Hacker News Reader

Comment by tyfon | Hacker News Reader

I've never used bitcoin but to my understanding it can take quite some time for a payment to be processed.

What happens if you buy something and pay, then when the merchant receives the bitcoins the value of them has dropped say 10%? Does the merchant take the loss in this system?

View on HN

Bitcoin alone is responsible for this reputation of slow transaction times among cryptocurrencies. Litecoin and Ethereum have the same blockchain architecture but are mined faster, so they are faster, but this linear improvement may not be sufficient for scaling. Minerless "DAG" coins like Nano may prove to be more practical in the long run because they scale exponentially.

re: Value change during a transaction: volatility is a fundamental problem with any currency. Adoption and volume will decrease volatility, which this announcement is moving towards :) Ripple's value-add to the banking industry, for instance, similarly relies on adoption in order to create an international pool of liquidity in order to facilitate efficient cross-border currency exchange. The more volume, the less volatility there is, which means less cost to banks using XRP as a payment "bridge". The same is desirable for merchants and consumers.

Nano takes like 4 seconds to confirm and has zero fees for transfer. It's really a no-brainer, but unfortunately this Bitgrail thing is going to muddy the waters for a long time.

BTC survived Mt. Gox and Bitfinex, NEM survived Coincheck, and fiat currencies survived many a bank robbery. Investor perception is low because the theft hurt early adopters, but I think that the technology will shine through sooner rather than later.

I'm with you for the long hodl.

your logic is funny, is like having a busy site with a postgress db, and saying postgress is at fault, look at my mysql wordpress site with 1.3 visitors per day, how fast it is.

Ethereum will colapse on itself sooner of later, no one managed to sync a node recently.

Hasn't the latest Geth update improved syncing? https://blog.ethereum.org/2018/02/14/geth-1-8-iceberg%C2%B9/

Where exactly is my logic funny? Mined blockchains scale linearly. Increasing block size, etc. can increase speed temporarily but can still only scale linearly. Busy site or not, the ledger needs an architectural change to handle exponential growth. Your busy Postgress db would also need to change in architecture from a single resource to something more distributed in order to scale exponentially. A less busy website of the same architecture also wouldn't handle exponential growth. What am I missing?

Edit: To clarify, yes, some mined blockchains feel faster now simply because they are less busy. They are still doomed to Bitcoin's fate at scale unless distributed systems techniques like sharding (or Nano's block lattice) are used. I'm not saying that the less busy blockchains are any better.

> To clarify, yes, some mined blockchains feel faster now simply because they are less busy.

Yes, and for now it's not proven in practice that Nano will also stay responsive when it hits the same scale/busyness.

Fair enough. In theory Nano's design can handle something like O(2^n) transactions per second where n is the number of nodes, while mined blockchains can handle O(n) transactions per second where n is roughly the block rate multiplied by block size. In practice who knows. The former has a fighting chance, at least.

Coinbase can easily take care of this. They can have a small reserve and sell the bitcoin as soon as the transaction is published/sent to the mempool. No need to wait for any confirmations. If the transaction happens to be fraudulent†, they just buy the bitcoin back and realize the earning/loss (in the long run they will tend to even out).

†Not sure what the attack vectors are here.

If they're interested in taking on the counterparty risk they could do this, but when the mempool is flooded you can replace pending transactions with new transactions with a higher fee (called "Replace by Fee").

but they don't seem to convert to fiat (at least from this article or the demo), so it's irrelevant.

I've made payments to merchants who acknowledge that the price can change over the process of sending a payment. The Bitcoin price is fixed at the point that the transaction is acknowledged in the mempool, so any fluctuation during the time taken for transactions to reach their required verifications is ignored.

This was my immediate question. I bought some Etherium from Coinbase a couple weeks back and by the time I got the funds, the value had dropped (!)25%. I don't particularly care because I'm long on my crypto holdings, but in what world is something like that going to be tolerable for merchants?

It sucks, but you can store cash there in advance, and purchase on the fly when desired.

Reply on news.ycombinator.com