Passwords

8tpercent@lemmy.world to Technology@lemmy.world – 1160 points –

We've all been there.

231

You are viewing a single comment

Fun fact: password controls like this have been obsolete since 2020. Standards that guide password management now focus on password length and external security features (like 2FA and robust password encryption for storage) rather than on individual characters in passwords.

Since 2017 at least; and IIRC years before that; that's just the earliest NIST publication on the subject I could find with a trivial Web search.

https://pages.nist.gov/800-63-3/sp800-63b.html

Verifiers SHOULD NOT impose other composition rules (e.g., requiring mixtures of different character types or prohibiting consecutively repeated characters) for memorized secrets. Verifiers SHOULD NOT require memorized secrets to be changed arbitrarily (e.g., periodically). However, verifiers SHALL force a change if there is evidence of compromise of the authenticator.

"Memorized secrets" means classic passwords, i.e. a one-factor authentication through a shared secret presumed to be known to only the right person.

I wouldn’t say obsolete because that implies it’s not really used anymore. Most websites and apps still use validation not too dissimilar from the OP, even if it goes against the latest best practices.

Yeah, the most recent one for me was creating a password at lemmy.world

I wouldn’t say obsolete because that implies it’s not really used anymore.

I'm not sure where you heard someone use the word "obsolete" that way, but I assure you that there are thousands if not millions of examples of obsolete technologies in constant and everyday use.

For today's 10,000 who have never seen it, https://xkcd.com/936/ succinctly explains why the whole mixed character types thing isn't favoured.

I'm still waiting on an XKCD that references #936 with the fact that we soon as we have reliable, functional quantum computing, all of the passwords from before that point in time will be completely and utterly broken. That the only way to make a password that a quantum computer would have a tough time breaking is if it was made by another quantum computer. Unless of course the comic has already been made and I just missed it, which is a complete possibility because this year for me has been utterly crap.

Some of them are broken by quantum computers, but not all of them. For example, SHA256. You can use Grover's algorithm to take sqrt(n) steps to check n possible passwords, which on the one hand means it can be billions of times faster, but on the other hand, you just need to double the length of the password to get the same security vs quantum computers. Also, this is the first I've heard of a hash that uses a quantum computer. Do you have a source? Hashes need to be deterministic, and quantum computers aren't, so that doesn't seem like it would work very well.

Maybe you're getting mixed up with using quantum encryption to get around quantum computers breaking common encryption algorithms?

Except you can run a dictionary attack on that and suddenly it's only 4 variables that are cracked way faster than the first password.

Except you can run a dictionary attack on that and suddenly it's only 4 variables that are cracked way faster than the first password.

People should be made aware of all the tools available to properly manage tons of passwords. Not even going too deep into "passkey" stuff or any modern shenanigans, but a password manager used to generate random passwords for each separate sites is such a simple step.

Yeah! And nowadays the industry is pushing towards password less authentication. Github just started rolling it out to beta users