this post was submitted on 26 Sep 2024
13 points (100.0% liked)
Technology
59587 readers
5370 users here now
This is a most excellent place for technology news and articles.
Our Rules
- Follow the lemmy.world rules.
- Only tech related content.
- Be excellent to each another!
- Mod approved content bots can post up to 10 articles per day.
- Threads asking for personal tech support may be deleted.
- Politics threads may be removed.
- No memes allowed as posts, OK to post as comments.
- Only approved bots from the list below, to ask if your bot can be added please contact us.
- Check for duplicates before posting, duplicates may be removed
Approved Bots
founded 1 year ago
MODERATORS
you are viewing a single comment's thread
view the rest of the comments
view the rest of the comments
Any password length (within reason) and any character should be allowed. It's going to be hashed and only the hash will be stored right? Length and character limits make me suspect it's being stored in plain text.
Rules here are 64 as a reasonable maximum. A lot of programmers don't realize that bcrypt and scrypt max at 72 bytes (which may or may not be the same as 72 characters). You can get around it by prehashing, but meh. This is long enough even for a reasonable passphrase scheme.
Minor note: 64 unicode characters is potentially much more than 72 bytes.
Then you're vulnerable to simple brute force attacks, which if paired with a dumped hash table, can severely cut the time it takes to solve the hash and reveal all passwords.
By any length I meant no maximum length. Obviously you don't want to use a super short password.
"What's your password?"
"The letter A."
Mine is the null string. They'll never guess it!
Some kind of upper bound is usually sensible. You can open a potential DoS vector by accepting anything. The 72 byte bcrypt/scrypt limit is generally sensible, but going for 255 would be fine. There's very little security to be gained at those lengths.
I don't know about a min length; setting a lenient lower bound means that any passwords in that space are going to be absolutely brutal force-able (and because humans are lazy, there are almost certainly be passwords clustered around the minimum).
I very much agree with the rest though, it's unnerving when sites have a low max length. It almost feels like advertising that passwords aren't being hashed and if that's the case there's a snowball's chance in hell that they're also salted. Really restrictive character sets also tell me that said site / company either has super old infra or doesn't know how to sanitize strings (or entirely likely both)...
The only justifiable reason I can see to have a length limit is because longer passwords would take more time to process and they don't want to deal with that.
Although it would only be on the order of a couple of extra microseconds and I'm not sure how much difference it would really make. But even on cyber security forums the max password length is 64 characters.
But it really doesn't, unless you're sending megabytes of text or something. Industry standard password algorithms run the hash a lot of times, and your entry will only impact the first iteration.
I usually set mine to 256 characters to prevent DOS attacks, and also so I don't need to update it ever. Most of my passwords are actually around 20-30 characters in length (I pick a random length in the slider on my password manager), because I don't want to be there all day if I ever need to manually enter it (looking at you stupid smart TV...).
That’s exactly what someone malicious would do though, either in a single password submission or DOS via the password maximum repeatedly. IMO there is no functional security difference between a 64 and a 256 character password, so the NIST 64 character max is reasonable.
You should probably have some safeguard to prevent jokers from uploading 14.2 gigabytes of absolute nonsense into your system's password field just to see if they can make it crash. But I think limiting it to, like, 8 kB ought to be quite lenient for anything with a modern internet connection.
As others have noticed, various hashing functions have an upperbound input length limit anyway. But I don't see any pressing reason to limit your field length to exactly that, even if only not to reveal anything about what you might be feeding that value into behind the scenes.