Anyway to stop full decryption and decompiling? about pyce HOT 2 OPEN

@VeNoMouS could you describe your attack model in more detail? Would the attacker have full root access to a system when you execute? Would they be able to do code injection? Would they start with only access to your .pyce files?

Typically, what you're talking about is restricted by key distribution techniques which is not within the scope of this module.

This module handles the low-level encrypt/decrypt operations, but where the keys come from is up to you or your organization. For example, you could run a key server on the Internet and only provide them to licensed users.

Obviously, your code has to decrypt somewhere before hitting the CPU. Attackers could get the code at any point during that window of execution.

For example, a cloud provider could read the RAM of your VM and take the decrypted code from pages in your memory.

What matters here is:

1. Your attack model
2. Your key distribution technique

from pyce.

To recap this, if you had the pyce and you had the key, you can simply decode the whole thing without execution... with what i outlined in the original post.

The whole "key distribution technique" argument is head in a paper bag response, your simply pushing the issue/problem of this flaw back onto the user.

I originally looked at this project to distribute private code to users without them reverse engineering my work...

But once ANYONE has a key you can simply reverse engineer all the way back to source,

pyarmor is a better approach in the fact decryption is closed source and only distributed in binary form

While yes, you can place traces around the stack and still reverse engineer but that will take far more time / effort than the ~3 lines of code I provided to convert back to source that it takes with this project.

I'm sorry but I fail to see how this project is designed for distributed code encryption, when it can be reversed to full source so easily.

from pyce.