“`html
Ethereum’s weekly All Core Developer calls can be quite challenging to follow, thus this “Checkpoint” series is intended for high-level summaries approximately every 4-5 weeks, contingent on developments in core advancement. View the earlier update here.
tl;dr:
The Fusaka enhancement is almost ready for deployment, and we will gain a clearer indication of “when” this week with the initial testnet upgrade set to activate just a few hours following this article. The main attractions for Glamsterdam have been finalized: enshrined Proposer-Builder Separation and Block-level Access Lists, but minor features still have about a week to be suggested for inclusion. Testing teams aim to push the gas limit past 60 million with Fusaka.
Fusaka
Fusaka is very likely to launch by year-end. The anticipated testnet upgrade timelines were revealed, the first of which will occur at 08:48 UTC today. For a comprehensive overview of what this upgrade entails click here.
| Network | Time (UTC) |
|---|---|
| Holešky | 2025-10-01 08:48:00 ✅ |
| Sepolia | 2025-10-14 07:36:00 |
| Hoodi | 2025-10-28 18:53:12 |
| Mainnet | tba (minimum 30 days after Hoodi) |
Testnets
Conducting the fork on public testnets is the final step in the procedure. It indicates that all integrations are seemingly complete, bugs have been resolved on private devnets, and we’re probably prepared to proceed.
The Holešky testnet will be phased out shortly after this fork – it encountered a configuration error during Pectra testing in February of this year that led to a fork and a subsequent prolonged period of non-finality. The validators on the testnet that committed a slashing offense by voting for the incorrect fork were so numerous that it resulted in an exceptionally long exit queue – an insupportable situation for a network intended for testing, where validators must be onboarded and offboarded efficiently.
This round of fork testing has incorporated significant insights from this mistake; The original
“““html
The misconfiguration was resolved through modifications that validate fork parameters. Non-finality had not been examined to such a degree previously, revealing that clients were unfamiliar with recovering from such a condition. In recent months, assessments have included a sequence of scheduled non-finality incidents to evaluate recovery.
Timeline
We’ll find out on ACD(Consensus) this Thursday (10/2) if the initial fork performed successfully, and on ACD(Testing) the following Monday (10/6) if the first Blob Parameter Only fork succeeded.
The audit competition will continue for an additional two weeks, and assuming that the testnets operate smoothly with no or minimal bugs, the upgrade mainnet will be announced after developers have had the opportunity to observe Hoodi’s upgrade for several days. Anticipate a mainnet upgrade date 30 days following the ACD immediately succeeding a successful Hoodi fork.
Glamsterdam
The primary features (also known as “headliners”) for Glamsterdam, the upgrade succeeding Fusaka, were selected from a range of proposals. These features include enshrined Proposer-Builder Separation (ePBS) and Block-level Access Lists (BAL). Additional smaller features are still being suggested for approximately another week and will be selected based on their readiness, necessity, safety, and compatibility with the main features. These proposals can be submitted by opening a GitHub pull request for an EIP against the Glamsterdam Meta EIP.
Timeline
While the emphasis is strongly on finalizing Fusaka, implementers are deeply engaged in testing both ePBS and BAL without the smaller features that have yet to be determined. Those interested in including a specific minor-feature EIP in Glamsterdam should submit their proposal within the next week. Proposers of an EIP should be ready to champion that EIP throughout the upgrade process. The estimate for this fork is still projected for some time in 2026.
Gas limit
We’ve observed a focused effort to scale the L1, which consists of boosting the gas limit. Since February, the gas limit has risen from 30 million to 45 million, and developers are striving to push that even further with Fusaka, aiming for potentially surpassing 60 million.
This limit operates independently of the fork and is
“““html
determined by validator settings, yet default client configurations can assist validator managers in understanding what has been thoroughly assessed and is secure for the network. Forks are instances when we anticipate validators will be upgrading, making it an opportune moment to refresh client defaults without necessitating a communication effort to validators.
source: https://etherscan.io/chart/gaslimit
Against all expectations, it appears that 2025 will witness two ethereum enhancements. As I compose this blog entry, the Pectra upgrade occurred merely four and a half months prior. Given the substantial implementation, testing, and notification required for these enhancements to be launched across 12 client organizations, I’m genuinely impressed and hope the developers and testing teams are obtaining the rest they require.
It cannot be emphasized enough how remarkable the adjustments in testing have been, owing to insights gained from Pectra. I stated it then and I’ll repeat it: Pectra has produced a notably superior quality of core developer (we should likely strive to ensure they’re better rewarded!).
It will be a larger challenge to maintain the speed of Glamsterdam. It feels like ages ago, yet Fusaka’s functionalities emerged from the Pectra upgrade dividing into two distinct forks because of complexity and scope. As a result, PeerDAS had an early advantage in both implementation and decision-making.
Although the procedure was made more efficient for the first time by officially selecting upgrade highlights even before the preceding fork was activated, there are still 23 minor features suggested for inclusion that need decisions, in addition to the winter festivities and the January slowdown. Whether this upgrade will be released in mid- or late-2026 remains uncertain.
Relevant ACD calls:
[ July 31st – September 29th ]
Source link
“`