AllCoreDevs - Consensus #185
Transcript
- Dima Gusakov | Lido
Good morning devs
- Kalo | Obol
gm
- Dima Gusakov | Lido
Reacted to "gm" with ❤️
- cayman
gm
- Chris Berry
Reacted to "gm" with ❤️
- jochem-brouwer
gm
- lightclient
gm
- Lewej | Optimum
gm
- Ben Edgington
gm
- Dima Gusakov | Lido
My concern here is that Lido was not able to properly test how our Oracles and tooling passes hardfork because of the loss of the finality
- Barnabas
Replying to "My concern here is t..." we might have unfinality on mainnet too. Why would this be an issue? We had no consensus issues
- Dima Gusakov | Lido
Replying to "My concern here is t..." We might. For sure. Tho we expect that it will go well on mainnet…
- Chris Berry
Will the new devnet be open everyone? It would be good to see a successful fork before we head to testnets
- Toni Wahrstätter
Will the fork transition bug be fixed? https://discord.com/channels/595666850260713488/874767108809031740/1534880634470924359
- Kalo | Obol
Reacted to "Will the new devnet ..." with 👍
- potuz
Is Terence here?
- Parithosh Jayanthi
nop
- Parithosh Jayanthi
Might need to be you @potuz
- potuz
Pinging him, but it’s essentially the same thing as LH
- Dima Gusakov | Lido
Reacted to "Will the new devnet be open everyone? It would be good to see a successful fork before we head to testnets" with 👍
- Barnabas
great edge case test guys
- Eitan Seri-Levi
Replying to "great edge case test..." Totally on purpose
- Barnabas
Reacted to "Totally on purpose" with 🥦
- Parithosh Jayanthi
Reacted to "Totally on purpose" with 🥦
- pk910
for mainnet I'd expect deposits coming in way earlier than the latest node restart? depositing right before the fork is more a devnet scenario, no? so clients shouldn't rely on all deposits coming on during the process runtime
- Barnabas
we could already have deposits now
- Nico Flaig
no, potuz that is not correct
- Dima Gusakov | Lido
38 days now
- jochem-brouwer
Why does the zoom 😂 emote (the big version, this small one looks fine) looks like someone crying if I emote it 😭
- Barnabas
Feels like non finality network would need to have stable cache implementations that will work.
- Barnabas
Ideally also trunk branches imo
- pk910
yes, we'll defnitly repeat the mass deposits there
- Barnabas
Reacted to "yes, we'll defnitly repeat the mass deposits there" with 👍
- Mario Vega
Reacted to "yes, we'll defnitly repeat the mass deposits there" with 👍
- Parithosh Jayanthi
Reacted to "yes, we'll defnitly repeat the mass deposits there" with 👍
- potuz
Reacted to "yes, we'll defnitly repeat the mass deposits there" with 👍
- Nico Flaig
Replying to "yes, we'll defnitly..." https://gist.github.com/nflaig/c02c00714d07df076bed23200805d2d6 has my malicious deposits documented, so can repeat there, maybe even scale it up a bit
- Nico Flaig
Reacted to "yes, we'll defnitly..." with 👍
- cayman
Lodestar ready for next week
- Chris Berry
Could we get a week before gloas fork on devnet? Just to give chance to get validators online etc.
- Kalo | Obol
Reacted to "Could we get a week ..." with ➕
- Dustin
Reacted to "yes, we'll defnit..." with 👍
- potuz
Is it really needed to be on trunk? there’s some paths that I’d like to keep longer on the devnet branches
- Barnabas
Replying to "Could we get a week ..." thats what devnet 8 was for
- Chris Berry
Replying to "Could we get a week ..." And we would like to do it again but with finality
- Barnabas
whats the absolute min time you would need?
- Chris Berry
Replying to "Could we get a week ..." We can do 1-2 days if we know. It might be better if we just get genesis validators, but I can follow up offline.
- Barnabas
Reacted to "We can do 1-2 days if we know. It might be better if we just get genesis validators, but I can follow up offline." with 👍
- Parithosh Jayanthi
Reacted to "We can do 1-2 days if we know. It might be better if we just get genesis validators, but I can follow up offline." with 👍
- Parithosh Jayanthi
- Nico Flaig
Replying to "https://github.com/e..." seems like it's missing stuff
- potuz
James has implemented this on Prysm has some PRs for it
- james
Replying to "https://github.com/e..." I pinged uzi on it
- james
Got a pr open but not merged yet
- Nico Flaig
lodestar has all remote signer stuff implemented besides the signed auth request for builder api
- Parithosh Jayanthi
Sepolia: epoch 351232, slot 11239424, unix (1790606688), Mon 28 Sep 2026, 14:44:48
- Barnabas
Reacted to "Sepolia: epoch 351232, slot 11239424, unix (1790606688), Mon 28 Sep 2026, 14:44:48" with 🚢
- Iván | ethrex
Reacted to "Sepolia: epoch 35123..." with 🚢
- Nico Flaig
Reacted to "Sepolia: epoch 35123..." with 🚢
- Mario Vega
Reacted to "Sepolia: epoch 351232, slot 11239424, unix (1790606688), Mon 28 Sep 2026, 14:44:48" with 🚢
- Stefan Starflinger
Reacted to "Sepolia: epoch 35123..." with 🚢
- Eitan Seri-Levi
Reacted to "Sepolia: epoch 35123..." with 🚢
- Parithosh Jayanthi
Hoodi: epoch 132352, slot 4235264, unix (1793036568), Mon 26 Oct 2026, 17:42:48
- kingy_sigp
that date is fine with us provided we get a stable devnet ofc
- Barnabas
also no objection = agreement
- Barnabas
Reacted to "Hoodi: epoch 132352, slot 4235264, unix (1793036568), Mon 26 Oct 2026, 17:42:48" with 🚢
- Dima Gusakov | Lido
I would rather have both dates a bit later. Like a week or so to have more room for the pending devnets testing
- Iván | ethrex
Reacted to "also no objection = ..." with 🤝
- kingy_sigp
Reacted to "also no objection = agreement" with 🤝
- FLCL
EIP-7610 is excluded, will it be postponed or fully cancelled
- Pablo Deymo | Lambda
Reacted to "also no objection = agreement" with 🤝
- Iván | ethrex
Reacted to "Hoodi: epoch 132352,..." with 🚢
- Nico Flaig
Reacted to "Hoodi: epoch 132352,..." with 🚢
- jochem-brouwer
- Miguel Tenorio
Reacted to "that date is fine with us provided we get a stable devnet ofc" with ➕
- Miguel Tenorio
Reacted to "I would rather have both dates a bit later. Like a week or so to have more room for the pending devnets testing" with ➕
- Chris Berry
Reacted to "I would rather have both dates a bit later. Like a week or so to have more room for the pending devnets testing" with ➕
- jochem-brouwer
- jochem-brouwer
- Pooja Ranjan
Reacted to "https://github.com/ethereum/EIPs/pull/11855" with 🙏
- cayman
https://gist.github.com/lodekeeper/ab3164dc62c81f2d172ae7ce0f0d2c55
- Barnabas
Reacted to "https://gist.github.com/lodekeeper/ab3164dc62c81f2d172ae7ce0f0d2c55 " with 🚢
- Nico Flaig
Reacted to "https://gist.github...." with 🚢
- cayman
Reacted to "https://gist.github...." with 🚢
- Barnabas
Replying to "https://gist.github...." small eip big win 👍
- Parithosh Jayanthi
https://github.com/tersec/EIPs/commit/6fc997491b1e3d1d1cb241bdea31a44c9984d87b
- Kalo | Obol
Reacted to "https://gist.github...." with 🚢
- Parithosh Jayanthi
Reacted to "https://gist.github.com/lodekeeper/ab3164dc62c81f2d172ae7ce0f0d2c55 " with 🚢
- Nico Flaig
Reacted to "small eip big win 👍" with ❤️
- jochem-brouwer
Antwoord verzenden naar "EIP-7610 is exclud..." To answer explicitly, cancelled. I will also ask authors to formally withdraw this EIP
- Barnabas
feels like EIP 8379 could be an optional EIP?
- Parithosh Jayanthi
- Kevaundray Wedderburn
- jochem-brouwer
frontrun :D
- Barnabas
I’d prefer if we also added 8192 + any nonfinality period.
- Barnabas
also this was proposed for glamsterdam LOL?
- Parithosh Jayanthi
Replying to "also this was propos..." Nah proposal date is 3d ago
- Dustin
One concern: 5 months means if it's nonfinalized that long it's dead. 2 weeks or etc doesn't
- jochem-brouwer
Can the EIP-8379 top up sync PR be linked here?
- Dustin
it makes it more structurally complicated
- Parithosh Jayanthi
Replying to "Can the EIP-8379 top..." https://github.com/tersec/EIPs/commit/6fc997491b1e3d1d1cb241bdea31a44c9984d87b
- jochem-brouwer
Antwoord verzenden naar "Can the EIP-8379 t..." Is there a PR against EIPs? Because now I am uncertain if this number is correct
- jochem-brouwer
Antwoord verzenden naar "Can the EIP-8379 t..." Ah found it, https://github.com/ethereum/EIPs/pull/12154 Number is correct 😃 Thanks!🙏
- cayman
Reacted to "I’d prefer if we als..." with 👍
- Parithosh Jayanthi
Reacted to "Ah found it, https://github.com/ethereum/EIPs/pull/12154 Number is correct 😃 Thanks!🙏" with ❤️
- Dustin
To be clear, I'm pretty open to this EIP
- Dustin
This is a mild-skepticism point
- Barnabas
currently we dropped pre merge history only no? Lets work on dropping pre fulu first?
- Dustin
It is actually useful for the CL/EL to be synced on this point
- Dustin
I agree with Kev here
- potuz
It doesn’t need to be synced, one need to cover the other
- potuz
The CL side is a bound for the EL
- potuz
Doesn’t need to be equal
- Barnabas
I think its also safe to say if we have 4-5 months non finality its game over
- Luca Donno | L2BEAT
Reacted to "I think its also safe to say if we have 4-5 months non finality its game over" with 🥲
- Parithosh Jayanthi
Reacted to "I think its also safe to say if we have 4-5 months non finality its game over" with 🥲
- jochem-brouwer
Heeft gereageerd op "I think its also s..." met 🥲
- Parithosh Jayanthi
https://github.com/ethereum/pm/issues/2187#issuecomment-5336129489
- Kalo | Obol
Reacted to "I think its also saf..." with 🥲
- jochem-brouwer
Where is this document you are referring to?
- potuz
Nahhh let’s go and fight on Devcon
- Parithosh Jayanthi
- Luca Donno | L2BEAT
Reacted to "Nahhh let’s go and fight on Devcon" with 😂
- Dima Gusakov | Lido
Replying to "https://forkcast.org..." Forkcast is DOPE
- Barnabé + Francesco 🚗
Reacted to "Forkcast is DOPE" with ➕
- Luca Donno | L2BEAT
Reacted to "Forkcast is DOPE" with 💯
- Barnabas
[S-tier] strongly recommend inclusion [A-tier] recommend inclusion if blockers are cleared (e.g., complexity assessment, impact analysis, path to adoption) [B-tier] stretch goal [D-tier] recommend rejection
- wolovim
Reacted to "Forkcast is DOPE" with ❤️
- Barnabé + Francesco 🚗
Reacted to "[S-tier] strongly recommend inclusion [A-tier] recommend inclusion if blockers are cleared (e.g., complexity assessment, impact analysis, path to adoption) [B-tier] stretch goal [D-tier] recommend rejection" with ❤️
- Oisin Kyne
Reacted to "[S-tier] strongly re..." with ❤️
- jochem-brouwer
Antwoord verzenden naar "[S-tier] strongly ..." What is C-tier?
- kingy_sigp
Reacted to "[S-tier] strongly recommend inclusion [A-tier] recommend inclusion if blockers are cleared (e.g., complexity assessment, impact analysis, path to adoption) [B-tier] stretch goal [D-tier] recommend rejection" with 👍
- Dima Gusakov | Lido
Reacted to "[S-tier] strongly recommend inclusion [A-tier] recommend inclusion if blockers are cleared (e.g., complexity assessment, impact analysis, path to adoption) [B-tier] stretch goal [D-tier] recommend rejection" with ❤️
- Dima Gusakov | Lido
Reacted to "[S-tier] strongly recommend inclusion [A-tier] recommend inclusion if blockers are cleared (e.g., complexity assessment, impact analysis, path to adoption) [B-tier] stretch goal [D-tier] recommend rejection" with 👍
- Barnabas
Replying to "[S-tier] strongly re..." we shouldn’t have C tier
- —
Ben Adams: S - C: should definitely go in D: No
- kingy_sigp
Reacted to "S - C: should definitely go in D: No" with 😂
- Parithosh Jayanthi
Reacted to "S - C: should definitely go in D: No" with 😂
- jochem-brouwer
Antwoord verzenden naar "S - C: should defi..." seems there is no rough consensus on tiered ranking lol
- Sam Jernigan
bye
- jochem-brouwer
Heeft gereageerd op "S - C: should defi..." met 😂
- jochem-brouwer
Heeft gereageerd op "[S-tier] strongly ..." met ❤️
Call summary
Targets
- •Sepolia Glamsterdam fork — Sep 28, 2026 (epoch 351232); pending confirmation at next ACDC - 00:27:50
- •Hoodi Glamsterdam fork — Oct 26, 2026 (epoch 132352); contingent on stable DevNets - 00:28:57
- •Hegota EIP tier rankings from all client teams — mid-September 2026 - 00:54:44
- •Hegota rough consensus on EIP scope — before DevCon - 00:55:36
Decisions
- •EIP-7610 cancelled from Glamsterdam; not postponed — superseded by EIP-8253 targeting Hegota - 00:30:21
- •Non-finality DevNet targeted for the week after next; all clients must be on trunk/unstable branches first - 00:23:04
- •PFI for Hegota: EIP-833 (Align Checkpoint with Epoch Boundary Block), EIP-8379 (Engine API Sync Status); PR-12188 (Reduced CL Block Retention Window) deferred to ETH R&D discussion - 00:54:23
Highlights
- Client Updates:
- ·Lodestar: deposit preprocessing optimization reduced fork-transition time from ~20s to ~500ms; already on trunk - 00:16:07
- ·Prysm: builder-deposit cache was removed from branch pre-fork; Bastin's optimization reduces recovery even without cache to one slot - 00:14:58
- Organizational:
- ·EIP-7610 formally cancelled from Glamsterdam (not postponed); superseded by EIP-8253 (PFI Hegota); authors asked to withdraw EIP-7610 - 00:30:21
- ·Hegota EIP tiering: all client teams to publish first-pass S/A/B/D rankings by mid-September; aim to reach rough consensus before DevCon; use Forkcast ranking tool - 00:54:44
- Testing Progress:
- ·Lido needs ≥1 day (preferably 2) on a stable DevNet before Glamsterdam activation to test oracles/tooling; Optimism also requested similar testing window - 00:10:34
- ·Non-finality DevNet (~1000 nodes) cost-constrained; likely 1–1.5 days max; pk910 will repeat mass-deposit stress test there - 00:23:28
- Eip Proposals Hegota:
- ·EIP-833 (Align Checkpoint with Epoch Boundary Block): fixes off-by-one in checkpoint root calculation; improves target-vote accuracy; pairs well with quicker slots - 00:33:39
- ·EIP-8379 (Engine API Sync Status): lets CL become sole sync source; reduces duplicate block gossip; narrows devp2p scope long-term; optionality rejected as impractical for Engine API - 00:36:28
- ·PR-12188 (Reduced CL Block Retention Window): halves retention constant due to EIP-8061 in Glamsterdam; discussion ongoing — potuz questions motivation; Dustin mildly open; thread to continue on ETH R&D - 00:42:28
- Fork Status And Schedule:
- ·DevNet 8 Gloas activation (6h ago) caused non-finality; finalization just recovered; root cause: client caching bug for builder deposits at fork transition (Lighthouse, Prysm, Teku affected) - 00:07:13
- ·Sepolia Glamsterdam fork proposed: epoch 351232, slot 11239424, Sep 28 2026 — no objections raised; decision deferred to next ACDC - 00:27:29
- ·Hoodi Glamsterdam fork proposed: epoch 132352, slot 4235264, Oct 26 2026 — contingent on stable DevNets - 00:28:57
- ·Non-finality DevNet targeting week after next; all clients should be on trunk/unstable branches by next week - 00:20:43
Action Items
- •Parithosh Jayanthi (EIP officer) - Review and approve Glamsterdam Meta EIP PR (move status draft→review) before merging EIP-7610 removal PR - 00:32:01
- •CL client teams - Review remote-signing-api PR #28; implement signed auth request for builder API - 00:26:15
- •All client teams - Publish Hegota EIP tier rankings (S/A/B/D) with rationale by mid-September - 00:54:44
Key decisions
Non-finality DevNet targeted for the week after next; all clients must be on trunk/unstable branches first
Non-finality DevNetTargeting the week after next for launch, contingent on client readiness on trunk branches.Hegota EIP tiering: all client teams to publish first-pass S/A/B/D rankings by mid-September
Client teams to rank Hegota EIPs to reach rough consensus before DevCon using the Forkcast ranking tool.
EIPs discussed
AI Disclaimer: Some content or metadata on EIPsInsight may be AI-inferred or automatically compiled. If you find any discrepancy, please contact us at dev@avarch.org.