The most exotic consensus in the series. The most vanilla build.

Week 3 of Proof of Support is live on Core DAO. Chain selection has its own article — I covered why Core won the shortlist here. Short version: Satoshi Plus consensus, Bitcoin miners voting with hashrate, BTC holders timelocking on L1 with OP_CLTV. Three parties, one validator election, EVM on top.

Here's what actually happened.

The thing that was supposed to bite didn't

Research flagged one config change going in: evm_version = "shanghai" in foundry.toml. Core DAO's docs require the Shanghai pin — Cancun opcode support isn't confirmed even after the Hermes fork. Two weeks on Base and Scroll had me defaulting to "cancun", and silently wrong bytecode is exactly the kind of wrong that wastes a morning.

forge build compiled clean on the first run. OZ Ownable and ReentrancyGuard are fine against the Shanghai target. 14 tests. No issues.

So. That wasn't the problem.

The first L1 in the series — and the CORE token problem

Base is an OP Stack rollup. Scroll is a ZK rollup. Core DAO is just an L1. No sequencer, no proving layer, no 7-day withdrawal window. After two weeks of rollup mechanics, the mental model simplification is real — "confirmed" means confirmed, full stop.

The catch is the gas token. CORE, not ETH. For a developer it's trivial — you need CORE in the deployer wallet, you get it, you move on. For a casual supporter landing on the app, it's a different story.

Getting CORE isn't smooth. There's one official bridge at bridge.coredao.org (uses LayerZero for its bridging tech!) that works. Symbiosis also lists a Core DAO bridge — I checked, it was dry, at least for certain pairs. In the end I swapped on a CEX from my Base wallet and sent CORE directly to the deployer. That's not a bad path if you know your way around exchanges, but it's not something a casual user clicks through in thirty seconds. It's probably the biggest real-world friction point for adoption this week.

Hermes finality is noticeable

One quick line on this: post-Hermes (November 2025), Core DAO has 2-block finality — about 6 seconds. Compared to Scroll's ZK proving latency, where confirmation means "in the sequencer, not yet on L1," the Core DAO frontend felt fast. Transactions landed visibly. No sitting and wondering.

Vibe Check

(Full methodology: how I'm scoring the chains)

Actual: 46/60. Estimated: 49/60. Delta: −3 weighted points (D4 and D6 each drop one raw point at ×1.5 weight: −1.5 each)

D4 and D6 are both down one — the verification docs had the wrong flags, and the testnet/mainnet API key split wasn't documented anywhere. Everything else landed where research said it would.

D5 is the strong spot: coreDao is built into viem, MetaMask connected immediately, zero custom chain config for mainnet. D7 stays a 5 — the deploy cost 0.083 CORE, roughly $0.002 at ~$0.027/CORE. A sendSupport() call is a fraction of that. Effectively free for supporters.

The consensus paradox

Satoshi Plus is the most interesting security mechanism I've touched in this series. Bitcoin miners include metadata in coinbase transactions to vote for Core validators. BTC holders timelock coins on Bitcoin mainnet using OP_CLTV — the coins don't move, they just declare a preference. Three independent parties produce a hybrid score that elects 31 validators for the day.

From inside the house — the Solidity contract, the viem adapter, the MetaMask prompt — I felt none of it. Same EVM. Same Foundry flags. Same gas estimates I'd have on any other EVM L1. The foundations are Bitcoin-native. The walls look identical.

I'm genuinely uncertain whether that's a feature or a missed opportunity. You want the security model to be invisible to the developer — that's what "it just works" means. But there's something odd about spending a week on the most exotic consensus design in the series and coming away with a build log that reads the same as the previous two weeks.

The app is live: https://proof-of-support.pages.dev

Week 4 picks up the non-EVM category. The build log will look different.