AllCoreDevs - Consensus #181
Transcript
- Kalo | Obol
gm
- Iván | ethrex
gm
- Barnabas
- Barnabas
https://notes.ethereum.org/@ethpandaops/glamsterdam-devnet-6
- Justin Traglia
Reacted to "https://dora.glamste..." with 🔥
- Marius Van Der Wijden (M)
Whcih clients?
- Iván | ethrex
Reacted to "https://dora.glamste..." with 🔥
- Stefan Starflinger
Reacted to "https://dora.glamste..." with 🔥
- Justin Traglia
Are there slashed validators?
- Nico Flaig
Reacted to "https://dora.glamste..." with 🔥
- Justin Traglia
Awesome.
- Parithosh Jayanthi
- terence
Ya it’s valid
- Barnabas
We also have spamming and fuzzing and blobs as well.
- terence
We’ll fix it for devnet7
- Barnabas
Reacted to "We’ll fix it for devnet7" with 👍
- Barnabas
alpha 12 ftw
- Dima Gusakov | Lido
Let’s include 7688!
- Stefan Starflinger
Reacted to "Let’s include 7688!" with 👍
- Parithosh Jayanthi
- cayman
Reacted to "Let’s include 7688!" with 👍
- Dustin
Reacted to "Let’s include 768..." with 👍
- Nico Flaig
Replying to "https://github.com/e..." default should be to keep it if there is no one pushing for removal
- Etan (Nimbus)
Reacted to "Let’s include 7688!" with 👍
- Nico Flaig
Reacted to "Let’s include 7688!" with 👍
- terence
Yes, keeping it is fine
- Barnabas
Reacted to "default should be to keep it if there is no one pushing for removal" with 👍
- Barnabas
Circuit breakers should be client specific imo
- terence
We haven’t even thought about circuit breaker yet
- Justin Traglia
I vote to leave flexible.
- Enrico Del Fante (tbenr)
We haven’t. But I believe we should some common baseline...
- terence
It’s the same as default today
- Justin Traglia
Reacted to "We haven’t. But I be..." with 👍
- terence
Ya. Same as today
- Eitan Seri-Levi
We haven’t implemented circuit breakers yet
- Enrico Del Fante (tbenr)
I won’t 🙂
- Marius Van Der Wijden (M)
-10 crop$
- Barnabas
maybe we could have a circuitbreaker breakout 😂
- Dustin
it's different than pre-ePBS because consensus chain doesn't break
- Dustin
it's not "fragile" in the same way
- Parithosh Jayanthi
Reacted to "it's not "fragile" in the same way" with 👍
- Barnabas
Reacted to "it's not "fragile" in the same way" with 👍
- Parithosh Jayanthi
- Enrico Del Fante (tbenr)
Reacted to "it's not "fragile" i..." with 👍
- Barnabas
prysm is the only client that has anything afaik
- Parithosh Jayanthi
- Barnabas
lgtm sendit
- Eitan Seri-Levi
yes
- Barnabas
Reacted to "yes" with 🔥
- Parithosh Jayanthi
- james
Have not had the chance to do it yet
- Justin Traglia
Replying to "https://github.com/e..." Ah the agenda has a wrong link to this btw
- Justin Traglia
Replying to "https://github.com/e..." It was linking to call notes from 4 years ago 🙂
- Parithosh Jayanthi
- nixo
Reacted to "It was linking to call notes from 4 years ago 🙂" with 😅
- Barnabas
Reacted to "It was linking to call notes from 4 years ago 🙂" with 😅
- Eitan Seri-Levi
Reacted to "It was linking to ca..." with 😅
- Jun Song
Reacted to "https://github.com/ethereum/beacon-APIs/pull/606" with 👍
- Barnabas
teku seems to have merged it in already
- Barnabas
so thats great
- Barnabas
Replying to "teku seems to have m..." https://github.com/Consensys/teku/pull/10867
- Barnabas
extend with optional
- james
- james
This one
- Barnabas
we need that tested asap i agree
- Eitan Seri-Levi
https://github.com/ethereum/beacon-APIs/pull/621 Another one that I think would be good to get in
- Barnabas
all beacon api prs needs to be merged by ACDT 😂
- pk910
I think we should merge it. the endpoints are already in use
- Jun Song
Reacted to "I think we should merge it. the endpoints are already in use" with 👍
- Barnabas
I’m gonna block dora page for devnet 6 till beacon api prs are not merged
- Jun Song
Reacted to "I’m gonna block dora page for devnet 6 till beacon api prs are not merged" with 😂
- Barnabas
Reacted to "I think we should merge it. the endpoints are already in use" with 👍
- Kalo | Obol
Reacted to "I’m gonna block dora..." with 😂
- Enrico Del Fante (tbenr)
Reacted to "I’m gonna block dora..." with 🥹
- pk910
Reacted to "I’m gonna block dora..." with 😂
- Justin Traglia
Reacted to "I’m gonna block dora..." with 😂
- Satyajit
Reacted to "I’m gonna block d..." with 😂
- Eitan Seri-Levi
lgtm
- Eitan Seri-Levi
Reacted to "I’m gonna block dora..." with 😂
- bfqm
Reacted to "I’m gonna block d..." with 😂
- Parithosh Jayanthi
- Barnabas
SFI decision will have to wait till implementations are done, and its running fine on a devnet.
- Etan (Nimbus)
Reacted to "SFI decision will ha..." with ❤️
- cayman
Lodestar has an implementation
- Barnabas
Reacted to "Lodestar has an implementation" with 🔥
- Etan (Nimbus)
Nimbus has an implementation
- Eitan Seri-Levi
Lighthouse has an implementation
- saulius
Grandine doesn't have implementation.
- Barnabas
Reacted to "Nimbus has an implementation" with 🔥
- Barnabas
Reacted to "Lighthouse has an implementation" with 🔥
- Marius Van Der Wijden (M)
If we put more work on the cl, we can start on frames for the el already
- Eitan Seri-Levi
Reacted to "If we put more work ..." with 😂
- Barnabas
Replying to "If we put more work ..." most of the cls are already on the devnet, and we had to delay launch of the devnet due to el’s not being ready just fyi
- Parithosh Jayanthi
Replying to "If we put more work ..." The thing is this EIP solves the 8192 deposit overflow issue as well
- Etan (Nimbus)
Replying to "If we put more work ..." SSZ frames!
- Justin Traglia
Reacted to "most of the cls are ..." with 🌶️
- Etan (Nimbus)
Reacted to "Lighthouse has an im..." with 🔥
- Etan (Nimbus)
Reacted to "Nimbus has an implem..." with 🔥
- Etan (Nimbus)
Reacted to "Lodestar has an impl..." with 🔥
- Dustin
Replying to "If we put more wor..." Also two ELs have kept doing fake invalids, through the last week or so
- Toni Wahrstätter
Replying to "If we put more work ..." The deposit cap point can be ignored
- Toni Wahrstätter
Because the constant could just be raised too, super simple with 8282. What this eip solves is stable 4788 proofs
- Greg K | Lido
Reacted to "If we put more wor..." with 🤔
- Barnabas
sounds like most clients are almost there tho, would be a waste to let it all go now.
- Dima Gusakov | Lido
Reacted to "sounds like most cli..." with 👍
- kasey
Reacted to "sounds like most clients are almost there tho, would be a waste to let it all go now." with 👍
- Satyajit
Reacted to "sounds like most c..." with 👍
- cayman
Reacted to "sounds like most cli..." with 👍
- kasey
When is devnet-7?
- Etan (Nimbus)
when is devnet 7 expected?
- Barnabas
devnet 7 launch date estimated in 2/3 w
- Etan (Nimbus)
From Nimbus side OK
- Nico Flaig
Reacted to "I’m gonna block dora..." with 😂
- cayman
Lodestar yes
- Dustin
Reacted to "sounds like most c..." with 👍
- Nico Flaig
Reacted to "Lodestar has an impl..." with 🔥
- Nico Flaig
Reacted to "Nimbus has an implem..." with 🔥
- kasey
3 weeks is better than 2 haha, it’s close but possible
- Nico Flaig
Reacted to "Lighthouse has an im..." with 🔥
- Eitan Seri-Levi
Lighthouse yes
- Barnabas
Reacted to "3 weeks is better than 2 haha, it’s close but possible" with 😂
- Barnabas
Reacted to "Lighthouse yes" with 😂
- Etan (Nimbus)
Reacted to "3 weeks is better th..." with 😂
- Nico Flaig
no beacon-api merges until 7688 is in a devnet, deal?
- Barnabas
We don’t need all clients for genesis for devnet 7
- Etan (Nimbus)
Reacted to "sounds like most cli..." with 😂
- cayman
Reacted to "no beacon-api merges..." with 😆
- Etan (Nimbus)
Reacted to "We don’t need all cl..." with 🎉
- Eitan Seri-Levi
Reacted to "no beacon-api merges..." with 😆
- saulius
Grandine maybe, it really depends if there will be left enough of time after ePBS works.
- Parithosh Jayanthi
Reacted to "Grandine maybe, it really depends if there will be left enough of time after ePBS works." with 👍
- kasey
IMO if it doesn’t go into devnet-7 it should be in devnet-8, I’m against DFI
- Parithosh Jayanthi
- Greg K | Lido
May I ask why to DFI?
- Dima Gusakov | Lido
But it is good for ETH in general
- Greg K | Lido
Reacted to "But it is good for..." with 👍
- Marius Van Der Wijden (M)
Just don't exit
- Justin Florentine (Besu)
it limits consolidation agility
- Parithosh Jayanthi
Reacted to "Just don't exit" with 🥲
- Justin Traglia
Reacted to "Just don't exit" with 😄
- Parithosh Jayanthi
Reacted to "Just don't exit" with 😄
- Chris Berry
Reacted to "it limits consolidat..." with 👍
- Louis
Reacted to "Just don't exit" with 😄
- Kalo | Obol
Reacted to "Just don't exit" with 😄
- Justin Florentine (Besu)
Replying to "Just don't exit" 33% of eth is staked. hold my beer
- Parithosh Jayanthi
Reacted to "33% of eth is staked. hold my beer" with 🤪
- Marius Van Der Wijden (M)
In my opinion we should not waste complexity for this "gap". It's just not important imho
- Greg K | Lido
Isn't it important allowing entities to cheat and exit faster than everyone else?
- Greg K | Lido
Especially when some of them create a product out of it
- terence
Replying to "In my opinion we sho..." I tend to agree with this. We should do only if it’s a “strong must have”. But ill leave this decision to whether that’s true
- Francesco
I can repropose but I am not going to champion the EIP, if someone wants to do it go ahead
- Parithosh Jayanthi
- Marius Van Der Wijden (M)
If clients have not looked into it at all we should dfi it
- Nico Flaig
it's on devnet-6 already
- Nico Flaig
5*
- Parithosh Jayanthi
https://ethereum-magicians.org/t/eip-8282-builder-execution-requests/28699
- terence
We do need to slash validators on devnet5 right
- Dima Gusakov | Lido
Replying to "We do need to slash ..." There are a few slashed on devote-6
- Dima Gusakov | Lido
Replying to "We do need to slash ..." So it will be tested there
- Francesco
Re: 8080, iirc this was discussed at interop mainly. The call transcripts have essentially nothing about it
- Justin Traglia
I support setting target=8, max=64.
- Toni Wahrstätter
Sounds like a good trade off
- Parithosh Jayanthi
Replying to "Re: 8080, iirc this ..." Thank you! I was just frantically searching where that was too
- Parithosh Jayanthi
Replying to "Re: 8080, iirc this ..." We can prep some doc indicating the issues then by next acd
- Kevaundray Wedderburn
Do we have a date for when we want to freeze changes and then start/finish external audits for 8282?
- cayman
Reacted to "I support setting ta..." with 👍
- Barnabas
it would be 3 constants?
- Justin Traglia
Replying to "it would be 3 consta..." The 3rd one is in the consensus-specs.
- Barnabas
Replying to "Do we have a date fo..." post devnet 7 for sure
- Barnabas
Replying to "Do we have a date fo..." so probably end of Jul
- Nico Flaig
Reacted to "The 3rd one is in th..." with 👍
- Kevaundray Wedderburn
Reacted to "so probably end of J..." with 👍
- Nico Flaig
Replying to "it would be 3 consta..." https://github.com/ethereum/sys-asm/pull/43#discussion_r3470524501 comment related to what I meant
- Marius Van Der Wijden (M)
Sounds a but brittle tbh
- Marius Van Der Wijden (M)
I would rather replace the contract
- Barnabas
maybe pr against the sys contract repo?
- Toni Wahrstätter
How often do we expect a single builder to deposit?
- Marius Van Der Wijden (M)
Reacted to maybe pr against the... with "👍"
- Toni Wahrstätter
64 sounds like more than enough
- Nico Flaig
Reacted to "I support setting ta..." with 👍
- Justin Traglia
Replying to "Sounds a but brittle..." My opinion is that dynamic limits are probably not necessary. I think it’s unlikely that we’ll change them later. It just took us a while to decide on the values.
- Justin Traglia
Reacted to "64 sounds like more ..." with 👍
- Nico Flaig
Replying to "64 sounds like more ..." to be clear that's the max, the target=8, but the expected rate of builder deposits in the normal case is likely far below 1 per payload
- Justin Florentine (Besu)
LUCID needs CL oversight!
- Toni Wahrstätter
Replying to "64 sounds like more ..." Right. I think that's fine
- Barnabas
Devnet 7 spec sheet is up: https://notes.ethereum.org/@ethpandaops/glamsterdam-devnet-7
- Toni Wahrstätter
Replying to "64 sounds like more ..." Maybe we could set the target to idk 16 or 32. What matters is that it's capped
- Parithosh Jayanthi
https://ethereum-magicians.org/t/mascot-needed-for-glamsterdam-upgrade-voting-closes-july-8/26008
- Barnabas
POLAR BEARS MUST WIN
- nixo
flamingo winning right now, polar bear second in lead
- Toni Wahrstätter
Polar bear!
- Luca | Vero
Reacted to "https://ethereum-mag..." with 🦫
- Iván | ethrex
Reacted to "POLAR BEARS MUST WIN" with 🐻❄️
- Satyajit
Reacted to "POLAR BEARS MUST W..." with 🔥
- nixo
Reacted to "POLAR BEARS MUST WIN" with 💯
- Satyajit
Reacted to "POLAR BEARS MUST W..." with 🐻❄️
- nixo
Reacted to "POLAR BEARS MUST WIN" with 🐻❄️
- Satyajit
Reacted to "POLAR BEARS MUST W..." with 💯
- Mehdi Aouadi
Reacted to "POLAR BEARS MUST WIN" with 💯
- Parithosh Jayanthi
Reacted to "POLAR BEARS MUST WIN" with 🐻❄️
- bfqm
Reacted to "POLAR BEARS MUST W..." with 💯
- Parithosh Jayanthi
Reacted to "flamingo winning right now, polar bear second in lead" with ❤️
- Justin Florentine (Besu)
where my hamsters at?
- Francesco
Reacted to "POLAR BEARS MUST WIN" with 🐻❄️
- Enrico Del Fante (tbenr)
Reacted to "POLAR BEARS MUST WIN" with 🐻❄️
- Marius Van Der Wijden (M)
I don't like people offering money on acd tbh
- NC
Reacted to POLAR BEARS MUST WIN with "🐻❄️"
- danceratopz
Reacted to "Polar bear!" with 🐻❄️
- Barnabas
Replying to "where my hamsters at..." polar bears ate them
- Nico Flaig
Replying to "64 sounds like more ..." no, you want it lower to not be able to dos the validator sweep
- Daniel Knopik
Reacted to "I don't like peopl..." with 👍
- Nico Flaig
Reacted to "POLAR BEARS MUST WIN" with 🐻❄️
- Satyajit
Reacted to "Polar bear!" with 🐻❄️
- Stefan Starflinger
Reacted to "POLAR BEARS MUST WIN" with 🐻❄️
- Nico Flaig
Reacted to "Polar bear!" with 🐻❄️
- Kalo | Obol
Reacted to "I don't like people ..." with 👍
- Justin Traglia
Reacted to "no, you want it lowe..." with 👍
Call summary
Targets
- •glamsterdam-devnet-7 launch — ~2–3 weeks out (mid-July 2026) - 00:10:42
- •EIP-8282 external audit — post devnet-7, ~end of July 2026 - 00:47:03
Decisions
- •EIP-7688 (progressive containers) targeted for devnet-7; not DFI'd; revisit if devnet-6 instability warrants it - 00:34:04
- •EIP-8080 (consolidation-based exits) DFI'd from Glamsterdam; authors may re-propose for Hegota - 00:36:07
- •ePBS circuit breakers should track missing payloads (not beacon blocks); no standardization required, client-specific implementations acceptable - 00:17:01
Highlights
- Client Updates:
- ·ePBS circuit breakers: Prysm uses consensus-block-based breaker (acknowledged as incorrect for ePBS); needs updating to track missing payloads, not beacon blocks - 00:14:46
- Organizational:
- ·Lucid (encrypted mempool EIP-8184) seeking CL spec guidance; Justin Traglia volunteered to help Besu team - 00:50:27
- ·Glamsterdam mascot vote open; voting closes July 8 - 00:51:37
- Testing Progress:
- ·Pre-fork slashing test planned on devnet-6 just before Gloas fork activation - 00:09:20
- ·EIP-8282 contract constants under discussion: target=8, max=64 preferred over current 32/256; reduces builder withdrawability delay - 00:44:04
- Fork Status And Schedule:
- ·glamsterdam-devnet-6 launched hours ago; Gloas fork in ~40 min; EL: Besu, ethrex, Geth; CL: all except Teku - 00:07:25
- ·glamsterdam-devnet-5 stable for 3 weeks; Lighthouse range-sync bug outstanding; Prysm peering bug resolved - 00:07:33
- ·Spec bug (consensus-specs #5399) confirmed valid by Terrence; patch targeting devnet-7 - 00:10:24
- Eip Proposals Glamsterdam:
- ·EIP-7688 (progressive containers): 3 clients yes (Nimbus, Lodestar, Lighthouse), 3 maybe (Prysm, Teku, Grandine); targeting devnet-7; DFI deferred - 00:26:52
- ·EIP-8080 (consolidation exit mechanism): DFI'd; authors encouraged to re-propose for Hegota - 00:35:14
- ·EIP-8282 (builder execution requests): old contract on devnet-6 for workflow testing; updated contract (fee-per-enqueue) targets devnet-7; external audit post-devnet-7 (~end of July) - 00:42:34
Action Items
- •All CL client teams - Review and approve Beacon API PRs (608, 580, 606, 621) by ACDT - 00:19:28
- •Nico Flaig / EIP-8282 authors - Open PR against sys-asm repo for EIP-8282 contract constants (target=8, max=64); signal when contract is final for audit scheduling - 00:47:00
Key decisions
EIP-7688 (progressive containers) targeted for devnet-7; not DFI'd; revisit if devnet-6 instability warrants it
devnet-7EIP-8080 (consolidation-based exits) DFI'd from Glamsterdam; authors may re-propose for Hegota
DFIePBS circuit breakers should track missing payloads (not beacon blocks); no standardization required, client-specific implementations acceptable
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.