Re: black-box implementation of CryptProtectData/CryptUnprotectData

[Michael Jung]

On Monday 04 April 2005 08:01, Mike McCormack wrote:
> Kees Cook wrote:
> > This patch implements a functional replacement for crypt32.dll's
> > CryptProtectData and CryptUnprotectData.  It does _not_ perform any
> > encrypt/decryption, but rather tracks the cipher/entropy/plain triplets
> > so that programs depending on the calls will operate correctly.
>
> Perhaps you could make it work "right" by using a key stored in ssh-agent?

Having a "correct" implementation for this would be cool for rsaenh, too. 
Currently persistent RSA keys are stored in the registry in plaintext.

Bye,