Native Account Abstraction #007
Transcript
- wolovim
gm
- Iván | ethrex
gm
- felipe
gm
- lightclient
gm
- marc | wolovim
- Stefan Starflinger
gm
- lightclient
have we run the spec tests on the network?
- lightclient
i like "nonce"
- Toni Wahrstätter
Nonce is better yeah
- Daniil Ankushin | Nethermind
Reacted to "i like "nonce"" with 👍
- Toni Wahrstätter
Might need an update in 8250, will check
- Iván | ethrex
Reacted to "i like "nonce"" with 👍
- Stefan Starflinger
Reacted to "i like "nonce"" with 👍
- Guru
Reacted to "Might need an update..." with 👍
- felipe
Reacted to "Might need an upda..." with 👍
- felipe
Reacted to "i like "nonce"" with 👍
- lightclient
which tracing api do you use?
- Justin Florentine
Reacted to "which tracing api do you use?" with 😂
- Stefan Starflinger
Reacted to "which tracing api do you use?" with 😂
- pk910
let me check the exact method, but it'S pretty much a trace block by hash call
- pk910
debug_traceBlockByHash with the call & state change tracer the state tracer still works, just the callTracer gives no results it'll need a frames array too
- felipe
It would be good to get benchmarking tests earlier in this dev cycle too
- spencer
sgtm!!
- Toni Wahrstätter
Can we iterate faster if not yet combined?
- Daniil Ankushin | Nethermind
Reacted to "Can we iterate faster if not yet combined?" with 👍
- Iván | ethrex
Reacted to "Can we iterate faster if not yet combined?" with 👍
- lightclient
probably
- spencer
Replying to "Can we iterate fas..." add EL EIPs?
- spencer
Replying to "Can we iterate fas..." :D
- Toni Wahrstätter
Ofc, focil profits from a combined one
- Toni Wahrstätter
Replying to "Can we iterate faste..." Exactly
- Stefan Starflinger
Replying to "Can we iterate faster if not yet combined?" i would say so
- marc | wolovim
- Iván | ethrex
Reacted to "max_fee: https://github.com/ethereum/EIPs/pull/12414" with 👀
- Iván | ethrex
I haven’t yet
- Stefan Starflinger
Reacted to "Can we iterate faster if not yet combined?" with 👍
- lightclient
- Justin Florentine
what do the wallets think?
- Iván | ethrex
Reacted to "what do the wallets think?" with 👍
- Guru
Yeah. The gas spill accounting tends to get messy
- Justin Florentine
agree this is overloading, and we should avoid unless this is an opportunity we can't miss
- Iván | ethrex
Reacted to "agree this is overloading, and we should avoid unless this is an opportunity we can't miss" with 👍
- spencer
feels like it simplifies the future (if we do 7999)
- lightclient
Reacted to "feels like it simplifies the future (if we do 7999)" with 👍
- Toni Wahrstätter
For which dimensions would we do that?
- Iván | ethrex
I like the idea of giving it a try implementing it, see how it goes, and decide then if we keep it or postpone it
- Toni Wahrstätter
Is it only state vs execution gas?
- lightclient
Replying to "Is it only state vs ..." and blobs
- Marc
same would be happy to include it
- Daniil Ankushin | Nethermind
Reacted to "same would be happy to include it" with 👍
- soispoke
no
- Guru
No
- Toni Wahrstätter
As a headliner Edit, ah no that was enc mempool
- Iván | ethrex
Reacted to "As a headliner Edit, ah no that was enc mempool" with 😅
- spencer
Reacted to "As a headliner Ed..." with 😅
- soispoke
probably more eyes from people like Maria would be really good
- lightclient
- Iván | ethrex
Reacted to "probably more eyes from people like Maria would be really good" with 👍
- soispoke
how much of Frames <> mempool is in the current devnet? thinking specifically about removing the 1 txn per sender limit and testing multiple concurrent private txns with MATCHA
- lightclient
- Alex Forshtat
- Justin Zen - ocl
sorry didn't have chance to create pr, we from ocl would like to propose introducing VERIFY_NONSTATE and VERIFY_STATE mode to separate heavy verification like signature verification with state check like recovered address from sig. it'll save L2 sequencer from doing all verification again. Happy to hear thought from everyone, thanks
- spencer
could we consider adding new eips to bogota-devnet-2 :D
- Iván | ethrex
Reacted to "could we consider adding new eips to bogota-devnet-2 :D" with 🙏
- lightclient
Replying to "could we consider ad..." like what 😆
- Iván | ethrex
Replying to "could we consider ad..." I’d do SETCODEFROM + 8151
- lightclient
Replying to "could we consider ad..." hmm that would be good
- Iván | ethrex
Replying to "could we consider ad..." + others that the sooner we add them the better
- spencer
Replying to "could we consider ..." i dont know if we are going against process but...
- Guru
Replying to "could we consider ad..." +1
- spencer
Replying to "could we consider ..." EIP-8131 Unified Transaction Content Floor (test benchmarking earlier) EIP-8279 Block Access List Byte Floor (test benchmarking earlier) EIP-8253 Bump nonce of zero-nonce storage accounts (critical to test earlier, requires fork transition) EIP-8298 SETCODEFROM EIP-4758 Deactivate SELFDESTRUCT (reduces test space, add earlier)
- Iván | ethrex
Replying to "could we consider ad..." I’d add 8151 to that list and would be ok for me
- lightclient
Reacted to "I’d add 8151 to that list and would be ok for me" with 👍
- Marc
Reacted to "I’d add 8151 to that list and would be ok for me" with 👍
- Daniil Ankushin | Nethermind
Reacted to "I’d add 8151 to that list and would be ok for me" with 👍
- soispoke
can we add MATCHA :)?
- Iván | ethrex
Reacted to "can we add MATCHA :)?" with 🔥
- Iván | ethrex
Replying to "can we add MATCHA :)..." 🤝
- Justin Zen - ocl
Reacted to "can we add MATCHA ..." with ❤️
- Justin Florentine
Reacted to "can we add MATCHA :)?" with 🍵
- Iván | ethrex
Reacted to "can we add MATCHA :)?" with 🍵
- Daniil Ankushin | Nethermind
Reacted to "can we add MATCHA :)?" with 🔥
- Daniil Ankushin | Nethermind
Reacted to "can we add MATCHA :)?" with ❤️
- spencer
lfg! :D
- Justin Zen - ocl
Sorry for unreliable internet :(
- felipe
cannot hear you :(
- spencer
Reacted to "can we add MATCHA ..." with 🍵
- Justin Zen - ocl
sure!
Call summary
Targets
- •bogota-devnet-2 scope finalized at Thursday ACDT; launch targeting before Mumbai travel - 00:59:02
Decisions
- •"nonce" adopted over "nonSequence" for Frame transaction RPC field naming; EIP-8250 to be updated accordingly - 00:11:28
- •PR #12502 (signature length bounding) to be merged post-call; included in bogota-devnet-2 - 00:43:32
- •PR #12530 (remove TXDiff opcode) tabled pending resolution of pre-state proxy slot use case - 00:53:29
- •Next devnet is bogota-devnet-2 (EL-focused); Hegota combined devnet deferred - 00:16:43
Highlights
- Organizational:
- ·"nonce" preferred over "nonSequence" for Frame transaction field naming; 4+ clients aligned, Besu abstaining pending further thought - 00:11:28
- ·Next devnet remains Bogota devnet-2 (EL-only); combined Hegota devnet deferred; scope to be decided at Thursday ACDT before Mumbai travel - 00:54:18
- Testing Progress:
- ·Bogota devnet-1 running but lightly stressed; spammer struggles with consistent Frame transaction throughput - 00:07:58
- ·RPC inconsistencies across clients blocking Frame transaction data retrieval; API specs not finalized - 00:08:26
- ·No clients have implemented tracing APIs yet; Explorer needs call traces (debug_traceBlockByHash) to inspect frames - 00:13:14
- Eip Proposals Bogota:
- ·EIP-7999 max_fee PR (PR #12414): unified fee cap across execution/state/blob dimensions; broad conceptual support, Daniil raised fee-access opcode interaction concern still unresolved - 00:19:46
- ·PR #12502: bound arbitrary signature object length in canonical hash to prevent relay-padding cost attack; no opposition, plan to merge post-call - 00:43:32
- ·PR #12530 (7906): proposes removing TXDiff opcode in favor of TXTrace; deferred after Alex raised pre-state proxy slot use case not coverable by TXTrace alone - 00:47:38
- Eip Proposals Hegota:
- ·Spencer proposed adding to bogota-devnet-2: EIP-8131, EIP-8279, EIP-8253 (nonce bump), EIP-8298 (SETCODEFROM), EIP-4758 (deactivate SELFDESTRUCT), EIP-8151; process/scope TBD at ACDT - 00:56:27
- ·Offchain Labs (Justin Zen) proposed VERIFY_NONSTATE + VERIFY_STATE frame modes to separate stateless signature verification from stateful checks, saving L2 sequencer re-verification - 01:02:22
Action Items
- •lightclient - Resolve Daniil's fee-access opcode concern on PR #12414, then raise on ACDE for final merge decision - 00:37:27
- •All client teams - Bring list of EIPs comfortable to include in bogota-devnet-2 to Thursday's ACDT call - 00:58:44
Key decisions
"nonce" adopted over "nonSequence" for Frame transaction RPC field naming; EIP-8250 to be updated accordingly
Standardizing field naming across clients for Frame transaction RPC data.PR #12502 (signature length bounding) to be merged post-call; included in bogota-devnet-2
bogota-devnet-2Bounding arbitrary signature object length in canonical hash to prevent relay-padding cost attacks.PR #12530 (remove TXDiff opcode) tabled pending resolution of pre-state proxy slot use case
Decision to remove TXDiff opcode in favor of TXTrace was deferred after concerns were raised regarding pre-state proxy slot coverage.Next devnet is bogota-devnet-2 (EL-focused); Hegota combined devnet deferred
Prioritizing an execution-layer focused devnet over a combined Hegota devnet for the next testing phase.
EIPs discussed
AI Disclaimer: Some content or metadata on EIPsInsight may be AI-inferred or automatically compiled. If you find any discrepancy, please contact us at dev@avarch.org.