Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Password Requirements: Myths and Madness

franzoni.eu

Post reply on HN
Print view View original

31–40 of 257 posts

Re: Password Requirements: Myths and Madness

#31

Post by lucb1e »

chadlavi wrote: post #9
janalsncm wrote: post #8

> A short password is easy to bruteforce. But, a website should aim at password strength, not at password length. Is there a standard metric for “strength” though? I built my own random password generator and considered adding a strength meter but all of the functions I could find online seemed like they had pretty arbitrary logic. Length is probably the most objective proxy for strength.

The author links to a recommended library in the blog post https://github.com/dropbox/zxcvbn

I was looking, last week, what to recommend to a customer and ended up at https://github.com/zxcvbn-ts/zxcvbn because the one from Dropbox seems to be completely abandoned and does not recognize e.g. 2020 as a year.

Other alternatives are also welcome if there is an even better one I missed! This was the best password strength estimator I could find.

Re: Password Requirements: Myths and Madness

#32

Post by _dan »

In the b2b world it's basically impossible to improve password policies. Most of the onerous examples only exist because some other entity (a customer, insurance company, parent company, etc) has demanded them. The problem is that the demand isn't being made by security professionals, it's being made by risk management people who are only interested in a simple way to mitigate risk - it's simply much easier for them to reach for an obvious and easily testable policy with character counts, expiry, no autocomplete, etc.

Even if you manage to convince one of them to accept something a little more forward-thinking and less user-hostile, it's only a matter of time before Big Customer X comes along with their antiquated requirements and you have to choose between doing what they want or losing that business.

Re: Password Requirements: Myths and Madness

#33

Post by cirrus3 »

joezydeco wrote: post #11

My favorite is this transaction: I forgot my password. Can you send me a link? Sure, here you go. Pick a new password. Okay, CorrectHorseBatteryStaple. Sorry, you can't use an old password

Old password detection isn't necessarily a sign of weak security, if they are doing it right, they are just comparing a hash of the new password to saved previous password hashes. It is probably debatable if it is better security to not save old passwords hashes at all to perform this check, or to let users re-use old passwords that may have been disabled since they are known to have been cracked.

Re: Password Requirements: Myths and Madness

#34

Post by the_af »

danrocks wrote: post #26
joezydeco wrote: post #11

My favorite is this transaction: I forgot my password. Can you send me a link? Sure, here you go. Pick a new password. Okay, CorrectHorseBatteryStaple. Sorry, you can't use an old password

Funny, sure. But you still can have the same transaction by hashing "CorrectHorseBatteryStaple" then comparing with the stored hash.

I think the OP is arguing you can guess a (possibly) current password this way.

Re: Password Requirements: Myths and Madness

#35

Post by patrickdward »

_dan wrote: post #32

In the b2b world it's basically impossible to improve password policies. Most of the onerous examples only exist because some other entity (a customer, insurance company, parent company, etc) has demanded them. The problem is that the demand isn't being made by security professionals, it's being made by risk management people who are only interested in a simple way to mitigate risk - it's simply much easier for them…

sad but this is too true

Re: Password Requirements: Myths and Madness

#36

Post by anewguy9000 »

danrocks wrote: post #21
anewguy9000 wrote: post #4

yep, most of the discussion about passwords completely miss the point. a random word, like "dog" or "pingpong" is fine if the pqsswords are salted and hashed appropriately. how often have your accounts been hacked this way? if an adversary is really banging on the hash, and they want it, any password under around 50 characters is as good as "dog", and no "complexity" meter is gona cut it. that xkcd comic that says 55…

> yep, most of the discussion about passwords completely miss the point. a random word, like "dog" or "pingpong" is fine if the pqsswords are salted and hashed appropriately. how often have your accounts been hacked this way? A word like "dog" or "pingpong" is favored in a rainbow-table type of attack. Nobody pre-hashes "109231oijoasdfnaisdfabatteryhorse123". And yes many passwords have been hacked this way [0]. [0]…

apoligies for the lack if clarity

how does a rainbow table crack "dog" with the salt "109231oijoasdfnaisdfabatteryhorse123"?

rainbow tables are as old as time and indeed still work on passwords with poor salting. for more complex (but not complex enough) passwords there are more modern approaches, like probabilistic candidate generation

Re: Password Requirements: Myths and Madness

#37

Post by chordalkeyboard »

anewguy9000 wrote: post #25
chordalkeyboard wrote: post #17

Earlier quoted context omitted.

short passwords are still a security weakness even when properly stored because the time necessary to brute force them is relatively low.

i agree. but most sites that enforce a policy (8 chars, symbols, etc) are bruted just as easily. we need to take a step, away from passwords, to secure against brute force in 2022

well the gp's point is that it literally is not 'just as easy' because the space of possible passwords increases greatly with each character, such that 'dog' is much, much easier to bruteforce than 'doggie12'.

Re: Password Requirements: Myths and Madness

#38

Post by cirrus3 »

LeoPanthera wrote: post #22
cirrus3 wrote: post #12

Earlier quoted context omitted.

Check out https://www.grc.com/haystack.htm

This particular tool is misleading at best. It counts only characters, and so misleadingly overrates passwords which are strings of words. Compare it with this one, which actually identifies words and other patterns, for a more accurate guess of the search space: https://www.bennish.net/password-strength-checker/

I only said check it out, there is a lot of content there useful to the wider topic and the page clearly states what it is and isn't.

>>> IMPORTANT!!! What this calculator is NOT . . .

It is NOT a “Password Strength Meter.”

Since it could be easily confused for one, it is very important for you to understand what it is, and what it isn't:

The #1 most commonly used password is “123456”, and the 4th most common is “Password.” So any password attacker and cracker would try those two passwords immediately. Yet the Search Space Calculator above shows the time to search for those two passwords online (assuming a very fast online rate of 1,000 guesses per second) as 18.52 minutes and 17.33 centuries respectively! If “123456” is the first password that's guessed, that wouldn't take 18.52 minutes. And no password cracker would wait 17.33 centuries before checking to see whether “Password” is the magic phrase.

Okay. So what IS the “Search Space Calculator” ? This calculator is designed to help users understand how many passwords can be created from different combinations of character sets (lowercase only, mixed case, with or without digits and special characters, etc.) and password lengths. The calculator then puts the resulting large numbers (with lots of digits or large powers of ten) into a real world context of the time that would be required (assuming differing search speeds) to exhaustively search every password up through that length, assuming the use of the chosen alphabet.

How can I apply this to my daily life? Answering that question is the reason this page exists. The whole point of using padded passwords is to adopt a much more you-friendly approach to password design. On June 1st, Leo Laporte and I recorded our weekly Security Now! podcast as part of Leo's TWiT.tv (This Week in Tech) audio and video podcasting network. You may download a shortened, 37-minute, excerpted version presenting the padded password and Haystack calculator concepts:

Re: Password Requirements: Myths and Madness

#39

Post by lucb1e »

janalsncm wrote: post #8

> A short password is easy to bruteforce. But, a website should aim at password strength, not at password length. Is there a standard metric for “strength” though? I built my own random password generator and considered adding a strength meter but all of the functions I could find online seemed like they had pretty arbitrary logic. Length is probably the most objective proxy for strength.

Password strength cannot reliably be measured by looking at a password.

To take one common trick: is VpttrvyJptdrNsyyrtu a strong password? It looks pretty okay, definitely less of a pattern than CatName1985, and an estimator like zxcvbn might also not know that it's just CorrectHorseBattery offset by one character on a qwerty keyboard. Now that you know this, you can start to guess my other passwords. Would you still say it is strong now? What if I'm not the only person on the planet using this trick, does that affect it?

The strength calculator you're looking for is (technically) the method of generating the password. If you generate 6 random words on a dictionary size of 15'000, that's log(15e3⁶, 2)≃83 bits of entropy. You can only know this by asking the user how they came up with the password, or by telling them to do it a certain way. The rest is just guessing, though I will say that something like zxcvbn does a surprisingly good job at it.

The caveat is that the generating method should not have any plausible chance of overlapping with other (insecure) in-use methods. E.g., if you roll a die between zero and a hundred quadrillion and use that as password, most of the time you'll roll a really long number. But if you roll 123123 or 12345678 and use that as password, you can expect it to be cracked. Technically, by excluding "weak" choices from the list of possible values, you are reducing the strength of the password because there are fewer remaining valid guesses, but given how attacks actually work, it makes sense to do this.

Re: Password Requirements: Myths and Madness

#40

Post by lolinder »

lucb1e wrote: post #13

> A short password is easy to bruteforce. But, a website should aim at password strength, not at password length. If it is "easy to bruteforce" (guess without any prior knowledge), then it is not "strong". > If I use four unicode code points outside the [multi-language characters table], my password is probably very safe already. Security through obscurity. This is safe until everyone does it. It's not a strong passw…

> Security through obscurity. This is safe until everyone does it.

If he's actually randomly selecting from the entire set of valid Unicode characters, then theoretically his four character password has orders of magnitude more entropy than an eight character alphanumeric password.

The problem is that from the website's perspective there's no way to know what set of characters the user is choosing between. They could be dealing with someone who's picking truly at random, but more likely someone just thought they'd be cute and throw a few popular emoji at it. By just looking at a single password I can't tell the difference, so it's safer just to impose a minimum length requirement.

Post reply on HN