>>1757738

Well intentioned statements aside, telling people not to look isn't going to solve his problem because ultimately someone will look.

Right now, the best possible way to secure Julian Assange's safety is to get him another batch of 'highly explosive informative shit' to act as replacement insurance files pronto.

This time, not encrypted with any NSA approved technologies. XOR'd one-time-pad is the only secure methodology I know, but the key for it would be the same length as the file set, making it hard to remember or hide.

>>1784612

SHA256 penetration won't occur for some time. I reckon the NSA are saving that mathematical backdoor for when they really want to suddenly deflate the value of BitCoin.

SHA1 family is pretty much largely broken, which doesn't inspire much confidence in the SHA2 family.

Biggest enemy right now is bad randomisation on wallet IDs.

When decrypting something - the whole point is incorrect passwords won't give you an error, because that would defeat the point of having encryption (brute force until machine goes ding!).

All AES will do is 'decrypt' the information based on the set of rules it has using the information (key and initialisation vector, AKA salt), which can include decrypting to absolute garbage if the password is wrong.

AES should only throw an error if the password, and salt, are not (seperately) either 16, 24 or 32 bytes (as AES divides messages into 16, 24 or 32 byte blocks). Some variations of AES (EG ECB) won't even accept a salt.

Some encryption programs might include a wrapper front-end (lowers security, warning!) to know when it has successfully decrypted a message by putting inside a container that is essentially an overglorified message basically saying 'this has been successfully decrypted'.

Summary: assuming you supply the right password length AND the right salt (if applicable) AND the right algorithm (there's a copyrighted/commercial variant, note) it will always decrypt 'successfully': whether or not it decrypts into the message you want is another thing entirely.

>>1825501

He might have been a tad smarter than I first assumed. As you correctly observe, why even give the extension?

.ZIP (newer versions) definitely support AES-256, but they don't encrypt file hierarchy (there's a workaround: yodawg, I herd you like .zip files). I believe 7zip however does.

If he's smart, he will have obfuscated the encryption algorithm entirely.

The best cryptographers often say 'the best encryption security is when the enemy doesn't even know it's there'. Can't crack or brute force what you don't know exists.