Feb 2025

1 commit
Add PGErrorVerbosity to typedefs.list
18 Feb 2025Daniel Gustafssonmaster18beta1REL_19_STABLE18beta1REL_18_STABLE18beta1
LLM analysis, 100% surepart of a posted patchsetsignificant changesfrom patchset #611 of 1 reviewer saw the final version

The commit diff exactly matches the first patch in #61: the PGErrorVerbosity alignment fix in src/include/utils/elog.h and its addition to src/tools/pgindent/typedefs.list. The substantive backtrace-control patch is omitted, as Daniel confirms in #62, making this a significant reduction in series scope despite the retained patch being unchanged.

  • Daniel Gustafssonsaw the final versiontrailerDaniel committed the cleanup and confirmed its application in #62, establishing final coverage of the retained patch from #61.

Apr 2024

1 commit
Revert "Add GUC backtrace_on_internal_error"
29 Apr 2024Peter Eisentrautmaster17beta1REL_19_STABLE17beta1REL_18_STABLE17beta1REL_17_STABLE17beta1revert
LLM analysis, 100% surewritten from the discussion, no matching patch0 of 5 reviewers saw the final version

No attachment implements this revert: the commit removes backtrace_on_internal_error from doc/src/sgml/config.sgml, elog.c, guc_tables.c, and guc.h. It implements the revert consensus recorded in #51, #53, and #55, with completion confirmed in #56.

  • Jelte Fennema-Niosaw an earlier versionLLM analysis, 94% sureJelte argued for reverting after rejecting unconditional backtraces and finding no naming consensus in #47, before the renewed design that became #65.
  • Robert Haassaw an earlier versionLLM analysis, 97% sureRobert rejected the two-value enum proposal in #39 and identified consensus for the eventual revert in #51, before the later redesign in #65.
  • Tom Lanesaw an earlier versionLLM analysis, 98% sureTom identified upgrade risks from changing a released GUC in #41 and explicitly supported a full revert in #53, before the later redesign in #65.
  • Michael Paquiersaw an earlier versionLLM analysis, 98% sureMichael analyzed the internal-error backtraces generated by tests and favored reverting in #50, then confirmed agreement in #55, before the later redesign in #65.
  • Andres Freundsaw an earlier versionLLM analysis, 94% sureAndres raised security and performance objections to backtraces without a disable control in #52 and accepted the revert reasoning in #54, before the later redesign in #65.

Credits across 2 commits

Author
Jelte Fennema-Nio1
Reviewer
Daniel Gustafsson1
Committer
Daniel Gustafsson1
Peter Eisentraut1