In-depth analysis of EIP 1559, why it is the most important improvement proposal of Ethereum
Editor's Note: This article comes fromBabbitt Information (ID: bitcoin8btc)Editor's Note: This article comes from
Babbitt Information (ID: bitcoin8btc)
Babbitt Information (ID: bitcoin8btc)
, by Hasu and Georgios Konstantopoulos, translated by Free and Easy, published with permission.
Moreover, in the opinion of the two researchers, even readers who don’t care about Ethereum at all should know about EIP-1559.
Because it involves elastic block size caps and congestion control systems, it may become an important part of ensuring Bitcoin's post-block subsidy (referring to sustainable mining).
Better user experience: The first-price auction mechanism used by blockchains like Bitcoin and Ethereum, while easy to implement and understand, has the problem of being inefficient. (If you want to know more,hereandherehere
and
here
Slack mechanism: The demand for block space can be unstable. The result is that some blocks are half full, while others are very crowded. The slack mechanism allows some blocks to be larger, while others can be smaller. This allows us to enforce a longer-term average block size limit, but takes into account variance from block to block.
Preventing Economic Abstraction: When users can pay transaction fees with any token (such as a stablecoin), it threatens the reserve status and monetary premium of the native token (ETH). By forcing transactions to consume a specific amount of ETH, EIP-1559 attempts to make economic abstraction more difficult.
BASEFEE + Tip
first level title
How does EIP-1559 work?
secondary title
As the first building block, the protocol specifies a minimum fee called BASEFEE (Base Fee).
Minimum fees are generally not enforceable because the protocol cannot prevent external price discovery regarding transaction inclusion. The protocol can always dictate a price, but if the miners and users agree on a lower price, the user can pay the miner within the agreement, and the miner can refund the user outside of the agreement.
EIP-1559 solves this by burning the entire BASEFEE (base fee), so it cannot be refunded.
First, they identify a GAS_PREMIUM (from now on, we'll call it a "tip") as the inclusion bid. Typically, it must be high enough to compensate the miner for the risk of mining an uncle block (e.g. 1 Gwei). It allows the old first-price auction mechanism between traders during times of congestion.
Second, the user determines a FEECAP, which represents the maximum amount the user is willing to pay (including tip) for inclusion. This is necessary because the BASEFEE (base fee) can actually move up or down, as we'll see later, and users who set a low FEECAP should be able to wait for later blocks to contain transactions.
In summary, BASEFEE (Base Fee) allows the protocol to enforce minimum fees without incentivizing the formation of off-chain markets. It lays the groundwork for more flexible block space utilization mechanisms such as uniform price auctions and elastic block sizes.
secondary title
Elastic block size upper limit
More flexibility in block size is naturally desirable, but there are known incentive issues with allowing unlimited block sizes. The ability to push up the uncle rate with larger blocks provides an incentive for miners to centralize power. Additionally, verification costs must be included so that the network can remain trustless.
This is where the elastic block size mechanism comes in. The goal is to allow miners to produce larger blocks, but the cost must be provable. Increased orphan block rates are not a justifiable cost per se, as it can be mitigated by increased centralization, but per-protocol fees are.
In the long run, the network adjusts BASFEE up and down to achieve the desired average block size. When blocks fall below the target, fees are reduced over time to encourage demand. Fees increase over time to dampen demand when blocks exceed the limit. The size of the change is determined by the distance from 10 million gas, but is capped at 12.5% in either direction per block.
better user experience
All in all, the BASEFEE adjustment mechanism allows users to automatically adjust the required transaction fee rate according to the utilization rate of the block space.
first level title
better user experience
In order to evaluate the UX of users and wallet providers, we have to consider the different congestion situations the system is in.
Congestion-free state: Users have no reason to increase their tip fees when blocks are below the 20 million gas max cap. (This still applies even if BASEFEE adjusts up and down over time)
As a result, inclusion in a block can be fully determined by BASEFEE.
Users who are willing to pay BASEFEE plus the minimum tip fee will enter the next block. As long as there is no congestion, users can purchase block space at a fixed price. It's like if you go shopping on Amazon today and see a fixed price for the item you want instead of bidding at an auction. The user then either accepts the transaction or abandons it. As a result, fee estimates for users and wallets become highly predictable.
This is an important realization. If blocks exceed 10 million gas for a while, transactions can become very expensive, eventually pushing demand down. How fast are transaction costs rising?
figure 1
Assuming that the BASEFEE of block 0 is 1 billion wei per gas, based on the ETH price of $240, a typical transaction of 210,000 gas will cost $0.0005. After only 10 blocks of 20 million gas, the transaction fee becomes $0.02, and after 100 blocks, the transaction fee becomes $657. This is the power of exponential growth!
image description
figure 1
Even in tip auctions, the size of BASEFEE remains predictable for traders. If the virtual starting value of BASEFEE in block 0 is 100, it can reach up to 100*112.5% at T1, 100*112.5%^2 at T2, and 100*112.5%^3 at T3, depending on And so on. The same applies to the case of reduced fees.
figure 2
The colored area in Figure 2 shows the potential value BASEFEE may receive in future blocks based on its initial value (100 in this case).
image description
The only way for BASEFEE to return to a more "normal" level is to mine blocks below 10 million gas.
image 3
Imagine that there have been 100 blocks of 20 million gas going on. In keeping with the previous example, the average transaction cost has increased to over $657. In order for BASEFEE to go back to the starting point, 89 empty blocks must be mined. In addition, we can mine 183 blocks of 5 million gas, 371 blocks of 7.5 million gas, etc.
image description
Thus, under high congestion, the typical pattern is a series of large blocks followed by a series of very small blocks. This makes sense because users with high time preference can now have their transactions confirmed faster. But in the end, BASEFEE exceeded the needs of the trading parties, and they had to wait for BASEFEE to drop again.
Another way to think about this is that the slack mechanism pulls some blocks worth of capacity from the near future into the present.
But it couldn't generate more capacity and eventually had to pay down the debt.
first level title
After commenting on how EIP-1559 deals with congestion, we can see that its slack mechanism is closely related to BASEFEE, the maximum amount of change between blocks. The slower it adjusts, the more gracefully the system can handle changes in demand.
A common theme of this variation is the periodicity of the day-night or weekday-weekend cycle. In its current form, EIP-1559 does not allow blocks to grow continuously overnight and during the day because these periods are so long that fees would grow into the billions of dollars and drop to near zero at night as quickly.
So the slack mechanism works on the minute-half hour timeframe, but not beyond that. A longer demand will pull the block size in one direction. For example, when the block is full, it will be pulled back more violently by the rising BASEFEE.
first level title
Many blockchains have deflationary monetary policies, meaning that the number of new coins they issue decreases over time (i.e. a common halving strategy). When issuance drops low enough, the incentive to underpin security comes from transaction fees. EIP-1559 is incompatible with this one-fee security model, since most transaction fees are not used to incentivize miners, but instead are burned.
Therefore, it can be said that EIP-1559 allows for a permanent block subsidy for miners to ensure the security of the protocol. But a better way to think about it is: this would make the permanent block subsidy preferable as a design option. The reason is that fee burning has a deflationary effect on the coin supply, allowing new coins to be issued elsewhere without inflating the overall supply.
Due to the fact that a protocol with a permanent block subsidy can lead to a more stable income stream for miners, it is fair to say that EIP-1559 has a positive impact on the long-term security and stability of Ethereum.
first level title
EIP-1559 largely addresses this issue. The BASEFEE portion of every transaction is denominated in ETH and is always burned. In this case, no matter who pays, a portion of the ETH supply will disappear. A miner is still free to choose a lower BASEFEE transaction, but only if he pays the difference from the ETH-denominated block subsidy. Miners earn less ETH in that block, while users keep more. Someone has to pay for this transaction with ether.
potential problem
We say "largely solves this problem" because EIP-1559 prevents the economic abstraction of BASEFEE, but it does not prevent tip fees. Since the tip fee is not burned, there is no way for the protocol to enforce the payment method or location, so there is the possibility of an off-chain market. However, as we showed earlier, in most cases it is not necessary for users to increase their tip fee above the minimum requirement.
first level title
potential problem
A major issue with EIP-1559 is whether miners will be able to manipulate BASEFEE, and if they will.
When BASEFEE is 0, miners receive all bids from users due to no burning. Also remember that the tip fee auction will only start when the demand for block space exceeds the available supply.
For it to be a stable monopoly, a miner activates a soft fork (MASF). In MASF, more than 50% of the computing power will be guaranteed to ignore any block larger than 10 million gas, thereby incentivizing a minority to follow the newly implemented rules.
Summarize
Since this attack vector exists on any network, we currently do not consider it a specific risk to EIP-1559.
first level title
Summarize
We found that EIP-1559 lives up to its promise to a large extent. It should make fee estimates more predictable except for very brief periods of high congestion, where the system falls back to the established first-price auction model. Since BASEFEE will grow exponentially to suppress demand, it won't last too long.
BASEFEE is an interesting building block that makes these benefits possible. Its ability to set minimum fees in the protocol opens up a new design space, from elastic block sizes, permanent block subsidies, better resistance to economic abstraction, to better auction models in the future.
