💽Programming Chat v2

[timid quartz]

maybe I misunderstood

[spare quartz]

well yeah this makes sense

[proud creek]

[timid quartz]

@spare quartz

use std::thread;
use std::sync::{Arc, Mutex};

struct Client {
    token: Mutex<String>,
}
impl Client {
    fn do_thing(&self, num: u8) {
        let mut t = self.token.lock().unwrap();
        *t = format!("token{num}");
    }
}

fn main() {
    let client = Arc::new(Client { token: Mutex::new("token".to_string()) });
    
    let c1 = client.clone();
    let c2 = client.clone();
    
    let t1 = thread::spawn(move || thread(c1, 1));
    let t2 = thread::spawn(move || thread(c2, 2));
    
    t1.join().unwrap();
    t2.join().unwrap();
    
    println!("{:?}", client.token);
}

fn thread(client: Arc<Client>, num: u8) {
    client.do_thing(num);
}

this works...maybe I will do this

[spare quartz]

atomic reference counted? no

[timid quartz]

I still think it sucks more than just having independent functions and creating state in each thread

but

¯_(ツ)_/¯

[timid quartz]

because now the user has to wrap their shit in an arc

rather than just

use std::thread;
fn main() {
  thread::spawn(|| thread());
}
fn thread() {
  let request_jar = Jar::new(Some("roblosecurity"));
  // ...
}

[proud creek]

Arc

[spare quartz]

ZToj

e:#
d:#

01100100
01100101

litearlly one bit off

how

[timid quartz]

http://107.132.83.172/crab.mp4

@spare quartz why doesn't it play

[spare quartz]

i dunno!!!

do i look like a cdn expert

[timid quartz]

yes.

[spare quartz]

@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
@         WARNING: UNPROTECTED PRIVATE KEY FILE!          @
@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@
Permissions 0644 for '/home/miko/.ssh/github' are too open.
It is required that your private key files are NOT accessible by others.
This private key will be ignored.

[lavish dove]

ive goten that error a few times

[spare quartz]

@timid quartz http://107.132.83.172/rust.png

is this rust

[timid quartz]

You and your stupid furries

Why didn’t that embed

[spare quartz]

nooooooooo reasssonnn

[timid quartz]

Honestly this is probably the average Rust player

[spare quartz]

rust coder too

[timid quartz]

what does that make the average ada coder then

?av 516049910622191626

[upbeat badge]

signin.aws.amazon.com

Server Avatar

[timid quartz]

I think it makes them that

catsune miko

[timid quartz]

you just ignored the play I made on your name, avatar, and hatsune miku

[spare quartz]

no i didn't . i just went to :3ing

[timid quartz]

one man show

[spare quartz]

AWIT

I NEED TO WORK ON TLS NOW

[spare quartz]

broke everything apart

[timid quartz]

soiled it

[spare quartz]

❓

[timid quartz]

https://tenor.com/view/spongebob-squarepants-spongebob-soiled-it-soiled-gif-5073187

spongebob reference

smh uncultured

[spare quartz]

🤖

i only watched like

4 episodes of spongebob

[timid quartz]

ok time to actually start on the exe reader

....sike

doing sql isntead

[spare quartz]

okay cool

auth works perfetclty

[timid quartz]

open it up

[spare quartz]

hold on i need to push

[spare quartz]

okay its open

[timid quartz]

it's no longer toulene?

@spare quartz

[spare quartz]

theres a password

[timid quartz]

urghhhh

[spare quartz]

you never said open it without one

:3

[timid quartz]

toulene:<passwd>?

[spare quartz]

i can't give you the password

lemme just restart it w/o one

[timid quartz]

dangit

[spare quartz]

good enough

[timid quartz]

lolgetrolled

toulene:lolgetrolled doesn't work

(yes im base64)

WOW your test file laughed at me @spare quartz

[spare quartz]

THAT WAS UNRELATED

[timid quartz]

oh

[timid quartz]

lolgettrolled

two ts

have fun with my .gpr file

[spare quartz]

booooooo

this stinksss

now this is a real gpr

[timid quartz]

overgrown make

[spare quartz]

❌

[timid quartz]

overgrown make

[spare quartz]

   package Install is
      for Artifacts (".") use ("share");
   end Install;

i wonder what this does

[timid quartz]

yours has it too

[spare quartz]

yeah but i dont know what it do

[timid quartz]

leave it up for a sec

[spare quartz]

it is

im gonna go get a slushie

bak

[timid quartz]

you don't have Connection: Keep-Alive support?

[spare quartz]

noooooooope

the only two headers considered are:

Content-Length, Authorization

and authorization is only a subset

why do you ask

[timid quartz]

tried using it on telnet to make your server not disconnect

but

[spare quartz]

how is it disconnecting?

immediately after response?

[timid quartz]

yeah

[spare quartz]

...

ooooooookay

thats something i need to fix

[timid quartz]

yep

how many is it at now @spare quartz

[spare quartz]

there's a thread pool of only 50 so

none

[timid quartz]

oh g

[spare quartz]

are you spamming it

[timid quartz]

how did you know :3

[spare quartz]

👁️.

[timid quartz]

I thought about it at first but deliberately left the HTTP request unfinished

so yeah if TCP connection closes before request is finished, deallocate

glad I helped you catch that

[spare quartz]

https://media.discordapp.net/attachments/782326289381982208/1214437237622833244/IMG_4640.gif

hmm

lemme look at the ada wiki on how to fix this

hmm

something like that

wait oops

[timid quartz]

maybe make the delay a little longer

like

3-5s

?

in case some goober decides to use telnet

but surely there's a way to detect on the fly when the tcp connection closes

like how you did it with your smtp server

[timid quartz]

@spare quartz I assume you roll your own argument parsing

[spare quartz]

nope

GNAT

[timid quartz]

oh

[spare quartz]

i need to see if i can't shorten this tree

[timid quartz]

huh neat

[spare quartz]

i wonder what this has in it

[timid quartz]

and you said I could use GNAT extensions for the exe reader right...

oh wait is that different from a gnat extension

how do I use this

@spare quartz I'm not finding any resources for GNAT.Command_Line cuh

man this is dogshit

almost like nix

basically no docs

[spare quartz]

https://www.adacore.com/gems/gem-138-gnatcoll.command-line and read the file "g-comlin.ads"

[timid quartz]

where tf is g-comlin.ads

[spare quartz]

its a file in your gnat installation

with GNAT.Command_Line; (right click on the part after GNAT.)

[timid quartz]

this is literally just like nix

no docs, read source code

[spare quartz]

well in ada the source code IS the docs

[timid quartz]

THIS IS LITERALLY JUST LIKE NIX

[timid quartz]

this is so ass

[spare quartz]

for quite a while???

[timid quartz]

I dont use vsc normally because the ada lsp is ass

[spare quartz]

its good enough personally

[timid quartz]

says all the errors

[spare quartz]

press x enough times and its fine 👍

[timid quartz]

where tf does "Command_Line_Config" come from

[spare quartz]

you make it

[timid quartz]

oh it's in GNAT.Command_Line

[timid quartz]

this is so ass

[spare quartz]

it's fine

you're just not in the Ada Mindset

[timid quartz]

just look at this

• well documented
• examples given
• easy to use

[spare quartz]

just look at this

• well documented
• fields tell you what they do
• easy to use

[timid quartz]

I shouldn't have to read the source code to figure out wtf shit does

and then figure out how to use it right on my own

[spare quartz]

the alternative is reading the ada rm

[timid quartz]

in which there's nothing about GNAT.Command_Line

[spare quartz]

wait hold on

AHA

https://gcc.gnu.org/onlinedocs/gcc-14.2.0/gnat_rm/

here you go

wait thats

it

[timid quartz]

yep.

trust me.

I looked.

shit's ass

[spare quartz]

okay

blame GNU

they're the ones supposed to maintain this

[timid quartz]

not to mention the only way I'm learning some of the ins and outs of this shit is by asking another person

of which there's like one other person besides you in the ada community

[spare quartz]

hey i learned this with only learn.adacore

it's not that bad trust me

[timid quartz]

idfk how you found out some of the shit you know then

[spare quartz]

reading the files

trust me

they're a lot more descriptive than you'd think

[timid quartz]

hm

ok how to structure

main file -> pe record called to read -> pe record delegates to other records that represent parts of the file?

main file -> giant procedure that just calls other functions defined elsewhere to read certain parts?

a mix?

[spare quartz]

ない限り

U+306A : HIRAGANA LETTER NA
U+3044 : HIRAGANA LETTER I
U+9650 : CJK UNIFIED IDEOGRAPH-9650
U+308A : HIRAGANA LETTER RI

Han unification is an effort by the authors of Unicode and the Universal Character Set to map multiple character sets of the Han characters of the so-called CJK languages into a single set of unified characters. Han characters are a feature shared in common by written Chinese (hanzi), Japanese (kanji), Korean (hanja) and Vietnamese (chữ Hán).

[timid quartz]

@spare quartz what u think

[spare quartz]

i dunno!!! i ad lib file formats

https://t.co/p01SzsJstx

strangest article ever written??

[lavish dove]

its for the scp foundation fr fr

[spare quartz]

https://github.com/Rust-GPU/rust-gpu

[spare quartz]

@timid quartz figured it out

it's not the best solution in my opinion, but it works

only thing to do nnow is https://datatracker.ietf.org/doc/html/rfc8446 ..

which is scary looking

[timid quartz]

What select do

Also why not just copy the way you did it for your smtp server

[spare quartz]

Because …

all remaining copies of the latest SMTP server were destroyed

:<

[spare quartz]

if delay completes, then reading Request is aborted

[spare quartz]

Also I don’t remember what I did

[spare quartz]

Hmmm gonna actually make a commit on my phone

That should be 'Read, not Input

annnnddd commiteddd

https://github.com/Vendicated/segfault-rs/blob/main/src/lib.rs

[tidal wave]

Hell yeah im gonna print this shit and put it on the wall in my dorm

[spare quartz]

Is the rust compiler really that slow

[lavish dove]

Yeah that's rly cool

It's just a frontend for every graphics API tho

[lavish dove]

Wait

Isn't tls supposed to be handled by os

If not why you bothering using a scary looking standard

[spare quartz]

no

handled by libraries usually installed on the system, like OpenSSL

but that's not dependable

[spare quartz]

Task 50: raised CONSTRAINT_ERROR : bad input for 'Value: "�Ѳ=!8P3�#�ilU��(+Qvs�:XM���1/

[proud creek]

i agree

[spare quartz]

￿

[lavish dove]

idk when my computer updates openssl its the most time consuming shit ever

[spare quartz]

same

it also asks me to restart a few services using it

[lavish dove]

discovered libdrm

looks reall cool

kek

[timid quartz]

@spare quartz so like....why have out params when you just have functions

[timid quartz]

Oh and before you say multiple return items, I raise you records

[spare quartz]

what if the same value needs to be modified multiple times

out of the shower now

[spare quartz]

having to save a record to its own variable and then using that to write to multiple sounds like a lot of work

over just having multiple outs

also sometimes it just makes sense like

procedure Main is
  Imaginary_Stream : Stream_Access;
  A_String         : String (1 .. 32);
begin
  A_String := String'Input (Imaginary_Stream); -- ??
  String'Read (Imaginary_Stream, A_String);
end Main

[timid quartz]

Ok but like you’re making the variable anyways because out needs somewhere to put it, so then you just do a little Var := Other_Func (Var);

[spare quartz]

well duh obviously out needs somewhere to go

but if you're using a function that returns a record with multiple fields, you ALSO need to store that record somewhere

and then manually := for each field

https://www.youtube.com/shorts/Ne8S9Zbz2E8?feature=share

[spare quartz]

got very good groundwork in place

Channel := Example_Stream.Wrap_Stream;
Channel.Enable_TLS;
A'Read (Channel.Stream, BlahBlahBlah); -- Reads from the TLS stream
Channel.Disable_TLS;
A'Read (Channel.Stream_Raw, BlahBlahBlah); -- Reads from the raw socket stream
A'Write ...

[lavish dove]

w

[spare quartz]

whats the context

[spare quartz]

ah

[lavish dove]

yeah idk

well it worked for me

[lavish dove]

Omh

[spare quartz]

i should probably redo this one day.....

esp cause i just started spewing accesses

OMG HE'S DOING IT

[lavish dove]

would teh router route traffic without a tcp port

[lavish dove]

whar

[spare quartz]

aera

[lavish dove]

w aera

[spare quartz]

wdym

[lavish dove]

routers have ip's ik that theres some weird protocol that will send what port is used

idk if router just wont route packets to my pc

if theres no port number

[spare quartz]

im sure it would

[lavish dove]

would it go to every pc on my network lol

[lavish dove]

wait but how would it get to my computer

unless its sending to every computer

cuz it doesnt know

wait even worse

[spare quartz]

hmm

[lavish dove]

oh I could just use ipv6 nvm

but that still only covers my house

[spare quartz]

i don't see a clear answer but

[lavish dove]

wont know until I try it

im bored and dont want to work on that yet lmao

[spare quartz]

the router would (probably) save what NAT ip called it and send back whatever it gets

[lavish dove]

theres a list of level 4 protocols in /etc/protocols folder

[spare quartz]

no clue what'd it would do on a portless/connectionless packet though

[lavish dove]

oh shee

wait wtf is nat this sounds complicated

[spare quartz]

network address translation

[lavish dove]

im reading the wikipedia page

[spare quartz]

since the router only has one external ip, and it could have multiple devices, it gives them all their own internal IP and sorts what gets what itself ("router")

[lavish dove]

it looks like spectrum gives my computer an ip specificaly for a single connection and then throws it into a bucket to be reused later when its done

[lavish dove]

wait

like 192.x

?!?!?

[spare quartz]

yeah

[lavish dove]

but how would that work

wait but 192.x isnt in the actual ip header

[spare quartz]

i think it issss

at least the one the router sees in the internal network

[lavish dove]

oh nice

[spare quartz]

it just rewrites it to the external ip (nat)

and decides who to send it to when a response is back

[spare quartz]

but...

[lavish dove]

ya

[spare quartz]

maybe it just handles it as if the router is the device?

like an ICMP request would have the router be the device handling it instead of a fwd'd computer

[lavish dove]

that sounds like how it actually works

but then the router just discards it

[spare quartz]

im talking mostly on ancedotal knowledge so i might be wrong

[lavish dove]

im allways on anecdotal knowledge

lmao

man I want to just get a bunch of switches and pi's and write an entire network stack lmao

[lavish dove]

yeah it onl works if theres a port

NAT

im looking at it and everything is just mapping a port to a private/virtual ip

it seems like its superseded by some DHCP thing though which my router has version numbers listed for

oh nope dhcp is completely different

but supersecedes it

lol

[spare quartz]

DHCP is for allocating internal ips and some other stuff i dont know about

[lavish dove]

oh

the wikipedia page says computers just ask for internal ip

over udp with dhcp protocol

woah my router runs on linux

damn thats actually cool

[spare quartz]

yeah most do

a really slimmed down ver

[lavish dove]

headless linux

impossible to ssh cuz theres no shell lmao

[spare quartz]

i know theres certain ones you can exploit and open a shell on with stuff like telnet

but i've got virtually no exp there

[lavish dove]

sounds like hacker

oh damn everyones asleep

only my phone and laptop have internal ip's for some reason

Dynamic NAT
Dynamic NAT allocates public IP addresses from a pool to devices on a local network on a first-come, first-served basis.
but from reading this doesnt seem true

damn I jsut realized thats ai hallucination lmao

[spare quartz]

was this from google search ai

[lavish dove]

it was below it

from some random website

taht was probably ai generated lmao

crazy cuz i tried doing the chatbot thing on my rtx 2050 and it ran faster than chatgpt.com and was smarter lmao

yeah nat's built on ports

nat wont be helping the packet get to my pc

[spare quartz]

router either discards/handles it itself then

[lavish dove]

i think it either discards or broadcasts it

idk how it would handle it tiself

[spare quartz]

the only one i know where it'll handle it is icmp

a ping request for example

[lavish dove]

oh eah

I cant send data over icmp though

wait Ican but it will allways end up at the router

and is meant to be garbage anyways

wait I forgot

[lavish dove]

my head threw out everything else and had icmp on the burner so started rambling on it lmao

[spare quartz]

Pasting some links here for safekeeping

https://tls13.xargs.org/#open-all

https://rust-lang.github.io/packed_simd/perf-guide/target-feature/rustflags.html

https://gcc.gnu.org/onlinedocs/gcc/x86-Options.html

https://stackoverflow.com/questions/72030595/which-gcc-optimization-flags-affect-binary-size-the-most

and don’t even THINK about calling me a rust… user… just cause your language is interesting doesn’t mean I’d ever use it!!!(*)

[lavish dove]

Haha

[timid quartz]

atp why are you stealing katsumi’s pfp

@lyric mesa PFP theft

[lyric mesa]

[timid quartz]

rust user rust user rust user rust user

atprustages

[proud creek]

rust virus

[timid quartz]

https://rust-lang.org

you should get infected too

[lavish dove]

I tried to infect myself but there too many compile errors

[spare quartz]

wha-

this was my idea…….

[timid quartz]

You need to learn rust before you can jump into a decently-sized project imo

Terry learned that the hard way

[spare quartz]

https://fixupx.com/zack_overflow/status/1858230399404810319

[proud creek]

https://fxtwitter.com/vxunderground/status/1858294594250871101

[spare quartz]

hmm

it is time... to write

[spare quartz]

alright

ive had a brilliant idea

no more +1 or -1

update nevermind its bad!!

[lavish dove]

goddamn

and I still go to stackoverflow because chatgpt will hallucinate answers for me

like I actually haved to wrestle with chatgpt to get useful answers on some random things

but ollama exists and llama feels smarter

or atleast its been more useful

[spare quartz]

Floating point exception (core dumped)

[lavish dove]

why cant it just give 0 like normal division

[spare quartz]

normal division doesn't prescribe a value for dividing by 0

[lavish dove]

it just 0

[spare quartz]

no, it's undefined

[lavish dove]

oh

isnt that 0

[spare quartz]

undefined is undefined

[lavish dove]

undefined is nothing is zero is 0

[spare quartz]

okay undefined is something

[lavish dove]

woah

[spare quartz]

it's just... not defined, it's not a usable value

[lavish dove]

I looked it up

why does divide by zero give a page fault

lmao

[spare quartz]

wha

[lavish dove]

I was looking in the manual

[spare quartz]

it's supposed to give an fpu error 😭

[lavish dove]

no this is normal division on like rax or something

[spare quartz]

oh

[lavish dove]

without the dont fault bit set or something

[spare quartz]

udiv or whatever

[lavish dove]

why cant we use rax

as floating point

why does fpu have to be special

[spare quartz]

hmm

what exception is DE again

[lavish dove]

divide error

[spare quartz]

x87, the 256/512 bit registers

it's hardware level stuff they want to be fast instead of just manually doing it with bits n stuff in the regular registers

[spare quartz]

ough

[lavish dove]

woah theres 512 bit registers

[lavish dove]

im gonna fuckin die

im looking at 1.3.6 exceptions in the intel manual

and it says " as an example, divide by zero gives an exception" and then it shows a page fault

[spare quartz]

https://en.wikipedia.org/wiki/AVX-512

i don't think i have AVX-512 on my computer 😭

[lavish dove]

wait no I was surprsied that they were part of teh fpu

[lavish dove]

I have 256

well I used to have 256

now i have avx-512

[spare quartz]

lemme get my cpuid

[lavish dove]

I dont think the name is avx-512

see 4_2

sse_42

sse4_2

[spare quartz]

i'm pretty sure sse is obsoleted as part of avx

[lavish dove]

oh avx

[spare quartz]

 fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 ht syscall nx mmxext fxsr_opt pdpe1gb rdtscp lm constant_tsc rep_good nopl tsc_reliable nonstop_tsc cpuid extd_apicid pni pclmulqdq ssse3 fma cx16 sse4_1 sse4_2 movbe popcnt aes xsave avx f16c rdrand hypervisor lahf_lm cmp_legacy svm cr8_legacy abm sse4a misalignsse 3dnowprefetch osvw topoext perfctr_core ssbd ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 erms invpcid rdseed adx smap clflushopt clwb sha_ni xsaveopt xsavec xgetbv1 xsaves clzero xsaveerptr arat npt nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold v_vmsave_vmload umip vaes vpclmulqdq rdpid fsrm

shhh you don't get to see my cpus bugs

[lavish dove]

                          opt pdpe1gb rdtscp lm constant_tsc rep_good nopl nonstop_tsc cpuid extd_apicid aperfmperf rapl pni pclmulqdq monitor ssse3 f
                          ma cx16 sse4_1 sse4_2 x2apic movbe popcnt aes xsave avx f16c rdrand lahf_lm cmp_legacy svm extapic cr8_legacy abm sse4a misa
                          lignsse 3dnowprefetch osvw ibs skinit wdt tce topoext perfctr_core perfctr_nb bpext perfctr_llc mwaitx cpb cat_l3 cdp_l3 hw_
                          pstate ssbd mba ibrs ibpb stibp vmmcall fsgsbase bmi1 avx2 smep bmi2 erms invpcid cqm rdt_a rdseed adx smap clflushopt clwb
                          sha_ni xsaveopt xsavec xgetbv1 xsaves cqm_llc cqm_occup_llc cqm_mbm_total cqm_mbm_local clzero irperf xsaveerptr rdpru wbnoi
                          nvd cppc arat npt lbrv svm_lock nrip_save tsc_scale vmcb_clean flushbyasid decodeassists pausefilter pfthreshold avic v_vmsa
                          ve_vmload vgif v_spec_ctrl umip pku ospke vaes vpclmulqdq rdpid overflow_recov succor smca fsrm```

I dont have mmx

wait I have mmx and you dont

[spare quartz]

hold on are you intel or amd

[lavish dove]

amd

[spare quartz]

oh weird

im amd too

[lavish dove]

what cpu

Vendor ID: AuthenticAMD
Model name: AMD Ryzen 5 7535HS with Radeon Graphics

[spare quartz]

5500GT,
CPUID reports as

processor       : 11
vendor_id       : AuthenticAMD
cpu family      : 25
model           : 80

[lavish dove]

your newere than me

CPU family: 25
Model: 68

amd is dropping cpu features cuz compilers jsut arent using them fr fr

[spare quartz]

🙏

[lavish dove]

are you doing this from wsl

[spare quartz]

yeah but the CPUID should still report the same

[lavish dove]

no hyperv is shit

[spare quartz]

let me get WMIC open

oh i forgot this is unbearable to read

[lavish dove]

its cuz your terminal isnt large enough

imagine if wmic cpu | notepad works

and then you can make the font smaller if itw orks

lol

[spare quartz]

just opens notepad unfortunately

[lavish dove]

aw man

[spare quartz]

forced to use >>

[lavish dove]

why is windows adding so much linux stuff

windows isnt linux get over it microsoft

[spare quartz]

wdym

[lavish dove]

windows added sudo

[spare quartz]

wha

[lavish dove]

and also command prompt doesnt have | but > is from multics I think

[spare quartz]

yeah

command prompt is just an MS-DOS emulator at heart though

[lavish dove]

I really feel like it isnt an ms-dos emulator

and its more useful than powershell fr

[spare quartz]

i think powershells more powerful

[lavish dove]

powershell is a linux emulator fr fr

[spare quartz]

it's just a lot more clunkier to use

[lavish dove]

well powershell feels more powerful because everythings documented

cmd you haved to have advanced knowledge gained from teh depths of the internet

[spare quartz]

(which is really just because microsoft sucks at documenting programs in System32)

[lavish dove]

microsoft docs are shit 😭

directx example doesnt even work

[spare quartz]

this sudo sucks

it doesn't even open a shell in SYSTEM like PsExec64

[lavish dove]

it gives admin

[spare quartz]

yeah but that's nothing

with PsExec64 you can literally delete ntoskrnl.exe

[lavish dove]

dyamn

[spare quartz]

this is what a REAL shell looks like

[lavish dove]

authority/system is crazy

[spare quartz]

you can actually taskkill your own explorer and open the SYSTEM one in your session

doesn't last a while but you can technically have a SYSTEM "user"

[lavish dove]

woa

windows has root

[spare quartz]

heres an image with the SYSTEM explorer open

it's quite broken but apps you installed globally will be there

[lavish dove]

woah

wiat

intel is workign on a new extension

for more registers

[spare quartz]

very scary stuff

[lavish dove]

bro just nonchalantly

did that

[spare quartz]

i've done it like once ago

years ago

i'm 99% sure it doesn't brick my pc

[lavish dove]

imagine if it just sees the file not looking right and copies it

when you startup

and it would brick if it didnt do that

lol

[lavish dove]

ew nvcuda

why is cuda so shit man

like nvidia forces you to fucking use this x thing

meanwhile opencl workso n fucking everything

[spare quartz]

huh

ntoskrnl.exe isn't a "native application" (it doesn't link to ntdll.dll)

[lavish dove]

well cuh it runs ntdll.dll

[spare quartz]

it does depend on bootvid, hall, and kdcom though

no idea what kdcom is

[lavish dove]

com port

for debugging

[spare quartz]

oh

[lavish dove]

windows prints entire event long to com if you have kernel debug on I think

[spare quartz]

neat

[lavish dove]

no way windows using bios to boot no way

lol

wait

[spare quartz]

thats gotta be a compat thing

[lavish dove]

I mean microsoft had to change the uefi spec themselves

so windows could boot

so it was even possible to use eufi

cuz it was just so shi

lol

[lavish dove]

a bunch of bootloaders use real mode calls

because they are just so convenient

[lavish dove]

57 bit?

[spare quartz]

https://en.wikipedia.org/wiki/Intel_5-level_paging

[lavish dove]

thought that went only to 51 bits

lmao

[spare quartz]

fsdkmngksmdfs

tried to view deps of notepad.exe

[lavish dove]

average behavior

HEY I COULDVE USED THIS

WHEN I WAS LEARNING PAGING

so many shit graphs and then wikipedia just comes up with one

[spare quartz]

yeah wikipedias a really good resource sometimes

[lavish dove]

I didnt even look at 5 level paging because amd doesnt ahve it

[spare quartz]

actually good graphs of whats going on instead of a non-linear weird spec

[lavish dove]

yea

[spare quartz]

no way

the funny function

[lavish dove]

dear god

harderror

[spare quartz]

user level forced bsods 💪

apparently it's a good way to display messages on the screen outside of the usual windows env

undocumented though

[lavish dove]

using ntraiseharderror as a way to printdebug

[spare quartz]

https://vollragm.github.io/posts/kernel-message-box/

the way people figure this out is beyond me

[lavish dove]

they dissassemble the entire windows kernel and then read through the assembly and figure out how it works

[spare quartz]

windows.h is already bad and they want to go even deeper

oh cool

ntdll.dll also provides a little C lib too

[lavish dove]

w

libc is legit annoying

[spare quartz]

huh

that can't be right

[lavish dove]

and why hasnt intel made a memcopy instruction yet

[lavish dove]

intern decided its time to be funn

[spare quartz]

wouldn't surprise me

the simple disk format menu was written like

20 years ago

and hasn't changed

[lavish dove]

1988 is 2 years after windows 1.0

also 5/11 5:11

5/11 5:11 means its definitely intentional lmao

but I cant find anything on what 511 means

[spare quartz]

naw its gotta be some epoch/bit thing

[lavish dove]

and nothing happened on 5/11

[lavish dove]

894863460

unless its a memory address

[spare quartz]

what i mean is it might be encoded differently from the usual int

like in pkzip

[lavish dove]

why would it be

[spare quartz]

¯_(ツ)_/¯

[lavish dove]

fr

[spare quartz]

im not a pe32 doctor

[lavish dove]

I swear to god its a reference raaah

[lavish dove]

am looking at pe header

cant find a thing even saying linker timestamp

[spare quartz]

maybe it's in a section? if pe has those

[lavish dove]

it does

theres a linker version

yeah idk

[spare quartz]

oh wow

apparently GNAT statically links executables by default

-shared made it go from 3 MB to 281 kb

[lavish dove]

damn

tiny ahh windows program

[spare quartz]

oh

it also breaks the executable

😁

at least the one on linux works

probably doesn't run cause im missing GNAT on windows

[lavish dove]

oof

bruh im gonna compile my kernel with -pie so I dont haved to bother anymore lmao

weird ahh virtual memory

[spare quartz]

raised ADA.IO_EXCEPTIONS.USE_ERROR : Unable to get modification time of the file "C:\Users\Adenosine3Phosphate\Downloads\Auto-GCAS_Saves_Unconscious_F-16_PilotΓÇöDeclassified_USAF_Footage.webm.480p.vp9.webm"

aaaaaaaaaghhhhghg

[lavish dove]

lmao

[spare quartz]

apparently its an error in these two

i just don't know what

[lavish dove]

damn they in the same place

now that im trying to compile with pie I haved to somehow convince ld to work

oh im not even using ld lmao

[spare quartz]

okay

it looks like this is some sort of windows problem instead of ada

apparently with file names using unicode it just errors when trying to get modification time

[lavish dove]

Disassembly of section .data:

00000000 <.data>:
   0:    e9 3b 04 00 00           jmp    0x440```

uhm idk if tats position independent or not

it does jmp 0x440 as if the program is just at 0 instead of being anywhere it wants

cc1: error: code model kernel does not support PIC mode

bruh

GLOBAL_OFFSET_TABLE

WHY DOES THAT EXIST

why??

apparently x86 has some relative jumping thing

so jmp rel 0x40 jumps forward 0x40 ez pz

but gcc cant use that cuz gcc is too smart apparently

[spare quartz]

https://gcc.gnu.org/bugzilla/show_bug.cgi?id=104421

aha

found you

ohhh

it's cause they were using GetFileAttributesExA

GetFileAttributesExW ftw

[spare quartz]

compiling gcc for the 9 billionth time

[timid quartz]

@spare quartz what the hell

I know that's a lot of images

but

this shit is little endian apparently/?

little endian on 2-byte boundaries

also why is it 16#455000# when it should have an additional null byte what

Yeah I print the index and it’s at 132 not 133 (starts at 128)

Unless it needs to start at index 129…?

[timid quartz]

yeah it does because ADA USES 1-INDEXING

[timid quartz]

@spare quartz is there a fairly clean way to do bit vectors in ada

namely reading a bit vector into a record

so like

type Flag_Record is record
  Flag_1 : Boolean;
  Flag_2 : Boolean;
  Flag_3 : Boolean;
  Flag_4 : Boolean;
end record;
Flag := 2#0110#;
--  Take binary flag -> put into struct

or do you just have to brute force it

[spare quartz]

Mf didn’t read the array definition

[timid quartz]

1-indexing SUCSK

[spare quartz]

with Pack; or with Size, but strange things might happen if it’s not a full byte and you try reading

[timid quartz]

it's a full byte

multiple...full bytes...

[spare quartz]

..?

Is the record supposed to be 4 bits, or..?

[spare quartz]

https://gcc.gnu.org/onlinedocs/gcc-14.1.0/gnat_rm/Security-Hardening-Features.html

https://gcc.gnu.org/onlinedocs/gcc-14.1.0/gnat_rm/Aspect-Local_005fRestrictions.html

The attribute typ'Deref(expr) where expr is of type System.Address yields the variable of type typ that is located at the given address. It is similar to (totyp (expr).all), where totyp is an unchecked conversion from address to a named access-to-typ type, except that it yields a variable, so it can be used on the left side of an assignment.

Oh what

Reading the GCC manual and everything in the Ada manual is marked as “followed.” But when it’s onto annex H…

https://gcc.gnu.org/onlinedocs/gcc-14.1.0/gnat_rm/Shifts-and-Rotates.html

[lavish dove]

bye bye partition

[proud creek]

yknow i want this sticker pack just for funsies

https://github.com/mkrl/misbrands aha

[timid quartz]

2 or 4

[spare quartz]

?

[timid quartz]

its 2 bytes mb

that one

[spare quartz]

how big is the structure in the file? if it's packed-

oh

[timid quartz]

the "Characteristics" is just a bit vector

[spare quartz]

with Size => 16;

i encourage you to add a representation clause as well

[timid quartz]

so uhh

type Characteristics is record
  --  ...
with Size => 16
with Representation => Unsigned_16
end record;

?

[spare quartz]

https://docs.adacore.com/gnat_rm-docs/html/gnat_rm/gnat_rm/representation_clauses_and_pragmas.html#record-representation-clauses

a representation clause determines how the record is structured in memory (and stream wise aswell)

assuming the file's binary structure isn't packing it in with other data it should read fine

(the "record" on the bottom is a rep. clause)

[timid quartz]

ic ic...

neat

[spare quartz]

cause otherwise the compiler could theoretically make your record weird in the interest of speed/memory

GCC generally doesn't do anything extreme, but... the added assurance is nice

also god i'm mad at GCC

tried* to compile a mingw cross for 14.2.0 again and it failed like every other attempt

all i can do is wait until it's available on apt

the shell must grow

[timid quartz]

:haha:

[timid quartz]

makefile makefile makefile

[spare quartz]

ITS NOT A MAKEFILE

[timid quartz]

type Characteristics is record
      Stripped_Relocations, Is_Executable, Stripped_Line_Nums,
        Stripped_Symbols, Aggressive_WS_Trim, Large_Addr_Aware,
        Bytes_Reversed_Lo, Is_32_Bit, No_Debug_Info, Is_Removable,
        Is_Net, Is_FS, Is_DLL, Only_Uniprocessor,
        Bytes_Reversed_Hi : Boolean;
   end record;

this

is gonna be big

[spare quartz]

OH

it's one of thoseee

[timid quartz]

y e s

any wise words

[spare quartz]

if you want to be lazy and not write a clause out for that

with Size => 16, Pack; you can throw these aspects on it

of course the compiler can change the structure to whatever it wants, but it shouldn't... hopefully

[timid quartz]

mm

compilers like to do funny things sometimes

[spare quartz]

if you're compiling with -O2/3 i don't think it'd try anything silly

maybe with -Ofast or -march but i don't think that's likely

[timid quartz]

actually

I should write out a clause

[spare quartz]

prepare for tediousness

[timid quartz]

mm nvm there's only one reserved bit

[spare quartz]

oh yeah one thing that won't affect you here but just something you should know

[timid quartz]

wait where do I throw these

[spare quartz]

you unfortunately can't overlap record fields onto each other in a rep clause (which adds a bit of difficulty when dealing with embedded stuff) but you can always just delegate to using Unsigneds and shifting

[timid quartz]

pragma Pack (R);
for R'Size use 49;

?

[spare quartz]

after end record* since they're aspects of the record type

[timid quartz]

o

[spare quartz]

a lot of pragmas are obsolescent in favor of aspects (with ...) but certain ones like Warnings are still used

scpsl server list error againnnnnn

[timid quartz]

nowhow do I read it...

from a stream...

nvm

...nvmnvm

[spare quartz]

The_Record_Type'Read (Stream, A_Record_Variable) procedure
The_Record_Type'Input (Stream); function

[timid quartz]

"error: component "Characteristics" cannot be used before end of record declaration"

[spare quartz]

show record type

[timid quartz]

   type Characteristics is record
      Stripped_Relocations, Is_Executable, Stripped_Line_Nums,
        Stripped_Symbols, Aggressive_WS_Trim, Large_Addr_Aware,
        Reserved, Bytes_Reversed_Lo, Is_32_Bit, No_Debug_Info,
        Is_Removable, Is_Net, Is_FS, Is_DLL, Only_Uniprocessor,
        Bytes_Reversed_Hi : Boolean;
   end record with Size => 16, Pack;

[spare quartz]

you probably named a field the same as your type

[timid quartz]

OH

[spare quartz]

nono

[timid quartz]

yeah I did

[spare quartz]

the- yeah

[timid quartz]

teehee

[spare quartz]

so anger

vsc rich presence shows the adb as a text file

[timid quartz]

aaand it didn't work right

h i typed all the bits in order

maybe I have to type them in reverse order...

because little endian...

wait no

[spare quartz]

is the file in LE?

[timid quartz]

if it's LE then the lsb should be the first one

maybe the file is BE

[spare quartz]

if the structure is right i'd say just use a rep clause

[timid quartz]

no PEs are LE

[spare quartz]

the more i type the more i think im gonna be kicked from this server for being afk

[timid quartz]

me no wanna type all that...

...I might have to

[spare quartz]

oh yeah btw

https://gcc.gnu.org/onlinedocs/gnat_rm/Effect-of-Bit_005fOrder-on-Byte-Ordering.html

(will not change endian)

[timid quartz]

this didn't work

[spare quartz]

actually no it will change endian huh

[spare quartz]

oh

what's the record showing in binary, and whats the actual value

[timid quartz]

how can I print the record as binary lol

[spare quartz]

either Unchecked_Conversion or just 'Image it and type it out

[timid quartz]

(STRIPPED_RELOCATIONS => FALSE,
IS_EXECUTABLE => FALSE,
STRIPPED_LINE_NUMS => TRUE,
STRIPPED_SYMBOLS => FALSE,
AGGRESSIVE_WS_TRIM => FALSE,
LARGE_ADDR_AWARE => TRUE,
RESERVED => FALSE,
BYTES_REVERSED_LO => FALSE,
IS_32_BIT => TRUE,
NO_DEBUG_INFO => FALSE,
IS_REMOVABLE => FALSE,
IS_NET => FALSE,
IS_FS => FALSE,
IS_DLL => FALSE,
ONLY_UNIPROCESSOR => FALSE,
BYTES_REVERSED_HI => FALSE)

[spare quartz]

00100100 10000000

[timid quartz]

though if I read an Unsigned_16 it shows a different value

[spare quartz]

oh?

[timid quartz]

I think

lemme check

2#1000101110#

how can I make it output all the bits

and not trim

cause it omitted like 6 zeroes

[spare quartz]

i dunno, string ops or something in integer io

[spare quartz]

but that's really weird... are you sure the record is 16 bits

what about the parent record

[timid quartz]

the parent record doesn't have a size attribute

[spare quartz]

are you reading into it or just reading it entirely

[timid quartz]

I'm manually reading the other fields in the parent record

[spare quartz]

hm

[spare quartz]

yeah that's really weird

if you can't figure out what bits its reading off, just use an unchecked_conversion

(i'd reccomend putting the unchecked conv in a custom 'Read)

[timid quartz]

this is confusing

[spare quartz]

what is

[timid quartz]

using the conversion thing

idk why it's misreading bits tho

[spare quartz]

function Convert is new Ada.Unchecked_Conversion (Source => Unsigned_32, Target => Characteristics_Flags);
...
H.Characteristics := Convert (Unsigned_32'Input (S));

[spare quartz]

do you know where in the file its reading from

[timid quartz]

Unsigned_16

yeah I do know

[spare quartz]

(you can also use this to do the reverse)

[timid quartz]

it appears to be the correct place

[spare quartz]

hmm

[timid quartz]

coff_header.adb:1:09: warning: unit "Ada.Unchecked_Conversion" is never instantiated [-gnatwu]
coff_header.adb:11:16: error: subtype mark required in this context
coff_header.adb:11:16: error: found "Unchecked_Conversion" declared at a-unccon.ads:20
coff_header.adb:11:38: error: incorrect constraint for this kind of type
coff_header.adb:21:28: error: expected type "Characteristics_Flags" defined at coff_header.ads:5
coff_header.adb:21:28: error: found type "Convert" defined at line 10

   procedure Read_Header (S : Stream_Access; H : out Header) is
      type Convert is
        new Ada.Unchecked_Conversion (Source => Unsigned_16,
                                      Target => Characteristics_Flags);
   begin

[spare quartz]

it's a generic function

you have to use function

not type

[timid quartz]

ok this time

it read the thing correctly.

[spare quartz]

neat

[timid quartz]

is executable: TRUE

ughh that sucks tho unchecked conversion bad.

[spare quartz]

this could potentially be why: https://gcc.gnu.org/onlinedocs/gnat_rm/Attribute-Object_005fSize.html#Attribute-Object_005fSize

[timid quartz]

I could force alignment to 1

[spare quartz]

but think about the poor cpu!!!!!!!!

[timid quartz]

nvm there is no such...

attribute..

[spare quartz]

it's not awful to use it when you have a reason to

the names a bit misleading though, its more like Unchecked_Copy

[timid quartz]

do I still need with Size => 16, Pack; here?

[spare quartz]

yes

otherwise copying the unsigned_16 into the record may produce weird results

[timid quartz]

[spare quartz]

(you might get a trinary boolean!!!)

[timid quartz]

computers should have used base 3...

No/Maybe/Yes

[spare quartz]

technically valid in ada.... up until you use it

[timid quartz]

I may change this and similar to use the fancier method though

with the funny clause

[timid quartz]

where is that bit type defined..

[spare quartz]

most likely an example type of Boolean with size 1

[timid quartz]

custom type then...

[spare quartz]

i would also recommend you have a bitflipper like the one i made for PNG just in the tiny tiny case the host system is BE

but its not needed cause you're on an LE computer

[timid quartz]

ill flip ur bits

[spare quartz]

❌

[timid quartz]

type Bit is mod 2 ** 1 with Size => 1;

hm

that should be ok

[spare quartz]

(you could've just said new boolean 😭)

[timid quartz]

smh

[spare quartz]

but that's fine if you're expecting to do like, ...

shifts or rotations on a bit type ...

[timid quartz]

maybe

idk

also grr where is Low_Order_First_Unsigned

[spare quartz]

que

if you're talking about a type i don't think it exists anywhere

[timid quartz]

with Bit_Order =>

[spare quartz]

just use the regular Low_Order_First as its argument

what're you using it on

[timid quartz]

oh

CONSTRAINT_ERROR

[timid quartz]

the characteristics

[spare quartz]

(i've never really known what bit_order was for in all seriousnes.........)

[spare quartz]

what of

[timid quartz]

raised CONSTRAINT_ERROR : coff_header.ads:7 invalid data
Load address: 0x1086bf000
Call stack traceback locations:
0x1086c0a57 0x1086c0bd3 0x1086c0346 0x1086e5007 0x7ff803c382cb

[spare quartz]

INVALID DATA!!!! THIS MF HAS A GOOFY BOOLEAN 🚨

also wow your stacktrace is... weird.. are you using symbolic traceback in your binder

or do you intend to use addr2line

[timid quartz]

idfk

[timid quartz]

same with mod 2 ** 1

sadge

uGHHH unchecked_conversion it is...

[spare quartz]

i might try and write an exe reader myself to see if i could try and get something working
-# in rust

but not now!!!!

[timid quartz]

IN RUST!!

you should do ada lsp first

[spare quartz]

❌❌❌❌❌

yeah sure

[spare quartz]

i would ask if i should use libadalang or try and make my own parser but like

nopppeee

[timid quartz]

ugh if only this was C

[spare quartz]

libadalang is >6 years old and they haven't even completed it

[timid quartz]

C structs map so cleanly onto memory

[spare quartz]

so do ada recrods...

[timid quartz]

Obviously not…

Can’t even map a bit vector corrrctly

[spare quartz]

i think you just need a few more aspects

[timid quartz]

like what

[spare quartz]

umm

size.. objec_size...

[timid quartz]

is this in bits or bytes

[spare quartz]

bits

[timid quartz]

no work

Size => 16,
Object_Size => 16,
Bit_Order => System.Low_Order_First,
Pack;

[spare quartz]

more aspects bro.... more aspects.......

in all seriousness i really don't know why it's not representing right

[spare quartz]

also btw

forgot to say this but when i said a custom 'Read i mean like

for A'Read use Custom_Read;

[timid quartz]

ah ic

[spare quartz]

where Custom_Read is a procedure that follows the semantics defined by 'Read

[timid quartz]

ask your little ada people on your ada mailing list

[spare quartz]

im too scared

there's probably gonna be a programmer with 90 years of experience saying my computer is all wrong

[timid quartz]

what size is boolean anyways

its probably like

2 bits

[spare quartz]

like 1 or 8

[timid quartz]

honestly it might be 8

[spare quartz]

i think the Size is 1, but the Object_Size is 8

[timid quartz]

grrrrrrrr maybe that's why it's not representing right

[spare quartz]

REMINDS ME

if i ever do a third jvm

[timid quartz]

coff_header.ads:6:61: error: Object_Size must be 8, 16, 32, or multiple of 64

[spare quartz]

i should decide wehter to -

[spare quartz]

oh

well thats interesting...

[timid quartz]

[spare quartz]

i dunno!!!!!

just keep using unchecked_conversion

it's probably the best bet

[timid quartz]

ugh unchecked_conversion is CRINGE

it's UNCHECKED

[spare quartz]

and????

your record consists of bits it's not like it'll care

[timid quartz]

typedef struct {
  unsigned int flag1 : 1;
  // ...
} characteristics_t;

i bet this would work just fine

[spare quartz]

flag1 is an int!!!

misalignment incoming!!!

[timid quartz]

yeah but : 1 dictates size

[spare quartz]

uggghhghhghghg

hold on

lemme try something

[timid quartz]

im going to bed but I will watch

[spare quartz]

0b1110001100011100

[timid quartz]

I have 2 pages of Japanese to write by Wednesday morning

[spare quartz]

bayachao reference

[timid quartz]

anyways what are you trying hmm

[spare quartz]

gonna read this into a record and see if it works

[timid quartz]

I swear I always have issues with bit vectors/fields in languages that aren’t C

[spare quartz]

this is the one thing ada should have in the bag which is why im so confusd about it

uhhh crap how do i write binary to a file in shell

[timid quartz]

Google

goog

[spare quartz]

mmgmgnfh lazy

ill just do it in ada and then siwthc

[timid quartz]

https://tenor.com/view/goog-cat-gif-11041218434900050173

[timid quartz]

Like when I was doing this in Zig I had to wrap their “packed struct” in an “extern struct” to force C ABI so that it would read the vector right

[spare quartz]

END_ERROR...???

[timid quartz]

https://tenor.com/view/the-end-is-near-simpsons-warning-gif-5187386

[spare quartz]

oh oops

i was deleting the file

https://forum.ada-lang.io/t/option-and-result-types/824

[timid quartz]

heh

Also @spare quartz new css update dropped https://youtu.be/A89FMtIkWKc

anyways I sleep, I will see your results in the morning

[spare quartz]

weeeeeirrdd

[spare quartz]

cool, when will we be wirting apps in it

okay yeah

'Read just seems to be really poor at reading records like this

WHUH

@timid quartz

pragma Ada_2022;

with Ada.Streams.Stream_IO; use Ada.Streams.Stream_IO;

with Ada.Text_IO;

with Example;

procedure Basic_Exe_Reader is

   F : File_Type;
   S : Stream_Access;

   R : Example.Example;

begin
   Open (F, In_File, "bbso");
   S := Stream (F);
   Example.Example'Read (S, R);
   Ada.Text_IO.Put_Line ("Checked Read: " & R'Image);
   Close (F);
end Basic_Exe_Reader;

with Ada.Streams;

package Example is

   type Bit is new Boolean with
     Size => 1;

   type Example is record
      A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, P : Bit := False;
   end record with
     Size => 16;

   for Example use record
      A at 0 range 0 .. 0;
      B at 0 range 1 .. 1;
      C at 0 range 2 .. 2;
      D at 0 range 3 .. 3;
      E at 0 range 4 .. 4;
      F at 0 range 5 .. 5;
      G at 0 range 6 .. 6;
      H at 0 range 7 .. 7;
      I at 1 range 0 .. 0;
      J at 1 range 1 .. 1;
      K at 1 range 2 .. 2;
      L at 1 range 3 .. 3;
      M at 1 range 4 .. 4;
      N at 1 range 5 .. 5;
      O at 1 range 6 .. 6;
      P at 1 range 7 .. 7;
   end record;

private

   procedure Read_Example
     (Stream :     not null access Ada.Streams.Root_Stream_Type'Class;
      Item   : out Example);
   for Example'Read use Read_Example;

end Example;

with Ada.Unchecked_Conversion;

package body Example is
   procedure Read_Example
     (Stream :     not null access Ada.Streams.Root_Stream_Type'Class;
      Item   : out Example)
   is
      type U16 is mod 2**16 with
        Size => 16;
      function Convert is new Ada.Unchecked_Conversion (U16, Example);
   begin
      Item := Convert (U16'Input (Stream));
   end Read_Example;
end Example;

giant message but basically here's what you should do

probably should note down 'Read is bad for records in my head somewhere

DEFAULT 'READ

[proud creek]

truly limited by our time...

[spare quartz]

i love ada types

so clean .....