Super Screencast Community v0.10.4

Super Screencast 0.10.4

Community and Pro use the same 0.10.4 application version.

• Adds the native macOS recorder and control panel, with display, camera,
  microphone, system audio, composition, and MP4/MKV recording adapters.
• Adds the native composition/recording preview and recording countdown window.
• Ships a macOS app with bundled runtime libraries and the website's app icon.
• Enables the private Pro provider on macOS, including cinematic composition,
  speaker notes, and a controlled whisper.cpp CPU speech helper.
• Adds selected-display pointer coordinates and permission-controlled mouse input
  for Pro cursor following and automatic cinematic click zoom.
• Keeps MP4 recordings recoverable as MKV until finalization, and refreshes the
  completed path before reporting status or creating automatic subtitle sidecars.
• Preserves the Linux Arch packages and adds a static Linux executable artifact.

The static Linux executable includes X11, Wayland, a software-rendered panel,
and built-in PipeWire client modules. It still needs host desktop services,
fonts, keyboard-layout data, and PipeWire configuration. SDL D-Bus IME and GPU
rendering are disabled in this static variant.
The static panel uses X11 and requires XWayland on Wayland desktops; Wayland
screen capture remains available through the desktop portal and PipeWire.

The macOS download is an Apple Silicon app with ad-hoc signing. It is not
Developer ID signed or notarized. Its Info.plist records the minimum macOS version
required by its bundled libraries; the source adapters target macOS 13 onward.
Screen, camera, and microphone access require macOS permission grants. Automatic
click zoom requires Input Monitoring and restarting capture. Independent cursor
smoothing is unavailable with the embedded capture cursor.
Virtual camera/microphone, RTMP, window/region capture, and global keyboard
annotations/hotkeys remain unsupported on macOS. Continuous system-audio quality
and long-duration A/V drift still need hardware validation; intermittent system
audio gaps were observed in the earlier development port.

Private Pro acceptance bundles use controlled LGPL media and CPU speech
libraries, the existing production public license trust, and the reviewed EULA.
They require a valid license and an explicitly installed speech model. Official
Pro publication retains its separate signing, dependency review, and hardware
acceptance requirements; Community releases do not contain private Pro code.