Meeting Time: Thursday, September 25, 2025 at 14:00 UTC (90 minutes)
ACDE TL;DW by @timbeiko [Copied from Eth R&D Discord]
!
And, lastly, this was my last ACD before my leave. @adietrichs will be taking over ACDE until EOY. See you after Fusaka 
!
Tweet thread for quick notes - https://x.com/poojaranjan19/status/1971213098733158747
The meeting covered updates on the Fusaka upgrade and associated bug bounty competition, with discussions about testnet schedules and client releases for the Shanghai upgrade. The team reviewed various Ethereum Improvement Proposals (EIPs) and their implementations, including Block Lab Access Lists, Beta Zero for EIP7732, and a new delegate mechanism for smart contract factories. The conversation ended with discussions about formalizing the role of EIP champions and exploring potential new proposals for future Ethereum forks.
The meeting focused on updates and plans related to the Fusaka upgrade and the associated bug bounty competition. Fredrik reported that the competition is progressing well, with more valid reports as the deadline approaches, and confirmed that no critical issues have emerged that could affect the mainnet upgrade timeline. Tim outlined the testnet schedule, which includes forking Holesky on October 1st, followed by forking Sepolia and Hoodi, with BPOs activated one week after each testnet’s Fusaka activation.
The team discussed the upcoming formal announcement of client releases for the Shanghai upgrade, with blog posts planned for tomorrow and additional releases expected by Monday. Several clients including Geth, Lighthouse, Nimbus, and Prism are still working on their releases due to various build and process issues, though Nethermind has already released today. The team confirmed that Panel Ops will run a Shadow Fork of Holesky tomorrow with the available releases, and discussed raising the default gas limit on mainnet to 60 million as part of the Shanghai release, which will be implemented once the mainnet releases are completed.
Tim announced that EIP authors should review and approve PRs moving their EIPs to last call for the testnets, as the mainnet launch is imminent. Toni provided an update on Block Lab Access Lists, noting no changes to the EIP and clarifying client handling of self-destructs and precompiles, with a target devnet release by October. Terence mentioned the release of Beta Zero for EIP7732 and progress on container types and stage transition functions, aiming for a devnet release by early November. Several Ethereum clients, including Besu, Geth, and Nethermind, have started working on BAL implementations. Tim suggested using the latest CL spec (Beta 0 from V1.6) as the basis for both BAL and PBS devnets to ensure compatibility with the Fuji testnet. The deadline for new proposals for Glamsterdam was set to coincide with the mainnet release of Fuji in the next month.
Tim invited participants to propose EIPs for Glamsterdam by opening a PR against the Meta EIP. Justin raised a question about managing early EIPs for CFI while juggling smaller PFI issues, to which Tim responded by advocating for limiting new proposals and not moving EIPs from CFI to SFI until late stages of implementation for SFI EIPs like BALs and EPBS. Tim suggested prioritizing EIPs and reviewing them when there is more breathing room in implementation, rather than rushing to promote them from CFI to SFI.
Tim and Wolovim discussed a proposal by Mark to formally define the role of an EIP champion and the associated responsibilities. Wolovim suggested requiring a primary point of contact for each proposed EIP in a given fork to improve communication and reduce friction. Tim agreed with the idea and proposed that the person who opens the PR for an EIP should be considered the champion. Wolovim offered to have the protocol support team identify champions for existing EIPs, and both agreed to review the PR before the next AllCoreDevs meeting.
The team discussed the role and identification of EIP champions, with a focus on whether champions should be explicitly named in Meta EIPs. They agreed to test the concept on ForkCAST for the Glamsterdam fork before potentially formalizing it in either the meta EIPIP or EIP header field. Tim noted that while champions aren’t necessarily authors, there’s value in having an explicit champion for coordination purposes. The group also briefly discussed the possibility of PFI EIP presentations, with Hadrien presenting his Missouri IP proposal for inclusion in Amsterdam.
Hadrien presented EIP-7819, which proposes a new delegate in RLP code to allow smart contract factories to instantiate objects equivalent to current proxies or clones using ERC-1167 or ERC-1967. He explained that this would reduce costs, create smaller objects (23 bytes vs. 70-100+ bytes), and improve upgradability by using existing delegation primitives. The proposal aims to avoid previous bugs and attacks associated with unconventional storage slots in clones and proxies, with the implementation already existing in execution clients.
The meeting focused on a proposal for a CREATE-like mechanism to create and update delegations, with discussions on its potential impact and implementation. Hadrien explained that the mechanism could reduce state growth by approximately 25% of future contract deployments, and the GAST team’s research showed that 89.1% of contracts are deployed from factories. Participants raised concerns about the trade-off of increased interaction costs and discussed the use cases for smart wallets. The group also briefly touched on the possibility of implementing SWAPN and DUPN opcodes without breaking changes to address the “stack too deep” error in Solidity.
nfCj%?55)nfCj%?55)YouTube Stream Links: