Does BitFury's new research findings impact Dash's PrivateSend

[Willy Woo]

I was having an conversation with one of the Monero devs on Twitter and it brought to light that BitFury's new research published in their recent whitepaper has methods to untangle some CoinJoin transactions.

(Page 2) : Shared send. Users organize into groups (via an intermediary) and tangle their coins in a single transaction (e.g., within the framework of the CoinJoin algorithm [3]). The small set of senders and recipients may be known, but it is not clear how the funds are distributed among them.

Conclusion (page 19) : Our computational experiments show that detection and analysis of shared send mixers is possible in real time for the most of bitcoin transactions. We also discover that mixing transactions occur quite often on the Bitcoin Blockchain; by our estimations, they constitute about 2.5% of all bitcoin transactions. Interestingly, about half of these transactions are able to be untangled. Namely, they can be uniquely split into two or more sub-transactions, allowing for the restoration of relationships among addresses referenced in the transaction.​

Here's the whitepaper.

Here's the Twitter conversation.

Does anyone know how the methods in the whitepaper impacts Dash's PrivateSend feature given it's an implementation of CoinJoin?

 

[TanteStefana]

In Bitcoin, you have to send your coins into a 3rd party to have them "mixed". It's been a while, but I'm pretty sure that your funds are not denominated when mixed. If this is new research, it only confirms Kristov Atlas's "Bitcoin Sudoku".

But this has nothing to do with the way Dash's private send works.

In Private Send, the user's funds are first denominated into units: 0.1001 dash, 1.001 dash, 10.001 dash, etc. Why the .001? This is for fees when the funds are used. Then, at the protocol level, meaning -- the funds never ever leave the control of the owner, your wallet sends a denomination into a pool, then pulls it back out under a different "name" or actually account number. There is no way to link the new account number to your wallet. 3 people are in each mix minimum. The MN could be running some kind of snooping software, and could keep track of which funds went to which address, theoretically. They might be able to see who sent the new account addresses to collect the funds. So that is why we have multiple mixing rounds. The chances that a group of colluding masternodes are able to follow the mixes becomes exponentially more difficult with each round. Even with massive control of the MN network, say 3000 out of 4000 masternodes, the chances of de-anonymizing 8 rounds is reduced to 9%. This sounds like a lot, until you realize how unlikely it is that any one entity owns 3000 MNs or 1/2 of all coins in existence when the entire core team owns less than 10%

Still, it's not a 100% surety. But it requires enormous control over the network.

How will this be improved?

In Evolution, mixing will be coming from the DAPI, the network in such a way that no masternode can see what IP address the inputs are coming from. By "blinding" the MNs, there is no way to follow the mixing. This plus instant 10 rounds via quarum should make it as close to 100% secure as monero in any practical sense of the word.

 

[TanteStefana]

I have more to say but g2g, be back

So to continue, the difficulty and cost of trying to buy that many MNs, it's simply impossible. Dash's daily volume is less than 1 million dollars. It's simply impossible to do, at least to do it without people noticing. At at current volumes, it's not happening.

In the early days, one person stood out as being a Dash Whale. We watched as he bought up coins, and his holdings went up to around 600,000 coins. It was an event to say the least. Later we learned it was one of bitcoin's, then litecoin's earlyish adopters, otoh, who is now one of our coolest community members. Sure, he didn't make it hard for us to watch, but if someone were buying up Dash to control the MN network, it would be noticeable.

IF a government wanted to buy into the network, they would have had to have done so early on, and believe me, nobody knew Dash would do so well in the early days, certainly not enough to get a slow bureaucracy to spend money. Governments do not move fast - EVER - except maybe if a leader is extremely confident and reckless in times of extreme war (like WWII) But such risk would never include funding, it would be deployment of assets.

And finally, the billions it would take to get a small percentage chance of denominating random inputs (no targeting, no assurance to hit the enemy) is not worth the money.

So where does that leave Dash today?

In most likelihood, there are no masternodes logging information, unless it's a hacker trying to see what they can do, which is kind of pointless because they can't possibly have enough MN to get anywhere.

Evolution will "blind" Masternodes, so that all these issues are mute

The reason I like Private Send is it's simplicity. It's not dependent on fancy cryptography, it's simply removing the ability to follow the coin. In many ways, it's like Monero that way. Monero offers up 3 addresses per transaction, only one of which is the "real" one. It is instant, that is good, but at the cost of enormous blockchain bloat, which they're trying to fix.

The other cost, which for me is the #1 reason I don't trust Monero as a currency, is the inability to follow the ins and outs, the coins generated since the beginning and what is being spent. What if someone found a way to insert coins into the chain, and spend them, nobody would know. This could undermine the whole thing. What if a Quantum Computer is able to figure out which were the true coins spent, thus unraveling the privacy of Monero?

Dash uses simple uncertainty, it's VERY good uncertainty, while retaining all properties that are important to a currency and being 100% quantum computer proof because everything is already known, you can't get more information that isn't there.

And of course, privacy is only one aspect of Dash. For some reason, probably because it was the first feature Dash implemented, that's all that people see. The populous is so far behind us, it's actually funny. They're more than 2 years behind. Instant transactions, Governance, Budget from the blockchain already implemented 9 months ago, soon we'll have ease of use, a decentralized API for anyone to plug directly into the network with. This will allow mobile wallets with assurance that they are always on the correct decentralized blockchain, not one maintained by a wallet provider. And that's not even scratching the surface, there is so much going on, it's hard to choose what to talk about!

 

[Willy Woo]

Awesome and thorough answer, thanks!