Topic: Password strength - page 2. (Read 2609 times)

My wallets password is : F4tF0cKM4rkT0oKMyCoIns 

Is that strong enuf to keep me safe ?

None of the above are very secure against a determined and well-funded attacker - not even the last one.
7 to 8 diceware words, on the other hand, is all you need to be very safe for years.

You might be interested in my NoBrainr script, which is a simple example of diceware applied to bitcoin address generation:
 https://bitcointalksearch.org/topic/nobrainr-a-secure-and-transparent-cold-address-generator-in-1024-bytes-308972

There are about a million words in English. Assuming my math is right, here's the number of passwords for x words:
1: 1000000 possibilities
2: 1000000000000
3: 1000000000000000000
4: 1000000000000000000000000
5: 1000000000000000000000000000000
vs a 14 character password of letters, numbers, and common symbols: 67822307284900000000000000
so as long as you've got a good list of words to randomly pick from, you need 5 words to beat the security of a typical password. With real word lists, however, I imagine it being closer to 7 or 8 words.

Seems I've caused some controversy, heh.  Can we at least agree that in order from weakest to strongest password strength, it would be:

• elephant
• 3l3ph4nT
• flying elephants with bow ties
• fLy1ng-3l3ph4nT5_wiTh*b0w.t13$

But it would be almost impossible to commit the last one to memory.

That is a good point. It must be a sign that they don't store it as salted hashes!:D On the other hand password for a web-service need not be as complex as a file encryption pass. Since they can stop brute force pretty easily.

I don't understand why some websites set a limit on the length of passwords, since the passwords are supposed to be stored as salted hashes (of the same finite size).

Many banks are like this as well.

Or worse, Schwab.com has a limit of 8, yes, eight, characters.

Random is overrated.  Its not pure random thats important, but that the combination is unique, and that it hasn't been used and found in an unsecured system.  What XKCD is trying to highlight is that length is better than complexity.  Given two passwords of 10 char, "(wRD9=K-]3" or "Complacent" have the same entropy for a brute force attack.  

However if there is a dictionary attack, then the latter is weak.  But if you'd used the first in a system that has poor password security, it would be just as weak as it would be included in dictionary attacks.  Thats what they are, long lists of all the compromised passwords found before.  Random just gives us a better starting point for being (far) less likely to have been used and included in a dictionary.

Assuming we don't reuse passwords and they aren't exposed through a poor site, a longer memorable password is always preferable.  Memorable passwords help avoid reuse, so help avoid that weakness.  If I use a "random", previously unused sequence of words, thats good enough for 99% of use cases.  Knowing that its unused is difficult of course (unless checking against dictionaries) so needs a bit of thought.

If you want to protect against supercomputers, you need to reduce the opportunity to access the secured system, not worry about password length.  If one can access your system for your high value wallet, you are already exposed to physical attack methods (keylogging, forensic data capture, good 'ol fashion lead pipe threat)

My answer hidden to avoid spoiling the fun for others.  Click "quote" on this post to see my answer.



<sub>Well,

I started by noticing that 7776 is pretty close to 96^2. That would mean that everytime x increases by 1 the exponent on the other side of the equal sign would have to increase by approximately 2.  Since the exponent is 16 on the other side of the equal sign, x must be somewhere around 8 (give or take 1).

How'd I do?</sub>

Microsoft has 16, and it has been like that for years

Well it doesn't hurt to have more probably in most cases 20 is enough tho. Better yet... some websites have a limit on like 24 characters

Please show me the math where it takes 10 to 12 random words from a list of say the diceware list to equal the entropy of 16 random characters.  If you are right and have the math to show it I will give you 100 mBTC.

Don't get excited how about we estimate it like this.  There are 80? 96 characters (upper, lower, number, and symbol) on standard keyboard.

A guestimation game.  Do you think 96^16 >= 7776^10?  If not what x do you think solves this equation 7776^x >= 96^16

For those playing along at home, don't grab a calculator right away.  Just take a guess based on the base and the exponents.

The comic didn't say 12 random characters.  Most people don't use 12 random characters for their password.  So was the problem the example in the comic or your understanding of the example in the comic?

which is useful against a set of random words how?

Here is a quote from from the website of none other than TrueCrypt


Shouldn't the TrueCrypt designers, of all people, have a better grasp of this concept?  

Actually, that XKCD comic is dangerous advice for Joe Average.  It is true only IF those words are chosen completely at random.

But that is not how Joe Average would go about choosing he words. He would go something like "Horse...uhm...Cart...uuhhhm...Galloping...Away!"

Or he would click on "random article" on Wikipedia 4 times.

None of these methods are random enough to create secure pass phrases.

Also, to resist brute force attacks by supercomputers (a realistic threat for high balance bitcoin wallets) you need something like 10-12 words and not 4.

Which is not that much easier to remember than 16 characters.

Just a friendly warning that comic is bullshit. You would need at least 8 random words to have about the same security as 12+ random characters. The entropy maybe higher, but every serious hacker uses rainbow tables, dictionaries or even phrases from common digitalized books...

I use KeePass with the default (20 characters). Should I increase it?

I assume you are talking about https://bitcoin.org/en/secure-your-wallet

Bitcoin.org is not an official website. (it actually says that literally on their about page: https://bitcoin.org/en/about-us )

But since this is a website where many new members come, I can agree with you that it should be adjusted. I would suggest to come up with a better explanation and just e-mail it to them. Perhaps they will change it. Most importantly it should say to keep different passwords for all websites.

I personally advise programs like KeePass and LastPass with 30(+) characters.

So the Bitcoin website itself states:


But as explained fantastically well by XKCD, it's actually not entirely true.  Random characters only make it harder to remember, not to crack.  As pioneers of this revolutionary cryptographic technology, shouldn't we of all people have a better grasp of this concept?  The responsibility is on us to make sure new users understand it correctly, but that won't happen until we get it right ourselves.  This part of the site should ideally be edited to be more accurate.