[krapp]

    Usernames cannot contain more than 32 characters 
    and they may only contain upper/lower case
    alphanumeric characters (A-Z, a-z, 0-9), dot (.), 
    hyphen (-), and underscore (_). 
    Passwords must contain between 8 and 32 characters.

My money is on "ineptly."

[terminado]

There's really not much rational for capping passwords at anything beneath 256 characters.

256 characters makes for a fairly sizable passphrase, and doesn't represent a substantial hit on storage space. In reality, even if they were stored as encrypted binary/base64 in a nosql file system of structured data files, 4096 is pretty much the de-facto floor for disk space occupied by non-zero-byte individual files on most modern file systems.

...variable data size being a concern in cases where the transformed value is encrypted rather than hashed.

[gherkin0]

> 256 characters makes for a fairly sizable passphrase, and doesn't represent a substantial hit on storage space.

They shouldn't be storing passwords at all so storage space should be a non-issue. My 20 meg password should hash down to the same small(er) value as your 15 character one.

[0942v8653]

On the other hand, you might not want to be hashing a 20meg password. It is fast on my computer but it's fair to limit at something more reasonable.

    $ python -c 'print "8 bytes\n" * (20 * 1024 * 1024 / 8)' > 20meg.txt; time shasum -a 512 20meg.txt 
    59cb7f88ad8d6229e6d3a74ee422dff57e17f168c6e6fa44ef32c3f07a73a6e455d8b55c1265d5212b9ed5475b6d9364286645200dada59aa16905a9ce748561  20meg.txt

    real	0m0.289s
    user	0m0.284s
    sys	0m0.004s

    $ python -c 'print "8 bytes\n" * (16 / 8)' > 16byte.txt; time shasum -a 512 16byte.txt
    b6043d3a520424d5ec17dc0c23ba3b591d74517e2b9faa0df2d69d13c89a5f372d6dc35f95836687ee05be18433277e1c4b67393eb2771b475d655a832b16654  16byte.txt

    real	0m0.045s
    user	0m0.040s
    sys	0m0.004s

[thaumaturgy]

Fast password hashing is bad anyway. Slow schemes are better (assuming they come from a thoroughly tested library written by someone that actually knows what they're doing).

[rev_bird]

I'm admittedly not familiar with the details of the hashing process, but it could be done client-side, no? Then the compute power required falls on the user, PLUS the 20 megs never gets sent over the wire.

[mkagenius]

No. Actually we would not want the hashing technique to be exposed in the source code.

[ColinWright]

That turns out not to be the whole story.

Hiding the technique to compute the hash is relying on security by obscurity. Ideally you want a system that is secure even if potential attackers know what methods you're using.

[GlennS]

If you do the hashing on the client side, then the hash itself effectively becomes the password.

[roblabla]

That would assume your users have JS enabled. I've seen something like that done before, but always with a fallback in case user has JS disabled.

[Retra]

If you're not sending 20 megs of data, you're not getting 20 megs of security. So why allow it if it doesn't add anything?

[ColinWright]

It doesn't add to the security, but I might find it easier to remember 20 MB of redundant and meaningful stuff than to remember 384 bits of literally random stuff. The entropy might be the same, but my memory is not a computer. I can remember vast amounts of material that is meaningful and use it as a password. I can't remember 384 bits that have no meaning.

The benefit isn't in the entropy, it's in the abilities of your users to remember their passwords/passphrases in the first place.

[a_t48]

When there will be multiple shorter passwords that hash to the same value, is there a point to a 20mb pass?

[ColinWright]

Yes. Ideally you want users to be able to remember their pass-phrases. To do so usually implies significant internal structure and/or correlations, so to get the necessary entropy they will be large. The fact that they hash to the same value as other things is effectively irrelevant.

[gherkin0]

I misspoke, I meant size.

[elektromekatron]

Depends. Can you guess them?

[a_t48]

If I'm an attacker who is running through hashes...yes. Faster than the 20mb one.

[bosdev]

There is a slight exception to this. If they are using an older statically typed language like C, it might make sense for them to have a limit on the buffer ready to store your unhashed password. Yes it seems crazy these days, but it might apply to some of the older systems which have password length limits.

[h0w412d]

BCrypt has a character limit of 72.

[Freaky]

Hard limit of 72, beyond which many implementations will silently truncate, and reduced entropy from each character beyond 55 bytes.

Probably a good idea to pre-hash. Or use scrypt.

[chucksmash]

I can't find it now but I seem to remember this came up in response to another breach ~24 months ago. At that time they made an announcement to the effect that from then on you'd no longer be able to have your password sent to you if you forgot it, but that you would instead need to use an account recovery key.

I took that to mean that prior to being pwnd they had been storing passwords cleartext and would no longer be doing so.

Also, the wording about allowed special characters seems to be incorrect. I personally have a non ./-/_ special character in mine. Unless they are doing something terribad like silently discarding noncompliant parts of the password.

Re: password length - at least 32 characters is respectable. I believe last time I used outlook.com they had a max length of 12-16!

[chucksmash]

Oh and on the topic of silently discarding portions of passwords, another outlook.com password deficiency (circa 2011, doubt it still exists):

When setting the password, max length was only enforced by a text input with a max length attribute. You could happily type more characters and everything would work as expected....until you went to log in. The max length on the password field on the login form was greater so those characters that were silently dropped when setting the password suddenly weren't.

[uxcn]

It used to be a salted MD5 hash, but it may have changed.

[krapp]

I would have assumed of course that the size limits were because the passwords were being stored in plaintext in fixed-length fields, but I guess they wanted to make sure they were 'complicated' enough? I guess salted md5 is literally better than nothing.

The character limits for usernames, though... smells like a SQL injection issue. Which is an obvious and completely naive thing to assert but they're using PHP so my immediate thought is that they're passing raw userdata into the database as strings.

[uxcn]

my immediate thought is that they're passing raw userdata into the database as strings

That was my first thought too. I'd guess that it's a vulnerability somewhere in the code for handling the forums.

I would be willing to bet that they could get rid of a lot of the attack surface just by using standard services for certain things.

[krapp]

Probably. If they're not using PDO then that needs to be their first priority, dead stop. After that, maybe looking at their captcha script, because those sometimes have issues if they're not well designed. I don't know where theirs comes from but it doesn't seem to use much obfuscation so it's probably old. After that, Twig.

Although judging by a screenshot of the recent hack[0] posted here[1] escaping (and XSS) may not be an issue.

[0]https://i.imgur.com/pl22srz.png

[1]https://news.ycombinator.com/item?id=9990221

[ProjectEuler]

Admin from PE here.

We've already been using PDO. As for overall privacy/security, please see https://projecteuler.net/privacy

[krapp]

And PHPass as well, fair enough.

Thank you for showing up and addressing my armchair criticisms. I appear to stand corrected.

[uxcn]

I genuinely hope the security hole is findable/fixable. Thank you guys for continuing to run an awesome service, despite asshats repeatedly trying to abuse it.

[sfrank2147]

Thanks for the response!