Ethereum Weekly Digest, July 27, 2025
Eth News and Updates
Eth R&D protocol call (All Core Devs)
Consensus layer focused protocol call (ACDC #161):
Fusaka DevNet: DevNet‑2 is closed and DevNet‑3 has just launched, with the Fulu fork expected within hours. Engineers are monitoring new bandwidth limits, MEV‑Boost v2 and client sync issues before freezing code in late August.
Glamsterdam headliner selection: Agreed to issue a CFI for three contenders—ePBS (EIP-7732), six‑second slots (EIP‑7782) and FOCIL (EIP‑7805)—with a final SFI decision due on the next ACDC call.
Six‑second slot study: Preliminary data show 95 % of block + attestation traffic arrives within 4‑5 s on high‑bandwidth nodes, suggesting six‑second slots are plausible. More long‑tail measurements, blob timing and issuance adjustments are required before the change can be scheduled.
ePBS free‑option problem: Flashbots quantified the builder’s “no‑blob, no‑pay” option—worth millions per day if the current ~8 s grace window remains. Potential mitigations include steeper slashing, shorter blob deadlines or fully decoupling blob and transaction fee markets.
Layer1
Background: With practical quantum computers on the horizon, Ethereum has launched a “Lean” refactor to drop classical crypto and move toward formally‑verifiable, quantum‑safe components. Within this effort, the Beam research track is standardizing a hash‑based post‑quantum signature stack (XMSS leaves → recursive SNARK aggregation → six‑opcode ZK‑VM). The PQ Interop program exists to prove those primitives can run in a real, multi‑client, peer‑to‑peer setting before they are ever proposed for main‑net hard‑fork. The inaugural DevNet‑1 will therefore start small—un‑aggregated XMSS signatures, 4 s slots, SSZ + Poseidon hashing, ≈1 k validators—then scale by 10× each iteration.
EIPs/ERCs
EIPs (Ethereum improvement proposals):
EIP-7990: Opcode for arbitrary bytecode execution
Client Release
Consensus Layer
Grandine v1.1.2: HTTP API: requested post-Electra attestation data always has its index set to zero
Execution Layer
Nethermind v1.32.3: a mandatory upgrade for all node operators of the Energy Web Chain
Reth v1.6.0: includes several RPC fixes and various stability fixes
Erigon v3.0.15: flag `experimental.commitment-history` was renamed to `--prune.experimental.include-commitment-history`
Ecosystem
EF Allocation Update Q2 2025: 83 projects, $8,374,108.64 total rewards
The Ethereum Torch lit: The Torch NFT will be symbolically passed from wallet to wallet to honor people and values
Ethereum Comments Protocol:A new, programmable social content primitive on Ethereum to put your thoughts onchain.
Research & Paper
Multi-Relay Inclusion Lists (mrIL) let Ethereum relays agree on a shared transaction list by ranking transactions by peer frequency. This improves censorship resistance, lowers builder costs, and encourages cooperation without needing consensus. It also enables sharing proposer commitments to reduce slashing risk and improve returns, boosting Ethereum’s neutrality and efficiency.
Extensive testing of Ethereum CL and EL clients showed that most can handle ≥72 blobs per block under ideal conditions, with Grandine sustaining 84 blobs for 16+ hours. Lighthouse, Teku, and Nimbus lagged, likely due to fewer validators per node. Reth required manual peering to reach full performance. Bandwidth-limited tests revealed that short gossip bursts, not average usage, are the main bottleneck—capping throughput at ~60 blobs under a 30 Mbps limit. Sustained tests also showed growing memory usage.
Stats
Fees:
Gas: 0.2 to 8.0 gwei, 1.0 gwei average; zero net issuance at 17.1 gwei
17.5k ETH net issuance this week
ETHUSD: $3,530 – $3,845, currently $3,737, all time high $4,878
ETH/BTC: currently 0.032 (Flippening at ~0.165)
L2 Total Value Locked: 10.87 M ETH, -4.1% in 7 days
ETH ETF: 495.47K ETH net inflow (July 21 – July 25)



