Major Bitcoin Core update changes default wallet protocols, risking temporary disruption across popular apps

Bitcoin Core v32.0rc1 has turned the Sept. 14–Oct. 10 window into a concentrated compatibility test for node operators, wallet providers and services that depend on Bitcoin Core's RPC interfaces.

The candidate was tagged with a verified signature on Sept. 14. The live release schedule lists Oct. 10 as the aim for the final v32.0 tag, leaving a 26-day elapsed interval. CryptoSlate's August preview recorded a Sept. 10 RC1 target, while the live schedule now shows Sept. 14, creating a four-day discrepancy without establishing that an unchanged deadline was missed.

The v32.0rc1 tag identifies prerelease software, not a production-final upgrade. It also does not signal a new consensus-rule activation. One change tied to draft BIP 323 alters how Bitcoin Core treats signaling bits and unknown-deployment warnings, but the proposal itself remains in Draft status.

Operators can begin with the high-level pattern in Bitcoin Core's most recent RC testing guide: exercise regularly used features in separate temporary data directories and compare the candidate with the prior release. The official download page lists 31.1 as the current baseline. That comparison can expose differences in node startup, wallet behavior and RPC responses without treating the candidate as a routine production update.

The largest performance change in the draft v32 release notes is parallel prefetching of transaction outputs during block connection. The setting defaults to eight workers, supports up to 16 and can be disabled. Running disk-bound validation with multiple settings can reveal whether faster block processing comes with unacceptable CPU, memory or storage-latency costs on an operator's hardware.

Wallet and service integrations face a separate breakage risk. Four RPCs will default to PSBTv2, while other interfaces remove deprecated fields or reject arguments that older versions tolerated. Teams that create, convert or fee-bump PSBTs should therefore trace those transactions through their downstream parsers and signers.

Fee handling also needs failure-path coverage. The default estimatesmartfee path combines block-policy and mempool estimators, may return a lower estimate and can error if either component fails. Operators should observe startup and sparse or unhealthy mempool conditions, then confirm that monitoring and explicit block-policy fallbacks behave as expected.

The HTTP server rewrite broadens the test surface beyond the node itself. It adds an 8,192-byte header limit, stricter malformed-header handling, a default ceiling of 16 RPC connections, new REST cache controls and immediate disconnection of unauthorized client addresses. Those changes can surface in reverse proxies, health checks, client pools and error handlers.

Rollback deserves equal attention. A rebuilt transaction index uses less than half the disk space, but older releases cannot read the new format, so a downgrade can trigger another rebuild lasting hours. Privacy-focused operators should also reproduce private-broadcast failure paths around the Tor fallback fix, the 10,000-entry queue, the 1,000-attempt limit and relay behavior under load. With the final tag still only an aim, these edge cases are the practical work of the RC window.