Slashdot Mirror


Password Power Rankings: a Look At the Practices of 40+ Popular Websites (helpnetsecurity.com)

Orome1 shares a report from Help Net Security: Nothing should be more important for these sites and apps than the security of the users who keep them in business. Unfortunately, Dashlane found that that 46% of consumer sites, including Dropbox, Netflix, and Pandora, and 36% of enterprise sites, including DocuSign and Amazon Web Services, failed to implement the most basic password security requirements. The most popular sites provide the least guidance when it comes to secure password policies. Of the 17 consumer sites that failed Dashlane's tests, eight are entertainment/social media sites, and five are e-commerce. Most troubling? Researchers created passwords using nothing but the lowercase letter "a" on Amazon, Google, Instagram, LinkedIn, Venmo, and Dropbox, among others. GoDaddy emerged as the only consumer website with a perfect score, while enterprise sites Stripe and QuickBooks also garnered a perfect score of 5/5. Here's a screenshot of how each consumer/enterprise website performed.

127 comments

  1. Uh by sexconker · · Score: 5, Interesting

    Didn't we just have a (absolutely stupid) story about how password complexity rules are bad?
    Which is it?

    (Hint: Password complexity rules are a good way to prevent the dumbest of passwords from being used.)

    1. Re:Uh by Anonymous Coward · · Score: 0

      They don't even prevent the dumbest of passwords from being used. password becomes Password1!

    2. Re:Uh by webnut77 · · Score: 2

      [0:root@yoda ~]$ pwqgen
      more2rival+Relish

    3. Re:Uh by Anonymous Coward · · Score: 0

      I'm not sure why you're confused. This article doesn't say anything about super-complex hard-to-remember passwords.
      It scores things like minimum length, preventing brute force attacks, and having an on-screen "password strength" guide".

    4. Re:Uh by geekmux · · Score: 4, Insightful

      Didn't we just have a (absolutely stupid) story about how password complexity rules are bad? Which is it?

      (Hint: Password complexity rules are a good way to prevent the dumbest of passwords from being used.)

      To clarify, the author of complex password policies that have lasted 15+ years had regret for one reason; the rules were too complex for users. In other words, he underestimated just how stupid and ignorant the masses are.

      Force complex passwords? Users write them down. Every time. And "hide" them in the same stupid place.

      Don't force complex passwords? Users create shitty passwords, and the Top 10 Shitty Passwords in 2017 are the same Top 10 Shitty Passwords used in 1987.

      Force password changes? Users change from Password1 to Password2. You'll be able to guess their password 5 years from now.

      Don't force password changes? Users never change them. Ever. Even if they are a victim of hacking or identity theft, they insist on keeping the same shitty password they used in high school. If you forced them to change it, they would have to write it down.

      Sorry, but it doesn't matter what NIST or any other standard recommends; All the password rules in the world won't prevent the masses from building a better idiot.

      TL; DR - The problem isn't password policies; it's stupid users.

    5. Re:Uh by lucm · · Score: 4, Funny

      minimum length

      What would be cool is minimum keystrokes instead. That way one could have a couple backspace in the password. Try to rainbow table that!

      --
      lucm, indeed.
    6. Re:Uh by UnknownSoldier · · Score: 2, Interesting

      > Force password changes? Users change from Password1 to Password2. You'll be able to guess their password 5 years from now.

      That is why I append a 4 digit to the passphrase, of the format MMYY, of when the password expires as a mnemonic for when it expires.

      Your crappy "password1" becomes "password0817"

      Good luck guessing the first part -- the pass phrase, along with the second part -- when it expires.

      > The problem isn't password policies;

      Incorrect. I've seen sites where they had a maximum password length, usually like 8 characters. Seriously, WTF. You are _intentionally_ making your passwords insecure???

    7. Re:Uh by Anonymous Coward · · Score: 0

      This is a lame assessment, they should evaluate how those sites store the passwords.

    8. Re:Uh by Anonymous Coward · · Score: 1

      Long time ago on some kinds of Unix (not Linux) systems it was possible to embed control characters or backspaces in your local account password. You could, in fact, do what you say!

      I've never seen any kind of web site that would accept such a thing though.

    9. Re:Uh by GuB-42 · · Score: 1

      (Hint: Password complexity rules are a good way to prevent the dumbest of passwords from being used.)

      No they are not. Checking over a list of common passwords is.
      Half of the time the rule abiding password just has an uppercase first letter and a number or symbol at the end. Yeah, it adds a bit or two of entropy but it isn't worth the annoyance.

    10. Re:Uh by viperidaenz · · Score: 2

      brute force is mitigated by account lockout. If someone has a local copy of your password hashes, restricting the available passwords is only going to help a brute force attack.

    11. Re:Uh by Anonymous Coward · · Score: 0

      Perhaps because browsers use the backspace for text edit instead of a stored character? And since when you *do* send special characters in a request, they're converted to hex reps?

    12. Re:Uh by Anonymous Coward · · Score: 0

      What pompous horseshit. Most passwords are discovered by leaks or hacks into hosting companies, not by watching over a user's shoulder or listening in on transmissions.

    13. Re:Uh by swillden · · Score: 1

      (Hint: Password complexity rules are a good way to prevent the dumbest of passwords from being used.)

      (Hint: Two-factor authentication is so dramatically more secure that you're far better off implementing it and letting people choose the passwords they want.)

      --
      Note to ACs: I usually delete AC replies without reading them. If you want to talk to me, log in.
    14. Re:Uh by Anonymous Coward · · Score: 1

      One thing to keep in mind that password database breeches give you password hashes (most of the time). So, the attacker still has to (1) figure out the salt for the hash (2) salt and hash potential password and (3) compare the hash to the hash in their breech list. Only when the two hashes match do they have a good idea (after all, the hashes can match because of a hash collision, theoretically) of what the password is.

    15. Re:Uh by Pieroxy · · Score: 2

      It would require probably less than 5 lines of JS to actually allow tabs, backspace and other special characters in a password (or otherwise text) field. And the transmission to the backend has been figured out since decades now. You can actually encode those the way you want, even left up down and right arrows and other special keys that do not have an ASCII counterpart (think caps lock, Scroll Lock and such and even mouse events !)

      The biggest issue here is that you're diverging from a perfectly universal way of entering a password on your site, which your users need to be aware of. Since the available mind-time for such bullshit is lower than ever these days, I'd say good luck.

    16. Re:Uh by Anonymous Coward · · Score: 1

      > Force password changes? Users change from Password1 to Password2. You'll be able to guess their password 5 years from now.

      That is why I append a 4 digit to the passphrase, of the format MMYY, of when the password expires as a mnemonic for when it expires.

      Your crappy "password1" becomes "password0817"

      Good luck guessing the first part -- the pass phrase, along with the second part -- when it expires.

      > The problem isn't password policies;

      Incorrect. I've seen sites where they had a maximum password length, usually like 8 characters. Seriously, WTF. You are _intentionally_ making your passwords insecure???

      My internet banking site makes you have an 8 character password, using numbers and uppercase letters. You can't have 7 characters, you can't have 9. It has to be 8. Furthermore, your "username" is 6 digits, no more, no less.

    17. Re:Uh by EzInKy · · Score: 1

      Have you considered changing banks?

      --
      Time is what keeps everything from happening all at once.
    18. Re:Uh by Bearhouse · · Score: 2

      Yes, you could do it on a PR1ME... could also embed backspaces in messages to other users which would then crawl backwards across the screen deleting themselves...ah...the 1980s

      https://en.wikipedia.org/wiki/...

    19. Re:Uh by Barefoot+Monkey · · Score: 1

      Didn't we just have a (absolutely stupid) story about how password complexity rules are bad?

      No, we didn't. We got a story about how one particular guide to making strong passwords wasn't good advice after all, according to the author. That advice guided people to making short passwords that included a capital, a lowercase letter, a digit and a special character, or variations thereof, and the author conceded that allowing long multi-word passwords is actually stronger than the short obfuscated ones that he recommended many years ago.

      This story is that some sites allow you to give yourself a ridiculously weak password (for example, a single lowercase letter).

      Both stories agree that password complexity rules are good.

    20. Re:Uh by Zumbs · · Score: 2, Informative

      To clarify, the author of complex password policies that have lasted 15+ years had regret for one reason; the rules were too complex for users. In other words, he underestimated just how stupid and ignorant the masses are.

      Force complex passwords? Users write them down. Every time. And "hide" them in the same stupid place.

      I'm registered at more than 50 sites (including work). How do you expect a sane person to remember that number of reasonably strong passwords? And change them at regular intervals?

      My point is that the strong password system may work well if you have a small number of passwords, but once the number of passwords increase beyond maybe a handful, the password system breaks. The problem is not stupid users; the problem is the notion of requiring users to remember many passwords. Something better is sorely needed.

      --
      The truth may be out there, but lies are inside your head
    21. Re:Uh by TheRaven64 · · Score: 2

      On the Psion Series 3, it was possible to set a password containing arbitrary ASCII characters. Unfortunately, I discovered after doing this, it wasn't possible to enter any of the special characters from the login screen...

      --
      I am TheRaven on Soylent News
    22. Re:Uh by TheRaven64 · · Score: 1

      There are some advantages in having a bank with known weak security practices. If your account is ever compromised, it's really easy to argue in court that it's their liability if you can point to a dozen places where they're not following industry best practices.

      --
      I am TheRaven on Soylent News
    23. Re:Uh by EzInKy · · Score: 1

      Guess that depends on how much you value your money. Personally, especially considering how much so many here harp on personally responsiblity, I'd rather not take the chance. Afterall, if you know your banks practices are weak, doesn't the onous fall on you?

      --
      Time is what keeps everything from happening all at once.
    24. Re:Uh by Applehu+Akbar · · Score: 1

      They don't even prevent the dumbest of passwords from being used. password becomes Password1!

      And worse, it becomes more difficult to use those highly random generated passwords available from password managers.

    25. Re:Uh by Applehu+Akbar · · Score: 1

      Corollary: a halfway decent password kept in a secure place is one that the same idiot will lose. I run into this with my IT customers all the time.

    26. Re:Uh by toonces33 · · Score: 1

      Well, there are password managers, of course, but most people can't be bothered. Speaking of which, there are websites which will not allow you to paste a password when you are creating the account, which of course makes it harder to use a password manager.

    27. Re:Uh by Anonymous Coward · · Score: 0

      Force 'complex' passwords and people will use the minimum complexity and write it down ... because they are guaranteed to be unmemorable ...and if the site says it must have at least 10 characters of which one must be uppercase, 1 digit, and one symbol, then that is exactly what all the passwords will have ... and you have invalidated the requirements

      A simple password strength check that actually measures how difficult it would be to crack is far better, people get to choose the password they want, but cannot use a poor password

    28. Re:Uh by jareth-0205 · · Score: 5, Insightful

      Tell me oh massive brained one, how many passwords do you hold in your head? And how many will you still know in a year's time when you haven't used some of them for a while? Also, how many do you think you'll be able to hold in your head when you're 60? 70?

      Passwords are a terrible solution for security, and a solution that we've never as a species had to deal with before. Remembering something that has absolutely no margin for error is hard for squishy brained organisms to do. Password managers are a solution but not exactly a widely spread well-known one, and they have their own issues.

      Also, in your better-than-thou rant you haven't taking into account that worldwide security measures have to *work with stupid people too*. Someone who isn't too clever deserves decent security too, not just you and your Mensa brethren.

    29. Re:Uh by mysidia · · Score: 1

      Tested by creating a new account on each website. Researchers attempted to create passwords with all letters (“aaaaaa”) or numbers (“111111”).

      Traditional Complexity rules ARE bad. Dashlane has a product to sell, and I'm beginning to think dashlane themself is a bad actor, because of the PR promoting whatever they think websites should enforce upon their users to encourage them to use Dashlane's password manager product.

    30. Re:Uh by mysidia · · Score: 2

      It's not a problem to rainbow table that. What kills rainbow tables is strong salting.

    31. Re:Uh by godefroi · · Score: 1

      If your password policies fail to account for reality, then they're the problem. Thus, password policies are the problem, because clearly they don't account for reality.

      --
      Karma: Poor (Mostly affected by lame karma-joke sigs)
    32. Re:Uh by Anonymous Coward · · Score: 1

      ...So, the attacker still has to (1) figure out the salt for the hash...

      You don't need to figure out the salt for the hash. In most password storage methods the salt is stored as a base64 encoded value along with the hashed password. If it isn't stored with the password then you are using 1 common salt for all passwords.

    33. Re:Uh by Ken+D · · Score: 1

      There was one website that I only needed to visit once a year to download my annual tax form, their passwords expired in some time less than one year, so every time I visited my password was expired and needed to be reset.

      So... you need to choose a hard password, that you will use exactly once, a year from when you choose it. ha!

    34. Re: Uh by Anonymous Coward · · Score: 0

      The Holy Login Screen of Antioch eh...

    35. Re:Uh by jareth-0205 · · Score: 2

      I remember reading an article once that was talking about how important your email password (and security of whatever email provider you have) is. It's basically the easy backdoor to almost everything we have online because pretty much everything uses email as a forgot password - so if someone gets into your email they can reset absolutely everything. Scary as fuck... and yet that's one of the ones that many probably don't usually use the crazy-complex passwords for because 'it's just email'.

    36. Re:Uh by Anonymous Coward · · Score: 0

      Obvious solution to improve security: Online random password generator

    37. Re: Uh by Anonymous Coward · · Score: 0

      Password manager. Moron. It's mentioned on slashdot constantly

    38. Re:Uh by green1 · · Score: 2

      This!, I'm getting tired of being told that / isn't a special character, or that my truly random password only had 1 uppercase and 2 numbers, but needs 2 uppercase and 3 numbers, or that my random password can't have the same character twice in a row, etc.

    39. Re: Uh by Anonymous Coward · · Score: 0

      Wait - everyone doesn't have multiple emails set up for various buckets of online registrations as well ?

    40. Re:Uh by green1 · · Score: 2

      I've seen sites where they had a maximum password length, usually like 8 characters. Seriously, WTF. You are _intentionally_ making your passwords insecure???

      I know a specific bank that has the following password rules for their online banking:
      - must be all numeric
      - must be between 4 and 6 digits long

      And this is a BANK!!!!

      Even better was when they sent out a newsletter which included a section on "staying safe online" which specified that you should always use a strong password of greater than 8 characters mixed upper and lowercase with numbers and symbols. I found that pretty ironic from a bank that won't let you use those very passwords on their own site.

    41. Re:Uh by green1 · · Score: 1

      And of course there's the sheer stupidity of storing all your passwords in one place, especially one accessible by multiple devices over the internet....

    42. Re:Uh by green1 · · Score: 1

      Yup, I have a lot of sites where my password is simply to hit the "forgot password" link. There's no point in even trying to remember the password on any site where the password expiry is more frequent than my visits to the site, or where I visit the site less frequently than once every 6 months or so. I'll use a strong password, that I can re-type once when entering it, after that it's gone from my memory.

    43. Re:Uh by tepples · · Score: 1

      brute force is mitigated by account lockout.

      What mitigates the denial of service caused by account lockout?

    44. Re:Uh by tepples · · Score: 1

      If it isn't stored with the password then you are using 1 common salt for all passwords.

      Or the salt and hash are stored in separate tables on physically separate machines.

    45. Re:Uh by hackel · · Score: 1

      Heh, thank you for posting exactly what I intended to when I read this article, even including the "Uh." These people are quite clueless.

    46. Re:Uh by Anonymous Coward · · Score: 0

      Yeah, because what every individual user wants is to have to maintain 500 unique passwords all of which are 16 characters long and are made up of mandatory caps, numbers and symbols and aren't in any language's dictionary.
      It's time to find another way.
      Most people never input passwords anymore anyway. They let their phone/tablet/browser remember for them.
      No. It's not secure to do that. People don't care about security. They care about convenience. If people who care about security want things to change they need to find a better way.

    47. Re: Uh by Anonymous Coward · · Score: 0

      Not pissing off people who know your username.

    48. Re: Uh by Anonymous Coward · · Score: 0

      Only lock for the specific attacking IP. In case of a botnet (multiple IPs failing login), block for all IPs and whitelist real IPs from logs or when they call in to complain.

      Or if it's a social media account, pretend to care but don't actually care.

    49. Re:Uh by 0xdeadbeef · · Score: 1

      Didn't we just have a (absolutely stupid) story about how password complexity rules are bad?

      Why is this idiot at +5?

      I wonder which one he is responsible for.

    50. Re:Uh by Anonymous Coward · · Score: 0

      I think it is that passwords are the wrong answer given that people are people.

    51. Re: Uh by tepples · · Score: 1

      A password stored in a password manager's file is only as strong as the file's master password. And don't password managers that synchronize new or changed passwords between machines cost money?

    52. Re:Uh by tepples · · Score: 1

      so every time I visited my password was expired and needed to be reset.

      Some sites are in fact using passwordless login, which is equivalent to resetting the password on every login.

    53. Re:Uh by tepples · · Score: 1

      Traditional Complexity rules ARE bad.

      Would it be bad to retain the "must contain a letter" rule if the password is long enough? This RC car shop has these rules: 8-15 characters with at least 1 letter and 1 digit, or 16+ characters with at least 1 letter.

    54. Re: Uh by tepples · · Score: 1

      whitelist real IPs from logs or when they call in to complain.

      That's still a DoS against the department that responds to "call in to complain."

    55. Re:Uh by Anonymous Coward · · Score: 0

      And you agreed to binding arbitration when you opened the account, page 217, paragraph 14. There went any hope of getting justice.

    56. Re:Uh by mysidia · · Score: 1

      Would it be bad to retain the "must contain a letter" rule if the password is long enough?

      Yes, because strong passwords don't need to use a letter, And if you estimate entropy PROPERLY, then
      there's no reason for the restriction --- it's just spurious.

    57. Re:Uh by Thad+Boyd · · Score: 1

      (Hint: Password complexity rules are a good way to prevent the dumbest of passwords from being used.)

      Comparing against a blacklist of common passwords, and having a few modest length requirements (and maybe an entropy counter), are good rules. I'm not convinced that complexity rules are.

      I used to work at GoDaddy. The security tutorial I had to pass on my first day actually recommended satisfying the "mixed-case/symbol" requirement by starting with an initial capital letter and ending with an exclamation point -- in other words, reducing a six-character password with mixed case and special characters to exactly the same complexity as a five-character, all-lowercase password.

      That's a dumb damn thing for a security tutorial to advise users to do, but in its defense, even if it didn't, users would just do that anyway.

      Requiring mixed-case passwords may prevent users from using password, but it just means they're going to switch to Password. Require numbers? Then it'll be Password1. Require symbols? Password1!. And so on.

      Meanwhile, complexity rules interfere with actual secure passwords. I once had a site reject K"Nb\:uO` as too weak but accept P@55w0rd without issue. And who can blame it? Just look how secure P@55w0rd is. It's got mixed-case, numbers, and a symbol. Not like K"Nb\:uO`. Look at it. No numbers anywhere!

    58. Re: Uh by Thad+Boyd · · Score: 1

      A password stored in a password manager's file is only as strong as the file's master password.

      Yes, but remembering one secure password is a lot easier than remembering fifty.

      And don't password managers that synchronize new or changed passwords between machines cost money?

      As far as I know, yes, though using a sync solution like Dropbox shouldn't be a problem if your password table is properly encrypted. (Especially if there's a separate keyfile that you don't include on shared storage and instead copy to every client device manually.)

      I sync my password file using an Owncloud instance that's only accessible from my LAN. It's not the most convenient solution, but it's secure enough that I'm confident I'm safe unless a sophisticated attacker targets me personally. And password complexity isn't about protecting you from a sophisticated, targeted attack, it's about protecting you against brute-force attacks.

    59. Re:Uh by Anonymous Coward · · Score: 0

      I'm registered at more than 50 sites (including work).

      I do this fine.
      Perhaps you meant horribly scrambled passwords?

      Strong passwords are long passwords. Period.
      Passwords don't gain anything from randomness. You can't dictionary attack a login screen on any decent site, never mind dictionary attack it if they never had any guess limits.
      An attacker has zero clue about your password. They cannot assume anything.
      They'll dictionary attack the most common passwords then fall back to brute-force after that fails.
      You could literally quote a Lord of the Rings quote for a password right now and it wouldn't even get dictionary-attacked.
      If you replaced one word with a nonsense word, it breaks every dictionary attack out there.
      You could take any syllable-complex word and just swap: ablesyll.

      The only thing that gains from randomness is encryption. And even then it is a specific type of encryption (deniable), not your bog-standard database encryption.
      Database encryption just needs to be a large range and again a long password, it still doesn't require randomness.
      Even with the best supercomputer, huge quotes with a couple mangled words and the weight of an electron as a suffix (or insert other number of interest in your area of work), the sun will burn out well before it is even hinted at.

      I've been doing this since I was 13. Never had issues with hackers, crackers, thieves or whatever else.
      Even on the few sites that have been hacked and DB stolen and clearly in plain sight, none of them since every one is unique.
      Relying on password managers is doing one thing and one thing only, making sure you will never be able to do this, or remember phone numbers, or remember where the keys are, and soon your own family members faces, what day it is, who the leader of your country is and eventually everything.
      Offsetting memory tasks to external devices kills your memory. Literally kills it.
      It's like any muscle in the body. Use it or lose it.

    60. Re:Uh by geekmux · · Score: 1

      What pompous horseshit. Most passwords are discovered by leaks or hacks into hosting companies, not by watching over a user's shoulder or listening in on transmissions.

      Most password databases that are hacked are full of weak passwords that are easily cracked. Those passwords are weak because users are too fucking stupid to remember strong passwords. Systems have to be purposely weakened because of stupid users.

      I stand by my original statement.

    61. Re:Uh by geekmux · · Score: 1

      Tell me oh massive brained one, how many passwords do you hold in your head? And how many will you still know in a year's time when you haven't used some of them for a while? Also, how many do you think you'll be able to hold in your head when you're 60? 70?

      I only know one insanely long passphrase. The rest of my passwords are unknown to me. Yes, that's correct, I don't know them. They are randomly generated and I max out the system limit every time I generate one. I use a password manager. Before they came along, I regularly managed a dozen different systems, so I got used to remembering several usernames and passwords.

      Passwords are a terrible solution for security, and a solution that we've never as a species had to deal with before. Remembering something that has absolutely no margin for error is hard for squishy brained organisms to do.

      OK, let's just stop with the species bullshit as if comparing our challenges to caveman ancestors is relevant. The concept of a password is decades old. If you're too stupid to protect and value a password, then perhaps you shouldn't be using the computer requiring one.

      Password managers are a solution but not exactly a widely spread well-known one, and they have their own issues.

      Also, in your better-than-thou rant you haven't taking into account that worldwide security measures have to *work with stupid people too*. Someone who isn't too clever deserves decent security too, not just you and your Mensa brethren.

      Password managers are a solution. If users are too ignorant to understand the value of them, it only reinforces my original statement. Yes, we do have to deal with a lot of passwords these days. We also have multi-factor authentication solutions available. Unfortunately, that would require extra effort, and users are also lazy, and still consistently undervalue the benefits of decent password security. They prefer to learn the hard way.

      There are two teachers in life. Wisdom is usually gained through education. Experience is usually gained through doing it the fucking hard way. 25 years of IT experience has shown me the latter prevails when it comes to users and security, and hasn't changed in decades.

    62. Re:Uh by viperidaenz · · Score: 1

      Send the user an email with an unlock link in it.

    63. Re:Uh by geekmux · · Score: 1

      If your password policies fail to account for reality, then they're the problem. Thus, password policies are the problem, because clearly they don't account for reality.

      The reality is privacy and security is often compromised because of shitty passwords. This fact is broadcast almost every single day when we hear of everything from stolen celebrity pictures to theft of IP.

      The average user takes the time and effort to lock doors and set alarms to prevent their house or car from being broken into, but then uses the same shitty password across all banks and social media, and ignores all advice to the contrary.

      I'd say the problem is stupidly obvious, and was summarized in my original statement.

    64. Re:Uh by syntotic · · Score: 0

      Africans boasting they can fish your PIN in the convenience store line, then going flat faced when you input it with your eyes closed, one hand over the other hand and barely moving the input hand within gloves. Try this with a super uber complex unbreakable password under scrutiny by some surveillance security camera in the nothing-will-happen cafeteria cyber coffee free wifi hotspot. How many times did you obviously move your both arms simultaneously to input yet one more shift key while aiming at the upper left corner, then at the... you get the idea? Simple passwords are a must. I ve forgotten sites because my ruled in password was impossible to remember without thinking it or to input with discrete keystrokes without watching the keyword. The simpler, the longer I do not have to change it and the better security I get. My FB and amazon and bank passwords have been a success, not so the google ones, to the point several sites I rather ASK to reset password every time I want to log in than actually have MY OWN password in use.

  2. Recent NIST recommendations? by Anonymous Coward · · Score: 0

    Glad to see that TFA gives no bonus to sites favoring hard-to-remember password with stupid special characters combinations...

  3. Password character requirements are a bad thing by Anonymous Coward · · Score: 1

    Passwords need to not be a single dictionary word or name followed by a one followed by an exclamation mark with the first letter capitalized. The really stupid thing to do is issue requirements for including upper case, lower case, numbers, special symbols, etc. and then often within a lower or upper limit and then with symbol restrictions. If a password can't be memorized, it's useless. The only thing that should be mandatory is that the password be long (passphrase), complicated, or both, and perhaps that it fails the "will a computer easily hack it?" test. There's no reason that websites can't do a check for this instead of requiring a password like "Biscuit1!" instead of "biscuit1". I can't tell you how many times I've been forced to create a less useful password on a website because the password security requirements wouldn't let me use any of the passwords I normally use.

    1. Re:Password character requirements are a bad thing by Nartie · · Score: 1
      I once tried to create a password on a site with some very complicated password rules. it rejected all of my 30 character random passwords no matter what I did to the generator. So I gave up and called tech support. It turns out that there are two sets of complicated password rules, and the one published on the web site is very wrong. For example, you had to use a special character. The web site helpfully listed half a dozen, but only three of them worked. My final password was only marginally better than Password, but the site accepted it so I gave up.

      I think the site was the post office, but my memory may be wrong.

  4. U2F to the rescue! by icknay · · Score: 4, Informative

    If you really want it locked down, U2F (2FA device standard) is the way to go. Currently only supported by technically leading sites: google, facebook, github, but jeez it's such a huge improvement over passwords or password managers. One neat side effect of U2F is that with it in place, the password can be super simple, since with U2F the password is not very important. See the U2F FAQ: https://medium.com/@nparlante/...

    1. Re:U2F to the rescue! by sn0wflake · · Score: 0

      I was unaware that Github had 2FA so I set it up ASAP after reading your comment. Now I'm wondering why Github didn't offer 2FA when I created my account.

    2. Re:U2F to the rescue! by Average · · Score: 2

      U2F really does whip the proverbial llama's ass. I wouldn't say, though, that your password is 'not very important'.... your password is still your second factor for a lost/stolen U2F key.

      It is slowly gaining market share. One major financial firm (Vanguard mutual funds/brokerage) has enabled U2F logins, hopefully more to follow.

    3. Re:U2F to the rescue! by Anonymous Coward · · Score: 0

      While Google supports U2F with Chrome, that's it. Even with the Firefox addon "U2F Support Add-on" it fails, as Google checks the useragent. When I altered that to represent Chrome, it failed with an error. It appears Google is using some sort of homebrewed variant incompatible with UNIVERSAL2F. Really disappointing.

    4. Re:U2F to the rescue! by Anonymous Coward · · Score: 0

      So I can use U2F on the phone and tablet devices I use to connect to most of the sites I do business with through their apps?

    5. Re:U2F to the rescue! by tepples · · Score: 1

      Now I'm wondering why Github didn't offer 2FA when I created my account.

      Probably because you still need to generate a password in order to push.

  5. Passwords not usually the only way in by Anonymous Coward · · Score: 2, Informative

    Many websites have good password policies - however, too many of them have entirely vulnerable account/password recovery systems.

    I am reminded of this story about a clever attacker who convinced GoDaddy to let them into the victim's account by means of the last four digits of a credit card number provided over the phone by PayPal's recovery process: https://medium.com/@N/how-i-lost-my-50-000-twitter-username-24eb09e026dd

    Securing a site against password-based attacks is a solved problem. Figuring out what to do when people forget their passwords is still hard.

  6. Don't make netflix fuck with the passwords! by Anonymous Coward · · Score: 0

    I like being able to use fuck for my netflix password. Why the fuck shouldn't I be able to?

    1. Re: Don't make netflix fuck with the passwords! by Anonymous Coward · · Score: 0

      f!ck will bypass that filter and is a stronger password. It's win-win.

    2. Re: Don't make netflix fuck with the passwords! by Anonymous Coward · · Score: 0

      A computer with a powerful GPU can brute force up to 8 digits pretty much instantaneously. Sure, the data has to be breeched first, but that happens like every month...

  7. Worst that can happen by esperto · · Score: 2
    The worst that can happen if some "bad guy" finds out someones netflix account is to make a mess on episodes that were seen/not seen.

    If said someone reuses their password across sites, it can be real bad, but password formation rules are useless against that type of bad password management, you can have the strongest password ever create by man, if you use the same across all your accounts and one dumb webmaster decides to save password as plain text and get invaded, you are fucked the same way!
    So children, use password managers, you can use the most simple of the passwords for your logins (albeit with a manager that would be dumb), as long as you use a different one for each.

  8. Stupid Admins by Anonymous Coward · · Score: 5, Insightful

    You can rant about stupid users all you want, they are the users you have. If you have rules that are not reasonably executable by the average user, then your rule is stupid.

    1. Re:Stupid Admins by geekmux · · Score: 1

      You can rant about stupid users all you want, they are the users you have. If you have rules that are not reasonably executable by the average user, then your rule is stupid.

      Cars have seat belts, and yet there are drivers that fail to use them. If drivers are too fucking stupid to protect themselves and understand the value of a seat belt, then they get what they deserve.

      Computers have passwords, and yet there are users who fail to protect them. If users are too fucking stupid to protect themselves and understand the value of protecting their identity online, then they get what they deserve.

  9. It's the end of the world! by Hentes · · Score: 1

    How could I ever avoid using 'a' as a password without a dozen BS rules that are different on every fucking site?

  10. dashlane who by gavron · · Score: 1

    Sites "failed Dashlane's tests." Good for them. Recent analysis by real cryptographers shows that password rules are worse than no rules.

    And now we have "Dashlane", a nobody who wants to "grade" sites on their "password creation policies."

    https://xkcd.com/936/

    Bye Dashlane and stop it with your self-serving PR memos. You are a disservice to oxygen-breathing password-users.

    E

  11. buck the trend by swell · · Score: 2

    I have a two character password for one important account. It wasn't important 15 years ago when I created it, but grew in value. Perhaps I should change it, but then I'd be among the millions of others using this service with 8+ character passwords. I'm pretty sure that if a hacker looked at my 2 character password, she would just assume that it was a fragment of some code.

    "GoDaddy emerged as the only consumer website with a perfect score" - I hope they've improved; for years they consistently locked me out of my account, requiring calls to tech support. There is a practical limit to the number of obscure requirements for account access. Other companies require phone confirmation (I won't give them my phone #), email or text confirmation, etc. Is it necessary or simply a means to gather more marketable information about users?

    Then there are companies who insist that your username or password is incorrect. Yes, the one you've been using all along. You have to go and create a new one (again, wait for a code via email). Then, when you use the same password, the system says you are not allowed to use the same password (it knew you had the correct password all along!). Somewhere behind the scenes is an Eichmann who delights in torturing users.

    --
    ...omphaloskepsis often...
    1. Re:buck the trend by Anonymous Coward · · Score: 0

      A brute-force attack on a low character count password is quick. Quick enough to try, even if someone doesn't expect to find any passwords of that length. The idea that a password could be assumed to be a fragment of code is misconceived, as password hash files contain only password hashes,not code.

      Please, change your password. It will do you good in the long run, and possibly help you avoid a devastating breach.

    2. Re:buck the trend by apoc.famine · · Score: 1

      It's quick enough to try, but does anyone? Human nature being what it is, I assume the people designing brute force attacks design for the common password requirements prevalent in the area that they're trying to force. I know that if I was designing a brute force attack, I'd probably start with 6 characters, because almost nobody allows less than that now.

      --
      Velociraptor = Distiraptor / Timeraptor
    3. Re:buck the trend by Anonymous Coward · · Score: 0

      A quick test, using an Intel i5 CPU shows that for Windows hashes, I can crack the entire character space for 4 characters and under within a minute. One minute.

      If someone was attempting to break into as many accounts as possible, they are going to make sure that they don't miss any passwords. Of course, your point about checking the requirements of the site is valid, but doesn't that mean your two-character password meets the requirements, and should therefore be checked?

  12. Usernames and e-mail adresses. by jondeanmack · · Score: 0

    How about websites adjust their software such that e-mail addresses are not required for registration nor use! Somebody would think that the internet designed to survive nuclear attack would ensure that one website doesn't need to rely on another site for e-mail service!

    1. Re: Usernames and e-mail adresses. by oobayly · · Score: 1

      We tried that with our website, and it was a nightmare. Our customers are some of the most technologically illiterate people I've come across (and proudly so), yet they still want to use our website to find trade vehicles to buy.

      We had some people who tried clicking the "what is my username link" and then getting confused about what email address to your in (hint, it's probably your work email address). Plus, they then would have had to reset their password too...

      In the end we removed usernames and required the email address to log in. It's slightly better, plus we know that a password reset request *should* go to the right place.

      The other problem is reading comprehension. They still don't click on the activation link that contains the UUID, instead they scroll down and click on the manual link that specially says "use this only if the link above didn't work". I've gotten several people inside and outside our office to read the emails, and bar some padding to separate the paragraphs nobody has come up with more concise instructions.

      Yet almost every one of these people will have a Facebook account, and they can manage that, but with us the first thing they do is call their broker and say "I can't log in, what's my email address and password", and then get grumpy because they have to reset their password because it's encrypted (hashed, but they seem to understand the with encryption).

    2. Re:Usernames and e-mail adresses. by tepples · · Score: 1

      How about websites adjust their software such that e-mail addresses are not required for registration nor use!

      Are you recommending use of a mobile phone number capable of receiving SMS as a substitute for an e-mail address? If not, then through what other mechanism would a user recover a forgotten password?

    3. Re: Usernames and e-mail adresses. by tepples · · Score: 1

      The other problem is reading comprehension. They still don't click on the activation link that contains the UUID

      That could be due to following security guidance in articles like this one:

      "Never follow links in e-mails because that confirms to spammers that you read their message."
      "Never follow links in e-mails because they could lead to phishing sites on typosquatted domains instead of the real thing."

  13. Don't care about your site you precious snowflake by FeelGood314 · · Score: 3, Insightful

    Seriously fuck you Help Net Security. I really don't care about the security of most sites enough to have to memorize a unique password for them and most sites actually do understand this. Further if it is a site that I do care about the security I want to be able a secure password that I can remember. TR0b@dor is hard as hell for me to remember and will likely be in the first million passwords a cracking program will try. Second for an online attack you need enough entropy to stop an attacker who is rate limited. So 2^30 is likely strong enough (that's 3 common English words). If someone gets your salted hashed password file you are going to need 2^60 bits of entropy. 6 English words. Making be choose a password that is anywhere between those two lengths is either a waste of my time or insufficient security.

  14. Needless complexity reduces security by Tony+Isaac · · Score: 2

    I've lost track of how many passwords I have on various sites. Each site has its own rules, that conflict with each other. There's no way I can remember them all. So what do I do? I send myself emails with password hints for each site, or save a list in a password-protected document, or let Chrome remember it, or write them on a sticky note.. If somebody figures out a way to hack Chrome's password vault, a LOT of people are in trouble! Somebody DID hack LastPass.

    When building security is very tight, and there's a need for a plumber to come and go, what do they do...somebody props open a door, of course! Passwords are no different. If you make them too hard, people take measures to remember them--measures that make them less secure than if the rules weren't there in the first place!

    1. Re:Needless complexity reduces security by apoc.famine · · Score: 1

      I've got a password protected file with login and password hints for a couple dozen sites in it. Not the frequent ones like amazon or my banking, the infrequent ones that I need like once every year or so. Car insurance website with the bizarrely shitty requirements where only some subset of the symbols are allowed, and some other crazy requirements. Student loan login so I can get my interest statement for taxes. I find that it's invaluable when I go to log in each year, as if I can at least remember the login, I can reset the password if I have to. Trying to reset it without the login is more likely to trigger the stupid security questions, which are random answers also in that document.

      --
      Velociraptor = Distiraptor / Timeraptor
  15. Good by Anonymous Coward · · Score: 0

    I hate it when websites take it in their own hands to tell me how good or bad my password is. 1234 is perfectly valid for most of them. They provide nothing of value and i won't give them anything more than that. For those that do matter (i give them CC info for example) I choose my own thing. No need for a dumb mid-level manager to tell me what's good and what's bad.

  16. Another leave out the details post by Anonymous Coward · · Score: 0

    for example AWS allows the account owner to set the password policy strength:
    Options are:
    Minimum password length: [ ]
    [ ] Require at least one uppercase letter
    [ ] Require at least one lowercase letter
    [ ] Require at least one number
    [ ] Require at least one non-alphanumeric character
    [ ] Allow users to change their own password
    [ ] Enable password expiration
    Password expiration period (in days): [ ]
    [ ] Prevent password reuse
    Number of passwords to remember: [ ]
    [ ] Password expiration requires administrator reset

    And at user creation you can assign an MFA device. So i'm thinking these guys didn't test a thing.

    1. Re:Another leave out the details post by green1 · · Score: 1

      for example AWS allows the account owner to set the password policy strength:
      Options are:
      Minimum password length: [8]
      [X] Require at least one uppercase letter
      [X] Require at least one lowercase letter
      [X] Require at least one number
      [X] Require at least one non-alphanumeric character
      [ ] Allow users to change their own password
      [X] Enable password expiration
      Password expiration period (in days): [1]
      [X] Prevent password reuse
      Number of passwords to remember: [32000]
      [X] Password expiration requires administrator reset

      *sigh* if only sites didn't chose options like this so I can read their useless blog...

  17. gratuitous Spaceballs clip by meglon · · Score: 1
    --
    Fascism: An authoritarian and nationalistic right-wing system of government and social organization. See also: NAZI's
  18. Why is requiring alpha numeric important? by bussdriver · · Score: 3, Insightful

    Requiring UPPERCASE doubles the space while 0-9 only adds 10 digits. It would be better to require mixed CASE than to require digits.

    Also, requiring a symbol then allowing ANY symbol would expand the space to typical symbols people use... probably only about 8 symbols cover 90% of passwords. A full brute force would expand to nearly all of unicode! Emjoii included.

    Requiring a SPACE might only add 1 digit but it would hint to people to add a whole WORD and I bet you get more in practice than requiring digits.

    Strength tests should include the domain name because I've seen some lists where the domain name was used. My own investigating found people will use dates, names, initials, their PIN #, phone, even part of their email address. That kind of easily guessed stuff does not show up in these checkers OR in the stats gathered from break ins. Sites really should not create an account password UNTIL you enter all your account information. The session ID is good enough for tracking logins it surely is good enough to setup an account before creating a password and account name. Everybody does it backwards.

    1. Re:Why is requiring alpha numeric important? by Anonymous Coward · · Score: 0

      Require a digit and a huge chunk of people will just append '1'.
      Require a symbol and a huge chunk of people will just append '!'.

      Requiring a space is an interesting idea.

    2. Re:Why is requiring alpha numeric important? by Anonymous Coward · · Score: 0

      ALLOWING doubles the space, requiring reduces the space. If you require 1 upper, 1 lower, 1 digit your combination calc looks like
      10 * 26 * 26 * 62 * 62 * ...

      If you just allow from those 3 sets, then it looks like

      62 * 62 * 62 * 62 * ...

    3. Re:Why is requiring alpha numeric important? by Anonymous Coward · · Score: 0

      Requiring UPPERCASE doubles the space while 0-9 only adds 10 digits. It would be better to require mixed CASE than to require digits.

      And this is what I hate.

      Requiring uppercase requires use of the shift key. The shift key is a keypress. It also requires some slightly uncomfortable finger acrobatics.

      A 10-character all-lowercase password is over twice as secure than a 8-character one with two uppercase characters. Yet both require ten keypresses.

      8 log2 52 = 45,6 bits of entropy
      10 log2 26 = 47,0 bits of entropy

      The exponent matters more than the base, as you can see.

      My password has no uppercase characters in it, nor digits. It is, however, plenty long. An unfortunate amount of web sites reject it because of the fact.

      Don't demand mixed-case passwords. Demand long passwords.

    4. Re:Why is requiring alpha numeric important? by bussdriver · · Score: 1

      I have no trouble pressing shift. What I have trouble doing along with all other humans is having a perfect memory for tons of strong passwords. (I use a keychain so I rarely deal with passwords other than a few.)

      My exploration of user patterns from years of looking at people I was admin for, is that capitalization is how they mostly handle the caps requirement or they go full capslock. Programmers love camel case.

      In terms of brute force space, it makes sense to force a larger character set while it also makes sense to require a good length too; we try to cut down on length because of human memory. It is easier to remember case than a few more symbols. Thing is what I've observed is likely representative of what a real study would show-- people are NOT even remotely random and fail to use the symbol space and prefer to go as short as possible. A smarter more statistical attack will do far better than brute force, adding caps or symbols won't grow the space by much in practice - but adding to the length a few characters probably doesn't practically help much more when they are NOT random. More dictionary word combinations are possible with 10 characters -- the increase in space actually allows for more common patterns. Such as using two 5 letter words from the dictionary or a 10 letter word - it gives the attacker more (not as much as it increases the combinations obviously.)

      I've seen 123456789 passwords too. my checker I wrote detected keyboard sequences as well as appending ! or 1... doubling a bad password twice is another one. The clever ones going unnoticed were quite stupid: password@domain.com by being cleverly unique but extremely predictable. Humans are strong at hacking around the rules to save energy. We probably only evolved to save on labor.

      Like I said, simply requiring the use of a SPACE character should do more than the few symbols people use when forced to do so. It would encourage the adding of words maybe phrases which would do more than adding 2 characters to the length. Then at least we have english's 60,000 vocab of which normal people maybe only know about 20,000 words but only using 2-4 words turns the problem into a 20,000 symbol set of 3 characters. As you said, the exponent matters more than the base... So then if any trend happens where we need to protect against phrases we then need password checkers to detect phrase passwords and require more than just 3 words. Emjoii use for cell users would probably end up with poop replacing or being next to ! Requiring a new password periodically results in a numbering scheme.

      So, the best policy would be require great length; allow all of unicode and phrase detection for imposing more rules on minimum wordcount and a pop culture phrase book.

      Me, I run new passwords against the dictionary of top 100 bad passwords after regex stripping away common hacks along with banning @ and all user account info and company info and many date formats. Sometimes I throw it at my hashed 1900 Webster dictionary then I don't allow any phrase list of dictionary words. It can frustrate new users something awful since it is hard to specify all the rules and they just want a few simple rules to hack around to get their horrible password accepted.

  19. In other news... by Chrisq · · Score: 1

    Shock as sites designed for professional use leave responsibility for choosing secure password to their clients!

  20. Re:Don't care about your site you precious snowfla by tlhIngan · · Score: 3, Interesting

    Seriously fuck you Help Net Security. I really don't care about the security of most sites enough to have to memorize a unique password for them and most sites actually do understand this. Further if it is a site that I do care about the security I want to be able a secure password that I can remember. TR0b@dor is hard as hell for me to remember and will likely be in the first million passwords a cracking program will try. Second for an online attack you need enough entropy to stop an attacker who is rate limited. So 2^30 is likely strong enough (that's 3 common English words). If someone gets your salted hashed password file you are going to need 2^60 bits of entropy. 6 English words. Making be choose a password that is anywhere between those two lengths is either a waste of my time or insufficient security.

    Exactly.

    Your website may not be important to me, so I won't give it a very important password. It may be important to you, but not to me. Especially if you insist on a username and password to do the most basic things.

    You want me to log in to download your free software? Sure, I'll create an account - with a wimpy password. I don't care if that software is your heart and soul and you missed your mother's funeral to release it on time. I just want the file.

    You want me to log in to comment on your article? Well, ditto. Same for forums as well.

    Hell, I fully expect those sites to be hacked, so why use a strong password? Might as well just make it "password" and be done with it - if someone's downloaded the password file then they have all the time in the world to crack it. I might as well assume your site has vulnerabilities that make it easy to steal the password file.

    Oh yeah, my Paypal, Amazon and bank passwords? They're nice and secure.

  21. Re:Don't care about your site you precious snowfla by Anonymous Coward · · Score: 0

    Maybe you should learn first how to correctly express a measure of entropy, and then we can discuss whether n user-chosen words have this entropy. (Little hint: They don't.)

  22. THE solution: expiry depends on complexity by Gunstick · · Score: 5, Interesting

    Hi

    you chose a password, there is a calculation performed how long a brute force/dictionary attack will take.
    Your password will expire after this time.
    Calculate the time using this calculator (take the botnet time): https://password.kaspersky.com...

    thisisanicepassword => 3 days
    this is a nice password => 40 years (maybe maximize on a top limit)
    12345678 => 1 second
    one two three four => 3 years
    correcthorsebatterystaple => 5 years (hmm, maybe they should add that to an exception list)
    h4Z7p8d0 => 51 seconds
    h4Z7p8d0x3 => 2 hours
    h4Z7p8d0x3w1 => 6 days
    h4Z7p8d0x3w1bd => 2 years

    --
    Atari rules... ermm... ruled.
  23. Re:Don't care about your site you precious snowfla by Gunstick · · Score: 1

    Maybe his usage of 'entropy' is not correct, but there are easy ways to have long and remeberable passwords.
    See diceware.
    5 words from a dictionary of 7776 is 7776^5
    That's quite a lot.
    Equivalent to a 14 lowercase letter password. But instead of memorizing 14 items, you only need to memorize 5

    --
    Atari rules... ermm... ruled.
  24. Correlation? Causation? by houghi · · Score: 2

    The most popular sites provide the least guidance when it comes to secure password policies.

    What if they are the most popular sites BECAUSE they are the least secure?
    Ease of use is pretty important and people would rather use a less secure, but easy platform than a more secure and complicate platform.

    Just recently somebody told me they did not use any protection on their phone, because it was to much trouble to use.
    They added fingerprint readers because people are too lazy to type in a 4 pin code.

    --
    Don't fight for your country, if your country does not fight for you.
    1. Re:Correlation? Causation? by Anonymous Coward · · Score: 0

      too lazy to type in a 4 [digit] pin code

      This is not just about being lazy. Imagine you were in an accident and need to urgently call a friend to come pick you up. (There are situations where 911 isn't the appropriate emergency number.) But you also had a contusion and don't remember the PIN. What do you do? Make three attempts so your phone gets locked forever?
      Yesterday I rebooted my phone. Only after I had pressed the button did I notice that I don't remember the exact order of the digits in my PIN. I was really lucky to still have the SIM card holder around that they had sent me by snailmail with the PIN printed on it. Horrible idea securitywise, but it saved me. When you're stuck in a shit town with the only internet connection being over your phone, you don't want to take any chances. I'm really happy my phone has a fingerprint reader so I don't need to remember the PIN all the time.
      Also, you will get older. Never trust your memory alone.

  25. More important to ALLOW strong passwords by arobatino · · Score: 2

    It's more important that a site allow strong passwords, by having long or no length limit, and no character restrictions. Amazon, Google, and LinkedIn, for example, may allow weak passwords, but unlike many sites, they also allow very strong passwords (no length or character restrictions AFAIK). If someone doesn't want a strong password (for example if they insist on trying to remember dozens of different passwords instead of using a password manager) forcing one will just make them write it on a sticky pad. Which may or may not be OK, depending on whether it's a secure environment.

  26. Re:Don't care about your site you precious snowfla by Anonymous Coward · · Score: 0

    take a deep breath, ass stain.

  27. ATM fees, postage, and money order fees by tepples · · Score: 2

    Have you considered changing banks?

    Yes. But when only one bank has ATMs within cycling distance, that makes every other bank much more expensive: withdrawing cash costs ATM fees, depositing checks costs postage, and depositing cash costs postage plus money order fees. In the city where and years when I attended college, there was only one bank.

  28. Re:Don't care about your site you precious snowfla by green1 · · Score: 1

    This is the biggest thing about security of websites. If your site doesn't handle my money, or my real life reputation, then it doesn't need a secure password.
    Imagine if every single store you ever visited required you to sign up with all your personal details and carry around a user card before you could walk in the door? Sure you'd put up with it for your favourite grocery store, the local hardware store, and maybe 1-2 others, but you'd quickly say enough is enough and just avoid the mall. The web is increasingly like this, every site wants all your personal details, and for you to remember a username and password just for their site, a site you may not ever visit again. It's insane.

    If a streaming video provider doesn't need a strong password, that's fine, the worst case is someone else watches a TV show instead of me, oh the horror! If my bank doesn't require a strong password, that's a problem, but for the rest, give it a rest!

  29. That requires JavaScript by tepples · · Score: 1

    There are two kinds of web-based random string generators: those that generate the password on the server and therefore allow the operator of the site to see every string that is generated, and those that generate the password on the client and therefore require the user to add the site to the browser's whitelist for running JavaScript.

  30. Twitter's 2FA is expensive by tepples · · Score: 1

    Hint: Two-factor authentication is so dramatically more secure that you're far better off implementing it

    Unless it's Twitter, which allows only the login method that's most expensive per use for many U.S. users.

    • YubiKey and other FIDO U2F devices: Not supported
    • Google Authenticator and other TOTP apps: Not supported
    • One-time random number through voice call: Not supported. This leaves out users of landlines or wireless home phone service.
    • One-time random number through SMS: Supported, but standard messaging and data rates apply. Cellular carriers in the United States tend to charge pay-as-you-go subscribers 10 cents per sent message and 10 cents per received message.
  31. using nothing but the lowercase letter "a" by cyberfunkr · · Score: 1

    The testing criteria is flawed.

    If websites did their security right, there is no issue with it just being "a".

    Once you salt, pepper, and hash that letter it becomes just as tricky to hack as "h&t3)__ner!1" -- 64 digits of random looking hex.

    A real indicator of a website's bad password storage is if there is a character limit. If they only allow password that are 12 characters or fewer, then you know they are saving the password in a recoverable format. You should also try doing a "Forgot Password"; if they can email your password back in plain text, I wouldn't trust that site with the recipe for a peanut butter and jelly sandwich.

  32. Don't make Netflix mess with my pants by tepples · · Score: 1

    Sure, the data has to be breeched first

    Why does the data need to start wearing pants?

  33. Usernames aren't supposed to be secret by tepples · · Score: 1

    That'd be difficult on sites that use a username as part of a user's public identity. For example, someone who reads the comments of all stories on the front page of Slashdot can see the usernames of all logged-in users who have commented on those stories.

  34. Re:Don't care about your site you precious snowfla by Gilgaron · · Score: 1

    Yeah at least when sites have custom logins the profile stays there, more and more are asking to log in via social media or gmail account... why would I want to link anymore information about be on the internet than absolutely necessary?

  35. Computing emtropy "properly" by tepples · · Score: 1

    And if you estimate entropy PROPERLY

    What's "properly"? Kolmogorov complexity isn't tractable to compute.

    1. Re:Computing emtropy "properly" by The+Phantom+Mensch · · Score: 1

      Does Kolmogorov complexity adequately describe what users actually choose as passwords when "complex" password rules are imposed? Most people will do something easy to remember involving pet and kid names mixed with birth dates and a few obvious special character substitutions, or variations on that theme. This should be your expectation when attempting to estimate the entropy in your passwords.

  36. Merge conflicts; keyfiles on mobile by tepples · · Score: 1

    using a sync solution like Dropbox shouldn't be a problem

    What I fear is that I would add two passwords on separate machines, and then the ownCloud or Dropbox client gets a merge conflict when it sees that both versions of the password vault file have changed.

    Especially if there's a separate keyfile that you don't include on shared storage and instead copy to every client device manually.

    How is that done on mobile, especially when iOS didn't have a user-accessible file system last I checked?

  37. How to identify websites with weak passwords by Anonymous Coward · · Score: 0

    See: http://cubicspot.blogspot.com/2012/08/how-to-identify-websites-with-weak.html

  38. Bullshit by allo · · Score: 1

    8+ - Good
    Alphanumeric required - Bad, you allow the attacker to skip testing all alpha-only / numeric-only passwords.
    Password strenght meter - We all know they don't work
    Logins cannot be brute forces - OK
    2-FA auth - doesn't have much to do with passwords

  39. Passwordless login, except to your e-mail by tepples · · Score: 1

    That's fine for passwords that don't affect the path to e-mail. In fact, some sites embrace passwordless login through one-time tokens sent through e-mail. But it wouldn't work for the password to the user's Internet connection (PPPoE, RADIUS, subscription hotspot with a captive portal, etc.) or to the user's e-mail itself.

    Nor does it work if your site has a lot of users such as jondeanmack, who expects to be able to register without providing a means of password recovery.

  40. Re:Don't care about your site you precious snowfla by Anonymous Coward · · Score: 0

    I wouldn't bother.

    In their snide remark about entropy above about the calculation being incorrect, the AC has merely just declared that they don't even have a vocabulary of 1000 words (~2^10).