9 ms·

I'm one of the maintainers of one of the implementations of eth2 live today (https://github.com/prysmaticlabs/prysm https://github.com/prysmaticlabs/prysm) and

by rauljordan2020 6y ago

I'm one of the maintainers of one of the implementations of eth2 live today (https://github.com/prysmaticlabs/prysm 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/ 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.

chrisco255 6y ago
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?
3np 6y ago
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?
- proto-n 6y ago
  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.
  - Zetaphor 6y ago
    "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
    - lupire 6y ago
      "attack" is a misnomer. 51% is a majority vote.
      - rualca 6y ago
        Doesn't the majority vote mean that they get to control the cryptocurrency? Sounds like an attack to me.
      - SilasX 6y ago
        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.
    - Spivak 6y ago
      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.
      - 6y ago
SubiculumCode 6y ago
How does this effect the viability of L2 side chain solutions?
- anonymoushn 6y ago
  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.
- mratsim 6y ago
  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.
6nf 6y ago
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?
- ArtTimeInvestor 6y ago
  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?
  - ThePowerOfDirge 6y ago
    False, the difficulty bomb will kill off eth1.
    - JumpCrisscross 6y ago
      > the difficulty bomb will kill off eth1 Wouldn’t it make sense for miners to switch to a new Ethereum chain without the bomb?
      - dylkil 6y ago
        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
        JumpCrisscross 6y ago
        > 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.)
ArtTimeInvestor 6y ago
How did those 6 Billion USD worth of value move to the beacon chain while it is not connected to the old chain?
- delaaxe 6y ago
  They were burned on the old chain
  - ArtTimeInvestor 6y ago
    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: ..."?
    - Erlich_Bachman 6y ago
      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/ https://ethereum.org/en/eth2/
      - ArtTimeInvestor 6y ago
        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?
        michaelsbradley 6y ago
        It involves something called the staking deposit contract: https://ethereum.org/en/eth2/staking/ https://ethereum.org/en/eth2/staking/ https://launchpad.ethereum.org/overview/ 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://beaconcha.in/validator/91676#deposits https://beaconscan.com/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/phase0 https://github.com/ethereum/eth2.0-specs/tree/dev/specs/phas...
TacticalCoder 6y ago
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).
- morelisp 6y ago
  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!
  - norswap 6y ago
    Well, the code never guaranteed that people would not merge the PoW chain to the PoS chain :)
    - koheripbal 6y ago
      At the end of the day, this consensus system will rule by consensus.
  - SilasX 6y ago
    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!
    - vvillena 6y ago
      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.
dvh1990 6y ago
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?
kylebenzle 6y ago
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.

chrisco255 6y ago

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?

3np 6y ago

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?

proto-n 6y ago
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.
- Zetaphor 6y ago
  "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
  - lupire 6y ago
    "attack" is a misnomer. 51% is a majority vote.
    - rualca 6y ago
      Doesn't the majority vote mean that they get to control the cryptocurrency? Sounds like an attack to me.
    - SilasX 6y ago
      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.
  - Spivak 6y ago
    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.
    - 6y ago

SubiculumCode 6y ago

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

anonymoushn 6y ago
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.
mratsim 6y ago
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.

6nf 6y ago

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?

ArtTimeInvestor 6y ago
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?
- ThePowerOfDirge 6y ago
  False, the difficulty bomb will kill off eth1.
  - JumpCrisscross 6y ago
    > the difficulty bomb will kill off eth1 Wouldn’t it make sense for miners to switch to a new Ethereum chain without the bomb?
    - dylkil 6y ago
      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
      - JumpCrisscross 6y ago
        > 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.)

ArtTimeInvestor 6y ago

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

delaaxe 6y ago
They were burned on the old chain
- ArtTimeInvestor 6y ago
  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: ..."?
  - Erlich_Bachman 6y ago
    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/ https://ethereum.org/en/eth2/
    - ArtTimeInvestor 6y ago
      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?
      - michaelsbradley 6y ago
        It involves something called the staking deposit contract: https://ethereum.org/en/eth2/staking/ https://ethereum.org/en/eth2/staking/ https://launchpad.ethereum.org/overview/ 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://beaconcha.in/validator/91676#deposits https://beaconscan.com/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/phase0 https://github.com/ethereum/eth2.0-specs/tree/dev/specs/phas...

TacticalCoder 6y ago

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).

morelisp 6y ago
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!
- norswap 6y ago
  Well, the code never guaranteed that people would not merge the PoW chain to the PoS chain :)
  - koheripbal 6y ago
    At the end of the day, this consensus system will rule by consensus.
- SilasX 6y ago
  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!
  - vvillena 6y ago
    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.

dvh1990 6y ago

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?

kylebenzle 6y ago

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.