chore(web): jsdom 29.1.1 → 30.0.1 #256

Merged

dominik.polakovics merged 1 commit from afk/255 into main

2026-08-01 04:28:35 +02:00

dominik.polakovics commented

2026-08-01 04:17:26 +02:00

Owner

Bumps the web/ SPA's only landable npm major: jsdom 29.1.1 → 30.0.1 (devDependency, vitest test environment only). This is dependency hygiene, not a security fix — no advisory reaches any version in the tree.

Closes #255

What changed

web/package.json: "jsdom": "^29.1.1" → "^30.0.1"
web/package-lock.json: regenerated by npm install; jsdom resolves to exactly 30.0.1 and every lockfile delta is confined to jsdom's own dependency subtree (@asamuzakjp/css-color 6.0.5, @asamuzakjp/dom-selector 8.3.0, undici 8.9.0, nested whatwg-url 17.1.0, updated engines ranges).

No web/src/**, test, flake.lock, or nix/ changes — jsdom 30's sole breaking change is the Node floor (^22.22.2 || ^24.15.0 || >=26.0.0), which native CI (Node 24 ≥ 24.16.0) and the pinned nixpkgs rev (nodejs 24.16.0) both satisfy, as pre-verified in the issue.

Verification (run in this worktree, Node 24.18.1)

npm ci — clean; zero warnings, no EBADENGINE, 0 vulnerabilities
npm run lint — pass
npm run format:check — pass
npm test — 66 files / 975 tests, all pass, no test edited; ToolPanel.test.tsx (the getComputedStyle()-sensitive suite flagged in the issue) passes unchanged, as predicted — the 30.x getComputedStyle() changes touch length/function serialization, not the keyword overflow-x/y values this repo reads
npm run build (tsc --noEmit && vite build) — pass

nix flake check — not run, and here is the honest accounting

This environment has no nix and no root to install it, so the hermetic gate was not executed locally. It will also not trigger on this PR: ci-nix.yml path-filters to **/*.nix, flake.lock, go.mod, go.sum — by documented design (ADR-0023 §2, comment at the top of the workflow), because the web derivation uses importNpmLock and follows web/package-lock.json directly, npm dep changes cannot rot the nix build. The one way this bump could break the nix build — the Node engines floor — was checked in the issue against the pinned flake.lock rev: nixpkgs ships nodejs 24.16.0 ≥ 24.15.0. If you want the gate exercised anyway before merge, run nix flake check locally on this branch.

🤖 Generated with Claude Code

Bumps the `web/` SPA's only landable npm major: **jsdom 29.1.1 → 30.0.1** (devDependency, vitest test environment only). This is dependency hygiene, not a security fix — no advisory reaches any version in the tree. Closes #255 ## What changed - `web/package.json`: `"jsdom": "^29.1.1"` → `"^30.0.1"` - `web/package-lock.json`: regenerated by `npm install`; jsdom resolves to exactly 30.0.1 and every lockfile delta is confined to jsdom's own dependency subtree (`@asamuzakjp/css-color` 6.0.5, `@asamuzakjp/dom-selector` 8.3.0, `undici` 8.9.0, nested `whatwg-url` 17.1.0, updated engines ranges). No `web/src/**`, test, `flake.lock`, or `nix/` changes — jsdom 30's sole breaking change is the Node floor (`^22.22.2 || ^24.15.0 || >=26.0.0`), which native CI (Node 24 ≥ 24.16.0) and the pinned nixpkgs rev (nodejs 24.16.0) both satisfy, as pre-verified in the issue. ## Verification (run in this worktree, Node 24.18.1) - `npm ci` — clean; zero warnings, **no EBADENGINE**, 0 vulnerabilities - `npm run lint` — pass - `npm run format:check` — pass - `npm test` — **66 files / 975 tests, all pass, no test edited**; `ToolPanel.test.tsx` (the `getComputedStyle()`-sensitive suite flagged in the issue) passes unchanged, as predicted — the 30.x `getComputedStyle()` changes touch length/function serialization, not the keyword `overflow-x/y` values this repo reads - `npm run build` (`tsc --noEmit && vite build`) — pass ## nix flake check — not run, and here is the honest accounting This environment has no nix and no root to install it, so the hermetic gate was not executed locally. It will also not trigger on this PR: `ci-nix.yml` path-filters to `**/*.nix`, `flake.lock`, `go.mod`, `go.sum` — by documented design (ADR-0023 §2, comment at the top of the workflow), because the web derivation uses `importNpmLock` and follows `web/package-lock.json` directly, npm dep changes cannot rot the nix build. The one way this bump could break the nix build — the Node engines floor — was checked in the issue against the pinned `flake.lock` rev: nixpkgs ships nodejs 24.16.0 ≥ 24.15.0. If you want the gate exercised anyway before merge, run `nix flake check` locally on this branch. 🤖 Generated with [Claude Code](https://claude.com/claude-code)

dominik.polakovics added 1 commit

2026-08-01 04:17:26 +02:00

chore(web): jsdom 29.1.1 → 30.0.1

ci / native (pull_request) Successful in 9m54s

Details

135badd0ee

jsdom 30's sole breaking change is a Node.js floor bump to
^22.22.2 || ^24.15.0 || >=26.0.0, which both build environments already
satisfy: native CI pins Node 24 (>= 24.16.0) and the pinned nixpkgs rev
ships nodejs 24.16.0. No flake.lock change needed — the web derivation
uses importNpmLock, so the nix build follows package-lock.json directly.

The getComputedStyle() changes in 30.x affect length/function
serialization only; this repo's two call sites read keyword overflow
values and the full vitest suite (975 tests, ToolPanel.test.tsx
unchanged) passes under the new jsdom.

Closes #255

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>

dominik.polakovics commented

2026-08-01 04:28:31 +02:00

Author

Owner

[autoland] verdict: pass

dominik.polakovics merged commit e9ad99a7a6 into main

2026-08-01 04:28:35 +02:00

dominik.polakovics referenced this pull request from a commit

2026-08-01 04:28:36 +02:00

Merge pull request 'chore(web): jsdom 29.1.1 → 30.0.1' (#256) from afk/255 into main