5 ms·

Researchers Locate Flaw In Bitcoin Protocol

59 points by pwg 15y ago

niklas_a 15y ago
Glad to see that the researches also propose a solution - doesn't seem like this is much of an issue. Not sure why it ended on the front page of YC.
- adrianwaj 15y ago
  HN subs are back in https://twitter.com/#!/bitrific https://twitter.com/#!/bitrific https://twitter.com/#!/bitrific/status/136398818303426560 https://twitter.com/#!/bitrific/status/136398818303426560 I had them removed for a number of months as so many articles were upvoted indiscriminately, but then reintroduced about a week ago. Seems articles will still be upvoted by followers, but they are more discretionary this time around. Although, I might be placing too much importance on whether having them there makes a difference.
- mquander 15y ago
  I assume it's on the front page of YC because it's an good summary of an interesting topic, which is more than you can say about most stories. I don't use Bitcoin or care whether it succeeds but I upvoted the submission because it got me to spend ten minutes thinking about this interesting problem with the protocol.
dmk23 15y ago
The bigger flaw is that there is lack of economic incentive to continue to donate computing power to keep the network running. At the current exchange rates and bitcoin mining yield it cannot even pay for the energy cost. As it stands today just around $300K worth of hardware (I cannot quite recall the reference off top of my head) is needed to be able to "out-compute" the existing network and introduce a corrupted / compromised version of the block that everyone will accept as real. That's not to mention that the transactions are not anonymous and just pseudonymous, which means if your identity is somehow disclosed (anytime in the future) all your bitcoin dealings are going to be completely public for anyone to see and use against you. Distributed crypto-currencies may have a future, but it would require a lot of work to make them viable.
- narcissus 15y ago
  "All your bitcoin dealings are going to be public" is really only true if someone gets a hold of your wallet, or if you are using the same Bitcoin address for each transaction. By default, every time you send or receive you are given a new 'default' address. If you don't use that, then yeah, a group of transactions can be linked to you. Otherwise your identity would have to be 'disclosed' for each different address separately. Unless of course someone gets your wallet, but then as far as I'm concerned, you've got bigger things to worry about it.
  - coderrr 15y ago
    "Otherwise your identity would have to be 'disclosed' for each different address separately." Actually, it's a lot worse than that. Most people will have a majority of their addresses linked without knowing it. http://coderrr.wordpress.com/2011/06/30/patching-the-bitcoin-client-to-make-it-more-anonymous/ http://coderrr.wordpress.com/2011/06/30/patching-the-bitcoin...
    - feral 15y ago
      I've done some work on this (http://arxiv.org/abs/1107.4524 http://arxiv.org/abs/1107.4524) and the phenomenon that coderr describes is very widespread in practice (we can examine this passively, from the transaction history) and its linking identities we can be pretty sure the users didn't intend to have linked. I'd say that anyone concerned about anonymity should be using a client with the functionality that coderr's patch provides, and be incredibly careful using Bitcoin in general.
    - narcissus 15y ago
      Thanks for the link: that's really interesting. I guess that's more a problem with the official client more than the protocol itself, though, right? If you were to create a new address and have all of those 'trackable' addresses send their balances to that new address, haven't you essentially removed this problem? I mean, all the people that sent to you in the first place can see where you sent the coins to, but they can't prove that you are holding that new account. Or is there something I'm missing here? For what it's worth, I completely see how this is a problem. However, I don't see it as a particularly hard problem to solve (even 'programatically'). EDIT: and it was until I read the other reply here that I realised that that is actually your blog post. Nice work!

59 points by pwg 15y ago

Glad to see that the researches also propose a solution - doesn't seem like this is much of an issue. Not sure why it ended on the front page of YC.

adrianwaj 15y ago
HN subs are back in https://twitter.com/#!/bitrific https://twitter.com/#!/bitrific https://twitter.com/#!/bitrific/status/136398818303426560 https://twitter.com/#!/bitrific/status/136398818303426560 I had them removed for a number of months as so many articles were upvoted indiscriminately, but then reintroduced about a week ago. Seems articles will still be upvoted by followers, but they are more discretionary this time around. Although, I might be placing too much importance on whether having them there makes a difference.
mquander 15y ago
I assume it's on the front page of YC because it's an good summary of an interesting topic, which is more than you can say about most stories. I don't use Bitcoin or care whether it succeeds but I upvoted the submission because it got me to spend ten minutes thinking about this interesting problem with the protocol.

dmk23 15y ago

The bigger flaw is that there is lack of economic incentive to continue to donate computing power to keep the network running. At the current exchange rates and bitcoin mining yield it cannot even pay for the energy cost. As it stands today just around $300K worth of hardware (I cannot quite recall the reference off top of my head) is needed to be able to "out-compute" the existing network and introduce a corrupted / compromised version of the block that everyone will accept as real. That's not to mention that the transactions are not anonymous and just pseudonymous, which means if your identity is somehow disclosed (anytime in the future) all your bitcoin dealings are going to be completely public for anyone to see and use against you. Distributed crypto-currencies may have a future, but it would require a lot of work to make them viable.

narcissus 15y ago
"All your bitcoin dealings are going to be public" is really only true if someone gets a hold of your wallet, or if you are using the same Bitcoin address for each transaction. By default, every time you send or receive you are given a new 'default' address. If you don't use that, then yeah, a group of transactions can be linked to you. Otherwise your identity would have to be 'disclosed' for each different address separately. Unless of course someone gets your wallet, but then as far as I'm concerned, you've got bigger things to worry about it.
- coderrr 15y ago
  "Otherwise your identity would have to be 'disclosed' for each different address separately." Actually, it's a lot worse than that. Most people will have a majority of their addresses linked without knowing it. http://coderrr.wordpress.com/2011/06/30/patching-the-bitcoin-client-to-make-it-more-anonymous/ http://coderrr.wordpress.com/2011/06/30/patching-the-bitcoin...
  - feral 15y ago
    I've done some work on this (http://arxiv.org/abs/1107.4524 http://arxiv.org/abs/1107.4524) and the phenomenon that coderr describes is very widespread in practice (we can examine this passively, from the transaction history) and its linking identities we can be pretty sure the users didn't intend to have linked. I'd say that anyone concerned about anonymity should be using a client with the functionality that coderr's patch provides, and be incredibly careful using Bitcoin in general.
  - narcissus 15y ago
    Thanks for the link: that's really interesting. I guess that's more a problem with the official client more than the protocol itself, though, right? If you were to create a new address and have all of those 'trackable' addresses send their balances to that new address, haven't you essentially removed this problem? I mean, all the people that sent to you in the first place can see where you sent the coins to, but they can't prove that you are holding that new account. Or is there something I'm missing here? For what it's worth, I completely see how this is a problem. However, I don't see it as a particularly hard problem to solve (even 'programatically'). EDIT: and it was until I read the other reply here that I realised that that is actually your blog post. Nice work!