A citation-ready description of how we evaluate client-side JSON tools. This page does not invent comparative scores — publish measured results only after running the protocol below.
Goal
Compare time-to-readable output and correctness for format / validate workflows on representative payloads, entirely in the browser.
Fixtures
- Small object (~1 KB) with nested arrays
- Medium API dump (~100 KB) with repeated keys
- Large pretty-printed document (~1 MB) for scroll/perf feel
- Deliberately invalid JSON for error messaging quality
Protocol
- Cold load the tool URL in a clean profile (Chrome latest, desktop).
- Paste fixture; measure time from paste to valid pretty-print (stopwatch or Performance marks).
- Confirm round-trip: minify → format preserves semantic equality.
- Record whether invalid input shows a precise parse error (line/column).
- Note whether payload leaves the device (network panel).
Reporting rules
- Publish raw timings and environment; no single composite “score” without formula.
- Disclose that code.live ships a JSON tool under test when comparing.
- Link this methodology from any future results post.
Related: JSON formatter · JSON topic hub · Editorial standards