[Research Report] will the eip-1559 proposal make eth appreciate or face risks?

Time:2020-10-23

[Research Report] will the eip-1559 proposal make eth appreciate or face risks?

[Research Report] will the eip-1559 proposal make eth appreciate or face risks?

Eip-1559 is probably one of the most anticipated upgrades to the Ethereum economic model, and it is also acknowledged that it is hard to understand because many articles focus on technology. In fact, it may be a shortcut to understand this matter from the perspective of gas.

[Research Report] will the eip-1559 proposal make eth appreciate or face risks?

Important background supplement. Eip-1559 is a proposal of eth that will bring significant changes to users, miners and wallet developers, and even affect the security of the whole Ethereum. Eip-1559 contains four design objectives:

  • Better user experience
  • Relaxation mechanism: make the unoccupied blocks smaller and allow some blocks to increase;
  • Higher security: separate transaction costs from network security.
  • Preventing economic inflation: it is mandatory to consume a certain amount of eth to pay the transaction fee and burn the transaction fee.

Main findings

  • The overall impression of the eip-1559 is negative. If miners are excluded, 60% of projects are usually positive, and neutral projects are twice as many as negative ones.
  • The main benefit of projects using eip-1559 is predictable gas prices, especially for projects that set prices for users, because eth must be consumed in every transaction.
  • The main problems of the project adopting eip-1559 are the negative impact on miners, the difficulty of thorough analysis of the proposal, and potential implementation or tool issues.
  • Appropriate tools, clear communication, good documentation and public test network can help the project to increase support for eip-1559, but many projects still hope to upgrade the EIP network before investing a lot of time to implement EIP.
  • Wallets and exchanges have received little response on behalf of participants. The subsequent community development should focus on these two types of projects.
  • During our conversation with the project, we repeatedly raised several questions and answered them in the FAQ at the end of the report.

background

In the past few months, eip-1559 has made a lot of progress, which is a proposal to reform the Ethereum charging market. Although this EIP can bring huge user experience and economic benefits to Ethereum, it changes several basic parts of the protocol, including block header, transaction format and transaction pool. The need for a better understanding of the impact of these changes on the project and the need to assess the community sentiment associated with eip-1559 was highlighted in the last implementer’s appeal. Therefore, it is decided to adopt a more structured method to carry out community development activities.

Not long after, Ethereum cat herders organized a community outreach questionnaire and held several 1-on-1 meetings with projects on Ethereum. The meeting enabled us to answer the project’s questions about eip-1559, collect their feedback on the proposal, and discuss what can be done to make it easier for the project to increase support for it. This report shares feedback from both meetings and tabular responses collected by Wednesday, 8 October 2020.

Overview of participants

A total of 25 projects shared their ideas on eip-1559. Fifteen of them shared information with them through the questionnaire, and another 10 had video calls with us. Most projects don’t want to be directly referenced in the final report, and a project wants to remain completely anonymous, so it won’t be reflected in this data report.

The projects involved in the study include gitcoin, argent, infraa, kyber network ethers.js 、POAP、TrueBlocks、Bitfly ( etherchain.org/beaconcha . in), nanopool and 0xbitcoin.

We count by project type: on chain apps, wallets, exchanges, miners and other tools or infrastructure frameworks. The breakdown is as follows:

[Research Report] will the eip-1559 proposal make eth appreciate or face risks?

It is worth emphasizing that the response rate provided by miners is high, while the participation rate of wallets and exchanges is low. The only exchange involved is DEX kyber network. Where appropriate, this report separates the miner’s feedback from other items for clarity.

General cognition and thinking of eip-1559

All the projects we contact are aware of eip-1559, and almost half of them are tracking the latest development of EIP. The latest way people get to know EIP is through Twitter. Other common methods of tracking are GitHub, discord, Ethereum magicians, and phone calls with eip-1559 implementers. Almost every project we talk to uses multiple resources to get the latest status of eip-1559.

[Research Report] will the eip-1559 proposal make eth appreciate or face risks?

When asked what their overall impression of eip-1559 was, 10 projects (42%) were generally positive about eip-1559, 4 (17%) were neutral or had insufficient opinions, and 10 projects (42%) had negative overall impression on EIP. Miners were the most negative, and 8 / 9 (89%) responded negatively. If the miners were excluded, 9 / 15 (60%) of the respondents had a positive impression of eip-1559.

Benefits of eip-1559

When asked what benefits they expect from eip-1559, respondents most often list the following two areas:

  1. Gas price predictability (mentioned 7 times)The project is generally satisfied that eip-1559 can set the correct gas price reliably. Projects that set gas prices for users are particularly positive.
  2. Eth combustion (mentioned 5 times)Even though respondents seem to ignore this benefit, they often mention that burning is an advantage of EIP (e.g., “people like eip-1559 because it burns eth”).

In addition to these two advantages, other benefits were mentioned, including lower costs, better incentives across the network, no extra transaction costs to miners, and faster packaging of transactions for users. Since eip-1559 is not intended to significantly reduce the cost of Ethereum, it is necessary to make this clear in future EIP announcements.

Risk of eip-1559

When asked about the risks of using eip-1559, the three most commonly mentioned risks are:

  1. Negative impact on Miners (10 items mentioned)Almost all the miners who answered the survey mentioned that eip-1559 would have a negative impact on them, and if they passed, they would consider mining through other chains. Some miners suggested that 1559 would be better as part of eth2.0. Other concerns include potential collusion among miners, miners’ failure to upgrade clients supporting EIP, and the risk that small miners will be expelled from Ethereum, resulting in the risk of centralized mining.
  2. Analysis of the difficulties of EIP (7 mentioned)The second major problem is the difficulty of analyzing eip-1559. This is proposed in the following situations: there is a lack of specific formal specifications for people to evaluate; eip-1559 involves many parts of Ethereum and the second-order effects are difficult to predict; and to test EIP with real eth running, it may not be able to correctly test the EIP incentive design.
  3. Implementation issues (6 items mentioned)Implementation issues at the customer and tool level are the third most mentioned risks. Specifically, respondents were concerned that without a standard API for eip-1559, it might make it more difficult to handle multiple clients than at present. There are also concerns about the tool disruption when eip-1559 comes online, and how tool developers can easily test its consistency with EIP.

Other noteworthy issues

Effectiveness issues

  • The adjustment period of eip-1559 does not match the peak usage we see on the network, which usually lasts from hours to days, as the cost per block increases by 12.5%. How can eip-1559 help with the issuance of the next uni token, trader clearing crisis, etc?
  • Eip-1559 tries to do too many things at once: improving the user experience and creating deflation. Why don’t we separate these two objectives into separate EIPs for analysis?
  • Will eip-1559 reduce block space? Nodes need hardware that can handle theoretically maximum blocks, but on average, they can only handle half the size.
  • Why would people use 1559 instead of continuing to use the old way of trading because it would be simpler?

Mining issues

  • Is there a way for miners to combine to maintain a high or low base charge, or to motivate users to offer more tips? How can we prove that?
  • The miners’ Union reduced the basic cost to zero and there was no loss. This will not have a negative impact on users and the ecosystem as a 51% attack.
  • Reducing incentives for miners can lead to more ASIC on the network.
  • The bribeproxy contract is likely to become common, with more out of agreement fees being passed on to coinbase.

Tool issues

  • tx.gasPrice Will the opcode be discarded? Does the smart contract that uses this opcode and wants to use eip-1559 need to be rewritten? It is often pointed out that although the workload of rewriting contracts may not be much, this requires a re-examination of the relevant projects, which can result in significant costs and delays.
  • Tool development will require a lot of work. In some cases, if the old transaction mode is abandoned and replaced by the latest one, it is difficult for users to access the old version when the new transaction type is not well accepted by users and they continue to use the old version.

Implementation issues

  • Keeping the original mode and 1559 will add complexity to Ethereum.
  • Will eip-1559 increase the risk of DoS attacks on Ethereum?
  • What is the impact of this on the order of transaction processing? What is the second order chain effect?
  • What is the impact of eip-1559 on various parts of the client stack (such as devpp2p, MemPool, database, etc.)?
  • Can the problem that eip-1559 attempts to solve be resolved at the client level without consensus changes?
  • For eip-1559, the implementation is too complex.

economic problem

  • In economics, there is no study to explain the impact on consumption costs.
  • Why increase the basic cost by 12.5% instead of other values?
  • Before eip-1559 was released on the main network, there was no incentive to take advantage of it.

political issue

  • What if miners don’t want to upgrade support for eip-1559?
  • How do we determine that there is enough support for eip-1559 to make such a big change?

It will be valuable for practitioners and researchers to try to address or refute these concerns.

Alternatives to eip-1599

When asked about the alternative to eip-1559, the only alternative known to some respondents was eip-2593, also known as the “escalator fee,” proposed by Dan Finlay of metamask.

Willingness to implement

Respondents were asked what changes were needed to support the eip-1559 project, how long they expected to take to make those changes, and whether EIP advocates could do something to make their work easier.

Here is a list of improvements that must be made to support eip-1559:

  1. Rewrite block reward calculation;
  2. Redo transaction processing logic, including update sequence and parse library;
  3. Update the software of mine pool;
  4. Rewrite the contract to use new opcodes and delete the ones that are not recommended;
  5. Ensure that third-party dependencies still work (i.e., chainlink Oracle);
  6. Redo the “gas abstraction” function, so that users can consider 1559 transactions;
  7. Update the client software;
  8. Change the API used to estimate gas prices;
  9. Redo UI to comply with 1559 transaction mode;
  10. Adapt to the change of RPC interface;
  11. Using 1559 enabled test network and other available resources for extensive testing;
  12. Change the API used to share chunks and transaction data with users.

Most of the changes mentioned are project specific. To protect the privacy of the project, the above list summarizes similar changes mentioned in the project and aims to qualitatively outline the impact of 1559 on existing projects.

When asked what would help them prioritize support for eip-1559, the most common answers are as follows:

  1. Test network (5 items mentioned)For most projects, having a common eip-1559 test network suitable for application use (for example, JSON RPC support for 1559) will be the first step. Some projects want to use eip-1559 supported public network testing, while others are willing to wait for 1559 enabled test networks (e.g., ropsten, goerli, etc.).
  2. Including network upgrade (mentioned 3 items)Several projects mentioned that they wanted to wait until eip-1559 was properly accepted by the network upgrade before it was implemented.
  3. Incentives (3 items mentioned)Some projects mentioned that incentives to use eip-1559 (e.g., lowering gas prices) would prompt them to implement them as soon as possible.

Other items mentioned in the project include updating tools (e.g. for eip-1559 ethers.js Or Web3. JS support), relevant RPC endpoints in the client, clear opcode definition and retirement schedule, public API endpoints for estimating the basic cost of the next block, appropriate documentation, clear communication, support channels, and the correct economic model for EIP.

Similarly, respondents were asked what caused them to delay the implementation of eip-1559. In this case, the most common response is “None” (referring to 7 items), indicating a strong desire to support EIP. According to the order of frequency, the following potential problems are raised:

  1. EIP / specification issues (3 items mentioned)If the customer finds significant problems in the specification early, the project will use this as a dangerous signal to delay support for eip-1559.
  2. Uncertainty of main network deployment (2 items mentioned)Just as some projects hope to include eip-1559 in network upgrades to give priority to their implementation, others point out that uncertainty surrounding the main network will cause them to delay implementation.
  3. Competitive priority (2 projects mentioned)Some projects mentioned that other high priority features they were developing might bring more value to their users than supporting eip-1559.
  4. Poor documentation and support (only 2 items mentioned)If it’s not clear how the project adds support for eip-1559, how to test it correctly, and where to go when problems arise, some of them delay adding eip-1559 support until the documentation is complete.
  5. Lack of community support (2 projects mentioned)Some projects have mentioned that if they think that eip-1559 will be opposed by some communities, they will not be willing to increase support for it because it may be a waste of energy and they do not want to participate in governance.

In addition, other potential issues mentioned include: poor test network support, lack of appropriate tools, possible long-term use of simple transactions and inability to test the economics of eip-1559 before deploying the main network.

Finally, it is worth noting that some respondents who are very opposed to eip-1559 mentioned that they would not implement it under any circumstances, of which 6 / 7 (~ 86%) came from miners.

Implementation schedule

We asked when the project started to add support for eip-1559, how much time they planned to invest in doing so, and when they would like the old legacy transactions to be supported once eip-1559 was enabled on the main network.

When will your project begin to support eip-1559?

[Research Report] will the eip-1559 proposal make eth appreciate or face risks?

When EIP exists in the existing test network, most respondents are willing to start implementing changes. Again, many projects have mentioned that EIP, while not requiring significant development changes, may result in the need to re audit certain parts of its code, which will increase costs.

How much time (for a full-time employee) do you need to spend on R & D to support eip-1559?

[Research Report] will the eip-1559 proposal make eth appreciate or face risks?

As you can see, most projects expect to take a relatively short time to add support for eip-1559.

Finally, after eip-1559 was launched, the project strongly hopes to see a longer period of time to support the old processing methods on the main network. More than half of the respondents said they would like to see traditional deals supported within 12 months of its launch in 1559. This confirms that recent changes to eip-1559 to achieve long-term support for legacy transactions may be accepted by the wider community.

After deploying eip-1559 on the main network, how long would you like to see support for transactions prior to 1559?

[Research Report] will the eip-1559 proposal make eth appreciate or face risks?

conclusion

According to the community survey, although eip-1559 is indeed controversial, more than 60% of respondents hope to see eip-1559 deployed on the main network next year, but there are also concerns about miners’ lack of support for EIP.

The number of concerns and issues raised highlights the need for more eip-1559 documentation and the potential to refute some of the controversies. Finally, respondents made it clear that frequent communication between multiple platforms, upgrading tools and support channels would greatly help them increase their support for eip-1559.

Comments and concerns

What is the motivation behind the eip-1559 proposal?

EIP details some of the motivations, as are vitalik’s latest FAQ.

Is there any objection to the proposal?

There are no large numbers of opponents, but the miners’ groups seem to be the most. It is worth noting that non miners are also opposed to the current proposal. Concerns include the lack of formal economic analysis or a preference for other alternatives, such as eip-2593.

Is there a simpler proposal to deal with transaction fees?

Another alternative is eip-2593, which can help send UX transactions better without incurring the basic cost of being burned.

Plan change

What are the obvious differences that users will see?

EIP makes it easier for developers to estimate the gas price of their transactions, and since blocks always have extra space, users should see their transactions packaged faster in most cases.

From the perspective of contract developers, what changes will it bring?

EIP will change the block header to add a basic fee field and change the transaction format by removing the gas price field and introducing the fee cap and prompt fields.

Planned release

How will it be released?

EIP will first be launched on the eip-1559 test network. Once proven usable, it will be upgraded through the network and, if accepted, deployed on the existing test network and main network.

Is it backward compatible?

No, EIP will need to activate network upgrade and update nodes to support this operation.

How to deal with the old trading mode, will there be a transition period?

The original version of EIP has a transitional period in which the old approach is phased out. The current version of EIP includes the old transaction in the block by treating the “gas price” of the old transaction as the “cost cap” and setting the difference between the “cost cap” and the “basic cost” as the miner’s “hint”. This means that old style deals may pay more for their tips, but will always be supported.

Gas price, basic cost, cost cap and promotion

Will eip-1559 increase or decrease gas?

Eip-1559 should not keep gas prices up or down. Gas price is a function of the demand for Ethereum block space. EIP will make gas prices easier to predict (and help users avoid paying too much), but will not reduce the “equilibrium price” of gas. In other words, eip-1559 won’t make our gas price from 300 to 30 Gwei.

What is the basic cost? What is the fee cap and tip?

  • Basic costIs the lowest gas price a transaction must pay to be included in a particular block. Depending on the full or empty setting of the block. If the block is more than 50% full, it will rise; if the block is less than 50%, it will decrease. The basic cost is the burned part of the transaction cost.
  • Cost capIt’s the maximum amount a transaction is willing to pay, including basic fees and miner tips. If the user suspects that the cost cap for the next block will increase, the user is allowed to set the cost cap above the current base cost. The extra part of the fee cap (fee cap minus basic fee minus tip) will be returned to the user.
  • TipsIt’s part of the transaction costs sent to the miners.

Willingness to implement and obstacles

Respondents were asked what changes needed to be made to support the eip-1559 project, how long they would like to make those changes, and whether EIP advocates could take anything to make their work easier.

This is a list of changes that must be made to support eip-1559:

  1. Rewrite block reward calculation;
  2. Redo transaction logic, including updating serialization and parsing library;
  3. Update the software of mine pool;
  4. Rewrite the contract to use new opcodes and delete the ones that are not recommended;
  5. Ensure that third-party dependencies still work (i.e., chainlink Oracle);
  6. Redo the “gas abstraction” function, so that users can consider 1559 transactions;
  7. Update the client software;
  8. Change the API used to estimate gasoline prices;
  9. Redo UI to match 1559 style transactions;
  10. Adapt to the change of RPC interface;
  11. Extensive testing / quality checking using 1559 test networks and other available resources;
  12. Changed the API used to share frozen and transaction data with users.

Most of the changes mentioned are project specific. To protect the privacy of the project, the above list summarizes similar changes mentioned in the project and aims to qualitatively outline the impact of 1559 on existing projects.

When asked what can help them prioritize eip-1559 support, the most common answers are as follows:

  1. Test network (5 items mentioned)For most projects, having a common eip-1559 test network suitable for application use (for example, JSON RPC support for 1559) will be the first step. Some projects are eager to use any public eip-1559 test network, while others prefer to wait for their already supported test networks (e.g., ropsten, goerli, etc.) to add eip-1559 support.
  2. Including network upgrade (mentioned 3 items)Several projects mentioned that they wanted to wait until eip-1559 was properly accepted by the network upgrade before it was implemented. The implementation of these projects does not guarantee that some projects on the main network are not worth it.
  3. Incentives (3 items mentioned)Some projects mentioned that incentives to use eip-1559 (e.g., lower gasoline prices) would prompt them to implement them as soon as possible.

Other items mentioned in the project include updating tools (e.g. for eip-1559 ethers.js Or Web3. JS support), relevant RPC endpoints in the client, clear opcode definition and retirement schedule, public API endpoints for estimating the basic cost of the next block, appropriate documentation, clear communication, support channels, and the correct economic model for EIP.

Similarly, respondents were asked what caused them to delay the implementation of eip-1559. In this case, the most common response is “None” (referring to 7 items), indicating a strong desire to support EIP. According to the order of frequency, the following potential problems are raised:

  1. EIP / specification issues (3 items mentioned)If the customer’s early efforts identify a major problem in the specification, the project will use this as a sign to delay the implementation of eip-1559 support.
  2. Uncertainty of mainnet deployment (mentioned 2 items)Just as some projects hope to include eip-1559 in network upgrades to give priority to its implementation, others point out that uncertainty surrounding mainnet will cause them to delay implementation.
  3. Competitive priority (mention 2Projects):Some projects have mentioned that other high priority features they are developing may bring more value to their users than supporting eip-1559 and will implement these priorities first.
  4. Poor documentation and support (only 2 items mentioned)If it’s not clear how the project adds support for eip-1559, how to test it correctly, and where to go when problems arise, some of them delay adding eip-1559 support until it’s in place.
  5. Lack of community support (reference 2Projects):Some projects mentioned that if they felt some opposition to eip-1559, they would not be willing to increase support for it, both because of the potential waste of energy and because they did not want to be involved in politics.

In addition, other potential problems mentioned include poor test network support, lack of appropriate tools, possible use of simple transactions for a long time, and failure to test the economy of eip-1559 before deploying the main network.

Finally, it is worth noting that some respondents who are very opposed to eip-1559 mentioned that they would not implement it under any circumstances. Among them, 6 / 7 (~ 86%) came from miners.

Implementation schedule

We asked when the project started to add support for eip-1559, how much time they planned to invest in doing so, and how long they would like to see legacy transactions supported once eip-1559 was enabled on the main network.

When will your project begin to add eip-1559 support?

[Research Report] will the eip-1559 proposal make eth appreciate or face risks?

When EIP exists on existing test networks, most respondents are willing to start implementing changes. Again, many projects mentioned that EIP, while not requiring significant changes in development, may result in the need to re audit certain parts of its code, which will increase process and cost.

The answer to another question also supports this point:“How much time (for a full-time person) do you expect to spend supporting eip-1559?”

[Research Report] will the eip-1559 proposal make eth appreciate or face risks?

As we can see here, most projects expect to take a relatively short time to add support for eip-1559.

Finally, with eip-1559 enabled, the project strongly hopes to see legacy transactions supported on the main network for a long time. More than half of the respondents said they would like to see traditional deals supported within 12 months of its launch in 1559. This confirms that recent changes to eip-1559 to achieve long-term support for legacy transactions may be accepted by the wider community.

After deploying eip-1559 on the main network, how long would you like to see support for transactions prior to 1559?

[Research Report] will the eip-1559 proposal make eth appreciate or face risks?

conclusion

This community expansion work shows that although eip-1559 is indeed stimulated by many projects, and more than 60% of the respondents hope to see eip-1559 deployed on the main network next year, there are also concerns that EIP miners have made inadequate designation and significant backtracking for EIP.

The number of concerns and questions raised highlights the need for more eip-1559 interpreters and the potential to refute specific concerns. Finally, respondents made it clear that frequent communication between multiple platforms, upgrading tools and support channels would greatly help them increase their support for eip-1559.

We hope the report will be useful to the community! In this extended study, we will open the questionnaire to other projects that want to share ideas, and try to contact more wallets and exchanges to increase their sample size. An updated version of this report will be released after contacting a sufficient number of other projects.

Frequently asked questions

Several questions were repeatedly raised in the interviews with the project. Here are questions and some answers for the community.

General comments and concerns

What is the motivation behind the proposal?

  • EIP details some of the motivations, as are vitalik’s latest FAQ.

Is there any objection to the proposal?

  • Yes, as this survey shows. There are no large numbers of opponents, but the miners’ groups seem to be the most. It is worth noting that non miners are also opposed to the current proposal. Concerns include the lack of formal economic analysis or a preference for other alternatives, such as eip-2593.

Are there any other simpler suggestions to deal with transaction costs?

  • Another alternative is eip-2593, which can help send UX transactions better without incurring the basic cost of being burned.

Plan change

What are the significant differences that end users can expect?

  • EIP should make it easier for application developers to estimate the gasoline price of their transactions, and since blocks almost always have extra space, users should see their transactions included more quickly in most cases.

From the perspective of smart contract developers, what changes will it bring?

  • The EIP will change the block title to add a base fee field and change the transaction format by deleting the gasoline price field and introducing the fee cap and prompt fields.

Release plan

How will it be launched?

  • EIP will first be launched on the interim eip-1559 test network. Once proven usable, it will go through the network upgrade process, and if accepted, it will be deployed on the existing test network and main network.

Is it backward compatible?

  • No, EIP will need to activate network upgrade and update nodes to support this operation.

How will the Internet handle old-fashioned transactions? Will there be a transition period?

  • The original version of EIP has a transition period in which the old transactions are phased out. The current version of the EIP includes the old transaction in the block by treating the “gasoline price” of the old transaction as the “cost cap” and setting the difference between the “cost cap” and the “basic cost” as the miner’s “hint”. This means that old style deals may pay more for their tips, but will always be supported.

Gasoline prices, basic costs, cost caps and tips

Will eip-1559 raise / lower gasoline prices?

  • Eip-1559 should not keep gasoline prices up or down. Gasoline price is a function of the demand for Ethereum block space. EIP will make natural gas prices more predictable (and help users avoid overpayment), but will not reduce the “equilibrium price” of natural gas. In other words, eip-1559 won’t make our gas price from 300 to 30 Gwei.

What is the basic cost? What’s the maximum cost? What is the tip?

  • Basic costIt’s a transaction that has to be paid_ Lowest_ The price of gasoline can only be included in a specific block. Depending on the full or empty setting of the block. If the block is more than 50% full, it will rise; if the block is less than 50%, it will decrease. The basic cost is the burned part of the transaction cost.
  • Cost capThe deal is willing to pay_ Highest_ Amount, including basic fees and tips for miners. If the user suspects that the cost cap of the next block will be increased, the user is allowed to set the upper limit of the cost to be higher than the current basic cost. The extra part of the fee cap (fee cap minus basic fee minus tip) will be returned to the user.
  • TipsIt’s part of the transaction costs sent to the miners.

about

ChinaDeFiIt’s ChinaThe first investment research brand focusing on defiEvery day, from nearly 900 content of more than 500 high-quality information sources in the world, we will find more in-depth and more systematic content to provide investment assistance suggestions to the Chinese market at the fastest speed. Welcome to communicate with Gavin and iris and join the chinadefi community, and continue to pay close attention to projects that are really using blockchain to create value at home and abroad.