Searching events reads the text index once, and durations through their own
- github
- github/openchronology/timeline-app
- commit
- main
- github/athanclark
The text match runs in a fenced subquery, so the planner no longer reads the whole index once per timeline; durations alone have a partial text index. At 1.76 million rows, searching only durations for a common word went from 955 ms to 58 ms. docs/benchmarks.md has the results of bench/library-search.mjs. Co-Authored-By: Claude Opus 5.5 Claude-Session: https://claude.ai/code/session_011fFY8eruEi1xAkkMBni13B https://github.com/openchronology/timeline-app/commit/8aa24f89fb4f4b13000b8f517e5467516d2f4fd6
Open Searching events reads the text index once, and durations through their own in the timeline