← product evidence + the ladder to state of the art

Nishi browser vs Chrome — visual parity benchmark (first honest run)

workstream: S21 experiential / S18 graphics — browser chrome-parity ratchetrun: browser-parity-2026-07-162026-07-16

Verdict: NOT YET MATCHED — and that is the honest baseline. Layout height: /experiential 64% of Chrome, Wikipedia 111%. Paint parity: 95.8%/81.7% coarse, 56.5%/45.1% tight. Two root bugs found by this benchmark were fixed and gate-proven today (3x row bloat from a fractional font-size parse; white void below dark pages). The instrument itself is gate-proven (identical=100%, half=50%, oracle-vs-itself=100.0% ceiling).
evidence
/experiential first screen — Chrome (left) vs the Nishi browser (right), same 1024px viewport, same live page. Structure matches (title, intro, verdict box, census table). Visible Nishi gaps: rougher typography, missing pill shapes, inline-label overlap, tighter table rows. open full size
evidence
Wikipedia "Benchmark (computing)" — Chrome (left) vs Nishi (right). The Nishi browser fetches, lays out and paints a real Wikipedia article: body text, blue links, the hatnote box WITH its warning-icon PNG decoded, and the OGRE screenshot JPEG decoded and painted. Visible gaps: inline links break text flow (spacing), the image sits left and overlaps text (no float), Wikipedia skin chrome is not rendered. open full size
evidence
Eaglercraft (Minecraft) launcher — Chrome oracle only. The Nishi browser fetched the page (131KB, 1s) and laid it out (under 20s) but could not FINISH PAINTING it within 170s (paint-time CSS cascade cost on Next.js-scale stylesheets). That is the honest current gap; the fix (per-box computed-decl index) is queued with evidence. open full size

Measured

whatresultmeaning
/experiential layout height1418px vs Chrome 2227px (64%)We now render DENSER than Chrome (missing padding/margin shorthands + line-height); before today's font-size fix we were 206% (4580px)
/experiential paint parity95.8% coarse (16x12, tol 32) · 56.5% tight (64x48, tol 8)Coarse = structure/colors match; tight = font raster + spacing differences dominate
Wikipedia layout height8428px vs Chrome 7582px (111%)Within 11% of Chrome on a real complex article
Wikipedia paint parity81.7% coarse · 45.1% tightDominant delta = inline link flow (measured: layout reserves bitmap-font widths, paint draws narrower vector font)
Calibration ceilingChrome-vs-itself = 100.0%The instrument is capture-stable; every point below 100 is real difference
Sensitivity liar-killerFirst-screen parity scored the PRE-fix browser HIGHER (96.3 vs 95.8)Proved single-axis parity lies; the benchmark is multi-axis by law (height + paint)
Root fix 1 (today)font-size "13.5px" parsed as 135px -> 3x-tall rowsFixed at the parser; nx_css_fontsize_frac_gate GREEN 4/4; page height 4580 -> 1418
Root fix 2 (today)white void below dark pagesCSS background propagation; frame-7 luminance 236 -> 15; no regression (bg-cascade gate GREEN)
Usability study gradeFAIL 25% -> CONDITIONAL 75%, 0 major findingsNN/g task-based grading (nx_ux_study_gate); PASS still refused while the inline-flow overlap remains
OraclesChrome + Edge captured; Firefox NOT INSTALLED on this machineEdge output byte-class identical to Chrome (same engine); Firefox pending install

Machine detail: api.json · related: /experiential