4/ Built and proven. Not live yet — on Bitcoin you buy computation with bytes, and right now the computation costs more bytes than the data does. Gas first, wire second. We publish the arithmetic because someone will check it. 572 ÷ 3.73, take 10%. Do the division.

V17 adds 1/ What we can't derive yet — and what V17 fixes. 🧵 2/ Holder identity. The blob names owners by compact ID, not address.You can rebuild the ID→address map from Bitcoin alone today, in about 17 minutes. V17 puts it directly on the wire. ~2,860 bytes. 3/ Per-band TVL. One already-paid-for slot in the format was carrying nothing. We filled it with exact ORDI custody — every band now self-certifies the vault. 4/ Per-owner ORDI attribution — an explicit ledger instead of a 1:1 inference. 5/ Token metadata: name, symbol, image CIDs, creation time. The token contract address comes free from the CREATE2 salt. 6/ Full parity for launchpad tokens — including ones that never traded — and creator-fee rows that currently lose track of which token they belong to. 7/ Why is all of this so cheap? Because we measured it: ~163,000 bytes of every inscription were padding. The entire verifiability program costs under 2% of a band. 8/ V17 also removes ~20% of settlement gas that was being spent decoding the same data twice — a compiler workaround, not verification work. More actions per epoch, identical guarantees. 9/ The goal was never a TPS number.It's that a stranger with a Bitcoin node can check every claim we make. Yet we still are at 102 TPS with full verifiability.