Sep 2025

2 commits
doc PG 18 relnotes: update description of skip scan item
18 Sep 2025Bruce MomjianREL_18_STABLE18.0
LLM analysis, 86% surelanded as postedfrom patchset #1461 of 1 reviewer saw the final version

#146 explicitly announces the attached committed patch, and the commit replaces the skip-scan paragraph with the wording developed in #145. The empty attachment prevents direct content verification.

  • Peter Geoghegansaw the final versionLLM analysis, 98% sureReported the misleading equality restriction in #139 and explained inequality and prefix-column cases in #144, leading to the final wording.
doc PG 18 relnotes: add mention of temporal foreign key constr.
16 Sep 2025Bruce MomjianREL_18_STABLE18.0
LLM analysis, 100% surelanded as postedfrom patchset #1361 of 1 reviewer saw the final version

The temporal foreign-key links and WITHOUT OVERLAPS/PERIOD explanation exactly match either duplicate attachment in #136; Bruce confirms application in #138.

  • Bruce Momjiansaw the final versioncommittercommitted the patch

May - Jun 2025

34 commits
doc PG 18 relnotes: add AFTER trigger user change item
7 Jun 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surewritten from the discussion, no matching patch3 of 3 reviewers saw the final version

Adds the AFTER-trigger incompatibility entry drafted inline in #129 and announced committed in #130; no attachment contains this addition.

  • Noah Mischsaw the final versionLLM analysis, 99% sureRequested the missing incompatibility entry in #118 and reviewed its role and SECURITY DEFINER semantics in #121 and #125.
  • David G. Johnstonsaw the final versionLLM analysis, 98% sureCorrected the execution-timing explanation and discussed trigger terminology in #120 and #126.
  • Tom Lanesaw the final versionLLM analysis, 99% sureSupplied the AFTER-trigger and event-queue wording in #124 and clarified non-deferred cases in #127.
doc PG 18 relnotes: modify async I/O item for other improvements
29 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #1171 of 1 reviewer saw the final version

The single change making the asynchronous-I/O operation list non-exhaustive exactly matches #117.

  • Melanie Plagemansaw the final versionLLM analysis, 100% sureReviewed #113 and requested an explicit indication that other operations also benefit in #116.
doc PG 18 relnotes: split apart log_connections item
29 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #1131 of 1 reviewer saw the final version

#113 attaches this commit's complete diff, including the log_connections split and asynchronous-I/O examples.

  • Melanie Plagemansaw the final versionLLM analysis, 100% sureIdentified the distinct connection-duration feature and missing asynchronous-I/O examples in the existing draft in #112.
doc PG 18 relnotes: move ANALYZE item,split ANALYZE/EXPLAIN item
28 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #1071 of 1 reviewer saw the final version

Matches #107, including placement of ANALYZE entries under Monitoring; the earlier #106 instead placed them under Utility Commands.

  • Bruce Momjiansaw the final versioncommittercommitted the patch
doc PG 18 relnotes: clarify multiplication item
28 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #1051 of 1 reviewer saw the final version

All numeric multiplication/division wording and reference additions match #105; differences are patch metadata rather than content.

  • Bruce Momjiansaw the final versioncommittercommitted the patch
doc PG 18 relnotes: add removal details to MD5 item
27 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

Uses Nathan Bossart's inline MD5 deprecation text from #101, with line wrapping adjusted; the commit explicitly credits him as Author.

  • Bruce Momjiansaw the final versioncommittercommitted the patch
doc PG 18 relnotes: fix markup
27 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

Removes the stray greater-than sign from the Logical Replication Applications title reported in #100 and confirmed fixed in #103.

  • Peter Smithsaw the final versionLLM analysis, 100% sureReported the extra greater-than sign in the rendered section heading in #100.
doc PG 18 relnotes: clarify btree skip-scan item
23 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #971 of 1 reviewer saw the final version

Exactly matches the spelling and equality-reference changes attached in #97.

  • Peter Geoghegansaw the final versionLLM analysis, 99% sureReviewed #86 and explained why absence of an equality condition matters in #95, prompting the follow-up wording.
doc PG 18 relnotes: update to current
23 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #961 of 1 reviewer saw the final version

Exactly matches Bruce's combined attachment in #96, incorporating Michael Paquier's injection-point patch from #90 alongside other updates; the commit explicitly credits Michael as Author.

  • Bruce Momjiansaw the final versioncommittercommitted the patch
doc PG 18 relnotes: adjust CREATE SUBSCRIPTION attribution
23 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #891 of 1 reviewer saw the final version

The one-line CREATE SUBSCRIPTION attribution replacement exactly matches #89.

  • vignesh Csaw the final versionLLM analysis, 100% sureReported the incorrect release-note attribution and supplied the correction in #88.
doc PG 18 relnotes: clarify btree skip scan item
23 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #861 of 1 reviewer saw the final version

Exactly matches #86, replacing the skip-scan cardinality sentence with a description of later indexed columns.

  • Peter Geoghegansaw the final versionLLM analysis, 100% sureExplained the ambiguity between skip scan and loose index scan and requested a functional description in #85.
doc PG 18 relnotes: mv. hash joins and GROUP BY item to General
10 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

Moves the hashing entry from Optimizer to General Performance following #74; no attachment contains this relocation.

  • David Rowleysaw the final versionLLM analysis, 100% surePointed out that the hashing changes do not affect the planner and requested General Performance placement in #74.
doc PG 18 relnotes: fix missing parens for crc32c()
9 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

Adds the missing parentheses to crc32c exactly as requested inline in #71, with application confirmed in #73.

  • Steven Niusaw the final versionLLM analysis, 100% sureReported the missing function parentheses and supplied the corrected sentence in #71.
PG 18 relnotes: adjust RETURNING new/old item
8 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #671 of 1 reviewer saw the final version

Exactly matches #67, simplifying RETURNING old/new behavior and explaining that aliases can be renamed to avoid conflicts.

  • jian hesaw the final versionLLM analysis, 99% sureIdentified confusing alias terminology and proposed the replacement explanation adopted with small edits in #67.
doc PG 18 relnotes: adjust pg_log_backend_memory_contexts()
8 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #621 of 1 reviewer saw the final version

Exactly matches #62, correcting the function parentheses and adding the missing contributor names and documentation reference.

  • David Rowleysaw the final versionLLM analysis, 100% sureFound duplicate parentheses and inconsistent authorship versus commit references in #60, reporting both in #61.
doc PG 18 relnotes: add pg_log_backend_memory_contexts() mention
8 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #601 of 1 reviewer saw the final version

Exactly matches #60, including the initially incorrect duplicate parentheses and attribution subsequently fixed separately.

  • David Rowleysaw the final versionLLM analysis, 100% sureReported that the one-based numbering change also affects pg_log_backend_memory_contexts() in #59.
doc PG 18 relnotes: adjust pgbench per-script reporting item
7 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #582 of 2 reviewers saw the final version

Exactly matches both hunks of #58: the pgbench per-script description and missing links for the earlier GROUPING SETS fixes.

  • Yugo Nagatasaw the final versionLLM analysis, 100% sureExplained that failed, retried, and skipped transaction counts were newly available in per-script reports in #56.
  • Richard Guosaw the final versionLLM analysis, 90% sureSupplied the GROUPING SETS fix references in #55 whose missing links are completed by this commit.
doc PG 18 relnotes: mention GROUP SET fixes
7 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #571 of 1 reviewer saw the final version

Exactly matches #57, adding references and a warning that previously incorrect GROUPING SETS results have been fixed.

  • Richard Guosaw the final versionLLM analysis, 100% sureRequested coverage of longstanding incorrect-result bugs and identified the relevant changes in #55.
doc PG 18 relnotes: adjust partition planning item
7 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #533 of 3 reviewers saw the final version

Exactly matches the one-line planning-description correction in #53; David's later approval in #54 postdates this commit.

  • David Rowleysaw the final versionLLM analysis, 100% sureCorrected the pruning description in #44 and supplied a query without joins in #52, prompting the final general planning wording.
  • Amit Langotesaw the final versionLLM analysis, 94% sureExplained the planning improvement for queries that do not prune partitions in #45, supporting the final broader description.
  • jian hesaw the final versionLLM analysis, 90% sureReported the spelling error in the earlier version of this planning item in #47, before the wording-only follow-ups.
doc PG 18 relnotes: small adjustments regarding options
6 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #511 of 1 reviewer saw the final version

All jsonb null, pg_dump option spelling, negative-option history, and row-level security wording changes match #51.

  • jian hesaw the final versionLLM analysis, 100% sureReported each option and jsonb terminology problem corrected by the patch in #46.
doc PG 18 relnotes: move partition locking item to General Perf
6 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

Relocates partition locking to General Performance as requested in #45 and confirmed in #49; #48 changes wording but does not contain this move.

  • Amit Langotesaw the final versionLLM analysis, 100% sureExplained that locking is an executor improvement and requested moving it out of Optimizer in #45.
doc PG 18 relnotes: adjust partition items
6 May 2025Bruce Momjianmaster18beta2REL_19_STABLE18beta2REL_18_STABLE18beta2
LLM analysis, 100% surelanded as postedfrom patchset #483 of 3 reviewers saw the final version

Exactly matches #48, correcting both partition entries introduced by #36.

  • David Rowleysaw the final versionLLM analysis, 100% sureReported that the planning work was unrelated to pruning and that locking occurs during execution in #44.
  • Amit Langotesaw the final versionLLM analysis, 100% sureConfirmed the execution-time locking correction and explained both partition performance changes in #45.
  • jian hesaw the final versionLLM analysis, 100% sureReported the misspelling in the planning entry in #47, which this replacement wording removes.
doc PG 18 relnotes: reword OAuth item
5 May 2025Bruce Momjianmaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surelanded as postedfrom patchset #401 of 1 reviewer saw the final version

Exactly matches the OAuth naming, authentication-method wording, and library-description changes in #40.

  • Jacob Championsaw the final versionLLM analysis, 100% sureCorrected OAuth capitalization and explained the authentication-method and token-validation terminology in #39.
doc PG 18 relnotes: add mention of pg_stat_reset_backend_stats()
5 May 2025Bruce Momjianmaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surelanded as postedfrom patchset #381 of 1 reviewer saw the final version

Exactly matches #38, separately documenting reset behavior for per-backend I/O and WAL statistics.

  • Bertrand Drouvotsaw the final versionLLM analysis, 100% sureIdentified inconsistent reset-function coverage and proposed the separate I/O and WAL descriptions adopted in #38.
doc PG 18 relnotes: adjust hash item
5 May 2025Bruce Momjianmaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surelanded as postedfrom patchset #371 of 1 reviewer saw the final version

Exactly matches #37, extending the hash-join entry to other hashing operations and Jeff Davis's memory savings.

  • David Rowleysaw the final versionLLM analysis, 100% sureIdentified omitted hashing improvements and supplied a measured memory-usage comparison in #32.
doc PG 18 relnotes: split partition optimizer item into two
5 May 2025Bruce Momjianmaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surelanded as postedfrom patchset #362 of 2 reviewers saw the final version

Exactly matches #36, splitting the combined partition entry into separate planning and locking descriptions.

  • Amit Langotesaw the final versionLLM analysis, 100% sureIdentified the distinct planning and execution improvements and proposed separating their entries in #2.
  • David Rowleysaw the final versionLLM analysis, 100% sureSupported the split and supplied user-facing descriptions for both groups of changes in #31.
doc PG 18 relnotes: adjust COPY and REJECT_LIMIT items
5 May 2025Bruce Momjianmaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surelanded as postedfrom patchset #301 of 1 reviewer saw the final version

The complete diff matches #30 after repairing attachment character-encoding artifacts; no substantive committer edits appear.

  • Atsushi Torikoshisaw the final versionLLM analysis, 100% sureReported COPY terminology, REJECT_LIMIT semantics, logging scope, and option-capitalization problems in #24.
doc PG 18 relnotes: move and clarify constraint items
5 May 2025Bruce Momjianmaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surelanded as postedfrom patchset #291 of 1 reviewer saw the final version

Matches the constraint reordering and NOT NULL clarification in #29 exactly after repairing attachment character-encoding artifacts.

  • Alvaro Herrerasaw the final versionLLM analysis, 100% sureExplained the NOT NULL restriction and proposed the constraint ordering adopted in #29, while his proposed merged entry was declined.
doc PG 18 relnotes: add commit for cancel key and protocol neg.
5 May 2025Bruce Momjianmaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

Adds cancel-key and protocol-negotiation references and attribution requested in #22; #28 explains which requested reference was omitted.

  • Jelte Fennema-Niosaw the final versionLLM analysis, 100% sureIdentified missing supporting references and contributor credit for the protocol changes in #22.
doc PG 18 relnotes: fix libpq wording
3 May 2025Bruce Momjianmaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

Implements the client typo, libpq connection-parameter terminology, and missing temporal-constraint link reported in #13; no attachment contains this combined correction.

  • Jelte Fennema-Niosaw the final versionLLM analysis, 100% sureReviewed the protocol wording and reported the typo, incorrect libpq terminology, and absent temporal-constraint reference in #13.
doc PG 18 relnotes: add GROUP BY column elimination item
3 May 2025Bruce Momjianmaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surelanded as postedfrom patchset #121 of 1 reviewer saw the final version

Exactly matches the GROUP BY unique-index column-elimination entry attached in #12.

  • jian hesaw the final versionLLM analysis, 100% sureReported the omitted unique-index GROUP BY optimization and explained its relationship to existing primary-key handling in #10.
doc PG 18 relnotes: move protocol version item to "server"
3 May 2025Bruce Momjianmaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surelanded as postedfrom patchset #111 of 1 reviewer saw the final version

Exactly matches #11, moving protocol 3.2 coverage to cancellation keys and retaining the client typo fixed later.

  • Jelte Fennema-Niosaw the final versionLLM analysis, 100% sureReviewed #8 and requested server-section placement and wording applicable to non-libpq clients in #9.
doc PG 18 relnotes: adjust libpq trace & potocol version items
3 May 2025Bruce Momjianmaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surelanded as postedfrom patchset #81 of 1 reviewer saw the final version

The complete protocol-version and libpq-trace diff matches #8 after repairing attachment character-encoding artifacts.

  • Jelte Fennema-Niosaw the final versionLLM analysis, 100% sureReported the missing protocol 3.2 announcement and libpq tracing reference in #7.
doc PG 18 relnotes: reword and reorder items
2 May 2025Bruce Momjianmaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surewritten from the discussion, no matching patch1 of 1 reviewer saw the final version

Combines broad release-note editing with Jacob Champion's ssl_groups compatibility correction from #3; it predates the first attachment and has no matching submitted patch.

  • Jacob Championsaw the final versionLLM analysis, 100% sureExplained that the old SSL setting remains an alias and requested moving the entry out of incompatibilities in #3.

Credits across 36 commits

Author
Bruce MomjianLLM23
Yugo Nagata2
Dean Rasheed1
Michael Paquier1
Nathan Bossart1
Reviewer
Bruce Momjiancommitter5
Reported by
David Rowley7
Jelte Fennema-Nio4
jian he3
Peter Geoghegan3
Yugo Nagata3
Jacob Champion2
Melanie Plageman2
Alvaro Herrera1
Amit Langote1
Bertrand Drouvot1
Dean Rasheed1
Michael Paquier1
Nathan Bossart1
Noah Misch1
Peter Smith1
Richard Guo1
Steven Niu1
torikoshia1
vignesh C1
Committer
Bruce Momjian36

N LLM means N of the credits come from LLM analysis, not from commit trailers.