CLI for testing DOM timings built upon Headless Chrome & Lighthouse 💜
npm i -g teti
$ teti --help
Usage
$ teti <url>
Options
-n number of tests to run (10 is default)
--custom custom User Timing (performance.mark) to be measured
--verbose output all data
➜ teti google.com
Results for http://google.com based on 10 requests:
Timing median mean p95 σ² MAD
domLoading 0.2 0.2 0.22 0.18 0.01
domInteractive 0.29 0.29 0.32 0.36 0.02
domComplete 0.67 0.68 0.83 3.34 0.03
first-paint 0.35 0.35 0.4 0.44 0.01
➜ teti https://www.theguardian.com/football -n 30 --custom "commercial boot"
Results for https://www.theguardian.com/football based on 30 requests:
Timing median mean p95 σ² MAD
domLoading 0.16 0.16 0.17 0.03 0
domInteractive 0.68 0.67 0.8 7.73 0.07
domComplete 3.42 3.58 4.94 284.16 0.3
first-paint 0.34 0.34 0.38 0.58 0.01
commercial boot 0.87 0.84 1.06 17.73 0.1
By default teti collects the following timings:
Based on collected data, the following statistical metrics are calculated:
- median
- mean
- p95 - 95th percentile
- σ² - variance
- MAD - median absolute deviation
MIT © Michał Jezierski