News for hackers. Stuff that matters.
Re:Evernote hacked by Terretta on Saturday March 02, 2013 @01:34PM (#5310462)
Your blog post says:

> "Avoid using simple passwords based on dictionary words"

And yet your password algorithm rejects highly secure pass phrases:

> "New passwords can contain letters, numbers and punctuation."

Disallowing spaces is particularly annoying for a company with a strong security requirement, as passphrases are simultaneously far more secure and far more memorable than the monkey rules your validation demand.

It gets worse.

If I type a long sentence without spaces such as "thisphraseisdefinitelynotinthedictionary", you give me an orange warning checkmark and suggest:

> "Try using a combination of letters and numbers"

If I replace that long phrase with "abcde1", just 5 lowercase letters and a digit, you give me a green checkmark! Same green checkmark for "abc123", the third most popular password online.

Have whoever wrote your validator needs to go read http://xkcd.com/936/ and really internalize the idea.

EDIT: See also:

1. How people choose passwords: http://www.troyhunt.com/2011/07/science-of-password-selectio...

2. Top 25 passwords: http://www.prweb.com/releases/2012/10/prweb10046001.htm

3. Twitter banned passwords: http://securitywatch.pcmag.com/none/284196-the-twitter-banne...

Re:Evernote hacked by Niten on Saturday March 02, 2013 @02:20PM (#5310636)
Anyone using this comic to imply that a passphrase is more secure than a short random password hasn't done the math. This is comparing a passphrase drawn from four of the 2048 most common words against not a random password, but one based on a mutated version of one of the 65536 most common words.

The example passphrase does have the equivalent of 44 bits of entropy:

log_2 (2048^4) = 4 * 11 = 44

However, if we take a random password formed by just seven of the 95 printable ASCII characters, we already have a password four times stronger than such a passphrase:

log_2 (95^7) ≈ 45.99

XKCD is great and all, but I wish Randall had been more clear that this comparison does not touch on the strength of a random password, because I have since seen an infuriating number of people point to it to claim that passphrases are more secure than random passwords. They are not.

Re:Evernote hacked by Terretta on Saturday March 02, 2013 @03:42PM (#5310942)
> to imply that a passphrase is more secure than a short random password

I noted to internalize the idea, the entropy idea versus human "random" passwords which aren't random at all. "Normals" are using their name with a 3 instead of an E, or some word with a 1 on the end. This tends to put them in rainbow tables or easy attacks. See my link #1 in GP comment for reference.

That's why I said "pass-phrases are stronger than monkey rules".

Meanwhile, "random" passwords don't necessarily come up stronger. Forget just seven. Let's use 1Password's random sixteen (16) char generator as an example[1]:

    2Fro%7gMfQbwBktm : 85.3 bits of entropy from 72 char set
    [xNyCHZc44RzPeMJ : 85.1 bits of entropy from 82 char set
Meanwhile, going by length and charset alone and ignoring whether the characters are words (using my passphrase from GP comment):

    thisphraseisdefinitelynotinthedictionary : 153.7 bits of entropy from 26 char set
The thing is, when you're attacking Evernote's database, you don't know someone's using XKCD's idea. So you're not looking at (2048^4) [2]. You're looking at (using the XKCD password):

    correct horse battery staple : 104.2 bits from 27 char set
If as an attacker we magically know "thisphraseisdefinitelynotinthedictionary" is words without spaces, and even assume they come from the 2048 word set (though neither "definitely" nor "dictionary" are in that 2048 word set), then by your formula:

    log_2 (2048^8) = 8 * 11 = 88 bits of entropy
markedly stronger than your example random password's 45.99 bits. And as I noted, "far more memorable", meaning such a phrase could be used by regular folks.

1. Quick entropy check: http://rumkin.com/tools/password/passchk.php

2. Manual entropy check: http://www.wolframalpha.com/input/?i=log_2%282048%5E4%29

Re:Evernote hacked by Niten on Saturday March 02, 2013 @04:28PM (#5311144)
Respectfully, you are flat out wrong. This "entropy check" doesn't prove what you think it does.

This password strength test is attempting to estimate the entropy of given passwords under the assumption that it is a word roughly obeying the character distributions of English text, using Shannon's approximation. It does not apply to randomly generated passwords, which violate these assumptions, as described in Appendex A of the NIST plublication linked on that very site.[1] As that document describes, the entropy of a randomly (not user!) selected password is not estimated in this manner, but calculated according to the same formula I provided: H = log_2 (b^L), where b is the number of letters in the alphabet (95) and L is the length of the password. (If the password had any less entropy than that, then by definition it would not have been pseudorandomly generated!)

Additionally, the algorithm employed by this site does not take into account that an intelligent password cracker is capable of exploiting the construction of passphrases.

In other words, you've taken an algorithm designed to approximate the entropy of a user-selected password and misapplied it to both randomly-generated passwords and user-selected passphrases. The fact that the algorithm is able to give you a number for these inputs does not mean it is any valid indication of how difficult such a password or passphrase would be to crack.

[1] http://csrc.nist.gov/publications/nistpubs/800-63/SP800-63V1...

EDIT: But you're right that passphrases can be easier to use by people. I personally think a password manager like keepass / lastpass / etc. is a better choice than trying to select a memorable password, though.

Re:Evernote hacked by Terretta on Saturday March 02, 2013 @05:06PM (#5311296)
First, all this is beside the point. Evernote hasn't understood the concepts of either entropy or human chosen passwords. Rejecting my passphrase and accepting "abc123" is wrong. That's my original post, and that's what you objected to. Computer generated random passwords that nobody's going to use on their mobile phone Evernote client, simply don't figure into normal human use.

Our job is to recommend things that can help real people use tech more safely.

> "the entropy of a randomly (not user!) selected password is not estimated in this manner, but calculated according to the same formula I provided: H = log_2 (b^L), where b is the number of letters in the alphabet (95) and L is the length of the password"

Yes, I'm aware of that. Using that site's check, "correct horse battery staple" comes out weaker at 104.2 bits, so I listed that weaker "lower bound"[1] for that phrase. I'm happy to use whatever formula comes up with less entropy for reasons discussed in [1].

I also don't care when sharing with less technical users if it's exact. I care if I can point them to a URL that gives a reasonable approximation, which that "quick check" does. For users who want to do math, I listed both approaches:

> 1. Quick entropy check: http://rumkin.com/tools/password/passchk.php

> 2. Manual entropy check: http://www.wolframalpha.com/input/?i=log_2%282048%5E4%29

The "manual" check is pre-filled with your suggested formula. It's interesting to compare the entropy check to http://www.passwordmeter.com which I think users will "solve" as if it were a password meter puzzle, in very predictable ways.

Meanwhile, to an attacker trying the whole character space, "correct horse battery staple" is log2(27^28) or 133.1 bits of entropy. And if you use H = log_2 (b^L) on the passphrase that Evernote wouldn't accept, it comes in at 188 bits of entropy.

In any case, the approximation is a more conservative "lower bound" than the formula you're suggesting as applied to character set ^ length.

> I personally think a password manager like keepass / lastpass / etc. is a better choice than trying to select a memorable password, though.

I agree. And I use 1Password and generate random passwords.

Btw, the two truly random passwords from 1Password (equivalent of keepass, lastpass, etc), if working with the H = log_2 (b^L) formula, give only 98 bits and 101 bits. Again in their case, the "quick entropy check" URL gives lower numbers, meaning it's a remains a reasonable "lower bound" check for casual users who don't grok formulas.

I tell non-technical family members and friends who can't be bothered with password minders to use sentences meaningful to them and unlikely to be in a book.

    This phrase is definitely not in the dictionary! : 227 bits or 286 bits
That's a pretty good password that my Mom can remember.

--

1. lower bounds: http://subrabbit.wordpress.com/2011/08/26/how-much-entropy-i...

Re:Evernote hacked by Volpe on Saturday March 02, 2013 @11:38PM (#5312549)
While "This phrase is definitely not in the dictionary" is true, every component is. If this sort of password became popular brute forcing passwords would just start using whole word combinations when cracking passwords. i.e a word becomes the equivalent of a character (though from a larger set of characters).

Also wouldn't a hash of a long passphrase be longer? (I am completely ignorant the details of hashing algorithms)... so if I'm cracking a table of hashed passwords, I could set a "passphrase" cracker on hashed passwords of above average length?

Re:Evernote hacked by rudedogg on Sunday March 03, 2013 @02:50AM (#5312902)
No, the hash is always the same length, a single character and a 4GB movie will generate a hash of the same length (using the same hash algorithm (md5, sha512, etc). I've casually read about it, I think that is the point of the hash, to take an input of any size or length and represent it as a fixed size.

The problem with this is hash collisions, weaknesses found in hash algorithms can make it so attackers could generate files to match another files hash (or password or w/e). I think the idea is that they create a harmful file, then add null characters or w/e so the hash algorithm will generate/assign the two different files the same hash.

Re:Evernote hacked by ajanuary on Sunday March 03, 2013 @07:25AM (#5313493)
The key idea behind any hashing algorithm is to take inputs of any length and turn them into an output of a fixed length.

Different problem domains have different additional requirements: password hashes want to minimise collisions (multiple inputs hashing to the same output), checksums want to make similar inputs produce wildly different outputs, hashtable/dictionary/map key hashes want to create an even distribution of outputs etc. But the thing common to all hashing algorithms is the output is the same size.

Re:Evernote hacked by vikstrous4 on Saturday March 02, 2013 @04:21PM (#5311114)
When cracking passwords attackers don't just use brute force. The most effective attacks are ones that exploit human patterns such as leet replacements, capital first letter, punctuation at the end, etc. Concatenated words in all lowercase with no spaces is another pattern that can easily be added to their list and probably already is there, so yes, you can assume that that they will be looking at the space of 2048^4 such passwords.
Re:Evernote hacked by Terretta on Saturday March 02, 2013 @04:38PM (#5311188)
> The most effective attacks are ones that exploit human patterns such as leet replacements...

I think you missed my first paragraph, where I made that point:

"Normals" are using their name with a 3 instead of an E, or some word with a 1 on the end. This tends to put them in rainbow tables _or_easy_attacks_.

Also, my example passphrase that I was annoyed Evernote wouldn't let me us isn't log2(2048^4), but ~ log2(5000^8) even assuming you know it is words.

This thread is getting of on the tangent of debating XKCD's particular formula. XKCD is not the point.

My original post asked Evernote Team to please grok the idea of that cartoon, which it's obvious they had not given their clearly wrong tips and rejection of a very strong password and acceptance of one of the weakest.

Re:Evernote hacked by m8urn on Saturday March 02, 2013 @11:08PM (#5312465)
The problem with your argument is that you are equating entropy with password strength. While entropy is a valid measurement of password randomness, it is not a direct measurement of how strong a password is. Take the password "vvvvvv.vvvvvvvvvv7vvvvvv" which has terrible entropy yet is very unlikely to be cracked. Why? because a cracker does not know that your password contains only three distinct characters and would still perform a brute force attack based on an assumption of higher entropy.

If you are talking about short passwords, entropy is critical in determining the strength of the password, but the true measure of a password's strength is the permutations required to perform a brute force attack. While range of character sets determine permutations, so does the password length. You can make up for one with the other, which is why "vvvvvv.vvvvvvvvvv7vvvvvv" is a very secure password.

That aside, I did address more of the math of the XKCD comic here: http://xato.net/passwords/analyzing-the-xkcd-comic/

Re:Evernote hacked by miles on Saturday March 02, 2013 @04:50PM (#5311230)
Thanks for igniting this discussion, Niten. While digging around, I stumbled onto this tool which others might find helpful:

https://github.com/lowe/zxcvbn

zxcvbn, named after a crappy password, is a JavaScript password strength estimation library. Use it to implement a custom strength bar on a signup form near you!

zxcvbn attempts to give sound password advice through pattern matching and conservative entropy calculations. It finds 10k common passwords, common American names and surnames, common English words, and common patterns like dates, repeats (aaa), sequences (abcd), and QWERTY patterns.

Sample results (including Tr0ub4dour&3 and correcthorsebatterystaple) and a demo can be found here:

http://dl.dropbox.com/u/209/zxcvbn/test/index.html

Re:Evernote hacked by rpearl on Saturday March 02, 2013 @08:44PM (#5312092)
I also ported zxcvbn to python for those of you looking to integrate it into things that don't have javascript support (a native app or something--that was the original motivation for it).

https://github.com/rpearl/python-zxcvbn/

I am a bit slow at pulling patches (sorry, I just get super busy) but I do get to it eventually.

Re:Evernote hacked by cynwoody on Sunday March 03, 2013 @03:44AM (#5313015)
Interesting.

xyzzy scores 'instant'.

My B of A pass phrase, Jer1m1ah Cl4rke, scores 31 years. Should I change it?

https://dl.dropbox.com/u/209/zxcvbn/test/index.html

Edit: I just checked 9223372036854775808L, and it answered 'centuries' with a suspiciously round crack time in seconds. Don't think so.

Re:Evernote hacked by mbreese on Sunday March 03, 2013 @04:46PM (#5315368)
You should probably change it now...
Re:Evernote hacked by luser001 on Saturday March 02, 2013 @05:04PM (#5311287)
Pretty cool tool! My passphrases (omitting spaces between words) get a score of 2. But if I drop the vowels, the score goes to 4 and crack time to 'centuries'.

I wonder if this is a good way to create passphrases. Anybody want to chime in?

Re:Evernote hacked by tempestn on Saturday March 02, 2013 @05:38PM (#5311404)
Originally was going to chastise you for typing your actual passwords into a random demo version that could have been modified in whatever way. But since it's all js I guess it's a simple matter to verify it's not transmitting anything. Although I guess it could be some really devious thing where it saves the info in a cookie to be snagged later or some such. Did you check the JS code? ;)
Re:Evernote hacked by rorrr on Sunday March 03, 2013 @12:10AM (#5312635)
Don't use phrases. Use one really strong password that you can remember with a tool like KeePass. Generate random long passwords (30 characters) for everything and store them in KeePass.

I've recently converted all my accounts to that - my passwords are practically unbreakable, at least with the current tech. I feel way more secure than with the shit I had to memorize before.

Re:Evernote hacked by patrickaljord on Sunday March 03, 2013 @10:21AM (#5313994)
If I enter the first 10 characters in regular order of my French azerty keyboard that are &é"'(-è_çà)= I get a top 4/4 score :). Not very good for security.
Re:Evernote hacked by miles on Sunday March 03, 2013 @01:55PM (#5314766)
The author of zxcvbn acknowledges this weakness (and others) here:

https://tech.dropbox.com/2012/04/zxcvbn-realistic-password-s...

zxcvbn currently only supports English words, with a frequency list skewed toward American usage and spelling. Names and surnames, coming from the US census, are also skewed. Of the many keyboard layouts in the world, zxcvbn recognizes but a few. Better country-specific datasets, with an option to choose which to download, would be a big improvement.

Re:Evernote hacked by borski on Saturday March 02, 2013 @08:14PM (#5312019)
zxcvbn is great. We actually use it for people registering on https://www.tinfoilsecurity.com
Re:Evernote hacked by wamatt on Saturday March 02, 2013 @05:29PM (#5311369)
>"I have since seen an infuriating number of people point to it to claim that passphrases are more secure than random passwords. They are not."

It appears you have assumed all passphrases must take the same methodological route as outlined in the xkcd comic.

However, 'noozle stroodle' has an entropy of 69.303 [1]

The above quotation, is a passphrase, and it's more secure than any 7char printable ASCII password.

While I accept your argument that not all pass phrases are necessarily more secure than seven lettered passwords, one cannot rightly make the antithetical claim either.

The correct judgement is: It depends

[1] According to the JS tester, http://dl.dropbox.com/u/209/zxcvbn/test/index.html

Re:Evernote hacked by jimminy on Saturday March 02, 2013 @07:54PM (#5311953)
I agree with you, but that's not how Randall got the entropy for the comic. His entropy numbers come from NIST SP 800-63[0], for "correct horse battery staple" spaces included. But NIST SP 800-63 has been shown to be invalid in metric for entropy, making the comic's numbers incorrect in real application.[1]

[0]: http://en.wikipedia.org/wiki/Password_strength#NIST_Special_... [1]: http://reusablesec.blogspot.com/2010/10/new-paper-on-passwor...

Re:Evernote hacked by catharsis on Saturday March 02, 2013 @03:07PM (#5310816)
This assumes that it is known that the password is four random words. The security of this method hinges on the fact that that information is not known.
Re:Evernote hacked by Niten on Saturday March 02, 2013 @03:36PM (#5310914)
That's not the case, though. Even knowing the general manner in which the passphrase is constructed, if the four words are randomly selected then the best an attacker can do is to brute force the space of 2^44 possible phrases, which is difficult enough to be considered secure. The very same can be said for the seven-character random password: the best an attacker can do, knowing how the password is constructed, is to brute-force a space of about 2^46 possible passwords, which is secure enough.

And this must be true in order for passphrases to be a reasonable choice. Because if people start using passphrases commonly, then this formulation will immediately be added to password cracking toolchains along with all the other common types of password.

On the other hand, the passphrases's actual "entropy" is probably lower than the advertised 44 bits, because patterns like the appearance of an adjective ("correct") before a noun ("horse") are common enough in the English language that they, too, could be factored into a smart password cracking tool.

Re:Evernote hacked by miles on Saturday March 02, 2013 @03:22PM (#5310868)
Exactly. In a thread about "900gage!@#" being cracked in a few hours[1], this same discussion came up. It's worth reading for those who are wondering about passphrases vs. complex passwords.

[1] http://news.ycombinator.com/item?id=4545893

Re:Evernote hacked by Niten on Saturday March 02, 2013 @03:49PM (#5310973)
This is, again, not about a random password, as moxie explains in that thread. And as he continues, with regard to the comic: "I think that's totally on the right track, but if people start to do that, chances are that they'll start to create exploitable patterns again".

Any password cracker smart enough to exploit the patterns in "900gage!@#" is also smart enough to exploit the construction of an English language passphrase. The passphrase is still secure enough (probably), but it is not more secure than the random password. And if there is any one thing to take away from that thread, it should be that it's foolish to assume the obscurity of your passphrase's formulation gives you any extra security whatsoever.

Re:Evernote hacked by miles on Saturday March 02, 2013 @04:10PM (#5311053)
Thanks for your reply, Niten! Sorry, I didn't mean to imply that "900gage!@#" was random, but many people would (wrongly) consider it complex. Users who are not generating and storing random passwords (with KeePass or the like) may make safer password decisions when thinking in terms of a phrase rather than a "complex" word. Of course, they'd be far safer still by using a good password manager and long, truly random passwords.
Re:Evernote hacked by pyre on Saturday March 02, 2013 @03:01PM (#5310797)
Passphrases are more secure than random passwords when you factor in the human element. A 64 character passphrase is more memorable than a 64 random character string and less likely to be written on a post-it note.
Re:Evernote hacked by apawloski on Saturday March 02, 2013 @03:08PM (#5310821)
Be careful -- you're starting to argue a slightly different point than what your parent made. You're not wrong, but you also haven't done anything to dismiss his claim. He's talking about a 7 character password vs a 4-word passphrase. Not equal sized, 64 character passwords.
Re:Evernote hacked by fiatpandas on Saturday March 02, 2013 @02:23PM (#5310648)
>Disallowing spaces is particularly annoying for a company with a strong security requirement, as passphrases are simultaneously far more secure and far more memorable than the monkey rules your validation demand.

I just don't understand the logic behind some of these password rules. Wouldn't it require more effort to explicitly disallow certain characters? Like, they wrote code somewhere that is specifically making sure your password doesn't have a spaces, and other arbitrarily chosen characters.

It doesn't make sense to restrict anything in passwords other than length (and of course testing that is meets certain complexity/length requirements-- smartly). Just set your DB field to be 30 characters, accept any character, and be done with it.

There must be some logical explanation behind why companies implement these rules. And banks are the worst. Because people far more experienced than me at programming (e.g. Evernote devs) make these decisions, so there must be some reason. Is the decision a defense against SQL injection techniques? Since injections usually contain spaces. But then again, this is 2013 so you think they would use prepared statements, or something.

Madness!

Re:Evernote hacked by pyre on Saturday March 02, 2013 @03:04PM (#5310806)
My understanding is that banks have these weird rules because of the legacy systems that they interface with. I've even heard of new systems built on top of old ones by writing an interface layer on top of the green-screen UI of the old system (i.e. automating the green-screen interaction / screen scraping).
Re:Evernote hacked by cynwoody on Sunday March 03, 2013 @04:18AM (#5313088)
I used to have an account with New England Merchants Bank. My ATM PIN was seven digits. But I noticed that the ATM screen would flash after the first four. A little trial and error revealed that that was all I needed to enter! Comparing notes with friends (this was before the internet) revealed that this little secret was common knowledge.

Subsequently, NEMB got subsumed by successively larger banks, until, finally, Bsnk of America took over. Suddenly, my ATM card wouldn't work! It turned out that they had kept my full PIN on file through the whole chain of acquisitions, but only with B of A did they start comparing ATM PIN entry against the full PIN! My card still worked with the full PIN (which I still remembered, it being an old phone number).

Re:Evernote hacked by entropic on Saturday March 02, 2013 @04:27PM (#5311140)
Some banks have IVR systems that allow users to log in to their account via telephone, so they only allow characters that can be entered via the touch-pad.

When you first create your password they translate the characters to the numerals on the phone and then hash it.

In my experience that is the most common reason why you'll see password for policies like: "Your password must be between 6 and 20 characters and only contain upper and lower case letters."

Re:Evernote hacked by sdevlin on Saturday March 02, 2013 @02:44PM (#5310723)
Since you're only going to store the hashed value, there's no practical reason to limit the maximum length of the password.
Re:Evernote hacked by georgemcbay on Saturday March 02, 2013 @04:32PM (#5311164)
"Since you're only going to store the hashed value"...

Look fellas, we've got an optimist over here!

Re:Evernote hacked by bproctor on Saturday March 02, 2013 @05:02PM (#5311279)
netteller (I think was the name?), an online banking system used by many smaller banks, including the bank I used to bank at in Tennessee, used to require passwords be exactly 6 alphanumeric characters. This didn't change until just a year or two ago. Scary.
Re:Evernote hacked by j_s on Sunday March 03, 2013 @06:50AM (#5313428)
See also: Steve Gibson's "Password Haystack" https://www.grc.com/haystack.htm

  > everyone has always believed or been told that passwords
  > derived their strength from having “high entropy”. [...]
  > But [...] when the only available attack is guessing, 
  > that long-standing common wisdom is not correct!
(Practical advice for normal users seems to be Steve's raison d'etre; I would be glad to hear anyone share their concerns about his recommendations.)
Re:Evernote hacked by hdra on Saturday March 02, 2013 @09:59PM (#5312281)
seriously though, its 2013, is there any reason to enforce character sets allowed in a password? I should be able to use any kind of character i want.
Re:Evernote hacked by apawloski on Saturday March 02, 2013 @02:56PM (#5310768)
I don't think UnoriginalGuy actually works for dropbox -- that first line is on the actual blog post too.