model-writtenrun 2436 - may be wrong, links go to the source messages. Commit-by-commit LLM analysis is in the Commits dialog.
- Euler Taveira proposed a separate max_replication_origins setting, initially defaulting to 10, because sharing max_replication_slots confused subscriber configuration and prevented independent capacity limits #1.
- Peter Eisentraut raised deployment and automation compatibility concerns #2, prompting Euler to add a -1 default that inherited max_replication_slots #3.
- Keisuke Kuroda questioned the parameter name #4, and Masahiko Sawada explained that the limit governed origins set up for tracking rather than all defined origins, while finding an unchanged error message and an inappropriate header include #5.
- Euler corrected those issues and pg_createsubscriber documentation #6, then adopted Amit Kapila's proposed max_replication_origin_sessions name #7 in the next revision #10.
- Masahiko identified another subscriber-upgrade documentation reference needing replacement #11; his earlier observations about checkpoint error terminology and startup behavior with a zero limit remained outside this patch #5.
- Amit argued that compatibility fallback would perpetuate confusion and maintenance burden #12, while Euler preferred a clean break with a default of 10 #13; Masahiko and Amit agreed that the existing updated pg_upgrade check addressed upgrade risks #15 #16.
- Peter challenged the meaning of replication origin sessions and the limit's scope #17 #21; Amit suggested max_active_replication_origins #18, Euler endorsed it and clarified that the limit was cluster-wide #19 #22, and the revised submission included an optional patch removing compatibility fallback #24.
- Vignesh C found that the renamed setting changed sorted query-result positions in pg_createsubscriber and repeated the missing upgrade-documentation correction #26; Euler fixed both in the next series #27.
- Vignesh requested fallback testing and clarification for manually managed origins #28, but Euler considered the existing assertion sufficient and deferred broader replication-origin documentation to keep the patch focused #29.
- Masahiko requested better grouping in postgresql.conf.sample #30, Euler reordered the setting #31, and Masahiko then squashed both patches into a single submission with a revised commit message, selecting the independent default of 10 #32.
- Amit identified a remaining documentation sentence made inaccurate by separating the settings #33, Masahiko proposed changing Most to These #34, and Amit approved that wording #35.
- Masahiko committed 04ff636cbc from #32 with that documentation correction and a shortened sample-configuration comment, then confirmed completion in the thread's final message #36; the entire submitted feature landed.