Hacker News Reader

Comment by JohnTHaller | Hacker News Reader

I wish they would refer to password hashes as password hashes and not call them 'encrypted passwords'. Encrypted passwords screams that you don't understand security and were storing real passwords instead of hashes. (Or, more properly, salted hashes.)

View on HN

Maybe we should just give up and tolerate the non-technical usage of those terms. The situation with "hash" vs. "encrypt" seems just as hopeless as with "hacker" vs. "cracker".

I realize "hash" is too technical, but "encrypt" is a technical term with a different meaning.

If they'd just used a vague handwavy word like "encoded" we'd be left wondering what they did to the passwords and it would still be just as legible to the laymen, instead of being given an actually incorrect piece of information.

And a "cracker" is a subset of hackers. So as much as it makes self-identified hackers cringe, it is correct to say that a hacker broke into a system.

"encoded passwords".

In ordinary usage, "encrypt" means "to change information from one form to another, especially to hide its meaning" (Merriam-Webster). It doesn't necessarily mean reversible, although in programmer jargon it does.

So hashing is a subset of encrypting, just like crackers are a subset of hackers.

I'd think "ordinary usage" would be by cryptographers and other cryptography-related personnel, not laypeople. Most of the people around where I live don't even know what cryptography is, and it's hit-and-miss whether or not they've heard of encryption.

From a theoretical point of view, hash functions are not only a subset of block ciphers: hash functions and block ciphers are equivalent. You can take a hash function and build a block cipher from it (use a Feistel network). You can also take a block cipher and build a hash function from it (use the Merkle–Damgård construction).

But even if you can build one from the other, saying something is 'encrypted' implies, in even the meanest definition, that it can be decrypted. If the layperson definition does not include this aspect, then it is, well, wrong. Even if you disagree on this point, I would expect a tech journalist to endeavor to use the correct, field-standard terminology, regardless, because as Pxtl says, anyone who is familiar with even the basics of cryptography will interpret "encrypted password" in an incorrect way (i.e., they'll see it as what Adobe did).

If tech journalists would use the right terms that would at least help lead the public towards a better understandings of this issue. I presume they say 'encrypted' as their generally means 'not plain text' to the lay person. To be fair, we do call these cryptographic hashes.

By the way, if anyone is wondering how VBulletin forums store passwords... It's md5(md5(password)+salt).

ugh, cos md5 squared is twice as secure as just once - why not md5 x 10 to be uber secure?

if my memory is correct this is because they don't have access to $password; they get md5($password) from the client and to store that in the database with a salt need to run md5() again.

oh ok I thought the usual MO was:

if (md5(password + salt) == stored value) ..

It is, but if you don't have the naked password available at the time that you got your hands on the salt? May as well still salt it.

Yes a cryptographic hash (like MD5 or SHA etc) stores a non-reversible string. Encryption means that it reversible but lets be honest here, when writing an article most laymen understand the general idea of encryption as opposed to hashing.

I'd really like to see how this hack works for general knowledge and if it's purely via the script itself and not a server attack.

Don't forget that most symmetric encryption algorithms can be used as a hash. Just encrypt a known plaintext (usually all NUL bytes) using the password as the key.

If you do that and only that, you open yourself up to related-key attacks. A better approach would be to use a well-known scheme like the Merkle–Damgård construction.

Reply on news.ycombinator.com