Comment by rkangel | Hacker News Reader

Comment by rkangel | Hacker News Reader

Parent Full thread

I think that crypto-currencies as a fairly direct replacement for traditional currencies is probably not the future. I don't think it's a 'fad', but I think it'll settle into a niche position in the long term.

The underlying problem that blockchains solve is 'distributed consensus'. This is a solution with a much broader range of applications. For example Maersk has a system for signing handover of shipping containers in ports (https://www.maersk.com/apa-tradelens). This is an international problem with a lot of it happening in countries with a lot of corruption (i.e. you can't rely on legal mechanism). Not being able to forge who is responsible for which container eliminates a lot of problems.

Ethereum does something even more interesting, which is that the network can agree on the result of computations (these are called dapps for "distributed apps"). These can be used to implement simple "smart contracts" for financial purposes but they have a much broader applications. To some extent I'm slightly underwhelmed by the things people are doing with them, but the potential is enormous.

I've been looking for years, and aside from cryptocurrencies, I can't find a single practical use for blockchains that couldn't be done better with more boring technologies.

The Maersk thing is a fine example. It's one company. They already have the trust relationships and legal power that make distributed-consensus approaches unnecessary. That "blockchain" is involved makes no practical difference. It was a shiny bauble that got a lot of consulting hours for IBM, and surely helped getting the project approved because Maersk execs were seeing "blockchain" in the news a lot when it was kicked off.

Yeah but now instead of organizing around a local division of the Maersk company and their software to validate operations - relying on trust that they'll handle things correctly on their end - the locals could just use a blockchain app with zero organization and zero people to trust beyond the protocol itself. Granted that's not needed, as Maersk is apparently doing a fine job now, but if that ever changes - it's there as a backbone, and it's unlikely to be replaced except by a better protocol. This just lowered the need for organization and cut out a middleman (Maersk) saving money on the whole process (at least once the protocol market has churned for a bit and settled into a nice boring reliable one around for a few years / decades, charging only the bare computation costs - which is hopefully/predictably the future of all crypto tech... boring and cheap as shit).

I have little trust in cryptocurrency beyond "people like hype markets", but a global consensus layer is an obvious step with many applications for whittling down every process to its barebones - with some guarantees at a protocol level that the savings aren't going to anyone in particular. This doesn't need to do anything new, it doesn't need to do anything flashy. It just needs to slowly devour every existing business and reduce them to essentially open source software, uncontrolled by any middlemen or power brokers, accessible by anyone for pennies. And it probably will!

This is such a perfect example of a pro-blockchain answer in that it's entirely based in fantasy.

The blockchain hype cycle peaked nearly five years ago: https://trends.google.com/trends/explore?date=all&geo=US&q=b...

There were many pilot projects that soaked up vast sums of money. But in reply to a comment saying I haven't been able to find any practical results, what do I get? No practical results, just the same sci-fi guff.

Again though, you shouldn't be expecting practical results already. The short term hype soaking up vast sums of money is just an overreaction to a technology that will still take many years to develop proficient programmer/user experiences and public trust before it sees real use. Being skeptical of every current project claiming they'll be the ones left standing a decade from now is only smart - it's a giant hype market built on dreams. But you shouldn't be so doubtful that a relatively-uncontrollable internet consensus layer won't be built upon with some interesting results down the line. And that has clearly been created, and proven with decent security already.

I absolutely should be expecting practical results, because most of the people trying to apply this technology are promising practical results. If blockchain proponents were just saying, "Hey, we're fucking around with some tech to see what happens," I wouldn't have any complaints. But that's not where we are. Instead, it's billions of dollars of investment.

Is it possible that there will be "a relatively-uncontrollable internet consensus layer won't be built upon with some interesting results down the line"? Sure. Lots of things are possible. But it's also very possible that after another decade of Bitcoin, et al, we'll still have just a bunch of stuff that's not adding at all to the world beyond a playground for scammers and some tooling for light financial crime.

Sometimes a small and ugly technology turns out to be the Internet. But most of the time, it's just a bit of garbage that never went anywhere.

I'm on the opposite side of yours. I believe that blockchain has a bright future as a currency, but not in logistic. Blockchain money is great because its inflation is very predictable, everyone can use it without permission (versus the slow banking system). I don't believe in blockchain in logistic because database corruption has never been an issue, the problem has always been that people did not insert any data or corrupted data in the database. Blockchain only insure that the database won't be corrupted (not that the data it uses are correct), so it doesn't solve the main issue of logistic.

The slow banking system? Don't crypto transactions (on bitcoin or ethereum) take 5-10 minutes to complete, whereas I can tap my credit card to make a transaction more or less instantly?

A transaction on the Bitcoin base layer takes 10 minutes to be confirmed once. There is second layer tech available to allow faster transactions (Lightning).

A transaction on the Ethereum base layer takes 12 seconds to be confirmed once. There is a vast variety of second layer tech available to allow faster transactions.

There are very different amounts of risk associated with accepting a transaction on the base layer of Bitcoin vs Ethereum after n blocks. For example, Coinbase accepts Bitcoin deposits after 3 confirmations (30 minutes) and Ethereum deposits after 35 confirmations (7 minutes).

Compare to traditional banking: Coinbase accepts ACH deposits instantly (up to a limit) and wires of any size can take 24 hours.

SEPA transfers are instant.

Secondly, is Bitcoin + Lightning in combo decentralized in practice?

> Secondly, is Bitcoin + Lightning in combo decentralized in practice?

I don't know. I'm much more interested in Ethereum, personally.

Time to transaction finality in Crypto vs Banking:

Solana: 0.12 seconds

Bitcoin: 1.2 seconds (on Lightning)

Ethereum: 12.0 seconds

Banking System: ~30 days (2 hours to 5 days for usable funds)

Payment Card Industry: 180 days (2 seconds to a few minutes for usable funds)

Tx finality is immediate in banks, payment cards and blockchains in the same manner. The difference is in network finality. Network finality in Eth2 is around 15 minutes.

Transactions are not finalized for 30-60 days at banks and even longer for payment cards. This is why you can charge-back. 'Time to usable funds' and finality are not the same thing. And no, Eth2 reaches finality in 12 seconds.

15 minutes till the tx is guaranteed to not reverse.

For example, let’s say you sell a car, the buyer sends you 150 ETH, you wait for 3 confirms and he drives away with the car. Then at some point the tx is removed from the blockchain. Now the buyer has your car and the 150 ETH as well. Why did it happen? Because you didn’t wait for tx network finality before giving away the car to the buyer. The amount of time that you should wait is around 15 minutes. Therefore Ethereum is a bad choice to buy coffee with :)

You're just plain wrong.

15 minutes is 75 confirmations.

The Ethereum network has never deviated by more than 2 confirmations (24 seconds) and with PoS this is even less likely to happen.

Finally, the likelihood of that happening AND a car buyer colluding with block producers to scam you is effectively zero.

You're FAR more likely to be struck by lightning in the 15 seconds it takes the driver to leave than you are to have their transaction reversed on the Eth2 blockchain.

Indeed, around 60-70 confirms and then you can give him the car knowing tx won’t reverse

Please look here for yourself: https://beaconcha.in/

I am still waiting for a compelling argument why distributed consensus about me buying some bread this morning is important. What does it gain that mere non-repudiation doesn't?

Reply on news.ycombinator.com