Stellar's Bell Upgrade Votes Nov 23, and Your Decoder Might Break
Stellar published a software checklist for its Protocol 30 'Bell' upgrade on October 23, with testnet voting November 3 and mainnet November 23. New contract addresses are one problem. The ledger header format change is the one that catches people off guard.
JUST IN: Stellar pushed a version checklist on October 23 for Protocol 30, the network's Bell upgrade. Testnet validators vote November 3. Mainnet votes November 23. Both at 17:00 UTC.
That leaves 31 days for wallets, exchanges and infrastructure operators to get their software sorted. And the checklist isn't optional reading, because Bell breaks two things quietly.
Start with addresses. A single smart contract can now carry an extra customer identifier, so one contract receives funds for multiple customers and still records who paid. Sounds boring. It's not. Custody shops running a pile of shared accounts could collapse that into something simpler. The JavaScript SDK v17.3.0, out October 21, handles the new W-prefix addresses, with encode, decode and validate tools plus transfer handling baked in.
But installing the library isn't the same as doing the work. Your app still has to recognize the address and tie that identifier to the right customer. Skip it and you're staring at a ledger full of payments nobody can attribute.
Now the part that'll bite the teams who think they're sitting this one out. The ledger header format changes. Decoders built on older definitions will choke on the new values. Indexers, explorers, anyone reading raw ledger data, that's you. This changes things for plenty of shops that aren't touching the address feature at all.
Bell also adds ML-DSA signature verification inside smart contracts. Big buzzword, narrow scope. Ordinary network transaction signatures stay Ed25519. So it's a building block for custom accounts, not a quantum migration. Anyone calling this a post-quantum upgrade is reading it wrong.
Millisecond ledger timestamps come along too. Doesn't speed up closing times today, it just clears a barrier for later. And it breaks code that assumes consecutive ledgers carry different whole-second stamps.
Loose ends remain. The Protocol 30 command-line release is still pending. The Soroban Rust SDK is a release candidate, not a final build. Dates can slide, and the foundation says so.
The market's verdict: XLM holders get a countdown, developers get homework. The address feature is a genuine win for custody and payments. The header change is where people get caught. Run infrastructure and haven't opened the checklist yet? You've got 31 days.