Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

A faster, more efficient cryptocurrency

news.mit.edu

Post reply on HN
Print view View original

281–290 of 839 posts

Re: A faster, more efficient cryptocurrency

#281

Post by westurner »

throwaway415415 wrote: post #234
westurner wrote: post #104

Earlier quoted context omitted.

Are there reasons that e.g. Bitcoin and Ethereum and Stellar could not implement some of these more performant approaches that Algorand [1] and Vault [2] have developed, published, and implemented? Which would require a hard fork? [1] https://www.algorand.com/ [2] https://dspace.mit.edu/handle/1721.1/117821

My understanding is that PoS approaches follow normal byzantine agreement theory which states that adversaries cannot control more than 1/3rd of the accounts (or money in the case of algorand). You can also delay new blocks more easily. Ethereum is scared or that so they are implementing some hybrid form. Bitcoin is doomed from my perspective, because of the focus on proof of work and the confirmation times. When you…

And what of decentralized premined chains (with no PoW, no PoS, and far less energy use) that release coins with escrow smart contracts over time such as Ripple and Stellar (and close a new ledger every few seconds)?

> Algorand has a very fast consensus mechanism and can add blocks as quickly as the network can deliver them. We become a victim of our success. The blockchain will grow very rapidly. A terabyte a month is possible. The storage issue associated with our performance can quickly become an issue. The Vault paper is focused on solving this and other storage scaling problems.

What prevents a person from using a chain like IPFS?

Ethereum Casper PoS has been under review for quite some time.

Why isn't all Bitcoin on Lightning Network?

Bitcoin could make bootstrapping faster by choosing a considered-good blockhash and balances, but AFAIU, re-verifying transactions like Bitcoin and derivatives do prevents hash collision attacks that are currently considered infeasible for SHA-256 (especially given a low block size).

There was an analysis somewhere where they calculated the cloud server instance costs of mounting a ~51% attack (which applies to PoW chains) for various blockchains.

Bitcoin is not profitable to mine in places without heavily subsidized dirty/clean energy anymore: energy and Bitcoin commodity costs and prices have intersected. They'll need any of: inexpensive clean energy, more efficient chips, higher speculative value.

Energy arbitrage (grid-scale energy storage) may be more profitable now. We need energy storage in order to reach 100% renewable energy (regardless of floundering policy support).

Re: A faster, more efficient cryptocurrency

#282

Post by solean »

throwaway415415 wrote: post #264
mihaifm wrote: post #62

Earlier quoted context omitted.

I would name Nano, a very underrated cryptocurrency. Uses DPoS (Delegated Proof of Stake) instead of Proof of Work, so it doesn't do mining, also it has zero fees and minimal transaction times. I've ran a full node for some time to support the network, the whole blockchain size was about 5GB, but the downside of zero transaction fees is that there is less incentive for people to run full nodes. I stopped mine out of…

If no fees, what are the incentives to accept transactions in a block?

Also, what stops you from spamming the network?

Re: A faster, more efficient cryptocurrency

#283

Post by omarchowdhury »

rmah wrote: post #246
koonsolo wrote: post #190

Earlier quoted context omitted.

I see this as a disadvantage of using banks. They act as if you can freely access your money in your account, but actually you can't. That money is loaned out for the most part, and it's a trick that keeps the current system going. But once you have a bank run, reality will kick in on how banks really work. At least the cryptocurrency that you own is really yours today. The numbers on your bank account are yours some…

Most people I've spoken with about crypto prefer the way banks work. The idea that losing your keys means losing your money, with no recourse, terrifies them.

> The idea that losing your keys means losing your money, with no recourse, terrifies them.

All the more incentive to not lose them!

Re: A faster, more efficient cryptocurrency

#284

Post by cjbprime »

good-idea wrote: post #242

I'm shocked and disheartened by how infrequently environmental concerns are brought up when discussing crypto. Proof of Work vs Proof of Stake, security, storage size - these are problems for the network and those that want to see it become a real part of our economy. (I am one of those) But the amount of energy being used to mine bitcoin is a real problem - it's not just a technical challenge (like the problems abov…

I'm not sure whether this is true, but I expect that electricity-from-coal for Bitcoins isn't a massive problem because doing so is simply unprofitable: the electricity costs more than the value of the Bitcoins, so it would be irrational to do.

I would expect that mining has become an arbitrage market that narrows in on the cheapest possible electricity anywhere in the world, and as such is dominated by excess "free" electricity from e.g. hydroelectric power that is generating more electricity than can be used (or transported).

Re: A faster, more efficient cryptocurrency

#285

Post by hanniabu »

naveedi wrote: post #38

Full disclosure: I work on the cryptocurrency in this article, Algorand. There are a lot of questions and speculation here about this paper and Algorand. I would be happy to try an answer them to your satisfaction. Some context may be helpful first, though. This paper is an innovation about one aspect of our technology. Algorand has a very fast consensus mechanism and can add blocks as quickly as the network can deli…

Can you provide more detail because from the article this doesn't seem so impressive.

> Vault reduced the bandwidth for joining its network by 99 percent compared to Bitcoin and 90 percent compared to Ethereum, which is considered one of today’s most efficient cryptocurrencies

1) Since when are these considered efficient? I don't think anybody in the know would say this. They're the most popular, but no means the most efficient.

Bitcoin is 250GB, so 90% of that is still 25GB to join the network, which is still ahhh enormous amount. And what's the baseline for comparison here? Were all of bitcoin's transactions replayed on an Algorand test network for this comparison? Or is this a metric from some test usage? If the latter then that's a huge issue since it grows in size.

2) On top of that you're saying it could accrue a terabyte a month in data. What type of usage is this under? Is that on current bitcoin transaction levels? 2017 transaction levels? A steady state tx/s? Is a backup of this data needed or is it throwaway and summarized in the latest blocks? If it's still needed then that's a decentralization issue because not many people will be maintaining full nodes.

3) What type of specs are affected by these changes? Can you still perform atomic swaps? That's a pretty standard requirement nowadays and would hinder the Blocknet and exchange interiperability.

Re: A faster, more efficient cryptocurrency

#286

Post by drak0n1c »

polyomino wrote: post #69

Hogwash. How can a user be sure that the chain is valid if they don't download the transaction set? If users cannot verify the amount of coin in the network this is useless.

Hedera Hashgraph uses the gossip protocol and simulated voting protocol to do it. It's possible, there are implementations like Hashgraph already doing a distributed ledger without a full blockchain.

Re: A faster, more efficient cryptocurrency

#287

Post by sparkie »

JoeAltmaier wrote: post #274
seanyesmunt wrote: post #273

Earlier quoted context omitted.

> In that sense, the coin would create an economic incentive to pollute I think if anything it creates an incentive to use renewable energy. If miners are able to pay less for power, they make more money. As the cost of renewables continue to drop, more miners will move to it.

Irrelevant. Renewable energy could be used to replace other sources, instead of peed away on this. A larger demand for energy likely means more polluting forms of energy will come back into service/stay in service to meet the demand.

Energy is created to meet demand. It is not consumed to meet supply.

The exception to the latter is where excess energy is created because a method of energy production is created to meet peak demand, but generates more energy than is needed off-peak. While some of it could be stored, storage is inefficient and it is better to sell the excess energy cheap, so that you can continue to expand your energy production capabilities.

Miners take advantage of off-peak energy excess, whilst at the same time, increasing the on-peak demand. This means that the provider of the cheap renewable energy source needs to invest in producing more, and cheaper energy to meet the higher peak demand.

Re: A faster, more efficient cryptocurrency

#288

Post by brazzy »

arisAlexis wrote: post #121
brazzy wrote: post #66

Earlier quoted context omitted.

> Also why would I trust them, when there is no need to? Because your alternative is to trust the crypto infrastructure, which is far less trustworthy.

what do you mean by crypto infrastructure? you mean the code or the miners? actually banks have failed in the past, this infrastructure not yet

I mean all of it in aggregate, and it has failed many, many times.

Re: A faster, more efficient cryptocurrency

#289

Post by warkdarrior »

packet_nerd wrote: post #187
jayd16 wrote: post #150

Earlier quoted context omitted.

>actually banks have failed in the past, this infrastructure not yet Man, what are you talking about? There's a new "$100 million in coins go missing" story every 6 months.

Crypto 101: never let anyone have your private key. No-one who religiously followed that rule has ever had their crypto stolen. Coins go missing from exchanges, which break the basic private key rule. People use exchanges for convenience because as clever as cryptocurrency is, it has no answer to the exchange problem.

Crypto 102: regular people will never use cryptocurrencies if they have to guard their private keys without a fallback in place.

Re: A faster, more efficient cryptocurrency

#290

Post by hanniabu »

sbenitoj wrote: post #252
hanniabu wrote: post #240

Earlier quoted context omitted.

I think it makes sense. You have an economy built on an economic model rather than an economic model based on an economy. With the latter you're constantly guessing and tweaking the model and roughly every 10 years there's a crash and the economists say okay, NOW we have it perfect and we'll never get it wrong again. Why do you need an unlimited money supply? The only thing causing these shocks is that the economy is…

Exactly, any supply of money will do, no need to permanently increase the amount in circulation — this is why the market settled on gold (which has an inherently low inflation rate due to the difficulties of mining) and impossible to pass off counterfeit gold at scald.

You need inflation for proper incentives. It's like a tax on the entire network to keep it supported. Without proper incentives, the game theory collapses.

The difference here is that inflation is predetermined and set instead of printing money on demand like we do in the traditional financial system.

Post reply on HN