encoded one though. Before I took a break, I decided my schema for including the key at the top of files and I haven't finished the function to write it yet. I'll get some more work in on it in a few hours, I think.

File: doki doki cat maid club2.jpg (1.34 MB, 5000x2894)

1.34 MB JPG

>>108840292
nice pic

File: bocchi2.png (1.45 MB, 1024x1024)

1.45 MB PNG

>>108842932
>perfectly fine with losing the ability to work on their own code if they for some reason lose access to their paid AI subscription
You can still work on code without AI, the same way you could work on it without an IDE. AI is just a tool for going faster.

This maid poster really is starting to piss me off.

File: 1778436431647941.jpg (1.32 MB, 1800x1012)

1.32 MB JPG

>>108845452
>namefag
>has shoyu in the name
>hides comment
>continues on with day

>>108845452
Transgender Appreciation General.

>>108845452
Why? What did maidposters do to you?

>>108845452
Maid posting board

File: CENK.png (57 KB, 133x175)

57 KB PNG

how the fuck do I teach myself to code these days? i don't care about getting a job but i have a lot of free time in my current stage of life. College was forever ago. I already know basic linux crap and bash scripting so like ???

>>108845775
Pick an easy project.
Start working on it using what you do know, if anything.
When you realize that you don't know how to do something, look it up.
Repeat until you're satisfied that you know enough.

File: perfumer.jpg (218 KB, 850x1286)

218 KB JPG

>>108845775
Put on a maid outfit and read books until you find something interesting you can apply the computer to, then learn whatever you need to learn to apply the computer. Don't try to learn every application of the computer first, that is a waste of time. Think of the computer as a microscope. You wouldn't read a whole microscope manual before trying to look at whatever interests you, you'd look at the relevant parts to be able to apply the microscope to see what you want to look at with it.

>>108845775
leetcode easy & equivilants are good starts. you don't need to set up an environment / worry about building and you have a clear objective that does something CS releated so there are lots of resources online to give you guidance on the problem and to compare your own code against

>>108845832
this is my go-to
if you're into web, a blog is good
my thing for desktop programs back then was a music player, i always wrote one in whatever i was learning
follow a tutorial, then figure out the next bits after that

>>108845452
>calls himself a maid
>won't even seek Medical Assistance In Dying

>>108845992
MAID is for reproductive aged white people who are mildly depressed.

>>108845596
Not all maidposters, just that one in particular.

>>108846114
Which is why we have a DIE initiative to make the MAID program more diverse, you agree with Diveristy, Inclusion, and Equity right?

>>108846213
What did that particular maidposter do to you?

There's a little bit of unix in all of us.

Like cancer.

Or microplastics.

File: ss+(2026-05-18+at+00-47-06).png (127 KB, 580x1084)

127 KB PNG

>>108840292
>What are you working on, /g/?
More SQL optimizations

>>108845775
Well, first decide what area of programming you want to do. Then open official or some other popular website dedicated to that language or technology and read the getting started guide.

>>108840292
Huh, when did /dpt/ turned into this?

>>108847097
It was a maid thread on and off for years, then maids won the culture war and since then it has always been a maid thread.

File: 177906017174440.png (203 KB, 1920x1057)

203 KB PNG

The other day an anon abused me for not using direct system calls.
You were right, this is infinitely faster than putchar.

File: 1779063643142674.png (222 KB, 1920x1055)

222 KB PNG

>>108847242
Wrong image.

>>108847192
Shut the fuck up it was never a maid thread, still isnt. it's a few schizos who want to have a culture war but nobody gives a fuck anymore, they didn't win anything other than a corpse of a """"daily"""" programming thread that gets less than 100 posts in a day now.

File: 1779064456636106.png (98 KB, 1920x469)

98 KB PNG

>>108847284
getchar() seems to be fully buffered by default.
putchar() is line buffered by default.

I genuinely hate you people char letting my fuck up instead of telling me about setvbuf().
You're sadistic.

File: bocchi1.png (1.77 MB, 1280x2238)

1.77 MB PNG

>>108847326
>it was never a maid thread
It is a maid thread now. It was a maid thread before that. It was a maid thread before that. It was a maid thread before that... It has been a maid thread for a long time.

>>108847365
meds

File: maids from bocchi.jpg (228 KB, 850x601)

228 KB JPG

>>108847378
>meds
No, maids!

Finished my prototype functions to produce Huffman encoded files.
Next I'll write some tests and I expect to find that all of my code is shit.
The file is just under 300 lines right now, but half or more is just comments and white space.

File: lilith.jpg (121 KB, 1063x1200)

121 KB JPG

>>108840292
AI made a weird slop architecture with a lot of bugs which were discovered when I started adding assets and testing, so I narrowed the requirements, since I have a better idea of how the game should work and regenerated a bunch of it.

File: __kirin_futaba_channel_an(...).png (92 KB, 512x512)

92 KB PNG

Trying to port a proprietary audio decomp from Assembly to Python despite not being versed in Assembly. And it probably involves bit math.

And I'm on the fence on how much I can trust AI to help me along the way

And on top of that I don't know if I'll be able to test my progress until this is all done

Pray for me

>>108847761
Why would you make a worse port?
If you need abstractions use C.

File: cat sakuya.jpg (169 KB, 600x600)

169 KB JPG

>>108847761
What did you decompile?

>>108840292
A novel way to harden ARC4 came to me in a dream the other night. Silly and probably of little practical use but it stuck with me, so I bashed it out in C then spent the evening running it through the dieharder test suite and it looks okay.

>>108840292
watching a video of someone discussing the bun rewrite, rust vs zig. it seems like zig is doomed because it was meant to be read and written by humans whereas rust is more suitable for LLMs. The idea of a human-centric programming langauge is interesting to me, but perhaps its not worth learning.

>>108848164
the demise of zig has nothing to do with llms lmafo why do you faggots have to make everything about AI all the time? zig is just a pointless memelang being developed at a snails pace by a single mitwit visionary like jai, odin, hare, etc

File: owarida.png (1.4 MB, 1254x1254)

1.4 MB PNG

>>108848164
>watching a video of someone discussing the bun rewrite, rust vs zig. it seems like zig is doomed because it was meant to be read and written by humans whereas rust is more suitable for LLMs. The idea of a human-centric programming langauge is interesting to me, but perhaps its not worth learning.
wow we live in weird times. ruby was designed and caught on for exactly that reason - being nice to read and write for humans. but ya that sounds
>>108847326
imagine being this mad because of cute maid pics. owari da

>>108848194
>ruby
I thought ruby was fake and gay? I wanted to learn it but everyone made fun of me for it, but I hear its easy to learn... Or at least easier than Python... Not that Python was hard to learn... But I'm curious about modern programming languages that are made to be simple and easy just cause I'm curious how much you could do with it.

>>108848254
The reason is libraries. Ruby had Ruby on Rails which made it popular for the web for a short period but eventually it wasn't enough as JavaScript caught up with better libraries and won. Python also has extremly important libraries so thats why it's exploding in popularity. Syntax is only a minor detail.

>>108840292
My own neural opology in c via tensorflow and CUDA and openvino along with optical pattern processing on a seperate box and allophonic speech on another, my primary goal is to have it train only via those inputs

It kind of works, it 'screams' when I go off camera and makes happy noises when I humm to it

>>108848193
ok so what is a non-meme alternative to zig that is meant for human readability?

>>108848280
What makes you think that human readability is even a metric that matters to anyone? AI sucks at Zig because zig is an incomplete language that is missing important features, missing libraries, smaller community, less training data. There is no evidence to suggest that the syntax and human readability has anything to do with it.

File: _560e223b-9c7b-4c33-8996-(...).jpg (245 KB, 1024x1024)

245 KB JPG

>>108848280
>ok so what is a non-meme alternative to zig that is meant for human readability?
golang if you don't want to manage memory. c if you do.

>NVME command queue is called submission queue and is abbreviated as sq
>NVME completion queue is abbreviated as cq

It's the small decisions made by clueless/tasteless people that make programming a miserable experience.

All languages are pointless garbage. Everything should be coded in C

>>108848280
>ok so what is a non-meme alternative to zig that is meant for human readability?
COBOL

File: 1779075519206459.png (12 KB, 322x194)

12 KB PNG

>>108848302
AI sucks at zig because there is not enough data on r*ddit for it to copy.
AI doesn't think it just uses a random number generator against a weighted list of what words usually come after the current word.

File: 1779076066483126.png (35 KB, 704x247)

35 KB PNG

AI IS A FUCKING GIMMICK.

File: noyes.png (1.53 MB, 1254x1254)

1.53 MB PNG

>>108848280
>ok so what is a non-meme alternative to zig that is meant for human readability?

File: 1550094465777.png (938 KB, 850x1200)

938 KB PNG

>>108848462
consider these space leaks

>>108848338
>AI sucks at zig because there is not enough data on r*ddit for it to copy.
I've also found ai is also complete trash at any lisp based language. It just can't fucking match up the brackets or the arguments correctly ever. It'd be hilarious if it wasn't so fucking pitiful.

File: 1779078160766660.png (12 KB, 323x198)

12 KB PNG

>>108848481
It's honestly funny how much of a meme AI is.
Unlimited energy consumption just to copy and paste from r*ddit.

>>108848499
I'm actually kind of curious about that sort of thing going forward. Right now AI has all the code that's ever been posted online. So is it even possible it can get better? We might be perma plateaued right here. All new code ai trains on will be mostly slop gen. I wonder if it'll start degrading.

>>108848553
The next era is AI cannibalism. I've already seen it with Gemini referencing your old chats. It quickly starts to break down.

File: ftt8ve4b76q21.png (1.02 MB, 1920x1080)

1.02 MB PNG

>>108848476
that's a lot to consider. I feel my developer productivity increasing... purely functionally speaking.

>>108848462
>langchans.jpeg
https://bli-aik.github.io/LangGalMemes/

>>108847790
Sometimes when you're decompiling all you're really after is to convert a set of assets, then whatever language you find easiest is best.
I decompiled the decompression/obfuscation of some PS1 game assets using Python once. It didn't have to be fast.

The thing about C++ is that you can just improve the language without asking anyone for permission.

#define cast(type) *(type *)&
#define autocast(type) constexpr inline operator type() const { return *(type *)this; }

>>108846349
Being an AI shill

>>108846349
not him, but maidniggers break the most basic courtesy of not being an avatarfag, and being extremely obnoxious with it

imagine going with your friends to your regular hangout spot but some queer shows up there, declares that spot is for queers and starts rubbing all over you while insisting you're all queers, even if you're not. and he keeps coming back everyday doing the same queer shit
that's the level of obnoxiousness the maidnigger is producing

>NaN is a number
T-thanks js/ts. I hate this shit.

>>108849381
As much fun as it is to poke fun that Javascript's ridiculous type system and the holy trinity or whatever, NaN being a number actually makes sense.
"is a number" means "is a float" and NaN is a valid float value.

>>108848781
I thought C++fags get irrationally angry when you do C-style casts.

>>108847340
Problem isn't that either are buffered. Problem is you still have to waste at LEAST three calls on every single byte (sure, it's not a syscall, but there's still calls and checks for single thread mode or lock acquisition like glibc's lll_lock), and *especially* considering that your conversion could be done entirely with SIMD instructions).

And setvbuf just introduces countless more branches.

>>108849560
Problem is you're just moving the goal posts now..

>>108849585
Lol no: >>108827727
>there's a whole bunch of userspace crap that is endemic for compiled languages too

>>108849621
Go away.

>>108849631
>>108840546

>>108848462
What's the top graphic in this picture? I actually want to see it in full

>>108848462
>not making your own waifu
Are you even a maid?

>>108848193
This.
Just once I'd like to see a language that can export used registers, or doesn't introduce new names for old symbols, or mandates the use of arenas. Instead we just get more layers around userspace crap.

>>108849560
Anecdotally, recently I tried replacing calls to fwrite() with calls to write() (manually buffered with a two page buffer) to try and fix a throughput issue, and startlingly fwrite() actually had higher throughput somehow.

>>108849813
Have you straced both?

>>108849842
Didn't think to at the time. Should've tho. I'll try it later if I remember.

>>108849849
You do that. My guess is that constant mode switching is so slow that it makes up for the slower L2 cache.
>although if you're writing sequentially you might be better off with mmap and ftruncate

File: 1779105082712630.png (155 KB, 905x1056)

155 KB PNG

Where is your God now?

Erlang is King

File: Unix-moment-x-windows.png (27 KB, 648x219)

27 KB PNG

>>108850101
>Where is your God now?

File: 1597602813134.gif (2.34 MB, 320x240)

2.34 MB GIF

>>108850125
>suddenly having Vietnam flashbacks of ten years ago, when I had worked on my own X connector library

File: Screenshot_20260518_142237.png (36 KB, 800x628)

36 KB PNG

>>108850101
God is semihysterically and unbuffed rn

Kill package managers. Behead package managers, Roundhouse kick a package manager into the concrete. Slam dunk a crate into the trashcan. Crucify filthy npm. Defecate in a package manager's parasite repos. Launch package managers into the Sun. Stir fry package managers in a wok. Toss package managers into active volcanoes. Urinate on a package manager's config. Judo throw package managers into a wood chipper. Twist package manager heads off. Report package managers to the IRS. Karate chop package managers in half. Curb stomp pip. Trap package managers in quicksand. Crush package managers in the trash compactor. Liquefy package managers in a vat of acid. EAT package managers. Dissect package managers. Exterminate package managers in the gas chamber. Stomp package manager skulls with steel toed boots. Cremate package managers in the oven. Lobotomize package managers. Stuff package manager heads in the garbage disposal. Drown package managers in fried chicken grease. Vaporize package managers with Magic. Kick all dependencies down the stairs. Feed package managers to other package managers. Slice package managers open with segfaults.

>>108850416
Based windows user

>>108850420
He's right tho. There is absolutely no need for package managers.
Especially software dev package managers. For shipping binaries it might be okay-ish

File: maidposting site.png (3.37 MB, 4000x2200)

3.37 MB PNG

>>108849105

>>108850567
frog board

>mainposting board
You need to be actively hunted down and lobotomized.

>>108850416
Based linux user

>>108849105
top fucking kek, dangerously accurate

I don't care maid or no maid, I just want vibe coders to stay in the vibe coding general and not shill that garbage here.

File: ellen joe.jpg (141 KB, 1103x1103)

141 KB JPG

>>108850670
Do you have the same problem with auto-complete in an IDE? All AI is, is an auto-completer which is orders of magnitude more powerful. Instead of auto-completing your line or your method, it auto-completes your project. I'm also not sure how mentioning AI in general is shilling anything? No actual models/companies were mentioned.

>>108850814
Prompting is not programming, so go away.

Also does anyone know how to create an appimage which actually works? on their documentation they link https://github.com/linuxdeploy/linuxdeploy
the above creates an old style AppImage which has fuse2 issues.

File: iroha.jpg (296 KB, 850x1207)

296 KB JPG

>>108849105
>frequent shitty dive bar with friends
>bar goes out of business due to low traffic
>bar is replaced with maid cafe
>maid cafe is very popular
>go to maid cafe
>see maids
>Why are there maids here?
>This is a bar!
Posting maids is not "avatarfagging". You are a zoomer and you don't know what an avatar is, because forums were before your time.

>>108850846
>Prompting is not programming, so go away.
Is accepting the results of your IDE auto-complete programming?

>>108850862
You're a fucking moron, kill yourself.

>>108850867
I accept your concession.

>>108850872
Yes, I might as well concede because arguing with a retard is like sisyphus pushing a boulder up a hill.

>total noob
>want to code some tools for personal use and also just for the heck of it
>decide on c++
Retard move or no? What would you use/do

>>108850918
You're better off with Perl or Python.

>>108850918
Haskell

>>108850918
Just use AI. Language doesn't matter.

File: 1759539504215486.jpg (11 KB, 230x230)

11 KB JPG

>>108850951

>>108850951
Kill yourself you retarded faggot.

>>108850951
>DON'T LEARN THAT'S BAD FOR YOU
>LEARNING IS BAD AND DUMB AND TRANS
>HAVE A SLOPBOT DO IT FOR YOU INSTEAD AND LEARN FUCKALL
What you said, distilled.

>>108851051
Both positions are valid. Learn all you want. Use the slopbot to automate boring things, or things you already know how to do. Use your knowledge to validate the results of the slopbot.

The human and the slopbot can be friends.

>>108850918
I'd go for Go/C#/Java personally. I've heard many horror stories about c++

>>108849105
make this the next OP btw, they're asking for it

>LEARNING
Sure. Learning. That's why you're stuck on Python.

We really need to lobotomize this autist.

File: wednesday.jpg (222 KB, 1357x1281)

222 KB JPG

>>108850918
Don't use AI. Language doesn't matter.
Just do it.

>language doesn't matter
Your processor disagrees.

>>want to code some tools for personal use and also just for the heck of it
>some weirdo maidfuck thinks the language is important
you niggers should all get a rope in minecraft

>language is important
Yes.

File: moe moe kyun.jpg (56 KB, 750x536)

56 KB JPG

>>108851467
The only thing that matters about language is the language waifu. See >>108848462

The waifus should be wearing maid outfits though.

File: maxresdefault.jpg (71 KB, 1280x720)

71 KB JPG

>>108851490
https://www.youtube.com/watch?v=YUT22fHzGYA

File: prog_maid_waifus.png (2.44 MB, 1672x941)

2.44 MB PNG

>>108851490
here's ai slop version! best I can do.

>C
>Bad: memory abyss
No, that's C++.

Also.

>>108845775
Watch that MIT/Harvard CS50 course which is free
then learn C to get your feet wet (no need to become a pro kernel hacker or whatever), write some small programs and tools, then move on to other languages

>>108851490
No wonder old devs stay loyal to C mommy.
They are using the guide: >>108848570

>late twenties
>start writing big boy projects after many years of depression and procrastination (instead of just killing myself)
>feel retarded, with permanent brainfog
did I get too old or is this just a side effect of jacking off daily?

File: 1430216867881.webm (502 KB, 320x240)

502 KB WEBM

>What are you maids working on?

reminds me of ancient times when i wrote a catgirl interface to monitor my pi cluster...ahh those were the days!

File: jake.png (425 KB, 2427x1359)

425 KB PNG

>>108852174
>did I get too old or is this just a side effect of jacking off daily?
ahh don't worry about it. I didn't start trying to get better until 30s. after 5 years of studying I finally feel mediocre at programming.

>>108852195
lol omega based. truly a master of ui / ux.

File: nekomimi cluster.webm (2.37 MB, 964x576)

2.37 MB WEBM

>>108852231
much fun was had back then
>2015
fuck im old lol

File: yasuna.jpg (100 KB, 798x1046)

100 KB JPG

>>108852037
Best post in thread!

>>108852417
Java should actually wear her ceremony cape, C# should have bigger ecosystem gravity.

>>108851351
women koders will never escape the allegations

File: cow maid.jpg (152 KB, 900x1382)

152 KB JPG

>>108852851
Most CS/math is invented by maids, and most programming is done by maids.

Maid autists must be lobotomized for the good of humanity. There is no other way.

>>108852851
without women koder the US wouldn't have been to the moon

>>108852851
>>108853059
black women koder*

We probably would've been to the moon sooner and more often without women koders.

>>108845856

no sniff raisin booze

>>108849849
So?

>>108850814
>Do you have the same problem with auto-complete in an IDE?
Yes.

>>108850814
>All AI is, is an auto-completer which is orders of magnitude more powerful.
It's demonstrably not, and that's not even a good idea.

File: 91Mb1aMLydL._AC_UF350,350(...).jpg (15 KB, 263x350)

15 KB JPG

Standardizes your path.

>Standardizes your path for the lowest common denominator for the sake of """""portability"""""
We should declare the C standard committee enemy of humanity.

>>108853988
>Rename it the non-standard library
>Suddenly autists are happy

>>108853998
The problem isn't the name, but its use as the standard interface for so many programs.

>>108853998
I'm a turbo autist and I am quite content with the C standard library.
Very portable, very convenient and has optimizations most people don't even think of.

>>108854008
Those programs are made by other people who will never be as smart as you, in fact the retards in this thread don't even deserve to hear your complaints and suggestions, you should write them in a diary somewhere and burn it before someone steals it

>>108854021
>Very portable
>optimizations
Doesn't compute, but then again you're an autist.

>>108854022
Nice try, autist.

>>108853913
Very based.

programming my own synth rn to learn audio programming in c. started with diving deep into portaudio. any other libs that people can recommend?

>>108854033
The C standard library has been optimized by thousands of people and man hours.
You will never beat that except in the tiniest most obscure use case.

You will waste hours to save a nanosecond.

File: bowsette.jpg (83 KB, 750x471)

83 KB JPG

>>108840292
Tested and tweaked the game engine. Have spent the last 4 hours generating art for the game.

>>108854184
>thousands of people and man hours
You are very much off with that number. Throw some more zeroes to make it right.

File: kronii.jpg (170 KB, 850x1159)

170 KB JPG

>>108854184
>You will waste hours to save a nanosecond.
Worth it if you're iterating over something trillions of times.

>>108854184
>The C standard library has been optimized by thousands of people and man hours.
Irrelevant, because you degenerate autists STILL don't understand that any implementation can be only as good as its interface, and the C standard interface fucking sucks because they built it for portability. It literally doesn't matter how many heuristics you slap onto your malloc because you'll never make up for the fact you cannot group allocations or reserve VAS.

And now accept the fact that you're a retarded autist and shouldn't be programming anything ever.

>>108854215
You are the most autistic person on the board and it isn't even close.

>>108854215
maidlocposter...

>>108854227
No, I just happen to be a non-autistic programmer surrounded by autistic retards. Rather than looking at how the hardware runs best your ilk feels good about simple interfaces that automatically lobotomize their implementations, and you don't like me pointing that out.

It's that simple.

>>108847945
Descent 1's audio decompression

Looking back, the comments on the original file helped, though I'm still stumped at where the audio data itself is fed and returned

>>108854235
>maids
Don't conflate me with that subhuman.

>>108850918
you didn't say what you want to build exactly, but yeah it's probably a bad choice. C++ is extremely complex, much more verbose than most languages, and the tooling is a nightmare to deal with.

File: 1779148368781793.png (25 KB, 646x122)

25 KB PNG

>>108854215
I'm gonna trust the old fellas on this one. You're just a screeching autist with delusions of grandeur.

>>108854341
>portability over efficiency
In other words *just* what I have been saying: >>108853988

And what's even better is that most functionality is already provided natively by the big four kernels (NT, Linux, Mac, BSD). Simple compile-time wrappers could easily allow you to use proper arenas or relative directory handles - it's just the bottom-feeding riffraff that cannot be bothered to provide a modicum of standardization for which these autists optimize.

>not that I assume for a moment turbo autists could ever comprehend this

>>108854370
You clearly have a founding PHD in computer science... why are you on /g/?

>>108854399
Because users deserve to be told - in simple terms - just how much autists have been fucking up things ever since. It's my big hope that at some point we will screen every single person for autism and prevent those affected from ever programming again the same way we prevent sex offenders from accessing the internet.
>although I wouldn't say no to putting them all into ADX and pitting them in life-or-death fights either for public entertainment

>>108854414
You should write a dissertation and submit it to Harvard then. Cut the snakes head off.

>>108854420
That could've worked in 1987, before POSIX and NT and Tanenbaum and Torvalds and Cutler (and specifically with Tanenbaum and Torvalds and Cutler). Now the interfaces (NOT IMPLEMENTATIONS!!!) have solidified, and the only course of action left is payback.

tranny thread

>>108854432
Although to further make my point: it's not like the implementations are still fluid either. You don't just reverse 30 years of optimizations for single-object interfaces; if they could do that they wouldn't have bothered with the RESOLVE_CACHED flag for openat2, and we would've already seen massive rewrites for batched mmaps via io_uring.

I JUST FINISHED MY PROGRAM

https://www.youtube.com/watch?v=1mCOT9aU7iI

it's a helper app to import your entire emulation library into Theseus/UIX Desktop, except CEMU. no support for CEMU yet.

check it out!

>>108854184
how is muslc different from libc, i heard someone say its a world of difference and i have to start fucking with docker which fucks with muslc

>>108854455
incel chud detected

File: 177915411238039.png (93 KB, 946x1037)

93 KB PNG

>>108854432
>>108854467
System call maxing it is. You did this.

>>108854767
How much more obvious do I have to make this?

>>108849560
>Problem isn't that either are buffered

>>108849919
>My guess is that constant mode switching is so slow that it makes up for the slower L2 cache.

So, no. I didn't do anything. *You* are the one who thinks that cutting off userspace bullshit takes precedence over proper kernel interface utilization, not the other way round.

>>108854779
How much more obvious do I need to make it that no matter what bullshit you say I and literally the entire industry will continue using the C Standard Library.

>>108854801
And that's why I do >>108854414. So that users can properly shame "you and literally the entire industry" into suicide.

>>108854184
I'm joining the loosely formed breakaway movement to leave UNIX slop behind (Blow, Muratori, Fleury, etc). Enjoy fighting the last war.

>>108854851
What is the purpose of leaving Unix?

>>108854851
>I'm joining the loosely formed breakaway movement to leave UNIX slop behind (Blow, Muratori, Fleury, etc). Enjoy fighting the last war.
lies. muratori left windows. and blow constantly rants about how much windows fucking sucks. windows is just a pile of shit, but game devs are stuck on it because thats where gamers are.

https://www.youtube.com/watch?v=NvZmBF9YL5o
https://www.youtube.com/watch?v=W5s965qdmyU

File: htbi.jpg (30 KB, 397x441)

30 KB JPG

>>108854215
>group allocations
Do you mean putting together already malloced memory into consecutive addresses? Couldn't you manually do that by calling malloc again and copying values to put them in the correct stack frame? They might not be actually adjacent, but they would at least be close.
>>or reserve VAS.
Why is it good to take the virtual address space and not allocate the physical memory?
Both are genuine questions.

>>108855108
I'm writing my own OS.

>>108855189
>I'm writing my own OS.
ah gomen. ganbatte.

File: 1581649854031.jpg (65 KB, 680x680)

65 KB JPG

>Java
>2 enums
>Enum A wants to have an EnumSet<B> as a member
>Enum B wants to have Enum A as a member
>All of this bullshit about enum members being null or otherwise not initializing properly in time for when it's needed
I hate how everything in Java has to be so pointlessly complicated to get working.

>>108855307
Turns out the "solution" was to move everything out into static { } blocks; now my members don't get to be final.
Thanks Pajeet,

>>108855307
You mean statically?

>>108855356
Yes.

This is the Unix philosophy: Write programs that do one thing and do it well. Write programs to work together. Write programs to handle text streams, because that is a universal interface.

>>108855432
Text is the JSON of operating systems. Utter shit that needs to go away.

>>108855701
Nobody wants to deal with your COM object bullshit or whatever it was, Ballmer.

>>108855432
I use Unix because I have to, not because it's better than Windows used to be.

>>108855126
>consecutive addresses
Doesn't need to be consecutive addresses, they just need to belong to the same lifetime. If you have, same, one million objects for each level, then you don't want to iterate through each object in order to free it - you want to discard the entire data in bulk/overwrite it.

>to put them in the correct stack frame
Stacks have limited size.

>Why is it good to take the virtual address space and not allocate the physical memory?
Because making sure that other, unknown, very miserable parts of your program (i.e. libraries or shitty modules) don't write their shit into a virtual address that is reserved for the growth of a structure allows you to essentially "always take the realloc fastpath" - where resizing doesn't require allocating a new memory block, copying the old data into the new block, and releasing the old block because virtual address space has run out. Instead you just keep on adding pages, as your program needs it, without having to commit a huge amount of memory at the very start, thus starving other threads.

>that's what this whole overcommit business on Linux was supposed to solve - they realized that programmers malloced big blocks they didn't end up using because they were just so afraid of realloc's relocations
>so now they just promise you memory even if it's not really backed by physical memory
>and now the out-of-memory killer haunts us all

>>108855432
strace.

>>108855970
>same
say*.

Also:
>Stacks have limited size
, on top of newer objects relying on older ones. The only mechanism stacks know to free memory is by manipulating the one stack pointer register per core; if you want to get rid of an object from 10 calls ago you have to re-arrange the state from the other 10 calls.

With a bunch of loose pages in the VAS you don't have those issues.

I heard there were gfs in here

>>108856020
No gfs here.

>>108856308
):

>>108855307
>2 enums
>Enum A wants to have an EnumSet<B> as a member
>Enum B wants to have Enum A as a member
sounds retarded
what are you actually trying to do?

File: output.png (76 KB, 2555x1369)

76 KB PNG

>>108856569
It's probably not necessary to go too much into the details, but I'm writing a plugin for a game (Old School Runescape/Runelite) and I'm trying to represent some stuff as enums so they can be used as EnumMaps/EnumSets. They're going to be used in computationally heavy enough code that I think using it over another kind of map will be quite beneficial.

I have a bunch of Ports and a bunch of Tasks that involve you travelling between those ports (which is the point of the minigame this is representing). So I want each Port to have a list (or EnumSet really, because I'm doing a lot of set operations) but also have the tasks know what ports they're actually travelling between.

In the end, it's trying to optimize a travelling salesman type problem, obviously using a bunch of heuristics to cut the search space down to actually make it tractable.

(Yes I'm aware that Task 603 and 604 are out of order. That's just how the input data dumped from the game itself is.)

>>108856678
why not just move the EnumSet<Task> outside of Port and into some EnumMap<Port, EnumSet<Task>>?
marginally more indirection in exchange for getting rid of a circular dependency, plus a bit more flexibility with the mapping