model-writtenrun 3390 - may be wrong, links go to the source messages. Commit-by-commit LLM analysis is in the Commits dialog.
- Tom Lane opened the thread after a report that postgres_fdw failed when a remote server rejected UTC, proposing either hard-wired server support for UTC, requesting GMT from postgres_fdw, or both #1.
- Laurenz Albe supported both changes for consistency and compatibility, Daniel Gustafsson considered the risk of accommodating broken installations low, and Etsuro Fujita agreed provided the postgres_fdw change stayed simple #2 #3 #4.
- Tom deferred server-side UTC handling because it would involve documentation and possibly changing the timezone default, making it unsuitable for backpatching or post-feature-freeze work; he suggested revisiting it when development opened for v18 #5.
- Tom attached a minimal postgres_fdw patch requesting GMT and documenting its hard-wired support, proposing backpatching because users might have no control over a remote server's broken timezone database #5.
- The attached change landed unchanged on April 21 in the supplied branch-equivalent commits ce1c30eceb and a3021aafce, whose messages specify backpatching to all supported branches; broader server-side UTC support remained deferred.