So is that possible to build an on-chain orderbook dex but using Prop AMMs' quotes and liquidity? Since we can get a composite-curve like this, it should be possible to create an orderbook trading platform, right? Prop AMMs are good for swaps and users are more familiar with

"aggregators countering abusive AMMs" My thought on this is that using JIT routing, see what happened in DFlow and Jupiter

One of the most thought-provoking talk in the BP25, although I haven't understand it (I will rewatch this again and again), pls do have a look on the whole speech

This article shows how to make Prop AMMs work on EVM by solving each bottleneck directly: • Update before swap → Builder-enforced Priority Lane (ToB inclusion) • Cheap updates → Priority Lane = local fee market (updates only compete with updates) • Expensive SSTORE →

So what's the reason to choose alt-DA, higher throughput? native composability?

privacy matters

非常好的 summary

Another "probably nothing" huh Mantle

Come here!

PBFT is far from perfect: multi-round voting, complex view-change logic, and strong assumptions about leader stability make it hard to fit long-running blockchains with frequent leader rotation. This is why the community keeps looking for PBFT alternatives. Minimmit raises the

一个新的场景： rollups 的升级由一个 security council 组成，其中包含 internal dev team，external team 以及擅长 audit 的 agent bots，未来的升级将由内部团队发起并通过 x402 支付给 audit agent bots，这些 bots 需要 8004 中的 registry 和 reputation 做支持，这样就是一个 crypto native

It's my first time to join EF event, really enjoyed, seeing so many talented people who are still working on Ethereum. And ofc Tesa's institution talk was impressed.

Base is huge, which means they can iterate quickly, 6 hardforks per year, that's what they can do but others can't.

That's what happens in this special period, traditional client design and implementation flow is dead, super fast iteration era is coming

How to be a better builder in this era, here are the things: 1. Improve your management skills, try to build agent orchestration capabilities on top of strong traditional engineering fundamentals 2. Make your codebase better (aka. more agent friendly), testing, consistency, and

six levels of automation for AI-driven commerce.

3 structural constraints to push agents to blockchain rails: --- 1. Identity and Access 2. Programmable Money and Micropayments 3. Deterministic Enforcement

Context engineering from scratch

x.com/i/article/2030…