Hacker News new | past | comments | ask | show | jobs | submit login

This article isn't very good. Half of it is about 51% attacks, which is the one attack blockchains were always transparent about, and the other half talks about how some smart contracts are buggy. Cool.



The blockchain developers were always transparent about 51% attacks, but it turned out to be far more plausible in a scaled network than was expected. The viability of 51% attacks are a result of the business deals surrounding the ecosystem, not any technical deficiency in the code.

Personally, I think the entire concept of cryptocurrency is flawed because it punts on the governance problem. Ditto for smart contracts. People / companies like to have the ability to dispute the outcome of a transaction, and you really can’t do that without vesting authority in a governing body to overrule the technology. But then, if you have a central authority that has the power to modify transactions, do you really need blockchain?


On the other hand, the vast majority of transactions in my life don't need government-muscled dispute protection. To me it makes more sense to opt-in to that in the occasions you deliberately want it.

Our financial system is its own brand of crapshoot. As we speak, I've just issued my 6th chargeback (6th month) to Amazon more an AWS account that someone hacked and changed all my details except for my payment info. I think about this every time I see someone nit-picking some imperfection in cryptocurrencies.

I just want a choice.


On the other hand, if someone hacks your crypto wallet and cleans you out, there’s no administrative remedy or deposit insurance.


Security technologies can only get better over time. I would argue that with hardware wallets, recovery seed backups well-protected & well-stored, M-of-N signatures, etc, you can get to a level of security that's already far far better than cash, and good enough for 99.99% of use cases.


All of those require me to have air-tight opsec all of the time. One screw-up and you can be totally cleaned out with no remedy other than the courts (which can take years!)

In the real world, I lose my actual wallet every 5 years or so. Worst case, I lose $40 cash and have to spend a couple hours reporting credit cards stolen.

The best feature of a bank is that you can outsource all that opsec to them in exchange for a trust relationship with the bank. The latter is far easier to manage day-to-day.


That "only" in first sentence invalidates it. There's enough historical data to disprove it.


Always remember: it's not actually about removing trust, it's about shifting trust. And when you look at it in that light, most of these technologies become a lot hairier then they're worth.


Could you elaborate on this? It's an interesting comment


One of the main selling points of cryptocurrencies is decentralization, of not needing to trust a central party in order to use the currency. But the cryptocurrency doesn't actually eliminate that trust, it simply moves it. It moves it into trusting the protocol design, trusting the implementation of the clients, trusting your local hardware and OS to not be compromised, and in the case of smart contracts, trusting the smart contract to not be buggy. What's worse is in the centralized case, as long as the government continues to function you have lots of options for what to do if a transaction goes badly. In the cryptocurrency case, for the most part you're SOL, because cryptocurrencies by design don't allow anyone to e.g. void a transaction.


> It moves it into trusting the protocol design, trusting the implementation of the clients, trusting your local hardware and OS to not be compromised

These are not central parties and they present risks with anything one does with a computer on a network.

> what to do if a transaction goes badly. In the cryptocurrency case, for the most part you're SOL, because cryptocurrencies by design don't allow anyone to e.g. void a transaction.

Transaction finality is a feature not a bug. Whatever conveniences that are enjoyed int traditional banking can be implemented in a layer on top of the base protocol.


> These are not central parties and they present risks with anything one does with a computer on a network.

If someone compromises my OS and steals my credit card details, the bank sends me a new card and handles the fraudulent charges on my behalf. I'm not out any money

If someone compromises my OS and steals my bitcoin private key, they can drain my entire wallet and I've lost everything, and have zero recourse.


Those examples ignore the other side:

If I'm a merchant and someone uses a stolen or cloned credit card, or simply performs a charge-back, they get my goods for free and I have basically zero recourse.

If I'm a merchant and someone pays me with bitcoin, I'm not out any money regardless, nor do I have to expend time/money/legal resources to ensure I keep that payment.


Show me merchants who trust Bitcoin enough to accept it directly, and not through an exchange.


>People / companies like to have the ability to dispute the outcome of a transaction

Crypto currency doesn't stop this. You can dispute cash transactions and you can dispute crypto transactions. If I pay for something in crypto and it is not delivered, I can sue, complain, etc just as I would with a credit card. What you can't do with crypto is run to visa and have them magically reverse the charge (possibly screwing over the merchant). But dispute away my friend.


This is actually valid. You can absolutely dispute a crypto currency just like you can dispute a cash transaction.

Except that only holds true when you perform a crypto transaction in the same way as you do a cash transaction, for local services. The moment you do it for something remote, you will run into all sorts of issues.

If a merchant in my country fails to deliver on a cash/crypto transaction, the dispute process is easy.

If a merchant in another country fails to deliver on a crypto transaction (posting cash is a whole other topic that we'll avoid here), what dispute mechanism do I have? International law is a minefield.


Look at the long wake of professionals and businesses left by Trump alone, who could out litigate most to get a cheap (or outright unpaid) bill. There are countless reasons to shift trust to a programmatic contract. Plus there are projects like Agrello which seek to implement the smart contract in a way that is also legally binding. Programmable money is here to stay, and there is a sea of social, economic, and technical benefit to be had from it.


The issue isn't just that some smart contracts are buggy. After all, all software should be expected to have bugs at some point. The bigger issues are that a)the source code can be very visible, making it easier to figure out how to exploit than traditional code running in an isolated server, and b)recovering from an attack can be incredibly difficult (such as needing to create a fork of the blockchain pre-attack).


Sure, but this has been known for years to everyone with even a tangential interest in smart contracts. It is in no way news.


Agreed, it's a sensationalist title.




Consider applying for YC's Spring batch! Applications are open till Feb 11.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: