"Post-Ethereum 1.0" sentient beings: DeFi congestion, breaking through Layer2

Editor's Note: This article comes fromCarbon chain value (ID: cc-value)Editor's Note: This article comes from

Carbon chain value (ID: cc-value)

Carbon chain value (ID: cc-value)

, Author: Bai Ye, reproduced by Odaily with authorization.

Unknowingly, we have entered a "post-ETH1.0 era". From the perspective of user experience, ETH1.0 can no longer meet the needs of users; only DeFi applications have brought it to unprecedented and unbearable congestion. However, ETH2.0, which has been conceived for several years, has been unable to give birth for a long time. In such a situation, where will the applications on Ethereum go? Can Layer2 become the savior of the moment? To this end, we invited Pan Chao, head of MakerDAO China, A Jian, editor-in-chief of Ethereum enthusiasts, Min Dao, founder of dForce, and Uncle Miao, chief researcher of Spark Mining Pool, to discuss these issues.

Q1: How do you view the congestion problem of Ethereum? What impact does it bring to DeFi applications? Will it cause users to migrate from the Ethereum network to other public chains such as TRON and Polkadot?

Pan Chao: Gas fees are actually an economic activity on the Ethereum network. According to the amount of Gas fee, the network restricts and balances the above activities in the form of price. But the gas fee on Ethereum is not governed, or it is actually very blind, because it does not set the price for the content of the above economic behavior, but treats it equally.

This led to the entire liquidity mining, and finally turned into a game that is very inclined to large capital. Because for big capital, it has a lot of funds, and the gas cost required to participate in the game is negligible. In addition, it can get very large returns and subsidies. But for small users, this cost is very high, which is actually very unfair to them. Although we see a lot of liquidity mining advertised as a fair concept, it is actually extremely unfair. It is a very pure "capitalism". The only factor that determines the outcome of the game is the size of the capital.

The next step is to create a more interesting, but not everyone wants to see the scene. The popularity of liquidity mining has led many centralized exchanges to list these tokens—after all, they are also afraid of losing customers. But the actual result is that large capital parties can take advantage of their capital cost advantages to obtain mining benefits and subsidies on DeFi. But for small capital parties and small users, the only economical way to participate in the game is to buy tokens on those centralized exchanges. This has created a very strange but very realistic situation: the liquidity mining of DeFi is completely inclined to large funds.

For many project parties working on DeFi applications, I think this is a situation that everyone does not want to see. Because the initial vision of DeFi is inclusive finance and open finance, which is to lower the threshold for everyone to enjoy financial services. However, due to the emergence of liquidity mining, this vision has developed in the opposite direction. I think that DeFi or liquidity mining must make some changes in the future, and introduce some mechanisms to limit these large capital parties in order to find a balance, otherwise such a game will not continue.

A Jian: On the surface, the high price of Gas will indeed cause users to migrate to other chains to play music and dance, which is a very natural phenomenon. But I want to ask: then what? Are these users really converted to long-term users of other chains? Or is it just for a short period of time, while there is still a wave of speculation, go to other chains to play? Or even more extreme, is the user's use of the chain either-or? Will they only pick one? Or is it that they don't actually have any loyalty at all, and they go wherever it's fun?

I think on the surface, this market is dominated by users. You don't need to hand over your funds to anyone, and you can choose what you want. But in fact, it is still a market where creators have the final say. In other words, everyone actually has funds in their hands, and then they want to find a place where they want to invest. The key is to create an environment where he can put funds. go in. So which chain can continuously burst out this kind of creativity, and can continue to create this kind of user spillover effect, which chain will never have to worry about the lack of users. Let's think about it, it's not once or twice to complain about the congestion of Ethereum, but the number of users of Ethereum has not decreased.

So from my personal observation point of view, I don't think the biggest constraint on the current development of DeFi and Ethereum ecology is the gas cost. I think the bigger problem is that there will always be some participants in this ecology. They will overdraw the reputation of the industry, overdraw everyone's expectations for the industry, destroy the reputation of the industry, and then profit from it. This is the biggest problem we face right now. Maybe we need to form stronger industry self-discipline, or have a better community culture to change this. Otherwise, what we see are always some applications that trigger speculation. It is actually a very bad thing to squeeze out those applications that have been cultivated for a longer time and have a more stable infrastructure.

Mindao: I ask a fundamental question: Why is Ethereum congested? I think this is the core of the problem. There is a similar question: Why is Beijing, Shanghai, Guangzhou and Shenzhen so congested, but many people return to these core cities after fleeing Beijing, Shanghai, Guangzhou and Shenzhen?

So some people joked that there is no traffic jam on Tron or EOS because there are not many applications on it, but there is no need to be jammed, just like there are few cars on the road, so of course there is no traffic jam.

In addition, I think there is another very critical point. Everyone knows that there are dozens or hundreds of building blocks in the entire DeFi of Ethereum, so you can understand why Ethereum has suddenly become congested recently. In fact, building blocks have reached this scale, and there are at least millions of possibilities for the various combinations among them. After the eruption of the composability between building blocks, it is the evolution and evolution of an exponential DeFi building block.

So this is also a very critical issue at the infrastructure level of Ethereum's recent congestion. DeFi has reached this stage, all the building blocks have been completed, and then everyone can play various combinations, just like Lego building blocks, you play various combinations, and then stack them on top of the combination.

In such a congested situation in Ethereum, is it possible for other new chains to seize some subdivided tracks? I think it's entirely possible. Including that we used to pay wages through stablecoins on Ethereum. In the near future, we may consider other options, including the exchange’s TRON’s USDT, because its handling fee is really low.

It can be imagined that if each transaction fee is 10 US dollars or even 70 US dollars when paying wages, the scene itself is very ironic. We know that one of the earliest slogans of the blockchain is to make small and micro payments and low-cost payments. But now it seems that blockchain payment should be one of the most expensive payment methods for small amount transfers. So I think other public chains definitely have the opportunity to do it in the subdivided field. Of course, a more realistic approach is that other public chains may help Ethereum achieve some expansion needs at the level of the second-tier network.

I think it is difficult for other building blocks, or other DeFi, to migrate away from Ethereum in a short period of time. In fact, this problem is very simple. For example, our DForce makes these protocols, but in fact, a lot of them depend on the underlying layer to build them. If I go to a new chain without any infrastructure, how can I build this building block? I think this is very difficult. For most public chains, this is also very difficult. Of course, for public chains such as TRON and Cosmos with relatively rich ecology, if they can achieve the initial cold start, I think there is still a certain opportunity.

In addition, I think that after everyone saw the DeFi boom this year, another group of high-potential competitors began to enter the public chain field. I think this may steal some traffic from Ethereum's DeFi. For example, Binance’s smart chain, FTX’s DEX’s Sonala chain, including Huobi, are also deploying on the DeFi public chain.

Centralized exchanges do DeFi by themselves. I think it is very likely to steal the market of other public chains. Because one of the biggest advantages of the exchange public chain is that the cold start of its assets is very easy. For example, Binance can transfer all 50 or 60 assets to its smart chain overnight, and then you have the basic building blocks. Do a few more core agreements, including AMM and lending. With the basic building blocks, I think it is relatively easy for future entrepreneurs to build. So I think that in the field of DeFi public chain, exchanges are competitors that cannot be ignored. You can take a closer look this year, especially exchanges are very smart now, they don’t engage in independent technologies, and many directly follow the 100% compatible route of Ethereum, so this is a very big variable in the competition in the public chain field this year.

Q2: The problem of Ethereum congestion is so prominent, is there any specific method that can immediately solve Ethereum congestion? Can the Layer 2 solution, which was hotly discussed in the community before, achieve this?

Uncle Meow: First of all, let’s define what the Ethereum congestion problem is.

If you look at it from the perspective of usage, you will find that Ethereum has been a very congested chain since 2018, because when each block is packaged, the usage of the gas limit will be relatively full, all at 80% In the ~100% range, there are no empty fasts or transactions that cannot be processed.

But why didn't we have such obvious feelings in the past two years? I think the main reason is that our urgency for packing is not so strong. Teacher Mindao mentioned earlier that the composability of DeFi building blocks has been magnified and exploded in the current liquidity mining. It has tens of thousands of combinations and even more possibilities. Under such circumstances, when a player or user participates in transactions on Ethereum, his purpose is different from what he expected to achieve last year or even two months ago.

For example, last month I might just transfer a sum of USDT to a friend, or simply call a Compond mortgage contract, or make a loan on AAVE, or mortgage some Dai on Maker. I will not be so urgent about this kind of behavior. I will feel like waiting for two blocks, three blocks, or even saying that I will go to eat after sending the transaction. I just need to come back and sit in front of the computer again, open my wallet and see that the transaction is packaged. It worked, and I think that's acceptable.

However, the recent liquidity mining has brought about an extremely high annualized rate of tens of thousands or even hundreds of thousands, and all income is calculated in seconds, not in blocks. No one wants to miss out on a profit with an annual rate of tens of thousands of years, so everyone will be very eager to participate, and there will be the same lane, I don’t want to queue but want to pay to jump in the queue.

So when we define Ethereum congestion, we also need to see that so many people want to do faster transactions, they want to complete more complex contract calls, and they want to complete their own transactions first. At this time, the gas price that we are seeing now has risen, forming a situation where the transfer cost of an ordinary user is very high. We call this situation the congestion of Ethereum.

Let's talk about whether there is an immediate solution to this congestion phenomenon. In fact, it will be very intuitive if you pat your head, that is, since the lane is very congested, then I will widen the lane, such as expanding the block size. But there is a cost to expanding the lane. This cost may even directly affect the overall security of the Ethereum consensus; moreover, this is not a one-and-done solution.

So it will lead to the next question: whether the layer 2 solutions currently being discussed in the community can solve the very difficult problem in front of us.

The Layer 2 solution is actually not a matter of a day or two. In fact, there are already many kinds of technical practice solutions. Let’s just talk about solutions like roll up, and they have evolved into 4 or more methods. There are some that focus on completing contracts in the second layer network, but it may give up user privacy. A scheme like zk-SNARK is based on the premise of privacy, using zero-knowledge proof to complete off-chain transactions, and then submit a proof to the first-tier network. Such a solution will have a short-term bottleneck, that is, what we can do on the Layer 2 network will be relatively limited.

But at least for now, I think that the various solutions of Layer 2 are feasible from usability to completing simple transfers, at least technically, they can be tried and used. We have also seen that there are actually some decentralized exchanges in China, they are trying to do this, and they are already in actual operation.

So I ask another question: Now that we have seen the congestion phenomenon of Ethereum itself, and at the same time face so many alternatives, where is the resistance that prevents us from advancing this matter?

In fact, this problem has many correlations. One is that most of the current assets are on the first-tier network, and the logic of various user contracts is also on the first-tier network. So when introducing the second-tier network, it seems simpler if we just solve the transfer congestion. But if you want to use the second-tier network to solve the current congestion problem of liquidity mining, it will be more difficult, because it will involve whether various contracts need to be migrated to the second-tier network. For example, if I can mortgage on the second layer network, I want to use Ethereum to lend stablecoins, and I want to use my stablecoins to participate in dForce’s liquidity mining, then how can I balance these two things at the same time.

At the same time, if each application is running on its own layer 2 network, it will also bring about the next problem: each application is using its own layer 2 network to speed up, so how to do the composability between each other? This will also be a bottleneck that hinders further innovation or the possibility of further combination.

Having said so much, my current conclusion is: First of all, there is no immediate, specific method for Ethereum to solve this congestion phenomenon, or there is no immediate method that can be implemented. Secondly, many Layer 2 solutions are technically ready. One of the obstacles to its implementation is the issue of agreement among various application portals and the distribution of benefits; at the same time, it is the issue of how to choose and judge the pros and cons of each program.

A Jian: Teacher Min Dao has given a very good answer to this question. Why are Beijingers unwilling to go to Xiongan? Is there so much fun in Xiongan? Are there any interesting people in Xiongan? Does Xiong'an have all the infrastructure, convenience and fun that I can enjoy in Beijing? Obviously not. That's why it's already so congested, every subway line in Beijing is full of people, but everyone is still willing to stay here. The reason is very simple, because what you need can only be found here.

Although the current Layer 2 solution has been developed for many years, it is actually not an easy task to achieve a good user experience and attract enough operators to participate. Just like Uncle Miao mentioned a few solutions, in fact, I personally recommend everyone to try it. Exchanges such as Loopring have launched a layer-2 network, and decentralized exchanges such as Uniswap and dYdX are planning to migrate to the layer-2 network. It is already a very promising solution in itself, and I think everyone can take this step bravely, just like you try different DeFi applications, try the Layer 2 solution.

In addition, we might as well think about it: Is the congestion of Ethereum a problem? If it is a problem, can it be solved? If it doesn't look like a solution in the short term, is there a solution in the long term?

In fact, as long as everyone has a strong demand, as long as you want to continue to expand the user base of Ethereum, such as an increase of hundreds of times, but you expect the Gas fee to remain low, this is unrealistic. As a global consensus, Ethereum itself has a cost, and it has to pay a high price. This price does not mean that the person who provides you with a global consensus is immoral, or that he is unwilling to do so, but is determined by objective factors. As Uncle Miao said, the structure of the blockchain itself cannot be infinitely expanded, otherwise your network will fall into centralization.

On the other hand, in fact, the person who pushes up the gas fee is the person who wants to use this chain like you, not the person who provides you with services. We have to correct it: it is not that the miners have problems, but that other people want to use this chain just like you, and their desire to use this chain is more urgent, and their economic value is higher, so he will cause you to use this chain. The cost of the chain is very high.

In general, the congestion problem is not entirely a technical problem, but should be regarded as a user experience problem. How can we improve the user experience as much as possible without discussing the ultimate technical problem of layer2 (its application itself will take a while)? First, users should know how quickly their transactions can be sent to the chain. Second, under the current circumstances, there should be a means to help users save Gas fees as much as possible.

In both directions, the community has proposed some tools dedicated to transactions. The gasnow recently launched by Spark Mining Pool is actually an indicator of Gas price. Its novelty is that it is directly connected to the transaction memory pool of the block producer. You can directly see how high the gas price is for the first 50, 100 or even 200 transactions in the transaction memory pool. In this way, you can be more targeted, because you can directly know how high the gas price is, and where your transaction will be inserted into the queue. This is more accurate than the original gas price indicators, and it actually improves the user experience.

In addition, there are now two methods in the community to help users save gas fees. One method is to use the contract of aggregation operation to pool the funds of many people to perform one operation, and when the operation is completed, the funds of many users are included in one operation. A typical example is the liquidity mining pool that has appeared recently, which combines the operations of many people to save that part of the fixed gas fee.

The other part is that you may have heard of the major transaction layer of the relayer of the original transaction, which is to let the relayer forward the user's transaction to improve the user experience. Because relayers are often able to grasp the appropriate gas fee more accurately, help users save this gas fee, and will not raise the inflated price.

Q3: When it comes to the congestion of Ethereum, in addition to the second-tier network, there is actually another problem that cannot be avoided, and that is Ethereum 2.0. At present, Ethereum has incubated the most prosperous DeFi ecosystem in the circle. However, its performance seems to be unable to carry such a large ecology. Everyone hopes to solve the problem on ETH2.0, but ETH2.0 has been delayed repeatedly. Currently, where are the main pain points of ETH2.0? How long will it take to actually realize ETH2.0?

Uncle Meow: Regarding Ethereum 2.0, everyone has been looking forward to and talking about it for a long time. According to the original milestone of Ethereum, the Mayoston 2.0 should have already been launched. Of course, there are many obstacles in the middle. In the end, everyone found that it is still quite difficult to realize a blockchain network with sharding, cross-sharding, and isomorphism.

Ethereum 2.0 is currently in the testnet stage. Let me first briefly talk about what kind of progress the official network will be, and what kind of way to achieve it, and what kind of goal will be achieved in the end.

The ultimate goal of Ethereum 2.0 cannot be achieved immediately after one-click upgrade or one-click deployment, it needs to be completed gradually. Finally, the current single-chain structure that relies on POW consensus is changed to a multi-chain structure that relies on POS to maintain company security.

Now the Ethereum 1.0 chain has carried a lot of funds, has a lot of contract application logic, and has a lot of users, so the entire upgrade and migration process of 2.0 is very cautious.

So 2.0 itself will be divided into different stages, we call it stage 0, stage 1, and stage 2. The main goal of phase 0 is to enable the beacon chain of POS. At this stage, users can only transfer Ethereum to the 2.0 chain one-way. At the same time, you cannot withdraw the Ethereum on 2.0 at this stage, and there is no sharding. The main goal of this stage is to start the beacon chain.

When the beacon chain is started, the main goal of phase 1 is to complete the communication between the shards and the beacon chain, and there will be many other technical problems in the middle, such as the need to convert Ethereum into a stateless shard slice, transfer various other information to the beacon chain, and may even open the function of transfor at this stage. There may be 1.5 in the middle, let's skip 1.5 and directly say that we have reached the second stage.

What Ethereum needs to do in the second stage is our ultimate goal: now there are 64 shards in the plan, and the shards can call each other, contracts can also call each other, transfers can also be completed, and users can also have different execution environment. It looks extremely complicated, as if it sounds like you don’t know what it is doing, so we might as well make a metaphor: it needs to build various viaducts, build various multi-lanes, and build subways; it needs to add more in a limited single point of time. Transactions that can be sent in parallel, contracts that can be called in parallel, so as to achieve an expansion effect.

As for how long it will take for Ethereum 2.0, no one can give a clear time, but it is more optimistic that the current 2.0 testnet has been running for a while and is relatively stable. If the testnet runs for as long as 3 months without any other abnormalities, the 0th stage of Ethereum 2.0 will be getting closer and closer to us.

https://ethfans.org/posts/desig-choices-in-eth2

Personally, I am relatively optimistic about whether Ethereum 2.0 can solve the current congestion situation. In figurative terms, the current Ethereum is not a highway, but a country road. But the goal of 2.0 is an 8-lane intercity highway, and it is not enough. It has 64 8-lane roads running at the same time. And we are constantly seeing this vision being realized step by step.

A Jian: The main challenge of Ethereum 2.0 is the specific mode of phrase two. Specifically, it is the problem of cross-shard communication. Cross-shard communication means that, because each shard itself has a different state, it must ensure that the state on each shard is verifiable to the verifier on another shard. In addition, it is necessary to ensure that the various shards are communicable. In other words, the operation initiated in one shard can change the state in another shard, and the entire change process can also be verified. Yes, and it's pipelined.

Ultimately, you can't let this scheme break scalability. In fact, using the beacon chain as an intermediary means that you put a lot of cross-shard information on the beacon chain, and even let the verifiers of the beacon chain actually execute these messages. But doing so is effectively indistinguishable from large blocks, because it completely destroys scalability. So I attach a website, you can simply take a look at where the difficulty of this problem is.

How ETH2.0 alleviates the congestion on the chain is actually highly dependent on the specific technical solutions for cross-shard communication. When this plan has not been determined, it is actually difficult to predict its actual shape. But I personally think that cross-shard communication is indeed more complicated than your transaction processing in the same shard, so it is obviously impossible to have the same user experience as a user whose target contract is in the same shard. Either you have to pay more, or you have to wait longer. This is also the psychological expectation that everyone needs to have.

Q4: The development of DeFi is changing with each passing day. It took only 3 years to realize from a concept to a locked position of more than 12 billion US dollars. Even the centralized exchanges have felt the shock. So, will the development of DeFi be constrained by the performance of Ethereum? How will the DeFi industry develop before ETH2.0 is actually launched?

Mindao: Why didn’t everyone feel that Ethereum was so congested before? I think it’s because the previous contract interaction method was very simple, just a very simple contract request, but now you can see that the complexity of our contract request has far exceeded to the previous extent. For example, if you deposit a sum of money on YFII, it will request the price of the following three or four underlying agreements at the same time, and the Gas fee will also be superimposed accordingly.

This bottleneck will affect the development of DeFi, which means that the composability of the DeFi protocol itself will be limited. For example, we can combine 3-4 ingredients to form a cocktail. Now there are not so many ingredients in this cocktail, you can’t make so many, it’s too expensive, or you can’t afford it at all. Then we can only use 2-3 kinds to combine. The playability here, or the innovation of the protocol may be hindered a lot.

The first question we have to consider when designing the mechanism is to evaluate the Gas fee. If the gas cost is so high that most people will not use it, then this direction or idea may be abandoned. Conversely, the Gas market is also guiding the design thinking of developers. I don't think it's a particularly good thing in a way. Because the reason why there are such interesting things in DeFi is because it can be freely created and combined freely, but you have to consider the Gas fee every time you create, I think it hinders the development of innovation in DeFi.

Now everyone is waiting for Ethereum 2.0. However, even according to everyone's optimistic estimate, it will take two or three years before it will be available. Sharding is a good research direction, but cross-shard communication in a sharded state is actually very difficult. In this industry, we have seen many projects working hard in this direction, but none of them have very good solutions, and I am not particularly optimistic.

Then back to the most realistic question: what should I do if Ethereum is very congested? What if Beijing is very congested? I think there is still a chance for the Layer 2 network to solve this problem. That is, instead of going to Xiongan, we dig down an underground city.

If it is possible to have a smart contract execution environment at the Layer 2 level, I think many applications are willing to migrate there. As long as a small cluster can be formed, it is possible to form a new ecology. At the same time, everyone can also enjoy the advantages of the infrastructure in the city, rather than completely moving out of the city. So I think the two-tier network should be a good way to solve this expansion problem to a limited extent.

We have seen that the entire infrastructure of Ethereum 2.0 solves the problem of expansion through partitioning, which can be expanded infinitely. The multi-chain expansion with Polkadot seems to be two different structures, but they are actually the same. It's just that Polkadot and Cosmos are solved through chains, and Ethereum is solved through slices. The second-layer network is expanded through the three-dimensional space level. Similarly, I am still in the same time and space, but the dimension of my space is extended a little bit more. I personally think that the current situation will indeed force various DeFi projects to consider this issue.

There is a particularly interesting phenomenon this year. In the past, no DeFi would say: I want to expand and try Layer 2. But this year, projects such as Compond and dydx have made attempts in this regard. This is indeed a pressing problem, and I think it can be partially solved.

Pan Chao: My point of view is that the performance of Ethereum will not be the biggest problem restricting the development of DeFi.