News for hackers. Stuff that matters.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by jrlocke on Sunday March 23, 2014 @05:37PM (#7455287)
What if there was a way to tag bitcoins as stolen and declare them null and void? Would this not disincentivize thieves? Could we prevent abuse of such a system?
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by fragsworth on Sunday March 23, 2014 @05:48PM (#7455345)
This is a really contentious issue. Depending on how late the coins are marked as stolen, some number of legitimate transactions may have been made with them.

What burden does this place on users to be certain that they're receiving "legitimate" coins?

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by saalweachter on Sunday March 23, 2014 @06:19PM (#7455482)
You could set up an independent organization to track and identify stolen coins, and then convince merchants and exchanges to voluntarily refuse transactions involving them. If a few large entities refused to do business with addresses linked to stolen coins -- and the tainted addresses were fairly few in number -- then a lot of actors would be forced to follow suit to maintain their relationships with the initial core group of boycotters.

If you want the BTC network itself to reject stolen bitcoins, there's no need to stop there: if you had community agreement, you could just as easily reverse thieving transactions and make the victims whole.

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by DomBlack on Sunday March 23, 2014 @05:42PM (#7455316)
The problem with the tagging approach, is how would you verify the tag was genuine? You would have to add some sort of central authority which every peer could implicitly trust and that would undermine the whole concept of bitcoin...
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by Lerc on Sunday March 23, 2014 @05:46PM (#7455333)
The term used for this is coloured coins. It's possible to do, and there are forms of coloured coins in use. The question is not of how to do it, but whether it is desirable to have coloured coins.

It's like adding a new rule to how a piece moves in chess. You could easily imagine some of the advantages caused by a change, but not all of the consequences.

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by waterlesscloud on Sunday March 23, 2014 @05:57PM (#7455384)
That's not what colored coins are. Colored coins are coins which have been used to mark ownership of another asset, like say a share of stock.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by TophWells on Sunday March 23, 2014 @06:22PM (#7455493)
Whether or not it's the same thing, it uses the same mechanics. You can link the output of one transaction to the input of the next, and construct a chain of transactions that says "This 0.125 BTC transaction uses bitcoins that were originally stolen from MtGox / bitcoins that were originally in the pile designated as marking ownership of my shares of stock".

The question is whether that's a meaningful thing to measure. If you have bitcoins that are more valuable than normal, because they're shares of stock, then you'll make sure they're not mixed with "normal" coins. That's the idea of the Coloured Coins project. But if you have bitcoins that are less valuable than normal, because they're stolen goods, then you'll try hard to disguise that fact. You'll immediately spend them on goods or currency, or deposit them into a Coinbase account and then withdraw them later (so you get different coins back), or just throw them into a mixer with the other coins that need to be laundered.

The person who ends up holding the tainted coins could be perfectly honest, and have acquired them from a perfectly honest source, who in turn acquired them from a perfectly honest source. Is it fair to tell them that their money is suddenly worthless, when they had no way of knowing anything was wrong?

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by Pacabel on Sunday March 23, 2014 @06:22PM (#7455494)
What prevents a particular mark from being interpreted as some measure of validity, though? If such marking can represent ownership of another asset, could it not represent some other concept, too?
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by waterlesscloud on Sunday March 23, 2014 @06:29PM (#7455525)
That's fine. But "colored coin" is a term that has a broader use than that.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by rdl on Sunday March 23, 2014 @08:02PM (#7455839)
The term is probably "coin validation" -- it breaks the finality of bitcoin transactions, creates a single point of regulatory exposure, and IMO will destroy bitcoin if widely adopted.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by r0muald on Sunday March 23, 2014 @05:42PM (#7455312)
I think that would require an authority deciding whether bitcoins were truly stolen. Otherwise, couldn't anyone could just claim a transaction was a theft?
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by atmosx on Sunday March 23, 2014 @05:44PM (#7455325)
It's a very interesting development as it shows that stolen bitcoins can be tracked indefinitely unlike FIAT.

Now, what you propose though, is against the freedom bitcoins is assumed to stand for (on the ethics side).

On the technical side that's very possible, as of today, if everyone plays along and I mean literally everyone, especially mixers.

How could one track down bitcoins flushed into public mixers? I guess they could do this as well, if the BTC were tagged and thus refused by the mixers or even worst hold captives and resend them to original issuer. However, you only need 1 TOR hosted mixer to lose track of the wallets.

On the other side imagine the practical chaos: I make a huge transaction of non-tangible goods in a situation where you can't prove that the service/product was delivered. Then I just tag these bitcoin that once belonged to my wallet as stolen if the other party doesn't give them back :-) you see where this is going right? :-)

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by gutnor on Sunday March 23, 2014 @05:58PM (#7455390)
> unlike FIAT.

Actually, unlike any commodity or any currency with a physical representation. What makes other currency not traceable is the notes and coins, not the fact that they are fiat or not.

side note: has "fiat" now become synonymous with non-crypto currency the same way hacker has become a swear word ?

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by Klinky on Sunday March 23, 2014 @06:32PM (#7455540)
Yes fiat has become a word of derision amongst the cryptocoin community, completely ignoring the fact that cryptocoins are essentially a form of fiat currency. Many crypto fans sidestep the fact that cryptos also don't have intrinsic value, and the networks aren't as a decentralized as one would think(handfuls of devs, early adopters, and pool and exchange operators hold a large sway over the networks and future of the coins).
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by baddox on Sunday March 23, 2014 @06:11PM (#7455442)
He doesn't say that fiat is the only untraceable form of currency.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by atmosx on Monday March 24, 2014 @03:17AM (#7457101)
I can't speak for others, but I have nothing against FIAT :-)
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by bjourne on Sunday March 23, 2014 @06:31PM (#7455535)
Aren't "mixers" basically acting as money launderers in that scenario? In the real world that is illegal and frowned upon so it should be in the bitcoin world too. Suppose I'm running Honest Business Inc. Since all transactions are public info, I can easily refuse transactions from mixer addresses. I can even decide to refuse transactions from address which are < N nodes separated from mixers and so on. If everyone in bitcoin land decided to freeze out blackmarket addresses it would make it very hard for crooks to use their stolen coins.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by JamesArgo on Sunday March 23, 2014 @06:40PM (#7455566)
Mixers maintain fungibility. If fungibility is lost, I don't see much future for Bitcoin.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by TophWells on Sunday March 23, 2014 @06:24PM (#7455509)
>It's a very interesting development as it shows that stolen bitcoins can be tracked indefinitely unlike FIAT.

You can track dollar bills by their serial number. It's basically the same thing as what's going on here, except that few people record the serial numbers of their bills.

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by atmosx on Monday March 24, 2014 @03:20AM (#7457106)
Hmm, not in real life. I can write a script that parses the blockchain and keeps track of the transactions.

How many people/authorities can do that with FIAT? Not to count the impossibility of actually tracking down FIAT that has not been used by a bank or authority that CAN actually control the unique number :-)

So in case 1 it's very easy to do, in second case is almost impossible outside very well defined scenarios (police trying to frame drug dealers, etc.)

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by annon on Sunday March 23, 2014 @05:41PM (#7455310)
Since there is a fixed number of bitcoins, there would still be incentive by causing faster deflation by removing coins from the pool.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by joeblau on Sunday March 23, 2014 @07:11PM (#7455682)

  > What if there was a way to tag bitcoins as stolen and declare them null and void?
When you say that sentence, replace the word bitcoins with gold. I say this because in essence, bitcoins are a finite resource that have perceived value. They also are harder to obtain the more people mine them. Now think about what would happen if someone stole gold and an authority (which doesn't exist) declared the stolen gold null and void. Someone, somewhere with lower ethics would be very happy to take the "null and void" gold because the fact that it's stolen has absolutely impact on it's value.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by gus_massa on Sunday March 23, 2014 @08:02PM (#7455838)
Gold is much easier to melt and cast in new “clean” ingots. It’s not possible to trace it, so nobody could declare some of it void. But each Bitcoin transaction is public, so it’s easy to see the origin of each Bitcoin.

But I think this idea has many technical and psychological problems . You essentially need to fork the Blockchain and convince the “51%” to use the new fork where these Bitcoins are illegal.

Doge recently changed some details of the mining procedure using a hard fork, but they have a strong central organization and they pretend that Doge is not a serious coin. An equivalent change in Bitcoin would be much more difficult.

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by sentenza on Sunday March 23, 2014 @09:24PM (#7456146)
Is it really not possible to trace gold? I was under the impression that for much of the gold out there you could use the "footprint" of contaminations in the gold to infer the age (since mining) and procedence to a certain degree.

Of course, if you melt it in a top-notch facility, they'd probably be able to heat out the contaminants (or remove them chemically?), but people trading in stolen or otherwise "hot" gold probably don't have access to such facilities.

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by gus_massa on Sunday March 23, 2014 @11:22PM (#7456504)
I guess that removing the contaminant traces is very difficult or impossible (unless you have a very hi-tech facility), adding a few contaminants looks easier. Anyway, mixing gold from different source should give a new contaminant composition that makes it very difficult to trace.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by mason240 on Sunday March 23, 2014 @08:19PM (#7455901)
> When you say that sentence, replace the word bitcoins with gold.

Let's suppose that this gold is Nazi gold, taken from Jews. You know that Nazi gold will be traced to you(because each gold atom has an address). Would you still accept that gold as payment?

I'm not an expert on bitcoin though, so my analoy could be way off.

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by Mchl on Sunday March 23, 2014 @05:41PM (#7455305)
Question 1: Who would have the power to do that?
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by orthecreedence on Monday March 24, 2014 @05:32PM (#7461706)
The US government, obviously.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by runeks on Monday March 24, 2014 @04:47AM (#7457267)
> What if there was a way to tag bitcoins as stolen and declare them null and void? Would this not disincentivize thieves?

It would both disincentivize thieves and honest users.

It was decided in a legal case over 300 years ago in Scotland that it would ruin the point of money if stolen money were - like all other goods - the property of their original owner: http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2260952

I think it makes sense to apply the same principle to bitcoins.

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by thefreeman on Sunday March 23, 2014 @06:15PM (#7455465)
there is no such thing as an individual "bitcoin". It's just a numerical balance associated with a wallet. There is no way to mark a certain subset of bitcoins as invalid.

At best you could flag wallets, and any wallets who've made transactions with those wallets, etc., etc. which is basically what the people on reddit are doing to "trace" the coins. But, as many have mentioned before, it is not a very accurate way to trace the transactions especially if a service like a tumbler is used.

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by wmf on Sunday March 23, 2014 @06:28PM (#7455523)
To understand tainting you really need to work on the level of outputs, not wallets.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by TophWells on Sunday March 23, 2014 @06:22PM (#7455497)
Correct.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by cbhl on Sunday March 23, 2014 @05:46PM (#7455332)
In theory, this is already possible. Anyone can patch the client to reject transactions containing such coins; the key is getting enough adoption for it to be meaningful. (I imagine starting by forking mining-specific applications like cgminer would be more effective than forking the client apps, since you really want to get 51% of the network to reject these transactions for it to mean anything.)
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by stormbrew on Sunday March 23, 2014 @06:59PM (#7455642)
cgminer never sees anything but the block header, it makes no decisions at all about which transactions to include in a block. The client is still the thing that does that.
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by iancarroll on Sunday March 23, 2014 @05:41PM (#7455308)
That would be centralizing the system, no?
Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by tinco on Sunday March 23, 2014 @05:41PM (#7455303)
Would it disincentivize exchange developers to have sane security architectures?

Who would declare them null and void? What's to stop any movement from declaring non-stolen bitcoin as null and void?

Re:Mt. Gox's Bitcoins are being sold on Bitstamp? by cLeEOGPw on Monday March 24, 2014 @03:59AM (#7457187)
You can modify clients to reject coins from the given address. If more than 50% of the network has modified clients, it will be very hard to spend the coins for thief.