BatchWebP Testing Methodology
Short answer: BatchWebP uses two verification layers. Static validation builds every public route and checks metadata, canonical URLs, headings, internal links, sitemap membership, PWA assets, and content regression rules. Browser end-to-end tests then upload real fixtures, wait for conversion, inspect output filenames and MIME types, fetch generated blobs, test previews, presets, Target KB, resizing, per-image quality, ZIP and CSV downloads, safety limits, error recovery, and legacy redirects.
Method last reviewed: August 2026. Test results describe the checked build and environment, not every browser/device combination.
Static site checks
- Exactly one title, meta description, canonical, H1, and manifest link per HTML page.
- Unique titles, descriptions, and canonical URLs across public pages.
- No broken internal page or asset links in the generated output.
- Only supported canonical routes in the sitemap.
- Regression guards for factual limits, platform disclaimers, and high-value guide structures.
Browser conversion checks
The end-to-end suite uses real JPG, PNG, WebP, GIF, AVIF, SVG, ICO, and BMP fixtures where the environment supports creation and decoding. For each public conversion route it verifies the generated extension, blob MIME, non-zero bytes, and visible preview. It also checks mixed batches, duplicate ZIP names, individual downloads, CSV content, queue clearing, damaged files, HEIC guidance, and oversized-file rejection.
What is not measured automatically?
| Not covered | Required follow-up |
|---|---|
| Perceived image quality | Human comparison at intended display size. |
| All browser/OS combinations | Manual compatibility matrix on target environments. |
| Production Core Web Vitals | CrUX, Search Console, or real-user monitoring after deployment. |
| Security certification | Independent review appropriate to the organization. |
| Platform upload behavior | Test inside the actual CMS, store, theme, plugin, and hosting stack. |
How recommendations are written
Product facts are tied to current source behavior. Platform claims link to primary documentation. Preset dimensions and quality values are labeled as BatchWebP starting points rather than universal requirements. Future benchmarks must disclose fixture types, sample count, browser version, operating system, date, input dimensions, output settings, and both median and outlier results.
Understand the CSV report or review browser-dependent behavior.