#development
1 messages ¡ Page 82 of 1
corellium

you release it and i give it to you 13 seconds after
omg i cant unzip this shit
aaron could pull the money from his personal bank
aaron buying corellium with the twitter money
nuh uh
this mf wouldn't even kick my old account

(isnt an old acc)
proof?
give me nexus src and i might
release zefram and the 2024 lamborghini aventador without cutting corners is yours
zefram first broski
flex your mod perms
spam ban
spam :/
(i am immune to gir spam detectors)
mod role flexed
@grave sparrow give me contributor access to github.com/captinc/Zefram
for no reason whatsoever
just make me sign an nda, if i leak it i legally have to give you a car 
official windows insider anniversary wallpapers
no it isnt
Text Text
good morning
gm
capts fat ass took all the spots 
Perfect he can refer me
capts fat ass took all the spots :trolley:
@pearl sail what did you say
That you and @grave sparrow are gay for each other
they are
i figured out how to get FLEXing working on jailed
idk how useful that would be tho since u cant really run tweaks
also it doesnt work on the lock screen/notification center
actually it may be able to and maybe i made an oopsie somewhere
wait what
u know
that thing where u hold on the status bar and u can see the views on the screen
There is already a way to load flex via a dylib in jailed ipas.
no
this works on the springboard
and in every app without needing to sideload that app
that's still pretty cool though
i didnt make it fully i just figured out how to dump the views on screen
and plus i already figured out how to overlay views on the springboard
so in theory someone can recreate it or it might not be that hard to port possibly
Text
just tried on rootful, same issue
you're not gonna say it 
sorry I just remembered that convo and found it too funny lol
the 2 other dickriders too saying "make something better"
Froggy did it so I don't need to
oh
how jailbreak ios 17 iphone 13 pro max

scoop
just tried google and gta car tracker, looks like every app is crashing
gm nicraze
Gm
the three horsemen of swiftui slander active in #dev
balance has been restored

fr
I was spelling the last part of shepâs word

đđđđ
@grave sparrow shep has been using php.
partially
i like money tho
shouldâve just quit imo 
true
thatâs reason enough to quit
ok fr
đđ
glad we can expect to be held accountable for such actions

python would be acceptable too depending on context
even substantial though
if i had to write a web backend itd be python
isnât swift also technically in the C family

mk
facts
they removed ++ and --
like its actually jut flat out good
fake ass language
they keep speeding it up with each update too
3.11 was like 25-40% faster than 3.10 on avg and 3.12 is 5% faster on avg
Faster cpython project đ
đ
like at this point im convinced apps could be run smoothly in python
swiftUI is probably worse
yeah legit
python+uikit bindings would be smoother than native swift+swiftui
syntax is also not regarded
php not use $ or ? challenge
I donât understand why people hate pythonâs indentation so much
Like just write good code
its dumb but
And itâs never a problem
js devs use actual functions challenge
fr
(they think => () is nice syntax)
javascript itself has very good overhead on modern JIT compilers
but, >js
when packaged with the DOM though it becomes dogshit
in terms of memory and performance
nah
v8 is good too
đŻ
not sure how good firefox's js engine is but its behind v8
also true
my first car was a v8
timing belt died within a year though 
skill issue
(its () => {})
Shut up
honestly yeah

i just added all the C and C++ post/prefix unary operators myself 
JS's syntax is nice for those arrow funcions though
really?
well realistically itâs either python or js
wasm
honestly I only use () => {} because function () {} is long
if it was like func it would be nice
tea
yea
typescript is ok
and plain js to be fair its just very lazy with typing so it requires strict development otherwise it quickly falls into gross paradigms
like returning a variable with different types depending on conditions in the same function

mfs do that
vanilla js is too lax
ts it too much work to setup, js with jsdocs and vscode set to type check it is great, but also a bit of work compared to just using a compiled language
Capital punishment
honestly maybe ive never even used it lmfao
although python is the same so
method overloadin
i cant really use that as leverage
lemme just return this const char* on a float type 
I would say it's nice for quick projects
even if you add types to the function?
maybe? i thought it was just hinting
idk if it errors or if theyâre just hints
not enforced
python with strong typing would be basically perfect
real
just return everything as void*
cursed
how and why
I hate it
you can kind of enforce it
if you annotate everything, you can have something checking the return values vs the annotations
match case is also nice
@ashen birch why is asn1 parsing so incredibly stupid
Python being actually good?
icream name leak
Please return to #jailbreak you pink name
@hasty ruin Hey, got a moment?
Struggling to flash my ipa onto Corellium, I fakesigned it with ldid but now I get this:
Failed to install app from opal.ipa: Code signature verification failed (No code signature found.)
It's decrypted
No idea tbh, I havenât used corellium in 3 years
I uploaded it as a .zip to /Applications because corellium doesnt let you upload folder
How can I unzip it? I tried unzip name.zip but it didnt owkr
What folder
Opal.app folder
I put it into a zip
and put it into /Applications
someone told me to put it there and uicache it
but im struggling to unzip it
ntrol
pyimg4 semi-does
and also agreed asn1 sucks
too bad itâll never be a real img4 parser bc itâs not âproperâ like img4lib is 
what is the latest crash log tool
i dont think cr4shed works anymore??
and how to make my tweaks support rootless
@ashen birch u must know this fr
[[crashcop]]
wtf
how do i include images in a theos app with makefile?
[desired_bundleid]_INSTALL_PATH = /Library/Application Support/
include $(THEOS)/makefiles/bundle.mk
i guess 
place stuff in folder called Resources in root
bundle?
[[FlexAll]]
Load FLEX loader dynamically.
just creates a folder at the given path more or less
the app is crashing on launch and i cant figure out why ae
specifying rootless packaging scheme translates that to /var/jb/Library btw so

bruh
it crashes on palera1n devices
wait a minute is it because its an ios 16 device and im using ios 15 sdk
good morning
can you spoof iPhone location without jb?
no
hbd!
bro i've literally never gotten into tweak dev
and haven't even messed with proc stuff in over a year tro
@vivid dew happy birthday big man!!!!
This can be a help: https://github.com/NightwindDev/Tweak-Tutorial/blob/main/p10_rootless.md
A tweak tutorial for beginners to the iOS jailbreak developer community! - NightwindDev/Tweak-Tutorial
why could I not find this /var/jb folder and only /private/preboot/jb-randomnumbers/
are you on palera1n or another rootless jb
dopamine
uhh it should be there then
Regardless from what I understand /var/jb/ is just a symlink to that folder so it doesnât really matter
I got my tweak mostly working with ROOT_PATH_NS and some other changes but that folder was nowhere to be found
ah ok
also make sure youâre compiling with a mac if you have one, and with Xcode 12 tool chain or above
Just some other prefs shit that has to be changed now, and commands that need to be switched
Yeah I had to get Xcode 14.3.1 cuz 15 wasnât working
weird
It would complain about the platform being null in some private framework tbdâs
Oh yeah that
Which it was, but had no idea how to fix so I just downloaded 14 
thatâs cuz they (Apple) changed iphoneos to ios for some reason
which broke everything
not exactly sure, I use Xcode 13 myself
oh one more thing
you have a flex loader installed, right? just to make sure
if not I can tell you where to get it
u mean like flex on device?
yea
I downloaded poomsmartâs version for rootless
didnât find any others that supported
yeah itâs only that one that seems to work
And for some reason sileo wouldnât load packages from nscake or dgh0stâs repo
Not sure if thatâs bc it checks for rootless supported packages
is there a DPKGArchitectures error
If so, yes youâre right, itâs because those repos donât work on rootless
ah ok thanks
I was using xina before this, only switched last night to dopamine
that phone been off for months
yeah itâs quite a transition lol
Now I feel like back in 2019 where I had no idea how to do anything
oh yeah havenât you not âtweak devvedâ for a long time now as well
I just realized that
yeah itâs been a long time
someone put a tweak bounty out for Akara to be updated so I thought why not
cool whatâd u make
[[Jade]]
worked with Timeloop (his design, my code) and I developed the tweak itself
Ahh yeah reminded me of that
yeah the timing was very unfortunate
RIP LQ đď¸
hey I mean at least his legacy lives on in a way
âThis course is hard. Like, really hard. Even if you've done all these things before, there is so much brain power to get through these exercises. Make sure you're practicing self-care by giving yourself the mental space necessary to learn these concepts.â Iâm in for a treat đ
gets to live a second life
haha
I worked on it far before those news, couldnâtâve predicted it even if I wanted to
so sad
Some of these animations in Akara are horrible
show us the first sneak peek into akara rootless
@radiant idol are u making use of the cc or fully replacing it
still hurts to think about it
replacing it because my goodness the stock ccâs class structure is annoying to use
my condolences :(
looks good lol
yeah I uh modify the existing design
itâs hell
make sure to check if the focus module is working
Computer science is necessary for reverse engineering iOS right?
that one is a pain in the rear end
I think iirc Linus was/is a cs student
thanking myself for adding comments to shit bc thereâs a few hooks to fix the ccâs normal behavior
it would help A LOT but I guess itâs not necessary
Hmmm Iâll have to check, but it should work as the backend of the cc should be the same
But I might have to take it out of support for the smaller buttons lol
yeah
that shit will not fit in there I think
yeah thatâs possible too
harder when youâre modifying the original cc tho
Interesting, I may take a dive into here đ https://btholt.github.io/complete-intro-to-computer-science/
thereâs one slight issue with the original cc, when closing the cc
good luck
what is it
For a small second the border of one of the views stays behind
And then closes
Lemme see if I have a video of it
wait I think ik what youâre talking about
i have the same issue, possibly
(if weâre having the same issue, that is)
whatâs your issue 
well I mitigated it by fading out when the cc is closing, but on the last frame of the cc closing, the menu just jumps up to the very top before disappearing
itâs supposed to animate going down
it does that, but then at the very end just jumps up
Oh yeah I think Iâve had that too
any fixes?
if I remember
and if I uh commented on it in my code possibly
lmao oke
Iâll hit u up when I get home from gym, then I can look
might have to do with some of the cc hooks
I love banging my head into the wall that is "kfd for M1 14.x"
lmfao
like somehow the way kfd is normally setup runs out of memory on it... despite this device having the most amount of memory an iOS device has ever had
and then when you reduce that kfd starts having extreme amounts of pain and suffering
(thanks opa for at least helping me get farther than where it was at but this is still the most annoying thing in existence)
Does someone know how to compile a build of dopamine? Do i need to do some extra steps as dopamine use extra entitlements?
4141414141414141
Couldn't you give taurine those entitlements that give you more memory
Or whatever
Have you tried that
@lapis vessel hi could u update the latest version of libimagepicker on github
not sure if the one on havoc has rootless support
(it does, at least according to Sileo)
the version on havoc is just a deb that @fading cave sent me actually haha, he might be able to make a PR for me to merge
ahh ok
classic dave
libimagepicker mfs who ever's gonna use it on their rootless tweaks:
Put the dylib from either Muirey's github (if its updated by the time u read this) or from your device on Havoc in $THEOS/lib/iphone/rootless/
I would except I'm currently banging my head into the wall of "14.x M1 kfd"
because guess who decided to buy a 2TB M1 iPad Pro for $900 (brand new)
Lol
thanks microcenter
Why does KFD not work
Increase the number of pages by a lot
To compensate for the extra RAM
rootless palera1n seems to work fine
anything specific you're looking for?
Not really
Do you have BioProtect
If not it's fine
@velvet path Is kfd just slow or is it broken?
so issue 1 is that for some unknown reason M1 14.x only let's us have 4096 IOSurface's (for reference: it normally has 16384) - this craters success rate by at least 75%
issue 2 is that we need to increase the pages a fuck ton - both to account for increased RAM and also to mitigate the success rate penalty introduced to solve the first issue
how many pages do we need? 131k pages... which both results in over 4GB of RAM usage and an over 3 minute exploitation time
how do you tell it not to use this shit codepath on non-M1
Wdym?
Like this was predictable
more RAM = slow data exploits
that's how PUAF works
I wonder how it is on 16, since there's vm swap enabled
problem here is trying to do 131k pages on other devices will likely fail miserably
and also, due to the weird M1 14.x quirk, we're facing an over 75% exploit success rate penalty (for equivalent page counts) compared to if we don't do that
Add M1 checks then
Sorry if this is dumb but why is there files like DriverKit.c and DriverKit.h where DriverKit.h includes DriverKit.c is this just for convenience?
Like why is it all not in one is what Iâm trying to ask
That makes sense, thank you.
you reimplemented the changes you lost?
also why the alt lol, I don't feel confident installing some random deb that some random account sent lol
Yes
I deleted my main
I also put them on twitter
If you trust that more
ok so update:
opa kfd app
- 131k pages worked 100% of the time
- 65.5k pages worked 0% of the time
Taurine app
- smith consistently panics
- physpuppet consistently is just stuck
(compiled with Xcode 14.2 on macOS 12.7)
Compile Taurine with it?
yes I compiled Taurine with the applicable changes
I need to inject FLEX on a Mac, I am running an iOS app via catalyst and I want to dig around in it. Please don't suggest "just use your phone" - I wouldn't be fucking with this if I didn't need to. Is there like a pre-made dylib that I can inject into an app and have FLEX be displayed? I already tried FLEXing but it didn't do anything. Does anyone know anything?
inject these into the ipa with sideloadly or whatever
ballin
@naive kraken Do Enterprise-signed IPAs not work with Dopamine and AppSync because you don't hook amfid/installd or what?
I'm confused
Done, but doesn't seem to be doing anything unfortunately, app just opens
im trying to do something with the chatgpt app 
https://f.igerman.cc/u/Ed8g2g.jpg but it just opens
not doing anything wrong am i?
looks right
your main still shows as deactivated though, not as "deleted user #69420" lol
No they don't work because Dopamine does not add non adhoc signed binaries to trustcache by design
Well it's gonna take 2 more weeks for that
Yeah okay
Not fixable right?
not really
Yeah it makes sense on second thought
you can add those to trustcache and that would work
but then app store apps and whatever else would also be added to trustcache
and that can cause unwanted side effects
there was an entire presentation at 0x41con where someone used frida to debug an app crash
the hider of root is gonna cry when his 600 app store apps get added to tc
lol true
and the whole conclusion was it happened because NSPredicate behaves differently when the app has platformization
yeah it was an attempt to mitigate recent NSPredicate exploits while not affecting apps too much
no, less stuff
some apps do
Weird
although it's a local check that consists of the framework running csops on the app itself
so the person that did the presentation was able to patch that to fix it
that's what I meant though lol
The problem is since iOS 15 you cannot run unsigned apps without platformization anymore because the only method left for bypassing code signing is the trustcache
CoreTrust bugs do allow this however
opa for how long are you reading or reverse engineering Apple's OSes? To know all this information
Potentially is there any way to, like, include missing frameworks (iOS 16+ stuff like RegexBuilder and AppIntents) into an app that doesn't support anything lower?
I mean, all the reversing I did before 2022 was just to get my tweaks to work
it was only 2022 where I actually got into kernel stuff
also kernel has source code for a lot of stuff
Where? On apples website ?
Which is decompiled? I don't think it's opensource :p
What.. It IS opensource đ
@tepid olive ellekit build seems fine so far but I also don't have a lot of tweaks installed
Yeah legally they cannot close source the kernel
Hm
Why does it take ages to look into the source of the kernel? Pages load soo slow
yeah that's it
i always ask cyp*n to test
these idiots have like 100 tweaks
well at least it didn't break crane
Yeah I have that on my test device
inb4 someone complains of a broken crane
I was getting rusty with arm64 asm and forgot that branches need to get the instruction offset added when rebinding
@naive kraken how did you start learning C & other programming languages that you use?
school
and well assembly I avoided by using decomp
only during my cellebrite internship I re-learned how to write assembly
So assembly is not really a must to be able to find some exploits?
tfw i learned assembly by making ellekit
by writing an assembler....
not at all
but it is good to know
you really only need it when the decomp fucks up
And how DO you find exploits? For example you know C, then what? Where do you look for an exploit? Could it be literally anywhere in the kernel source? Or are there certain maps where you'd have a higher chance on finding an exploit.
I never found an exploit
Oh okay, you used an exploit I guess.
and even if I did, no one would know 
Used CoreTrust exploit for TrollStore
Well essentially you are looking for code bugs
the most exploit related thing I did was replace oobPCI with kfd in Dopamime
For the rest I just used exploits that were already public
That may or may not be exploited for memory leaks
It is not a very glamorous job
It is tedious
Hm
Can I see what you've changed? I cannot see it on GitHub.
the private techniques are interesting
Oh it's private... Okay nvm
no it's public
I donât have a job in that profession so I canât spill anything
I replaced the private techniques with the ones from Fugu15_Rootful
But evelyne has access to the original ones which I guess she was referring to
lol
Coolstar has different exploits right?
Which was supposed to be used in Cheyote
no there weren't any PAC or PPL bypasses for cheyote
1984 pac brute force
CoolStar should have considered hardware breakpoint hooking.......
6 per process is fineeeee
naah
At the time it would have been cool
So CoolStar would've been using fugu15 too?
also it breaks a lot of apps and stuff like that
no just mcbc
mhm
What's mcbc?
15.0 - 15.1.1 do not need PAC/PPL bypasses to jailbreak
well....
multicast_bytecopy
page signing
Oh I remember that one, also used in TS or something with nonce setter.
I figured that out btw
you can mmap an executable segment of a dylib over any address you want
the magic is you have to use fcntl right before to load the code signature and you have to map the entire segment in
If you don't do either of those things, it will cause the page to be mapped as non execute
Do you have a poc of that
you essentially need to reimplement what dyld does
yeah but essentially you have to make a dylib with custom segments and stuff like that which I don't have anything for
Omw to make a JIT compiler that signs pages with a dev cert on the fly
You just need a fast wifi 
Sounds crazy
well it is being done in Taurines oldABI solution
I kinda wanna make sure at some point that even page signing still causes spinlocks
sign a dylib that has a modified executable segment of like libobjc in it and then map it on top of the original one
or something like that
spinlocks can be fixed most likely by wiring down the original page using krw before mapping on top of it btw I just haven't been able to figure out how to find the correct vm_page struct in kmem yet
because dsc vm map is fucked somehow
I don't think that's open source
@naive kraken
It's vm_remap and mmap
lord
that's iOS 14 though I'm fairly sure that doesn't work on 15
again there is strong boundary and code signature checks in mmap now
you can't partially map an executable segment
no
Since Cheyote exists
A guy on twitter posted that ellekit 1.1 safe modes and in his account I saw he retweeted this
Come on discord
twitter links donât embed
you have to use fx
also that link doesnât even load
it does for me
K buddies whats the new way of showing an alert after respring 
i see the springboard lockstate notif doesnt work anymore on ios 15
@hasty ruin
@hasty ruin
Like how nexus does it, also might want to work something out with them abijt the like 2 tweaks doing an alert respring loop bug
@naive kraken how did you go from making tweaks to learning jailbreak dev? any particular resources you used or..? i feel like there's a pretty big gap between those 2 things
a modern jailbreak is just a system wide tweak and a daemon that can run stuff with kernel primitives for it, Crane is more similar to a jailbreak than you might think actually
having to go from bypassing whatever mitigations to getting tweak injection is not really a common topic, there isn;t much documentation on it other than viewing past jailbreak source code
nobody feels like documenting on it either so its just gonna stay like that
you gotta figure it out on your own

- bypass dyld checks
- hook posix_spawn in launchd to add
DYLD_INSERT_LIBRARIESto every process spawned from it - in that library hook posix_spawn again to do the same thing
- dlopen a tweak loader
- profit
at least if you didn't have to add everything to TrustCache aswell to bypass codesigning
is that an issue with dopamine given you have the coretrust bug
CoreTrust bug doesn't allow tweak injection though
stuff signed with it cannot inject into system processes
only trustcached shit can
oh i see
and for writing to trustcache you need a PPL bypass
That is what Linus bypassed right?
I wish I looked into this sooner I feel like Iâm too old to start learning now đ
Indeed Fugu15 has a PPL bypass
Yeah Iâve spent most of the day looking through the code, pretty fucking crazy to be honest
^
@naive kraken Always wondered, can you platformize by setting the trustlevel to 7 in the vnode?
trustlevel and platformization are two different things
I don't know, I tried messing with the trust level before but it didn't do what I wanted it to do
so I'm unsure where this check is even enforced
So the vnode has a platformization field? Where is TF_PLATFORM inherited from (unless its checked on every process launch)
Yeah I think it's from vnode
is this how you do this ?
no qoutes?
@radiant idol ?^
and yes the VM can ssh into the device
no quotes
make do THEOS_DEVICE_IP=192.168.1.124
uhh how do i specify a user in it
It just uses root
nope
do i need to enable root login?
how tho
where ssh config

welp
how do i retsrat open ssh?
killall sshd doesnt work
or systemctl
nvm
needed sudo
đđđ
@radiant idol will it auto respring?
depends
add INSTALL_TARGET_PROCESSES = SpringBoard to the makefile and it will respring
depends if you're targetting for springboard or another app tho
what is muis
just Music
Music**
oh
music
no, it does
you dont need DEBUG=0 if you have finalpackage enabled
bc it does music and SB stuff
i havenât saw this channel in days
remove the second line
and change Music to Music SpringBoard
hi jez
yep
days i mean years
What could cause this to happen:
While building module 'std' imported from /Users/jalilfaquiri/theos/sdks/iPhoneOS14.0.sdk/System/Library/Frameworks/CoreFoundation.framework/Headers/CoreFoundation.h:19:
In file included from <module-includes>:2:
/Users/jalilfaquiri/theos/toolchain/MyToolchain.xctoolchain/usr/bin/../include/c++/v1/ctype.h:38:15: fatal error: could not build module 'Darwin'
#include_next <ctype.h>```
shitty sdk
shit tc i think
or sdk idk
one of them

no
@radiant idol how do i make it so i dont have to type my login every time
i have no other way to make tweaks unless i steal my brothers iphone x
.
that i bought 6 years ago in year 7
or just go through https://iphonedev.wiki/SSH_Over_USB
wtf
that page changed so much
fr
like this ?
yes
nice ip
its a local one
moyai
no you fucked it up
oh
do it again, ut don't run the second one until the first is finished
for the file name one does it matter what i put?
leave it blank
?
run the ssh-copy-id command like before
on my phone or mac?
mac
whos that?
my bro
cryptic? capt? hydrate?
@grave sparrow
also i didnt run the ssh-copy-id on the mac itself i did it thru ssh, would that cnage anything?
(its a mac VM btw)
just get like an old cheap iphone 6 or 7 or smth
Anyone here that got hikari to work with building rootless tweaks on iOS 15 for arm64e?
@hasty ruin
i don't use hikari on rootless
dont you do obfuscation
lol ok
Mild understatement
you write swiftui
there is nothing more cursed than that
nope
i can confirm
how long did it take until your decomp crashed
like 4 minutes
đĽ

this has the side effect of obfuscating your source code though
fr
swift symbols đ
rust syntax đ
its really not bad
i mean maybe the go button is a little too green
but other than that its literally just a normal windows program
i dont expect anything more for a jailbreaking tool
FR
@radiant idol would this work? (chaning the alpha thing (maxValueImageView.alpha = 0.45;
minValueImageView.alpha = 0.45;)
UIView *visualElement = MSHookIvar<UIView *>(self, "_visualElement");
UIView *minValueImageView = MSHookIvar<UIView *>(visualElement, "_minValueImageView");
UIView *maxValueImageView = MSHookIvar<UIView *>(visualElement, "_maxValueImageView");
[UIView animateWithDuration:0.5 delay:0 usingSpringWithDamping:0.7 initialSpringVelocity:0.2 options:nil animations:^{
minValueImageView.transform = CGAffineTransformScale(CGAffineTransformIdentity, 1.0, 0.7);
maxValueImageView.transform = CGAffineTransformScale(CGAffineTransformIdentity, 1.0, 0.7);
maxValueImageView.alpha = 0.45;
minValueImageView.alpha = 0.45;
visualElement.transform = CGAffineTransformScale(CGAffineTransformIdentity, 1.0, 1.5);
} completion:nil];
}
edit the float in memory 
me no know how
but how can i like chnage the color of it
its an image
can i add like a filter or or somthing like in css
Anyone have an explanation for this ? (skip to 00:12)
UIView *visualElement = MSHookIvar<UIView *>(self, "_visualElement");
UIView *m = MSHookIvar<UIView *>(visualElement, "_minValueImageView");
UIView *t = MSHookIvar<UIView *>(visualElement, "_maxValueImageView");
self.tintColor = nil;
visualElement.backgroundColor = [[UIColor whiteColor] colorWithAlphaComponent:0];
[UIView animateWithDuration:0.5 delay:0 usingSpringWithDamping:0.7 initialSpringVelocity:0.2 options:nil animations:^{
// m.hidden = YES;
// t.hidden = YES;
m.transform = CGAffineTransformScale(CGAffineTransformIdentity, 1.0, 0.7);
t.transform = CGAffineTransformScale(CGAffineTransformIdentity, 1.0, 0.7);
visualElement.backgroundColor = [[UIColor whiteColor] colorWithAlphaComponent:0];
UIColor *whiteColor = [UIColor whiteColor];
t.tintColor = whiteColor;
m.tintColor = whiteColor;
visualElement.tintColor = nil;
visualElement.transform = CGAffineTransformScale(CGAffineTransformIdentity, 1.0, 1.5);
} completion:nil];
}
I think I still have the version on my laptop, I can send it to you or @lapis vessel if it is still needed?
Could u do a merge request on the GitHub so it gets updated there
its for some reason one of the more popular ones so i tried to clean it up a bit
n then sig introduced me to the wonder of inetcat
yeah I had a brief look at it last night
looks good
rip iphonetunnel
why cant i compile 
make[2]: *** [/Users/brurmonemt/test/.theos/obj/iphone_simulator/debug/Test.dylib] Error 1
rm /Users/brurmonemt/test/.theos/obj/iphone_simulator/debug/Test.dylib.9f8f30ac.unsigned
make[1]: *** [internal-library-all_] Error 2
make: *** [test.all.tweak.variables] Error 2```
mike finish your sentence
@plain python sorry for the ping, but do you happen to know how Taurine selects the number of pages that kfd uses (does it use the max pages that are defined within kfd itself, or is there a thing somewhere that tells it to use a specific number of pages?)
Reason is that Iâm currently trying to compile Taurine with the relevant changes that were made that allow opaâs kfd app to work on M1 14.x, which requires both:
- reducing the number of IOSurfaceâs created from 16384 to 4096 (change already made)
- increasing the number of pages from 2048 to 131072 (Iâve changed the max number of pages, but canât find if Taurine just uses the max number of pages or uses a specific number - opaâs kfd app just defaulted to the 2048 option for instance but had a toggle in-app)
And right now, Taurine still doesnât work when compiled with those changes (smith panics, physpuppet just leads to a hang of sorts), so this is the only potential lead I have
the first 0x800 is the number of pages
I love Xcode
apparently my Apple ID got revoked or something
just use a diff one
is the reduction from 16384 really necessary, doesn't that hinder success rate
yes and yes - that's why we need 131072 pages
this appears in logs, and trying to go over 4096 throws some iokit out of memory error or something like that
how does that not happen for iphones though
oh and did I mention that increasing the number of pages to 131072 also increases:
- RAM usage to over 4GB
- exploitation time to around 2-3 minutes
hello developers
ok new problem: exploit now succeeds but post-exploit fucking dies
wonderful
how amazing
so it's dying on amfi
any devs that can help with an extremely rudimentary question
is the run action in xcode release or debug
debug
how do I build for release in a way that actually wants to install
does using command k then making the .app to .ipa work
i have a metapackage that installs multiple tweaks; is it possible to make this metapackage edit the preferences for these tweaks prior to respring or can the preferences only be changed manually through settings.app?
postinst can edit the plist files of the tweaks? not sure
exactly what i was thinking, but how exactly would i make postinst edit those files
@velvet path In the Taurine scheme, you can build for release
well I just left home and won't be back for a few hours so
can i use defaults like on macos?
rename the default plist and copy over an edited one you want bundled with your metapackage
idk thats prolly a bad idea
if the defaults command works on ios then thereâs the solution
or plutil
does it have to be a real one or does segfaulting springboard count

A real one
ok defaults works
___CFNotificationCenterGetDistributedCenter_block_invoke
__text:0000000180432C98
__text:0000000180432C98 var_s0 = 0
__text:0000000180432C98
__text:0000000180432C98 PACIBSP
__text:0000000180432C9C STP X29, X30, [SP,#-0x10+var_s0]!
__text:0000000180432CA0 MOV X29, SP
__text:0000000180432CA4 ADRL X8, _kCFAllocatorSystemDefault
__text:0000000180432CAC LDR X0, [X8] ; ___kCFAllocatorSystemDefault
__text:0000000180432CB0 ADRL X1, __kCFXNotificationConfigurationStandardDistributedConfiguration
__text:0000000180432CB8 BL __CFXNotificationCenterCreate
__text:0000000180432CBC ADRP X8, #___hostCenter@PAGE
__text:0000000180432CC0 STR X0, [X8,#___hostCenter@PAGEOFF]
__text:0000000180432CC4 LDP X29, X30, [SP+var_s0],#0x10
__text:0000000180432CC8 RETAB
@tepid olive
+36 is 0x0000000180432CBC
so hooking __CFXNotificationCenterCreate is probably what's broken
I don't see anything bad in that function however
000000018040D8C4 __CFXNotificationCenterCreate ; CODE XREF: ___CFNotificationCenterGetLocalCenter_block_invoke+20âp
__text:000000018040D8C4 ; ___CFNotificationCenterGetDistributedCenter_block_invoke+20âp ...
__text:000000018040D8C4
__text:000000018040D8C4 var_60 = -0x60
__text:000000018040D8C4 var_58 = -0x58
__text:000000018040D8C4 var_50 = -0x50
__text:000000018040D8C4 var_40 = -0x40
__text:000000018040D8C4 var_38 = -0x38
__text:000000018040D8C4 var_30 = -0x30
__text:000000018040D8C4 var_20 = -0x20
__text:000000018040D8C4 var_10 = -0x10
__text:000000018040D8C4 var_s0 = 0
__text:000000018040D8C4
__text:000000018040D8C4 PACIBSP
__text:000000018040D8C8 SUB SP, SP, #0x70
__text:000000018040D8CC STP X24, X23, [SP,#0x60+var_30]
__text:000000018040D8D0 STP X22, X21, [SP,#0x60+var_20]
__text:000000018040D8D4 STP X20, X19, [SP,#0x60+var_10]
__text:000000018040D8D8 STP X29, X30, [SP,#0x60+var_s0]
__text:000000018040D8DC ADD X29, SP, #0x60
__text:000000018040D8E0 MOV X20, X1
__text:000000018040D8E4 MOV X19, X0
yeah wtf
Impossible @sacred orbit
Not without a PPL bypass
@naive kraken Without many tweaks, this is the log
Pastebin.com is the number one paste tool since 2002. Pastebin is a website where you can store text online for a set period of time.
A bad syscall it seems
wtf
Genuinely no idea now
Does EXC_CRASH (SIGSEGV) point to a segmentation fault?đ
So something tried to access a memory location but was denied access?
Or am I on the completely wrong page đ
is there any way to include additional files in a deb package that i can move to the correct location with postinst
i still wonder if replacing dyld shared cache is possible, i even the found the value i was searching for, replaced the file with it with mdc with the changed value (the value deals with springboard) and tried to respring. but it was expected that a respring wont work. so im wondering how i can âreloadâ springboard or anything that will trigger the reload.
evelyn told me that a userspace reboot should work but how
that is evelyn
Does anyone know if hypothetically someone wanted to enable some of the InternalUI stuff on ios 14, what additional changes need to be made as well as the ones requierd on ios 13?
It seems like adding the Internal keys to the SystemVersion.plist isn't enough on ios 14 for anything to happen
it is not gonna work
Correcting myself
Idk why I said that
The cache is never reloaded until reboot, unless you somehow get the vm system to reload it with krw
@naive kraken Maybe you know something about that
@naive kraken It's Choicy
LMAO
Just disabling a tweak safe modes
wtf
yeah it works for me too
Just found a Choicy bug ig...
Let's see if anyone hooks CFX
@naive kraken No tweak hooks it
Unless it's obfuscated
this is complete nonsense....
What lang?
I need help with my code
@cinder pivot I fuck with your nickname
I just need to decompile it and make it to a dylib
đđđđđđ
U got a Mac book tho
You need to decompile a code?
I need help really bad
How
I donât have a MacBook
Itâs for iOS
Thatâs why
@cedar stag @pearl sail so can yâall help
Donât have my mac rn
Sorry mate Iâm not near mine
@tepid olive i think i have a sb crash for u
in anlytics an imporment i says springboardd
@ me if u want me to dm crash log thing
also i have alot of jetsamEvent crashes ig
please
Hello guys is it a good idea to buy a 13 Pro on 15.6.1 I know it has no bugs atm but are there any exploits for this specific version?
Iâm asking here because the devs probably know better about the exploits/bugs (sorry not sorry)
it has KFD and MDC but idk about the next coretrust bug
But I could update to iOS 17.0 which has the bug right? With delay ota
it does have the coretrust bug, however it doesn't have KFD
it does have new MDC but i don't know if we can use it for a new trollstore
So what would you do on 15.6.1
i spent a while trying to figure out why it crashed on launch just to realize i was compiling for arm64e instead of arm64 
and then i used the wrong build command and kept using the old build with the wrong arch so it was still crashing

Theos? If yes just define ARCHS= arm64 arm64e in the makefile.

