aaaAAAAAAAAAAAA so the game is happy if i do THIS, but if i do it the proper way (parsing the HTTP h

aaaAAAAAAAAAAAA
so the game is happy if i do THIS, but if i do it the proper way (parsing the HTTP headers and returning back the actual substring), it connects, then IMMEDIATELY disconnects, like the value failed to validate or something

Aqua•9/3/23, 11:38 PM

just raw html and php and good ol redirects

khitiara•9/3/23, 11:38 PM

it was like 2005-2008 everyone was slow internet users back then

khitiara•9/3/23, 11:39 PM

ajax itself isnt inherently bad but its gone way too far

Kai•9/3/23, 11:39 PM

I'm aware the conversation has moved on but I believe this is a valid XML tag because I don't even know
syntax: STag ::= '<' Name (S Attribute)* S? '>'STag ::= '<' Name (S Attribute)* S? '>'
<ᾯ #x9 :='&&ᾙ' #xA #xD><ᾯ #x9 :='&&ᾙ' #xA #xD>

khitiara•9/3/23, 11:39 PM

also ajax is a bad name imo but thats seperate

LLyris the Kitori aaaAAAAAAAAAAAA so the game is happy if i do THIS, but if i do it the proper way...

Lyris the KitoriOP•9/3/23, 11:39 PM

the parsing code looks like its returning the right data

khitiara•9/3/23, 11:39 PM

maybe

KKai I'm aware the conversation has moved on but I believe this is a valid XML tag be...

Aqua•9/3/23, 11:40 PM

xml has so many "extensions" its just unbelievable

Aqua•9/3/23, 11:40 PM

in some extension it probably is indeed valid syntax

Kai•9/3/23, 11:40 PM

this is just standard XML lmao

khitiara•9/3/23, 11:40 PM

yeah lmao

khitiara•9/3/23, 11:40 PM

i think that actually is valid xml

khitiara•9/3/23, 11:40 PM

not useful but valid

LLyris the Kitori the parsing code *looks* like its returning the right data

Lyris the KitoriOP•9/3/23, 11:46 PM

am i accidentally creating a copy of the header somewhere and returning uninitialized stack memory to the game?

Lyris the KitoriOP•9/3/23, 11:46 PM

does creating a string view copy?

Lyris the KitoriOP•9/3/23, 11:46 PM

i sure hope not

Lyris the KitoriOP•9/3/23, 11:52 PM

this has to be something small and silly

Lyris the KitoriOP•9/3/23, 11:56 PM

the new one should be more correct since im following the original functions behaviour a lot closer (you give back a pointer to the memory inside of the passed in header, no memcpy/malloc takes place, before i gave back the pointer of a string literal)

LLyris the Kitori does creating a string view copy?

Maxine•9/3/23, 11:57 PM

No that's why it's called a view

Lyris the KitoriOP•9/3/23, 11:58 PM

exactly, so the memory should still be valid when i pass it back to the game, as im passing it back a pointer to its own memory

Maxine•9/3/23, 11:59 PM

Is it not length prefixed

Maxine•9/3/23, 11:59 PM

A string view

Maxine•9/3/23, 11:59 PM

No wait I'm dumb ignite me

Maxine•9/3/23, 11:59 PM

Ignore

Lyris the KitoriOP•9/3/23, 11:59 PM

Maxine•9/3/23, 11:59 PM

Fftffff

Lyris the KitoriOP•9/4/23, 12:02 AM

heres the code so anyone can take a closer look, yes that return value is correct (it needs to return the offset from headerAddrheaderAddr to the newline after the returned

value

value

)

code.cpp2.41KB

Lyris the KitoriOP•9/4/23, 12:02 AM

i feel like im just missing something obvious

Lyris the KitoriOP•9/4/23, 2:05 AM

WOW thats a fucking red herring and a half

Lyris the KitoriOP•9/4/23, 2:05 AM

so LBP has this thing called the digest key which it uses to verify the data is both not corrupted and not tampered with

Lyris the KitoriOP•9/4/23, 2:05 AM

LBP PSP verifies this digest

Lyris the KitoriOP•9/4/23, 2:06 AM

if the digest is wrong, it refuses to connect

Lyris the KitoriOP•9/4/23, 2:06 AM

HOWEVER, there seems to be a bug in the implementation

Lyris the KitoriOP•9/4/23, 2:06 AM

if the digest is invalid eg. its not in the right format, or missing altogether

Lyris the KitoriOP•9/4/23, 2:06 AM

the check is bypassed entirely on the client

Lyris the KitoriOP•9/4/23, 2:06 AM

my original hack code

Lyris the KitoriOP•9/4/23, 2:06 AM

which faked X-exe-vX-exe-v and X-data-vX-data-v

Lyris the KitoriOP•9/4/23, 2:06 AM

had a bug

Lyris the KitoriOP•9/4/23, 2:07 AM

strcasecmpstrcasecmp returns 0 if equal
NOT 0 if different, like i thought

Lyris the KitoriOP•9/4/23, 2:07 AM

so what was happening

Lyris the KitoriOP•9/4/23, 2:07 AM

was when it checked the HTTP headers for the digest

Lyris the KitoriOP•9/4/23, 2:07 AM

it was getting returned 205205

Lyris the KitoriOP•9/4/23, 2:07 AM

for the digest

Lyris the KitoriOP•9/4/23, 2:07 AM

which is clearly wrong

Lyris the KitoriOP•9/4/23, 2:07 AM

and it failed to parse

Lyris the KitoriOP•9/4/23, 2:07 AM

and the digest check was skipped

Lyris the KitoriOP•9/4/23, 2:07 AM

HOWEVER

Lyris the KitoriOP•9/4/23, 2:07 AM

my replacement proper code

Lyris the KitoriOP•9/4/23, 2:07 AM

was ACTUALLY returning the digest when it asked for it

aaaAAAAAAAAAAAA so the game is happy if i do THIS, but if i do it the proper way (parsing the HTTP h

Similar Threads