Ethereum targets Oct. 6 for Glamsterdam on Sepolia

ACDC #186 meeting notes and subsequent reporting from Ethereum protocol researcher Christine D. Kim show that the date remains conditional. Developers had not completed a stable Glamsterdam activation on a private development network when they selected the Sepolia schedule.

The testing plan has since moved forward by another iteration. Kim said on Sept. 11 that attention had turned to Glamsterdam-Devnet-11, which is expected to launch on Monday, Sept. 14. Earlier plans had identified Devnet-10 as the next major test.

No activation dates have been confirmed for the Hoodi testnet or Ethereum mainnet. Developers have discussed a possible December mainnet release, but testing results will determine whether that schedule remains practical.

During the Sept. 3 All Core Developers Consensus meeting, participants agreed on Sepolia epoch 351232 for the proposed activation. Kim reported that the corresponding time would be Oct. 6 at 13:53 UTC. The meeting was held before developers had demonstrated stable performance across the private test networks used for Glamsterdam.

Selecting the epoch gives client teams, infrastructure operators and application developers a common planning target. It does not make the activation final. Developers can postpone the fork if the next testing phase uncovers a major fault or if client teams cannot prepare reliable releases.

The caveat remains relevant after Devnet-9 experienced finality problems. According to the meeting material, the network included approximately 1,000 validator nodes, making it the largest Glamsterdam devnet by validator count at that stage.

Finality requires enough validators to agree on the state of the chain. When a test network fails to finalize, developers must determine whether the cause involves client software, validator participation, network configuration or an interaction between separate protocol changes.

Devnet-11 will test fixes before Sepolia

The original plan called for Devnet-10 after faults appeared during previous trials. Kim’s latest update now identifies Devnet-11 as the next test that developers are watching, indicating that the private testing sequence advanced beyond the earlier plan.

A stable Devnet-11 would give Ethereum client teams another environment for testing the combined Glamsterdam specifications. Layer-2 teams, staking providers and other infrastructure operators need working client implementations before they can safely test their systems against the proposed fork.

Client diversity makes the process more complex. Ethereum operates through several independently developed execution and consensus clients, and the upgrade must work across different client combinations. A fault confined to one implementation can still interrupt a test network when affected validators hold enough weight.

The ACDC #186 agenda records requests from Lido and Optimism for at least one stable day before a fork. The agenda listed client fixes and successful interoperability as matters requiring confirmation before Sepolia.

A failed or unstable Devnet-11 would not automatically cancel the Oct. 6 activation. Developers would need to assess the cause and the time required for repairs. A serious issue could lead them to reconsider the date during an All Core Developers meeting.

Consensus and EIP-8037 bugs extended testing

Earlier Glamsterdam trials exposed faults on both sides of Ethereum’s architecture. Ethereum Foundation developer operations engineer Stefan Starflinger reported that Devnet-8 revealed a consensus-layer problem involving blocks that repeated a parent hash.

“You could get the whole network to stop,” Starflinger said while describing the test scenario.

The issue affected the system responsible for block agreement. Devnet-9 then suffered non-finality, prompting engineers to investigate more edge cases across a larger validator set.

On the execution side, Ethereum Foundation researcher Maria Silva reported an implementation problem involving EIP-8037. The proposal changes how Ethereum charges gas for creating new state, including new accounts, contracts and storage entries.

EIP-8037 separates state-creation costs from normal execution costs through a multidimensional gas model. Its published specification says the design seeks to control state growth as Ethereum raises its block gas limit. The proposal remains under peer review.

The discovered issue required execution clients to revise their implementations and led to specification work. EIP-8037 has been tested alongside the upgrade’s other protocol changes.

Testing serves a different purpose from approving each proposal individually. Developers must confirm that all selected changes operate together across multiple clients, validator configurations and transaction patterns.

Hoodi and mainnet dates depend on test results

Developers have declined to schedule Glamsterdam on Hoodi while Sepolia remains conditional. Hoodi is expected to serve as the second public testnet stage, giving staking operators and protocol teams another environment that more closely represents mainnet conditions.

Teku developer Enrico del Fante supported waiting before fixing the Hoodi date. During ACDC #186, he cited the recent Devnet-9 problems and favored allowing more testing time after the Sepolia decision.

A December mainnet activation remains a possible target, not a confirmed launch window. Scheduling Sepolia for early October preserves enough calendar time for another public testnet phase and client release preparation, provided testing progresses without lengthy delays.

Developers have not published a mainnet epoch, activation timestamp or final client release schedule. No formal deadline has been announced for deciding whether Oct. 6 remains suitable for Sepolia.

The immediate procedural event is the planned Devnet-11 launch on Sept. 14. Client teams will examine finality, cross-client behavior and the fixes introduced after earlier tests before deciding whether Sepolia can proceed under the current schedule.