Issues

• error: use of undeclared identifier INT_MAX

  bug
  timescale/pg_textsearch#586 · Filter by author dbeecham opened on Oct 23, 2026

• Compressed posting reads past segment end, causing "Invalid logical page" on a fresh index

  timescale/pg_textsearch#584 · Filter by author Murlors opened on Oct 23, 2026

• Weighted Sparse Vector Search with WAND

  enhancement
  timescale/pg_textsearch#574 · Filter by author hhkbble opened on Oct 20, 2026

• Index scan scores memtable rows with the exact document length, segments and <@> with the quantized one, so ranking changes on spill

  timescale/pg_textsearch#573 · Filter by author firu-daniel opened on Oct 20, 2026

• Parallel index build stores HOT-descendant TIDs instead of HOT-chain root TIDs; index scans silently omit matching rows

  timescale/pg_textsearch#571 · Filter by author MatthiasF999 opened on Oct 20, 2026

• VACUUM crash between bitmap page writes leaves a stale alive_count

  timescale/pg_textsearch#569 · Filter by author me1t0ut opened on Oct 19, 2026

• Ordered BM25 index scan returns a wrong top-N (rows out of order, better-scoring rows dropped) before the memtable is spilled — v1.3.1 and v1.5.1

  timescale/pg_textsearch#557 · Filter by author ld-873233 opened on Oct 12, 2026
  1

• Feature request: a numeric, non-superuser function for the dead documents of an index

  timescale/pg_textsearch#555 · Filter by author thijsdroog opened on Oct 9, 2026

• Literal query text with an index on a plain column of a partitioned parent: the leaf index is rejected

  timescale/pg_textsearch#554 · Filter by author thijsdroog opened on Oct 9, 2026
  1

• Two bm25 indexes on the same expression: the query aborts when the planner picks the other one

  timescale/pg_textsearch#553 · Filter by author thijsdroog opened on Oct 9, 2026
  1

• Partitioned table: the parent index name (and the implicit <@> 'text' form) scores with the statistics of the first partition only

  timescale/pg_textsearch#552 · Filter by author thijsdroog opened on Oct 9, 2026
  2

• BM25 index scan computes <@> again for each row when the connection sets its role and search path after login

  timescale/pg_textsearch#551 · Filter by author coroluca opened on Oct 7, 2026

• AlloyDB build reporting 1.0.0: invalid segment headers and corrupt tombstone pages; upgrade guidance

  timescale/pg_textsearch#549 · Filter by author bachner opened on Oct 5, 2026

• Issue performing "Build from source" for Windows

  bug
  timescale/pg_textsearch#534 · Filter by author nitin065 opened on Oct 1, 2026
  2

• Skip cache totals scan when the captured snapshot endpoint no longer matches

  timescale/pg_textsearch#520 · Filter by author tjgreen42 opened on Sep 29, 2026
  1

• Optimize Boolean queries with cache-backed memtable evaluation

  timescale/pg_textsearch#499 · Filter by author tjgreen42 opened on Sep 17, 2026

• Benchmark Boolean, phrase, mixed, and COUNT workloads after query support lands

  enhancementperformancescope:medium
  timescale/pg_textsearch#496 · Filter by author tjgreen42 opened on Sep 16, 2026

• Reduce peak memory during large BM25 index builds

  performancescope:large
  timescale/pg_textsearch#498 · Filter by author tjgreen42 opened on Sep 16, 2026

• Investigate poor read-only disjunction performance in PlanetScale TIN benchmark

  performancescope:large
  timescale/pg_textsearch#497 · Filter by author tjgreen42 opened on Sep 16, 2026
  2

• Confine at-capacity segments to the deepest reachable level

  enhancementperformancescope:medium
  timescale/pg_textsearch#484 · Filter by author tjgreen42 opened on Sep 3, 2026

• Enforce segment format limits consistently across build and merge

  bugscope:medium
  timescale/pg_textsearch#473 · Filter by author tjgreen42 opened on Aug 29, 2026

• On-disk uint32 format limits cap a fully-merged segment at a 4 GiB string pool / 268M terms. Is segment splitting (or a wider format) the intended path for very large corpora?

  enhancement
  timescale/pg_textsearch#456 · Filter by author YaserJaradeh opened on Aug 18, 2026
  1

• validate_queries.sql: MS MARCO query 158 fails by 1.5 score points (pre-existing tokenization discrepancy)

  bug
  timescale/pg_textsearch#361 · Filter by author tjgreen42 opened on May 11, 2026
  2

• Phrase queries with positional information

  enhancementscope:large
  timescale/pg_textsearch#314 · Filter by author tjgreen42 opened on Apr 15, 2026
  1

• Feature request: publish debian packages

  developer relationsenhancement
  timescale/pg_textsearch#304 · Filter by author ibotty opened on Apr 8, 2026
  2