Ethereum core developers choose stability over speed

This is a segment from the 0xResearch newsletter. To read full editions, subscribe.

The move to replace Holesky with Hoodi as Pectra’s main testing grounds was one of four options debated on Thursday’s All-Core Developers call. Some protocol teams specifically related to staking infrastructure lobbied for a quicker, but riskier option, and were effectively overruled by core devs.

With Holesky hobbled by a validator exit queue of over a million — which could take more than a year to clear — developers had to find an alternative test environment for Pectra, particularly for exit-related features. Four options were considered:

• Option A: Launch a new testnet (Hoodi), replacing Holesky.
• Option B: Use DevNet 6 as a temporary testnet.
• Option C: Spin up a short-term shadow fork of Holesky.
• Option D: Implement a hard fork that clears Holesky’s exit queue.

For protocols like Lido and EigenLayer, Option D — despite its hacky flavor — was the clear preference. Since staking infrastructure and integrations were already built around Holesky, this option would have allowed Lido to skip redeployment of tooling, oracles and other dependencies, and focus only on Pectra-specific testing.

Lido’s Ivan Metrikin estimated that Option A, migrating to a new testnet, will require two months — several weeks for contract deployment and tooling, followed by additional time for testing.

Matt Nelson at Eigen Labs echoed this concern, noting that every week spent redeploying infrastructure delays their ability to ensure compatibility with Ethereum’s next hard fork. Given that Lido and EigenLayer collectively manage billions in staked ETH, their green light for Pectra’s mainnet rollout is critical.

The topic of ether staking is one likely to come up at Blockworks’ DAS session, Ethereum’s Appeal for Institutional Builders on March 20.

Core Devs prioritize stability (and Fusaka)

Despite these concerns, Ethereum client teams opposed Option D. Their reasoning was twofold:

Hard-forking Holesky would require “hacks” that introduce risk. Making one-off modifications to the execution queue could inadvertently create code paths that affect mainnet stability.

Any delays from debugging a Holesky fix would eat into Fusaka development time. Core teams argued that an extra two weeks spent on a Holesky patch could spiral into months of cumulative delays.

Geth’s Marius Van Der Wijden even suggested that client teams could “chill for a bit” while waiting for LST teams to complete their testing on Hoodi — a comment that drew pushback from those in favor of a faster Fusaka path to mainnet.

As a consequence, the Pectra mainnet fork is now officially delayed to mid-May. Hoodi is set to go live Monday, March 17, with Pectra activating there on March 26. Ethereum’s rule of thumb is to wait at least 30 days before choosing a mainnet block, but liquid staking teams like Lido and EigenLayer now face an extended migration timeline.

Meanwhile, client teams will use the extra time to refine their code, laying the groundwork for Fusaka.

The deadline for submitting EIPs for Fusaka is now March 24, giving teams a short window to finalize proposals. By March 31, core teams are expected to share their feedback, leading to a final scope decision on either April 3 or April 10.

To streamline future ACD calls, Ethereum Foundation coordinator Tim Beiko suggested moving more EIP proposal presentations to an asynchronous process, with only those requiring deeper technical discussions being reviewed live. Plus there’s a new EIP status: declined for inclusion (DFI), reserved for EIPs that teams feel should be deferred to a future fork, but not rejected outright.

Source