The stroke-width saturation rung: direct overflow lands, precision aliases stay split

[softmarshmallow]

Outcome

This is the stroke-width percentage-saturation capability rung.

• Four Chromium 149 cells land: attribute/CSS twins for the main saturation
  matrix and for the cap/transform matrix. The primitive corpus moves
  280 → 284; all four additions are exact and add no tolerance.
• An unambiguous direct extreme percentage now resolves to the maximum finite
  width and crosses rframe unchanged. Its ink remains painter-shaped: join,
  miter limit, cap, transform, topology, and an ordinary dash cycle all matter.
• The former broad percentage-overflow refusal graduates. A narrower
  stroke-width percentage precision alias refusal replaces it, so the
  refusal register remains at 56.
• Both stroke-width checklist twins deliberately remain unchecked. Valid
  Chromium-honored source values can collapse to one pinned-Stylo percentage
  fact, and that class has no own checklist row: this is the The stroke-paint rung: the paint grammar celled as twins, the context refusal proven load-bearing #81
  SPLIT outcome, not a full-grammar tick.

Measured verdict

Chromium evaluates a pure percentage in the browser's percentage path, with
the authored percentage multiplied by the normalized-diagonal basis before
division by 100. Positive intermediate overflow becomes the maximum finite
width; it is not semantic stroke absence.

The discriminating matrix establishes:

• an unambiguous finite 5e36% branch and direct saturated 3.4e38% branches;
• miter limit 2 paints while the next representable value drops;
• round/bevel, dash, native box/ellipse, open-corner, and closed-contour routes;
• round and square caps drop under one-axis compression but both paint under
  two-axis compression, each with an independently visible pass;
• attribute and CSS source twins are pixel-identical.

The remaining source-precision gap is also measured rather than inferred.
Adjacent authored values can share a computed Stylo bucket while Chromium
gives them distinct used widths. A non-f32 decimal around 57384.265625%
differs by 864 pixels, and non-identity percentage math reaches further
results after Stylo erases its operation history—including length terms that
cancel to zero before a pure percentage remains. These cases refuse by the
stable precision-alias name across attribute, inline, stylesheet, and inherited
ingresses. Identity calc(N%) remains admitted; genuinely mixed
length/percentage values retain their established mixed-calc refusal.

Contract and consumer

rframe continues to carry the exact finite f32 width, cap, join, miter, and
dash facts. Only the derived direction-free coverage reach widens to a finite
f64 value; the backend-free architecture guard remains green.

n0 keeps painter input unchanged. Its damage projection computes wide reach,
also encloses sequential-f32 affine arithmetic, intersects the finite frame,
then outward-encodes RectF. Rect/Ellipse coverage additionally encloses the
actual composed box world (0..w / 0..h); Path keeps its absolute-coordinate
route. Exact cancellation and box-composition bit regressions guard both seams.

The generic damage-union representation remains separately tracked by
#87. The pre-existing fill-only box-world analogue found by this
audit is #88. The text font-basis patrol remains
#82.

Scratch probe testimony

Run from fixtures/web-first/ under the pinned Chromium capture posture:

just probe /private/tmp/stroke-width-extreme-percentage-probe.ts
just probe /private/tmp/extreme-width-architecture-probe.ts
just probe /private/tmp/stroke-width-percentage-precision-probe.ts
just probe /private/tmp/stroke-width-percentage-saturation-discriminator-audit.ts
just probe /private/tmp/stroke-width-detector-false-negative.ts
just probe /private/tmp/stroke-width-mixed-zero-provenance.ts

The final discriminator SHA-256 is
5270450f65f7e29ebb0a8e9d8730d89a0a0991fb423f5a551addb9174533e15f;
the raw-f64 and mixed-zero matrices are respectively
0e13545f3b02ecd0870c4c2a6fd6fdff2fab84f0c59320f51df61ccc17a65ba3
and
c3cc70ca099462e3906d0019fc724fa9d773ccb0a5a5c50ae26f0b3c4d509ab5.
All reproduce under Chromium 149.0.7827.55.

Verification

• just bake — 284/284 verified, never-overwrite
• just gate — 4/4
• just status — fresh and idempotent
• cargo test -p rframe -p n0 -p websem
• cargo test -p n0_cli
• cargo check -p rframe -p n0 -p websem -p n0_cli
• focused clippy for rframe tests plus n0/websem libraries with warnings denied
• saved verify-rung roles rerun after the grown fixes: TICK+LAW PASS, REPRO PASS
• git diff --check

No checklist/FLIP file changes, no conformance score produced or inspected.

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Actions	Updated (UTC)
nothing	Ready	Preview	Aug 20, 2026 11:09am

[coderabbitai]

Warning

Review limit reached

@softmarshmallow, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 49 minutes

Limit details: You’ve used the included review currently available.

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

How can I continue?

Wait for the limit to reset, then comment @coderabbitai review or push new commits to the PR.

An organization admin can change what happens after included review limits in Billing.

How do review limits work?

CodeRabbit enforces per-developer PR review limits within each organization.

For paid Pro and Pro+ reviews, CodeRabbit uses a developer's included PR review attempts over the past 7 days to set the current hourly allowance. At typical activity levels, the full plan allowance applies. Higher sustained activity can lower the allowance until earlier attempts leave the 7-day window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 5e79f70d-08d2-4f7d-86af-95f82ee6fa7e

📥 Commits

Reviewing files that changed from the base of the PR and between 07694a6 and ebf6509.

⛔ Files ignored due to path filters (10)

• fixtures/web-first/chromium/svg-stroke-width-percentage-saturation-attr.png is excluded by !**/*.png
• fixtures/web-first/chromium/svg-stroke-width-percentage-saturation-cap-attr.png is excluded by !**/*.png
• fixtures/web-first/chromium/svg-stroke-width-percentage-saturation-cap-css.png is excluded by !**/*.png
• fixtures/web-first/chromium/svg-stroke-width-percentage-saturation-css.png is excluded by !**/*.png
• fixtures/web-first/svg-stroke-width-percentage-saturation-attr.svg is excluded by !**/*.svg
• fixtures/web-first/svg-stroke-width-percentage-saturation-cap-attr.svg is excluded by !**/*.svg
• fixtures/web-first/svg-stroke-width-percentage-saturation-cap-css.svg is excluded by !**/*.svg
• fixtures/web-first/svg-stroke-width-percentage-saturation-css.svg is excluded by !**/*.svg
• fixtures/web-first/unsupported/svg-stroke-width-percentage-overflow.svg is excluded by !**/*.svg
• fixtures/web-first/unsupported/svg-stroke-width-percentage-precision-alias.svg is excluded by !**/*.svg

📒 Files selected for processing (14)

• crates/n0/src/glyphless.rs
• crates/n0_cli/README.md
• crates/rframe/README.md
• crates/rframe/src/stroke.rs
• crates/rframe/tests/stroke_reach.rs
• crates/websem/src/svg.rs
• crates/websem/tests/strokes_contract.rs
• crates/websem/tests/unsupported_corpus.rs
• docs/wg/consolidation/svg-engine-of-record.md
• fixtures/web-first/README.md
• fixtures/web-first/STATUS.md
• fixtures/web-first/oracle-bake.json
• fixtures/web-first/primitives.json
• fixtures/web-first/unsupported/README.md

✨ Finishing Touches 💡 1

🛠️ Fix failing CI checks 💡

• Create stacked PR
• Commit on current branch

📝 Generate docstrings

• Create stacked PR
• Commit on current branch

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch rung/stroke-width-percentage-saturation

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}