Ethereum Foundation Update - March 2025

Ethereum’s weekly All Core Developer meetings can be overwhelming, hence this “Checkpoint” series strives to provide concise high-level insights with a planned frequency of every 4-5 meetings, depending on developments in core engineering. Refer to the first update here. Future updates will be published here on the Ethereum Foundation blog.

Recent meetings have concentrated on testnet enhancements for Pectra in preparation for the Pectra mainnet fork, alongside simultaneous development of the forthcoming fork, Fusaka.

Pectra

Both prominent testnets, Holešky and Sepolia, experienced the Pectra fork, which unveiled configuration challenges demanding urgent resolution. The complications on both testnets pertained to their specific configurations and would not have occurred on mainnet, indicating that the fork code and setup were designed with mainnet considerations, overlooking essential configuration adjustments essential for a smooth fork on testnets. A new testnet, Hoodi, has been launched to evaluate Pectra capabilities, and the Pectra Audit Competition will conclude on March 27th.

Holešky

The testnet has been restored, and the post-mortem report is accessible here. Holešky operates as a permissionless testnet with an open validator pool, necessitating significant coordination among independent entities to achieve finalization after a majority of the network transitioned to a non-canonical chain. Although the network’s finality was restored, the situation led to numerous validator exits, causing the exit queue to extend to a year, with many Pectra features reliant on exits needing an alternative platform for testing.

In light of this, a new enduring testnet, Hoodi, has been initiated and is set to implement the Pectra fork on March 26th. Liquid staking protocols and other validator operators can utilize Hoodi for Pectra feature evaluations.

Sepolia

The Pectra incident on Sepolia was swiftly addressed, with the post-mortem available here. The configuration issue on Sepolia was less critical, did not lead to the significant forking experienced by Holešky, and the task of organizing a solution was much quicker, owing to Sepolia’s permissioned validator set.

Timeline

The latest meeting addressed when the core developers believed it would be wise to establish a date for the mainnet activation of Pectra. Developers concurred that the prerequisites for setting a date included a successful Hoodi fork, with adequate time for observation, and liquid staking protocols effectively testing Pectra features on the new testnet.

Provided unforeseen complications do not arise, a mainnet fork epoch is anticipated to be selected within the next 2-3 ACD gatherings.

History expiry

Since not all consensus clients have adopted the new deposit snapshot format at this time and there remains a risk that these snapshots could be inconsistently shared across the network, the “drop day” regarding history expiry will be postponed until Pectra is delivered on Mainnet. This is due to Pectra encompassing EIP-6110, an enhancement that eliminates reliance on pre-merge history. May 1st will now serve to test history expiry on Sepolia.

Fusaka

While core developers are mainly centered on launching Pectra, devnet testing is advancing for the two significant attributes planned for inclusion (”SFI”) in Fusaka: PeerDAS and EOF. Additional EIPs have been submitted for consideration (”PFI”), but the core developers are keen to keep the fork as streamlined as possible to expedite the launch of PeerDAS, which means that any EIPs categorized as PFI will encounter challenges to achieve SFI status.

There’s some increasing resistance against the intricacy introduced by EOF and alternative proposals, with steadfast backing on the opposing side with no agreement for removal currently.

Chronology

Fusaka timetables are indicated as follows:

March 24: cutoff for EIPs to be Proposed for Inclusion (”PFI”)
March 31: cutoff for client teams to communicate their preferences regarding scope, which includes which EIPs they believe should be Declined for Inclusion (”DFI”)
April 10: Fusaka scope finalized

In February’s ACD recap, also known as “Checkpoint”, the dominant attitude regarding prioritization was inclined towards scaling and pace enhancement. Certainly, the urgent need for PeerDAS, increases in gas limits, and more blob storage has taken a significant position among immediate priorities. However, the strain of consecutive testnets requiring recovery has slightly redirected the focus towards a wish to deliver ‘promptly but with utmost caution’.

Preliminary discussions have also circled around strengthening protocol security, enhancing testing protocols, and establishing standardized settings. The upcoming months are expected to unveil new procedures aimed at diminishing the likelihood of issues during a mainnet upgrade.