EIP-1559 Community Extension Report

source|Ethereum Cat Herders

source|

Author | Tim Beiko & Pooja Ranjan

This report was co-authored by Tim Beiko and Pooja Ranjan, with thanks to William Schwab, Alita Moore and Edson Ayllon for their efforts on this report.

• key result

• Projects' overall impression of EIP-1559 was evenly divided between positive and negative (a few projects were neutral). If we exclude miners, 60% of projects are generally supportive, with twice as many projects as neutral than negative.

• For projects, the main benefit of EIP-1559 is the predictability of gas fees (especially for those projects that set gas fees for users), and another benefit is that some ETH will be destroyed in each transaction.

• Some of the main problems of EIP-1559 include: it has a negative impact on miners, it is difficult to conduct a comprehensive analysis of the problem, and there are some potential implementation and tooling problems.

• Proper tools, good communication, good documentation, and a public testnet will allow various projects to increase support for EIP-1559, but there are many projects that want to upgrade the network before investing a lot of time to implement this EIP .

• During our calls with various projects, several issues came up repeatedly. These questions will be answered in the FAQ section at the end of this report.

background

backgroundEIP-1559In the past few months,[1] (a proposal to reform the Ethereum gas market) has made a lot of progress. While the proposal could bring significant user experience and economic benefits to Ethereum, it changes several fundamental parts of the protocol, including block headers, transaction formats, and transaction pools. existlast implementers meeting

, highlighted the need to better understand the impact of these changes on projects and to gauge community perception of EIP-1559. This led us to decide to take a more organized approach to community outreach for this particular EIP.Shortly after, Ethereum Cat Herders organized aCommunity Outreach Questionnaire

• , and held several 1:1 meetings with projects built on Ethereum. At the meeting, we answered some questions about EIP-1559D from various projects, collected their feedback on the proposal, and discussed what efforts should be made to get them to provide support for the proposal. As of Wednesday, October 8, 2020, this report collects and shares your feedback from the meeting and lists the answers.recent explanatory articlesandand。

• FAQ written by Vitalik

Survey Respondents Overview

A total of 25 projects shared their views on EIP-1559. Fifteen of them participated through questionnaires and another 10 shared with us via video call. Most projects do not wish to be directly mentioned at the end of the report, and one project wishes to participate completely anonymously so its data does not appear in the report.

The open participating projects are: Gitcoin, Argent, Infura, Kyber Network, ethers.js, POAP, TrueBlocks, Bitfly (etherchain.org/beaconcha.in), Nanopool, and 0xBitcoin.

We require projects to be grouped into the following categories: on-chain applications, wallets, exchanges, miners, and other tools or infrastructure.

The following is a breakdown of the projects participating in the survey:

Statistical graphs obtained by dividing according to each project

It is worth emphasizing that among the respondents, the proportion of miners is high, while the proportion of wallets and exchanges is relatively low. The only exchange participating is a decentralized exchange, Kyber Network. Where appropriate, the report will separate feedback provided by miners from other projects for clarity. Subsequent community outreach will focus on exchange and wallet projects.

Common perceptions and perceptions of EIP-1559

All projects we contacted expressed knowledge of EIP-1559, and nearly half of them have been following its latest developments. The most common way for everyone to follow up on the latest developments of EIP-1559 is to check Twitter. Other popular channels are: Github, Discord, Ethereum Magicians, and the EIP-1559 Implementers Conference. Almost every project obtains progress updates on EIP-1559 from different sources.

When asked about their overall impression of EIP-1559, 10 projects (42%) had a good overall impression of EIP-1559, 4 projects (17%) were neutral or unable to express their views due to lack of understanding, and 10 projects (42%) have an overall negative attitude towards EIP-1559. Miners were the group with the most negative attitudes, with 8 out of 9 (89%) responding negatively. If we exclude miners, 9 out of 15 (60%) respondents are positive about EIP-1559.

Benefits of EIP-1559

• When asked what benefits are expected from EIP-1559, respondents most frequently cited the following two:

• Gas fee predictability (7 mentions): In general, these projects are expecting that EIP-1559 will allow them to reliably set the correct gas fee. Those projects that specifically set gas fees for users, especially support EIP-1559.

Burning ETH (mentioned 5 times): Burning the base fee is one of the benefits of EIP-1559, which is often mentioned even when the respondents themselves do not value this benefit (eg, "People like EIP -1559 because it destroyed ETH").

In addition to these two, other benefits they mentioned are: lower gas fees, better incentives for everyone to join forces in the network, no need to pay extra transaction fees to miners, and faster packaging transactions for users. Given that EIP-1559 is not intended to significantly reduce Ethereum fees, it is necessary to make this clear in future communications about EIP-1559.

Risks posed by EIP-1559

• When asked what risks they perceived from EIP-1559, the three most frequently mentioned risks were:

• Negative impact on miners (10 mentions): Almost all interviewed miners mentioned that EIP-1559 would have a negative impact on them and would consider mining on another chain if the proposal passed. Some miners have suggested that 1559 is best as part of Eth 2.0 only. Other problems with mining include: possible collusion among miners; reluctance of miners to upgrade their clients to support EIP-1559; and risk of centralization of mining if small miners are kicked out of Ethereum.

• Difficulty analyzing EIP-1559 (7 mentions): The second major issue (or the first if you exclude the miner's answer) is that it is difficult for everyone to analyze EIP-1559. This is due to several backgrounds: first, it lacks a formal specification or proof of the mechanism, making it difficult for people to independently evaluate and criticize the proposal; second, EIP-1559 involves many aspects of Ethereum and it is difficult to predict secondary effects; Finally, it is impossible to properly test the EIP-1559 incentive design until EIP-1559 is actually applied to the mainnet and ETH is actually staked.

Implementation Issues (6 mentions): Client and tool level implementation issues were the third most cited risk. Specifically, respondents are concerned that if there is no standard API for EIP-1559, it may be more difficult than it is currently to handle multiple clients. There may also be issues with tool crashes when EIP-1559 is put into use, and whether tool developers can easily implement conformance tests against EIP-1559.

other worries

These issues are not necessarily mentioned often, but are noteworthy and should be included in this report for the attention of implementers and researchers. Note that we have not attempted to refute any of these questions or justify them. All reasonable questions asked throughout the interview are listed here.

• Effectiveness Concerns

• Since the base fee will increase by 5% per block, it will increase by 100,000 times after 30 minutes. And between the adjustment period of EIP-1559 and the usage peak, there will be a mismatch on the network (usually lasting a few hours to a few days). How can EIP-1559 help with the next UNI token launch, solve the market maker liquidation crisis, etc.?

• EIP-1559 tries to do too many things at once: improve user experience and create deflationary pressure. Why don't we split these two goals into two EIPs for easier analysis?

• Does EIP-1559 reduce the net free space for blocks? Nodes will need hardware capable of processing the theoretically largest blocks, but on average, only half as large.

What if people keep using legacy transactions instead of EIP-1559 because it's easier for them?

• mining concerns

• Is it possible that miners will collude with each other to determine the base fee? Or incentivize users to pay higher gas fees? How can we prove it?

• If miners/stakers collude with each other to change the base fee to 0, they have nothing to lose. This will not have negative impacts on users and the ecosystem like a 51% attack.

• Reducing incentives for miners could lead to more ASICs (Application Specific Integrated Circuits) on the network.

BribeProxy contracts may become common, which transfer more off-protocol fees to Binance.

• tool concerns

• Will the tx.gasPrice opcode be deprecated? Do smart contracts using this opcode that want to take advantage of EIP-1559 need to be rewritten? It is often noted that while rewriting a contract is not a huge amount of work, the project needs to re-audit the contract, which can cause significant costs and delays.

For some tools, it takes a lot of work to implement. In some cases, if they removed legacy transactions in the latest version, it would be difficult for users to go back to the old version when they were not very receptive to the new way of doing transactions and wanted to continue using the old version.

• implementation concerns

• Keeping these two transaction types (legacy transactions and EIP-1559 transactions) brings more complexity to the Ethereum network.

• Will EIP-1559 increase the risk of DoS attacks on Ethereum?

• How will this affect transaction ordering? What are second order effects?

• What impact does EIP-1559 have on various parts of the client stack such as devp2p, mempool, database, etc.?

• Does the problem that EIP-1559 tries to solve at the client layer no longer require changes at the consensus layer?

The implementation of EIP-1559 would be too complex to be put into use at all.

• economic concerns

• No economic research has been done on the impact of destroying the base fee.

• Why increase the base fee by 5% instead of other values?

Before EIP-1559 is launched on the main network, there is a lack of incentives to solve the problems of EIP-1559.

• Anxiety about community ideology

• What if miners don't want to upgrade their clients to support EIP-1559?

How can we be sure that EIP-1559 will have enough support to make such a drastic change?

It is valuable for implementers and researchers to grapple with these concerns.

Alternatives to EIP-1559EIP-2593When asked about alternatives to EIP-1559, some respondents indicated that the only alternatives were

, or “Escalator fees” proposed by Dan Finlay of MetaMask.

Achieving wishes and obstacles

Respondents were asked what changes they would need to make to their projects in order to support EIP-1559, how much time they expected to spend making those changes, and whether EIP advocates could offer any help moving the process forward.

• The changes required to support EIP-1559 by the interviewed projects are listed below:

• Rewrite the calculation of block rewards;

• Modify the logic of processing transactions, including updating serialization and parsing libraries;

• Upgrade mining pool software

• Rewrite contracts to use new opcodes, remove deprecated opcodes;

• Ensuring dependencies on third parties still work (e.g., Chainlink’s oracles)

• Think about transactions that apply to 1559, and modify the "gas abstraction" function for users;

• Upgrade client software;

• Change the API used to predict gas prices;

• Modified UI to work with 1559 transactions; adjusted for client side RPC (Remote Procedure Call) interface changes

• Use the 1559-capable testnet and other available resources to do extensive testing and quality assessment

Modifications to the API used to share block and transaction data with users

Most of the changes mentioned above are project-specific. In order to protect project privacy, the list synthesizes similar modifications mentioned by interviewed projects, and aims to give a qualitative overview of 1559's impact on existing projects.

• When asked what would help them prioritize support for EIP-1559, the most common answers were as follows:

• Testnet (5 mentions): Having a public testnet suitable for application use (e.g. JSON RPC with 1559 support) is the first step for most projects to start their implementation.

• Network upgrade integration (3 mentions): Several projects mentioned that they wanted to start their implementation work after EIP-1559 was widely accepted through the network upgrade. These projects don't feel it's worth implementing something that isn't guaranteed on mainnet.

Incentives (3 mentions): Several projects mentioned that if there were incentives to use EIP-1559 (e.g. lower gas prices), it would push them to implement it earlier.

These projects also mention other assistance, including updated tools (e.g. ether.js or web3.js supporting EIP-1559), related RPC endpoints in the client, clear opcode definitions and deprecation timelines, a user A public API to predict the base fee of the next block, complete documentation, good communication, support channels, and a reasonable EIP economic model.

• Likewise, respondents were asked what would cause them to delay implementing EIP-1559. The most common answer was "no" (mentioned 7 times), showing a strong willingness to support the EIP. Here are some of the potential concerns mentioned, in order of frequency:

• EIP/specification issue (mentioned 3 times): If the client's early work finds a major problem in the specification, the project side will use this as a red flag and delay the implementation of supporting EIP-1559 until the problem is resolved.

• Uncertainty in mainnet deployment (mentioned 2 times): Similar to some projects that hope that EIP-1559 will be integrated into the network upgrade and will give priority to their implementation, other projects also pointed out that the uncertainty surrounding mainnet integration will delay their implementation work.

• Prioritization (2 mentions): Some projects mentioned that they were working on some other feature implementations that might bring more value to their users than supporting EIP-1559, so work on those implementations might be prioritized .

• Lack of documentation and support (mentioned 2 times): If it is not clear to the project side how to provide support for EIP-1559, how to test it correctly, and where to report problems, some Projects may delay providing EIP-1559 support until these conditions are met.

Lack of community support (2 mentions): Some projects mentioned that if they knew there was opposition to EIP-1559 in the community, they would be reluctant to support EIP-1559, because on the one hand, the effort may be wasted, and on the other hand On the one hand, they don't want to get involved in the fights within the community.

In addition to these, other concerns that were mentioned included poor testnet support, lack of proper tooling, the possibility of using legacy transactions where possible, and the impossibility of testing EIP-1559 before mainnet deployment.

Finally, it's worth noting that some of the interviewees who strongly opposed EIP-1559 mentioned that they would never implement it under any circumstances. 6/7 (~86%) of such respondents are miners.

Implementation timeline

We asked these projects: when will they start supporting EIP-1559, how much time are they prepared to dedicate to doing it, and how long after EIP-1559 goes live on mainnet they expect to see legacy transactions supported.

When is your project going to start working on EIP-1559 support?

Most of the respondents hope to start implementing these changes after the EIP is launched on the current testnet. Many projects mentioned that although EIP will not have a major impact on project development, they still need to re-audit some codes of the project, which will increase the consumption of time and cost.

This point is also supported by an answer to another question: "How much time (for a full-time person) are you prepared to dedicate to supporting EIP-1559?"

We can see here that most projects want to spend relatively little time on providing EIP-1559 support.Finally, after EIP-1559 is launched on the mainnet, the project party strongly hopes to see the longest possible legacy transactions supported. More than half of the respondents mentioned that after EIP-1559 goes live on the mainnet, they would like to see legacy transactions over 12 months supported. This demonstrates that EIP-1559'sa recent change

- Long-term support for legacy transactions is likely to be widely welcomed by the community.

After EIP-1559 is deployed to the mainnet, how long do you expect to see transactions supported before 1559?

in conclusion

This community expansion survey shows that although a large number of project parties are excited about EIP-1559, 60% of the respondents hope to see EIP-1559 deployed to the main network next year, but some project parties are worried about this. This EIP still has many loopholes and will be severely regressed by miners who oppose this EIP.

So many concerns and questions emphasize that EIP-1559 needs more people to explain and refute some concerns. Finally, respondents made it clear that frequent communication across platforms, updated tools, and support channels would greatly help them provide EIP-1559 support.We hope this report is constructive for the community! usOur questionnaire will be made public

, other projects can continue to share their ideas, and we are also trying to reach out to more wallet and exchange projects to increase the sample size of these two categories in this extended study. We will publish an updated version of this report when a sufficient number of additional projects are interviewed.

common problem

When interviewing with these project parties, some questions are often raised, and the following questions and some answers are provided for the community.

Common Comments & Concerns

• What is the purpose behind this proposal?

The EIP document and Vitalik's recent FAQ list some of the purposes

• Anyone object to this proposal?EIP-2593。

Yes, as this survey shows. Although there is no detailed list of opponents, the group of miners seems to be the most opposed. It is worth noting that there are also non-miners who object to the current version of the proposal. Their concerns included lack of a formal economic analysis, or a preference for alternatives such as

• Is there another, simpler proposal for dealing with transaction fees?EIP-2593Another proposal is

• , which helps improve the UX of transaction sending, but does not introduce a mechanism to destroy the base fee.

change in plan

• What noticeable differences can end users expect?

This EIP should make it easier for application developers to predict the gas price of their transactions, allowing users to see their transactions included in blocks faster most of the time, since blocks almost always have extra space.

• What changes will it bring to smart contract developers?

• This EIP will modify the block header - add the base fee field, change the transaction format - delete the "gas price" field and introduce the "fee cap" and "tip" fields.

Launch plan

• How will it be rolled out?

This EIP will first be rolled out on the short-term EIP-1559 test network. If it proves to be viable, it will go through the network upgrade process, and if accepted, it will be deployed on the existing testnet and mainnet.

• Is it backward compatible?

No, this EIP requires a network upgrade to activate, and nodes need to be upgraded to support it.

• How will the network handle legacy transactions, will there be a transition period?

• The original version of EIP had a transition period during which legacy transactions were slowly deprecated. The current version of EIP changes the "gas price" to "fee cap", and sets the difference between "fee cap" and "base fee" as the "tip" for miners, thus packing the legacy transactions into blocks. This means that these legacy transactions will likely need to overpay their tip, but will always be supported.

Gas Prices, Base Fee, Fee Caps and Tips

• Will EIP-1559 increase/decrease gas prices?

EIP-1559 does not make gas prices ever higher or lower. Gas price is a function of demand for Ethereum block space. This EIP will make gas prices more predictable (and help users avoid overpaying for gas), but it won't drive the "equilibrium price" of gas down. In other words, EIP-1559 will not reduce the gas price from 300Gwei to 30Gwei.

• What are base fee, fee cap and tip?

• The base fee is the minimum gas price that must be paid for a transaction to be included in a block. It depends on how full the block is. If more than 50% of the block space is occupied, the base fee will increase, and if it is less than 50%, the base fee will decrease. The base fee is the burned portion of the transaction fee.

• The tip is the part of the transaction fee that is sent to the miners.

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.