- Glamsterdam
- Hegotá
- Revisiting fork scope after devcon discussions
- summary by @adietrichs
- pointer: FOCIL decisions will be made on ACDC, EL core devs are asked to participate
- open discussion
- next steps
- Frame transactions updates
- breakout summary
- minimal leanSPHINCS frames proposal by @ilitteri
- Devnets next steps
- Revisiting fork scope after devcon discussions
AllCoreDevs - Execution #249
Transcript
- Pooja Ranjan
gm gm
- Lean F | ethrex
Gm
- marc | wolovim
gm
- Ben Edgington
Need music
- Kamil Salakhiev
Reacted to "Need music" with 👍
- Barnabas
Replying to "Need music" start it
- Barnabas
Reacted to "Need music" with 👍
- Barnabas
mid january without spec change ?
- Stefan Starflinger
cool
- Barnabas
can we schedule hoodi then?
- Rahul (EF)
yes a lot of people wanted to know hoodi dates
- Will Corcoran
Reacted to "can we schedule hoodi then?" with ➕
- Iván | ethrex
Reacted to "can we schedule hoodi then?" with ➕
- Stefan Starflinger
Replying to "can we schedule hoodi then?" 24.12
- Toni Wahrstaetter
Reacted to "can we schedule hoodi then?" with ➕
- Fredrik
Reacted to "24.12" with 🎅
- Will Corcoran
Reacted to "24.12" with ✝️
- spencer
Reacted to "24.12" with 🎅
- Povilas Liubauskas
Reacted to "24.12" with 🎅
- danceratopz
Reacted to "24.12" with 🎅
- Barnabas
I’d propose hoodi for EPOCH: 140800 UTC: 2026-12-03 06:50:00
- Stefan Starflinger
is that a sunday?
- Justin Florentine (Besu)
so nobody wants to use this time to fix the spec?
- Dima Gusakov | Lido
Mid Jan is terrible option for Lido
- Stefan Starflinger
Replying to "Mid Jan is terrible option for Lido" whats better?
- Dima Gusakov | Lido
We would appreciate end of January at least
- Ansgar Dietrichs
Replying to "Mid Jan is terrible ..." it’s just the earliest possible
- Ansgar Dietrichs
Replying to "Mid Jan is terrible ..." I think late Jan / early Feb is more realistic
- Fredrik
I’d guess end of jan is more likely than mid
- Parithosh Jayanthi
Reacted to "I’d guess end of jan is more likely" with 👍
- Stefan Starflinger
Reacted to "I’d guess end of jan is more likely than mid" with 👍
- Toni Wahrstaetter
Replying to "so nobody wants to use this time to fix the spec?" spec changes would be more delay than jan (incl new audits, devnets, testnets, etc)
- Barnabas
Replying to "is that a sunday?" ah the page view messed it up
- Fredrik
Reacted to "We would appreciate end of January at least" with 👍
- Greg K | Lido
Reacted to "Mid Jan is terrible option for Lido" with 👍
- Parithosh Jayanthi
Reacted to "24.12" with 🎅
- Stefan Starflinger
Replying to "I’d propose hoodi for EPOCH: 140032 UTC: 2026-11-29 20:54:48" 30-11?
- marc | wolovim
Replying to "We would appreciate ..." For hoodi or mainnet?
- Stefan Starflinger
Replying to "We would appreciate end of January at least" mainnet i think
- Justin Florentine (Besu)
Replying to "so nobody wants to u..." yes i understand that, whats the cost of that delay? seems quite small to me
- Dragan Rakita
Reacted to "24.12" with 🎅
- Dima Gusakov | Lido
Replying to "We would appreciate ..." Mainnet. Because of the governance viability. The vote is 10 days, and people will not vote for the upgrade right after New Year
- Iván | ethrex
Reacted to "We would appreciate end of January at least" with 👍
- marc | wolovim
Reacted to "Mainnet. Because of the governance viability. The vote is 10 days, and people will not vote for the upgrade right after New Year" with 👍
- Iván | ethrex
Reacted to "Mainnet. Because of the governance viability. The vote is 10 days, and people will not vote for the upgrade right after New Year" with 👍
- Toni Wahrstaetter
Replying to "so nobody wants to use this time to fix the spec?" it delays shipping all the features that come in glam. Feels like a huge cost to me. Ofc, I'm always assuming we provide every app enough time if code changes are needed.
- Barnabas
Replying to "so nobody wants to u..." what spec change you proposing that we wanna fix ?
- Greg K | Lido
Reacted to "Mainnet. Because of the governance viability. The vote is 10 days, and people will not vote for the upgrade right after New Year" with 👍
- Toni Wahrstaetter
This would mean hoodi this year, mainnet towards end of jan, still allowing enough time for validators to upgrade. I like it.
- Potuz
how long before the Hoodi devnet do we want to have releases out?
- Dustin
would sepolia remain viable under current plans?
- Toni Wahrstaetter
we shouldn't be overly aggressive in setting the mainnet timeline when apps would need more time (which doesn't seem like the case right now), but pre-committing to a hoddi date makes sense.
- Parithosh Jayanthi
Replying to "would sepolia remain..." Yes it would be
- Iván | ethrex
@Rahul (EF) will this extra 2-week delay proposal impact on having Glamsterdam on mainnet for late January?
- Justin Traglia
We still need to make a decision on the CL about binding to the payload. One proposal (which is fairly popular with devs) is a breaking change.
- spencer
from what i looked at, 8037 affected around 11% of broken contracts, and rest are gas repricings related
- felipe
Reacted to "can we schedule hoodi then?" with ➕
- Barnabas
anyone took a look how to fix try catch?
- cayman
Reacted to "We still need to mak..." with 👍
- Justin Traglia
Replying to "We still need to m..." Maybe @cayman can provide more details. I'm a bit uncomfortable scheduling hoodi before a decision on this is made.
- Iván | ethrex
Is this reg a new finding in ePBS?
- Dustin
Reacted to "We still need to m..." with 👍
- Barnabas
Replying to "anyone took a look h..." wasn’t this a solidity issue?
- Dustin
Replying to "We still need to m..." agree
- Luis Pinto
Replying to "anyone took a look h..." Move refunds to the end?
- Rahul (EF)
Tbf, there is some 8037 spec feedback from app teams (e.g. try/catch) but that is separate to the security issues
- Dustin
Replying to "We still need to m..." (re scheduling Hoodi before this is addressed on CL side)
- spencer
Replying to "anyone took a look..." there is a work around for new/upgradeable contracts
- Radek | solidity
Replying to "anyone took a look h..." There are a couple of soolutions but all of them hardcode gas AFAIK
- Justin Traglia
Replying to "We still need to m..." Ivan, not super new. We've been discussing it for a few weeks now. There is a non-breaking solution, but it's sort of ugly.
- Radek | solidity
Replying to "anyone took a look h..." so next repricing will be a problem again
- Iván | ethrex
Reacted to "Ivan, not super new. We've been discussing it for a few weeks now. There is a non-breaking solution, but it's sort of ugly." with 👍
- Rahul (EF)
Replying to "@Rahul (EF) will t..." all the highs and mediums were fine with mid jan so they wouldn't care I guess Having hoodi will give them a chance to test
- Iván | ethrex
Replying to "We still need to mak..." Thanks for the context
- Justin Traglia
Reacted to "Thanks for the con..." with ❤️
- Ansgar Dietrichs
justin do you have your hand up again btw?
- Tanguy
Replying to "anyone took a look..." not to go too much into details, but indeed, to have a try catch that can not be abused with making the inner call OOG to force the execution to go in the catch, one HAVE TO somewhere make assumption about gas costs, that's a inherent issue of ethereum, that is both before and after glamsterdam
- lightclient
pls no opcode for state gas
- Iván | ethrex
Reacted to "all the highs and mediums were fine with mid jan so they wouldn't care I guess Having hoodi will give them a chance to test" with 👍
- cloud1992
The vast majority of those affected can fix everything by mid-January. Let's not delay this any further.
- lightclient
Replying to "pls no opcode for state gas" GAS is the root of all evil 😂
- Parithosh Jayanthi
I dont think any change to the spec is just the weeks spent in figuring out the fix. Instead it’s restarting the full fork shipping process. It will be months of delay and quite likely we’ll just introduce another issue.
- Dragan Rakita
Reacted to "GAS is the root of all evil 😂" with 💯
- cayman
Replying to "We still need to mak..." Spec proposal here https://github.com/ethereum/consensus-specs/pull/5820
- Mario Vega
Reacted to "I dont think any change to the spec is just the weeks spent in figuring out the fix. Instead it’s restarting the full fork shipping process. It will be months of delay and quite likely we’ll just introduce another issue." with 👍
- Rahul (EF)
Reacted to "The vast majority ..." with 👍
- lightclient
Reacted to "24.12" with 🎅
- lightclient
Reacted to "24.12" with ✝️
- Guru
Reacted to "I dont think any cha..." with 👍
- Parithosh Jayanthi
Replying to "I dont think any cha..." So if all the affected contracts can fix stuff by jan, then that is likely the sane timeline for me
- lightclient
Replying to "can we schedule hoodi then?" merry grindmas :)
- Francesco D'Amato
Reacted to "pls no opcode for state gas" with 👍
- Dragan Rakita
Reacted to "can we schedule hoodi then?" with ➕
- Mario Vega
Replying to "I dont think any cha..." This is my main concern, we need to run the contract analysis all over again, and will probably hit other broken contracts
- Trent Van Epps
It’s better if chat commentary unmutes and says things out loud
- Dustin
the thing with the CL side thing is... the status quo do-nothing doesn't break things either, but there's ... for "reasons" pressure to do something
- Caspar Schwarz-Schilling
Reacted to "I dont think any change to the spec is just the weeks spent in figuring out the fix. Instead it’s restarting the full fork shipping process. It will be months of delay and quite likely we’ll just introduce another issue." with 👍
- Iván | ethrex
Reacted to "merry grindmas :)" with 🎄
- Iván | ethrex
Reacted to "24.12" with 🎅
- Luis Pinto
Replying to "pls no opcode for st..." Yes I also don’t like it, just saying that without it and refunding in place, how can anyone make sense of the try catch problem?
- spencer
changing the spec and tests is easy and fast... its everything after.. are there more contract breaks, benchmarking rounds etc
- Justin Florentine (Besu)
Reacted to "GAS is the root of all evil 😂" with 💯
- Kevaundray Wedderburn
Reacted to "pls no opcode for state gas" with 👍
- Iván | ethrex
Reacted to "pls no opcode for state gas" with 👍
- Parithosh Jayanthi
Reacted to "GAS is the root of all evil 😂" with 💯
- Iván | ethrex
Reacted to "pls no opcode for state gas" with 🙏
- Łukasz Rozmej
Reacted to "We would appreciate end of January at least" with 👍
- Radek | solidity
Replying to "anyone took a look h..." links to OZ and GnosisSafe solutions for similar problem. https://github.com/safe-fndn/safe-smart-account/blob/247b8ce3bdbf204a120e77f5b369d137424ddc58/contracts/Safe.sol#L197 https://github.com/OpenZeppelin/openzeppelin-contracts/blob/7ee8e000e9c534cbb055895731808c6cc679cb18/contracts/metatx/ERC2771Forwarder.sol#L344-L374
- Parithosh Jayanthi
Reacted to "changing the spec and tests is easy and fast... its everything after.. more contract changes, benchmarking rounds etc" with 🫡
- Francesco D'Amato
It might also break more contracts than what is currently known, because some contracts might work with the higher gas costs due to taking advantage of the refund, and would not work if you take away the refund
- Mario Vega
Reacted to "It might also break more contracts than what is currently known, because some contracts might work with the higher gas costs due to taking advantage of the refund, and would not work if you take away the refund" with 👍
- Potuz
Reacted to "It might also break ..." with 👍
- Toni Wahrstaetter
Reacted to "It might also break more contracts than what is currently known, because some contracts might work with the higher gas costs due to taking advantage of the refund, and would not work if you take away the refund" with 👍
- Rahul (EF)
Replying to "It might also brea..." also you need to get feedback from apps again if the spec changes actually helped them etc
- Łukasz Rozmej
Do we have any writeup on what different spec changes would break?
- felipe
Reacted to "merry grindmas :)" with 🎄
- Parithosh Jayanthi
Reacted to "also you need to get feedback from apps again if the spec changes actually helped them etc" with 👍
- Caspar Schwarz-Schilling
Reacted to "It might also break more contracts than what is currently known, because some contracts might work with the higher gas costs due to taking advantage of the refund, and would not work if you take away the refund" with 👍
- felipe
Reacted to "I dont think any change to the spec is just the weeks spent in figuring out the fix. Instead it’s restarting the full fork shipping process. It will be months of delay and quite likely we’ll just introduce another issue." with 👍
- felipe
Reacted to "So if all the affected contracts can fix stuff by jan, then that is likely the sane timeline for me" with 👍
- Mario Vega
Replying to "Do we have any write..." I think we’ve been focusing so far on what the current specs break and not visualized entirely what other spec changes break, it’s another full analysis IMO
- Justin Florentine (Besu)
do most people think breaking try/catch in solidity is comparable to similar backward-incompatible changes? like difficulty vs. prevrandao for example? because to me thats a step change in impact.
- Radek | solidity
Replying to "anyone took a look h..." I can prepare some deeper analysis for next week if someone is interested. Can also discuss it with Juicebox
- Barnabas
I think 2w from now for a decision would be too late. I think it would be good to make a final decision on hoodi next ACDC
- Iván | ethrex
Reacted to "I think 2w from now for a decision would be too late. I think it would be good to make a final decision on hoodi next ACDC" with ➕
- Francesco D'Amato
Replying to "Do we have any write..." It’s a lot of work to do this, would have to basically redo all of the scanning and triaging etc...
- lightclient
Reacted to "I think 2w from now for a decision would be too late. I think it would be good to make a final decision on hoodi next ACDC" with ➕
- Toni Wahrstaetter
Replying to "do most people think breaking try/catch in solidity is comparable to similar backward-incompatible changes? like difficulty vs. prevrandao for example? because to me thats a step change in impact." it's not that every try/catch breaks but only those where gas_left is used to count used gas
- Rahul (EF)
Reacted to "it's not that ever..." with 👍
- Hubert (ChainSecurity)
Honest, dumb question: Is it an option to bring frame transactions into Glamsterdam? That would fix the UX issues that Uniswap is rightly concerned about.
- Rahul (EF)
Reacted to "it's not that ever..." with ➕
- felipe
Reacted to "pls no opcode for state gas" with 😅
- Lean F | ethrex
What’s the deadline to directly drop the EIP if we don’t see spec changes feasible and we don’t want to launch without spec changes?
- Justin Florentine (Besu)
Replying to "do most people think..." i think if it breaks inconsistently depending on use, thats broken
- Dustin
Replying to "Honest, dumb quest..." that would be a major delay
- felipe
Reacted to "It might also break more contracts than what is currently known, because some contracts might work with the higher gas costs due to taking advantage of the refund, and would not work if you take away the refund" with 👍
- Francesco D'Amato
Replying to "do most people think..." Well we do know there’s just a handful of contracts that are broken by this
- Hubert (ChainSecurity)
Reacted to "that would be a ma..." with 👍
- Radek | solidity
Is anybody interested in a deeper analysis on try/catch OOG problem?
- Łukasz Rozmej
Replying to "Do we have any write..." Yeah but at least having some summary on the invariants and some examples what breaks
- Luis Pinto
Reacted to "Is anybody interested in a deeper analysis on try/catch OOG problem?" with ➕
- Łukasz Rozmej
Replying to "Do we have any write..." Doesn't have to be complete
- Toni Wahrstaetter
Replying to "Is anybody interested in a deeper analysis on try/catch OOG problem?" yeah we did a lot of analysis internally as well as chainsecurity and dedaub
- Felix (Geth)
Reacted to "Is anybody interested in a deeper analysis on try/catch OOG problem?" with ➕
- Potuz
CL devs are here, EL devs can come to ADCD!
- danceratopz
Reacted to "Is anybody intereste..." with ➕
- Iván | ethrex
Reacted to "CL devs are here, EL devs can come to ADCD!" with ➕
- Toni Wahrstaetter
Replying to "Is anybody interested in a deeper analysis on try/catch OOG problem?" ahh. I read it as "if this was done?".. sry
- Toni Wahrstaetter
Replying to "Is anybody interested in a deeper analysis on try/catch OOG problem?" I'm interested though if you have such an analysis
- Mario Vega
Replying to "Do we have any write..." Not having a complete analysis on contracts is what brought us here in the first place
- Iván | ethrex
Replying to "CL devs are here, EL..." I (try to) go to every one 😄
- Fredrik
If we do end up making spec changes then we should imo plan for 3-4 months time to validate all the changes after the spec changes has been made (implementing changes, launching a new sepolia, running the bug bounty, running audits, doing tests, etc) to just make some spec changes and quickly yolo it through to mainnet by skipping the upgrade process seems like it would be a massive risk compared to what we know about glamsterdam as it is today imo.
- saulius
Replying to "CL devs are here, EL..." *ACDC
- Łukasz Rozmej
Replying to "CL devs are here, EL..." ABCD?
- Dustin
Replying to "CL devs are here, ..." well, also, 8037 and the binding thing are kind of separate -- each can be seaprately ok or not
- Iván | ethrex
Reacted to "*ACDC" with ➕
- Parithosh Jayanthi
Reacted to "If we do end up making spec changes then we should imo plan for 3-4 months time to validate all the changes after the spec changes has been made (implementing changes, launching a new sepolia, running the bug bounty, running audits, doing tests, etc) to just make some spec changes and quickly yolo it through to mainnet by skipping the upgrade process seems like it would be a massive risk compared to what we know about glamsterdam as it is today imo." with 🫡
- felipe
Reacted to "ABCD?" with 🔠
- danceratopz
Reacted to "ABCD?" with 🔠
- Pablo | Lambda
Reacted to "ABCD?" with 🔠
- Justin Traglia
Reacted to "*ACDC" with ➕
- Hubert (ChainSecurity)
Reacted to "ABCD?" with 🔠
- cayman
Reacted to "CL devs are here, EL..." with ➕
- Greg K | Lido
Reacted to "If we do end up making spec changes then we should imo plan for 3-4 months time to validate all the changes after the spec changes has been made (implementing changes, launching a new sepolia, running the bug bounty, running audits, doing tests, etc) to just make some spec changes and quickly yolo it through to mainnet by skipping the upgrade process seems like it would be a massive risk compared to what we know about glamsterdam as it is today imo." with 👍
- Barnabas
Reacted to "If we do end up making spec changes then we should imo plan for 3-4 months time to validate all the changes after the spec changes has been made (implementing changes, launching a new sepolia, running the bug bounty, running audits, doing tests, etc) to just make some spec changes and quickly yolo it through to mainnet by skipping the upgrade process seems like it would be a massive risk compared to what we know about glamsterdam as it is today imo." with 👍
- Potuz
Replying to "CL devs are here, EL..." lol don't pick on the dyslexic
- Radek | solidity
My zoom decided to update so I can’t read the messages 🙂 I will read them after the call on forkcast.
- Justin Florentine (Besu)
Replying to "CL devs are here, EL..." @Łukasz Rozmej it's spelled ACAB
- Marius van der Wijden
Also any kind of change will necessitate another round of reviews and outreach to the community. We have just gotten them to update their contracts once, doing it again isn't great
- Iván | ethrex
+1 to Dragan
- felipe
Reacted to "If we do end up making spec changes then we should imo plan for 3-4 months time to validate all the changes after the spec changes has been made (implementing changes, launching a new sepolia, running the bug bounty, running audits, doing tests, etc) to just make some spec changes and quickly yolo it through to mainnet by skipping the upgrade process seems like it would be a massive risk compared to what we know about glamsterdam as it is today imo." with 👍
- felipe
Reacted to "If we do end up making spec changes then we should imo plan for 3-4 months time to validate all the changes after the spec changes has been made (implementing changes, launching a new sepolia, running the bug bounty, running audits, doing tests, etc) to just make some spec changes and quickly yolo it through to mainnet by skipping the upgrade process seems like it would be a massive risk compared to what we know about glamsterdam as it is today imo." with ➕
- Iván | ethrex
Reacted to "ABCD?" with 🔠
- Marius van der Wijden
I don't care too much about timelines. I just don't think that breaking other invariants will be better than what we have right now
- spencer
I can work on a doc for next week on numbers regarding spec changes, just so the visability is more public. Happy to also check what these new spec changes might break, but it will be "partial impact" analysis.
- Mario Vega
Reacted to "I don't care too much about timelines. I just don't think that breaking other invariants will be better than what we have right now" with 👍
- Barnabas
I would have expected explicit changeset by today if we were to consider any actual spec changes. I haven’t seen any concrete proposals that were backed by multiple people.
- spencer
I do think we should keep the current spec though
- Lean F | ethrex
Replying to "If we do end up maki..." Isn’t it a no go then?
- Dragan Rakita
Reacted to "CL devs are here, EL devs can come to ADCD!" with 😆
- danceratopz
Reacted to "I can work on a doc ..." with 💙
- Barnabas
Reacted to "I don't care too much about timelines. I just don't think that breaking other invariants will be better than what we have right now" with 👍
- Dragan Rakita
Reacted to "I don't care too much about timelines. I just don't think that breaking other invariants will be better than what we have right now" with 👍
- felipe
Reacted to "+1 to Dragan" with ➕
- Łukasz Rozmej
I'm fine with going with the current spec, it is just super hard to be sure if it is an optimal decision.
- Dragan Rakita
Reacted to "I would have expected explicit changeset by today if we were to consider any actual spec changes. I haven’t seen any concrete proposals that were backed by multiple people." with 💯
- Dragan Rakita
Reacted to "I would have expected explicit changeset by today if we were to consider any actual spec changes. I haven’t seen any concrete proposals that were backed by multiple people." with ➕
- Iván | ethrex
Reacted to "I'm fine with going with the current spec, it is just super hard to be sure if it is an optimal decision." with 👍
- Hubert (ChainSecurity)
Replying to "I can work on a do..." What are the proposals for changes?
- Łukasz Rozmej
Reacted to "I can work on a doc for next week on numbers regarding spec changes, just so the visability is more public. Happy to also check what these new spec changes might break, but it will be "partial impact" analysis." with 💙
- Toni Wahrstaetter
Replying to "I can work on a doc for next week on numbers regarding spec changes, just so the visability is more public. Happy to also check what these new spec changes might break, but it will be "partial impact" analysis." e.g. delaying the refund after tx exec, or refunding into reservoid first
- jonny (geth)
Reacted to "I can work on a doc for next week on numbers regarding spec changes, just so the visability is more public. Happy to also check what these new spec changes might break, but it will be "partial impact" analysis." with 💙
- Toni Wahrstaetter
Replying to "I can work on a doc for next week on numbers regarding spec changes, just so the visability is more public. Happy to also check what these new spec changes might break, but it will be "partial impact" analysis." both have trade-offs though
- Dragan Rakita
Replying to "I'm fine with going ..." To be blunt, you needed to prepare yourself today to be sure
- spencer
Yeah all of this information is not fully public, so we have some that know more details than others :)
- Hubert (ChainSecurity)
Reacted to "e.g. delaying the ..." with 👍
- spencer
Reacted to "e.g. delaying the ..." with 👍
- spencer
Replying to "I can work on a do..." What does chain security recommend we do for Glamsterdam? :D
- Ben Edgington
Replying to "CL devs are here, EL devs can come to ADCD!" We'll likely be discussing the future of FOCIL on next ACDC, so please show up if you care.
- felipe
Reacted to "We'll likely be discussing the future of FOCIL on next ACDC, so please show up if you care." with 💙
- Toni Wahrstaetter
Reacted to "ABCD?" with 🔠
- Iván | ethrex
Reacted to "We'll likely be discussing the future of FOCIL on next ACDC, so please show up if you care." with 🫡
- spencer
Replying to "I can work on a do..." Rather are you aligned with keep the spec or change the spec?
- Justin Florentine (Besu)
thnx for the explaner: is this a use case that TSTORE can't support?
- Barnabas
Replying to "I'm fine with going ..." yeah, those people that are pushing for the changes should be fully convinced themselves that their own changeset is net benefit vs the current design
- Caspar Schwarz-Schilling
Reacted to "We'll likely be discussing the future of FOCIL on next ACDC, so please show up if you care." with 💙
- spencer
Reacted to "To be blunt, you n..." with 😂
- Radek | solidity
Reacted to "thnx for the explane..." with 👍
- Barnabas
Replying to "I'm fine with going ..." otherwise what are we even discussing
- Francesco Risitano (Tau)
Why not just refund into state gas instead of execution gas?
- Dragan Rakita
Reacted to "yeah, those people that are pushing for the changes should be fully convinced themselves that their own changeset is net benefit vs the current design" with 👍
- Will Corcoran
Reacted to "To be blunt, you needed to prepare yourself today to be sure" with 👍
- felipe
Reacted to "yeah, those people that are pushing for the changes should be fully convinced themselves that their own changeset is net benefit vs the current design" with 👍
- Dragan Rakita
Reacted to "otherwise what are we even discussing" with 💯
- felipe
Reacted to "otherwise what are we even discussing" with ➕
- Will Corcoran
Reacted to "yeah, those people that are pushing for the changes should be fully convinced themselves that their own changeset is net benefit vs the current design" with 👍
- Łukasz Rozmej
Replying to "I'm fine with going ..." I need smarter people than me on it 😛
- Marius van der Wijden
Replying to "thnx for the expla..." Sure, but that would require contracts to update as well
- Francesco D'Amato
Many patterns are broken by this anyway
- danceratopz
Reacted to "We'll likely be disc..." with 🫡
- Hubert (ChainSecurity)
Replying to "thnx for the expla..." I think partially it is also old contracts, deployed before TSTORE
- Jihoon
Replying to "CL devs are here, EL..." We'll also be discussing PQ FOCIL in the next FOCIL breakout on the 24th. Please drop by to share opinions and concerns!
- Francesco Risitano (Tau)
Reacted to "Many things are broken by this anyway" with 👍
- Iván | ethrex
Reacted to "We'll also be discussing PQ FOCIL in the next FOCIL breakout on the 24th. Please drop by to share opinions and concerns!" with 👍
- Caspar Schwarz-Schilling
Reacted to "CL devs are here, EL devs can come to ADCD!" with 😆
- danceratopz
Reacted to "We'll also be discus..." with 💙
- Ben Edgington
Reacted to "We'll also be discussing PQ FOCIL in the next FOCIL breakout on the 24th. Please drop by to share opinions and concerns!" with 👍
- felipe
Reacted to "We'll also be discussing PQ FOCIL in the next FOCIL breakout on the 24th. Please drop by to share opinions and concerns!" with 👍
- marc | wolovim
Reacted to "ABCD?" with 🔠
- Hubert (ChainSecurity)
Replying to "Yeah all of this i..." Not sure if people have seen, but our public report is here: https://www.chainsecurity.com/security-audit/eip-8037-eip-8038-impact-assessment/ (this doesn't contradict your point)
- Łukasz Rozmej
Do we have a writeup on what breaks?
- Barnabas
I’d be happy to make a decision today. We haven’t heard any concrete proposals to make a spec change.
- danceratopz
Reacted to "Not sure if people h..." with 👍
- Iván | ethrex
Reacted to "I’d be happy to make a decision today. We haven’t heard any concrete proposals to make a spec change." with 👍
- Iván | ethrex
Reacted to "I’d be happy to make a decision today. We haven’t heard any concrete proposals to make a spec change." with ➕
- Dragan Rakita
Replying to "I'm fine with going ..." It is about investing time, in reth team we did think about alternatives and even asked question in discord r&d eip8037 chat. Conclusion is that eip8037 is good and it is important to ship.
- Will Corcoran
Yeah, @marc | wolovim is it possible to do a Zoom poll?
- Toni Wahrstaetter
Reacted to "I’d be happy to make a decision today. We haven’t heard any concrete proposals to make a spec change." with ➕
- jonny (geth)
Reacted to "Why not just refund into state gas instead of execution gas?" with ➕
- Caspar Schwarz-Schilling
Reacted to "I’d be happy to make a decision today. We haven’t heard any concrete proposals to make a spec change." with ➕
- Barnabas
Replying to "I’d be happy to make..." some people are not fully sold, but I don’t see a reason to further delay the decision.
- Dustin
Replying to "I’d be happy to m..." regardless of 8037/etc, the CL binding thing Justin mentioned is still there
- Trent Van Epps
Reacted to "I’d be happy to make a decision today. We haven’t heard any concrete proposals to make a spec change." with 👍
- Toni Wahrstaetter
Replying to "I’d be happy to make a decision today. We haven’t heard any concrete proposals to make a spec change." I tend to agree with this
- Trent Van Epps
Reacted to "I’d be happy to make a decision today. We haven’t heard any concrete proposals to make a spec change." with ➕
- Will Corcoran
Replying to "Yeah, @marc | wolovi..." This feels 90-10
- Justin Traglia
Reacted to "Yeah, @marc | wolo..." with ➕
- Barnabas
Reacted to "Yeah, @marc | wolovim is it possible to do a Zoom poll?" with ➕
- Pablo | Lambda
Reacted to "I’d be happy to make a decision today. We haven’t heard any concrete proposals to make a spec change." with 👍
- jonny (geth)
Reacted to "Do we have a writeup on what breaks?" with ➕
- Pablo | Lambda
Reacted to "I’d be happy to make a decision today. We haven’t heard any concrete proposals to make a spec change." with ➕
- felipe
Reacted to "I’d be happy to make a decision today. We haven’t heard any concrete proposals to make a spec change." with ➕
- Dima Gusakov | Lido
Well, any certainty today will help
- Marius van der Wijden
Reacted to "Well, any certaint..." with ❤️
- Łukasz Rozmej
Reacted to "I do think we should keep the current spec though" with 👍
- Iván | ethrex
Reacted to "Well, any certainty today will help" with ❤️
- felipe
Reacted to "This feels 90-10" with 👍
- Miguel Tenorio
Reacted to "Well, any certainty today will help" with ➕
- spencer
Reacted to "Well, any certaint..." with ❤️
- Lean F | ethrex
Reacted to "Well, any certainty today will help" with ❤️
- felipe
Reacted to "Well, any certainty today will help" with ❤️
- Iván | ethrex
Reacted to "Well, any certainty today will help" with ➕
- felipe
Reacted to "Well, any certainty today will help" with ➕
- marc | wolovim
Replying to "Yeah, @marc | wolovi..." Havent zoom polled before, will pass you the full baton
- Pablo | Lambda
Reacted to "Well, any certainty today will help" with ❤️
- Toni Wahrstaetter
Replying to "I’d be happy to make a decision today. We haven’t heard any concrete proposals to make a spec change." Spec changes are quite costly and won't move the needle. 90-95% are affected by pure repricing and not by something a spec change could even tackle. And spec changes need to be understood before we understand their blast radius
- spencer
Reacted to "Not sure if people..." with ❤️
- Mario Vega
Reacted to "Well, any certainty today will help" with ❤️
- Mario Vega
Reacted to "Well, any certainty today will help" with ➕
- Iván | ethrex
Ethrex goes for keeping things as they are and decide on Hoodi today
- felipe
Reacted to "Ethrex goes for keeping things as they are and decide on Hoodi today" with 👍
- Mega | ethlambda
Reacted to "Well, any certainty today will help" with ❤️
- Toni Wahrstaetter
Penciling in hoody without a strong commitment might be a good compromise?
- Dustin
"Penciling in" is in principle not useful
- Barnabas
Replying to "Penciling in hoody w..." which was ansgar’s proposal
- Dustin
its whole purpose is to do some kind of behavioral nudge and create hard-to-override defaults with less justification
- Dustin
that's why people do it
- Toni Wahrstaetter
Replying to ""Penciling in" is in principle not useful" it get's the process going, otherwise we keep having the same discussion in 2 weeks and there is the need to decide quickly
- jordan
Reacted to ""Penciling in" is in..." with 👍
- Josh Rudolf
Reacted to "Penciling in hoody without a strong commitment might be a good compromise?" with 👍
- Dustin
Replying to ""Penciling in" is ..." the process can get moving without fake promises which, yes, people will regard as commitments regardless
- Dustin
Replying to ""Penciling in" is ..." and kind of betrayals if not met. which is, again, the point of such an approach typically
- jonny (geth)
I'd like to understand why we can't refund to state gas
- Ansgar Dietrichs
oops
- Dima Gusakov | Lido
Reacted to "oops" with 😂
- Dragan Rakita
Reacted to "oops" with 😂
- Ansgar Dietrichs
Replying to "oops" apologies my Erigon friends
- Potuz
it's not "small"
- Greg K | Lido
Reacted to "it's not "small"" with 🌶️
- Iván | ethrex
Reacted to "it's not "small"" with 🌶️
- Mario Vega
Reacted to "it's not "small"" with 🌶️
- Kevaundray Wedderburn
Replying to "it's not "small"" Is this caymans change?
- Gary Schulte
Reacted to "it's not "small"" with 🌶️
- Mega | ethlambda
Reacted to "it's not "small"" with 🌶️
- Justin Traglia
Replying to "it's not "small"" Yes.
- Barnabas
Replying to "it's not "small"" can’t we just add the rlp to the CL ?
- Kevaundray Wedderburn
Reacted to "Yes." with 👍
- Trent Van Epps
Scheduling hoodi still makes sense?
- Dustin
Replying to "it's not "small"" Barnabas: not a decision to be made in ACDE
- Dima Gusakov | Lido
Reacted to "it's not "small"" with 🌶️
- cayman
Replying to "it's not "small"" thats the discussion we haven't yet had...
- marc | wolovim
Reacted to "Scheduling hoodi still makes sense?" with ✏️
- jonny (geth)
Reacted to "it's not "small"" with 🌶️
- Barnabas
Replying to "it's not "small"" or ask the EL via engine api?
- Dustin
Replying to "it's not "small"" (personally I disagree with adding it but separately, processwise)
- Potuz
Replying to "it's not "small"" cc Terence for the "small" comment above
- Luis Pinto
Some analysis on the try catch would be nice IMO
- Barnabas
Reacted to "Some analysis on the try catch would be nice IMO" with 👍
- Dustin
Replying to "it's not "small"" engine API isn't part of the gossip val check
- Radek | solidity
Replying to "Scheduling hoodi sti..." I think we should wait for next ACDC
- Dustin
Replying to "it's not "small"" being proposed
- Ben Edgington
Replying to "it's not "small"" I hear different views of the size depending on who I ask 🤷
- Dustin
Replying to "it's not "small"" so it's fake news to propose that right now
- Barnabas
Replying to "it's not "small"" right but it wouldn’t be a consensus breaking change no?
- Hubert (ChainSecurity)
Replying to "Some analysis on t..." what would you like to know?
- Kevaundray Wedderburn
Replying to "it's not "small"" For reference: https://github.com/ethereum/consensus-specs/pull/5820
- Radek | solidity
I can commit to do it next week if it’s not too late
- Łukasz Rozmej
Reacted to "e.g. delaying the refund after tx exec, or refunding into reservoid first" with 👍
- marc | wolovim
Reacted to "I can commit to do it next week if it’s not too late" with 💙
- marc | wolovim
Reacted to "what would you like to know?" with 👆
- Lean F | ethrex
This decision at least for us is based on the idea that by January (mid or end) users have upgraded and moved away from the issues that this is bringing
- Dustin
Replying to "it's not "small"" But the general point is, ACDE is not the place to bulldoze this just because @barnabas thinks it's a "just"
- Iván | ethrex
Reacted to "This decision at least for us is based on the idea that by January (mid or end) users have upgraded and moved away from the issues that this is bringing" with 👍
- Łukasz Rozmej
Replying to "I can work on a doc ..." @Toni Wahrstaetter Can I read about those trade offs somewhere? I feel under informed
- Radek | solidity
Replying to "Some analysis on the..." Sorry for dump question. Where are the links to chainsec and dedub full reports?
- Luis Pinto
Have invariants changed on the try catch with in-place state refunds. And at which levels
- Justin Florentine (Besu)
Did anyone look at our Option 2 on the agenda?
- Guru
Replying to "I'd like to understa..." Wouldn’t this mean that a storage slot can still be written after a call that ran out of gas. This was one of the invariants
- Barnabas
Replying to "it's not "small"" I mean we have to either make a consensus breaking change, trash sepolia, test another round of devnets, or “just” add rlp to cl temporarily for one fork.
- Mario Vega
Reacted to "both have trade-offs though" with 👍
- Barnabas
Replying to "it's not "small"" thats where my “just” comes from
- Hubert (ChainSecurity)
Replying to "Some analysis on t..." https://www.chainsecurity.com/security-audit/eip-8037-eip-8038-impact-assessment/
- spencer
Replying to "Did anyone look at..." Removing means we can't increase gas limit safely
- Potuz
can we say something about the times from release to shipping Hoodi? this can be set now and will affect decisions on the CL
- spencer
Replying to "Did anyone look at..." ^^8037 that is
- Dustin
Replying to "it's not "small"" we don't really have to do any of it but sure
- Toni Wahrstaetter
Replying to "I can work on a doc for next week on numbers regarding spec changes, just so the visability is more public. Happy to also check what these new spec changes might break, but it will be "partial impact" analysis." yeah i have to check where that doc was. I think someone presented something in mumbai on that
- Ben Adams
Replying to "it's not "small"" ACDE is not the place to bulldoze this just because @barnabas thinks it's a "just" Correct, that's ACDT 😉
- Potuz
Reacted to "ACDE is not the plac..." with 🤣
- Barnabas
Reacted to "ACDE is not the place to bulldoze this just because @barnabas thinks it's a "just" Correct, that's ACDT 😉" with 🤣
- Toni Wahrstaetter
Replying to "I'd like to understand why we can't refund to state gas" yeah, refunding into the reservoir but still during execution only solves a small subset of issues. the main issue around that invariant was the fact that it happens during exec and not after the tx is done
- Greg K | Lido
Reacted to "For reference: https://github.com/ethereum/consensus-specs/pull/5820" with 👀
- Hubert (ChainSecurity)
Reacted to "For reference: htt..." with 👀
- Radek | solidity
Reacted to "https://www.chainsec..." with ❤️
- Ben Edgington
Reacted to "ACDE is not the place to bulldoze this just because @barnabas thinks it's a "just" Correct, that's ACDT 😉" with 🤣
- cayman
Reacted to "ACDE is not the plac..." with 🤣
- Kevaundray Wedderburn
I think it would be good to talk about removing EIPs in light of this information
- Ben Adams
Replying to "I think it would be good to talk about removing EIPs in light of this information" Adding? Still haven't gone through them all
- Iván | ethrex
Reacted to "I think it would be good to talk about removing EIPs in light of this information" with 🫡
- Iván | ethrex
Reacted to "I think it would be good to talk about removing EIPs in light of this information" with ➕
- Dragan Rakita
I love these summaries!
- Kevaundray Wedderburn
Reacted to "Adding? Still haven't gone through them all" with 😁
- Iván | ethrex
Reacted to "I love these summaries!" with ➕
- Mario Vega
Reacted to "I love these summaries!" with ➕
- Potuz
Replying to "I think it would be ..." you're on fire today Ben
- Potuz
Reacted to "Adding? Still haven'..." with 😁
- Dragan Rakita
Reacted to "I think it would be good to talk about removing EIPs in light of this information" with 🫡
- Ben Adams
Reacted to "you're on fire today Ben" with 🔥
- Francesco D'Amato
Reacted to "yeah, refunding into the reservoir but still during execution only solves a small subset of issues. the main issue around that invariant was the fact that it happens during exec and not after the tx is done" with 👍
- spencer
Reacted to "I love these summa..." with ➕
- spencer
Reacted to "I think it would b..." with ➕
- Dragan Rakita
Reacted to "you're on fire today Ben" with 🔥
- Kevaundray Wedderburn
Replying to "I think it would be ..." He’s killing me over here 😂
- spencer
Reacted to "Adding? Still have..." with 😁
- Will Corcoran
Reacted to "I think it would be good to talk about removing EIPs in light of this information" with ➕
- Trent Van Epps
Reacted to "I love these summaries!" with ➕
- Barnabas
Lets do 3.
- soispoke
keyed nonces
- Toni Wahrstätter
Reacted to "Lets do 3." with 🚢
- soispoke
and assertions
- Iván | ethrex
Reacted to "keyed nonces" with ➕
- soispoke
A réagi à "Lets do 3." avec 🚢
- Justin Florentine (Besu)
2 + sphincs
- Mislav | Ethlabs
keyed nonces, recent roots and assertions not on pq critical path imo? but all very good
- soispoke
Répondre à "keyed nonces, rece..." that's the diff between 2 and 3
- frames & friend & pq EIPs
) So I think thats 3! We could always do an EL only fork afterwards for all the removed EIPs, assuming PQ CL takes a while!
- Mislav | Ethlabs
Replying to "keyed nonces, rece..." yes
- frames & friend & pq EIPs
) So I think thats 3! We could always do an EL only fork afterwards for all the removed EIPs, assuming PQ CL takes a while!" with ➕
- soispoke
A réagi à "For me... remove a..." avec ➕
- spencer
Reacted to "Lets do 3." with 🚢
- Francesco D'Amato
Replying to "keyed nonces, recent..." Didn’t recent roots become out of protocol?
- Guru
Option 3 + other PQ related including 8151?
- Mislav | Ethlabs
Replying to "keyed nonces, rece..." yeah but they depend on keyed nonces imo - so if keyed nonces go they do as well. at least thats my latest read
- Csaba
2 or 5, both are just conflict free :-)
- soispoke
Répondre à "keyed nonces, rece..." yeah it did, still is an EIP that is part of the bundle though
- spencer
Reacted to "Option 3 + other P..." with 😁
- Toni Wahrstätter
We should be mindful with considering Frames without the add-ons. They would change the tx object which might hurt adoption on the app layer in case the tx changes again later on. Therefore, prefer keeping frames+.
- spencer
Reacted to "We should be mindf..." with 💯
- Mario Vega
Reacted to "We should be mindful with considering Frames without the add-ons. They would change the tx object which might hurt adoption on the app layer in case the tx changes again later on. Therefore, prefer keeping frames+." with 💯
- Vansh
Reacted to "We should be mindf..." with 💯
- Luis Pinto
Replying to "I'd like to understa..." This could be a good compromise, if the reservoir never gets drained everything is good. Only when it drains problems come up. At least gas turns into monotonic decreasing again
- Dragan Rakita
Reacted to "We should be mindful with considering Frames without the add-ons. They would change the tx object which might hurt adoption on the app layer in case the tx changes again later on. Therefore, prefer keeping frames+." with 💯
- danceratopz
Reacted to "We should be mindful..." with 💯
- soispoke
A réagi à "We should be mindf..." avec 💯
- Csaba
Replying to "2 or 5, both are jus..." Actually 3 or 5, I’ve misread the chart
- Rimeeeeee
Reacted to "We should be mindf..." with 💯
- Guru
Reacted to "We should be mindful..." with 💯
- Kevaundray Wedderburn
If we dont want to reduce the scope, then I would push for prioritising the PQ EIPs
- Kevaundray Wedderburn
Yeah exactly
- Mislav | Ethlabs
Replying to "We should be mindf..." on that i would add that even this scope isn't yet enough for wallets to adopt frames in good time.
- prior
- Kevaundray Wedderburn
Replying to "If we dont want to r..." And once implemented, we could consider just shipping the PQ EIPs
- Barnabas
Replying to "We should be mindful..." what scope would be ideal for wallets?
- Francesco D'Amato
Reacted to "We should be mindful with considering Frames without the add-ons. They would change the tx object which might hurt adoption on the app layer in case the tx changes again later on. Therefore, prefer keeping frames+." with 👍
- felipe
Reacted to "He’s killing me over here 😂" with 😅
- Francesco D'Amato
Replying to "I'd like to understa..." The main cases of contracts broken due to try/catch remain broken
- Mislav | Ethlabs
i would advise against freezing! the product/ux paths for wallet adoption of pq haven't been fully thought out yet. a technically viable solution won't work if wallets don't adopt
- soispoke
A réagi à "what scope would b..." avec ➕
- Guru
Replying to "I'd like to understa..." ^ yes
- Toni Wahrstätter
Replying to "We should be mindful..." Frames+ (frames + recent roots + 2d nonces + tx assertions)
- Mislav | Ethlabs
Replying to "We should be mindf..." my pov is that we need to sort out the key management for a smooth pq transition. either something like 8130 in the default account oath or some similar solution.
- Mislav | Ethlabs
Replying to "We should be mindf..." also shared to me by wallets in conversations.
- soispoke
Répondre à "We should be mindf..." does it need an EIP?
- Toni Wahrstätter
Replying to "We should be mindful..." Agree yeah, this might not require tx object changes though? Also, everything while making it 8288 future-compatible ofc.
- spencer
Replying to "i would advise aga..." i think its freezing the EIPs and not the exact spec
- soispoke
A réagi à "Agree yeah, this m..." avec 👍
- lightclient
there is really no way we will have sig agg in hegota, but we should be future compatible for it
- Mislav | Ethlabs
Replying to "We should be mindf..." i think it could expand frames default path without a new eip.
- soispoke
A réagi à "there is really no..." avec 👍
- Barnabas
yeah agree, no point in shipping pq features if wallets can’t use those pq features.
- Iván | ethrex
Replying to "there is really no w..." We could, at builder level
- lightclient
Reacted to "We should be mindful with considering Frames without the add-ons. They would change the tx object which might hurt adoption on the app layer in case the tx changes again later on. Therefore, prefer keeping frames+." with 👍
- Vansh
Reacted to "there is really no..." with ➕
- Barnabas
We need to get wallet devs on the devnets asap
- felipe
Reacted to "there is really no..." with 👍
- Mislav | Ethlabs
Replying to "We should be mindf..." or could be an eip itself
- Iván | ethrex
Then care about mempool aggregation later because it doesn’t need a hard fork
- lightclient
Replying to "there is really no way we will have sig agg in hegota, but we should be future compatible for it" it’s not ready at all
- lightclient
Replying to "there is really no way we will have sig agg in hegota, but we should be future compatible for it" even the aggregation technique
- Toni Wahrstätter
Replying to "We should be mindful..." I see yeah that's nice. It would need a fork but not a tx obj change. That's great
- lightclient
Replying to "there is really no way we will have sig agg in hegota, but we should be future compatible for it" not even considering who does it
- soispoke
Répondre à "We should be mindf..." ok yeah feels like an important one to figure out quite fast
- Ben Adams
Should still go through the remaining PFI items even if its to officially DFI them
- Mislav | Ethlabs
Replying to "We need to get wal..." +1
- lightclient
Reacted to "Should still go through the remaining PFI items even if its to officially DFI them" with 👍
- marc | wolovim
Reacted to "We need to get wallet devs on the devnets asap" with 💙
- Barnabas
Reacted to "Should still go through the remaining PFI items even if its to officially DFI them" with 👍
- soispoke
A réagi à "Should still go th..." avec 👍
- Mario Vega
Reacted to "We need to get wallet devs on the devnets asap" with 💙
- Potuz
Reacted to "We need to get walle..." with 💙
- Iván | ethrex
Replying to "there is really no w..." I think it could be ready for Hegota
- Kevaundray Wedderburn
Brief analysis was here if folks are interested: https://github.com/kevaundray/pq-chain-analysis/tree/main (Also posted in discord)
- Nico
Reacted to "We need to get wal..." with 💙
- lightclient
Replying to "there is really no way we will have sig agg in hegota, but we should be future compatible for it" no it can’t
- Iván | ethrex
Replying to "there is really no w..." Until now there were just a pair of hands working on it
- Francesco D'Amato
Reacted to "I see yeah that's nice. It would need a fork but not a tx obj change. That's great" with 👍
- Nico
Replying to "We need to get wal..." We will finish the leanSPHINCS design in the next few weeks
- Mislav | Ethlabs
Replying to "i would advise aga..." yupyup. depending on how key mgmt is implemented it can be just a PR on 8141
- jonny (geth)
Reacted to "This could be a good compromise, if the reservoir never gets drained everything is good. Only when it drains problems come up. At least gas turns into monotonic decreasing again" with 👍
- Barnabas
Replying to "Should still go thro..." only 3 more left 8368, 8372, 7907
- lightclient
Replying to "there is really no way we will have sig agg in hegota, but we should be future compatible for it" it needs to be ready in next 2-3 months. we literally haven’t even finalized sphincs
- Mislav | Ethlabs
Replying to "i would advise aga..." doesn't have to be a separate eip
- Dragan Rakita
Reacted to "Brief analysis was here if folks are interested: https://github.com/kevaundray/pq-chain-analysis/tree/main (Also posted in discord)" with 👍
- Nico
Replying to "We need to get wal..." and get the on click upgrade for EOAs as soon as the devnets are up
- lightclient
Reacted to "We need to get wallet devs on the devnets asap" with 💙
- Iván | ethrex
Replying to "there is really no w..." Tom isn’t here to give more details on the status, but as I heard in the first PQ CL Breakout it has pretty good progress
- Mislav | Ethlabs
Replying to "We should be mindf..." happy to share my view (still wip)
- Iván | ethrex
Replying to "there is really no w..." Also de FV for the aggregation verifier is almost done
- Dragan Rakita
Reacted to "We need to get wallet devs on the devnets asap" with 💙
- Justin Florentine (Besu)
EL ONLY FORK BABBBBY!!!!!
- Guru
Reacted to "yupyup. depending on..." with 👍
- Mario Vega
Reacted to "EL ONLY FORK BABBBBY!!!!!" with 🔥
- soispoke
+1 Lukasz
- felipe
Reacted to "We should be mindful with considering Frames without the add-ons. They would change the tx object which might hurt adoption on the app layer in case the tx changes again later on. Therefore, prefer keeping frames+." with 💯
- Barnabas
we could ship EL and CL fork separate, if we rip out FOCIL
- spencer
CL would add pre-reg?
- Stefan Starflinger
Reacted to "EL ONLY FORK BABBBBY!!!!!" with 💪
- Gary Schulte
Reacted to "EL ONLY FORK BABBBBY!!!!!" with 🔥
- Caspar Schwarz-Schilling
Reacted to "EL ONLY FORK BABBBBY!!!!!" with 🔥
- Kevaundray Wedderburn
Reacted to "EL ONLY FORK BABBBBY!!!!!" with 🔥
- Justin Florentine (Besu)
Replying to "we could ship EL and..." yes but don't rip out FOCIL
- Vansh
Reacted to "EL ONLY FORK BABBB..." with 🔥
- soispoke
A réagi à "yes but don't rip ..." avec ➕
- Iván | ethrex
Reacted to "EL ONLY FORK BABBBBY!!!!!" with 🔥
- Iván | ethrex
Reacted to "EL ONLY FORK BABBBBY!!!!!" with 💪
- Barnabas
Replying to "we could ship EL and..." if we keep focil, its gonna be more of an EL issue imo. FOCIL + Frames
- Dragan Rakita
Reacted to "EL ONLY FORK BABBBBY!!!!!" with 😂
- Mislav | Ethlabs
Replying to "We need to get wal..." is leanSPHINCS in this case in the protocol or a verifier contract?
- soispoke
Répondre à "we could ship EL a..." wdym by EL issue?
- Toni Wahrstätter
Shipping frames early, which is the first step for users to start the pq migration, might be worth reducing some scope.
- Iván | ethrex
Yes, we agree with this in our team. The main advantage of doing this is offering users the possibility to start migrating asap
- Rimeeeeee
Reacted to "EL ONLY FORK BABBB..." with 🔥
- Kevaundray Wedderburn
Reacted to "Shipping frames early, which is the first step for users to start the pq migration, might be worth reducing some scope." with 👍
- spencer
Lets do 3 + CL pre-reg :D
- Ansgar Dietrichs
lol I was getting bullied for saying July as aggressive target, just to now hear March :-)
- Iván | ethrex
Reacted to "Lets do 3 + CL pre-reg :D" with 👍
- jonny (geth)
Replying to "I'd like to understa..." which try/catch cases do you mean? if it's fixed budgets, those break from the repricing under the current spec too. so isn't the question whether refilling the reservoir is better than what we have now?
- Toni Wahrstätter
Reacted to "lol I was getting bullied for saying July as aggressive target, just to now hear March :-)" with 🚢
- Vansh
Reacted to "lol I was getting ..." with 😂
- Guru
Reacted to "Shipping frames earl..." with 💯
- Iván | ethrex
Reacted to "lol I was getting bullied for saying July as aggressive target, just to now hear March :-)" with 🚢
- soispoke
march doesn't feel realistic
- Dustin
July was aggressive. March is sheer fantasy
- Caspar Schwarz-Schilling
Reacted to "lol I was getting bullied for saying July as aggressive target, just to now hear March :-)" with 🚢
- Potuz
FWIW prereg only CL can be done in one week
- Justin Traglia
Reacted to "July was aggressiv..." with 💯
- Iván | ethrex
Replying to "lol I was getting bu..." We weren’t prepared then 😄
- Łukasz Rozmej
Replying to "lol I was getting bu..." What is The Point of reducing scope if we are not aggressive
- cayman
Reacted to "FWIW prereg only CL ..." with 🛳️
- soispoke
A réagi à "FWIW prereg only C..." avec 👀
- marc | wolovim
Reacted to "FWIW prereg only CL can be done in one week" with 🛳️
- Łukasz Rozmej
Reacted to "I'd like to understand why we can't refund to state gas" with 👍
- Iván | ethrex
Reacted to "FWIW prereg only CL can be done in one week" with 🛳️
- felipe
Reacted to "yeah, refunding into the reservoir but still during execution only solves a small subset of issues. the main issue around that invariant was the fact that it happens during exec and not after the tx is done" with 👍
- Iván | ethrex
Replying to "FWIW prereg only CL ..." Let’s gooo
- Barnabas
Replying to "Lets do 3 + CL pre-r..." cl pre reg doesn’t require hardfork
- spencer
What is the realistic CL PQ ship date?
- soispoke
Répondre à "FWIW prereg only C..." 1 week?
- Dustin
Replying to "FWIW prereg only C..." No hardfork can ship (including testing, testnets, etc) in one week
- Yann Vonlanthen
Reacted to "FWIW prereg only CL can be done in one week" with 🛳️
- Csaba
Replying to "there is really no w..." There is good progress, but the design space is still being explored. It does not make too much sense to rush it IMHO because it is a paradigm change. Even if technically we could have a version of it working, it might just not be the right approach. Also the FV is good, but until the underlying algo is final it does not give much assurance.
- Dustin
Replying to "FWIW prereg only C..." people want a month just do do their defi/L2/etc preparation
- soispoke
Répondre à "What is the realis..." depends of what we even mean by CL PQ
- Toni Wahrstätter
Reacted to "there is really no way we will have sig agg in hegota, but we should be future compatible for it" with 👍
- Iván | ethrex
Replying to "FWIW prereg only CL ..." If Potuz means just deploying the contract then it can be done as soon as the contract is ready and audited
- Łukasz Rozmej
If it isn't SFI'd it is not removed
- Guru
Reacted to "happy to share my vi..." with 👍
- Iván | ethrex
Replying to "FWIW prereg only CL ..." You just need the hard fork later to consume the pre-registrations
- soispoke
Répondre à "What is the realis..." Gasper/DC/aggregation, etc...
- spencer
Reacted to "Gasper/DC/aggregat..." with 😅
- Ansgar Dietrichs
Replying to "lol I was getting bu..." I think July is aggressive (and realistic!) fwiw
- Ignacio
Replying to "FWIW prereg only C..." With PoP that means freeze of leanXMSS design.
- Mashal @ Octant
Reacted to "lol I was getting bullied for saying July as aggressive target, just to now hear March :-)" with 😂
- Iván | ethrex
Reacted to "With PoP that means freeze of leanXMSS design." with 👍
- Mashal @ Octant
Reacted to "lol I was getting bullied for saying July as aggressive target, just to now hear March :-)" with 🚢
- spencer
Replying to "What is the realis..." fair :D
- Iván | ethrex
Replying to "FWIW prereg only CL ..." That’s right
- Dragan Rakita
GTA VI is going to delay things a lot 🙂
- Joss
Reacted to "GTA VI is going to..." with 😂
- Barnabas
Reacted to "GTA VI is going to delay things a lot 🙂" with 😂
- Louis
Reacted to "GTA VI is going to delay things a lot 🙂" with 😂
- Łukasz Rozmej
Replying to "lol I was getting bu..." July should be realistic with current scope
- lightclient
Replying to "there is really no way we will have sig agg in hegota, but we should be future compatible for it" it’s just too fast to have the required confidence in the scheme. there is also no reason to push it too quickly, we don’t need aggregation on the EL
- Rimeeeeee
Reacted to "GTA VI is going to..." with 😂
- marc | wolovim
Reacted to "GTA VI is going to delay things a lot 🙂" with 🫠
- Mega | ethlambda
Reacted to "FWIW prereg only CL can be done in one week" with 🛳️
- spencer
Reacted to "GTA VI is going to..." with 😂
- Vansh
Reacted to "GTA VI is going to..." with 😂
- soispoke
A réagi à "GTA VI is going to..." avec 😂
- Guru
Reacted to "GTA VI is going to d..." with 😂
- Ansgar Dietrichs
Reacted to "GTA VI is going to delay things a lot 🙂" with 😂
- Justin Traglia
Reacted to "GTA VI is going to..." with 😂
- Mashal @ Octant
Reacted to "GTA VI is going to delay things a lot 🙂" with 😂
- spencer
Replying to "GTA VI is going to..." GTA VI came before PQ :O
- Dustin
Replying to "FWIW prereg only C..." sure the contract can be deployed
- Ansgar Dietrichs
Replying to "GTA VI is going to d..." as long as we are pq ready before GTA VII
- Dragan Rakita
Reacted to "GTA VI came before PQ :O" with 😂
- Kevaundray Wedderburn
Reacted to "GTA VI is going to delay things a lot 🙂" with 😂
- Iván | ethrex
Reacted to "There is good progress, but the design space is still being explored. It does not make too much sense to rush it IMHO because it is a paradigm change. Even if technically we could have a version of it working, it might just not be the right approach. Also the FV is good, but until the underlying algo is final it does not give much assurance." with 👍
- Dragan Rakita
Reacted to "as long as we are pq ready before GTA VII" with 💯
- spencer
Reacted to "as long as we are ..." with 😂
- Mario Vega
Reacted to "GTA VI is going to delay things a lot 🙂" with 😂
- Mario Vega
Reacted to "GTA VI came before PQ :O" with 😂
- Fredrik
Replying to "lol I was getting bu..." That would mean the deadline for all clients to be feature complete is in April
- Ryan McPeck
“Enshrined leanSPHINCS (in some form)” - have we considered enshrining ML instead? A lot more tooling / adoption
- Ben Edgington
Replying to "What is the realistic CL PQ ship date?" Mid-2028. Don't quote me.
- Potuz
Replying to "FWIW prereg only CL ..." I mean more than deploying the contract. the CL code to consume the data can be coded and be ready and tested in less than a week
- Łukasz Rozmej
Replying to "lol I was getting bu..." We can do January
- lightclient
Reacted to "July was aggressive. March is sheer fantasy" with 💯
- felipe
Reacted to "GTA VI is going to delay things a lot 🙂" with 😂
- Iván | ethrex
Replying to "there is really no w..." I agree that we could ship without it, though. Just saying that it could be possible if it’s ready in time
- Ryan McPeck
Replying to "“Enshrined leanSPHIN..." Also we need to accelerate PQ as much as possible.
- Barnabas
Greg, you can’t count on anything, the world moves fast, we need to make sure we are secure post quantum computers. Nobody expected that AI gonna explode the speed it did. We need to keep up or get left behind (get exploited).
- Iván | ethrex
Reacted to "GTA VI is going to delay things a lot 🙂" with 😂
- lightclient
Reacted to "We can do January " with 😂
- Łukasz Rozmej
Replying to "lol I was getting bu..." Testing is bigger question
- Iván | ethrex
Reacted to "as long as we are pq ready before GTA VII" with 💯
- Justin Florentine (Besu)
Replying to "“Enshrined leanSPHIN..." yes, ML was DFI'd crypto experts prefer the hash based approach
- Iván | ethrex
Reacted to "as long as we are pq ready before GTA VII" with 😂
- Ansgar Dietrichs
Reacted to "That would mean the deadline for all clients to be feature complete is in April" with 👍
- Ryan McPeck
Reacted to "yes, ML was DFI'd crypto experts prefer the hash based approach" with 👍
- Iván | ethrex
Reacted to "I mean more than deploying the contract. the CL code to consume the data can be coded and be ready and tested in less than a week" with 🔥
- lightclient
i feel like we should also have ml-dsa
- Csaba
Replying to "there is really no w..." I think it is important that we work it out, and get confidence it our design (that’s why I’m working on it :-) ).
- soispoke
we can still reduce the scope to what seems to be essential and locked in, with the understanding that the rest of scoping also depends on coordination with CL
- spencer
we can all just use ambire wallet :)
- Nico
Replying to "“Enshrined leanSP..." Yes we did work on lattice and we decided not to move forward
- Łukasz Rozmej
Reacted to "GTA VI is going to delay things a lot 🙂" with 😂
- lightclient
Replying to "“Enshrined leanSPHINCS (in some form)” - have we considered enshrining ML instead? A lot more tooling / adoption" but there is no reason to also support ml-dsa? there is so much tooling and support for it
- Iván | ethrex
Replying to "lol I was getting bu..." Yes I think it’s clear that client development is not the bottleneck in this case
- Ryan McPeck
Replying to "“Enshrined leanSPHIN..." @Nico is there a writeup on that? On why that decision was made. Can you send it to me
- Iván | ethrex
Replying to "lol I was getting bu..." Testing needs time
- Ryan McPeck
Reacted to "but there is no reason to also support ml-dsa? there is so much tooling and support for it" with 💯
- Iván | ethrex
Replying to "lol I was getting bu..." And we need to engage with the community on breaking changes
- felipe
Reacted to "i feel like we sho..." with 👍
- soispoke
people that would want to use lattices could use them with leanPCS down the road
- felipe
Reacted to "people that would ..." with 👍
- Barnabas
+1 potuz, community is gonna be the slowest by far!
- soispoke
Répondre à "people that would ..." but having a standard (hash based) seems like a sensible option
- Iván | ethrex
Replying to "there is really no w..." Agree 💪
- Iván | ethrex
Reacted to "I think it is important that we work it out, and get confidence it our design (that’s why I’m working on it :-) )." with 🔥
- Mislav | Ethlabs
definitely fine with not enshrining via precompiles. evm validation is good imo.
- Justin Florentine (Besu)
Reacted to "i feel like we should also have ml-dsa" with 👍
- Dragan Rakita
Reacted to "+1 potuz, community is gonna be the slowest by far!" with ➕
- Iván | ethrex
Reacted to "+1 potuz, community is gonna be the slowest by far!" with ➕
- Iván | ethrex
Reacted to "people that would want to use lattices could use them with leanPCS down the road" with 👍
- Mislav | Ethlabs
Replying to "definitely fine wi..." if the only enshrinement is leanSPHINCS being cheaper, that's fine. as long as there isn't a bigger dependency in frames
- Nico
Replying to "“Enshrined leanSP..." Sure TLDR we allow custom verifiers so people can do whatever they want. Yet we believe that hash based sig is the most conservative option
- Potuz
Replying to "What is the realisti..." I think targeting EOY 2027 so that we actually ship mid 2028 is reallistic
- spencer
Reacted to "EL ONLY FORK BABBB..." with 🔥
- Justin Florentine (Besu)
i would reconsider the DFI decision on MLDSA precompiles
- Potuz
SPHINCS does not add new crypto assumptions (we already have Keccack), while Lattices would.
- soispoke
Répondre à "What is the realis..." to me it sounds like we could ship some user facing features in the meantime :)
- soispoke
A réagi à "SPHINCS does not a..." avec 👍
- Dragan Rakita
Reacted to "SPHINCS does not add new crypto assumptions (we already have Keccack), while Lattices would." with 👍
- Iván | ethrex
Reacted to "SPHINCS does not add new crypto assumptions (we already have Keccack), while Lattices would." with 👍
- cloud1992
Reacted to "I think targeting EOY 2027 so that we actually ship mid 2028 is reallistic" with ❤️
- Potuz
Replying to "What is the realisti..." nah you want FOCIL :)
- Iván | ethrex
Replying to "SPHINCS does not add..." And this is a very important distinction
- Toni Wahrstätter
What timeline for shipping frames are we targeting, and would it make sense to think about the fork scope knowing how much time we want to give outselves. This might impact dropping eips vs not.
- Mislav | Ethlabs
+1 on what Matt is saying
- felipe
Reacted to "SPHINCS does not a..." with 👍
- Mega | ethlambda
Reacted to "SPHINCS does not add new crypto assumptions (we already have Keccack), while Lattices would." with 👍
- Dragan Rakita
Reacted to "+1 on what Matt is saying" with ➕
- Justin Florentine (Besu)
i think having a second tier option in frames is still more important than other aspects of frames
- Potuz
Reacted to "+1 on what Matt is s..." with ➕
- Justin Florentine (Besu)
crypto agility kinda needs more than 1 option
- Iván | ethrex
Let’s just stop discussing lattices as an option pls
- lightclient
Replying to "Let’s just stop discussing lattices as an option pls" why??
- Alexander Hicks
If the concern is standards there is also a sphincs standard...
- lightclient
Replying to "Let’s just stop discussing lattices as an option pls" everyone else is supporting it
- soispoke
Répondre à "If the concern is ..." doesn't it have crazy conservative parameters?
- marc | wolovim
@Nico relay in chat maybe - lost your audio
- Mislav | Ethlabs
Replying to "Let’s just stop d..." i don't understand this? ml-dsa is native in many hardware devices we want users to use as signers
- Potuz
I agree with Matt, but I think there's a difference between scheduling for Hegota or not... I think whatever the first class signature scheme is in L1 should be enshrined in Hegota, and everything else could come later
- Dragan Rakita
Reacted to "@Nico relay in chat maybe - lost your audio" with 👍
- felipe
Reacted to "I agree with Matt,..." with ➕
- Nico
Sorry I crashed
- Iván | ethrex
Because in distinction to SPHINCS, they are based on assumptions that could be broken in the future
- Kevaundray Wedderburn
@Mislav | Ethlabs do you think 1M gas is too much? Or did the entities you spoke with indicate this?
- Justin Florentine (Besu)
Replying to "If the concern is st..." sphincs contains multitudes
- Csaba
Reacted to "GTA VI is going to delay things a lot 🙂" with 😂
- Mislav | Ethlabs
Replying to "@Mislav | Ethlabs ..." yes expense was mentioned as a blocker
- Csaba
Reacted to "as long as we are pq ready before GTA VII" with 💯
- soispoke
Répondre à "If the concern is ..." multiple sphincs standards?
- Iván | ethrex
Please give the floor to Antonio
- Dragan Rakita
Reacted to "yes expense was mentioned as a blocker" with 👀
- lightclient
Replying to "If the concern is standards there is also a sphincs standard..." yeah but it isn’t the standard we are using
- Kevaundray Wedderburn
Replying to "@Mislav | Ethlabs do..." Did they say what cost would make sense for them?
- Iván | ethrex
Reacted to "If the concern is standards there is also a sphincs standard..." with 👍
- Barnabas
Replying to "Let’s just stop disc..." btw whats wrong with lattices?
- Mislav | Ethlabs
This is not about the underlying cryptography at all folks. It's about product and UX. It doesn't have to be lattice or hash at all.
- Marc
Reacted to "+1 on what Matt is saying" with ➕
- lightclient
it’s not just about hash, it’s also about what is industry standard and what isn’t
- Mislav | Ethlabs
Replying to "This is not about ..." It's about giving the user ability to add a signing scheme which their iPhone can sign
- Potuz
Now I want lattices... this all thing of AI will soon announce the end of the world is landing badly with me
- Marc
Reacted to "It's about giving the user ability to add a signing scheme which their iPhone can sign" with ➕
- Dragan Rakita
Reacted to "It's about giving the user ability to add a signing scheme which their iPhone can sign" with ➕
- Vansh
Reacted to "It's about giving ..." with ➕
- Kevaundray Wedderburn
Replying to "@Mislav | Ethlabs do..." I think we won’t be able to get as cheap as ecdsa anymore, but im interested in what they think a reasonable cost will be as this will also be useful for SPHINCs
- Nico
Replying to "This is not about th..." Will be doable is already doable today
- Mislav | Ethlabs
Replying to "This is not about ..." if lattices are unsafe I am sure Apple will react in time
- Nico
Replying to "This is not about th..." It's just a gas cost question
- Nico
Replying to "This is not about th..." if lattices are unsafe I am sure Apple will react in time
- soispoke
Répondre à "This is not about ..." agreed, but industry standards can change as well
- Nico
Replying to "This is not about th..." Open for a bet Mislav 😁
- Will Corcoran
Replying to "This is not about th..." Bookmark this
- Mislav | Ethlabs
Reacted to "Open for a bet Mis..." with 😄
- Mislav | Ethlabs
Replying to "This is not about ..." hahah
- soispoke
A réagi à "Open for a bet Mis..." avec 😄
- Mislav | Ethlabs
Replying to "This is not about ..." yeah I am not even requesting a precompile
- Francesco D'Amato
Replying to "Now I want lattices...." In a few weeks AI will announce Gasper is a bit broken
- Potuz
Replying to "Now I want lattices...." looool
- soispoke
A réagi à "In a few weeks AI ..." avec 😂
- Mislav | Ethlabs
Replying to "This is not about ..." Just that Frames don't prefer any paths for SPHICS
- Derek Chiang | Ethlabs
To support ML-DSA cheaply we’d need to ship leanPCS right?
- Mislav | Ethlabs
Replying to "This is not about ..." if SPHINCS is cheaper bcs it has a precompile and users make that tradeoff thats fine imo
- Nico
Replying to "This is not about th..." It is doable with both Falcon (around 500k) and Dilithium (around 1M gas)
- felipe
Reacted to "In a few weeks AI ..." with 😅
- Nico
Reacted to "if SPHINCS is cheaper bcs it has a precompile and users make that tradeoff thats fine imo" with 👍
- soispoke
Répondre à "To support ML-DSA ..." yes
- Nico
Replying to "This is not about th..." Eactly
- soispoke
Répondre à "To support ML-DSA ..." well or a precompile
- Potuz
Reacted to "This is not about th..." with 👍
- Mislav | Ethlabs
Replying to "This is not about ..." Yeah but afaik Frames as-is have enumerated signing schemes in the design?
- Potuz
Reacted to "It's about giving th..." with ➕
- Ben Edgington
Reacted to "In a few weeks AI will announce Gasper is a bit broken" with 😂
- Greg Colvin
Replying to "Greg, you can’t coun..." Yes, @Barnabas . But subroutines are important to the VM -- yes, it is not fit for purpose, 80 years behind the relavant computer science, missing a feature provided by most all of our competiitos -- and versions of them have been getting delayed for 10 years now as champions move on or get fired. This version is a tiny change. I am the last champion alive. Please, please let's get this done.
- spencer
ctrl+alt+DFI
- Mislav | Ethlabs
Replying to "This is not about ..." With an ARBITRARY left as a catchall
- Greg K | Lido
Reacted to "well or a precompile" with 👍
- Nico
Replying to "This is not about th..." https://github.com/ZKNoxHQ/ETHDILITHIUM https://github.com/ZKNoxHQ/ETHFALCON Could go in arbitrary yes
- soispoke
A réagi à "nah you want FOCIL..." avec 😁
- Ignacio
Replying to "ctrl+alt+DFI" hotkey added
- Iván | ethrex
Reacted to "hotkey added" with ➕
- Derek Chiang | Ethlabs
Replying to "To support ML-DSA ch..." How much cheaper (or more expensive) would ML-DSA be with leanPCS+aggregation compared to with a precompile? I don’t have a good mental model for this
- Iván | ethrex
There be likely no time for the introduction I wanted to do, so I’ll post it in the chat: I’m working on for a minimal solution to having native leanSPHINCS signatures on top of frame transactions in Hegota, which is extensible to EIP-8288 (i.e. compatible with leanPCS and later aggregation), ethereum/EIPs#12672. It is currently in development, during the rest of the week I’ll be completing the specs, but happy to have some early discussions in the PR. I’m aiming to proposing it for Hegota next ACDE
- Fara Woolf
Reacted to "Open for a bet Mislav 😁" with 😄
- Derek Chiang | Ethlabs
Reacted to "It's about giving the user ability to add a signing scheme which their iPhone can sign" with ➕
- Nico
Good news regarding the frame breakout we figured out a way to do multisig with PQ
- Iván | ethrex
Reacted to "Good news regarding the frame breakout we figured out a way to do multisig with PQ" with 🔥
- Mario Vega
Reacted to "Good news regarding the frame breakout we figured out a way to do multisig with PQ" with 🔥
- felipe
Reacted to "Good news regardin..." with 🔥
- spencer
Reacted to "Good news regardin..." with 🔥
- Rimeeeeee
Reacted to "Good news regardin..." with 🔥
- Pablo | Lambda
Reacted to "Good news regarding the frame breakout we figured out a way to do multisig with PQ" with 🔥
- Mislav | Ethlabs
Replying to "This is not about ..." Yeah my take: Ideally we just have one path (which today is ARBITRARY) and everything goes through that.
- Nico
Replying to "Good news regarding ..." TLDR : one owner sends the frame transaction from their own PQ account and the safe checks the other owner inside execTransaction through 1271
- Mislav | Ethlabs
Reacted to "Good news regardin..." with 🔥
- Potuz
Reacted to "Good news regarding ..." with 🔥
- Nico
Replying to "Good news regarding ..." So that way only the sender SIG is verified in the validation prefix. No changes required to Safe nor to EIP 8141
- Kevaundray Wedderburn
Replying to "To support ML-DSA ch..." These are orthogonal I’d say. We can just have leanPCS and not have leanSPHINCs enshrined but it means that users will now need to make a stark proof for leanSPHINCs/ml-dsa etc
- Nico
Replying to "Good news regarding ..." https://github.com/nconsigny/SPHINCS-/blob/main/docs/safe-pq-owners.md
- Barnabas
we should only do it if we end up shipping focil
- Barnabas
otherwise lets keep it separate
- Iván | ethrex
The draft proposal I mentioned https://github.com/ethereum/EIPs/pull/12672
- Derek Chiang | Ethlabs
Replying to "To support ML-DSA ch..." Technically users can submit raw signatures and mempool nodes can wrap them in STARKs right?
- Stefan Starflinger
yeah, FOCIL was the main reason to merge them early
- soispoke
Répondre à "yeah, FOCIL was th..." is :)
- Kevaundray Wedderburn
Replying to "To support ML-DSA ch..." Yep, so in this world, we would have leanPCS with frames and then mempool nodes would have a canonical contract that uses leanPCS to make a stark verifier that can be used to verify those raw signatures
- soispoke
keep them separate until decisions are made on the CL side
- Iván | ethrex
Reacted to "keep them separate until decisions are made on the CL side" with 👍
- Iván | ethrex
Reacted to "keep them separate until decisions are made on the CL side" with ➕
- Barnabas
Reacted to "keep them separate until decisions are made on the CL side" with 👍
- Łukasz Rozmej
Wat would we consider for the next batch?
- Barnabas
can we do bogota-devnet-3 scoping?
- Mislav | Ethlabs
To summarize my view on the mutliple schemes: I don't have enough knowledge on cryptography to say whether or not lattices will have issues. I am inclined to trust the mathematicians here. But even if hashes are the solution, we have no guarantee that Apple, Google, and the rest of the relevant OS or hardware producers will choose the same hash based signatures as we did. Ethereum UX suffered for years due to the choice of the k1 curve and all of us who worked on product/tech for smart accounts have PTSD from it. The end result was big centralization where Privy/Dynamic and others took passkey sigs on their server and served the k1 sigs back to the chain - because the native solution was tricky/expensive. That's why I say I don't consider it a cryptography decision, but a product/ux decision.
- Mislav | Ethlabs
Replying to "To summarize my vi..." Frames are on a good path here btw, not a critique of work already done in any way.
- soispoke
Répondre à "To summarize my vi..." right but I don't think we should wait and see what Apple and Google will chose to make decisions ensrhine one standard that is best for Ethereum, let anyone use any signature scheme they want later (leanPCS)
- Derek Chiang | Ethlabs
Replying to "To support ML-DSA ch..." Right, it sounds like we are more or less convinced that with leanPCS we can support ML-DSA cheaply. But I still don’t know if we can support ML-DSA cheaply enough with precompiles alone
- Iván | ethrex
Testing team opinion on adding EIPs/freezing would be useful maybe for some client teams to decide
- Kevaundray Wedderburn
Replying to "To support ML-DSA ch..." I’d need to think a bit more about the UX of this for app developers. From the user perspective it looks good
- Mario Vega
Reacted to "Testing team opinion on adding EIPs/freezing would be useful maybe for some client teams to decide" with 👍
- Łukasz Rozmej
Reacted to "Testing team opinion on adding EIPs/freezing would be useful maybe for some client teams to decide" with 👍
- Mislav | Ethlabs
Replying to "To summarize my vi..." I agree with the first part. Second one depends on how much later it is.
- Nico
Reacted to "right but I don't think we should wait and see what Apple and Google will chose to make decisions ensrhine one standard that is best for Ethereum, let anyone use any signature scheme they want later (leanPCS)" with 👍
- Potuz
recall that we'll need a lot of attention of the testing team once we deploy G on Hoodi and before deploying it on mainnet
- Iván | ethrex
Reacted to "recall that we'll need a lot of attention of the testing team once we deploy G on Hoodi and before deploying it on mainnet" with 👍
- Mislav | Ethlabs
Replying to "To summarize my vi..." Maybe 8288 light in the block header is enough for now?
- Kevaundray Wedderburn
Replying to "To support ML-DSA ch..." If you mean an EVM precompile, the main cost would basically be calldata right. So for ml-dsa-44 thats about 2KB
- Mislav | Ethlabs
Replying to "To summarize my vi..." And we can have centralized aggregation for a while
- Ben Adams
We shouldn't let wanting to do PQ make us timid on improving Ethereum in other respects
- Ryan McPeck
Replying to "To support ML-DSA ch..." @Derek Chiang | Ethlabs are you thinking leanPCS via 8288 aggregation? My concern with that is timeline. Imo we need to accelerate PQ as much as we can
- soispoke
Répondre à "To summarize my vi..." ah well, I'm definitely all for leanPCS asap but definitely against centralized aggregation
- Iván | ethrex
Reacted to "We shouldn't let wanting to do PQ make us timid on improving Ethereum in other respects" with 👍
- Łukasz Rozmej
Reacted to "We shouldn't let wanting to do PQ make us timid on improving Ethereum in other respects" with 👍
- Derek Chiang | Ethlabs
Replying to "To support ML-DSA ch..." Yeah I’m just trying to figure out if there’s a realistic path to supporting ML-DSA cheaply in a short timeline
- marc | wolovim
Frames has some spec change proposals, so there’s changes to potentially implement without adding new eips
- soispoke
Répondre à "To summarize my vi..." I don't think losing all CR is worth it
- Mislav | Ethlabs
Replying to "To summarize my vi..." Yeah I would love to go deeper into this. Some tradeoff space exists for sure, my fear is we wait for leanPCS for a long time and there is UX confusion when transitioning
- Derek Chiang | Ethlabs
Replying to "To support ML-DSA ch..." Because if it depends on leanPCS, then it sounds like it won’t be for a while
- Barnabas
maybe we can ship some cleanup EIPs?
- Jihoon
Reacted to "We shouldn't let wanting to do PQ make us timid on improving Ethereum in other respects" with 👍
- Ben Adams
Reacted to "maybe we can ship some cleanup EIPs?" with 👍
- Barnabas
Replying to "maybe we can ship so..." removing bloom filter, remove storage clear refund,
- Iván | ethrex
Reacted to "maybe we can ship some cleanup EIPs?" with 👍
- Ben Adams
Reacted to "removing bloom filter, remove storage clear refund," with 💜
- Toni Wahrstaetter
Keeping devnets separate keeps the path towards EL-only fork open. Not adding more allows to move and ship faster, but not against adding things that simplify the protocol or harden it.
- soispoke
Répondre à "To summarize my vi..." yeah, I really think leanPCS is less of a huge problem once aggregation and leanSPHINCS ship but happy to also discuss more
- Stefan Starflinger
Replying to "maybe we can ship some cleanup EIPs?" don't these EIPs need to from CFI to SFI for us to include them?
- Mislav | Ethlabs
Replying to "To summarize my vi..." People might start aggregating themselves and posting atomic 4337 bundles with aggregated proofs and who knows what. Lesson is if we don't offer it, more centralized solutions will. That's my core fear. Though I agree there are tradeoffs and its not as simple.
- Kevaundray Wedderburn
Replying to "To support ML-DSA ch..." Yeah this is my understanding too. Not advocating for it, but at the very least we would need aggregation at the block level
- Potuz
I would stronly prefer to keep devnets separated
- Iván | ethrex
Reacted to "Keeping devnets separate keeps the path towards EL-only fork open. Not adding more allows to move and ship faster, but not against adding things that simplify the protocol or harden it." with 👍
- Barnabas
Reacted to "I would stronly prefer to keep devnets separated" with 💯
- Potuz
Reacted to "Keeping devnets sepa..." with 👍
- Barnabas
Replying to "I would stronly pref..." agreed
- Mislav | Ethlabs
Reacted to "yeah, I really thi..." with 👍
- Kevaundray Wedderburn
Replying to "To support ML-DSA ch..." Btw we did some experiments with ml-dsa here: https://github.com/eth-act/ml-dsa-fv
- Mislav | Ethlabs
Replying to "To summarize my vi..." Will reach out for sure
- Nico
Replying to "To support ML-DSA ch..." Is 50 cents cheap enough Derek ?
- Nico
Reacted to "Btw we did some experiments with ml-dsa here: https://github.com/eth-act/ml-dsa-fv" with ❤️
- soispoke
I don't think anyone is advocating for merging layers on devnets for now
- Justin Florentine (Besu)
Reacted to "I don't think anyone is advocating for merging layers on devnets for now" with 👍
- Iván | ethrex
Reacted to "I don't think anyone is advocating for merging layers on devnets for now" with 👍
- Toni Wahrstaetter
Replying to "I don't think anyone is advocating for merging layers on devnets for now" right, I guess the next question would be if we want to consider EL-only forks or not
- Derek Chiang | Ethlabs
Replying to "To support ML-DSA ch..." For mainnet i’d say so yeah. It’s not ideal but seems good enough for the short term
- Nico
Replying to "To support ML-DSA ch..." Right now ML DSA would be ~50 cents and at 200M gas it would be less
- Nico
Reacted to "For mainnet i’d say so yeah. It’s not ideal but seems good enough for the short term" with 👍
- Pablo | Lambda
Reacted to "I don't think anyone is advocating for merging layers on devnets for now" with 👍
- Derek Chiang | Ethlabs
Replying to "To support ML-DSA ch..." You mean with a precompile? Or with a normal contract?
- soispoke
feels pretty early to decide that, and kind of unnecessary to decide that early
- Nico
Replying to "To support ML-DSA ch..." Agreed long term it could / should be better with leanPCS
- Derek Chiang | Ethlabs
Reacted to "Btw we did some experiments with ml-dsa here: https://github.com/eth-act/ml-dsa-fv" with ❤️
- Justin Florentine (Besu)
Replying to "I don't think anyone..." it's always a tool in the toolbox but agree not worth planning around now
- Dragan Rakita
Shoutout to Mislav for joining the call and thinking about wallets/ux/product and pq
- Potuz
I +1 this, also it would help the CL if we fork the EL earlier cause we'll have some concrete data on SHPINCS impact on the P2P stack if people start sending those txs
- lightclient
Reacted to "Shoutout to Mislav for joining the call and thinking about wallets/ux/product and pq" with ❤️
- Mislav | Ethlabs
Reacted to "Shoutout to Mislav..." with ❤️
- Barnabas
Reacted to "I +1 this, also it would help the CL if we fork the EL earlier cause we'll have some concrete data on SHPINCS impact on the P2P stack if people start sending those txs" with 👍
- marc | wolovim
Reacted to "Shoutout to Mislav for joining the call and thinking about wallets/ux/product and pq" with ❤️
- Mislav | Ethlabs
Replying to "Shoutout to Mislav..." thank you!
- soispoke
A réagi à "Shoutout to Mislav..." avec ❤️
- Toni Wahrstaetter
Replying to "I don't think anyone is advocating for merging layers on devnets for now" it might affect when we can ship frames and start the pq migration
- Fabio Di Fabio
Reacted to "I +1 this, also it would help the CL if we fork the EL earlier cause we'll have some concrete data on SHPINCS impact on the P2P stack if people start sending those txs" with 👍
- Dragan Rakita
Reacted to "thank you!" with ♥️
- Potuz
no! that was never the case
- Iván | ethrex
Reacted to "Shoutout to Mislav for joining the call and thinking about wallets/ux/product and pq" with ❤️
- soispoke
A réagi à "Will reach out for..." avec ❤️
- marc | wolovim
No no current path requires stable devnet appearance before qualifying for SFI
- Trent Van Epps
Replying to "Shoutout to Mislav f..." Great work by Ethlabs 🫡
- Dragan Rakita
Reacted to "Great work by Ethlabs 🫡" with ♥️
- Nico
Replying to "To support ML-DSA ch..." With the verifiers yes https://github.com/ZKNoxHQ/ETHDILITHIUM
- Ben Adams
Reacted to "No no current path requires stable devnet appearance before qualifying for SFI" with 👍
- Vansh
Reacted to "Shoutout to Mislav..." with ❤️
- Stefan Starflinger
Replying to "No no current path requires stable devnet appearance before qualifying for SFI" when devnet if not CFI anymore?
- Barnabas
I’d like to see some progress, so lets do cleanup eips and updated frames spec for bogota devnet 3
- Derek Chiang | Ethlabs
Reacted to "With the verifiers yes https://github.com/ZKNoxHQ/ETHDILITHIUM" with 👍
- Barnabas
and I don’t think we need to wait for acdt for making this decision
- Mislav | Ethlabs
Reacted to "Great work by Ethl..." with ♥️
- Barnabas
no new devnet is a bad status quo imo
- Francesco D'Amato
Reacted to "Shoutout to Mislav for joining the call and thinking about wallets/ux/product and pq" with ❤️
- Ben Adams
Reacted to "no new devnet is a bad status quo imo" with 👍
- Ryan McPeck
Replying to "To support ML-DSA ch..." No 4m or 8m gas cost is too much
- marc | wolovim
Replying to "No no current path r..." Thats the intention behind the priority ordering exercise, but a lot has changed since we made this list before devcon: https://forkcast.org/upgrade/hegota/priority-order/
- spencer
We can all focus on hardening frames + PQ EIPs
- Nico
Replying to "To support ML-DSA ch..." 1.5M
- Derek Chiang | Ethlabs
Replying to "To support ML-DSA ch..." Speaking from a user’s perspective, it’d be good if the protocol can guarantee the security of popular signature types e.g. ML-DSA / SPHINCS, as opposed to having to trust contracts (FV-ed or not)
- Kevaundray Wedderburn
Replying to "To support ML-DSA ch..." @Ryan McPeck when you say too much, do you mean the absolute cost or are you referring to to the fact that the tx has a 16M gas limit and 1M takes up 1/16 of it?
- Nico
Replying to "To support ML-DSA ch..." Experimental: packed verifier (branch exp/packed-verifier) @Ryan McPeck
- Iván | ethrex
If they’re just cleanup EIPs (i.e. non-risky), I’m ok with that
- Derek Chiang | Ethlabs
Reacted to "1.5M" with 👍
- Ryan McPeck
Replying to "To support ML-DSA ch..." Too much for the user. If it is temporary, that is one thing. But in general that is too much of a cost to the user
- Kevaundray Wedderburn
Replying to "To support ML-DSA ch..." @Derek Chiang | Ethlabs does it make a difference if we enshrine an ml-dsa contract?
- Nico
Replying to "To support ML-DSA ch..." AI found optimisations that are correct so now the cost is 800k if you are OK to use keccak instead of shake and 1.6M if you want to stick to NIST
- Derek Chiang | Ethlabs
Replying to "To support ML-DSA ch..." @Ryan McPeck even at 1.5M gas / $0.5 ?
- Kevaundray Wedderburn
Replying to "To support ML-DSA ch..." Or do they want the guarantee that if it has a bug then the protocol will fork?
- marc | wolovim
Optional ACDT todo: teams make a formal proposal for next devnet scope so folks have concrete options to consider
At a glance
Detailed notes
Hegota Scope
- •Scope options debated (1–5); rough landing zone between option 3 (core frames bundle only) and 4 (frames + select EIPs); option 1 (enshrine PQ, drop frames) has zero support00:51:33
- •Besu/ethrex lean toward freezing at Hegota devnet 2 scope + adding leanSPHINCS/signature aggregation; Nethermind prefers not reducing unless it accelerates shipping00:57:46
- •EIP-8151 (EC Recover deprecation) flagged as high-risk: preliminary scan shows many contracts depend on it; may not make Hegota00:58:52
- •Greg Colvin: repeated EIP removals/delays cause planning hardship for champions; argues for sliding fork dates rather than dropping EIPs01:05:54
Hegota Devnets
- •Decision: no cross-layer DevNet until ACDC decides on FOCIL; EL-only DevNets remain separate01:27:13
- •Default: pause at current EL DevNet scope; possible exception for cleanup EIPs if strong ACDT consensus Monday01:36:43
Fork Status And Schedule
- •EF/ChainSecurity/DDOB contacted 50+ affected companies; all highs fix by mid-Dec, mediums by Jan 1000:10:49
- •Glamsterdam mainnet target: mid-Jan earliest (no spec changes); late-Jan/early-Feb more realistic00:14:44
- •Hoodi fork week of Dec 14 (Dec 16–17) penciled in; go/no-go decision deferred to next ACDC00:44:49
Frames Transaction Updates
- •Frames breakout: two open spec changes — (1) forward-compat with EIP-8288 signature list, (2) tree-root hashing scheme; target spec-complete before holiday break01:21:32
- •ethrex (Iván) proposing minimal leanSPHINCS-on-frames EIP (ethereum/EIPs#12672); 8288-compatible, no prover/aggregation yet; targeting PFI next ACDE01:23:37
Pq Cryptography Discussion
- •MLDSA/lattice vs. leanSPHINCS debate: EF crypto team holds hash-based is most conservative; MLDSA verifier costs ~800k–1.5M gas with current optimizations01:09:23
- •Mislav (Ethlabs): wallet/UX concern — leanSPHINCS non-standard may repeat secp256k1 adoption problems; frames should remain scheme-agnostic01:13:44
Glamsterdam Spec Change Debate
- •No spec change has majority support; any change restarts full audit/devnet/testnet cycle (~months delay)00:17:50
- •Marius: removing mid-tx refunds (most-cited candidate change) breaks contracts relying on intra-tx gas return; significant unknown blast radius00:32:19
- •Client poll: Geth, Nethermind, Reth, ethrex favor current spec + schedule; Erigon prefers spec change; Besu reluctantly accepts current spec00:40:09
Action items
- •EL devs favoring spec change: produce concrete candidate change with impact analysis before next ACDCSpec change proponents (Erigon, Besu)
- •All client teams: submit one-paragraph Hegota scope position (option 2–5) by end of next weekAll EL client teams
- •Spencer: publish partial impact analysis doc for proposed Glamsterdam spec changesSpencer (testing)
Agenda
From the call issue #2285ShowHide
Key decisions
Glamsterdam will proceed with current spec (no changes); Hoodi fork targeting week of Dec 14 — formal go/no-go at next ACDC
SFIThe Glamsterdam spec is frozen to avoid audit and development delays; the Hoodi fork is tentatively scheduled for mid-December 2026.Top Hegota scope option (drop frames, enshrine PQ directly) has zero support — effectively DFI'd; remaining options 2–4 open for two-week client team review
DFIOption 1 for the Hegota scope, which involved dropping frames to enshrine post-quantum cryptography directly, was rejected due to lack of support.No cross-layer Hegota DevNet until FOCIL decision is made on ACDC; EL DevNets remain separate
Hegota DevNetCross-layer devnet activity for Hegota is paused pending a decision on FOCIL at the consensus layer meeting.EL DevNet scope paused at current level by default; ACDT Monday may greenlight cleanup EIPs only
EL DevNetThe scope for the Execution Layer DevNet is frozen at its current state, with potential exceptions only for minor cleanup EIPs.
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.