{
  "authorship": "Five self-authored fictional cases, not vendor results or actual observations.",
  "cases": [
    {
      "id": 1,
      "text": " 1. Seek to the second verse without holding a handle  Fictional setup: Mira reviews an invented two-minute lyric video. The current sketch shows a timeline handle that must be dragged. Her task is to revisit the verse at 00:48 after pausing at 00:12. These times describe the exercise, not an observed recording.  Proposed route: allow a click or tap on the timeline to move playback, and provide a separate precise position field if the timeline cannot conveniently select the required point. Mira records whether seeking keeps the player paused or resumes it. That playback-state choice belongs in the handoff because an unexpected restart can interrupt lyric review.  Expected record: starting position 00:12, paused; requested position 00:48; final position and playback state pending observation. The route is not complete merely because a marker moves visually. The reviewer must check that the actual media position changes too. If the click only selects the handle and still requires dragging, retain the unresolved item. No vendor interface was opened for this fictional exercise. "
    },
    {
      "id": 2,
      "text": " 2. Adjust volume with separate step controls  Fictional setup: Arun's proposed player has a volume slider whose only documented pointer action is dragging. He wants a quiet rehearsal setting without muting the song completely. A mute button reaches a different outcome and therefore is not the entire proposed replacement.  Proposed route: add labeled decrease and increase buttons, with a visible current setting. In this invented design, each activation changes the setting by five percentage points. The chosen step is an authored example, not a W3C-prescribed increment or a measured sound level. Displayed percentages are not acoustic loudness measurements.  Expected record: start at the fictional setting of 50, activate decrease twice, and expect a displayed setting of 40. Save the actual result only after implementation. Check both directions and the range boundaries; do not infer the maximum or minimum from one successful middle value. Also record whether changing volume affects only this player. Keyboard controls and the pointer route need their own checks, even when they share the same underlying value. "
    },
    {
      "id": 3,
      "text": " 3. Reorder a review playlist using move buttons  Fictional setup: Bea's review page contains three invented versions: Draft A, Approved B and Instrumental C. The sketch asks her to drag Approved B above Draft A. She needs a route that does not require maintaining contact while moving an item.  Proposed route: give each item a clearly associated Move earlier and Move later action. From the starting order A, B, C, activating Move earlier on B should produce B, A, C. At the first position, the unavailable earlier action should not silently wrap B to the end unless that behavior is deliberately specified and explained.  Expected record: capture the complete order before and after the operation, not only the selected item's label. Check that playback selection, approved-version labeling and the underlying lyric revision stay attached to the correct item. A cosmetic row movement with the wrong audio still attached would not satisfy Bea's task. The example proposes an ordering model; it does not claim any music service implements it. Test persistence separately if the page promises to save the order. "
    },
    {
      "id": 4,
      "text": " 4. Enter a precise position through a pointer-operated field  Fictional setup: Leon wants to compare the opening of a chorus at 01:07. His small player makes that point difficult to select on a narrow timeline. A developer proposes a time field with an Apply button as a second route.  Proposed route: tap the labeled field, enter the position using the available on-screen keyboard, then activate Apply. Record the accepted format and what happens for an invalid or out-of-range entry. A field that can only be edited through a physical keyboard would not demonstrate the intended pointer-only route in this scenario.  Expected record: the authored input 01:07 means sixty-seven seconds in this exercise. The player should show the resulting position or an understandable validation message; silently substituting an unrelated value would hide an unresolved mismatch. Keep invalid input from being mistaken for a completed seek. None of these actions has been executed on a real player here. The review sheet therefore contains the proposed sequence and expected outcome, with the observed result left blank for the implementation reviewer. "
    },
    {
      "id": 5,
      "text": " 5. Establish who controls a native-looking slider  Fictional setup: Nia receives an embedded player that appears to use ordinary browser media controls. A teammate suggests closing the dragging review because the controls look familiar. Nia has no implementation evidence yet.  Proposed route: identify whether the relevant behavior is determined by the browser and unmodified by the page author. Record the embedding code or implementation documentation, browser version and control being considered. If the page draws its own slider or changes its pointer behavior, revisit the exception reasoning rather than carrying forward an assumption from appearance.  Expected record: ownership and modification status remain unresolved until supported. Nia can still document useful pointer alternatives for viewers even while the exception decision is pending. A finding about one native control does not decide whether a custom playlist elsewhere on the same page needs an alternative. This case teaches the boundary of a review, not a certification result. No browser behavior, assistive technology compatibility or vendor compliance is asserted from the invented scenario. "
    }
  ]
}