Synthetix founder: Why did Synthetix choose the Optimism expansion solution?
source|blog.synthetix.io
source|
Author | Kain Warwick
The rise of DeFi has sent a clear signal: Ethereum needs to achieve expansion before ETH2.0, but to maintain composability in the DeFi ecosystem, it needs to coordinate on the same expansion plan.
Fortunately, the market is very good at predicting and solving such difficulties, and we have already launched a lot of expansion technologies. We've had conversations with teams on several major scaling variants, and after evaluating each of them, I believe Optimistic Ethereum is the scaling solution most likely to reach consensus in the community.
This article will detail why I believe so, and I will also respond to the concerns in the community about launching Optimistic Ethereum. It will also give the design of the governance process needed to implement the Synthetix variant of OVM on the mainnet, and explain when another scaling technology can replace Optimistic Ethereum. This article aims to ensure that community members have the information needed to reach a consensus regarding the migration of Synthetix to Optimistic Ethereum.A rollup-centric ethereum roadmap[If you haven't read Vitalik's article "Ethereum Roadmap Centered on Rollup" (
), this article is about its discussion, but Vitalik's is top-down, and mine is bottom-up, focusing on what DeFi projects need to expand, and why Optimistic Ethereum can meet these requirements. ]
social consensus
When I first read about Optimistic Rollups, our gas fees were not that high. It looked like an elegant solution to both a technical and a social problem, but I've realized that no matter which scaling option we choose, we'll need to collaborate with other projects. Uniswap’s Unipig demo gave Optimism a great opportunity to gain community consensus as it partnered with one of the most well-known projects on Ethereum. Therefore, I chose to participate in this solution, not only to provide feedback to Optimism on how to solve the difficulties unique to DeFi, but also to assist them in leading social cooperation in the community.
secondary title
technical factors
For Synthetix, Justin Moses is both its blessing and its curse. He has fostered a no-compromise, rigorous engineering culture at Synthetix, but he dislikes cognitive load to a degree uncommon outside the mollusc world. This means that deploying Synthetix to Layer 2 requires a delicate balance between minimizing risk and reducing any modifications to the codebase. Running two parallel, different codebases on Layer1 and Layer2 during a migration is something we would never agree to because the migration won't even happen; and if it means rewriting contracts in another language , it is even more unlikely to happen.https://twitter.com/kaiynne/status/1166234541616316417?s=21Synthetix is one of the most complex smart contracts built on Ethereum, which undoubtedly increases the difficulty of maintaining different code bases. This is us porting the Havven network to EOS.
first discovered in a failed attempt.
We also need to prove to the community that this technology is feasible and worthy of our further investment of resources before trying to build consensus around it as our scaling solution, and the OVM transaction demo also helps to reinforce this. However, there are still community members who have reasonable concerns about this method, so even though the community consensus has clearly favored Optimistic Ethereum, we have not yet reached the point where we can use SIP to test it.
But before I get into the specific tradeoffs, I want to provide an overview of the existing scaling directions for executing smart contracts:
Fast blockchains, aka "Ethereum Killer", other Layer 1 architectures, very fast blockchains
ETH 2.0, see you in 2032 (just kidding)
Status/Payment Channel, i.e. "You can send tokens, what else do you want?"
Side chain, that is, the set of xDAI
Plasma, represented by Omisego, is also called "Late but Arrived"
Use zero-knowledge ZKrollup and other solutions if you love solidity.
Optimistic Ethereum Optimistic Rollup, high-energy early warning
Lightning, laugh without saying a word
If there are other solutions that I missed, I'm sorry, I'm really looking forward to hearing about them on twitter.
Most of the above solutions were ruled out due to the need to keep using the same code base during the migration phase. Of course, many solutions claim to be compatible with the EVM, but this is not as simple as it sounds - although Optimism breaks through this limitation, it still requires a small modification of the contract. But based on this we can quickly rule out these solutions: fast blockchains, ZKrollups, Lightning, state channels, and Plasma. Even with rapid progress on ZKrollup, all current variants would require contracts to be rewritten in a new language. It's not insurmountable, it's just that the tools for these languages are still very immature, which greatly increases the implementation risk.
Some fast blockchain proponents may get emotional reading this, yes some of these projects are EVM compatible, allowing solidity contracts to be deployed on it, but most of them have other issues that we feel will weaken them feasibility. Including those very novel consensus mechanisms have not been tested repeatedly to prove their feasibility, or the security is greatly compromised.
Ubiquitous computing is currently the second most difficult requirement for state/payment channels and Plasma.
Then there are only side chains and Optimistic Rollup left. We ruled out sidechains like xDAI because we need to insure assets worth hundreds of millions of dollars, and several orders of magnitude more in the future. If there is any objection, feel free to break your head with me
After looking at all the options, we feel Optimistic Ethereum presents the best trade-offs and their team is very capable of executing their roadmap.
Phased Optimism
Obviously Optimistic Ethereum is not yet live, so there is still a significant execution risk, which is why I have not yet submitted a SIP to close Synthetix on Layer1. However, considering the trade-offs and risks among the above options, we think that Optimism has proved that it is not only worth a lot of work for Synthetix to migrate, but also to drive other DeFi projects to participate. This is why I decided we needed to leverage our position as one of the oldest DeFi protocols and take the risk of an early implementation. I know this will help build consensus in the community. So far, the general acceptance of the community is high, and no one has objected to the allocation of resources to Optimistic Ethereum.
However, as we get closer to a possible mainnet migration, we have three major concerns.
1) Fraud Proof
2) Centralization
3) Withdrawal Delay
The most critical technical issue is fraud state transition. Some say that fraud proofs are not included at this stage of implementation, and in fact, fraud proofs are, but auto-generated fraud submissions are not yet supported. The Optimism team opted for a phased sequential approach in order to test specific features like deposits and withdrawals before adding complexity. However, this can be stated in the testnet roadmap to avoid user confusion. Although the function of automatically submitting fraud proofs will be launched before the mainnet is launched. incomplete fraud proof function
In the case of , it is obvious that no funds on the mainnet can be deposited into the Optimistic Ethereum network. I would personally vote against any SIP that proposes moving to a lower security network, and I encourage everyone in the community to do the same. This is one of the reasons why xDAI and other POA (Proof of Authority) networks are considered unusable, i.e. low security.
Another major concern is that Optimistic Ethereum appears to be decentralized, but it has centralized parts. I believe this perception is misguided, but it certainly makes sense. Over the past few years, the community has devoted a lot of time and resources to improving the decentralization of the protocol, and it would be a poor trade-off to fall back now for gas fees and throughput. But this is not the case. The sequencer greatly improves the user experience with the least sacrifice.
These concerns stem from misunderstandings about orderers in the Optimistic Ethereum network. Everyone needs to know that the sorter does not have to visit Optimistic Ethereum, it exists to improve user experience. Many people think that a sorter means a single point of failure, which is not ideal, but in fact this kind of failure has long existed in the poor UX of L1. Of course, the block times for falling back to Layer 1 are not ideal, but there are many potential solutions, including backup orderers that can step in if the active orderer fails or is attacked. All these worries are only temporary, as the goal is always to move to a network of orderers as quickly as possible.
The last major concern is the impact of withdrawal delay on cross-layer composability - this concern will be faced by all other expansion solutions except for other Layer1 architectures. These delays mean that during the challenge period, funds will be locked in the Layer 2 network, but there are still several alternatives. One is to build a network of validators that fund both sides of the bridge and take the risk of providing timely withdrawals to earn fees. In fact, Connext has already started work on this, here are more details. This doesn’t completely solve the composability problem, as funds still need to be confirmed on mainnet before another transaction can be initiated, but this is fundamentally critical to all scaling solutions, and we cannot escape it. This is why I think all major DeFi protocols should launch their parallel versions on Optimistic Ethereum as soon as possible. This will make almost all current DeFi transactions happen on Layer 2. It is very likely that staking will be fully migrated to Layer 2, while transaction contracts will continue to run in parallel on Layer 1 and Layer 2.
To be clear, all information about a fully functioning mainnet needs to be made public before the SIP to “Start Optimistic Ethereum Mainnet Migration” is proposed. I am personally confident that this condition will be met in the coming months.
After mainnet migration
How Synthetix migrates to Optimistic Ethereum is still up for debate within the community. First, we must come to a consensus on the need for a parallel and limited version of the network to be released on the Layer 2 mainnet. Once we have consensus here, we need to evaluate the results of this mainnet launch and formalize the rest of the migration.
While Synthetix's DAO will likely be funded in the first few weeks after the mainnet migration, we will need a SIP proposal to split 1% of the final protocol rewards to Optimistic Ethereum. If we distribute a portion of the protocol rewards to this parallel network, we can monitor adoption and allow the market to price migration risk. It is expected that the return will be lower than in Optimistic Ethereum due to the reduced gas fee, but it is also possible that the platform risk will lead to higher returns, because most SNX holders will choose to migrate later.
My opinion is that the minting of sUSD by staking SNX on Optimistic Ethereum is not much different from the current sUSD minting of SNX by staking escrow. We decided to allow minting of sUSD via escrowed SNX to maximize the value of available collateral. For those SNX that have been migrated to Layer2, if you want to migrate them back to a certain point, this is possible, which means that these SNX will be regarded as valid collateral in the network, just with the SNX status on Layer1 different. Therefore, the sUSD minted on Layer 2 and the sUSD on Layer 1 should be interchangeable. Of course there are objections to this, for reasons including that the implementation would be very complex. We have to take the approach that makes the most sense for the community as a whole.
If this migration works, we'll have a fairly powerful lever to influence other parts of the migration: just keep moving a larger percentage of the inflation reward until it's all sent to Layer 2 and all active, staked SNX All happen on Layer2. At that time, of course we need to support the exchange of Synth on Layer1 and Layer2. Therefore, there are a large number of interrelated dependencies to deal with in this process.
I have publicly stated that I believe Optimistic Ethereum will be the way out for DeFi between now and Eth2.0. If the rest of the Synthetix community is so convinced, we need to plan how to join this network and migrate away from Layer 1 entirely. We should of course proceed with caution, but the warriors at Synthetix have never backed down from betting that they are not sure, and I believe this is one of the biggest challenges we have encountered.
To carry out this work, we will need a set of SIPs and SCCPs outlining the changes proposed in each release and the rationale behind them. This will ensure maximum transparency and consent from all token holders.
backup plan
Generally speaking, Ethereum scaling solutions and the competition of smart contract platforms are multi-billion dollar businesses. In this high-stakes game, there are many competing teams, so despite our confidence in Optimism, there may be other teams with better solutions than them. If this happens, we must be prepared to shift the center of our work to this more competitive technology, especially if we find that multiple DeFi projects have achieved social consensus and decided to switch from Optimistic Ethereum. to another program.
We must prepare for the worst-case scenario, that the launch of Optimistic Ethereum fails or is aborted. In this case, we must quickly transfer resources to seriously investigate other solutions, while also optimizing the existing Layer1 system. In fact, we have already started preparing for emergencies. Debt snapshot The fast implementation of the SIP is an example of an optimization that has been on hold for many months, but we sped up the implementation in order to solve the pressing gas problem on Layer1. Thankfully, we got a temporary reprieve, but it can't last long. If for any reason we cannot reach a consensus on migrating to Optimistic Ethereum, I think we as a community must come together to elect another scaling solution and unite to move forward with it. While I don't think it's likely, it would be rash to pretend it's impossible.
This article aims to answer some of the questions the community has had, but also hopefully explain how we got here and where we are in the migration process. Nothing is immutable, even an approved and implemented SIP may be rolled back due to changing circumstances. However, I firmly believe that if we have a consensus on SIP and hope to build the most viable network for DeFi scaling, the entire community needs to push it as hard as possible. I am confident that our community can make it happen, we are stronger than ever and looking forward to 2021, which of course includes Synthetix running on a fully functional Optimistic Ethereum mainnet.
Disclaimer: ECN's translation work aims to deliver high-quality information and learning resources for the Chinese Ethereum community. The copyright of the article belongs to the original author. Reprints must indicate the source of the original text and ethereum.cn. For long-term reprints, please contact ethereumcn@gmail.com Authorize.
