jordan's avatar
jordan 1 week ago
BIP47DB IS LIVE A month after the white paper went live at https://bip47db.github.io, Max Tannahill has created the tool and it is live: Inscribe and browse BIP47 payment codes from your browser. Mainnet ready, fully client-side, no server to trust. The protocol whitepaper has been refined to v1.6 with community feedback, esp @MightyMercurian. If you missed the first tweet: BIP47 reusable payment codes solve real privacy problems but depend on directory services like Those are centralised. If they go down, wallet recovery from seed alone breaks for connections you've made. Max @maxtannahill · 5h BIP47DB inscribes the directory itself onto Bitcoin. The Publisher tool walks you through inscribing a batch of payment codes: – Fetch from (or paste manually): Sign a single message with your BIP47 wallet (@AshigaruFanClub/@SamouraiWallet) – Cosign the funding tx with @SparrowWallet: Up to 5,000 codes per batch. Validates the UTXO covers fees before you sign anything. The Browse tab decodes inscriptions back into a searchable database. Anyone can sync from a @mempool API endpoint and verify every record cryptographically. There is no trust in the indexer, no trust in the publisher. The whole site is static HTML + JS. The repo is open source. The single CORS proxy he runs (for the fetch convenience) only relays one specific endpoint and rejects unknown origins. If Max disappeared tomorrow, anyone could host this themselves in an hour. The April tweet got a healthy mix of "this is cool" and "Ordinals are cancer.", and he's very grateful for all the technical feedback. As for BIP110 signallers? He doesn't care what you think. Try it at Read the spec at Issues / PRs at Mainnet inscriptions are now possible but no batches have been written yet. I'm waiting for the first one to be by the community or Ashigaru, not me. Have at it and make sure to join Freedom Tech Friday on the Ungovernable Misfits feed on May 15th @ 9 AM EST to get the full breakdown of this important tech!

Replies (5)

Why ordinals? Why not just in OPRETURN? To save on fee? I'm just confused why outputs need to be inscribed as an ordinal and not just a new protocol? If ordinals is just abstracting main chain for a shared delusion then why not abstract something new? I'm not hating. Just very curious. It's a key design decision.
I read it. It didn't answer really why this couldn't be done with a new shared delusion instead of ordinals. Ord indexes are not backwards compatible. Ie I have to reindex the whole thing on any upgrade. Can I build a index that only gets these addresses and not have to index every sat? It takes weeks for a ordinals database to sync, especially if you're indexing sats.
jordan's avatar
jordan 6 days ago
In that case, I would compare your codebase with bip47db using Claude or Codex and determine if the architecture is a fit 👍