7 ms·
Its a nice write up, but it gets many points wrong. 1) Bitcoin is not anonymous. Its pseudonymous since all transactions take place in public between pseudonym
by anologwintermut 14y ago
Its a nice write up, but it gets many points wrong.
1) Bitcoin is not anonymous. Its pseudonymous since all transactions take place in public between pseudonyms (ECDSA keys). This is a big difference, one that hasn't been examined too well, and what has been written on it is not encouraging[0].
2) Bitcoin is not the first currency to prevent double spending without a third party. That minimally goes back to 2006 and a paper "Compact E-cash"[0] where double spending a coin reveled the user's identity and allowed for prosecution.
The problem Bitcoin actually does solve is you don't have to trust the bank to not devalue your currency.
3) Bitcoin does not solve the Byzantine generals problem. Bitcoin is assumed to be correct if 51% of the computation power is honest. If everyone is equal, this means that bitcoin only requires that the majority of the generals are honest. The Bzyantine generals problem has no solution if even 1/3 of the generals are malicious[2]: this is a rather famous result.
How is this possible? Bitcoin isn't dealing with a fixed n Bzyantine generals, its dealing with a peer to peer system where anyone can join and you need to prevent sock puppet accounts. It's a completely different problem.
[0]F. Reid and M. Harrigan, “An analysis of anonymity in the Bitcoin system,” in Privacy, security, risk and trust (PASSAT), 2011 IEEE Third Internatiojn Conference on Social Computing (SOCIALCOM). IEEE, 2011, pp. 1318–1326.
[1]http://cs.brown.edu/~anna/papers/chl05-full.pdf http://cs.brown.edu/~anna/papers/chl05-full.pdf
[2]http://research.microsoft.com/enus/um/people/lamport/pubs/byz.pdf http://research.microsoft.com/enus/um/people/lamport/pubs/by...
- wyager 14y ago>This is a big difference, one that hasn't been examined too well, and what has been written on it is not encouraging[0]. Random numbers is just about as close as you can get to true anonymity. In fact, I'd say that it is as close as you can reasonably expect to get no matter how you try to achieve it. It's also trivial to dis-associate your Bitcoins from a particular cryptographic identity, through the use of mixing services.
- anologwintermut 14y agoIf I temporarily change my name to a random number and then go to 3 brothels, a drug dealer, and then my church, I can assure that after I change my name back and run for congress, it will come up. The problem is that all your transactions take place with the same public key or set of public keys. Sure, you can add more keys to that mix, but someone can probably always untagle the web. Contrast this to say Chaum style e-cash where you get complete anonymity. To withdraw a coin , you have the bank signs a serial number in a such a way that they don't know what they are signing. You can then spend the coin by providing the signature and serial number to a merchant who asks the bank if that serial number has been deposited yet. If not, then the coin is neither counterfeit nor double spent. No one in this system can de-anonymize you since they can't know the serial number you used when you withdrew a coin.
- wyager 14y ago>If I temporarily change my name to a random number There's nothing temporary about it. Unless you voluntarily publish something associating a particular address with yourself, there's basically no way for someone else to make that association. Making the association is completely infeasible if proper precautions are taken. >The problem is that all your transactions take place with the same public key or set of public keys. No, they don't. In fact, with the default client, it tries to minimize the number of times you re-use any given public key. >Sure, you can add more keys to that mix, but someone can probably always untagle the web. No, they can't. The "web" is comprised of a lot of cryptographically random probabilistic association, so the most you can possibly say is something like "there is a .0001% chance that this address belongs to the owner of this other address, and a 99.999% chance that it belongs to someone else." Nothing that would even remotely hold up in court, or even let someone associate you with a purchase. It literally does not get any more anonymous than this in a distributed identity-based system.
- trhtrsh 14y agoWhen an 'anonymous' purchase gets mailed to your house, that creates evidence that you are the anonymous person.
- wyager 14y agoAnd? How does any other digital coin system solve this problem? In the end, you can always associate an identity with a purchase. Bitcoin simply allows you to have an infinite number of disconnected identities.
- anologwintermut 14y agoIn most other systems, you don't have an identity that is visibly associated with the transaction. Each transaction is unlinkable to any other transaction or your identity: you are not trying to prevent the breadcrumb trail between each of your transactions from following you home, you are actually not leaving crumbs.
- nazgulnarsil 14y agowhere double spending a coin reveled the user's identity and allowed for prosecution. because identities are a secure system? This also allows for malicious attacks where an innocent third party is made to look like a double spender and have their identity revealed.
- anologwintermut 14y agoYou can't frame a third party unless you have their private key. This is called Exculpability and there is a proof of that property in the paper. I agree it has problems ( especially if you can use double spent coins to buy a ticket to Venezuela and avoid extradition). But the point is Bitcoin wasn't the first paper to solve the double spending problem without just asking the bank, not that its solution was inferior. There is a good deal of academic research on e-cash schemes and some of it might actually be worth studying (despite the fact that it was never really deployed) instead of ignoring it and thinking Bitcoin is the be all and end all.
- jaimebuelta 14y agoI think the pseudonymous problem you are stating is a huge problem for wide adoption. If you are able to associate a key with a person, you know all the transactions done by some one (of course, with that wallet) You can take extra measures to have multiple wallets, etc, but it can be difficult to truly understand the risks and it could need an extra effort that, simply, is not going to be assumed by lots of people. I'm pretty sure that wallets associated with service and product providers will be easy to figure out, so you can learn A LOT of sensible information about someone... I guess that the risk will be low, but the consequences could be huge.
- zanny 14y agoCurrently, using any accepted online transaction policy forcibly immediately associates a real identity with it every time. Bitcoin makes acting anonymously possible, it just requires work. Those that want the anonymity can put in the work to be anonymous, whereas everyone else gets the ease of use of associating keys with people.
- anologwintermut 14y agoBy accepted online transaction systems I assume you mean visa, paypal, dwolla, etc. In that case two points 1) If you are willing to accepted centralized systems , than my second citation gets you complete anonymity. 2) At least with those systems, your information isn't public to everyone.