Topic:   (Read 1524 times)

This. It's not a question if I want to help or not, I don't have the discretion.

That is not going to work because the person is going to say, "I can't reveal private information about our customers without a search warrant (unless I am a phone company and you are the NSA)."

Yes, the stupid ones get caught. But I'd say they are the minority. In this digital internet age, criminals are increasingly automated and in remote locations. Try sending some bitcoins to known compromised addresses. The coins will be stolen immediately. I realized my email was compromised when I watched a password reset email arrive and was instantly deleted. These things are automated and can be done by criminals from all across the globe. Tracking bitcoin thefts is going to become pointless for most thefts. They'll be stolen and mixed automatically.

I saw a recent post from etotheipi that might be an applicable extension of this discussion.
https://bitcointalksearch.org/topic/directional-time-locked-money-storage-278992

I agree that criminals are rarely perfect or ominscient, the design of the system helps to promote an advantage in this regard. The kind that requires a solution to prevent anonymous transactions that are not permitted by increasing the security and safety. Projects like Trezor help with the security aspect but second and third layers of protection are needed to prevent the abilities of theft to occur.
Protocol implementation is the strongest and most efficient solution.

As for coin tainting that was just a suggestion based on the technical projects mentioned, there arou e multiple solutions to this problem and finding an effective one will take some experimentation as you mentioned.
Finding the right peg to fit into the hole is what advanced crytography is for.
Thanks for your interesting insight

Even though I'm sure this kind of tracking will end up existing, as it's one of the most convenient ways for banks and governments to know almost everything about you, I'd rather reverse the problem.

Bitcoin thievery, in my opinion, mostly happens because people don't understand fully how it works, and fail to identify risky moves they might be doing. The protocol is fairly complex, the tools very non-user friendly, and even the QT client that has been promoted as "official" has critical flaws (silently re-using any address of yours as a change one, for example).

EDIT : A recent example here : https://bitcointalksearch.org/topic/theft-from-paper-wallet-flhippy-notes-compromised-278220
Letting other people generate private keys for you, even if they swear they don't keep those, looks like a risky move that shouldn't be allowed.

I'm confident once we get proper, user-friendly tools, handling all the security aspects (such as moving an address to an archive storage once its unspent output is cleared, making sure never to reuse it), it will be even harder to steal coins than it is to steal money today (stolen checks, credit cards numbers, etc.).

As for scamming, I'd like to wait and see how it ends up with Trendon, before thinking any further.

Certainly, but this is just one of several method the thief has available to create deniability.  Any sophisticated thief will use multiple methods to maximize deniability and anonymity while minimizing the associated costs.


True.


Possibly.  It's difficult to predict what direction future bitcoin use will go, as well as the future security of phone wallets.


Absolutely.  It's also good to be aware of what weaknesses and failures any potential "solution" has.


Agreed.

Forcing the thief to lose half their "earnings" is still useful, though.

I think it's worth remembering that criminals are rarely perfect, omniscient adversaries. Many of them are kind of stupid, that's why they get caught.  The kind of thefts we see in the Bitcoin world today can be quite sophisticated and anonymously done, but as security improves that'll get harder and people getting mugged on the street for their phone wallets will get more common. It's worth considering what can be done about that sort of thing.

I'm not dead set against coin tainting, but any system would need to be as decentralised and democratic as Bitcoin itself. So far nobody has really figured out a workable system for that. Advanced cryptography may allow the circle to be squared.

If bitcoin makes it mainstream, I fully expect governments to have programs in place to do exactly this (although perhaps not exclusively for the purpose of tracking stolen coins).

Unfortunately, it doesn't take much for a thief to get around it if they are willing to keep only a small percentage of the stolen bitcoins.

Thief takes 100 stolen bitcoins.  They search the blockchain for 500 addresses that currently hold unspent outputs.  They create another 500 addresses of their own and send a random small amount to each address from a pool of bitcoins that aren't stolen.  They then send 0.1 BTC to each of the 1000 addresses.

The thief now has 50 stolen bitcoins, and any trace comes back to someone that says "I have no idea where they came from.  They just showed up in my wallet one day out of nowhere."

I was reminded of the colored coin system where the coins could be traced through multiple routes the question on how to retrieve them is a protocol related challenge.

However this seems similar to the crowdfunded idea Maaku is posting about

https://bitcointalksearch.org/topic/m.2979759

Of particular interest would be the parts regarding the accounting if you cannot reclaim the coins you can destroy them.

•Destruction of coins, tokens, or assets when no longer needed by a special class of non-spendable, prunable output script.
•Restricting the conditions by which a transaction or sub-transaction may be selected for inclusion by specifying validation scripts, which are run when the enclosing block is validated. Introspection of the block chain from within the bitcoin scripting environment is enabled by the introduction of new opcodes.
•Running accounting servers as private chains with centralized rather than distributed consensus, in which off-chain assets can be issued, transferred and traded in the same way they are in the public chain, with the private block chain providing an audit log.
•Execute an arbitrary number of trades from different accounting servers and/or the public chain in an atomic transaction, using either the public chain or an agreed upon timestamping service for the commit phase.
•Public chains or private accounting servers configured to “observe” other chains to enable much faster but secure cross-chain trade, compared with the existing slow, multi-phase protocols involving revelation of hashed secrets. This requires the ability to extract proofs from the observed chain in order to validate conditional transactions.
•Restrict the usage of a custom asset by assigning to it rotatable signing keys which that must sign all transactions involving the restricted assets prior to inclusion (support for KYC regulatory compliance).

That said the idea of having a service that is able to provide proof of transactions and tracking the details would be interesting.
The implementation of a protocol  that is compatibile with providing this feature to the bitcoin client is interesting discussion.

Pardon this post being a bit long.

As a sidepost it looks like there is some development in colored coins here
The key element of bitcoin is the adaptability of several projects creating something substantial
https://bitcointalksearch.org/topic/m.2909618

I feel that would be setting up a rough precedent and opening doors that should remain closed.

Then people will use mixing services that actually respect peoples privacy (Tor based).
Also would you do that for dollar bills too? Every dollar that every got stolen gets a big red mark on it, nice idea, wont take long untill all dollars are red

What would happen if the coins are sent to a mixing service