Quote from: jackbox on March 18, 2015, 05:28:27 PM

Quote from: molecular on March 18, 2015, 05:21:34 PM

Quote from: Erdogan on March 18, 2015, 10:30:16 AM

Quote from: molecular on March 18, 2015, 06:55:03 AM

Quote from: BitcoinIsLiberty on March 17, 2015, 09:26:35 PM

Quote from: molecular on March 17, 2015, 08:36:58 PM

Quote from: RoadStress on March 16, 2015, 11:26:35 PM

Question: How many different passphrases does Trezor support?

over 9000

no, wait

all of them!

How hard are the passphrases to crack compared to bip38? I know there is some type of key lengthening....

I'm mostly curious how hard short passwords are to crack if you have the seed.

That'd be interesting to know, yes.

I know one thing: on each bruteforce attempt you have to query the utxo set to see if there are funds, because there's no other way to know wether or not you have found the correct passphrase. In fact you also have to check wether there address has been used in the past (and now has 0 balance so it's not in the utxo set any more) and/or query a couple more addresses in case the first one was left untouched. That alone should slow things down quite a bit.

The passphrase has to be entered via the pc, and there is no concealment from a rogue pc pin-cod style.

That's a different attack.

The question was how long would it take (in relation to passphrase length) to bruteforce the passphrase if someone finds your paper seed backup.

That would depend on the length and complexity of the password and if anything about it was known to the attacker. With a strong password it could take thousands of years. Once quantum computers are mainstream it could possibly be brute forced within minutes but by then Bitcoin protocol would have changed to a quantum alogorithm.

Let me rephrase the question then: how long would it take on commodity hardware worth 10,000$ to guess an 8-letter (uppercase, lowercase plus 10 special characters) with known seed words?