v26.47.0 release notes

[bosconi]

Release notes for the v26.47.0 train. Dates in RC snapshots are provisional until the release ships. One commit is added per snapshot; the final snapshot sets the real dates.

• Content (all from rc.1): 1 Feature (TRY_CAST), 8 Improvements, 9 Bug Fixes.
• Dates: provisional Cloud 2026-10-22, Self-Managed 2026-10-23, matching the cadence rule applied to the v26.47.0-rc.1 tag (commit date 2026-10-16, a Friday): the first Thursday strictly after it, Self-Managed the day after.
• Operator compatibility: row added for v26.47, release date 2026-10-23.
• Branch base: main, whose newest section is ## v26.45.1 — v26.46.0's section is still on its own unmerged branch (v26.46.0 release notes #39690). Expect a top-of-list conflict when both merge; v26.47.0 goes above v26.46.0.

⚠️ Open questions for the release owner

• GA announcements — the largest judgment call in this snapshot. Three docs PRs remove preview markers: #39831 (source versioning, CREATE TABLE ... FROM SOURCE), #39836 (replacement materialized views), #39830 (Kafka sink headers, noted as "GA in v26.47"). None changes code, so all three were held; drafted wording for each is in omitted.md. Which are actually GA in v26.47.0 and should be announced? Note v26.46.0 published Kafka sink message attribution as private preview — if the header goes GA here, this release should say so rather than repeat the preview framing. Also, at the rc.1 tag the docs still label the sink-id header private preview; #39829, which removes those callouts, is not in this snapshot.
• SUBSCRIBE soft assertions (#37611, #39764, both held): drafted as Bug Fixes, then pulled back. Both failures come from soft_assert_or_log!, which panics only where soft assertions are enabled and otherwise logs; neither PR says what a release build does. Are soft assertions on in the environments customers run? If yes, both are publishable as drafted in omitted.md.
• Console Freshness page (#39340, held): a complete new Monitoring > Freshness page, gated behind freshness-page-CNS164, off in both Cloud and self-managed, and the LaunchDarkly flag is not yet created. If it is turned on for v26.47.0, this likely deserves a Feature heading.
• PostgreSQL sources ingesting during a snapshot (#39419, #39479, held): adding a large table to an existing source would no longer stall the tables already hydrated — but storage_persist_sink_description_lookahead is Duration::ZERO at the tag. Publish if the lookahead is being enabled.
• Coalesced materialized-view sink batches (#37591, held): measures ~67% fewer blob writes and ~54% smaller state diffs, at ~10 ms added p50 latency on small writes, but compute_mv_sink_shared_batch_max_bytes is 0 at the tag. Publish (with the latency cost stated) if the staged rollout reaches customers in this release.
• MCP 2026-07-28 conformance checks (#39694, #39695, held): enable_mcp_protocol_2026_07_28 is false at the tag. If the revision is being served, these belong together as one Improvement.
• Refresh strategy columns (#39597, published as an Improvement): this renames aligned_to → aligned_to_expression and at → at_expression and changes their type. It is a breaking change for anyone selecting them, in a schema documented as unstable — it may warrant more prominence than a bullet.

Snapshot drafts (with PR links) in mz-skills:

• rc.1: release_notes.md / omitted.md

Borderline PRs omitted: rc.1: 14 — omitted.md#borderline

🤖 Generated with Claude Code