Erigon 3.7 turns on parallel state roots and keeps receipts by default
The release schedules Glamsterdam on Sepolia for 6 October, requires x86-64-v2 CPUs, drops Polygon support flags and makes null JSON-RPC arguments an error, matching geth.
Ethereum
Geth 1.17.6 schedules Amsterdam on Sepolia for 6 October
Sepolia operators must upgrade before the fork. The release also removes Holesky, drops Go 1.24, and adds parallel block execution driven by block-level access lists.
Geth 1.17.5 stops random churn from dropping its most useful mempool peers
CHAMP shields a small set of peers per quality dimension - two of them anchored to transactions that actually reach the chain - while the rest keep churning at the same rate.
Stream compression cuts an Ethereum node's inbound gossip by a quarter
A research post measured Teku traffic on mainnet: keeping Snappy's history across messages removes repeated topic strings and attestation data, with the same decoder bound.
Formal verification does not remove trust from a client, it shrinks it
An ethresear.ch post reframes the question as the Trusted Computing Base: everything trusted rather than proven, which is what humans still have to check.
EIP-8298 drops the runtime-only restriction on SETCODEFROM and lets initcode use it
The instruction sets the calling account's code hash to that of a source account with existing deployed code; the update removes the wording that confined it to runtime.
Core devs move ten EIPs onto the inclusion lists for the next fork
EIP-8081 is updated with the decisions from ACDE246, adding proposals that remove refunds and bloom filters and reserve an opcode.
Ethereum research asks which hashes people have actually attacked
With Flock handling binary circuits, circuit-friendly hashes lose their reason to exist - and the comparison turns to cryptanalytic maturity rather than speed.
EIP-8383 would cut the beacon block retention window by three quarters
The draft lowers MIN_EPOCHS_FOR_BLOCK_REQUESTS from 33,024 epochs to 8,192, arguing the existing figure comes from a worst-case weak-subjectivity calculation nobody actually uses.
Simulations put one-dimensional Ethereum fees against EIP-7999
An Ethereum Research post benchmarks post-Glamsterdam single-base-fee configurations against multi-dimensional pricing, reporting execution gas and annual state growth for each.