Hacker News Reader

Nano is an interesting cryptocurrency, but it’s very benefits are attack vectors. Currently Nano has no way of dealing with a precomputed PoW attack which means a single attacker has the ability to effectively ‘DDOS’ the network. This is briefly discussed in their white paper, and AFAIK I’m the only one to have proposed a complete comprehensive solution to their problem, but the implementation appears to be so complex and would require a network split that the nano team has moved on to less comprehensive mitigations. Currently any actor, including myself with my trivial resources could easily DDOS the entire network, which is a major concern I hold to the cryptocurrency and cannot advocate its usage at least until this has been appropriately addressed.

Nano implemented Dynamic Proof-of-Work in V19 to mitigate spam attacks causing congestion on the network [1]. When the network gets congested due to a spam attack (or other reasons) the PoW required to send a transaction is dynamically raised. This should prioritize real transactions that will do the new, higher PoW, while the spam transactions that have a lower precomputed PoW have to wait to get processed until the network's PoW requirement dynamically lowers. Doesn't this mitigate the precomputed PoW attack vector? This tweet [2] also shows this dynamic PoW feature in action. [1]https://medium.com/nanocurrency/dynamic-proof-of-work-priori... [2] https://twitter.com/GenMeasures/status/1149855971457454081

The current mitigation enables attackers to raise the difficulty at whim and effectively disables any cached PoW's by large entities. This hurts Nano's primary feature - speed - for the end user. Also, while it certainly makes it harder for an individual attacker to attack the network, it is still very doable given enough time and resources. My proposal (which was shortly after the blog post you've cited came out) included a concept of leniency while making any possible attack exponentially more difficult both to plan and execute over a period of time while maintaining a constant ability for entities to precompute PoWs. However the biggest issue would be the introduction of synchronicity in an asynchronous network through probability and is not exactly trivial to implement. I've since moved on as I don't have the time or motivation to further pursue this (I am not exactly very hyped on the idea of cryptocurrency anymore).

Also worth noting that NANO distributed most of the supply via a CAPTCHA faucet. No token sale or mining. I think NANO is unique among cryptocurrencies with regards to its initial distribution.

This sounds incredibly interesting.

How does it work privacy-wise? Can I see other peoples wallets and who they've sent money to, received money from, etc, like Bitcoin or is it closer to Monero in that regard?

Currently, Nano is like Bitcoin in terms of privacy. So Nano is not private. However, the Nano devs are interested in adding privacy, if possible. This quote is supposedly from Colin LeMahieu in 2017

"I love the concept of privacy in the network and it's a hard thing to do right. Any solution used would need to be compatible with our balance-weighted-voting method which means at least we'd have to know how much weight a representative has even if we're hiding actual account balances. To be fully anonymous it would have to be hide accounts, amounts, endpoints, and also timing information; with advanced network analysis the timing is the hardest thing to hide. Hopefully some day we can figure out an efficient privacy solution though the immediate problem we can solve is making a transactional cryptocurrency so we're focusing on that." [1]

[1]https://www.reddit.com/r/nanocurrency/comments/am0dp0/lets_t...

>He did this by having each account have its own blockchain, rather than using one blockchain for all transactions. The result is Nano transactions take under 1 second (fully confirmed) and there is no fee for transactions. It's also scalable with 1465 confirmations per second (aka transactions) reached on the test net the other day.

What are the trade-offs being made by using this rather than a blockchain? I'd imagine the consistency guarantees would be lower?

>There is no mining involved, so Nano is environmentally friend and voting is done using a system called ORV (open representative voting) which is similar to delegated proof of stake, but doesn't require actually staking and risking coins.

If there's no staking/risk involved, what's the disincentive for bad actors to fork the chain? The whole point of staking (and putting coins at risk) is to disincentivize that from happening.

Nano refers to their system as a block lattice rather than a blockchain. For just transferring value between users, a block lattice seems to be superior to a single blockchain like Bitcoin in terms of scalability, free transactions, and transaction speed. I'm not sure what a consistency guarantee is, but I've never heard of a Nano double spend if that's what you mean.

Since there isn't a single chain in Nano, I don't think Nano can be forked (in the BTC blockchain sense) by bad actors. Instead, a bad actor would need to have 51% of the online supply of Nano delegated to representatives they control. At that point, the bad actor could start blocking transactions. Of course, there are risks with any decentralized system. For BTC, this risk is about controlling hash rate (which can be done by controlling mining pools, not just mining hardware).

To get proper answers from the devs to these questions, your best bet would be to post them on the official Nano forum at https://forum.nano.org/

> I'm not sure what a consistency guarantee is

Without the consistency part, its trivial to make an asset that is billions of times faster/scalable/etc than nano.

I meant I don't know what specifically they meant by "consistency guarantee". Does that mean a double spend? Or a 51% attack? The Nano main net has been running since 2015 without any double spend incidents or a 51% attack.

Usually the proposers of a system define some security/consistency garuntees that they think their system meets, followed by some arguments why that is true. Users can then decide if these properties meet their needs.

"51% attack" is an attack. The corresponding security property would be something along the lines of "A malicious party that controls < 50% of the network hash power can make a transaction that is confirmed by the network n times and then make another conflicting transaction on an alternative chain, and have that chain eventually become the canonical chain, with probability negligible in n" (i probably messed up details but that is the gist).

A consistency garuntee isn't a specific attack, but the general properties of the system, what it can do and what it cannot do. The fact that nobody has done a double spend is pretty meaningless. We don't know if that's because nobody has tried/cared or if that's because its really hard.

To put it another way: say there were two engineers who designed two bridges. Someone asks the engineers, is your bridge safe? Engineer 1 says: we made a careful design and extensive testing, as long as no more than 100 tons is on the bridge, it won't collapse. Engineer 2 says: the bridge has been there for 5 years. In that time not a single person has died from the bridge collapsing on them. Which bridge would you trust more?

What would a consistency guarantee look like for a cryptocurrency? A whitepaper? If so, Nano's whitepaper can be found here (though I've heard it needs to be updated) https://content.nano.org/whitepaper/Nano_Whitepaper_en.pdf

If not a whitepaper, then what do you consider Bitcoin's (or any cryptocurrency's) consistency guarantee to be?

I think you're fixating a bit too much on the medium rather than the content. It could be a whitepaper, it could be a forum post, it could be a message written in the sand at the beach.

That said, the whitepaper does contain arguments in that direction. It would be nice to see more in depth formal arguments though, as well as some independent analysis.

I hate that there still hasn't been a thorough security / plausibility review on Nano; at least I haven't seen one. No one seems to take it very seriously. Maybe it's too good to be true.

A security audit of Nano was completed by Red4Sec in July of 2019. You can find an overview of the audit here[1] and the full Red4Sec report here[2]. "Red4Sec has been able to determine that the overall security level of the asset is optimal."

[1]https://medium.com/nanocurrency/nano-protocol-security-audit... [2]https://content.nano.org/Nano_Final_Security_Audit_v3.pdf

This seems to be an vulnerability analysis of the Nano source code. I meant the monetary security and game theoretic plausibility of the protocol itself.

E.g. how do we know which blocklattice is the correct one? (in Bitcoin it's the longest chain, which contains most energy and thus can't be created from thin air). Are there enough incentives to keep the network running and decentralized? Etc.

Each Nano account is its own blockchain. These accounts interacting form the block lattice. So it's not as though there are competing block lattices in Nano. In Bitcoin there can be a sidechain that can overtake the main chain (if the sidechain becomes longer than the main chain) and then cause a reorg of some blocks. But I don't think anything like this is possible in Nano. You'd have to ask the devs though to know for sure on the official Nano forum at forum.nano.org

In terms of the incentives to keep the network running without mining fees or proof of stake rewards, the Nano community and devs think that merchants and other services using Nano will save significant amounts of money by not having to pay credit card processing fees. This will incentivize those merchants and service providers to run their own node (which only costs around $60 per month) in order to keep the network healthy and decentralized. The lack of mining fees and proof of stake actually disincentivizes the kind of centralization we are seeing with Bitcoin mining.

Security wise Nano is DOA. Currently any attacker can effectively DDOS the network with trivial resources. I’ve personally proposed a complete solution to this problem over a year ago but it’s implementation is complex and would require a network split and my proposal has effectively been ignored in favor of other mitigation’s that fall far short of protecting the network. I’m not a security researcher and hell I’m not even in the CS field anymore so if I could figure out and implement an network wide attack on my own the barrier to entry is low (to be clear I have not executed any such attack). If you want to read further, the issue is a precomputed PoW attack. Nano’s very benefit - speed - is it’s biggest pitfall.

I'd be interested to know more about your proposal to solve the precomputed work attack vector. Also, I think if you can easily create an attack that would DDOS network, you should absolutely try it on the Nano test net. That's what the test net is there for. To my knowledge, the precomputed spam attack has been mitigated by the Dynamic PoW feature that was added in v19. See my other reply with more details on that. I would like to know more about your proposal. You can email me at purplewumpus@protonmail.com

I simply do not have the time nor motivation to conduct an attack on the Nano testnet. I've responded to your other reply with the technical shortcomings of the mitigation, my proposal is still open somewhere in github's issues but I've since moved on from cryptocurrency as a whole.

Interesting. What's the incentive to devote computation to make the system work if there aren't mining fees?

Merchants using Nano for transactions will not have to pay credit card processing fees. This will save them a significant amount of money. They will thus have an incentive to run a representative (a full node) that voting weight can be delegated to from other users in order to keep the network decentralized.