Reporting window: September 28, 2026, 00:00:00 to October 5, 2026, 00:00:00 WITA (Asia/Makassar), end exclusive.
Hegotá gained a conditional commitment to shorter slots while losing an issuance-policy proposal. On October 1, AllCoreDevs Consensus advanced Quick Slots to CFI, reserving judgment on the proposed ten-second timing. Meanwhile, the Tapered Issuance Burn authors withdrew from Hegotá consideration and proposed a separate consensus-building process. The editorial significance is the distinction between questions that implementation can resolve and questions that need broader agreement. Ahead of Sepolia, the immediate engineering evidence came from a merged fork-choice correction and new parallel-execution tests.
Quick Slots advances with an explicit evidence requirement
EIP-8198: Quick Slots moved beyond the previous week's specification work. Barnabé Monnot identified ten seconds as the target and said the eight-second EIP text needed updating. He also reported ongoing Prysm prototyping and outreach to contracts that depend on slot timing.
The objections concerned delivery as well as timing. Fredrik warned that ecosystem uncertainty could delay Hegotá and divert work from post-quantum security. Alex Stokes closed with CFI, additional validation of twelve-to-ten seconds, and an opportunity to reconsider before SFI. Watch measured timing budgets and contract compatibility. CFI here supports further work without settling the activation parameter.
Issuance leaves scoping while the deposit guard advances
Jérôme de Tychey's October 1 post withdrew EIP-8363: Tapered Issuance Burn from Hegotá consideration. He attributed the decision to objections that fork scoping was the wrong venue for an issuance-policy change. The authors retained their motivation and proposed forums and workshops beginning at Devcon. Read the withdrawal as a change of process, not evidence that Ethereum has settled issuance policy. The next useful signal is an agreed problem statement and treatment of objections, rather than another scope vote.
A narrower change advanced. NC's revised EIP-8365: Disallow new 0x00 validators keeps the deposit guard and removes forced exits of existing validators. The September 30 merge also specifies that pending pre-fork deposits do not receive an exemption. After NC presented the reduced scope, ACDC moved the proposal to CFI. Review deposit-boundary behavior separately from any future retirement proposal.
Sepolia preparation produces narrower correctness evidence
At ACDC, Barnabas Busa reported preparations for a short-lived Sepolia shadow fork. Stokes reaffirmed the October 6 upgrade and said Hoodi had no agreed date. The preparation report alone does not establish that the shadow fork completed successfully.
On September 28, consensus-specs merged mkalinin's correction to Gloas empty-slot support discounting. The previous logic combined support across PENDING, FULL, and EMPTY parent variants instead of using the direct parent's support. The author credited Francesco's fast-confirmation formal-verification work with finding the bug. This establishes a specification correction with regression coverage; deployment readiness still requires client adoption.
On October 2, execution-specs merged tests for divergences between parallel and sequential BAL execution. The suite includes an invalid transaction followed by a transaction that could incorrectly make the earlier transaction valid. PR author fselmo reported matching verdicts across seven clients, including Nimbus with its parallel path forced. The report also warns that sequential fallback can pass these fixtures by construction. Watch authoritative parallel-path results; aggregate passes alone cannot establish parallel-execution correctness.
Anti-correlation penalties return with a different mechanism
Oisín Kyne's revision of EIP-7716: Anti-Correlation Attestation Penalties returned the proposal from Stagnant to Draft on September 29. The revision replaces the earlier counter with an excess-over-moving-average mechanism. The penalty multiplier applies to timely-target penalties when a validator misses both source and target; source penalties remain unchanged.
The companion consensus-specs change merged on September 30. Kyne's PR describes the proposal as PFI and identifies CFI as a subsequent objective. The useful next evidence is independent scrutiny of calibration and client behavior. A feature-specification merge does not establish fork inclusion or network adoption.
Sources and further reading
- AllCoreDevs - Consensus #188 — corrected transcript; 00:07:42–00:10:02, 00:23:28, 00:27:53–00:29:28, 00:37:35–00:39:32, 01:02:49–01:03:05 — fetched 2026-10-05.
- EIP-8198: Quick Slots — fetched 2026-10-05.
- EIP-8363: Tapered Issuance Burn — post 271 — fetched 2026-10-05.
- EIP-8365: Disallow new 0x00 validators — fetched 2026-10-05.
- Update EIP-8365: reduce scope to the deposit guard — fetched 2026-10-05.
- Fix empty slot support discount for Gloas — fetched 2026-10-05.
- feat(test-specs, tests): tests for BAL parallel execution — fetched 2026-10-05.
- EIP-7716: Anti-Correlation Attestation Penalties — fetched 2026-10-05.
- Update EIP-7716: Move to Draft — fetched 2026-10-05.
- Add anti-correlation attestation penalties (EIP-7716) — fetched 2026-10-05.