Live data from Hacker News

Blockchains: How They Work and Why They’ll Change the World

spectrum.ieee.org

81–90 of 141 posts

Re: Blockchains: How They Work and Why They’ll Change the World

#81
post #35

My understanding is that blockchain truth is determined via a majority vote. What I'm really not clear on is how someone with a botnet can't simply take over a blockchain. With such large sums of cash at stake, it seems like a serious hacker would be much more interested in controlling a blockchain than doing silly DDOS attacks on some corporate website that slighted them.

Many of these systems are designed such that if you do have the computational power to attack the integrity of the chain, it is actually more profitable to put that power to use for legitimate mining.

Having 51% of the computational power does not let you just make up transactions on your own or transfer funds to your own wallet. What it does allow you to do is pay a counter-party, receive some product from them, and then go back and create a new chain where you paid yourself, instead of the counter-party.

The ethereum whitepaper has a nice explanation of this. https://github.com/ethereum/wiki/wiki/White-Paper

Re: Blockchains: How They Work and Why They’ll Change the World

#82
post #56

It's an easy pitch. Not for all industries or markets, but broadly speaking, removing the need for trusted third-parties is a major leap forward. And that's how those (trustless) decentralized ledgers should market themselves. It only needs to come down to that one simple thing: they increase the surface on which supply and demand can interact. The resulting friction generates wealth. Lots of markets remained untappe…

This is the point so many here are missing... blockchain allows free markets to exist where they didn't before. Look at Sia, there's so many people trying to make money hosting files that you can get 5 Tb of storage on the network for $10 a month right now, less than 10% the cost of Azure or S3. Sure that discount will shrink as more people start using the service, but this is allowing everyone to participate in the storage economy instead of the oligopoly of a few key players that we had before. When everyone can participate in the market the competition will lower costs and raise efficiency, improving the system for everyone.

Re: Blockchains: How They Work and Why They’ll Change the World

#83
post #70

Earlier quoted context omitted.

> I'm on the hype train That’s the intellectual blind spot which is making this so hard to understand: you’re invested in the technology and want to see it succeed, which means you’re inclined to minimize every downside and boost every possible use without asking whether something is better in addition to being possible. Being old enough to remember when the internet was starting to go mainstream, there were op-eds l…

I think the key thing is that even without understanding the technology most consumers could see "before I couldn't get access to this information/correspondence/map/ticket/catalogue/advice/news/opinion/porn and now I can" as an obvious life enhancement in a way which having anonymous miners as counterparties rather than the company providing the service and their bank isn't (not even if expressed in terms of a 3% fe…

Exactly – during the 90s I heard from a ton of companies looking to get on the web. They needed help understanding the technology but in most cases there were clear benefits from making things easier to find, customize, etc.

I’ve seen the same trend for blockchain evangelists to have gone deep into that technology but have only a cursory understanding of the businesses they’re trying to enter. A great example is the retail purchase model which always comes up in Bitcoin discussions, which ignores the fact that most people think credit cards are just fine and like things such as fraud protection more than the hypothetical chance of a fractional overhead reduction. There might be room there but it’s an uphill fight and the value is capped at the lowest Visa, et al. are willing to drop their merchant fees to.

I think your comment about selling tech is also important because the model puts some unusual constraints on it: you can sell to developers by focusing on tech details but how many developers are paid to work on problems which require trustless distributed systems and low transaction volumes? Most big systems require the opposite.

Re: Blockchains: How They Work and Why They’ll Change the World

#84

I had to stop reading at: "In 2009, an anonymous hacker (or group of hackers) going by the name of Satoshi Nakamoto unveiled the first entirely digital currency" In what sense was Satoshi a hacker? Even the modern use of "hacking out code" is a stretch here. Also, it clearly was not the first digital cash. e-Gold and many others predate it.

Technically, I believe that has always been the definition of a hacker, even back in the 80’s. Since then it’s been movies and popular culture that have got it wrong. I believe that what most people think of a hacker, i.e someone who digitally breaks into a system, is actually a ‘cracker’.

I think it's time to stop fighting this fight--nobody I know, even among the security/tech literate, uses the word cracker. We may just have to live with the fact that "hacker" means two easily conflatable things, at least until another term actually catches on for one or the other.

Re: Blockchains: How They Work and Why They’ll Change the World

#85

All the good bits of blockchain, we already had in git: a transaction ledger with hashes for tamper-proofing. Git came out in 2005, based on work going back a few decades. What we're seeing in practice is a lot of "Blockchain(tm)" products that are just this, and not actually a blockchain. Remember how Estonia has blockchains in everything and talks them up at every opportunity? Their local "KSI Blockchain" is just a…

> All the good bits of blockchain, we already had in git

This is not true at all and shows a common misunderstanding of why the blockchain is useful.

The big breakthrough with blockchains and crypto currencies has very little to do with it it being a ledger with signed hashes. There are two main "good bits" of the blockchain:

- it provides a consensus algorithm to agree on a set of updates to the ledger. Not in git.

- it provides a mechanism for anyone to take part in this consensus algorithm while preventing sybil attacks. Also not in git.

Re: Blockchains: How They Work and Why They’ll Change the World

#86

Earlier quoted context omitted.

Confirmation Bias at work. You're invested, so you want it to take off. Yet none of those blockchain enthusiasts could tell me a single USP that blockchains had over storage in a normal database on any concrete use case, with the sole exception of cryptocurrencies. And the reason is simple: all those applications have none of the properties and problems the blockchain solves - which is an elegant niche solution tailo…

> Yet none of those blockchain enthusiasts could tell me a single USP that blockchains had over storage in a normal database on any concrete use case, with the sole exception of cryptocurrencies. > And the reason is simple: all those applications have none of the properties and problems the blockchain solves - which is an elegant niche solution tailor made for mutually mistrusting entities working together on a distr…

> If you find it difficult to imagine use scenarios for blockchain, look for deficient markets caused by the need for centralized trust: monopolies and oligopolies. In each, you'll find an opportunity to create a more competitive market through blockchain.

How? This is just vague handwaving that doesn't actually explain how the blockchain can solve real problems. As of now, bitcoin is the only demonstrable use case for a blockchain, and while I don't think anyone could convincingly argue that bitcoin is useless, it's not going to cause an economic revolution like the advent of the internet.

Re: Blockchains: How They Work and Why They’ll Change the World

#87

This is an open question, if you were to use blockchain to prove how much energy someone produced with their solar panel, doesn't this depend on a trusted configuration? Can't they feed in false data before the block chain part?

Yes and yes. This is the big pitfall i see so many ignoring - the real challenge if you're trying to interface with the outside world is establishing what the internal token means in the outside world without depending so heavily on a centralized weak link that the decentralization of the blockchain's maintenance is pointless. If the actors in a system don't trust each other enough to make changes to the database unilaterally, why would they trust that the other actors aren't lying about their on-chain/off-chain holdings?

Re: Blockchains: How They Work and Why They’ll Change the World

#88
post #61

Earlier quoted context omitted.

While this is correct, for that to be true a person would need to be using the same wallet address for every transaction which is contrary to default wallet behavior / best practices. I think people conflate the side effect of KYC regulation (mass identity linking) with an inherent weakness of blockchains when it's not. I can buy BTC directly from a person to a new wallet address and it is pseudonymous at that moment…

That’s only true if you posit that people actually follow those best practices at scale (is there any reason to believe that’s not false?) and that an attacker doesn’t know how to use a database. Otherwise you’re just betting that they’ll NEVER link your wallet IDs. That last part is guaranteed to be false. Beyond the obvious capabilities of a state-level attacker, think about why Equifax has that data in the first p…

The best practices I'm talking about are built into the client software and a user doesn't need to know about them (new address per transaction already happens in most wallets). Again, totally not claiming linking ids isn't a problem, its just being touted as an impassable blockchain problem which is the basis of this discussion. The problem exists because governments are forcing all onramps to link IDS, not because of an inherent flaw of the system. Even with that 'flaw' other proposals/experiments already exist and work really well, and will probably be integrated directly into bitcoin in the future. An equifax-like service could exist that establishes credibility based on holdings or the public history of an address you control, not necessarily an identified individual. There are some clear benefits to that, one being your identifying information is not put at risk.

Re: Blockchains: How They Work and Why They’ll Change the World

#89

From the article: Can a blockchain find people offering rides, link them up with people who are trying to go somewhere, and give the two parties a transparent platform for payment? Can a blockchain act as a repository and a replay platform for TV shows, movies, and other digital media while keeping track of royalties and paying content creators? Can a blockchain check the status of airline flights and pay travelers a…

- A very high computational cost to mine new blocks - A very high storage cost for each node

Those complains are most of the time coming from free lunch seekers. The network have a value BECAUSE we inject value into it, and that's either direct investment or some energy spent during the mining.

Re: Blockchains: How They Work and Why They’ll Change the World

#90

From the article: Can a blockchain find people offering rides, link them up with people who are trying to go somewhere, and give the two parties a transparent platform for payment? Can a blockchain act as a repository and a replay platform for TV shows, movies, and other digital media while keeping track of royalties and paying content creators? Can a blockchain check the status of airline flights and pay travelers a…

Looking at all these pessimistic comments I can't help but wonder if this is what the op eds in the newspaper looked like while the internet was being built. "its too slow, and even with our best compression transferring a movie would take a month. Not gonna happen. Computers are too expensive for most households." etc. If you want to keep middle men in between most things that we could potentially do programmaticall…

> high computational cost? Security feature

This is a cargo cult claim, repeated by shallow-thinking crypto-promoters. The high computational cost is the reason that proof-of-work systems are doomed to the dustbin of history, as it undermines economies of scale, and ensures perpetually increasing costs.

Fortunately there are other ways to solve the problem, but I feel comfortable that one can safely disregard anything said after the claim that the high computational costs are a good thing; because only a complete fool would ever consider such an argument to be valid.

> If you go the traditional route, you will always have a company in the middle, taking their cut or selling you out behind the scenes.

Instead, in the crypto world, you have... a distributed autonomous organization taking their cut; and they continuously make changes to the protocol which can change that cut or sell you out behind the scenes.

Again. Only extremely shallow thinkers believe that current crypto is worth a damn.

These are the same people who would've looked at the first ever web search engines and thought "this search engine will win", because they're not smart enough to differentiate between a good idea generally, and a terrible implementation.

Today's cryptocurrency implementations are terrible; and are beloved only by dumb, out of touch nerds who repeat nonsensical shit they've heard. Unfortunately, these nerds think they 're geniuses because they rode a bull market in their asset class, and history shows that such people will literally never be capable of recognizing that they're just lucky idiots, nothing more.

Post reply on HN