The result

After the bugs were fixed and the contracts were in their proper directories, the test suite settled at:

79 passing (36s)

That’s the 49 baseline tests, the 21 new Resolver tests (text + contenthash + ERC-165), the 7 new burn tests, plus a couple of additional checks I’d forgotten about.

More importantly, the four ENS interface IDs now report true from supportsInterface. That means:

• MetaMask can show profile cards for .dex domains using its existing ENS code path.
• IPFS-aware browsers (Brave, Opera) can navigate to .dex domains pointing at IPFS sites with no further integration work.
• OpenSea can display avatar text records for .dex domain NFTs by reading the resolver the same way it reads ENS resolvers.
• The official ENS app could, in principle, resolve a .dex name. (It won’t, because its UI hardcodes .eth, but the resolver interface is interchangeable.)

None of these integrations required me to talk to anyone. No business development calls. No partnership announcements. No SDK to publish. The wallets and browsers don’t even know .dex exists. They just call text() and contenthash() and addr(node, coinType) on whatever resolver they’re pointed at, and the bytes that come back are formatted exactly as their existing parsers expect.

This is what standards do. This is what nine years of ENS engineering bought, and what I got to inherit for free by typing the right four function signatures.