Jul 2026

1 commit
Doc: Clarify DROP SUBSCRIPTION behavior after SET (slot_name = NONE).
17 Jul 2026Amit KapilamasterunreleasedREL_19_STABLE19beta3REL_18_STABLE18.6REL_17_STABLE17.11REL_16_STABLE16.15REL_15_STABLE15.19REL_14_STABLE14.24
LLM analysis, 82% surelanded as postedfrom patchset #312 of 2 reviewers saw the final version

The doc/src/sgml/ref/drop_subscription.sgml change implements the wording agreed in #28 and #29, followed by the patch posting #31 and push confirmation #33. An unchanged landing is strongly indicated, but the extraction error prevents direct attachment-to-commit comparison.

  • Jeff Davissaw the final versionLLM analysis, 96% sureApproved Amit Kapila's proposed documentation wording in #29 before it was packaged as #31.
  • Nisha Moondsaw the final versionLLM analysis, 100% sureVerified that both branch variants in #31 applied and built successfully from master through PostgreSQL 14 in #32, with the omitted reviewer credit acknowledged in #33.

Jun 2026

2 commits
Avoid errors during DROP SUBSCRIPTION when slot_name is NONE.
18 Jun 2026Jeff Davismaster19beta2REL_19_STABLE19beta2
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #181 of 1 reviewer saw the final version

Matches the first attached patch in #18, including construct_subserver_conninfo(), delayed connection-string construction, and regression tests. The separate subtransaction-based error-handling patch was excluded, as confirmed in #19.

  • Chao Lisaw the final versiontrailerIdentified the tablesync-cleanup regression in #11 and reviewed the revised helper-based implementation in #15, after which this component received only light changes.
Avoid errors during ALTER SUBSCRIPTION.
17 Jun 2026Jeff Davismaster19beta2REL_19_STABLE19beta2
LLM analysis, 100% surepart of a posted patchsetmoderate changesfrom patchset #140 of 1 reviewer saw the final version

Lands only the ALTER component of #14, leaving both DROP patches outstanding. Beyond that subset, subscriptioncmds.c adds a DISABLE branch that skips old connection information, with corresponding test additions, explicitly reported in #18.

  • Chao Lisaw an earlier versiontrailerEndorsed the explicit-parameter design in #11, caught the assertion bug in #13, and approved its correction in #15, before the committer added DISABLE handling.

May 2026

1 commit
Check retain_dead_tuples for ALTER SUBSCRIPTION ... SERVER.
15 May 2026Jeff Davismaster19beta1REL_19_STABLE19beta1
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #121 of 1 reviewer saw the final version

Exactly matches the retain_dead_tuples component of #12 in subscriptioncmds.c: ALTER SUBSCRIPTION SERVER checks the publisher and uses new_conninfo. The other three patches were subsequently resent in #14.

  • Chao Lisaw the final versiontrailerExamined the retain_dead_tuples change and identified its interaction with the following ALTER patch's assertion in #13; the committed component itself remained unchanged.

Credits across 4 commits

Author
Jeff DavisLLM3
Amit Kapila1
Reviewer
Chao Li3
Reported by
Chao Li1
Hayato Kuroda (Fujitsu)1
Jeff Davis1
Committer
Jeff Davis3
Amit Kapila1

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