News for hackers. Stuff that matters.
Re:Ethereum will use around 99.95% less energy post merge by bsedlm on Tuesday May 18, 2021 @10:57AM (#27195485)
> the initial distribution of the coin has to be wide enough

I don't understand how proof of stake works to the depth I understand proof of work. But this reassures me that it's feasable that they'll accomplish the same distributed consensus.

So then, could I say that Ethereum proof of stake will allow the owners of the coins (ether) to be independent from the owners of the mining operations?

or uhmm...

is the independence between the computaional costs of the "mining" and actual minted ether?

Re:Ethereum will use around 99.95% less energy post merge by kemonocode on Tuesday May 18, 2021 @11:02AM (#27195563)
Ethereum proof of work made it so the owners of the coins could be independent from the owners of the mining operations, even if in practice many miners end up keeping most of the block rewards themselves and only reinvesting what they need in new infrastructure and to maintain what they already have. Proof of stake makes it so the miners and holders are now the same (you stake the coins that you have, or you pool them up with others), however the cost to wreck the chain is much greater than it would have been 3-4 years ago.

The whole idea is that Ether is so spread out now, it'd be unfeasible for someone to snatch up enough of it for an attack, in a similar way to how an ever-increasing difficulty makes it harder for a hostile actor to coordinate enough of it to make such attack.

Re:Ethereum will use around 99.95% less energy post merge by jyu on Tuesday May 18, 2021 @11:48AM (#27196268)
Maybe it is spread out now, but won't there be centralized aggregators of eth so some point in the future a handful of POS nodes control a disproportionate amount of power? Is it so hard to imagine that coinbase or some other exchange accumulates enough eth to sway transaction validation?

Seriously, please answer if this is wrong!

Re:Ethereum will use around 99.95% less energy post merge by swensel on Tuesday May 18, 2021 @01:34PM (#27198176)
I do think there will be some centralization at the exchange level. As of April 2021, Kraken had 600,000 ETH staked for ETH2 [1].

It's not in the interest of Kraken or Coinbase to disrupt one of these PoS networks, but there is some barrier to entry for staking ETH2 or other PoS coins on your own, vs staking them on an exchange. In the case of ETH2, if your staking node goes down, you get slashed and lose some ETH. If there isn't slashing (not all PoS coins have that), I don't see what guarantee of network security or uptime there is.

I'd be curious what PoS coin experts think about this part. It seems like PoS / staking can lead to centralization. PoW has energy concerns for sure, but it has so far demonstrated decentralization pretty well.

I'm legitimately curious about this. I'd love for PoS to be feasible and am trying to understand it more.

[1] https://en.cryptonomist.ch/2021/04/20/ethereum-2-0-600-thous...

Re:Ethereum will use around 99.95% less energy post merge by xur17 on Tuesday May 18, 2021 @03:52PM (#27200140)
I definitely do think centralization is a risk of PoS over PoW.

> It's not in the interest of Kraken or Coinbase to disrupt one of these PoS networks, but there is some barrier to entry for staking ETH2 or other PoS coins on your own, vs staking them on an exchange. In the case of ETH2, if your staking node goes down, you get slashed and lose some ETH. If there isn't slashing (not all PoS coins have that), I don't see what guarantee of network security or uptime there is.

This is actually one place where ETH has tried to incentivize independent staking - the penalty for downtime is equal to the incentive for mining. In an extreme example, if you are down for 6 months of the year, and up for 6 months, the downtime costs should cancel out the earnings from the other 6 months. One caveat to this is that there are larger penalties for correlated downtime (ex: if a large portion of the network is down). This is to de-incentivize centralization of mining.

That said, as someone fairly technical that could run his own staking node, I am seriously considering using a centralized service, or at the very least using a vps. This makes me think that the the majority are going to be independently run on a slew of hosting providers and via centralized hosting providers.

There are also some interesting "decentralized" options like Rocketpool that haven't launched yet, but will allow staking via smart contracts against a pool of random nodes.

And then at the end of the day, the choice in staking providers should allow the network to at least react to centralization risks. Say a locality forces a provider to censor transactions in some way - I imagine folks will move their funds to a provider in another country, or switch to something like Rocketpool, effectively working around the issue.

Re:Ethereum will use around 99.95% less energy post merge by nootropicat on Tuesday May 18, 2021 @03:10PM (#27199533)
>but there is some barrier to entry for staking ETH2 or other PoS coins on your own, vs staking them on an exchange

This is true, but those exchanges either charge fees or are going to charge them, making it more profitable to stake at home. Because offline penalties depend on correlation to how many other people are offline, it's actually safer to stake at home. Eth was accumulated primarily by devs that understood its value before everyone else. Devs are in general paid well. As a result, it's nothing special for those individuals to own thousands of eth - it was even possible to buy eth below $100 as recently as in 2020. Those 600k eth on Kraken aren't that much relatively.

Right now it's very early and many people aren't staking because there are much higher returns elsewhere, and before the merge withdrawals aren't possible, so you can't even go back if something better appears. I fully intend to stake at home once extreme yields elsewhere stop - in the long run staking is likely to have the highest yield on eth.

>PoW has energy concerns for sure, but it has so far demonstrated decentralization pretty well.

Mining is extremely centralized in China. Mining has infinite economies of scale + less efficient miners are pushed out, so the most efficient entity/location is certain to control all hashpower eventually (not necessarily China). https://www.nasdaq.com/articles/bitcoin-mining-hash-rate-dro...

Re:Ethereum will use around 99.95% less energy post merge by 0134340 on Tuesday May 18, 2021 @01:53PM (#27198450)
Centralization meaning those who have more gain more? I'd love to hear about any currency that doesn't have this feature/bug. Those who gain power tend to be able to acquire more by bootstrapping from prior power; it's pretty much universal. Even PoW has its own form of centralization in that those who have more can become richer and more easily gain therefore leading to maturation (centralization) of the currency.
Re:Ethereum will use around 99.95% less energy post merge by swensel on Tuesday May 18, 2021 @02:04PM (#27198605)
My understanding is those who have more coins on a PoS network have more stake / power. This can matter if there are things like on-chain governance / voting rights, depending on their stake. Those with more stake would also get more staking rewards (it's like an APR % return based on the total staked), and if they stake their rewards as well, then they'd have even more total stake on the network. There also are concerns with those having a majority of the stake in a network being able to disrupt or attack the network (things like slashing based on poor behavior can prevent bad actors from wanting to do that, as they would lose some or all of what they had staked in that case). There are also different kinds of PoS though, and I'm not an expert on it.
Re:Ethereum will use around 99.95% less energy post merge by 0134340 on Tuesday May 18, 2021 @02:50PM (#27199267)
Yes, that's a feature or bug, however you see it, that's universal in capitalism, even communism or socialism what implementations I've read about. I guess if one isn't happy with those that have power in one system, a person should switch to another system. It's just so universal that I don't think there's any other way around it other than switch systems as none are perfect, all insofar as I can see are susceptible to the power of consensus.

We've tried to mitigate it with constitutions in the political world and it helps to some extent but many would agree that there is still an exploitable hole in that those with power can use their own to gain more or mitigate risk. And any time you mitigate that feature/bug too much you run the risk of decreasing reward for work and stake, thereby delegitimizing the system itself or in the case we speak of, your currency. So pick your poison.

Re:Ethereum will use around 99.95% less energy post merge by randomhodler84 on Tuesday May 18, 2021 @05:56PM (#27201785)
It is in their interests — remember when CZ asked if it was possible to rollback BTC before the hack? Imagine if you actually gave the exchanges this power. Nothing to stop them staking with customer funds. The exchanges do it now with other PoS shitcoins. PoS is the death of decentralization.
Re:Ethereum will use around 99.95% less energy post merge by 0134340 on Tuesday May 18, 2021 @01:56PM (#27198499)
Yes, that's the problem with validation through concensus which is universal in currency. With enough power (nodes) you can delegitimize other stake holders.