Aug 2021

2 commits
doc: interval spill method for units greater than months
3 Aug 2021Bruce MomjianREL_14_STABLE14beta3REL_13_STABLE13.4REL_12_STABLE12.8REL_11_STABLE11.13REL_10_STABLE10.18REL9_6_STABLE9.6.23
LLM analysis, 100% surepart of a posted patchsetlight changesfrom patchset #961 of 3 reviewers saw the final version

Takes only the two doc/src/sgml/datatype.sgml hunks from #96, adapting older SGML context and changing “rounded” to “truncated” to describe existing stable-branch behavior. Bruce confirms this documentation-only application in #99.

  • Tom Lanesaw the final versionLLM analysis, 100% sureIn #90, he requested explicit documentation of truncation in older branches and corrected storage terminology and conversion wording, which informed #91.
  • Justin Pryzbysaw an earlier versionLLM analysis, 98% sureHe identified the need to update the documented behavior in #23 and challenged ambiguous fractional-input wording in #25, before subsequent changes to the documented spill policy.
  • Bryn Llewellynsaw an earlier versionLLM analysis, 88% sureHe challenged the explanation of fractional-year behavior in #36 and the claimed accuracy of conversions in #65, before the documentation was narrowed to existing spill behavior.
interval: round values when spilling to months
3 Aug 2021Bruce Momjianmaster15beta1REL_19_STABLE15beta1REL_18_STABLE15beta1REL_17_STABLE15beta1REL_16_STABLE15beta1REL_15_STABLE15beta1
LLM analysis, 100% surelanded as postedfrom patchset #962 of 7 reviewers saw the final version

The complete diff matches #96 after removing index metadata: datatype.sgml documentation and twelve rint() conversions across backend datetime.c and ECPG interval.c. The final attachment itself is an exact resend of #91.

  • Zhihong Yusaw an earlier versionLLM analysis, 100% sureHe found missing ECPG conversions in #21, tested the original reproducer in #34, and reviewed later rounding and regression changes in #55 and #67, before substantial scope reductions.
  • Justin Pryzbysaw an earlier versionLLM analysis, 99% sureIn #23 and #25, he identified compatibility and documentation problems with changing established fractional-unit spilling, before the implementation was substantially revised.
  • Bryn Llewellynsaw an earlier versionLLM analysis, 98% sureBeyond reporting the original bug, he examined proposed semantics and explicitly requested limiting the changes to the original defect in #72 and #77, before the larger changes were removed.
  • Dean Rasheedsaw an earlier versionLLM analysis, 98% sureHe demonstrated the fractional-week regression and discussed the corrected decade case in #59, then revisited arithmetic consistency in #79, before the final scope reduction.
  • Tom Lanesaw the final versionLLM analysis, 100% sureHe challenged broad behavior changes, supplied documentation corrections in #90, and examined the final rounding approach in #95, supporting rounding while questioning whether it fully solved order dependence.
  • Robert Haassaw the final versionLLM analysis, 99% sureHe objected to perceived behavior changes in #92 and sought clarification while attempting to apply the final-content patch in #94, without subsequently recording approval.
  • John W Higginssaw an earlier versionLLM analysis, 85% sureHe challenged the arithmetic guarantees of the proposed patches in #78 and supplied timestamp-addition counterexamples in #80, before the arithmetic changes were dropped.

Credits across 2 commits

Author
Bruce MomjianLLM2
Reported by
Bryn Llewelly1
Bryn Llewellyn1
Committer
Bruce Momjian2

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