Vitalik Buterin
@vitalik.eth
Yoav talking about AA at EthSingapore. Really important discussion on why AA is NOT "just meta-transactions", and why the extra steps that ERC-4337 takes compared to more naive approaches are necessary if you care about decentralization.
7 replies
13 recasts
93 reactions
Scott
@scotthconner
Paymasters for EOAs I feel like would give us 80% of the power. All of the bullet points on the slides could be describing a 4337 relayer. What am I missing?
2 replies
0 recast
0 reaction
Ahmed
@0xahmed
At least right now with bundlers, there’s a standard for transactions. Soon there’ll be a open bundler network
1 reply
0 recast
0 reaction
Scott
@scotthconner
Im still not convinced the step back is worth the trade off. Everyone agrees it's temporary until enshrined AA so it seems like a lot of throwaway work.
1 reply
0 recast
0 reaction
Ahmed
@0xahmed
I don’t see it being enshrined anytime soon until there’s major take off on L2s and then on L1
1 reply
0 recast
0 reaction
Scott
@scotthconner
Why? Imagine if we had put the effort into enshrinement instead of all this temporary, centralizing infra? We could be done.
1 reply
0 recast
0 reaction
Vitalik Buterin
@vitalik.eth
Protocol core devs are and have been super busy the last 1-3 years on the merge, EIP-4844, and client code firefighting. Were AA an EIP like verkle trees, I really do expect enshrinement would take almost as long, and we would not have been able to kick off an ecosystem for years longer.
1 reply
0 recast
4 reactions
Scott
@scotthconner
Definitely not questioning protocol priorities. ERC 4337 is explicitly outside the protocol. The 4337 mission seems to be onboarding. 4337 may enable easier use, but it's the strong use cases themselves that will bring in users, who will overcome friction relative to the use case value.
1 reply
0 recast
0 reaction
