Hacker News Reader

Comment by rauljordan2020 | Hacker News Reader

rauljordan2020

I'm one of the maintainers of one of the implementations of eth2 live today (https://github.com/prysmaticlabs/prysm) and can help offer more context on this. Ethereum proof of stake has been live since December 1st and it currently secures over 6 billion USD worth of value https://beaconcha.in/. Currently, this is a chain that lives in parallel to the proof of work chain we know as Ethereum today. The idea is that over this coming year, the entire state of Ethereum will be "merged" into this new chain that uses proof of stake consensus via a docking process. "Escalating the eth2 merge" means there is more expedited work from researchers and developers when it comes to performing this critical decision. Since eth2 consensus is already live and running correctly with a large set of validators, miners cannot censor this new chain and therefore cannot really stop the merge.

View on HN

Do you think the headline is correct in labeling this as a 51% attack? The other argument would be that developers are trying to push through a contentious fork (mostly but not only contested by miners and mining pools), and mining nodes are planning to not adopt the proposed change?

Afaik there's no plan to attack, they are planning a "show of force", i.e. get more then 51% in a pool for a while without doing anything with the power. This is separate from either adapting or not adapting the fork itself and happens before the fork.

The original chain has to be forked either way, with or without eip 1559 because the "ice age" coded into the current version (it's in there for this exact reason). So if the change goes through and some miners decide to block it, there's not going to be an "original chain" and a "fork", but two forks, one supported by devs and minority of miners, and one supported by miner majority.

"We're not going to 51% attack the network. We're just going to allocate 51% of the hashrate into a single pool to demonstrate that we could 51% attack the network"

Why should this threat of violence be treated differently than the act itself? Miners are threatening to do the one thing their entire existence is supposed to prevent. They're cutting off their noses to spite their faces, instead of enjoying another year or so of profitability

Because whether it’s an attack or not is more to do with the framing than act. If you get all the miners together and can amass 51% of the hash power then you have consensus and have enough control of the network to simply not accept the EIP.

If something like this isn’t possible and you don’t actually need consensus to push through changes then the whole “distributed” part is just a fiction and we’re really talking about a centrally controlled network. Miners are just basically threatening to vote against the EIP if it comes to it.

Threatening to vote against a bill if it reaches the floor in its current form isn’t an attack or violence. It’s just a message that if you want it to pass you should probably change it.

This isn't something new, Bitcoin went through this in 2017 (relevant keywords: UASF, segwit, BCH).

You don't need a miner majority to do fork the chain, both versions (with and without fork) will exist and whichever is most valued by users/the market tends to "win" and bring miners back in because they like money.

Unlike your voting on a bill analogy, the reality of blockchains is that both universes can co-exist.

Calling it a 51% "attack" is like saying Biden "stole" the election by getting at least 51% of the votes. The whole point of a cryptocurrency is that there is no central authority. Vitalik can provide a vision, but he doesn't, and shouldn't want, to control it. If the compute power disagrees with him, then that's just how it's gonna be.

But when we elevate from PoW to PoS the whole point is that compute power won't matter anymore. So a more appropriate analogy would be that if 51% of horse carriage drivers wanted that cars shouldn't be able to drive on roads, then cars shouldn't be able to drive on roads.

> Calling it a 51% "attack" is like saying Biden "stole" the election by getting at least 51% of the votes.

A better comparison would be that a billionaire decided to take over just because he had more cash.

"attack" is a misnomer. 51% is a majority vote.

Yes, but, in fairness, the assumption behind blockchain is that it works as long as 50%+1 of the computation is power is honest, not that any block constituting 50%+1 is necessarily honest/good/correct.

Doesn't the majority vote mean that they get to control the cryptocurrency?

Sounds like an attack to me.

The ice age mechanic remains one of the most significant and overlooked aspects of Ethereum. It's really good.

What really surprises me is how in bad faith the PoW miners are acting. Ethereum was, from day one, planning to transition to PoS once the research and proof-of-concept was working. If anything the Ethereum PoW miners have been very lucky in that Ethereum is kinda late on its roadmap to switch to PoS (at least compared to when first announced).

You're saying a significant number of people who gathered behind the maxim "code is law" for money may have signed onto what the code was doing and not what the long-term social plan was? Quelle surprise!

Yep, the Ethereum Classic crowd made the same mistake in the context of the DAO "hack".

DAO: Code is law, code is law, trust the code over any natural language document that describes our smartcontracts.

"Hacker": Okay, if the smartcontract says that, then it follows that I'm entitled to these ETH if I do this...

DAO: What? No! That's not fair! That's not what we meant! Oh, come on, do-over!

ETH Classic: No? That defeats the entire purpose of Ethereum, for code to be law.

DAO: Not when it comes at the cost of significant wealth for the core team!

This is the classical distributed consensus fallacy. Law is only law if there's consensus around it. The "original" law still lives on in the form of ETC, but it happens that the other one generates way more consensus around itself, and thus that's the more valuable one. Distributed consensus means no one can change the rules unless there's a "majority" defined by PoW, PoS, PoA, or something else, that chooses to change the rules. And there's nothing preventing anyone from proposing any kind of change and trying to get a majority behind it.

Which means there's as many "legitimate" chains as there are ways to define a majority, and as blockchain entries increasingly denote external assets things are about to get very messy. In the end disagreements will probably be adjudicated by traditional governments based on regular contract law (in the best cases) or individual physical force (in the not-exactly-so-formal ones), and after a few instances of that maybe we can be done with the whole foolishness.

This is mostly true, but I'd argue legitimacy is harder to achieve than it seems, because of network effects.

That said, the current NFT craze is idiotic. As you point out, there's lots of people trying to try external value to a blockchain using what is more or less nothing more than a "super pinky promise". And that's supremely stupid. Today's valuable blockchains have value by themselves. Linking blockchains to external properties, assets, certificates or capabilities is still an ongoing problem. Only after the identity problem is solved and standardized will NFTs linked to external things be worth anything.

So here’s a legitimate question because I haven’t seen it asked before. Whats the value of a paper “certificate of authenticity” at the moment. My understanding is that with art if you can tie it to the original issuer, then it has value. And indeed I know there is some fraud law around forging such things. So NFTs seem fairly well suited to provide hard to forge art provenance with minimal effort.

They also seem well suited for being able to transfer ownership of physical goods and also transfer things like warranties. With an NFT tied to a serial number, you could ensure only one owner with that serial numbers item can claim warranty coverage, ensure resellers don’t try and pass off something used as new, ensure stolen goods can be tracked as such, etc.

With both paper certificates and NFTs, their worth depends on the issuer. The NFT craze is stupid, but as you point out, coupling real art with a NFT certificate is a really neat use case for NFTs. A NFT minted by an author and/or a reputable source such as an auction house can act as a unique, impossible-to-duplicate certificate that helps claim ownership of some real-world possession.

It's a given that at some point this will be tested in a court, with favorable results.

Maybe the litmus test is does the object linked to the NFT have real world value without the NFT. If you wouldn’t pay for the item without an NFT, then perhaps it’s not legit.

Well, the code never guaranteed that people would not merge the PoW chain to the PoS chain :)

At the end of the day, this consensus system will rule by consensus.

Question: Will there be a chain split - will we end up with 'ETH classic' and 'ETH proof of stake' after the upgrade or how does it work?

I think this is guaranteed.

As long as one person keeps running the old software, the old chain is alive.

Even if nobody does it, somebody could start running it again at any time.

The different chains will have different market caps. And the general public will forget about chains that have sufficiently small market caps.

I wonder what will happen to Ether that is in custody. Like in the Greyscale Ethereum Trust or the 21Shares Ethereum ETP.

Has anybody ever checked the TOS of these companies on what will happen if multiple forks carry significant value?

False, the difficulty bomb will kill off eth1.

> the difficulty bomb will kill off eth1

Wouldn’t it make sense for miners to switch to a new Ethereum chain without the bomb?

Yes, miners can turn off the bomb, but that requires that they coordinate and determine a schedule for the hardfork that turns off the bomb (you need a hardfork to turn it off).

This scenario is definitely possible, but it requires a little more work than "one guy keeps running an old eth1 node".

isn't it trivial to patch?

Then they will be on a fork of a fork of a fork (ETC -> ETH -> ETH_POW -> ETH_POW minus Difficulty bomb)

who will take this chain seriously

> who will take this chain seriously

If it has 51% of Ethereum’s current hashpower, somebody. (Ethereum Proof of Stake is the same number of forks away from Ethereum Classic.)

The Difficulty Bomb has been delayed at least once, so delaying it once again or defusing it, it not unimaginable.

How would they be on a fork, if the "original" dies because of the difficulty bomb? Since the fork that wasn't bombed would survive, it would be considered the original, wouldn't it?

How is it defined? Like somebody releases a new version with the bomb removed. Most miners adapt it. Why is it then a fork? Because the bomb is a major feature? Otherwise, wouldn't ever software upgrade be considered a fork?

If I understood the conversation from the last dev call correctly, they're talking about combining EIP-1559 and the difficulty bomb fix into the same fork specifically to avoid this

"original" and "fork" have no meaning inside the ETH universe. These terms are used in the outside social milieu.

Stuff like this is exactly why I have chosen to allocate 0% of my cryptocurrency investments into Ethereum. Such a cluster

If I understand right the ETH chain will be integrated as a shard. This would mean it would just live on inside the new Ethereum ecosystem until it is ice aged - so one could say it's not a fork but a wrap of some sort.

Unfortunately not a single resource I found actually describes this process in an easy to understand fashion - pretty amazing considering we are talking about $200+ billion dollars....

Yes but there is already an 'ETH classic'. So there will be three, and the names should actually be 'ETH' (proof of stake), 'ETH for Miners' (proof of work), and 'ETH classic' (existing, for another group that didn't like a previous important change).

Reminds me this: https://en.wikipedia.org/wiki/Schism#/media/File:Christianit...

It would be more difficult these days to get much marketcap on the minority chain, now that Ethereum hosts tokens representing a variety of off-chain resources like fiat money and bitcoins. Those things will have to pick one chain whose claims will be honored. Same for some NFT projects. Anything involving data feeds won't necessarily support the minority chain either, though in principle they could.

This is weird. So NFTs don't do anything? I mean you can always fork ETH and then transfer ownership of all NFTs to yourself on the fork. So basically the value proposition of NFTs is tied to the chain they exist on. It's not the NFT that gives it value, it's the combination of NFT+ETH that gives it value. That's extremely absurd.

Vitalik and friends can just sell their coins on the proof of work chain and thus make it unattractive.

However you could change the protocol / software and block the addresses of people who don't support the PoW chain.

It would be a bit messy.

However if you are sitting with a massive ETH mining rig, you would like to do something with your hardware.

> However if you are sitting with a massive ETH mining rig, you would like to do something with your hardware.

You would like to profit from your hardware*.

That's an important distinction, as what make you profit from your hardware, isn't running it on ETH in particular, it's running it on a cryptocurrency which is used. They could run it on ETC currently if they want to keep mining, but they don't as ETH represents more value. So whether they keep running on the fork is more about whether the cash will be there to mine on it and my impression is that it won't be there.

Currently there's quite a big smear campaign on the environment impact of cryptocurrency on the world. This is important for many people, thus will impact whatever they decide to do.

There was also always a preference toward the forks supported by the official developers. It's normal, as long as we have no reason to doubt their decisions, why change? In this case this is even more true as it's not questioning the decision, the intention is clear that it's about the miner hardware.

Let not forget also that this fork will be supported by peoples that proved that they could orchestrate a 51% attack for their own gains... which is pretty much the fear that was always there on cryptocurrencies...

There's no reason to believe people would go toward a fork, thus there won't be much value in mining on that fork versus any alternative cryptocurrency. If I was a miner, I would start selling my graphic cards... they are still sold for quite a bit of money currently and it's just a matter of time before the price crash and it become easy again to buy a new graphic card.

So one can ask. Who has the biggest say in a cryptocurrency?

The owners of the coins, the developers, the miners?

Actually I would argue it would be the exchanges: What the major exchanges choose to trade as Ethereum is what will continue.

It's a big hodge-podge and that's what makes it great.

Ultimately society/the market decides which version is worth more.

Do they have enough ETH to run a 51% attack with a PoS system?

Unlikely as there is an ice age built-in to eth1 which will steadily increase the time between blocks until none are produced. There should also be a hard fork which includes code which just stops the old chain on block x. Anyone that wanted to continue with the old chain would need an additional hard fork.

This will not be the first time this has happened to eth.

They manage to avoid it most of the time via time bombs in the code

Nope, wont happen. https://medium.com/dragonfly-research/ethereum-is-now-unfork...

Excellent work on prysmatic! Can you clarify about this "quick merge" proposal? Would the idea be to shift "ETH1" validation to the ETH2 Beacon nodes? At first two chains would be running and be validated by the same nodes but then much later the final merge of Beacon and ETH1 chains would occur? Is that the plan?

But this is an odd conundrum. If Ethereum is decentralized, any such change must be achieved by a vote. Circumventing such a vote is detrimental to the eco-system in the long term, wouldn't you agree?

How did those 6 Billion USD worth of value move to the beacon chain while it is not connected to the old chain?

you intentionally stake your ETH on the new chain. So 6 billion from individuals and organizations that intentionally "moved" their coins.

They were burned on the old chain

Interesting.

I see how that is possible. Anybody can probably burn ETH they own by sending it to an invalid address or something.

And how was it created on the beacon chain? Was it simiply hard programmed into the software "We start with a bunch of ether at the following address: ..."?

There is a well-thought out and well-designed process for all stages of this. You can read up on that on the https://ethereum.org/en/eth2/

Been poking around for a while but did not find the part where it explains how ETH moves from the old chain to the new prior to the merge.

Do you know where it is described?

It involves something called the staking deposit contract:

https://ethereum.org/en/eth2/staking/

https://launchpad.ethereum.org/overview/

Picking any active validator on the Beacon Chain, we can see that there is an Eth1 deposit and a corresponding Eth2 deposit:

https://beaconcha.in/validator/91676#deposits

https://beaconscan.com/validator/91676#deposits

If you're interested in the technical details, see:

https://github.com/ethereum/eth2.0-specs/tree/dev/specs/phas...

You send your ETH to a specific contract, from which it can't be moved (whether you want to call that "burning" is up to you, but it's probably a bit confusing). Each validator must also run an Eth full node, and uses the state of that contract to determine who can act as a validator on the Beacon chain.

Are there any plans to do staking Polkadot/Cardano style? I'd like to stake my ETH but have no interest in running a full validator

This is the first time I've seen someone praising Cardano without me looking for it.

Could you elaborate?

Not OP, but I can give some elaboration on why Cardano is a really interesting project:

1. Staking in Cardano does not lock up your funds, and is really decentralized, getting more every day. Because of that, more than 70% of all coins are currently staked. This brings less volatility to the price and ensure longevity of the chain. The interest is not too big too, everything seems as sustainable as it can be.

2. Cardano supports assets natively instead of having them just being backed by smart contracts (this is/will be still possible though), which means that there are no extra gas fees to send/receive stablecoins or assets other than Ada.

3. Babel fees are part of 2, it means that the transaction fees can be paid by native tokens, not only ada. This is a huge improvement from the BSC or Eth, where you need BNB/ETH to pay transaction fees even if you're only dealing in stablecoins.

4. Smart contracts are finally coming in Q2, which will bring DeFi to Cardano (this explains the price mooning in late feb, early march, because of the mary hardfork).

5. Forks and updates in the chain are painless, last one I didn't even had to send money to a new wallet or anything like that.

6. New developments and features to the chain and ecosystem are being "baked into the protocol", in a sense. Project Catalyst is the first step for this, financing some projects, but in their roadmap they plan on giving away all control of the direction of the coin, and only work as a service provider to the ecosystem using the same infrastructure as all the other projects.

I'm really bullish on this coin, not only on price but also on the kinds of problems it can solve. Can answer more questions if you have any :)

Is Project Catalyst a Cardano's version of The DAO? I'm newbie in crypto.

It's possible. See, for example, Rocket Pool:

https://www.rocketpool.net/

Already possible

Sounds like trying to land a helicopter on a boat in rough seas.

First, they need to keep the miners happy until the second before they are turned off.

Second, "escalating" the deployment of code seems like an always bad idea for such a massive system.

How does this effect the viability of L2 side chain solutions?

Eth2 and L2 are independent solutions that are both happening. After Eth2, low tx throughput on Ethereum will likely still cause transactions to be expensive enough that users strongly prefer to use L2.

L2 can scale Eth2 as well. In fact Vitalik mentions often that scaling can go 100x with Eth2 and 100x with L2 for 10000x total scaling.

Reply on news.ycombinator.com