Eth 2.0 AMA Phase 2 | Vitalik: Eth 2.0 has no unsolved research challenges

Editor's Note: This article comes fromUnitimes, translated by Jhonny; the original text comes fromEthhub, reprinted by Odaily with authorization.

, reprinted by Odaily with authorization.

The Ethereum 2.0 research team conducted a 12-hour AMA (Ask and Answer) event on Reddit on July 15, 2019. This AMA was mainly conducted by Vitalik Buterin, the founder of Ethereum. Authors Danny Ryan, Justin Drake, and Carl Beekhuizen answer questions from Reddit users.

Question: What do you think are the biggest unresolved challenges in Eth 2.0 research before implementing Phase 1 or Phase 2 of Eth 2.0 in the future?

Vitalik Buterin: I really don't think there are currently unsolved research challenges. On the research side, it's mostly "how can we do this more elegantly, with fewer lines of code and fewer edge cases".

Danny Ryan: There needs to be a better understanding of the incentives and various actors that may be involved in stateless and highly abstract execution models. Both the Ethereum Foundation's eWASM team and ConsenSys' Quilt team have done a lot of work in order to better understand related designs and actively build prototypes to review related ideas.

Ethical-trade: I didn't intend to ask questions, but I just want to thank you in the name of all investors and enthusiasts. While there will always be some who will complain, most of us are very supportive of your efforts. We are so grateful for the decentralized future you are building.

Issue: Previously, the unofficial launch date for Phase 0 (ie the Beacon Chain) was January 2020. Do you think this date is realistic?

Carl Beekhuizen: Thanks for noticing that this date is unofficial :) Launching (beacon chain) early next year is my personal goal, but there is still a lot to do before then! Things we need to get done: long-running testnet (however that is defined), formal verification of the deposit contract, and a ready (Eth 2.0) client, but for now it looks like everything All will be completed on time. We also don't want to rush out a buggy software client just to meet some arbitrary date.

Danny Ryan: The various (Eth 2.0) client teams are doing a great job and continuing to push the envelope. I look forward to exciting progress in the coming months, but I also anticipate that the final journey may be long. Early 2020 (beacon chain launch) is realistic and our goal.

dmihal: My biggest concern with Eth 2.0 is that it will kill composability. Don't most dapps (decentralized applications) eventually choose to build in the same shard chain? Like being on the same shard chain as MakerDAO so they can use Dai?

• Justin Drake: Composability between shards is certainly uncharted territory, but there are reasons to be optimistic:

• In order to facilitate cross-shard transactions, all shards are designed to be homogeneous (this is different from Polkadot or Cosmos);

• There are design patterns that abstract the boundaries between shards. These design patterns will be easier to use in programmable execution engines.

• Each shard is designed to be friendly to achieve "fast optimistic finality" due to the verifier's proof of the shard, which is similar to the block confirmation on the current Eth 1.0 chain. That said, in practice, all shards can act as one logical blockchain due to the fast finality of individual shards.

The UI (user interface) layer can also abstract the boundaries between shards.

nicknle: I just wanted to say...the whole team and project are amazing. You have to deal with a lot of noise from the market, but it is really inspiring to see your unwavering optimism and positivity for the good of the world and pushing the boundaries of technology, finance and society. Keep up the good work, guys.

Danny Ryan: Thank you! I'm just trying to keep things moving forward and not be too concerned about market volatility and all that noise.

Question: Why are there so many teams building Eth2.0 clients? I know the importance of client diversity, but don't you guys think 6 clients would be more stressful? Supporting this many clients also requires allocating more resources financially. Which client do you think will be Geth and Parity client for Ethereum 2.0?

• Justin Drake: There are a few things to note about client diversity:

• There are not only 6 clients currently under development, but 8;

• I expect consolidation in clients - some clients won't survive 2020;

• I expect clients to specialize -- a client will focus on browsers (like Lodestar), or focus on resource-constrained devices (like Nimbus), or focus on enterprises (like Artemis), or focus on For prototyping (such as Trinity) and so on;

• At (beacon chain) startup, at least two clients are required to be production-ready. I expect the first mover advantage to be strong;

To some extent, all of the above situations once happened on the old Eth1.0 chain.

Vitalik Buterin: We definitely don't want to see a duopoly! As far as I am concerned, the ideal situation is that no one client will have more than 1/3 of the entire network's nodes, so that any software error in any client will not cause the entire network to stop functioning. Although if no single client has more than 1/2 of the entire network, that's fine and gives us the same benefits. I expect a power-law distribution of nodes for the network, and there is a good chance that some of these clients will not be heavily used on mainnet.

Danny Ryan: I think the reason why so many clients will do this hard work is because Eth2.0 is not only a purely technical challenge, but also an opportunity to leave a footprint in the field of Ethereum and cryptocurrency. I'm glad to see that there are so many excellent teams doing this difficult work, but recently, I'm more focused on finding contributors to do value-added work beyond the core client implementation, such as formal verification, protocol academic Analytics, testing, light clients, web3 interfaces and developer tools, a validator client capable of plugging into any underlying node with a strong user experience (UX), and more.

Question: How do you counter the criticism that Eth 2.0 is too complex?

Vitalik Buterin: Over the past year, it has become quite simple. If you count the words of the Eth2.0 specification, the current specification will have fewer words than the Ethereum Yellow Paper. There are many things in Eth 2.0 that are much simpler than Eth 1.0. But there's certainly some complexity that remains, and I'm very focused on minimizing that complexity.

Justin Drake: Although the road of research is a bit tortuous and difficult to keep up with, the final specification can be said to be simple and clear. Expect more educational material that emphasizes the simplicity of the current Eth 2.0 design. Phase 0 needs to specify about 1024 lines of code (assuming SHA256 and BLS12-381 primitives). I expect Phase 1 and Phase 2 combined to be 1024 lines of code (assuming WASM as a primitive).

Question: What happens when I stake 32 ETH and get slashed? If my pledged ETH is slashed to less than 32ETH, will I be expelled from the ranks of validators?

Justin Drake: Validators are evicted when they are slashed. There is another eviction mechanism where you are also evicted when your accumulated (non-slashing) penalties bring your balance below 16 ETH.

Danny Ryan: You will also lose a certain amount of ETH. The minimum loss amount currently set is 1 ETH. If you have committed other behaviors that meet the confiscation conditions in the recent period, there will be additional penalties. If more validators have been slashed recently, you lose more ETH. And if about 1/3 of the verifiers were slashed recently, the maximum penalty will occur, that is, you will lose all the pledged ETH.

Question: What are your thoughts on the current plan for migrating from Eth1.0 to Eth2.0?

Vitalik Buterin: The current method is to transfer Eth1.0 as an execution environment into the Eth2.0 chain. In practice, this means that we will need to perform a hard fork on the Eth1.0 chain to adjust some gas fees (the gas fee for reading storage or account opcodes will rise to 2,000-10,000), after which At some point, we will transfer the state root of the Eth 1.0 chain to the Eth 2.0 system at a certain block height. I do think the gas fees for opcodes that read storage/accounts will go up, and contract developers should be aware of this and plan accordingly.

Question: Will staking be easy for someone who holds 32 ETH but has little knowledge of crypto? Can "ordinary people" also participate and reap the benefits? Does staking pose a risk of losing ETH to participants? What I want to know is, unless we do something "deliberately" to damage the network (such as taking your validator node offline due to a problem with your internet provider), can we participate in staking without risk?

Justin Drake: I expect that there will be an industry for staking in the market, such as infrastructure including staking pools (centralized staking custody services such as Coinbase, and other decentralized staking pools), and There are some plug-and-play "validator in a box" solutions. In terms of risk, the penalty should be small for validator nodes that sometimes go offline for short periods of time.

Question: Assuming that Ethereum reaches 1 million TPS, the size of the ledger will grow by 1TB every day, is there any solution for this? What are your projections for ledger size growth? What is the corresponding solution?

Danny Ryan: Eth 2.0 with sharding is expected to handle about 10 MB/s of data availability. This is the consensus data in the shard chain, and the data is guaranteed to be available within 2 weeks or so.

The current way of dealing with state and state execution is to adopt a "stateless" method, that is, the block must include Merkle witnesses of the relevant state to complete the execution of the transaction. This would reduce the amount of state any consensus node would have to store, but it does raise other questions about the size of the state, who stores the state, how users get it, etc.

In fact, the entire state enforcement is abstracted so that we can try multiple solutions to deal with this problem (such as state rent, letting users/dapps store their own data, etc.). State rents, which Alexey Akhunov has been working on for the past year or so, will likely come into play.

Question: I've heard a lot of hype about staking rewards, what is the penalty when a validator is slashed? Will the network have different penalties for different misbehaviors (such as the verifier client being offline, or repeated signatures, etc.)? Is the current "60.8% reduction for 18 days offline" still valid?

Carl Beekhuizen: I think you are confusing slashing (slashing) with inactivity leak (punishment for validators not being active).

Inactivity leak means that when the validator node is offline for 18 days and the beacon chain stops the "finalizing" process, the amount you stake will be "reduced by 60.8% within 18 days";

https://notes.ethereum.org/9l707paQQEeI-GPzVK02lA#Why-are-the-Casper-incentives-set-the-way-they-are

Slashing means that when the verifier conducts malicious behavior, the amount of its pledge will be reduced. Assuming there is nothing wrong with the client software, slashing is very unlikely to happen to you. The minimum slashing penalty is 1 ETH, but the penalty grows with the number of people slashed at the same time as you. For more information on penalties see:

Question: What is the most exciting research currently being done on Eth 2.0 research progress?https://ethresear.ch/t/layer-2-state-schemes/5691Vitalik Buterin: Recently I have
most exciting research. Karl Floersch et al. in Optimistic Virtual Machine (OVM)

Research on this aspect is also very interesting.

Question: Are researchers satisfied with the current economic incentive mechanism of Eth 2.0?

Vitalik Buterin: I don't think we need to worry about specific numbers at the moment (Note: V God refers to the total number of ETH pledged in the beacon chain and the corresponding annual increase rate of ETH and the annual rate of return of the verifier), regardless of the verifier Whether the corresponding rewards are sufficient, the network will start.

Question: How many Eth 2.0 client teams must deliver a completed client before the Beacon Chain goes live?

Justin Drake: At least 2, ideally 3.

Question: I'm considering staking during phase 0, but I'm a bit worried about inactivity leak (penalty for inactivity). I was wondering how can I stop/suspend being a validator if I change my mind? I'm asking this question because in some cases I might be offline for a while and I don't want my funds to keep dwindling because of it.

Vitalik Buterin: When you are offline, you are only penalized significantly if more than 1/3 of the other validators are also offline at the same time. Other than that, the offline penalty is very small, as long as you are online for more than 50-67% of the time, then you will be in a net profit state (excluding computer costs, etc.). This incentive mechanism is very loose, and the goal is to encourage amateur participants to join in and promote decentralization.

Justin Drake: The way to "pause" being a validator is to first withdraw (this may only take half an hour, but it may also take days/weeks as there is an exit queuing mechanism), and later (want to re-participate in validating ) to activate it again.

MysticRyuujin: Do the various Eth 2.0 client teams feel that their client implementation will be robust enough, secure enough, and easy to use enough, so that nerds like me can safely run the node software they develop and pledge 32ETH on it, And will not be fined or lose ETH because of bugs in the client?

My biggest concern is that I don't do malicious things myself, but lose money because of bugs in the client. I've been running Geth, Parity, Trinity, and EthereumJ (Harmony) nodes for years, and these clients have been around for far longer than any Eth 2.0 client, and they still have issues, some bug, and sometimes crashes.

Vitalik Buterin: I hope so! A key point of the (Eth 2.0) incentive design is that the penalty (whether for being offline or for being slashed) is only high if many other validators are offline at the same time. So any bug that doesn't attack all nodes at the same time will only cost you very little money.

Question: I know it's early days, but is there a rough estimate of when the phase 1 and phase 2 specifications will freeze?

Carl Beekhuizen: Phase 2 is not yet clear, but I expect the phase 1 specification to freeze in the fourth quarter of this year.

Problem: 1024 shard chains, which means there are more than 131,000 validator slots... What if there are not enough validators when the shard goes live?

Carl Beekhuizen: Since there are 1024 shard chains, 128 validators are required in each committee, so at least 131,072 validators are required in order to crosslink each shard. If the number of validators is less than this number, some shards are skipped, so that each committee has 128 validators.

Danny Ryan: The system can handle down to 64 validators quite naturally. In this case, system security will be significantly reduced, but the protocol can continue to advance technically.

Question: Please briefly explain why it is necessary to develop a new chain instead of continuing to develop the current old chain?

• Justin Drake: Building an Eth2.0 chain on top of an Eth1.0 chain would be a poor design decision for the following reasons:

• We will be limited by the gas limit of the current Ethe 1.0 chain, which will seriously affect the performance of the new chain (for example, it will not be possible to achieve 1024 shard chains and pledge 32 ETH for verification);

• We will mix consensus layer and application layer. This means that the consensus layer will be subject to the DoS vector of the application layer (such as high gas prices);

• We will be at the mercy of the old chain's EVM (Ethereum Virtual Machine), and it is notoriously difficult to securely program complex contracts into the old chain's EVM.

• Eth 1.0 chain cannot support BLS12-381;

There are other reasons as well.

Question: I don’t know if my understanding is correct. Does it mean that the more verifiers there are, the longer it will take to finalize?

Carl Beekhuizen: In the Eth2.0 chain, more verifiers should not lead to longer finalize time. By using BLS signature aggregation and by forming validators into committees, we are able to accommodate hundreds of thousands (or even millions) of validators.

Question: I know that it is estimated that approximately 10 million ETH will need to be staked to provide sufficient security for the network. Suppose a large organization acquires more than 50% of the pledged ETH and uses it to launch an attack on the network, what will happen to the network? What I understand is that in this case, the party with less total stake will be slashed, so can this malicious attacker effectively kill the entire network?

Carl Beekhuizen: If a large participant wants to obtain a large amount of pledged ETH to control the entire network, he will need to purchase more than 2/3 of all circulating ETH, and when he buys, it will push the price of ETH up, so that First, the cost of this attack is very high; second, you will be fined only when your behavior is proved to be malicious by 2/3 of the verifiers. One of the great things about the PoS mechanism is that this kind of attack can be handled very well. As a community, we can join forces to fork the malicious attacker from the current chain through a hard fork, so that the malicious attacker has no voting rights. The attackers just spent a lot of money temporarily shutting down the network.

In the PoW chain, if the attacker purchases an ASIC mining machine with 51% of the computing power, there is no way to remove the attacker from the system.

Question: As an ETH miner, my main concern is how to prepare for Eth 2.0. Any tips or advice on how to continue contributing to (and monetizing) Ethereum?

Justin Drake: There will be no mining in Eth2.0. You may be able to lease out the computing power of these GPU mining rigs through platforms such as Golem.

Question: When will the issuance rate of ETH drop significantly? For example, the increase rate has dropped by 10 times?
Vitalik Buterin: When the PoS chain starts to host the PoW chain for security (this may happen in phase 1 or phase 2), then it is safe to put the PoW

The mining reward of the chain is reduced by about 4 times. When the PoW chain stops running completely, the issuance rate will be further reduced.

Question: Which currently running project/product could improve Ethereum's scalability in the short term?

Justin Drake: Layer 2 expansion solutions (Plasma, Rollups, state channels, etc.) can improve scalability in the short term. Unfortunately, building a Layer 2 solution in 2019 still requires a lot of technical complexity, but it's not impossible!

Question: Is the issuance rate of Eth2.0 enough to incentivize validators due to competition from other DeFi products? Will the increase rate be adjusted accordingly in the future?

Vitalik Buterin: As far as I know, the annualized interest rate of staking ETH on the Compound platform is 0.02%, so we have a great competitive advantage. And the design of "Staking ETH on Compound platform can get 3% interest rate, lending DAI interest rate is 6%" is misleading, because 6% is based on USD (US dollar), and 3% yield is based on ETH's. But with the emergence of more and more staking forms (such as Plamsa, Trubit and other margin games), the pledge rate of ETH will rise.

Question: During phase 0, can Eth 2.0 coins be transferred between wallets? If so, are you concerned that the exchange will list Eth 2.0 coins?

Carl Beekhuizen: The transfer of Eth 2.0 coins will only be possible in phase 1, when exchanges will list Eth 2.0 coins. Listing on exchanges will help maintain the parity between ETH coins and Eth 2.0 coins.

Question: What is the current design for moving Ether tokens from Eth 1.0 to Eth 2.0?

• Carl Beekhuizen: It depends on what you mean by "Ether token":

• The verifier can send 32 ETH to the deposit contract, and then the deposit will be transferred to the beacon chain, so that the verifier can participate in the verification on the beacon chain;

• For people who just want to transfer ETH from the Eth 1.0 chain to the Eth 2.0 chain, the design of this aspect has not been finalized, but there may be a dedicated bridge between the two chains, and it may be possible to achieve this through a mortgage contract. First, the purpose;

It is too early to talk about the transfer of ERC20/721 tokens between the two chains. If the Eth 2.0 chain has an execution engine like the Eth 1.0 chain, then this transfer will be very convenient, but even without an execution engine, it may be possible to transfer it by simply copying the state root of the ERC20 token.

Question: Sorry for the stupid question, but I haven't paid attention to Ethereum 2.0 for several months. Is there an upcoming release date? Regarding staking, how much ETH should be pledged? Can we now know the corresponding rewards? Which aspect of Ethereum 2.0 are you most proud of? Do you think Binance Chain will be a competitor to Ethereum? Thank you in advance for your time and effort!
Justin Drake: Phase 0 may start as early as January 3, 2020. I think it's likely to be in the first quarter of 2020. Need to pledge 32

ETH, the validator's reward depends on the number of all validators, the performance of individual validators themselves, and the gas market. I find it unsurprising that validators earn around 10% per annum. I feel most proud to have been a part of building this infrastructure that I believe will revolutionize the world in a positive way. Binance Chain is not a competitor of Ethereum.

Question: When Sharding starts, will there suddenly be 1024 shard chains? Or will the number of shard chains gradually grow with usage? I feel that starting a large number of shard chains at the beginning will lead to a lot of unused space/capacity.

Carl Beekhuizen: All 1024 shard chains will be started at once. Gradually increasing the number of shard chains would (probably) introduce unnecessary complexity. Shards with a lot of unused capacity will have lower gas prices and thus attract more users.

Question: I thought that the Eth 1.0 chain only needs to be upgraded to the Eth 2.0 chain, and there is no need to exchange tokens on the old chain for tokens on the new chain. Will the number of Eth 2.0 coins in circulation depend on how many people convert Eth 1.0 coins to Eth 2.0 coins?

Carl Beekhuizen: The Eth 2.0 chain is a complete change to the Eth 1.0 chain, so it is better to implement a brand new chain than to upgrade. The number of Eth 2.0 coins in circulation does depend largely on how many people convert Eth 1.0 coins to Eth 2.0 coins. Of course, the Eth 2.0 chain will also issue new coins as rewards for validators, which will also increase the supply of new coins. But eventually, all Eth 1.0 coins will be converted into Eth 2.0 new coins.

Question: As some other blockchain platforms have recently achieved thousands of transactions per second (although these platforms have not implemented sharding technology), can you imagine that a single shard chain in Eth 2.0 can also be compared with these "new & Fast Blockchain Platforms" Compete? If not, how do these blockchain platforms achieve their high performance? Could the same be true for Eth 2.0?

Vitalik Buterin: Are these blockchain platforms really capable of processing thousands of transactions per second? Usually my "gut feeling" is that they're boasting 10x the throughput they really are.

If a chain is not active, then I don't trust any numbers on that chain, because high numbers that are easily achieved in the lab don't hold up in practice. During the test period, Ethereum itself can also achieve 500-1,000 TPS, and Plasma can also achieve hundreds of thousands of TPS during the test period. If these blockchain platforms are really able to achieve more than 1,000 TPS in practice, then why are there so few nodes running on the server?

Question: Will phase 0 improve scalability? Through Sharding, how much transaction volume will Ethereum be able to process per second?

Carl Beekhuizen: The goal of phase 0 is to track validator state and generate randomness, so it doesn't make much sense to ask about its scalability aspects.
In terms of TPS, there is no clear answer. A basic calculation is that if each shard chain has the same throughput as the current Eth1.0 chain, then the entire Eth2.0

The throughput of the system will be: 16*1024=16,384 TPS (assuming there are no cross-shard transactions).
However, the above numbers are meaningless, because the Eth 2.0 network will be compatible with Layer 2 expansion solutions (such as Rollups and OVM, etc.)

Used together, this promises very high throughput.

Question: If Eth 2.0 introduces the PoS mechanism, will there be some legal obstacles that need to be clarified? Might its token be considered a security?