EIPsInsight

Loading Experience

Preparing your insights...

Loading0%
ACDT2026-02-16_070

AllCoreDevs — Testing #070

2026-02-16 5 decisions 356 transcript lines

Transcript

Ask about this call

Answered from this call's summary and transcript only.

AI-generated — verify against the recording.

Call summary

Targets
  • March 4 - Glamsterdam bal-devnet-3 with EIP-8037, 7954, ETH/70, optional ETH/71 — 00:34:16
Decisions
  • BAL DOS protection constraint enforced outside EVM at block receipt time — 00:18:10
  • BAL DOS protection formula targets DevNet-3, not DevNet-2 — 00:18:55
  • DevNet-3 launch moved from February 25 to March 4 — 00:31:36
  • ETH/71 scoped as optional for DevNet-3; clients can participate without it — 00:36:49
  • Perf-devnet-2 requires CL chain relaunch due to finalized state loss — 00:45:03
Highlights
  • Bal Updates:
    • ·BAL DOS protection: hard cap on items added via PR 11234; formula may simplify with EIP-7981 — 00:09:31
    • ·JSON-RPC methods simplified to use block number/tag/hash; return JSON instead of RLP — 00:16:41
    • ·ETH/71 adds BAL query method; scoped as optional for DevNet-3 — 00:22:40
  • State Bloat:
    • ·Perf-devnet-2 dead: all EL clients died, finalized chain without available blocks — 00:44:14
  • Epbs Updates:
    • ·EPBS devnet-0: All CL clients need 1-2 weeks; Feb 18 target unrealistic — 00:38:01
    • ·Lighthouse fixing gossip/import bug; ready for local interop by week end — 00:39:06
    • ·Consensor reference implementation passed all tests; epoch transition working — 00:43:44
  • Testing Progress:
    • ·NFT-devnet-10 (ETH/70) launched; Geth/Nethermind confirm sync impossible without ETH/70 for large receipts — 00:20:52
    • ·Besu syncing NFT-devnet-10 externally; expects to join network mid-week — 00:23:00
    • ·EIP-8037 testing concerns: many existing tests will break, need refill assessment — 00:30:17
  • Fork Status And Schedule:
    • ·Fusaka blob-devnet-0 running with 20 nodes; Lighthouse/Prism ready, Teku/Nimbus in progress — 00:06:01
    • ·Glamsterdam bal-devnet-2 had bugs over weekend; reverted to stable commit, 100% participation restored — 00:09:05
    • ·DevNet-3 scope: EIP-8037, EIP-7954, ETH/70, ETH/71 (optional), simulate_v1 RPC — 00:32:00
    • ·DevNet-3 launch pushed from Feb 25 to March 4 due to testing needs — 00:34:16
Action Items
  • All EL teams — Review/approve BAL DOS protection PR 11234 — 00:15:03
  • All EL teams — Review execution-apis PRs 726 and 727 for DevNet-3 — 00:16:41
  • Barnabas — Publish DevNet-3 specification document — 00:33:23

Meeting chat72

  • jochem-brouwer

    gm

  • felipe

    gm

  • jochem-brouwer

    Heeft gereageerd op "gm" met 👋

  • Mario Vega

    gm

  • jochem-brouwer

    Heeft gereageerd op "gm" met 🔥

  • Barnabas

    https://github.com/ethereum/pm/issues/1920

  • Barnabas

    https://dora.blob-devnet-0.ethpandaops.io

  • Marius van der Wijden

    Whats the tldr on blob-devnet-0?

  • Marius van der Wijden

    And the purpose is to test full peerdas?

  • Marius van der Wijden

    Seeign some orphans and 90% participation

  • Barnabas

    https://github.com/ethereum/EIPs/pull/11234

  • Marius van der Wijden

    Whats the max gas limit this could support?

  • Marius van der Wijden

    Ah okay

  • Marius van der Wijden

    cool

  • Marius van der Wijden

    (github doesn't load for me atm)

  • Dustin

    https://www.githubstatus.com/ says "All Systems Operational"

  • jochem-brouwer

    It should not be possible to construct a block which violates this constraint (because then you have invalid EVM execution) given that the formula is correct. So if we make this part of consensus then there should thus be tests of specifically sized BALs with items to test that invalidaton

  • Barnabas

    https://github.com/ethereum/execution-apis/pull/726

  • Barnabas

    https://dora.nft-devnet-10.ethpandaops.io

  • Barnabas

    try to sync with kurtosis — network: nft-devnet-10

  • jochem-brouwer

    Also need eth/71

  • jochem-brouwer

    Yes =)

  • Karim T. (matkt)

    Protection against DOS attack will depend on result regarding prefetch no ?

  • Mario Vega

    Which Engine API methods?

  • Jen

    What’s the timeline for devnet-3?

  • Jen

    Replying to "Which Engine API met..." https://github.com/ethereum/execution-apis/pull/726 I believe

  • Toni Wahrstätter

    Replying to "Which Engine API met..." one sec, I'll pull up the PR

  • Mario Vega

    Reacted to "one sec, I'll pull u..." with 🙏🏼

  • Toni Wahrstätter

    Replying to "Which Engine API met..." https://github.com/ethereum/execution-apis/pull/727

  • Mario Vega

    Reacted to "https://github.com/e..." with 👍🏼

  • Toni Wahrstätter

    8037 was scoped

  • Toni Wahrstätter

    and another one iirc

  • Dustin

    there's no https://notes.ethereum.org/@ethpandaops/bal-devnet-3 yet?

  • Maria Silva

    The contract size

  • Barnabas

    Replying to "there's no https://n..." will make it right after the call

  • g1

    devnet-3 - EIP-8037: State Creation Gas Cost Increase - EIP-7954: Increase Maximum Contract Size

  • spencer-tb

    Reacted to "devnet-3 - EIP-8037:..." with ❤️

  • felipe

    Reacted to "devnet-3 - EIP-803..." with ❤️

  • Barnabas

    Reacted to "devnet-3 - EIP-8037:..." with ❤️

  • Mario Vega

    Reacted to "devnet-3 - EIP-8037:..." with 👍🏼

  • Mario Vega

    My main concern is 8037 on the testing side

  • Ben Adams

    Was this one https://github.com/ethereum/EIPs/pull/11306

  • Karim T. (matkt)

    Is it devnet3 and devnet2 at the same time ? We wanted to do benchmark on devnet 2

  • Ansgar Dietrichs

    there was also `eth_simulateV1` RPC support

  • Marius van der Wijden

    Can we either 1) push devnet-3 to first week of march 2) remove new eips (8037 and 7954) from devnet-3

  • Marius van der Wijden

    Replying to "Can we either 1) ..." I would prefer to move the deadline

  • the ACDE summary (looks right)

    eth/70 part of devnet-3 spec; clients without support can still participate

  • Dustin

    Replying to "Is it devnet3 and ..." would this mean supporting devnet-2 specs at the same time as devnet-3? these inevitably involve pointless back/forward porting of the same things to keep known-obsolete spec-based devnets running

  • Ansgar Dietrichs

    Reacted to "I would prefer to mo..." with 👍

  • spencer-tb

    Reacted to "I would prefer to mo..." with 👍

  • Jared Wasinger

    Reacted to "I would prefer to mo..." with 👍

  • Mario Vega

    Reacted to "I would prefer to mo..." with 👍

  • spencer-tb

    Replying to "Can we either 1) pu..." I’d be happy with this

  • g1(shadow)

    Reacted to "I would prefer to ..." with 👍

  • felipe

    Reacted to "I would prefer to ..." with 👍

  • g1(shadow)

    Reacted to "from the ACDE summ..." with 👍

  • Karim T. (matkt)

    Don’t really have preference just wanted to understand if we want to do benchmark on devnet 3 now . Before we said devnet 2

  • felix (eest)

    Reacted to "I would prefer to ..." with 👍

  • Marius van der Wijden

    Which client does not have eth_simulateV1 at this point?

  • Marius van der Wijden

    We have it in the codebase at least

  • Jen

    Reacted to "from the ACDE summar..." with 👍

  • Karim T. (matkt)

    We have it in besu but not passing test, so maybe better to postpone also

  • Marius van der Wijden

    Reacted to "We have it in besu..." with 👍

  • Marius van der Wijden

    eth/71 is also optional (only used for sync improvements, which no one has iimpl yet)

  • Marius van der Wijden

    Reacted to "will make it right..." with 🔥

  • Barnabas

    epbs-devnet-0 branch name?

  • Barnabas

    we have ethpandaops/geth:epbs-devnet-0 for EL image to test your CL locally.

  • Barnabas

    Replying to " epbs-devnet-0 bran..." @Gaia ?

  • Gaia

    Replying to " epbs-devnet-0 bran…" gloas-devnet-0

  • Barnabas

    Replying to " epbs-devnet-0 bran..." could you please rename to epbs-devnet-0?

  • Gaia

    Replying to " epbs-devnet-0 bran…" yes np

  • Gaia

    Replying to " epbs-devnet-0 bran…" lol just realized im using my wife’s zoom profile

Key decisions

  • BAL DOS protection constraint enforced outside EVM at block receipt time

  • BAL DOS protection formula targets DevNet-3, not DevNet-2

  • DevNet-3 launch moved from February 25 to March 4

    SFI
  • ETH/71 scoped as optional for DevNet-3; clients can participate without it

  • Perf-devnet-2 requires CL chain relaunch due to finalized state loss