#27 accurate benchmarking
–
- github
- github/openchronology/timeline-app
- github/openchronology/timeline-app#27
- issue
- state:closed
There should be db seeds for sqlite, postgres, and .ochx for large swaths of timeline data to better measure performance handing limits on those various platforms. The initial benchmark suite should measure latency based on variations in timeline size. There should be a different chart for each operation - CRUD for moments, changing a viewport, etc. The initial benchmark suite should simply test how the server replies from the postgresql database, sqlite db, and inline browser mapping first - benchmarking the user interface should come later. These different backings should be different color lines on the same chart, sharing the same y axis The benchmarks should also use benchmarking best practices - a library for benchmarking if possible - like multiple calls and measuring standard deviation, etc. https://github.com/openchronology/timeline-app/issues/27