ETH2

Hello ETH2.0 AMA

Hi there 🙂

ECN finally lands on Medium, with our first Ask-Me-Anything summary script. The #Hello, ETH2.0# AMA lasted for 2.5 hours from 21:00 pm to 23:30 pm (UTC+8) last night, livestreamed on WeChat in English & Chinese and also on Twitter in English documentation. It’s really informative and insightful as an Ethereum 2.0 primer, with five eth buidlers from the wide ecosystem spectrum.


Guests

  • Afri Schoedon, Pioneer of ETH2 Multi-Client Testnets

 

Topics

Eth2 launch roadmap / multi-client testnets / layer2 scalability / PoS / community building


 

Q&A

Note: the answers from Steve, A Jian and Jason are translated from Chinese into English by ECN. In case of any discrepancy, the Chinese version shall prevail.

Q1 Ethereum has been interesting and exciting global users with its rapid growth. Now with the long-anticipated milestone getting closer, the calls for Ethereum 2.0 are growing louder. Could you explain what Ethereum 2.0 looks like in plain words? How does it differ from current eth1 main chain?

Raul:

To me, eth2 is a total redesign of the Ethereum blockchain to maximize for decentralization, censorship resistance, and scalability at the same time. It takes all the earnings we have had over the past few years and turns them into a much more sustainable approach to achieve its original goal of becoming a “world computer”.

A Jian: (translated)

Danny Ryan said “eth2.0 is built for Ethereum, and finally it will become Ethereum itself”. Ethereum 2.0 is to build the base layer of the next-generation Ethereum blockchain ecosystem.

Eth2 differs from the current Ethereum in two respects: 1) use PoS (Proof of Stake) algorithm to drive the blockchain operation; 2) to improve scalability by “beacon chain + multi-shard chain” architecture.

In plain words, blocks in eth2 are not produced by PoW miners, but confirmed and finalized by stakers who deposited their ETH. Besides, “beacon chain + multi-shard chain” architecture separates the state inside PoS and that of transactions. Instead of executing transactions, beacon chain will handle finalization and communicate with shards. Shards could be considered as complicated blockchains resembling the current eth1 chain, which excute transactions.

 

Q2 It’s been years since Ethereum 2.0 plan was initiated. There are many opinions and wild guesses occurring across the community about its official launch time. Could you share with us the current development progress of Ethereum 2.0? What are the major challenges to be solved? And when do you expect Ethereum 2.0 to go live?

Afri:

It’s good to talk about launch roadmap without coming up with a concrete dates. One could say it’s January the 3rd or July the 30th. But the truth is, it won’t be neither of the dates. We cannot tell when the mainnet genesis will happen. This is because there are so many moving parts.

But I can give a realistic view on the open issues and milestones. First of all, all clients have to implement the final version of the spec, v0.12.1. Currently only one client has a release out currently, and the other four clients are still working towards a stable release. We will see a first multi-client testnet with the final v0.12.1 spec be launched potentially end of this month.

If this goes well, we will be able to prepare the public multi-client testnet launch under mainnet conditions early July. This is the long awaited milestone, sometimes mistakenly referred to as “official” multi-client testnet. But if it goes well, we can do the final public testing in July.

There might be a couple of smaller developer testnets on the side, but the main focus should be the public one. If the public multi-client testnet turns out to be stable over the period of two or three months, we can start working towards a real mainnet launch.

First of all this requires a deposit contract. A ceremony to determine the contract that is to be used as the one and only bridge to ETH2 deposits needs to be held. How this will look like is not entirely defined yet. I know Carl is preparing that. We should ask him soon. I expect the earliest possible date for such a ceremony would be end of August or rather mid September.

Once the ceremony is done and the contract is deployed, technically the rush into ETH2 has officially started. Even without official launch date, people could start making the first validator deposits on the ETH1 chain as this is an entirely permissionless process.

I would suggest to schedule the minimum genesis time for the beacon chain mainnet to be exactly 90 days after the genesis time of the official multi-client testnet. This would give us 8 weeks of testing and 4 weeks of buffer time in case we would have to fix something.

Regardless of the deployed deposit contract, we can always push back the genesis time for mainnet if needed. But given the steps outlined above, the earliest possible mainnet genesis time would be somewhere in October. This does not take into account that we need at least 16k deposits before we see a mainnet genesis event. This could additionally take some weeks to ramp up.

So, to sum up…. November if everything goes well. October if we are very optimistic and pushy about the date. Pessimisticly speaking, it would be 2021 in case anything needs to be patched in the spec and adapted by all clients…. I don’t think we will launch in December.

 

Q3 (1/3) Interoperability has been one of Ethereum’s edges. The question for Raul is that by far is there anything more happening around client interoperability? And why client interoperability matters so much to eth2?

Raul:

Client interoperability matters so much because that’s the way we can catch critical bugs in a single implementation. For example, if all of eth2 ran using our Prysm implementation, we would never find a critical problem regarding validator rewards and penalties that we have found thanks to client interoperability.

 

(2/3) The multi-client testnet is considered to be the ultimate hurdle before we see real eth2. As the initiator of eth2 multi-client testnets Schlesi and Witti, could Afri share with us the primary targets of multi-clients, and Witti’s current running state?

Afri:

The first phase of Ethereum 2.0, the phase 0, is the beacon chain. The old ETH1 clients will not be able to run a beacon chain. For the first time, a variety of newly implemented clients will be working together on a brand new blockchain with a new, unique approach to networking and consensus.

Before such a mainnet can be launched, we need testnets that mimic mainnet conditions as good as possible. This requires us to have stable, long-term, and persistent testnets up and running that are supported by not only one client but multiple clients, ideally, all clients. The Schlesi testnet was one of many steps in that direction, the Witti testnet is another, and soon we will launch Altona.

Witti is running fairly stable. It currently has seen around 150k slots and good livelyness. The participation rate of validators is hovering at 71%.

All clients still experience bugs and introduce new optimizations. I suspect we still need a lot more multi-client testing.

 

(3/3) One more question for Afri. It’s said that the final multi-client testnet has to remain stable for at least 2 months before beacon chain release, so what are the indicators for stability measurement? When is the public multi-client testnet expected to go live?

Afri:

Indicators for stability are: no consensus issues and no permanent forks obviously, all clients should agree upon the correct chain. Also, we need to finalize proper networking, there are a lot of rough edges and hiccups between the different clients. I hope we can ban ENR into the backend of the clients before we launch and have proper multi-address format for wiring nodes.

Further indicators are attestations and validator activity. If we maintain above 80–90% activity and the chain stays in perfect liveliness with high finality, we should be good to go.

As I outlined earlier, the soonest possible launch date for a public multi-client testnet would be mid-July.

 

Q4 (1/2) After the official launch of eth2, with the support of layer1 solutions like PoS and sharding, Ethereum is heading into a new scalability epoch.The adoptions of Rollup techniques fuel the transaction speed while reducing the transaction fees significantly, bringing much better experience to users. As the CTO of Loopring, the first and ever decentralized protocol with working ZK-Rollup, could you please describe how is ZK-Rollup designed to optimize our user experience?

Steve: (translated)

Zk-Rollup is a scalability solution applied to Ethereum, often known as a layer2 solution. It can support more transaction processings with faster speed and lower cost. It executes all computations off-chain and only needs to submit a small zero knowledge proof (cannot be faked) to verify. Since the proofs will be verified on Ethereum, and sufficient data will be stored to clarify the sate of off-chain accounts, ZK-Rollup will share the same security level as Ethereum.

Loopring chose zkSnark as its zk proof framework (tested by ZCash for years), with smallest verifying data size so far, which means lowest on-chain cost. And zkSnark doesn’t require many off-chain computing resources to generate proofs.

cr: Loopring

There is a relayer system in ZK-Rollup. First, the relayer will organize all account information off-chain into a Merkle tree where the hashes are computed in pairs until we get the Merkle root hash. Given that any change in arbitrary leaf value results in different root hash, the root hash will be the only display of the whole tree’s current state.

Then the relayer will collect a transaction set(transfer or buy&sell)off-chain and generate zk proofs for those transactions. The proof generation process: the last stored root hash R1; every transaction is signed by users; update the Merkle tree according to those transactions, and verify the changes; get the updated root hash R2.

Finally, the relayer submits critical data (e.g. balance), zk proof and updated root hash R2 onto the blockchain. The pre-deployed smart contract will verify if the proof is valid based on the recorded root hash R1 and transaction data input. If it’s valid, the root hash R2 can be updated because the relayer has conducted honest changes to the off-chain Merkle tree.

Here lies the key point of this scheme: it slashes the resources consumption and improve the overall blockchain performance by only keeping data storage and zk proof verification on chain, and moving computations off chain. And more important, the security and validity of processing asset off-chain can be guaranteed by on-chain verification, so that we don’t need to compromise security for better performance. Theoretically, Loopring could bear 2025 tx/s at a cost of $150 per million transactions with the same security level as Ethereum mainnet.

 

(2/2) Could Steve and Jason further share your insights of how Ethereum ecosystem will benefit from the combination of Layer1+Layer2 scalability? And what are the potential driving forces for DeFi?

Steve: (translated)

The real application of layer2 scalability is indispensable for Ethereum ecosystem. Recently some fraud schemes lead to a surge of gas price, and the network has been slowed down, which would be damaging for mass adoption of Ethereum. For example, Reddit has already decided to use Ethereum as its token distribution platform. With the tremendous user base of Reddit flooding onto Ethereum, the traffic would be too much for the network to handle.

Fortunately Loopring has implemented a payment system, Loopring Pay, based on ZkRollup, which was officially launched on June 7th. Now it has already supported 10+ token transfers on layer2 with zero fees, which helps mitigate issues like network congestion and expensive fees. While for DeFi apps, if we can handle transfers on layer2, then the layer1 bandwidth coud be eased.

Jason: (translated)

To begin with, it’s a matter of choice. The rising layer2 scalability solutions endows developers and users with more choices, while avoiding the potential single layer2 failure. Take ZK-Rollup and Optimistic Rollup schemes for example. Seen from token transactions, ZK-Rollup is a better choice for DEX apps and users with faster transaction speed and settlement. Maybe that’s why Loopring adopted ZK-Rollup. While Optimistic Rollup is more friendly with logical implementation of smart contracts, which allows developers to build more delicate DeFi apps. So we saw Synthetix with Optimistic Rollup. More choices popping in, more prosperous the ecosystem will be.

Second, at a higher level, the security and scalability of Ethereum get improved. Layer1, the Ethereum main chain, serves as the basic security guarantee for data. On one hand, scalable layer1 improve on-chain data processing while lowering the possibility of spam transaction attacks. On the other hand, with more solutions deploying on layer2, layer1 can alleviate its scalability burden and focus more on security side. That is, Ethereum will become more reliable while ensuring the user experience, which encourages more users to build on Ethereum, including DeFi apps, where security comes as a priority.

Third, I’d like to talk about value capture and network effects. The combination of layer1+layer2 enables Ethereum to cover the demands of most applications in finance or other spaces. With PoS and booming DeFi, there will be more staking assets on Ethereum, which might eventually make Ethereum the “settlement layer”. Meanwhile, greater network effects will come along with thriving finance, which would be strong positive feedback by driving the further development of Ethereum ecosystem including DeFi.

 

Q5 It’s been known that when eth2 phase 0 happens, Proof of Work Ethereum will turn into a Proof of Stake blockchain where staking economy will be a major player. What are your insights of staking economy? How could various stakeholders participate in eth2 staking, such as average investors, miners, mining pools and DeFi services etc.?

Raul:

The staking economy, to me, will define the security of Ethereum. We wish to make staking easy for most people, including hobbyists that want to run a validator at home. If most staking only happens on a few centralized exchanges, we believe the chain will not be as secure or trustless as it should be.

Most stakers can get started today with experimenting on eth2 by joining one of the public testnets. At Prysmatic Labs, we run a public testnet called Onyx that you can join here: https://prylabs.net and monitor here: https://beaconcha.in

A Jian: (translated)

Eth2 staking distinguishes itself from existing staking model due to its unique protocols. To decentralize, the protocol layer places many curbs on staking, and this is exactly where service providers could come in.

Now when speaking of eth2, it’s more about the coming phase 0, which won’t bring disruptive changes to the whole ecosystem at once. So it’s totally fine to keep watching.

I suggest that if you want to participate in eth2 staking, understanding how its protocols are designed would be helpful for your assessments, no matter you choose to DIY or use third-party services.

Jason: (translated)

On the market side, staking behavior results in reduced ETH market circulation which will definitely stimulate a rise in ETH price. This is what all ETH holders would like to see. As integral parts in Ethereum ecosystem, most average investors, miners and pools are driven by earnings. After all stakeholders need to make their ends meet in the first place, and then they could afford to sustain Ethereum’s growth.

Therefore, players in the Ethereum ecosystem will form a community with shared interests. Everyone is expecting Ethereum to play safe, which potentially empowers and pushing Ethereum forward. There is a funny but vivid expression for how staking economy will affect Ethereum: too big to fail.

A more straightforward way for average investors to participate is purchasing ETH, or staking their ETH via staking services provided by exchanges or wallets. If you want to run your own PoS node, it’s not that difficult.

This is what DeFi servers would like to see, as implementing fractional reserves is a little bit tricky. One thing is that managing your asset with DeFi is still risky. For another, other public chains are relatively smaller in volume, and sometimes staking earnings will fail to cover the loss when its price goes down. However, maybe DeFi servers could see net earnings by staking with extra asset in their pools.

Steve: (translated)

Personally, I’m optimistic about eth2 staking giving rise to the ETH price. And it’s likely more than 10 million ETH will be locked in staking.

According to a staking survey from ConsenSys, with a sample of 287: 32.8% users will run their own validator nodes, and 33.1% will use a third party staking services.


cr: ConsenSys

Loopring is also considering to provide our own staking pool, as users will gain no excess earnings by leaving their assets in smart contracts. But staking will do. However, we promise that users’ assets will still remain non-custodial via smart contracts, which means users have the right to know and control their assets position.

 

Q6 A Jian and Jason are representatives from Chinese Ethereum communities EthFans and Unitimes. We wonder what are the challenges for community builders on the verge of Ethereum 2.0? What could be best practices to engage community members or even more mainstream population into the next generation of Ethereum?

A Jian: (translated)

A big challenge lies in the complexity and durative roadmap of eth2 system, which requires more patience to tell what really matters at the current stage and what really values in a long term. Otherwise, there’s time reserved for us to popularize the complicated eth2 system with more delicate efforts.

I think community builders should keep a good sense of the whole changing ecosystem, to deliver interesting and exciting stuff in this ecosystem. Technologies might be dry, but there are always marvels and novelties happening around the Ethereum ecosystem. The ecosystem as a whole is diversified and fascinating.

Jason: (translated)

The biggest challenge facing the Ethereum community is that currently there are still not many fans and contributors for the blockchain open source technology.

On the one hand, the technical threshold is a bit high, which means people should have a good mastering in various disciplines to better understand blockchain. On the other hand, such technology is too advanced to be popularized and even some are still biased towards it.

Let’s say Ethereum, with the biggest and strongest developer community, still has a long way to go in building Chinese Ethereum community.

One thing is that in the past few years, most users have been wandering in secondary market for earnings, staying out of the Ethereum ecosystem construction. For another, it’s a bit hard for some potential Ethereum learners to get over with the language barrier, less community maturity and rapid iteration of Ethereum technologies. Eth2 or eth1.x might be around before someone could understand eth1. And in this case, we still need to carry forward education on Ethereum and developer community building.

Now we have EthFans, Unitimes, ECN and EthPlanet to work as Chinese community contributors. Eth2 is a brand new start as well as an opportunity for Chinese Ethereum community to thrive. We still get plenty to do such as organizing online or offline seminars, workshops and developing educational resources.


 

Free Talk

Q1.We knew that ETH2.0 will introduce Casper consensus. What’s the advantage of it? What problems will it fix?

Raul:

A specific problem Casper solves is that allows for more secure proof of stake consensus than those which exist out there today. Most other proof of stake chains make the compromise of being delegated proof of stake. That is, the protocol only supports a small number validators (maybe 20–100). Ethereum 2.0 Casper allows for a maximum of 4 million validators, encouraging decentralization and increasing security.

Other protocols do not allow a large number of validators because they cannot scale their consensus algorithm if that happens. Ethereum 2.0’s casper allows you to have scalability and a high number of validators.

A Jian: (translated)

The consensus of Eth2.0 can be divided into two components:

  1. LMD-GHOST fork choice rule, which considers the latest signature of validators as the agreement of the fork chain, and selects the fork with most signatures as the calonical chain.

 

Q2.How can we handle the centralization risk of staking?

Raul:

We can help with this problem by making it easier for people to run their own node. Eth2.0 nodes are easier to run than eth1 or EOS nodes. This will help encourage decentralization and reduce exchanges holding all of the stake.

 

Q3.Has Ethereum ever considered to adopt Lighting Network used by Bitcoin as a layer2 solution? Why is ZK-Rollup better than LN?

Steve: (translated)

We can put it this way: ZK-Rollup is a more efficient and faster solution to achieve the asset consistency than LN. Intermediate nodes are required in LN while ZK-Rollup has its own relayer. ZK-Rollup is not likely to be deployed on Bitcoin because it needs support from smart contract and specific elliptic curve computations.

 

 

The End.


Yet it will never end. See you next time💗