model-writtenrun 4526 - may be wrong, links go to the source messages. Commit-by-commit LLM analysis is in the Commits dialog.
- David Rowley proposed standardizing user-facing documentation on “an SQL,” following its existing majority usage, while avoiding widespread code-comment changes because of backpatching costs #1.
- Peter Eisentraut supported that convention while distinguishing pronounceable abbreviations such as SASL and SHA; Dave Page proposed also fixing user-visible error messages, and Daniel Gustafsson supported that scope #2 #3 #4.
- David submitted a patch covering documentation, four error messages and abbreviations including SSPI, FSM, MCV and SRF, while leaving source comments and translator material alone #5.
- Alvaro Herrera supplied a multiline search for missed occurrences, and David confirmed that the documentation example was already covered and code comments remained outside the cleanup #7 #10.
- David proposed additional SASLResponse, SHA1 and MIT article changes inline; Isaac Morland corrected the MIT pronunciation assumption, and David accepted that correction before committing #8 #9 #21.
- Tom Lane questioned the backpatching cost and durability of the convention, while David defended consistency in public-facing text; Tom ultimately declined to block the change #11 #13 #14 #15.
- David committed the initial attachment with the SASLResponse and SHA1 additions and updated pgbench test expectations as 04539e73fa, then announced the push and promised periodic checks for new inconsistencies #5 #23.
- After the initial landing, Geoff Winkless, Michael Paquier and Andrew Dunstan discussed why “a SCRAM” remained appropriate, without producing another attached correction #25 #26 #27.
- The April 2022 follow-up 7bdd489d3d applied the established convention to four documentation occurrences, including new SQL/JSON text and older occurrences obscured by SGML tags, without a matching attachment in the supplied thread.
- In April 2023, David submitted three new ECPG wording corrections and Michael Paquier endorsed them; the attachment landed unchanged as d866f0374c #28 #29.
- In April 2024, David submitted one new correction in the dependency documentation and suggested checking closer to branch creation; it landed unchanged as b1b13d2b52, with the supplied discussion ending in his push confirmation and no attachment still awaiting commitment #30 #31.
- In April 2026, David continued the established cleanup with a78cf591a3, correcting one occurrence in doc/src/sgml/xfunc.sgml without a matching attachment or review in the supplied thread.