#development
1 messages · Page 186 of 1
In sizeof
no you can just directly use a string constant
yeah string literals in elle literally just create a data section like this and then return a pointer to the data section
hm
so yes null bytes
pretty sure that works
this is just C then
if you need the length you use sizeof
fr
No null bytes
hm wb immutable string slices created from alloc'd strings
those don't have fixed length
thus sizeof not resolvable at compile time
runtime sizeof
?
My tasty null bytes
yeah
the distinction is made in rust since a string is always an allocated owned buffer
sizeof is a compile time operation so you can’t actually get that in elle lmao
you can use strlen tho
if it's non allocated string slice you can definitely use that but it acts like an object with length and ptr
C is kinda fake tbh
idk why sizeof isn't exclusively compile time
yes but it’s not a transpiler to C
i swear it is
your language very closely resembles C
.
lol i know
once i get better semantics i will abstract it to be more like zig
do you guys know if theres a runtime sizeof
strlen is similar but it only returns the size of data that is allocated in the specified buffer rather than the whole size of the buffer
don't make the same confusing mistake as c
which is? to make sizeof a dynamic function?
yeah
yes very unpopular
why in hell would you EVER need that
youre passing around functions by byte pointer to the start of the blob anyway
when would you ever need the size of a function
just type your function in asm
I did in the end
Wasn’t even my code it’s reversed asm
That chops itself up a lot and copies specific functions as shellcode to other places
why would you actually need this?
like what applications does it have
It’s checkra1n shellcode
It has to calculate the size of itself (being the payload) as well as specific portions of itself (for copying as additional shellcode for the next stage bootloader, plus it also copies specific instructions as binary patches for the next stage bootloader)
wonderful
opinions on a $ operator
malloc(1024 * $double) = malloc(1024 * 8)
i dont like the whole word sizeof
make it € 
💀double
implement emojis /j
there you can do this now
double *zBuffer = malloc(size * $(double));
wrapped in () so its easier to read when its a pointer type
this is literally all i did lmfao
i could turn this into sizeof pretty easily
but i like $ more
sizeof makes much more sense
$ makes me think variable declaration/retrieval
not size
thats fair
idk, if anyone has a better idea for a single character keyword for it ill change it
if not ill just go with sizeof
maybe @ or %
it cant be !, &, or *, and ^ looks wrong
already used for inline IR
sz
actually not a bad idea
but that would mean you cant declare any functions as sz anymore
you can declare it as a variable tho because thats checked before the global scope is checked for functions
maybe ill just go with @ for now
and change it later
actually
ill make inline IR $ again and then i can make # the sizeof
do sizeof please
making stuff really short makes readability harder
yes, that counts for you too, statur
s#$sz*(double)
(i will do sizeof)
now its even more like C
who wouldve guessed the ideal low level language is just.. C
yea but then i dont get distinct syntax highlighting for it when using C++ syntax highlighting
Because its not c++ 🤷♂️
no as in
sizeof is a special keyword in c and c++
so the syntax highlighter colors it differently
oh i
read what you said wrong
ok ill just do size thatll make it easier
Yay 🎉
wait i have an idea
just so it doesnt block a function from being called size
ill just make it #size(double *)
and its intuitive too
because # is usually related to preprocessor directives
#size nice tbh
it actually gets a special color in c++ syntax highlighting lmao
that wasnt on purpose
great
i implemented proper validation
are there any alternatives to headers.cynder.me
without having to download ktool and all that
thanks
@radiant idol
he's right
That's because you switch to c++

how does one show alerts in swiftui app
by using .overlay perhaps ?
like email already registered, request timedout
.alert
the default alert design is horrendous
looks like it got killed in ios18
nah i like raw dogging my shii
Button("Save") {
addItem()
}```
Raaagh peardb code
I believe ispresented is optional
ik but .alert is gonna be deprecated + the ui/ux design for it is ugly
but the ui/ux is bad yk isn’t .overlay better for my situation
All else fails you overlay a custom view in a little rectangle in the same place alert would
wym all else fails ?
i’m planning to make a small alert rectangle that pops from bottom n fades away
Fire
so .overlay valid for my situation ?
In my humble probably not correct opinion yes
i put the donut elle code into a donut shape
pub fn main(){double A =0;
double B = 0; double *a = malloc(19200);
char*b=malloc(2400);char*c=".,-~:;=!*#$@";printf(
"\x1b[2J");while 1 {memset(a,0,19200);memset(b,32,2400)
;double d = cos (A); double e = sin (A); double f =cos(B);
double g=sin(B);for double h=0;h<6.28;h+=0.07{for double i=0
;i<6.28;i+=0.02{double j=sin(i);double k=cos(i);double l=sin(h);
double m= cos(h);double n= m+ 2;double o=1 / (j * n*e
+ l*d+ 5); double p= j*n *d- l*e;int x=30+ 36*o*(k
*n*f- p* g ); int y= 12+ 12 * o * (k * n * g + p*
f);int q = x + 100 * y; int r = 8 * ((l * e -j*m
* d) * f - j * m*e- l *d - k * m *g) ;if 24 > y&&
y > 0 && x >0&& 100 > x&&o >a[q] { a[q]=o; if r>0 {
b[q] = c[r]; } else { b[q] = c[0]; } } } } printf("\x1b[H");for
long s = 0; s < 2400; s++ { if s % 100 { printf!("%c", b[s]);
} else {printf!("\n");}}A += 0.004;B +=0.002;}return 0;}
//~::;;;;;;*********################*******!!====;;;:~~~
//~~~~:::====!!!!!!!!!**********!!!====::::-----
//....,,,~~~~;;;========;;;:~~~~~---....
//....,,,,------------,,,****///
it compiles
it looks a bit less wide in my editor lmao
looks very clean in discord before its sent
wait did you make that from scratch or did you translate the c version into elle
i took the C version, turned it into elle and deobfuscated it so i could understand how it works, then i decided i wanted to put it into a donut shape so i re-minified the elle code
this is the original unminified version https://github.com/acquitelol/elle/blob/rewrite/examples/donut.elle
i basically just took this and made it as little as possible and then donutified it
i need to fix arithmetic tokenizing tomorrow for the infix operator
has there ever been a public ios userspace pac bypass
@fading shell
wipe the jailbreak and the device.
I already installed glance and it seems to work fine, but when installing tweak by tweak I found respringloop WITH THESE TWEAKS:
facebooknoads
iclaner pro
jade
also would it be behind SPTM or is that just a baseless rumor
because if it is ive wasted like 6 hours of my time 💀
you know you don’t need to fully reinstall the jb when you have a respring loop
yep
you can just force restart and then launch the jailbreak with tweak injection disabled
ok good
but I'm a bit extreme

I wanted to give glance a chance, but it only brings me problems
The only way for it to work well with all my tweaks is with oldabi
?? wtf
I wanna remake the icon tint thing
Well the tint shouldn’t be that hard but the dark mode conversion might be
the first person to do so and make it a tweak will make a lot of money
I don’t wanna race
I’ll probably work on it but if a paid version comes out before mine I’ll just publish mine for free, otherwise it’ll probably be free anyways
Oooo extending functionality with something like this or more would be nice as well https://fxtwitter.com/Gavmn/status/1800982973392494968
Would love to see Apple take dark mode icons one step further and sample the primary icon color to use as a subtle tint on the dark background.
Yes
lol then my 8 hour work of symbolicating IOUSBDeviceFamily was useless
I don’t understand why you went for IOUSBDeviceFamily for a PAC bypass
IOUSBDeviceFamily is locked behind PAC instructions
Every kext is…
I noticed it while looking at the disasm
The whole kernel has PAC throughout
i was not specifically looking for a PAC bypass, i looked at it because i was wondering if i could get a chain from there to kernel
the thing i wanted to do probably wont work
no
before you were able to replace PAC keys with KRW
But that's behind SPTM now
But that's not relevant, you can bypass userland PAC by finding a bug in it's implemetation
Oh hm interesting
it looks good on github LMAO
Donut 🍩
looks nice
Nice obfuscation
?
?
?
done by hand because idk how to get software to do it while respecting the syntax
@granite frigate @grim sparrow I miss making apps for jailbreakers so much, like, whatever u make for jailbroken iphones, SOMEONE will have a use case for it no matter how niche. but for normal apps u need to get like 10x out of the box with thinking and it hurts my head
lol
idk i dont have any app ideas
like ever

happens to the best of us buddy
amy i love this image so much
Its a classic
same APIs and tools that we use. buddy how about you dump UIView ^& UISearchbar and get back to me on that
real
@grim sparrow I wonder how much CPU energy is wasted checking for if String(data: data, encoding: .utf8) returns nil because it turns it that it literally just doesn't
yeah but in most cases it just doesn't and I think that's a lot of wasted cpu energy tbh
like i'm 100% serious when I say that most String(data: data, encoding: .utf8) should just be force unwrapped
@grim sparrow OHHH I CONFUSED IT WITH string.data i'm so stupid omfg
💀
yeah string.data is the one that can never return nil if it's utf8
Yeah…
do u fw actual profiling or do u just use the xcode perf tab btw
Profiling
I like it until I need to do it to a process I don't own :(
i tried to profile discord one time and oughhh
Yeah it’s only good for ur own app
Chat how do I build for iOS 11
tried xcode 15 and setting deployment target to 11.0, fails when I try building
tried xcode 13, crashes when I try opening or creating a project
Use theos 
you get rid of scene delegate stuff?
The debugger section in the sidebar should show you where it crashed
No Xcode itself crashes

no
Idk
i used xcode 13 for ios 9+
Wtf.
what macos you on?
try xcode 14
I use xcode 16 with 18 sdk for 9
trol
try xcode 11.7
try xcode 3.2
Try xcode 5
A tweak that replicates those motion sickness dots from the new update would be cool
its so nice lol
pain
i hate intents
so much
I have NO CLUE what changed between beta 6 and 7
but SOMETHING BROKE IT
lemme look at commits
is it just
is it literally just i didnt recompile the new binary and so the library thinks it isnt working
and then the entire thing dies
Hold on
TIL there are intents on iOS
just updated to sequoia wonder if this is doable 😮
oh wtf
thats jj lol
i literally forgot about the point of arm64e 😭
Yeah and they SUCK
#!/usr/bin/env zsh will work on both rootless and rootful right
It will
@oak meadow https://www.youtube.com/watch?v=ZR8KivbEzLI they basically took your idea and 10x'ed it lmfao
Apple Intelligence is the most important announcement from Apple in nearly a decade. A lot of what was announced is not new — but no matter how jaded you are to all of this, I’m asking you to take a step back and hear me out because the intention and purpose behind Apple Intelligence is new- and it’s a game changer. Quite literally, after this, ...
should iirc libiosexec handles it ?
Yea it does work
It gets rid of the error bit though
posix_spawn(&pid, "/bin/zsh", NULL, NULL, (char *const *)args, NULL);```
This would work to spawn uialert right
Or I could just yoink the code from uikittools 
i do not know what uikittools is doing
it sends a message to SB using a private CF API
someone help me dream up syntax for lambda functions in elle
im currently thinking ^(int a, int b) -> int: a + b
Get rid of that ^ I’m begging
I HATE that thing so much
i dont really understand what the code is doing is what i mean
NO
no
THEY SUCK
no we dont
^ should only be used for xor 🙏
pls tell me how the parser can differentiate between (int) type casting and (int a): a * 2 lambda functions
without looking super far ahead
making a pr to fix this one typo
DO IT HOW JS DORS IT
ANYTYIJG BUT OBJC BLOCK SYNTAX
JS DOESNT HAVE TYPE CASTING LIKE THAT
ITS BEEN 3 YEARS AND I STILL DONT KNOW THAT STUPID SYNTAX
EVERY TIME I HAVE TO GO ON GOSHDARNBLOCKSYNTAX TO REMEMBER
what about this
errx(1, _("JSON formating failed: %s"), error.localizedDescription.UTF8String);
ITS FORMATTING
TWO T
but instead of { return a - 1 } its just a - 1
(i found this random image online when searching c lambda function lmfao)
also yes im aware its just a C++ lambda
server back s0n?
W
and all my apple tv
poweredge and xserve raid No though because my dad said no 240v
(we'd need to upgrade to 250a service)
@naive sedge so what did you change in a backup to get iPhone mirroring working 
so im powering this shit off a 120v 20a circuit
I may explode the pushmatic breaker from the 60s
💀
Big cable
pub fn main() {
int res = a(1)(2);
}
fn add(int x) {
return [](int y) -> x + y;
}
fun i guess
Here we go!
"u VS the person she says not to worry about"
seconded
@radiant idol hear me out
pub fn main() {
function *someLambda = fn(int x) {
// full block with return available
}
int res = add(1)(2); // 3
}
fn add(int x) {
return fn(int y) -> x + y; // single statement only
}
im concerned as to how im gonna put variables from the declaration scope into the scope of the lambda
considering under the hood itll just make a new standalone function with tmp_counter name based on the parent function
but we'll see
This s scares me
lmfao
perhaps something like this to keep it more consistent
pub fn main() {
function *someLambda = fn(int x) -> {
// full block with return available
}
int res = add(1)(2); // 3
}
fn add(int x) {
return fn(int y): int -> x + y; // single statement only
}
ok yea i think this is what im gonna go with
inside of statement parsing the fn keyword is literally never used so this is perfect
i think i know how to do this actually
- create a standalone function with the existing arguments of the parent function first then the new arguments after
- store the lambda in some hashmap in the compiler ctx
- when the lambda is called, check if its in the hashmap and retrieve it
- insert all of the parent function's arguments before the new arguments that the function can actually take
- call the temp function like normal
so internally
fn add(int x) {
return fn(int y): int -> x + y; // single statement only
}
will make
fn add.1(int x, int y) {
return x + y;
}
fn add(int x) {
return add.1; // store this somewhere so that you can insert `x` when it comes to running the lambda
}
that outlet is crying lmao 😭
THE CABLE MANAGEMENT
It should be grateful I didnt plug in my poweredge 2850
(pentium 4 xeon)
THEREs everyone
Sarafine got a memory wipe??
Oh yeah I threw big sur at it
Let me
make a monterey USB
and reinstall macos
do it do it do it
I accidentally installed windows on my Mac
how do you accidentally do that
Took the hard drive out so I didn't have to wait like 3 days for it to transfer files before I could use it and apparently the drive I stuck in had windows on it
I thought it had arch Linux but I guess that's a different drive I was thinking of
guys is this overengineered enough
ifeq (run,$(firstword $(MAKECMDGOALS)))
RUN_ARGS := $(wordlist 2,$(words $(MAKECMDGOALS)),$(MAKECMDGOALS))
$(eval $(RUN_ARGS):;@:)
endif
DIST_PATH = dist/$(RUN_ARGS)
IR_PATH = $(DIST_PATH)/target.ssa
TMP_PATH = $(DIST_PATH)/tmp
ASM_PATH = $(DIST_PATH)/out.s
EXEC_PATH = $(DIST_PATH)/$(RUN_ARGS)
run: $(EXEC_PATH)
$(EXEC_PATH): $(ASM_PATH)
cc -o $@ $<
$(ASM_PATH): $(TMP_PATH)/out.new.s
mv $< $@
rm -rf $(TMP_PATH)
$(TMP_PATH)/out.new.s: $(TMP_PATH)/out.tmp.s
sed -E 's/Lfp([0-9]+):/_Lfp\1:/g' $< > $@
$(TMP_PATH)/out.tmp.s: $(IR_PATH)
mkdir -p $(TMP_PATH)
qbe -o $@ $<
$(IR_PATH): examples/$(RUN_ARGS).elle dist/compiler
rm -rf $(DIST_PATH)
mkdir -p $(DIST_PATH)
dist/compiler $< $@ > /dev/null
.PHONY: run
run:
clear
@$(EXEC_PATH)
compile:
mkdir dist
rustc -o dist/compiler src/main.rs
No
what more can i add
cmake
right now building is as simple as make run <name>
cmake is the father of all build system overengineering
you don't need dependencies to justify using cargo 
After all it's just another build system
it'd probably be nicer bevause built in functionality like cargo run
and you can make whatever build scripts you want with build.rs
oh god i forgot i need to write my whole build system in rust
ok i can do this
build.rs my beloved
cargo time
i dont know what approach to take for lambdas 😭
if i capture all of the variables defined from the parent function they wont be in scope when calling the lambda because %x.23 is a temporary or whatever
You should make your build system in objective c
ill stick with a makefile for now because its really simple
as simple as this can get..
its as simple as running gmake run compile to get a compiler executable and then gmake run <file> to compile and run ./examples/<file>.elle
i gave up making lambdas
but i did make fake constants
look
const int MAX_INT = 2147483647;
const int MIN_INT = -1 * MAX_INT;
function w $MAX_INT() {
@start
%r.v0.1 =w copy 2147483647
ret %r.v0.1
}
function w $MIN_INT() {
@start
%MAX_INT.2 =w call $MAX_INT()
%tmp.3 =w copy -1
%tmp.4 =w copy %MAX_INT.2
%tmp.5 =w mul %tmp.3, %tmp.4
%r.v5.6 =w copy %tmp.5
ret %r.v5.6
}
at compile time it turns the literal expressions into a funcall
i could also make it into a data section and then make it a pointer derefence but that wouldnt work very well because you cant put ssa form into data constants
thats pretty cool i think
in rust what you have to do if the closure outlives the function is you have to move all the variables used into the closure instead of just taking refefences
so probably that's how you have to do it
closures are such a clusterfuck in non-reference counted languages
i gave up and decided elle will just not have lambda functions
i made array literals tho!!
pub fn main() {
char *test[] = {"abc", "meow", "test"};
for (int i = 0; i < 3; i++) {
printf!("test[%d] = %s\n", i, test[i]);
}
return 0;
}
abc meow test
meow
why are you using char *
ew
also make size_t arrlen(...) or something
ill make it a #size thing
cursed
should #size return the size in bytes or the size in bytes divided by the size of the type
so if you do
long test[] = {1, 2, -39};
should it return 3 or 24
always size of bytes
if it returns size of bytes then instead of
const long MAX_SIGNED_LONG = 9_223_372_036_854_775_807;
const long MIN_SIGNED_LONG = -MAX_SIGNED_LONG - 1;
pub fn main() {
long test[] = {MAX_SIGNED_LONG, MIN_SIGNED_LONG, -39};
for (int i = 0; i < #size(test); i++) {
printf!("test[%d] = %ld\n", i, test[i]);
}
return 0;
}
you would need to do
const long MAX_SIGNED_LONG = 9_223_372_036_854_775_807;
const long MIN_SIGNED_LONG = -MAX_SIGNED_LONG - 1;
pub fn main() {
long test[] = {MAX_SIGNED_LONG, MIN_SIGNED_LONG, -39};
for (int i = 0; i < #size(test) / #size(long); i++) {
printf!("test[%d] = %ld\n", i, test[i]);
}
return 0;
}
``` to get the behavior youre aiming for
is it the same in C?
ill make it size in bytes i suppose
literally just a difference between
Instruction::Copy(Value::Const(
ty.clone(),
data_section.unwrap().size as i64
/ buffer.unwrap().size() as i64,
)),
``` and
```rs
Instruction::Copy(Value::Const(
ty.clone(),
data_section.unwrap().size as i64,
)),
``` lmao
is that asm in rust-ic syntax
no its IR in rust-ic syntax
close enough
should it be a # directive or just a function
i dont like how in c sizeof looks like a function
because its not
its a macro
and yet nothing is special about it
at least user-made macros are all caps usually
ok @radiant idol ```c
const long MAX_SIGNED_LONG = 9_223_372_036_854_775_807;
const long MIN_SIGNED_LONG = -MAX_SIGNED_LONG - 1;
pub fn main() {
long test[] = {MAX_SIGNED_LONG, MIN_SIGNED_LONG, -39};
for (int i = 0; i < #arrlen(test); i++) {
printf!("test[%d] = %ld\n", i, test[i]);
}
return 0;
}
someone remind me to document all of this behavior in the morning before i push
because its like 1am and i cba to do it now
nice
i will forget to remind you if that helps
@placid kraken HAPPY BIRTHDAY!!!!!!!!!!!!!!!!!!!!!
happy birthday
Happy birthday
Bro is younger than me and is making a whole ass programming language
I’m cooked
same
@placid kraken congradulations on being born
import Foundation
class ErrorHandler: ObservableObject {
static let shared = ErrorHandler()
@Published var latestError: Error?
private init() {}
func storeError(_ error: Error) {
latestError = error
}
func clearError() {
latestError = nil
}
}
``````swift
.overlay(
(ErrorHandler.shared.latestError != nil) ? AlertView() : nil
)``` and the view in overlay doesnt update even if the value of latesterror changes
help ?
happy birthday!!!
that’s crazy
i thought you were like 20
happy birthday though
Would it be possible to backport RCS support with a tweak?
Well rcs isn’t even live in ios 18 yet, so you’re asking way to early
True true
It’s decent
happy birthday, does this make me the youngest dev now?
happy bday!!!
there’s nothing for it rn but it’s the lowest version that doesn’t have anything so likely is the next version to get something
yeah i guess so lmao
WHY DOES EVERYONE THINK THAT OMG
you have too much knowledge for your own age
and are more mature than others your age
(low bar
)
so
.
Happy birthday!
happy birthday @placid kraken
tyyy
Happy birthday!
has someone shared this already https://youtu.be/lQjaaAqgxp4
Skip brings Swift app development to Android. It is a tool that enables you to build genuinely native apps for both iPhone and Android from a single iOS codebase. Find out more at https://skip.tools
pretty cool
Happy birthday @placid kraken
Happy birthday @placid kraken
TrollStore 🥱
no it's alfiestore
Nah @hasty ruin is like 12
how do i hook/edit a dyld function in lldb
.
most likely the wrong tool for this but
lldb is all i know
(macOS)
i needed to wait for it to be loaded
did you get it to work in the end 😭
anyone know a way to bypass tethering throttle on linux
ik about pairvpn for macos but linux does not have
ive tried mangling ttl
ive seen some things about hosting ad hoc network from pc and using ssh proxy but i cant download necessary packages without internet lol
and that'll cut my speed in half
if you mean to use another debugger, i doubt something like gdb even works on macos lmao
Happy bday!
it broke many years ago according to saagarjha
also on the newest macOS version it can't work
they added in an entitlement you need for thread_* functions that gdb relies on iirc
do we not know what that entitlement is
or is there another reason we can't just grant it
@lusty jacinth

I didn’t tho
It wouldn’t even work unless I have the role
i just tested it, it wouldve
o
yea
just make it listen to a notification
forgot which one tho
Nice
Yeah, the solution’s here #development message
I actually went up that chain once
I regret wasting that time
we all have

What project
Dw about it
Worked for me. Thank you iCraze.
He keeps being lazy
OK ICRAZE
icraze telling you that he can vouch for himself
(he’s been visiting the mental institution)
🐓
I did just now
can't believe it spans 4 years
Yeah fr
actually i think its here #development message
Happy birthday
:3
i'm trying to use libkrw in an app, and i've added libkrw.tbd to my xcode project alongside everything else.
changed the tbd for rootless to use
install-name: '/var/jb/usr/lib/libkrw.0.dylib'
when i run, app crashes and crash log says
Exception Type: EXC_CRASH (SIGABRT)
Exception Codes: 0x0000000000000000, 0x0000000000000000
Termination Reason: DYLD 1 Library missing
Library not loaded: /var/jb/*/libkrw.0.dylib
any help would be greatly appriciated, don't want to be using kfd for overwriting when i have a full jailbreak
that means you don't have libkrw installed from procursus
?
file is there too, checked with filza
you should be linking against the tbd in the dev package
have to install windows for the single reason that pairvpn does not have a linux version
you chose the one vpn without linux support
hm what's a good way to hook a dyld swift function globally on macOS
it is the only program capable of bypassing my hotspot throttle
i just tried using it on my mac and using an ethernet cable from my macbook to my pc and it worked but i barely get 200 kb/s
pain
@radiant idol
But how much skibidi did they have combined
Wine
i still could not watch this
ok but who said configuring network interfaces works through wine?
(yes a tunnel is an interface)
Happy birthday @tight carbon
Thanks 🤩
Ah hell I'm an idiot
I added an ifeq for a makefile variable that has a default and was wondering why the value never changed
Hbd @tight carbon
vpn dont work in wine
of course it didnt the vpn isnt ip68 water resistant
yo do yall know how to allow lldb to insert and make breakpoints into any process
getting error: 9 sending the breakpoint request errors
how are you attaching to the process?
lldb's process attach --name "name" --waitfor
it is a normal non-debug program that i cannot launch myself
Could be wrong but iirc it has to have get-task-allow entitlement. Does the process have it
i.e is it signed with a dev cert
Compilation ?
like without the -g flag on gcc nor break to debugger on app start
or whatever else
I was gonna say you have to give the debugger some entitlement but I think that was gdb not lldb
since lldb is built in
is there any plausable way to do this without get-task-allow? would like to not modify the program if at all possible
Still it needs to have it though, xcode/mac just handles it automatically. Hence why you can't debug appstore apps etc
Yeah
This is an ios app and you are debuggin with a mac right ? I don't have a mac so idk much but i don't think its possible
oh, no this is a mac app running on macOS
can't launch it? launchconstraints or just missing perms?
additional args required when launching I don't have
i mean frida uses tfp
though that might require SIP being off
ofc it being an instrumentation tool you can't expect to set hw breakpoints. there's a memory access monitor function i believe
@wooden yarrow howd you get greymatter to show up on macos beta
bro I am getting swarmed
😭
most likely via some debugging magic
getting trolled through all platforms
didnt know
mfs even friended on discord and bluesky
oop
lmao
naw it's fine
sorry to hear that
nah i was asking cus someone showed the siri UI on 18 beta
and hes gatekeeping it
what Siri ui
@mattwagner Don’t want Apple to patch the bug that made this possible sorry
oh, yeah even the mf who has access to that dmed me
lmao
(I am not replying to any of them)
"Don't want apple to patch the bug" is crazy
jjtech in hack different got iphone mirroring working on sequoia
using lldb
what is greymatter
apple intelligence
i wonder if apple can see who has waitlisted and if he is the only one lol
Thx
С днем рождения!
a swiftui developer figure out how to activate it 🤣 it’s probably a plist he modified with a backup
Plx what have u gotten yourself into
poor plx
@wooden yarrow Helloo! Sorry for bothering, I was wondering if it Is possible that you could share your findings about iPhone Mirroring, Greymatter & Siri (at least a hint) i'll only tell a friend (we are both part of the Jailbreak server) and we wont post about it bc we aren't doing it for clout/attention we just wanted to try Apple's cool new stuff early because it is hard to wait till fall. It's alright if you don't. Thanks anyway! Have a nice day!
@wooden yarrow Helloo! Sorry for bothering, I was wondering if it Is possible that you could share your findings about iPhone Mirroring, Greymatter & Siri (at least a hint) i'll only tell a friend (we are both part of the Jailbreak server) and we wont post about it bc we aren't doing it for clout/attention we just wanted to try Apple's cool new stuff early because it is hard to wait till fall. It's alright if you don't. Thanks anyway! Have a nice day!
@wooden yarrow Helloo! Sorry for bothering, I was wondering if it Is possible that you could share your findings about iPhone Mirroring, Greymatter & Siri (at least a hint) i'll only tell a friend (we are both part of the Jailbreak server) and we wont post about it bc we aren't doing it for clout/attention we just wanted to try Apple's cool new stuff early because it is hard to wait till fall. It's alright if you don't. Thanks anyway! Have a nice day!
@wooden yarrow Helloo! Sorry for bothering, I was wondering if it Is possible that you could share your findings about iPhone Mirroring, Greymatter & Siri (at least a hint) i'll only tell a friend (we are both part of the Jailbreak server) and we wont post about it bc we aren't doing it for clout/attention we just wanted to try Apple's cool new stuff early because it is hard to wait till fall. It's alright if you don't. Thanks anyway! Have a nice day!
@wooden yarrow Helloo! Sorry for bothering, I was wondering if it Is possible that you could share your findings about iPhone Mirroring, Greymatter & Siri (at least a hint) i'll only tell a friend (we are both part of the Jailbreak server) and we wont post about it bc we aren't doing it for clout/attention we just wanted to try Apple's cool new stuff early because it is hard to wait till fall. It's alright if you don't. Thanks anyway! Have a nice day!
(this was in response to the og deleted message)
@wooden yarrow Helloo! Sorry for bothering, I was wondering if it Is possible that you could share your findings about iPhone Mirroring, Greymatter & Siri (at least a hint) i'll only tell a friend (we are both part of the Jailbreak server) and we wont post about it bc we aren't doing it for clout/attention we just wanted to try Apple's cool new stuff early because it is hard to wait till fall. It's alright if you don't. Thanks anyway! Have a nice day!
@wooden yarrow Helloo! Sorry for bothering, I was wondering if it Is possible that you could share your findings about iPhone Mirroring, Greymatter & Siri (at least a hint) i'll only tell a friend (we are both part of the Jailbreak server) and we wont post about it bc we aren't doing it for clout/attention we just wanted to try Apple's cool new stuff early because it is hard to wait till fall. It's alright if you don't. Thanks anyway! Have a nice day!
Спасибо
@wooden yarrow Helloo! Sorry for bothering, I was wondering if it Is possible that you could share your findings about iPhone Mirroring, Greymatter & Siri (at least a hint) i'll only tell a friend (we are both part of the Jailbreak server) and we wont post about it bc we aren't doing it for clout/attention we just wanted to try Apple's cool new stuff early because it is hard to wait till fall. It's alright if you don't. Thanks anyway! Have a nice day!
Does anybody know if CydiaSubstrate is also widely used on android ? I saw it had support for it from the website but idk if its mainstream like it is on ios
what
yes
substrate is for code injection basically
but obviously the more popular implementationwas xposed
@wooden yarrow Helloo! Sorry for bothering, I was wondering if it Is possible that you could share your findings about iPhone Mirroring, Greymatter & Siri (at least a hint) i'll only tell a friend (we are both part of the Jailbreak server) and we wont post about it bc we aren't doing it for clout/attention we just wanted to try Apple's cool new stuff early because it is hard to wait till fall. It's alright if you don't. Thanks anyway! Have a nice day!
@wooden yarrow Helloo! Sorry for bothering, I was wondering if it Is possible that you could share your findings about iPhone Mirroring, Greymatter & Siri (at least a hint) i'll only tell a friend (we are both part of the Jailbreak server) and we wont post about it bc we aren't doing it for clout/attention we just wanted to try Apple's cool new stuff early because it is hard to wait till fall. It's alright if you don't. Thanks anyway! Have a nice day!
@wooden yarrow Helloo! Sorry for bothering, I was wondering if it Is possible that you could share your findings about iPhone Mirroring, Greymatter & Siri (at least a hint) i'll only tell a friend (we are both part of the Jailbreak server) and we wont post about it bc we aren't doing it for clout/attention we just wanted to try Apple's cool new stuff early because it is hard to wait till fall. It's alright if you don't. Thanks anyway! Have a nice day!
I just installed theos on my mac and it's not working ```
bash: /Users/shorty/theos/toolchain/Xcode.xctoolchain/usr/bin/clang: No such file or directory
Making all for tweak Eliza…
bash: /Users/shorty/theos/toolchain/Xcode.xctoolchain/usr/bin/clang: No such file or directory
bash: /Users/shorty/theos/toolchain/Xcode.xctoolchain/usr/bin/clang: No such file or directory
bash: /Users/shorty/theos/toolchain/Xcode.xctoolchain/usr/bin/clang: No such file or directory
bash: /Users/shorty/theos/toolchain/Xcode.xctoolchain/usr/bin/clang: No such file or directory
bash: /Users/shorty/theos/toolchain/Xcode.xctoolchain/usr/bin/clang: No such file or directory
bash: /Users/shorty/theos/toolchain/Xcode.xctoolchain/usr/bin/clang: No such file or directory
bash: /Users/shorty/theos/toolchain/Xcode.xctoolchain/usr/bin/clang: No such file or directory
bash: /Users/shorty/theos/toolchain/Xcode.xctoolchain/usr/bin/clang: No such file or directory
==> Preprocessing Tweak.x…
==> Preprocessing Tweak.x…
==> Compiling Tweak.x (arm64)…
bash: /Users/shorty/theos/toolchain/Xcode.xctoolchain/usr/bin/clang: No such file or directory
make[3]: *** [/Users/shorty/theos/makefiles/instance/rules.mk:400: /Users/shorty/projects/Eliza/.theos/obj/debug/arm64/Tweak.x.604acd60.o] Error 127
rm /Users/shorty/projects/Eliza/.theos/obj/debug/arm64/Tweak.x.m
make[2]: *** [/Users/shorty/theos/makefiles/instance/library.mk:52: /Users/shorty/projects/Eliza/.theos/obj/debug/arm64/Eliza.dylib] Error 2
make[2]: *** Waiting for unfinished jobs....
Check this maybe? (I do NOT know what I'm talking about)
The toolchain directory is empty but idk how to populate it
It doesn't appear to be touched at all in the instller
And the guide doesn't say anything
is project orion or does the makefile contain PREFIX var?
thank god I didn't enable vibration while sleeping
my phone would have turned into a actual vibrator
lmao
oh the Makefile does contain a prefix varible
thats why
Do I just remove it or does it serve a purpose?
I don't think so
well its my tweak and I never knew what it did and it's in my inital commit
idk maybe I stole someone else's makefile and didn't realize it
or just forgor
anyways it looks like something is screwy in my SDK
theres no other explanation
just download the fresh sdk
that doesn't remotely make any sense
me or the error?
both
this uhh still doesn't allow me to do anything
send the makefile at this point
https://github.com/WilsontheWolf/eliza/blob/master/Makefile this with line 3 removed
thats fine
add SHELL+=" +X" to the front of make command and messages=yes to the end of make command
Which toolchain did you select when installing Theos
macos so xcode is automatic?
The fix in the Theos server for the issue is:
rm -rf $THEOS/toolchain/*
curl -sL https://github.com/kabiroberai/swift-toolchain-linux/releases/download/v2.3.0/swift-5.8-ubuntu20.04.tar.xz | tar -xJvf - -C $THEOS/toolchain/
this macos tho?
yeah that's what's confusing me
you can't use that toolchain on macos
because I have seen this issue occur multiple times in the Theos server and they pointed to this as the fix, but obviously that won't work on macOS
shoud i join the theos server and ask there?
yeah
thats actually new in iOS 15 btw
What is?
for night
@radiant idol
nsstring methods getting NS_FORMAT_ARGUMENT
ah
that's really weird...
- I never use theos so Im magnatudes behind on updates
- I'm using xcode 14.1
- Im using 16.2 sdk not 16.5
just dropping this here for future reference
april 2023 
I'm on 16da692ea93ecc80ecd7d7e904a62e378a0d58be
❯ git rev-list HEAD --count
2041
Dec 30, 2023
turns out the SDK was too new for my XCode
*Xcode
smh if you're gonna target 16.5 then why not have xcode 14.3.1+
You’ve probably seen posts sharing about the new Siri UI in iOS 18 on X, Threads, etc., but iOS 18 Developer Beta 1 currently hasn’t added any new features related to Siri. Good news! You can experience the new Siri user interface early with a few tips below. ⚠ To clarify, this only activates the new Siri user interface; features related to...
utterly tragic
confused
screenshot from the article
OH
Catalina
i love userdefaults so much i wonder when they'll make it so that you can't modify it on macOS lol
i could see them restricting the ability to modify system domains
I doubt it, I just see them not shipping things like this behind a .plist
they've done it since the beginning, i dont expect it to change anytime soon
They fr just copied it directly from Aaron’s tweet
given this is on macOS there's always gotta be some bootarg, some entitlement that lets you have your way
you can even still modify rootfs with authenticated-root disabling
being granted arbitrary entitlements probably feels soooo good as a process
???
what were you thinking while typing this
must feel so good as a phone to get unlocked with face id
Not when it's my face
oh
@christian left ?
i mean think about it
someone just gave you the powers of god
Can confirm
Trollstore has a god complex confirmed
Wait till he finds about pid 0
can confirm
hey, anyone has experience with appium development to control phones remotely via Py scripts for automation purposes?
vtables?
do u have virtual functions or sumn
no lmao
wyd
there are ways to make vtables in pure C
which means i can also implement a way to do it in elle
idk
elle+
i love const int NULL = 0;
const int NULL = 0;
const int BUF_SIZE = 256;
fn input(char *message) -> char * {
long stdin = fdopen(0, "r");
char *buf = malloc(BUF_SIZE * #size(char));
printf(message);
fgets(buf, BUF_SIZE, stdin);
buf[strlen(buf) - 1] = '\0';
char *result = malloc(strlen(buf) + 1);
strcpy(result, buf);
return result;
}
fn random(int lower, int upper) {
return rand() % (upper - lower + 1) + lower;
}
pub fn main() {
srand(time(NULL));
int lower = atoi(input("Enter the lower bound: "));
int upper = atoi(input("Enter the upper bound: "));
printf!("Rolled %d\n", random(lower, upper));
return 0;
}
elle--
dkkd
Free your memory 
i love memory leaks
that does remind me that the input function i have is actually flawed
its the old version that doesnt free the memory allocated in the buffer
malloc’ing and never free’ing

if you never malloc, you never have to free 
its not that big of a deal
because its not an infinite loop
If you have 128gb of memory, you'll never have to free
the OS will claim all of the allocated memory when you exit the program anyway
im aware
Yes ik but best practice
only time you dont free is when you return said allocated object frfr
omg i actually forgot that was a thing
me when
because i was trying to
ok well here you go nightwind i think that is no longer leaking memory
const int NULL = 0;
const int BUF_SIZE = 256;
fn input(char *message) -> char * {
long stdin = fdopen(0, "r");
char *buf = malloc(BUF_SIZE * #size(char));
defer free(buf);
printf(message);
fgets(buf, BUF_SIZE, stdin);
buf[strlen(buf) - 1] = '\0';
char *result = malloc(strlen(buf) + 1);
strcpy(result, buf);
return result;
}
fn random(int lower, int upper) {
return rand() % (upper - lower + 1) + lower;
}
pub fn main() {
srand(time(NULL));
int lower = atoi(input("Enter the lower bound: "));
int upper = atoi(input("Enter the upper bound: "));
printf!("Rolled %d\n", random(lower, upper));
return 0;
}
Good
Technically you should free the two input calls at the end as well but I’ll let that slide
how tf does C have such a robust system for sizeof
One time for a computer science project, I had a thing leaking memory and I didn't know why and it was making our ci fail, which would have reduced our grade so I just deleted the test that lead to that code being run and then the tests were green
like you can pass practically anything to it and itll report a valid size
horror
@radiant idol @radiant idol do you know if C pointers are fat
or if they hold any other metadata other than size
i do not know
The crazy part is we got our mark back and was like 70% - doesn't compile, I emailed my professor saying it compiled and then we got 100%
because i know slices in rust are just fat pointers
It definitely wasn't a 100% project
lmao
im hoping to use elle for my NEA
maybe itll work maybe it wont
considering i already started the project
"import solution"
from random import randint
lower = int(input("Enter lower bound: "))
upper = int(input("Enter upper bound: "))
print(f"Rolled {randint(lower, upper)}")
Ew
Tbh you should add implicit type casting
it already does for the suite of numbers
Do it for everything
as in it automatically casts ints to doubles and whatever
but if you pass the result from input to rand itll segfault or just fail to compile because its expecting int but getting char *
even if you make it a long as the type taken into random itll be unintended behavior because itll just treat the char * as a raw number and do arithmetic with it as if it was just a long
so theres no sweet solution here because you dont pass around a whole string you pass a pointer to the start of it
Tbh reading different types from stdin natively could be handy
i could make like input_to_int quite easily
atoi doesnt validate it just returns 0 if the string cant be accurately converted to a number
ok i did it like this
fn input_to_int(char *message) -> int {
char *res = input(message);
defer free(res);
if (strspn(res, "0123456789") != strlen(res)) {
long stderr = fdopen(2, "w");
fprintf!2(stderr, "Invalid input: '%s'. Falling back to 0\n", res);
return 0;
}
// Ensure the free call is done after we get the integer
int converted = atoi(res);
return converted;
}
i forgot strspn exists
i had to look at the c string.h defs because it has so many random utility functions
and now you dont need to do this anymore because it handles that automatically
and they're all slow
guys i wanna make a way to make interfaces so that i can implement c functions in elle to get information like the variadic index and whatnot
but i dont know what keyword to use
interface sounds too long
true
@radiant idol interface, extern, declare, or proto
which keyword sounds the best
external
sure
but that doesnt really fit when everything else has such short names
like pub and fn
pub external fn printf(char *formatter, ...);
You could do extern. It can be confused with extern in c though
external public func printf(const char *formatter, ...);
oh btw nightwind
you know that idea i had where you can have any variation of
function
like funct
yes
LOOK
fuct
i realized
i did it lmao
const int NULL = 0;
const int BUF_SIZE = 256;
external fn printf(char *formatter, ...);
external fn fprintf(long fd, char *formatter, ...);
fn input(char *message) -> char * {
long stdin = fdopen(0, "r");
char *buf = malloc(BUF_SIZE * #size(char));
defer free(buf);
printf(message);
fgets(buf, BUF_SIZE, stdin);
buf[strlen(buf) - 1] = '\0';
char *result = malloc(strlen(buf) + 1);
strcpy(result, buf);
return result;
}
fn input_as_int(char *message) -> int {
char *res = input(message);
defer free(res);
if (strspn(res, "0123456789") != strlen(res)) {
long stderr = fdopen(2, "w");
fprintf(stderr, "Invalid input: '%s'. Falling back to 0\n", res);
return 0;
}
// Ensure the free call is done after we get the integer
int converted = atoi(res);
return converted;
}
fn random(int lower, int upper) {
return rand() % (upper - lower + 1) + lower;
}
pub fn main() {
srand(time(NULL));
int lower = input_as_int("Enter the lower bound: ");
int upper = input_as_int("Enter the upper bound: ");
printf!("Rolled %d\n", random(lower, upper));
return 0;
}
that was surprisingly a lot easier than i expected
now you dont need to put the weird func!(a, b) syntax @radiant idol
confusion
ill keep func.(a, b) tho for argument length
now that i declared the interface for functions like printf i can declare what index the variadic arguments start
so the user no longer needs to include it at the call site
ah nice
the compiler just inserts it automatically
Best thing about witnessing a language getting developed is you are automatically learning it 
lmfao real
HES UNBANNED
hey
okay i haven't done development since ios 14
how the hell do i port tweaks to rootless
Super easy
Build with THEOS_PACKAGE_SCHEME=rootless
(Need to clean the build directory when switching between rootful/rootless builds)
And wrap any paths with the macros provided by theos within libroot.h
(I don’t remember the exact names tbh bc I still use the old rootless.h macros)
Wait you were banned?
what about, uh, launchdaemons or whatever
isn't it root_path ("/us/lib/ mytweak")1; ?
they’ll be moved to be within /var/jb
If you have a path within a plist, I’m not sure how you’d handle that tbh
There
Yeah that’s one of the things that was never clarified
lovely
need logos for plists that just converts paths based on macros

Does anyone know how to always display the "swipe up to unlock" text on the lock screen on devices with X gestures?

is this meant to be serious
DreamBerd is a shitpost about LLMs
The compilation instructions are "copy the readme into chatGPT, then copy your code into chatGPT and ask it what would be printed out"
Some nerd made an actual interpreter for it though that's mostly feature complete
ty
@grave sparrow you dont happen to have procursus strap 
discord ruined the files that are in ALL scripts and i cant get them any more
woe
i dont wanna compile tihmware for partialzip 😭
is the curl package dependant on those discord files 😭
bigboss ahh setup
what u need
we need linux and windows build script for your pzb
How can I track touches and touch movements on the screen without consuming them or breaking system gestures?
I tried -[SBSystemGestureManager shouldSystemGestureReceiveTouchWithLocation:] but that is only called for new touches and isn't called for movements
Adding a system gesture recognizer as mentioned in #development message also doesn't work because I need to consume the touch if I want to track its movements
-[BKTouchContactSet count] in backboardd seems to get called reliably with touches and touch movements but I don't really know how portable this is across iOS versions
And I want to avoid hooking backboardd if possible
I have been told that this is an xy problem but I am trying to treat touch position like a mouse cursor position for a passthrough window that is higher than every other window (I am porting/remaking a desktop app for iOS, so this is reasonable). I think tracking touches on the screen is a logical way to do this
UIDragInteractionDelegate
UIGestureRecognizerDelegate
[gestureRecognizer state] == UIGestureRecognizerStateEnded
gestureRecognizer velocityInView:self.view
CGPoint Translation ivar's
just a few things to investigate
I assumed these wouldn't work because they involved consuming the touch but I'll take a look
Thanks
I think it applies to moving the touch around prior to consumption?
just look into velocity stuff I guess
could just ask for a curl deb? lol
and i have latest curl from procursus, wdym it's broken
refuses to say why they aren't in the server, already reported by me, already fixed by nick
anyway, what's the new url for the apple silicon bootstrap
so i can update the dumb script i had
I didn't know it changed
i remember being told it was outdated
i'm still using that one that was uploaded to discord
look at procursus root 
subdomain for apt
ye i see now
hello friends
^ non systemd users
im the same lol
im only in 6 servers
well 7 now but when i leave this one itll be 6
systemd so good tho
fr
yeah this thing
What is it
what is
Sure why not
The only jailbroken device I have right now is an iPad on iOS 17 running rootless palera1n
What device/iOS version?
14 Pro Max on 16.6.1 but its not roothide
