Yeah, I also wonder why there is no full (modern) VMProtect devirtualizer yet now that AI can basically reverse engineer anything. Maybe there is but it's not public.

not only good quality programs, also some games and romhacks.

>>110025095
If the palest imitation possible of Photoshop the Rustrannies have been shilling the past week is any indicator: very poorly, as it doesn't seem to deal with decompiling unobfuscated code either.

ai can deobsfucate code easily since every single obsfucator still has to run on the cpu in some way even if its through layers of retardation

>>110025095
At this point you'te better off just making a 1:1 clone of the original program.
There's no gameplay mechanics or physics in a game.

>>110028792
*in a photoshop.
But the post do apply to westernoid games

>>110028766
as far as i understand a lot of these decomps are of poor quality, as in the code is hallucinated and barely usable because machine code doesn't match patterns properly or something

>>110028766
If the code is VMed, does it restore the original function before applying the VM or does it recereate the VM interpreter?
I find it hard to believe it does a standard decompilation in that case since a simple function that was originally 100 lines of assembly can get bloated into millions of instructions.

>>110031441
you put a vm in front of the vm and capture the full program, usually there's a small bit of translation that needs to occur with func address randomization which is nonstandard (as in not being done by the windows kernel) and applied by the vm, but that's also trivial for ai based decompilers to grab

I've done two decomps:
A commercial film emulation plugin done maybe 3 or 4 months ago - deltas between the recomp and original are basically just floating point noise because of slight differences in how very basicath functions are done in the languages
The second was started after that but stalled and astra did a breakthrough - that was a decompile of a firmware for a digital audio filter. It involved astra pulling patents and also some other firmware. The extraction was successful and I now have a vst version of the device - it's not exact 1:1 because I didn't want to emulate the full analogue-digital-analogue conversion path. But it's the same device - filter and soft clipper stuff was all extracted.

>>110025095
The CIA reverse engineering tool leak, its OpenSource now: Ghidra