Until now, only password surveys have been analyzed, and I knew the results would be almost useless when compared to a real dataset of passwords because who in their right mind would donate one of their own passwords, or anything like it, to a survey about password statistics?
Python’s keyring lib supports Windows win32crypto, Mac OS X’s SXKeychain, KDE’s KWallet, Gnome’s keyring, and encrypted or unencrypted password files. When your application wants to store or fetch data from the keyring, it will just work.
My suggestion is to come up with a system that makes sense to you. If it asks for your favorite pet’s name, maybe the answer could be “nail biting” or “chocolate lab”. You just have to remember how you translated the secret question.
The model has strong principals, but the goals blow my mind. I’ll explain in a minute, check out the principles first.