evidence · 1.5 KiB

2026-09-27-shell-acceptance.md

Committed at docs/evidence/operator-reports/2026-09-27-shell-acceptance.md. This is what grounds the requirements that cite it.

Open raw source

Operator real-finger report, 2026-09-27

Evidence class: operator report from real-finger use on the physical board. There is no camera recording and no native capture. The board was running the combined installed system built from master at the time (see docs/evidence/card-shell/transparent-corners/installed-system/).

The operator, verbatim: “the rounded cards seem fine … launch splash seems fine … new apps seem fine shade drag is fine brightness should be a slider”.

FeatureReportChange and task
Large rounded overview cards: flicking, open/close, radiusfinethe-overview-shows-large-rounded-cards E.1
Launch splashfinelaunching-an-app-shows-a-splash 6.1–6.3 stay open: they require captures
New drawer appsfinefeat/more-apps (no OpenSpec change)
Shade drag-to-close, including dragging the panelfineshade drag-to-close evidence
Brightnessworks; the control should be a slider, not a stepperfollow-up

Not covered by this report: the video-card overview flow, the timeout and failure splash states, and line-out audio.

Brightness slider physical acceptance, 2026-09-28

The operator, verbatim: “confirming physical acceptance of the-brightness-control-is-a-slider”. This covers the Settings slider (drag, tap-to-jump, 3% floor), the Shade slider (horizontal drag doesn’t close the shade; a drag elsewhere still does), and the current value showing when either sheet opens. Evidence class: an operator’s real-finger acceptance on the physical board.