6 ms·

> It seems at least plausible that he attempted to steal them and is now backpedaling since people aren't buying his malleability story. Or that they lost the

by Blahah 13y ago

> It seems at least plausible that he attempted to steal them and is now backpedaling since people aren't buying his malleability story. Or that they lost the keys to some wallets and have managed to crack one.

sillysaurus3 13y ago
If that were the case, then there's no reason to frame it as if they thought there weren't any bitcoins left in unused company wallets. They could just tell the truth.
- Blahah 13y ago
  That's true, it would actually make them seem less cretinous if they said they had cracked a lost key. I guess the fact that they have to have been so staggeringly incompetent to have 200,000 bitcoins they didn't even know about makes me feel like there must be some other explanation than the one they are offering. Attempted theft is up there.
  - erichurkman 13y ago
    Wouldn't someone publishing that keys can be cracked in realistic timeframes deal a blow to the currency itself? Sure, it would have required tons of computing power, but they still would have cracked the key in several weeks' time.
    - garrettgrimsley 13y ago
      Well not exactly, as I understand it you could attack the passphrase protecting the key. If you remember any details about this passphrase then you can dramatically reduce the strength? (I'm not sure of the word to use here. Someone help me out) of said passphrase. But yes, if someone was able to crack keys in a reasonable amount of time then bitcoin would crash overnight.
      - erichurkman 13y ago
        Ah okay, you're right. And yes, if you knew anything about the key you could drastically reduce the key space to search.
      - dTal 13y ago
        Was the word you were looking for "entropy"?
        garrettgrimsley 13y ago
        That would be it!
  - deftnerd 13y ago
    Karpeles said last winter that they were going to make their keys more secure by breaking them into pieces using a Shamir's secret sharing algorithm. This would allow a key to be more secure because it would take several pieces, but not all of them, to reconstitute the private key. My assumption is that they either made a mistake with the algorithm, or they lost enough chunks of the keys that they can't reconstitute the private key. This could have been as simple as a banks safety-deposit box being inaccessible because it's seized or losing the pieces. If they were stored digitally, it could be as simple as a media problem (Organic dyes in CDR's degrade. USB drives aren't infallible, etc). The reason they got this wallet open is because they probably recovered the private key in some ancient backup they forgot they had from before they split the keys into pieces and deleted what they thought were the only copies of the unsecured keys. This is just my completely unsubstantiated theory, but it seems to match the current facts.
    - olalonde 13y ago
      Mark stated that the missing coins were most likely lost due to transaction malleability. I don't know why he would risk lying for no good reason.
      - apalmer 13y ago
        The 'good' reason would be the 400 Million dollars... Not saying he is or isnt lying, but we cant just throw around the 'there is no good reason for him to say xyz' in this situation...
        olalonde 13y ago
        Yes, that's what I meant: stealing them would have been a good reason to lie but covering up a form of negligence (media failure) with another form of negligence (transaction malleability), which was the theory I was trying to refute, doesn't seem especially useful (hence the "no good reason to lie").
mcintyre1994 13y ago
Am I missing something or does that seem really really implausible http://bitcoin.stackexchange.com/questions/2847/how-long-would-it-take-a-large-computer-to-crack-a-private-key http://bitcoin.stackexchange.com/questions/2847/how-long-wou...
smtddr 13y ago
>>Or that they lost the keys to some wallets and have managed to crack one. No cracking. They just found a wallet.dat somewhere, that's all. If they lost the keys(as in, lost the wallet.dat with the keys in it) and were able to generate a valid private key given only a public bitcoin address... bitcoin is dead, game over, pack it up. Now instead, if we're talking about a passphrase to a wallet.dat...
- saraid216 13y ago
  > If they lost the keys(as in, lost the wallet.dat with the keys in it) and were able to generate a valid private key given only a public bitcoin address... bitcoin is dead, game over, pack it up. Yeah. This wouldn't be a fail on the MTGox league; this would be a "Satoshi Nakamoto and all the cryptographers who took a look at the code failed". Cracking a wallet is supposed to be something that requires the resources of a state-sized entity.
  - FatalLogic 13y ago
    >Cracking a wallet is supposed to be something that requires the resources of a state-sized entity. A galaxy-sized entity, maybe? The only known way to find a private key from a public key is brute force. That's way beyond the abilities of a state, unless they've made a massive breakthrough in quantum computing. http://i.imgur.com/fYFBsqp.jpg http://i.imgur.com/fYFBsqp.jpg
    - nullc 13y ago
      Obpedantry: In theory rho allows you to recover private keys with work-factor 2^128 (times some small constant depending on how much storage you wish to use), which is significantly less than the 2^256 you might expect from "brute force". Still completely infeasible.
      - teraflop 13y ago
        But first you have to obtain the public key by finding a preimage of the SHA-256-hashed address. And SHA-256 (with the full number of rounds) currently has no known preimage attacks that are even marginally better than brute force.

by Blahah 13y ago

sillysaurus3 13y ago

If that were the case, then there's no reason to frame it as if they thought there weren't any bitcoins left in unused company wallets. They could just tell the truth.

Blahah 13y ago
That's true, it would actually make them seem less cretinous if they said they had cracked a lost key. I guess the fact that they have to have been so staggeringly incompetent to have 200,000 bitcoins they didn't even know about makes me feel like there must be some other explanation than the one they are offering. Attempted theft is up there.
- erichurkman 13y ago
  Wouldn't someone publishing that keys can be cracked in realistic timeframes deal a blow to the currency itself? Sure, it would have required tons of computing power, but they still would have cracked the key in several weeks' time.
  - garrettgrimsley 13y ago
    Well not exactly, as I understand it you could attack the passphrase protecting the key. If you remember any details about this passphrase then you can dramatically reduce the strength? (I'm not sure of the word to use here. Someone help me out) of said passphrase. But yes, if someone was able to crack keys in a reasonable amount of time then bitcoin would crash overnight.
    - erichurkman 13y ago
      Ah okay, you're right. And yes, if you knew anything about the key you could drastically reduce the key space to search.
    - dTal 13y ago
      Was the word you were looking for "entropy"?
      - garrettgrimsley 13y ago
        That would be it!
- deftnerd 13y ago
  Karpeles said last winter that they were going to make their keys more secure by breaking them into pieces using a Shamir's secret sharing algorithm. This would allow a key to be more secure because it would take several pieces, but not all of them, to reconstitute the private key. My assumption is that they either made a mistake with the algorithm, or they lost enough chunks of the keys that they can't reconstitute the private key. This could have been as simple as a banks safety-deposit box being inaccessible because it's seized or losing the pieces. If they were stored digitally, it could be as simple as a media problem (Organic dyes in CDR's degrade. USB drives aren't infallible, etc). The reason they got this wallet open is because they probably recovered the private key in some ancient backup they forgot they had from before they split the keys into pieces and deleted what they thought were the only copies of the unsecured keys. This is just my completely unsubstantiated theory, but it seems to match the current facts.
  - olalonde 13y ago
    Mark stated that the missing coins were most likely lost due to transaction malleability. I don't know why he would risk lying for no good reason.
    - apalmer 13y ago
      The 'good' reason would be the 400 Million dollars... Not saying he is or isnt lying, but we cant just throw around the 'there is no good reason for him to say xyz' in this situation...
      - olalonde 13y ago
        Yes, that's what I meant: stealing them would have been a good reason to lie but covering up a form of negligence (media failure) with another form of negligence (transaction malleability), which was the theory I was trying to refute, doesn't seem especially useful (hence the "no good reason to lie").

mcintyre1994 13y ago

Am I missing something or does that seem really really implausible http://bitcoin.stackexchange.com/questions/2847/how-long-would-it-take-a-large-computer-to-crack-a-private-key http://bitcoin.stackexchange.com/questions/2847/how-long-wou...

smtddr 13y ago

>>Or that they lost the keys to some wallets and have managed to crack one. No cracking. They just found a wallet.dat somewhere, that's all. If they lost the keys(as in, lost the wallet.dat with the keys in it) and were able to generate a valid private key given only a public bitcoin address... bitcoin is dead, game over, pack it up. Now instead, if we're talking about a passphrase to a wallet.dat...

saraid216 13y ago
> If they lost the keys(as in, lost the wallet.dat with the keys in it) and were able to generate a valid private key given only a public bitcoin address... bitcoin is dead, game over, pack it up. Yeah. This wouldn't be a fail on the MTGox league; this would be a "Satoshi Nakamoto and all the cryptographers who took a look at the code failed". Cracking a wallet is supposed to be something that requires the resources of a state-sized entity.
- FatalLogic 13y ago
  >Cracking a wallet is supposed to be something that requires the resources of a state-sized entity. A galaxy-sized entity, maybe? The only known way to find a private key from a public key is brute force. That's way beyond the abilities of a state, unless they've made a massive breakthrough in quantum computing. http://i.imgur.com/fYFBsqp.jpg http://i.imgur.com/fYFBsqp.jpg
  - nullc 13y ago
    Obpedantry: In theory rho allows you to recover private keys with work-factor 2^128 (times some small constant depending on how much storage you wish to use), which is significantly less than the 2^256 you might expect from "brute force". Still completely infeasible.
    - teraflop 13y ago
      But first you have to obtain the public key by finding a preimage of the SHA-256-hashed address. And SHA-256 (with the full number of rounds) currently has no known preimage attacks that are even marginally better than brute force.