Author: Christine Kim, Vice President of Galaxy Digital Research Team Translation: Shan Oppa, Bitchain Vision

On February 28, 2024, Ethereum developers held their 182nd All-Core Developer (ACDE) conference call through Zoom.ACDE conference call is a biweekly conference series for developers to discuss and coordinate changes to the Ethereum Execution Layer (EL).This week’s meeting was chaired by Danny Ryan, a researcher at the Ethereum Foundation (EF).The developers discussed test updates for the Dencun upgrade and several candidate EIPs for Pectra.The most intensely discussed EIP proposed inclusion in Pectra is related to code changes related to account abstraction.Account abstraction (AA) is designed to introduce a higher degree of programmability for externally owned accounts (EOA), which are accounts controlled by users on Ethereum rather than smart contract code.

Dencun Update

Barnabas Busa, an engineer at the Ethereum Foundation Development and Operations (DevOps), shared the latest developments in the final test of the Dencun upgrade.The Ethereum Foundation announced on Tuesday, February 27 that the upgrade is now officially scheduled to be activated on the Ethereum main network on March 13, 2024.As discussed last week’s ACD conference call, developers are testing the final version of client software on the mainnet shadow fork, a test network that mirrors the blockchain state of the Ethereum mainnet andActivity.Busa said developers have already conducted different types of “spam tests” on the mainnet shadow fork.The nodes performed very stably in these tests, with network participation rates remaining close to 100%.While no issues occur, Busa notes that spam testing does seriously affect nodes in terms of computer resources (particularly memory and CPU occupancy).

Busa then reminds attendees that the Goerli testnet will be deprecated soon.Anyone using the test network should move its operations to other Ethereum test networks by April 17.Busa said he has noticed that some of the large validator node operators on Goerli have retired their machines.This led to a delay in Goerli’s network finalization on February 28, but the Goerli network appears to have recovered.Ryan points out that Goerli’s network participation rate is already quite low, hovering around 70%.“Honestly, I don’t expect [the participation rate] to last until April 17,” Busa said.“But this is still worthy of attention.”

Busa asked his team when should stop Devnet 12, a dedicated test network launched last November for client teams to test their Dencun upgrade implementation.In case you need to test any last minute client version for Dencun, the developer agreed to shut down Devnet 12 shortly after Dencun upgraded to launch the Ethereum mainnet.

Traceability of Pectra Upgrade EIP

Next, the developers discussed two retrospective Ethereum Improvement Proposals (EIPs) for Pectra upgrades.Traceability EIP is a code change that traceable to the Ethereum protocol that already exists to a large extent but needs clarification to resolve specific edge cases.The first traceability EIP, EIP 7610, extends a rule that limits the creation of smart contracts to addresses with pre-existing storage space.For more background information about this code change, see the previous meeting minutes here.

One concern about EIP 7610 is whether it will affect Verkle, a code change that developers are preparing for an upgrade after Pectra.Geth developer Gary Rong explained how EIP 7610 does not pose any problem with Verkle upgrades in the future.Danno Ferrin, a Hedera Hashgraph engineer and Besu client maintainer, raised some unresolved concerns about how EIP 7610 might affect Verkle, saying he will share these concerns on the Ethereum Magician discussion board for Ethereum Improvement Proposal 7610.

The second traceability EIP discussed by developers is EIP 7523, which will officially prohibit empty accounts from appearing in the state of Ethereum and Ethereum test networks.Ryan said he will carefully check who is performing the analysis to ensure that any accounts on the Ethereum network (main or testnet) will not be affected after the rule is implemented and will discuss the issue again on the next ACDE call..

Account abstraction for Pectra

Next, the developers discussed potential account abstraction EIPs included in Pectra.On February 28, a number of developers participated in a dedicated AA meeting discussing the macro goals of the program and the various EIPs that can be implemented in the short and long term to achieve these goals.Ethereum co-founder Vitalik Buterin said of AA’s goal, “The long-term [target] is this fundamental desire that we must eventually have some type of account system that allows key rotation and [On the other hand] key depreciation to enable us to resist quantum computing. III. Allow batch processing…[and] allow sponsor transactions and some other smaller functions, of course, the first two goals are obviously not possible through EOA, thus presenting a fairly clear case of moving the ecosystem to a place that goes beyond the Ethereum account center, but then the discussion turns to what are the actual means of achieving these goals, and some less clear specifics, as well asWhat is a short-term roadmap that actually makes a profit for what people want in the short term, but at the same time is compatible with those long-term [goals].”

In the short term, developers are evaluating three main AA EIPs, namely EIPs 3074, 5806 and 7377.Developers participating in the conference call differed on the pros and cons of EIP 3074 and 5806.One of the points of the debate is to what extent EIP 3074 requires users to double signatures of transactions and rely on decentralized means to initiate transactions, as well as other related complexity and security of EIP 3074 compared to 5806Sexual debate.Developers generally believe that EIP 7377 is the least controversial AA EIP because it is orthogonal to the other two AA EIPs in terms of use cases.EIP 7377 is designed to help users easily migrate their assets from Ethereum accounts to new smart contract wallets, while the other two EIPs focus primarily on creating new AA features that support bulk transaction authorization and gas fee sponsorship.

The developers did not agree on these three EIPs and agreed to continue discussing them in the coming weeks.

Other EIP proposals from Pectra

In addition to account abstraction EIPs, the developers briefly discussed several other EIPs that were proposed to include Pectra upgrades:

EIP 7623: Increase calldata gas fees: This proposal suggests increasing the regular transaction costs on Ethereum that are primarily used for data availability.By adjusting the calldata gas fee on Ethereum, this EIP reduces the number of call data transactions that can be reasonably placed into a block, thereby reducing the maximum size of the block.Reducing the block size allows more blob transactions.Danny Ryan recommends developers involved in the discussion to review the EIP in the coming weeks.
EIP 2537: BLS12-381 Curve Operation Precompilation: This proposal introduces a new cryptographic signature scheme to Ethereum, which has been approved for inclusion in the Pectra upgrade.Antonio Sanso, one of the authors of the proposal, raised some questions about its implementation.Danny Ryan recommends logging the questions and distributing them to developers for further discussion outside of the call.
EIP 5920: PAY Opcode: This proposal creates a new operation that allows the user to send ETH to the address without triggering any address’s function.Geth developer Marius van der Wijden said that after further discussion of the EIP with other teams, it was found that the proposal’s testing was more complex than expected.Van der Wijden also pointed out that the norms of the proposal are not yet perfect.Ferrin added that the PAY opcode is currently specified to use the same code number as the other opcode (AUTH opcode) and therefore needs to be corrected by its author.
EIP 7609: Reduced temporary storage pricing: This proposal recommends lowering the temporary storage opcode price for common smart contract use cases such as maintaining reentrable logs.Van der Wijden and Ryan both agreed to collect data on how temporary storage opcodes are used after the Dencun upgrade is launched and then rediscuss its pricing.
EIP 7639: Stop providing historical data before proof of stake: This proposal sets a timetable for the execution layer (EL) clients so that they stop providing historical data before the merge upgrade.The motivation for this code change is to reduce the amount of data that Ethereum nodes need to store permanently.The proposal also promises nodes to build and retrieve them from external sources in a standardized manner.Teku developer Mikhail Kalinin noted that the EIP relies on another EIP (EIP 6110), which was approved for inclusion in the Pectra upgrade on a previous ACD conference call.The developers agreed to review EIP 7639 in more detail in the coming weeks.

Engine API and JSON RPC changes

In addition to the above topics, Ethereum core developers also discussed changes to engine API and JSON RPC.

Teku developer Mikhail Kalinin raised some questions related to the implementation of a confirmation rule, a CL mechanism that confirms whether a block will remain under a specific assumption in about 12 seconds (one slot).Canonical chain and finalize.This is a powerful feature because many applications built on Ethereum can operate with information from early block acknowledgments.However, to expose data about early block acknowledgments, some changes are required to the Ethereum Engine API and JSON RPC.Due to limited call time, Ryan recommends discussing the changes in more detail on the ACD call next week or next week.

Light client breakout room meeting

Ryan reminds developers that there will be a dedicated meeting next Wednesday (March 6) to discuss the light client roadmap for Pectra upgrades.For background information on light client discussions, see previous meeting minutes.

New Ethereum client version proposal

Finally, van der Wijden proposed a suggestion to build a new version of Ethereum client to save 550GB of bandwidth for nodes during the initial synchronization process.Van der Wijden said he is preparing a formal EIP for the new version, but a draft of its specification can be found here.Ryan encourages developers to view drafts and ask any questions on Discord.