Summary of the latest meeting of Ethereum core developers: The next dedicated testnet for the Pectra upgrade will be launched next week
Original Title: 《Ethereum All Core Developers Execution Call #190 Writeup》
Author: Christine Kim
Compiled by: Luccy, BlockBeats
Editor’s Note:
The Ethereum All Core Developers Execution (ACDE) call is held bi-weekly, primarily to discuss and coordinate changes to the Ethereum Execution Layer (EL). This was the 190th ACDE call, where topics such as EIP 7251 and 7702, the Ethereum Object Format (EOF), and PeerDAS were highlighted.
During the meeting, Beiko also proposed a new EIP discussion template and emphasized the importance of the Ethereum developer resource page "eth-clients." Christine Kim, Vice President of Research at Galaxy Digital, took detailed notes on the key points of this meeting, which BlockBeats has compiled as follows:
On June 20, 2024, Ethereum developers gathered on Zoom for the All Core Developers Execution (ACDE) call #190. The ACDE call is a series of meetings held every two weeks, hosted by Tim Beiko, the Ethereum Foundation's Protocol Support Lead, where developers discuss and coordinate changes to the Ethereum Execution Layer (EL).
This week, developers presented their progress on implementing various code changes, including Ethereum Improvement Proposals (EIPs) 7251 and 7702, as well as the Ethereum Object Format (EOF) and PeerDAS.
Beiko also shared a proposed template for EIP discussions, and EF Development Operations Engineer Parithosh Jayanthi introduced a new GitHub page called "eth-clients" for configurations of Ethereum mainnet and testnets. Finally, Beiko reminded the team that the next ACDE call would take place on July 4, during the U.S. holiday. Although he would not attend, EF researcher Alex Stokes would host the meeting in his place.
Pectra Devnet 1
At the start of the meeting, developers discussed Pectra Devnet 1. The developers aim to launch the next dedicated testnet for the Pectra upgrade next week. Compared to the previous testnet Devnet 0, the new testnet primarily has the following specification changes:
- Added EL-triggered validator merges
- Replaced EIP 3074 with EIP 7702
The EL client teams are advancing their Devnet 1 implementations. Representatives from each major Ethereum EL team shared their latest progress:
- Besu: The first version of their Devnet 1 implementation is ready, but there are still some missing tests, and the team has questions about the specifications for gas pricing operations.
- Nethermind: They are reviewing their Devnet 1 implementation, merging various code changes and software branches, and writing test cases.
- Erigon: They have a draft implementation of EIP 7702 and are reviewing the latest changes to EIP 7251, which include the addition of EL-triggered validator merges.
- Reth: Their Devnet 1 implementation should be completed before the next ACD meeting.
- EthereumJS: The team is working on the implementation of EIP 7702 and expects to have Devnet 1 ready in about a week.
- Geth: The team is also working on the implementation of EIP 7702.
As the client teams shared their progress, a developer named "Racytech" asked whether the implementation of EOF would be excluded from Devnet 1. Beiko confirmed that EOF would not be tested on Devnet 1 and that the implementation of EOF should be based on the latest Pectra specifications rather than the final Deneb specifications. EF Development Operations Engineer Barnabas Busa emphasized that this strategy would make debugging EOF more challenging, as the Pectra specifications are changing and debugging is happening simultaneously. "Even so, this is still the right approach because we want the teams to work on the core Pectra EIPs and fix those bugs. So, this is more reasonable than basing it on a previous fork and then trying to base it on Prague at the last minute while discovering all those bugs," Beiko said. Prague is the name of the EL for the Pectra upgrade.
EIP 7702 Update
Although developers are working hard to implement EIP 7702 in Devnet 1, there may be significant changes to the code to support new features, particularly the ability for users to revoke EIP 7702 transaction authorizations.
A Geth developer named "Lightclient" proposed new changes to support EIP 7702 revocation. He explained, "It's not easy to see what authorizations you have active at any given time, and it's hard to know which have been revoked and which have not. There are some design patterns that can avoid these situations, but in reality, we may not know how wallets will use this feature. To better defend against this, I remember Vitalik mentioning in the last call, or in a breakout room, that rather than making authorizations a temporary transaction that must be included in the transaction, we could save these authorizations so you can see at any time what code has been assigned to the account."
A developer named "Arik" expressed that he preferred the current version of EIP 7702 over the changes proposed by Lightclient. Arik stated that there would be some specific use cases under the new EIP 7702. Lightclient countered these comments, stating that all use cases supported by the old version of EIP 7702 are still retained, and a robust revocation mechanism has been added. However, Lightclient also pointed out that these changes make the implementation of the EIP in clients more complex. Beiko noted in the meeting chat that finalizing the design of EIP 7702 is the "number one unresolved specification issue" for Pectra and suggested that developers organize another breakout discussion call to reach a consensus on its revocation mechanism. "If by the next meeting we have a version that everyone is generally satisfied with, whether it's this version or an adjusted version, that would be great to include as part of Devnet 2," Beiko said.
Besu developer Daniel Lehrner questioned whether developers should spend time implementing the current version of EIP 7702 in Devnet 1 if the specifications might change in Devnet 2. He said, "These proposals are quite different, so we might implement something that we will have to partially discard in a few weeks due to 7702 changing again." Richard Meissner, co-founder of the crypto wallet Safe, stated that in his view, it would be helpful for wallet teams to implement some version of EIP 7702 on the testnet. Beiko added that the implementation of the current form of EIP 7702 would not significantly delay the progress of the client teams in his view, and trying to remove it from the Pectra specifications now could complicate Devnet 1.
EOF Update
Regarding EOF, Busa stated that his team would test the EOF implementation after implementing other Pectra code changes. The readiness of EOF is being tracked on GitHub. Currently, two EL client teams, Besu and Reth, have implemented all EOF EIPs, while other teams are still working to implement all EOF EIPs. A representative from the Erigon team noted that additional test cases simulating Ethereum blocks containing EOF transactions would be useful. EF testing team member Mario Vega mentioned in the chat that his team is releasing more EOF tests and will share more in the next EOF breakout meeting. Independent Ethereum protocol developer Danno Ferrin stated that since the EOF specifications are now finalized, the EOF calls have changed from weekly to bi-weekly, and the current focus of EOF work is on testing and client implementation.
PeerDAS Update
Busa briefly updated on the progress of PeerDAS implementation work. He stated that the next PeerDAS Devnet would be launched in one to two weeks. As discussed in the last ACDC meeting, PeerDAS will be developed based on the finalized Dencun specifications rather than the Pectra specifications.
Announcements
Beiko proposed a new default template for EIP authors to use in Ethereum Magicians posts. This new template is designed to facilitate high-quality reviews of EIPs.
Jayanthi introduced the "eth-clients" GitHub page, which provides various resources for Ethereum developers, including configuration specifications for Ethereum testnets and mainnets. Currently, the page is maintained by representatives from all CL client teams. Jayanthi called for representatives from the EL client teams to volunteer to help maintain this GitHub page.
Lightclient inquired about the progress of the client teams in handling historical data expiration. No updates on this were shared during the meeting.
Beiko stated that he would not host the next ACDE call on July 4, during the U.S. holiday. At that time, EF researcher Alex Stokes would host the meeting in his place.