Repository navigation
Conversation
## Description Updated Greek translation. ## Specific Changes proposed - Added some missing translations on el.json. - Updated translations-needed.md. ## Requirements Checklist - [ ] Feature implemented / Bug fixed - [ ] If necessary, more likely in a feature request than a bug fix - [ ] Change has been verified in an actual browser (Chrome, Firefox, IE) - [ ] Unit Tests updated or fixed - [ ] Docs/guides updated - [ ] Example created ([starter template on JSBin](https://codepen.io/gkatsev/pen/GwZegv?editors=1000#0)) - [ ] Has no DOM changes which impact accessiblilty or trigger warnings (e.g. Chrome issues tab) - [ ] Has no changes to JSDoc which cause `npm run docs:api` to error - [ ] Reviewed by Two Core Contributors Co-authored-by: Adam Waldron <acwald@gmail.com>
## Description Update VHS to v3.14.1 for a VTT timing fix detailed here: videojs/http-streaming#1537 ## Specific Changes proposed Update the VHS package to v.3.14.1. ## Requirements Checklist - [x] Feature implemented / Bug fixed - [ ] If necessary, more likely in a feature request than a bug fix - [ ] Change has been verified in an actual browser (Chrome, Firefox, IE) - [ ] Unit Tests updated or fixed - [ ] Docs/guides updated - [ ] Example created ([starter template on JSBin](https://codepen.io/gkatsev/pen/GwZegv?editors=1000#0)) - [ ] Has no DOM changes which impact accessiblilty or trigger warnings (e.g. Chrome issues tab) - [ ] Has no changes to JSDoc which cause `npm run docs:api` to error - [ ] Reviewed by Two Core Contributors --------- Co-authored-by: Essk <sarah@teaandbiscuits.net>
## Description Double clicking the document pip window requests fullscreen but this fails. The gesture on the pip window can't initiate fullscreen. Fixes #8877 ## Specific Changes proposed Change the default double click action to just exit pip. This is the same behaviour as before without the error, as requestFullScreen() calls exitPictureInPicture(); --------- Co-authored-by: Gary Katsevman <git@gkatsev.com>
## Description It is useful to have methods for appending and removing `<source>` elements to the `<video>` element, as they are sometimes required to enable certain playback features, for example, using [Airplay with MSE](https://webkit.org/blog/15036/how-to-use-media-source-extensions-with-airplay). ## Specific Changes proposed Add new methods-- `addSourceElement()` and `removeSourceElement()` to the player and tech. The former will take a source object and create and append a new `<source>` element to the `<video>` element, and the latter will take a source url and remove any `<source>` element with a matching `src`. ## Requirements Checklist - [ ] Feature implemented / Bug fixed - [ ] If necessary, more likely in a feature request than a bug fix - [ ] Change has been verified in an actual browser (Chrome, Firefox, IE) - [ ] Unit Tests updated or fixed - [ ] Docs/guides updated - [ ] Example created ([starter template on JSBin](https://codepen.io/gkatsev/pen/GwZegv?editors=1000#0)) - [ ] Has no DOM changes which impact accessiblilty or trigger warnings (e.g. Chrome issues tab) - [ ] Has no changes to JSDoc which cause `npm run docs:api` to error - [ ] Reviewed by Two Core Contributors
Localization is not applied correctly in fieldset labels and select options. As a result, the text track setting modal dialog is only half translated. - add `localize` at `label` level in `TextTrackFieldset` - add `localize` at `option` level in `TextTrackSelect` - add test cases
Fixes that Escape being pressed isn't closing menus due to incorrect `event.key` name.
Change examples to use https and not http. Co-authored-by: mister-ben <1676039+mister-ben@users.noreply.github.com>
Generated CHANGELOG for v8.19.2 re-added some older commits. Removes these. Only changes from v8.19.1 are these: videojs/video.js@v8.19.1...v8.19.2 --------- Co-authored-by: André M. <34163393+amtins@users.noreply.github.com>
## Description n/a ## Specific Changes proposed n/a ## Requirements Checklist - [ ] Feature implemented / Bug fixed - [ ] If necessary, more likely in a feature request than a bug fix - [ ] Change has been verified in an actual browser (Chrome, Firefox, IE) - [ ] Unit Tests updated or fixed - [ ] Docs/guides updated - [ ] Example created ([starter template on JSBin](https://codepen.io/gkatsev/pen/GwZegv?editors=1000#0)) - [ ] Has no DOM changes which impact accessiblilty or trigger warnings (e.g. Chrome issues tab) - [ ] Has no changes to JSDoc which cause `npm run docs:api` to error - [ ] Reviewed by Two Core Contributors
## Description On desktop, a user can hover over the progress bar while content plays, which makes it possible to seek to a relatively precise location without disrupting playback. On mobile there is no hovering, so in order to seek during inline playback the user can only tap a location on the progress bar (very hard to do precisely on a small screen) or scrub to try to hone in on a specific location (can be very clunky because seeks are constantly being executed). This PR adds a feature to treat scrubbing on mobile more like hovering on desktop-- while scrubbing, seeks are disabled and playback continues, only when the user finishes scrubbing is a single seek executed to the desired location. One key use-case for this feature is thumbnail seeking integrations on mobile, where the user can scrub through different thumbnail images until they find their desired seek location. ## Specific Changes proposed This behavior is similar to the existing `enableSmoothSeeking` behavior in that the `PlayProgressBar` slider visibly updates with the scrubbing movements, but differs in a few ways: - Playback continues while scrubbing, no seeks are executed until `touchend`. - The seek bar's `TimeTooltip` component displays the target seek time while scrubbing, rather than the `CurrentTimeDisplay` (which continues to show the current time of the playing content).
bump vhs version (3.16.1) Co-authored-by: Dzianis Dashkevich <ddashkevich@brightcove.com>
## Description - Add "Playing in Picture-in-Picture" translations to zh-TW.json - Deduplicate object keys in zh-TW.json - Update translations-needed.md - Auto-generated - The np.json section removed from translations-needed.md was a duplicate. ## Specific Changes proposed - Add missing translation for "Playing in Picture-in-Picture" to zh-TW locale. - Remove duplicate keys 'Picture-in-Picture' and 'Exit Picture-in-Picture' from zh-TW.json to eliminate redundancy. ## Requirements Checklist - [ ] Feature implemented / Bug fixed - [ ] If necessary, more likely in a feature request than a bug fix - [x] Change has been verified in an actual browser (Chrome, Firefox, IE) - [x] Docs/guides updated - [x] Has no DOM changes which impact accessiblilty or trigger warnings (e.g. Chrome issues tab) - [x] Has no changes to JSDoc which cause `npm run docs:api` to error - [ ] Example created ([starter template on JSBin](https://codepen.io/gkatsev/pen/GwZegv?editors=1000#0)) - [ ] Reviewed by Two Core Contributors --------- Co-authored-by: Harisha Rajam Swaminathan <35213866+harisha-swaminathan@users.noreply.github.com>
## Description Update for the translation Co-authored-by: Harisha Rajam Swaminathan <35213866+harisha-swaminathan@users.noreply.github.com>
Do not allow to register new Player component, if any instance of the current component still exists. Fixes #8925 Co-authored-by: mister-ben <1676039+mister-ben@users.noreply.github.com>
## Description
Pressing the arrow and page keys seeks, currently by a set interval.
This change makes this configurable.
## Specific Changes proposed
Adds options that can be included in player options which set the
increments to used.
```
{
controlBar: {
progressControl: {
seekBar: {
stepSeconds: 2,
pageMultiplier: 4
}
}
}
}
```
Defaults remain as 5 seconds for left/right and 12x multiplier for page
up/down.
## Requirements Checklist
- [x] Feature implemented / Bug fixed
- [ ] If necessary, more likely in a feature request than a bug fix
- [x] Change has been verified in an actual browser (Chrome, Firefox,
IE)
- [x] Unit Tests updated or fixed
- [ ] Docs/guides updated
- [ ] Example created ([starter template on
JSBin](https://codepen.io/gkatsev/pen/GwZegv?editors=1000#0))
- [x] Has no DOM changes which impact accessiblilty or trigger warnings
(e.g. Chrome issues tab)
- [x] Has no changes to JSDoc which cause `npm run docs:api` to error
- [ ] Reviewed by Two Core Contributors
…n (#8911) ## Description - Configure GitHub Actions workflow for secure publishing - Enable automatic provenance generation during npm publish - Add integrity verification through Sigstore transparency logs Following the recent Lottie-Player supply chain attack, it's crucial to enhance package security. NPM provenance provides cryptographic proof that this package was built from this repository using GitHub Actions, making supply chain attacks significantly harder. More info in my blog post https://medium.com/exaforce/npm-provenance-the-missing-security-layer-in-popular-javascript-libraries-b50107927008 ## Specific Changes proposed Changes the workflow github to publish provenance attestation on https://www.npmjs.com/package/video.js ## Requirements Checklist - [x] Feature implemented in CI/CD - [ ] If necessary, more likely in a feature request than a bug fix - [ ] Change has been verified in an actual browser (Chrome, Firefox, IE) - [x] Unit Tests updated or fixed - [ ] Docs/guides updated - [ ] Example created ([starter template on JSBin](https://codepen.io/gkatsev/pen/GwZegv?editors=1000#0)) - [ x Has no DOM changes which impact accessiblilty or trigger warnings (e.g. Chrome issues tab) - [x] Has no changes to JSDoc which cause `npm run docs:api` to error - [ ] Reviewed by Two Core Contributors
Problem: the vjs-user-inactive class triggering the hiding of the tooltip is added much later on touch devices, therefore the mouse tooltip stays visible on its last seek target position whenever the user taps the screen to access the controls. This change ensures that the mouse tooltip is only visible when the user is actually scrubbing.
## Description Projects using newer versions of sass which have deprecated `lighten()` will get deprecation warnings. This updates sass, and switches to `color.adjust()`. This function is available since dart sass 1.23.0. Note this would increase the output CSS size, as it will output different colour syntax, .e.g. ``` < background-image: linear-gradient(-180deg, #fff 88%, rgb(114.9141509434, 132.7028301887, 159.3858490566) 100%); --- > background-image: linear-gradient(-180deg, #fff 88%, #73859f 100%); ``` Closes #8876 ## Requirements Checklist - [x] Feature implemented / Bug fixed - [ ] If necessary, more likely in a feature request than a bug fix - [x] Change has been verified in an actual browser (Chrome, Firefox, IE) - [x] Has no DOM changes which impact accessiblilty or trigger warnings (e.g. Chrome issues tab) - [x] Has no changes to JSDoc which cause `npm run docs:api` to error - [ ] Reviewed by Two Core Contributors
…579) Co-authored-by: Casey Occhialini <1508707+littlespex@users.noreply.github.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
…ource (#9605) Co-authored-by: Claude <noreply@anthropic.com>
… available (#9624)
v10 now lives in videojs/video.js; sync the MoQ branch to its main (6f56667). Conflict resolutions and follow-ups: - spf exports/entries: keep `./moq` beside upstream's `adapters/*` paths. - tsconfig: the single `playback/engines` project now covers the MoQ engine; drop its standalone tsconfig. - measure-size.js: take upstream's removal (workspace measurement). - track-switching / hls-video media-tracks: keep both sides. - MoQ engine: replace the removed `shareSignals`/`onSignalsReady` with `defineExternalSignals`; the adapter and tests read `engine.state`/ `engine.context` directly, and `MoqEngineSignals` is gone. - MoQ behaviors import `defineBehavior` from `define-behavior`. Claude-Session: https://claude.ai/code/session_01DaH4iixnBQbNchnez2h4ZG
The API-stability check from videojs/video.js requires an @internal tag on public exports no reference page documents. Claude-Session: https://claude.ai/code/session_01DaH4iixnBQbNchnez2h4ZG
Codex Review SummaryThis comment shows the latest Codex review activity on this pull request.
ℹ️ About Codex in GitHubYour team has set up Codex to review pull requests in this repo. Reviews are triggered when you
Codex reacts with 👀 while any review is running, comments if it has suggestions, and reacts with 👍 once all reviews finish with no findings. |
📦 Bundle Size Report🎨 @videojs/html
Small changes (43, ≤ 300 B)
Presets (7)
Media (20)
Players (5)
Skins (53)
UI Components (62)
⚛️ @videojs/react
Small changes (27, ≤ 300 B)
Presets (7)
Media (23)
Extensions (2)
Players (5)
Skins (34)
UI Components (39)
🧩 @videojs/core
Small changes (44, ≤ 300 B)
Entries (76)
🏷️ @videojs/element — 1 small size change
Entries (2)
📦 @videojs/store — 1 small size change
Entries (3)
🔧 @videojs/utils — 1 small size change
Entries (13)
📦 @videojs/cdn — 1 small size change
📦 @videojs/cloudflare-video — 1 small size change
📦 @videojs/dash-video — 1 small size change
📦 @videojs/google-cast
📦 @videojs/hlsjs-video
📦 @videojs/media
Small changes (1, ≤ 300 B)
Entries (3)
📦 @videojs/mux — 1 small size change
📦 @videojs/mux-audio
Entries (2)
📦 @videojs/mux-data
📦 @videojs/mux-video
Entries (2)
📦 @videojs/native-hls-video
📦 @videojs/playerjs-video
📦 @videojs/shaka-video — 1 small size change
📦 @videojs/spf
Small changes (3, ≤ 300 B)
Entries (63)
📦 @videojs/spotify-audio — 1 small size change
📦 @videojs/tiktok-video — 1 small size change
📦 @videojs/twitch-video — 1 small size change
📦 @videojs/other
Small changes (2, ≤ 300 B)
Entries (3)
📦 @videojs/vimeo-video — 1 small size change
📦 @videojs/wistia-video — 2 small size changes
Entries (2)
📦 @videojs/youtube-video
ℹ️ How to interpretEach entry is independently bundled, minified, and brotli-compressed. Initial size includes its static import graph; lazy dynamic chunks are reported separately. Entries are not additive because their dependency graphs overlap. Preset rows represent realistic combined bundles. Changes of 300 B or less across initial, lazy, and total size are collapsed, not discarded. Run |
There was a problem hiding this comment.
36 issues found across 2641 files
Prompt for AI agents (unresolved issues)
Check if these issues are valid — if so, understand the root cause of each and fix them. When an issue isn't valid or won't be fixed in this PR, reply in its thread with the reason and then resolve the thread. If appropriate, use sub-agents to investigate and fix each issue separately.
<file name="apps/e2e/suites/browserstack/README.md">
<violation number="1" location="apps/e2e/suites/browserstack/README.md:13">
P3: This says connecting 'retries up to three times', but `connectWithRetry` makes three attempts in total (one initial attempt plus two retries, after 10s and 20s). Reword it to say up to three attempts, each with a 90-second timeout.</violation>
</file>
<file name="internal/design/README.md">
<violation number="1" location="internal/design/README.md:27">
P3: This adds a new `player/` area for a single draft record. The layout rule in this file says to add a new area only when at least two durable records are likely, and otherwise use the nearest existing area. Move `extensions.md` into an existing area (or drop it from this PR) and remove this row.</violation>
</file>
<file name="internal/design/spf/evaluation-axes.md">
<violation number="1" location="internal/design/spf/evaluation-axes.md:94">
P3: This line is inaccurate: `pnpm size` measures every package entry point, not only SPF, and it reports brotli sizes, not gzip. Reword it to say “every package entry point, brotli-compressed”.</violation>
</file>
<file name="internal/design/element/reactive-element-lit-compat.md">
<violation number="1" location="internal/design/element/reactive-element-lit-compat.md:44">
P3: Restore 'custom media elements' here. 'Media components' is ambiguous in this repo, where it also refers to the Video/Audio components, and it no longer matches `CustomMediaMixin`.</violation>
</file>
<file name=".github/workflows/pr-review-request.yml">
<violation number="1" location=".github/workflows/pr-review-request.yml:24">
P1: This request job is the only gate before `pr-review.yml` runs with `ANTHROPIC_API_KEY` and a write-scoped token on fork PRs. For fork PRs, the `pull_request` run uses the PR head's copy of this file, so a fork can change `run` to succeed and trigger the review without maintainer approval. Enforce the gate in `pr-review.yml`, for example by checking the PR author's permission or the approval state there, rather than relying on this file.</violation>
</file>
<file name="packages/adapters/shaka-video/src/live.ts">
<violation number="1" location="packages/adapters/shaka-video/src/live.ts:99">
P3: This reverts an accurate reference to the wrong media. The fallback derivation mirrors hls.js, not native HLS, which does not expose HOLD-BACK to JS. Restore the original 'hls.js media' wording.</violation>
</file>
<file name="apps/sandbox/app/constants.ts">
<violation number="1" location="apps/sandbox/app/constants.ts:1">
P2: Removing `minimal` from `SKINS` breaks the SPF MoQ sandbox. Its skin `<select>` in `templates/spf-moq-player/index.html` still offers `minimal`, and `oneOf(..., SKINS, ...)` in `main.ts` now rejects it. Selecting it silently keeps the previous skin, and `?skin=minimal` silently falls back to `default`. Update the option to `neutral` (and the `skin=` doc comment in `main.ts`), or add the remaining skins to the select.</violation>
</file>
<file name="AGENTS.md">
<violation number="1" location="AGENTS.md:66">
P3: This bullet is dense and duplicates the classification rules already encoded in `site/scripts/check-api-stability.ts`. Shorten it to the rule plus the enforcing command, and point readers to the script for details.</violation>
</file>
<file name="internal/design/spf/features/source-replacement.md">
<violation number="1" location="internal/design/spf/features/source-replacement.md:24">
P3: The Cleanup contract bullet in this Status section still names `engine.test.ts` for the in-place replacement test, but this PR moves that test to `hls-video.test.ts`. Update it to `hls-video.test.ts` so the doc has one location for this test.</violation>
</file>
<file name="apps/e2e/suites/browserstack/connect.ts">
<violation number="1" location="apps/e2e/suites/browserstack/connect.ts:19">
P3: Every error is retried, including deterministic ones like rejected credentials or invalid capabilities. Those fail the same way on each attempt, so they cost up to the full retry delay and timeout budget. Consider retrying only on timeout or queueing errors.</violation>
</file>
<file name="packages/adapters/dash-video/src/adapter.ts">
<violation number="1" location="packages/adapters/dash-video/src/adapter.ts:47">
P3: The new static `engineName` is not covered by any test. Add an assertion in the dash-video adapter tests that the engine name resolves to `'dashjs'`, as the hlsjs-video tests do.</violation>
<violation number="2" location="packages/adapters/dash-video/src/adapter.ts:47">
P2: The Node export uses `src/server.ts`, whose `DashAdapter` lacks this static, so `DashAdapter.engineName` is undefined there despite the shared declaration. Add the same static to the server adapter.</violation>
</file>
<file name="packages/adapters/native-hls-video/src/fairplay.ts">
<violation number="1" location="packages/adapters/native-hls-video/src/fairplay.ts:16">
P2: `NativeHlsDrmErrors` and `NativeHlsDrmErrorContext` are exported from the package entry, and the doc comment says consumers branch on these context values. Marking them `@internal` hides a supported contract from API docs and stability checks. Remove the `@internal` tags unless this is meant to make them private, and then also stop exporting them from `index.ts`.</violation>
</file>
<file name="internal/design/spf/features/video-abr.md">
<violation number="1" location="internal/design/spf/features/video-abr.md:40">
P3: The updated Composition path is correct, but the Behaviors table directly below still points to `switchVideoQuality` in `behaviors/quality-switching.ts`, and neither exists anymore. The ABR behavior is `switchVideoTrack` in `behaviors/track-switching.ts`. Update the row so this section matches the new engine path.</violation>
</file>
<file name="packages/adapters/playerjs-video/src/props.ts">
<violation number="1" location="packages/adapters/playerjs-video/src/props.ts:8">
P3: This sentence is garbled: "are they, `controls`, and `preload` also written onto the embed URL" has no clear subject. Reword it to name the props directly, e.g. "For the services the host recognizes (...), `autoplay`, `defaultMuted`, `loop`, `controls`, and `preload` are also written onto the embed URL in the service's own spelling." `muted` is in the `Pick` but is not described.</violation>
</file>
<file name="packages/adapters/spotify-audio/src/adapter.ts">
<violation number="1" location="packages/adapters/spotify-audio/src/adapter.ts:515">
P2: With `autoplay` on, clearing `src` leaves `paused` as `false`. `#resetState()` sets `#paused = !this.#autoplay`, and the new early return drops the only update that would correct it, so the embed's pause is never reflected. Keep the paused state in sync when the source is cleared, for example by setting `#paused = true` on the cleared path.</violation>
</file>
<file name="apps/sandbox/templates/react-spotify-audio/main.tsx">
<violation number="1" location="apps/sandbox/templates/react-spotify-audio/main.tsx:16">
P3: `mediaProps` includes `muted`, which `SpotifyAudio` does not accept. `props.ts` deliberately omits it, so it reaches the iframe as an inert `muted` attribute. Consider leaving it out of the spread for this template, or confirm the HTML sibling's parity is intended.</violation>
</file>
<file name="packages/adapters/mux-video/src/adapter.ts">
<violation number="1" location="packages/adapters/mux-video/src/adapter.ts:76">
P3: The early-return check now calls `createMuxVideoURL`, so the dev-only min/max resolution warning fires here as well as in the `source` setter. A misconfigured URL logs it twice per assignment and again on every re-sync. Consider building the canonical form once and passing it through, or skipping the warning for the comparison.</violation>
</file>
<file name="packages/adapters/playerjs-video/src/index.ts">
<violation number="1" location="packages/adapters/playerjs-video/src/index.ts:5">
P2: These protocol types are tagged `@internal` in `player-api.ts`, but this entry point re-exports them as public API. Either drop the re-export block or remove the `@internal` tags. The index comment says the player API stays internal, so dropping the block matches it.</violation>
</file>
<file name="packages/adapters/mux-video/src/spf/mixin.ts">
<violation number="1" location="packages/adapters/mux-video/src/spf/mixin.ts:93">
P2: Keep the raw-string equality short-circuit. Setting `src` to the value already in place now re-derives the source when `source.src` is a non-canonical Mux URL with no `playbackId`, and that drops `drm` entries. Compare against `value` as well as the canonical form.</violation>
</file>
<file name="packages/adapters/shaka-video/src/media-tracks.ts">
<violation number="1" location="packages/adapters/shaka-video/src/media-tracks.ts:100">
P2: Dispatching `change` on `videoRenditions` here also runs `#switchRendition`, which re-issues `engine.selectVideoTrack` for the pinned rendition on a metadata-only refresh. That contradicts the comment that the pinned track should be left alone. The event is also fired synchronously inside Shaka's `trackschanged` handler, unlike the queued `change` the base list uses. Route metadata notifications around the selection listener, or skip re-selection when the pinned track is already applied.</violation>
</file>
<file name="packages/adapters/native-hls-video/src/chapters.ts">
<violation number="1" location="packages/adapters/native-hls-video/src/chapters.ts:59">
P3: This adds a second fetch of the same multivariant playlist that the live mixin already requests on `loadstart`. The cost is one extra request per source load. Consider sharing one playlist fetch between the two mixins, or accept the duplicate and note it in the mixin docs.</violation>
<violation number="2" location="packages/adapters/native-hls-video/src/chapters.ts:69">
P2: Native HLS chapter discovery ignores `crossorigin="use-credentials"`: both the playlist fetch and chapters document request omit `credentials`, so cross-origin protected streams cannot load their chapters. Pass `credentials: 'include'` to both requests when the element uses that mode.</violation>
</file>
<file name="packages/adapters/shaka-video/CHANGELOG.md">
<violation number="1" location="packages/adapters/shaka-video/CHANGELOG.md:8">
P3: This Shaka release note advertises an hls.js-specific fix from the hlsjs-video changelog. Remove the unrelated entry so Shaka users are not told it contains a fix it did not ship.</violation>
</file>
<file name="internal/design/spf/conventions/config.md">
<violation number="1" location="internal/design/spf/conventions/config.md:60">
P3: This `ts` example is no longer valid TypeScript: the interface declaration was removed, leaving property signatures inside a standalone block. Restore a type or interface declaration.</violation>
</file>
<file name=".agents/plans/moq-engine-implementation.md">
<violation number="1" location=".agents/plans/moq-engine-implementation.md:163">
P3: This temporary implementation plan remains in the PR despite the repository instruction to delete plans before merge. Remove `.agents/plans/moq-engine-implementation.md` instead of carrying this updated note forward.</violation>
</file>
<file name=".github/workflows/e2e.yml">
<violation number="1" location=".github/workflows/e2e.yml:46">
P2: This filter omits packages that `check:projects` packs into generated projects, including `core`, `html`, `react`, and the mux adapters. Changes to those packages alone skip the install/typecheck/build integration test; include these package paths in the filter.</violation>
</file>
<file name="apps/sandbox/templates/cdn/main.ts">
<violation number="1" location="apps/sandbox/templates/cdn/main.ts:168">
P1: The CDN loader routes `compat` and `scaffold` to the default bundle, so their generated skin tags are never registered. Add matching compat/scaffold bundle imports in each player branch, including live variants.</violation>
</file>
<file name=".github/scripts/changelog-prose-open-pr.sh">
<violation number="1" location=".github/scripts/changelog-prose-open-pr.sh:73">
P2: An already-open prose PR exits before this label is applied, so legacy bot PRs still won't trigger the site cherry-pick when merged. Add the label to the existing PR before returning.</violation>
</file>
<file name="packages/adapters/playerjs-video/src/source.ts">
<violation number="1" location="packages/adapters/playerjs-video/src/source.ts:12">
P3: The serialization sentence is malformed (`as given becomes 'true')`), making the option behavior harder to understand. Rewrite it as “Values are stringified as given (`true` becomes `'true'`)”.</violation>
<violation number="2" location="packages/adapters/playerjs-video/src/source.ts:78">
P2: HTML `<playerjs-video>` ignores `source.engine.playerJs.referrerPolicy`: this code removes it, while the HTML template reads only the separate `referrerpolicy` attribute. Apply the source option to the generated iframe or document that HTML callers must set the attribute.</violation>
</file>
<file name="packages/adapters/hlsjs-video/src/cap-level.ts">
<violation number="1" location="packages/adapters/hlsjs-video/src/cap-level.ts:44">
P3: This cap matches pixel area, so a wider but shorter rendition can still fit; qualify “wider renditions” as wider renditions at the same height.</violation>
</file>
<file name="packages/adapters/hlsjs-video/src/adapter.ts">
<violation number="1" location="packages/adapters/hlsjs-video/src/adapter.ts:23">
P2: `PlaybackType` now accepts every string, but engine selection only treats `'native'` specially; a typo such as `'natve'` silently selects MSE when supported. Keep this union closed or validate unknown preferences.</violation>
</file>
<file name=".agents/skills/implement-spf-use-case/references/workflow.md">
<violation number="1" location=".agents/skills/implement-spf-use-case/references/workflow.md:438">
P2: This workflow directs contributors to expose engine variants through `hls.ts`, but that module explicitly omits engines; the package exposes them through `@videojs/spf/engines/<name>`. Update this reference and the earlier `@videojs/spf/hls/*` reference to use the engine subpath.</violation>
</file>
<file name="packages/adapters/twitch-video/src/adapter.ts">
<violation number="1" location="packages/adapters/twitch-video/src/adapter.ts:529">
P2: `#nextMuted` lets `defaultMuted` override the live mute state on every reset. After a user unmutes with `defaultMuted=true`, a content-only source swap reuses the iframe but resets `#muted` to `true`, so the adapter reports a mute that never happened; preserve the live mute value for reused embeds.</violation>
</file>
<file name="packages/adapters/hlsjs-video/src/preload.ts">
<violation number="1" location="packages/adapters/hlsjs-video/src/preload.ts:60">
P2: Setting `autoplay` on the public `HlsJsAdapter` only updates the video element; it never invokes this delegate setter, so loading remains deferred until `play`. Forward autoplay changes to the HLS delegate or rerun its preload initialization from the facade.</violation>
</file>
Note: This PR contains a large number of files. cubic selects up to 200 of the highest-priority eligible files for this review, so some files may not have been reviewed.
Tip: Get a Slack DM when cubic finishes reviewing your PRs. Connect Slack
View guided diff | Turn on auto-fix | Re-trigger cubic
|
|
||
| steps: | ||
| - name: Request review | ||
| run: echo "Requesting a Claude review of ${{ github.event.pull_request.head.sha }}." |
There was a problem hiding this comment.
P1: This request job is the only gate before pr-review.yml runs with ANTHROPIC_API_KEY and a write-scoped token on fork PRs. For fork PRs, the pull_request run uses the PR head's copy of this file, so a fork can change run to succeed and trigger the review without maintainer approval. Enforce the gate in pr-review.yml, for example by checking the PR author's permission or the approval state there, rather than relying on this file.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. When an issue isn't valid or won't be fixed in this PR, reply in its thread with the reason and then resolve the thread. At .github/workflows/pr-review-request.yml, line 24:
<comment>This request job is the only gate before `pr-review.yml` runs with `ANTHROPIC_API_KEY` and a write-scoped token on fork PRs. For fork PRs, the `pull_request` run uses the PR head's copy of this file, so a fork can change `run` to succeed and trigger the review without maintainer approval. Enforce the gate in `pr-review.yml`, for example by checking the PR author's permission or the approval state there, rather than relying on this file.</comment>
<file context>
@@ -0,0 +1,24 @@
+
+ steps:
+ - name: Request review
+ run: echo "Requesting a Claude review of ${{ github.event.pull_request.head.sha }}."
</file context>
| else await import('@videojs/cdn/live-video'); | ||
| } else { | ||
| if (skin === 'minimal') await import('@videojs/cdn/video-minimal'); | ||
| if (skin === 'neutral') await import('@videojs/cdn/video-neutral'); |
There was a problem hiding this comment.
P1: The CDN loader routes compat and scaffold to the default bundle, so their generated skin tags are never registered. Add matching compat/scaffold bundle imports in each player branch, including live variants.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. When an issue isn't valid or won't be fixed in this PR, reply in its thread with the reason and then resolve the thread. At apps/sandbox/templates/cdn/main.ts, line 168:
<comment>The CDN loader routes `compat` and `scaffold` to the default bundle, so their generated skin tags are never registered. Add matching compat/scaffold bundle imports in each player branch, including live variants.</comment>
<file context>
@@ -162,20 +162,20 @@ async function loadCdnPlayer(skin: Skin, live: boolean) {
else await import('@videojs/cdn/live-video');
} else {
- if (skin === 'minimal') await import('@videojs/cdn/video-minimal');
+ if (skin === 'neutral') await import('@videojs/cdn/video-neutral');
else await import('@videojs/cdn/video');
}
</file context>
| @@ -1,4 +1,4 @@ | |||
| export const SKINS = ['default', 'minimal'] as const; | |||
| export const SKINS = ['default', 'neutral', 'compat', 'scaffold'] as const; | |||
There was a problem hiding this comment.
P2: Removing minimal from SKINS breaks the SPF MoQ sandbox. Its skin <select> in templates/spf-moq-player/index.html still offers minimal, and oneOf(..., SKINS, ...) in main.ts now rejects it. Selecting it silently keeps the previous skin, and ?skin=minimal silently falls back to default. Update the option to neutral (and the skin= doc comment in main.ts), or add the remaining skins to the select.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. When an issue isn't valid or won't be fixed in this PR, reply in its thread with the reason and then resolve the thread. At apps/sandbox/app/constants.ts, line 1:
<comment>Removing `minimal` from `SKINS` breaks the SPF MoQ sandbox. Its skin `<select>` in `templates/spf-moq-player/index.html` still offers `minimal`, and `oneOf(..., SKINS, ...)` in `main.ts` now rejects it. Selecting it silently keeps the previous skin, and `?skin=minimal` silently falls back to `default`. Update the option to `neutral` (and the `skin=` doc comment in `main.ts`), or add the remaining skins to the select.</comment>
<file context>
@@ -1,4 +1,4 @@
-export const SKINS = ['default', 'minimal'] as const;
+export const SKINS = ['default', 'neutral', 'compat', 'scaffold'] as const;
export const PLATFORMS = ['html', 'react', 'cdn'] as const;
export const STYLINGS = ['css', 'tailwind'] as const;
</file context>
| * `context` values carried by the `MediaError`s key exchange produces. The message is prose meant for a person; this is | ||
| * the part to branch on. | ||
| * | ||
| * @internal |
There was a problem hiding this comment.
P2: NativeHlsDrmErrors and NativeHlsDrmErrorContext are exported from the package entry, and the doc comment says consumers branch on these context values. Marking them @internal hides a supported contract from API docs and stability checks. Remove the @internal tags unless this is meant to make them private, and then also stop exporting them from index.ts.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. When an issue isn't valid or won't be fixed in this PR, reply in its thread with the reason and then resolve the thread. At packages/adapters/native-hls-video/src/fairplay.ts, line 16:
<comment>`NativeHlsDrmErrors` and `NativeHlsDrmErrorContext` are exported from the package entry, and the doc comment says consumers branch on these context values. Marking them `@internal` hides a supported contract from API docs and stability checks. Remove the `@internal` tags unless this is meant to make them private, and then also stop exporting them from `index.ts`.</comment>
<file context>
@@ -12,6 +12,8 @@ export const FAIRPLAY_CONTENT_TYPE = 'application/vnd.apple.mpegurl';
* `context` values carried by the `MediaError`s key exchange produces. The message is prose meant for a person; this is
* the part to branch on.
+ *
+ * @internal
*/
export const NativeHlsDrmErrors = {
</file context>
| // to complete, and completing on an update from the paused embed would put the cleared state right back. | ||
| if (this.#src && !this.#loaded) this.#onLoaded(); | ||
| // The embed retains the old entity after clearing, so its snapshots no longer belong to a source. | ||
| if (!this.#src) return; |
There was a problem hiding this comment.
P2: With autoplay on, clearing src leaves paused as false. #resetState() sets #paused = !this.#autoplay, and the new early return drops the only update that would correct it, so the embed's pause is never reflected. Keep the paused state in sync when the source is cleared, for example by setting #paused = true on the cleared path.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. When an issue isn't valid or won't be fixed in this PR, reply in its thread with the reason and then resolve the thread. At packages/adapters/spotify-audio/src/adapter.ts, line 515:
<comment>With `autoplay` on, clearing `src` leaves `paused` as `false`. `#resetState()` sets `#paused = !this.#autoplay`, and the new early return drops the only update that would correct it, so the embed's pause is never reflected. Keep the paused state in sync when the source is cleared, for example by setting `#paused = true` on the cleared path.</comment>
<file context>
@@ -510,9 +511,11 @@ export class SpotifyAdapter extends MediaPlayedRangesMixin(EventTarget) implemen
- // to complete, and completing on an update from the paused embed would put the cleared state right back.
- if (this.#src && !this.#loaded) this.#onLoaded();
+ // The embed retains the old entity after clearing, so its snapshots no longer belong to a source.
+ if (!this.#src) return;
+
+ // Later loads (`loadUri`) never re-fire `ready`, so the first update completes the load.
</file context>
|
|
||
| ### Bug Fixes | ||
|
|
||
| * **hlsjs-video:** read live edge start from hls.js live sync position ([#9409](https://github.com/videojs/video.js/issues/9409)) ([4676472](https://github.com/videojs/video.js/commit/4676472d9136580a5d45e0e8d2929a7e124ad122)) |
There was a problem hiding this comment.
P3: This Shaka release note advertises an hls.js-specific fix from the hlsjs-video changelog. Remove the unrelated entry so Shaka users are not told it contains a fix it did not ship.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. When an issue isn't valid or won't be fixed in this PR, reply in its thread with the reason and then resolve the thread. At packages/adapters/shaka-video/CHANGELOG.md, line 8:
<comment>This Shaka release note advertises an hls.js-specific fix from the hlsjs-video changelog. Remove the unrelated entry so Shaka users are not told it contains a fix it did not ship.</comment>
<file context>
@@ -1,5 +1,80 @@
+
+### Bug Fixes
+
+* **hlsjs-video:** read live edge start from hls.js live sync position ([#9409](https://github.com/videojs/video.js/issues/9409)) ([4676472](https://github.com/videojs/video.js/commit/4676472d9136580a5d45e0e8d2929a7e124ad122))
+
+
</file context>
| ```ts | ||
| interface HlsVideoEngineConfig { | ||
| // Combined from `SwitchVideoTrackConfig`, `SegmentLoaderActorConfig`, `TextTrackActorsConfig`, … | ||
| { |
There was a problem hiding this comment.
P3: This ts example is no longer valid TypeScript: the interface declaration was removed, leaving property signatures inside a standalone block. Restore a type or interface declaration.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. When an issue isn't valid or won't be fixed in this PR, reply in its thread with the reason and then resolve the thread. At internal/design/spf/conventions/config.md, line 60:
<comment>This `ts` example is no longer valid TypeScript: the interface declaration was removed, leaving property signatures inside a standalone block. Restore a type or interface declaration.</comment>
<file context>
@@ -53,10 +53,11 @@ Worked example: `bufferDuration` is used by the load-segments dispatcher (for `r
```ts
-interface HlsVideoEngineConfig {
+// Combined from `SwitchVideoTrackConfig`, `SegmentLoaderActorConfig`, `TextTrackActorsConfig`, …
+{
// Sub-configs reuse the canonical types from their owning module
bandwidth?: Partial<BandwidthConfig>; // network/bandwidth-estimator
</file context>
| { | |
| type HlsVideoEngineConfig = { |
| @@ -160,7 +160,7 @@ promotion). Zero page errors. | |||
| real relay. | |||
There was a problem hiding this comment.
P3: This temporary implementation plan remains in the PR despite the repository instruction to delete plans before merge. Remove .agents/plans/moq-engine-implementation.md instead of carrying this updated note forward.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. When an issue isn't valid or won't be fixed in this PR, reply in its thread with the reason and then resolve the thread. At .agents/plans/moq-engine-implementation.md, line 163:
<comment>This temporary implementation plan remains in the PR despite the repository instruction to delete plans before merge. Remove `.agents/plans/moq-engine-implementation.md` instead of carrying this updated note forward.</comment>
<file context>
@@ -160,7 +160,7 @@ promotion). Zero page errors.
- **Adapter seam**: `MoqMediaOptions.engineConfig` forwards engine config
(transport factory, ABR/latency tuning) through the element
- constructor; the adapter still owns `onSignalsReady`. Without it there
+ constructor. Without it there
was no way to reach `createMoqTransport` from outside SPF.
- **Skin layout parity**: the wrapper's canvas is now styled by a shadow
</file context>
| * parameters: each key here is a query parameter written onto the embed URL for whichever service serves it — anything | ||
| * in [Gumlet's](https://docs.gumlet.com/video/embed-stream), [Bunny Stream's](https://bunny.net/docs/stream/embedding), | ||
| * or [Livid's](https://support.livid.com/article/46-advanced-embedding-parameters) parameter lists, say. A parameter a | ||
| * service reads from the URL hash, like Livid's `#t`, belongs in `src`. Values are stringified as given becomes |
There was a problem hiding this comment.
P3: The serialization sentence is malformed (as given becomes 'true')), making the option behavior harder to understand. Rewrite it as “Values are stringified as given (true becomes 'true')”.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. When an issue isn't valid or won't be fixed in this PR, reply in its thread with the reason and then resolve the thread. At packages/adapters/playerjs-video/src/source.ts, line 12:
<comment>The serialization sentence is malformed (`as given becomes 'true')`), making the option behavior harder to understand. Rewrite it as “Values are stringified as given (`true` becomes `'true'`)”.</comment>
<file context>
@@ -0,0 +1,91 @@
+ * parameters: each key here is a query parameter written onto the embed URL for whichever service serves it — anything
+ * in [Gumlet's](https://docs.gumlet.com/video/embed-stream), [Bunny Stream's](https://bunny.net/docs/stream/embedding),
+ * or [Livid's](https://support.livid.com/article/46-advanced-embedding-parameters) parameter lists, say. A parameter a
+ * service reads from the URL hash, like Livid's `#t`, belongs in `src`. Values are stringified as given becomes
+ * `'true'`), an array repeats the parameter once per value, and `null` removes a parameter the embed URL would
+ * otherwise carry.
</file context>
| * while still excluding anything genuinely wider. | ||
| * Rounds the 16:9 width up so every standard rendition fits its own cap. For `'480p'`, the exact width is 853.33, while | ||
| * ladders ship 854×480. The exact area would put that rendition over its own cap; rounding admits it while still | ||
| * excluding wider renditions. |
There was a problem hiding this comment.
P3: This cap matches pixel area, so a wider but shorter rendition can still fit; qualify “wider renditions” as wider renditions at the same height.
Prompt for AI agents
Check if this issue is valid — if so, understand the root cause and fix it. When an issue isn't valid or won't be fixed in this PR, reply in its thread with the reason and then resolve the thread. At packages/adapters/hlsjs-video/src/cap-level.ts, line 44:
<comment>This cap matches pixel area, so a wider but shorter rendition can still fit; qualify “wider renditions” as wider renditions at the same height.</comment>
<file context>
@@ -39,10 +39,9 @@ export interface RenditionCapController {
- * while still excluding anything genuinely wider.
+ * Rounds the 16:9 width up so every standard rendition fits its own cap. For `'480p'`, the exact width is 853.33, while
+ * ladders ship 854×480. The exact area would put that rendition over its own cap; rounding admits it while still
+ * excluding wider renditions.
*/
export function resolutionToPixelArea(resolution: MediaResolution | undefined): number {
</file context>
| * excluding wider renditions. | |
| * excluding wider renditions at the same height. |
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: d8e9f72642
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| #resetState() { | ||
| this.#listenerId = createListenerId(); | ||
| this.#requests.clear(); | ||
| this.#ready = false; |
There was a problem hiding this comment.
Clear pending play intent when replacing the source
If play() is called before the embed reports ready, replacing or clearing src leaves #playRequested set because #resetState() never resets it. When a subsequent source reports ready, #syncToEmbed() sends it a play command even with autoplay disabled and no play request for that source. For example, starting A, clearing it while it loads, and then selecting B unexpectedly starts B. Cancel the outgoing source's pending play intent when its load is discarded.
Useful? React with 👍 / 👎.
| get isFullscreen() { | ||
| return this.#isFullscreen; |
There was a problem hiding this comment.
Derive fullscreen state from the actual fullscreen element
After requestFullscreen() succeeds, exiting through Escape, browser controls, or document.exitFullscreen() leaves #isFullscreen permanently true: only the adapter's own exitFullscreen() and load reset clear it. This also affects the player fullscreen feature, whose isFullscreen() helper falls back to the media's flag after the document leaves fullscreen. Read the target's actual fullscreen status or synchronize the flag on fullscreenchange so external exits restore the correct state.
Useful? React with 👍 / 👎.
Co-authored-by: Claude <noreply@anthropic.com>
…eports (#9566) Co-authored-by: Rennan Ribeiro <rennanribeiro1@gmail.com> Co-authored-by: Claude Opus 5.5 <noreply@anthropic.com>
…s-moq # Conflicts: # packages/spf/package.json # packages/spf/src/playback/behaviors/resolve-catalog.ts # packages/spf/src/playback/behaviors/subscribe-selected-tracks.ts # pnpm-lock.yaml
# Conflicts: # packages/html/src/define/tests/ssr-safety.test.ts
Summary
Syncs
mmcc/moqto videojs/video.jsmain(905a6d1). v10 moved into that repo after GA, and the fork'supstreamremote now points there. Merge only: land by fast-forward push, never squash or rebase.Upstream reorganized SPF (#9391, #9392, #9395), so the MoQ engine needed porting:
shareSignals/onSignalsReadyremoved upstream. The MoQ engine now declares itsuser*TrackSelectionslots withdefineExternalSignals. The adapter and tests readengine.state/engine.contextdirectly, andMoqEngineSignalsis gone.playback/enginestsconfig project. It now coversengines/moq, so the standaloneengines/moq/tsconfig.jsonis deleted../moqsits beside upstream's newadapters/*paths.measure-size.jsfollows upstream's removal.defineBehaviorfromdefine-behavior.track-switching.ts(confineToActiveSwitchingSetand upstream'spreferAudioPolicy) and in hls-videomedia-tracks.ts.mmcc/moq(feat(spf): render MoQ audio through the @caprid/moq-media worklet core #72, feat(spf): speak moqt-22 through @caprid/moq-core #75, fix(spf): pin @caprid/moq-media 0.1.0 and address #75 review #77:@caprid/moq-core/moq-media) and six new upstream commits (SoundCloud adapter, hls.js 1.7.3). Conflicts: spfpackage.jsondeps anddefineBehaviorimports kept both sides, and the SSR-safety test keeps both the simple-moq-video and soundcloud-audio cases.SimpleMoqVideoElementis tagged@internal, as upstream's new API-stability check requires.Testing
pnpm build:packagesandpnpm typecheckpass.pnpm lintpasses.@videojs/spf: 2343 pass, 1 skipped (fewer than before because feat(spf): speak moqt-22 through @caprid/moq-core #75 moved the transport code and its tests to@caprid/moq-core).@videojs/html: 497 pass.@videojs/installation: 312 pass.pnpm check:workspaceandpnpm -F site check:api-stabilitypass.🤖 Generated with Claude Code
https://claude.ai/code/session_01DaH4iixnBQbNchnez2h4ZG
Summary by cubic
Syncs the
mmcc/moqfork tovideojs/video.jsmain (currentmainhead). Upstream reorganized SPF (#9391, #9392, #9395), so the MoQ engine was ported to the new structure:shareSignals/onSignalsReadyare replaced bydefineExternalSignalstrack-selection slots, and the standaloneengines/moqtsconfig andmeasure-size.jsare removed.Merge note
SimpleMoqVideoElementis tagged@internalfor the upstream API-stability check.maincontent, including the fullCHANGELOG.md, the minimal-to-neutral skin rename, and the newcompat/scaffoldskins.Written for commit be530c1. Summary will update on new commits.