1.4 KiB
1.4 KiB
Roadmap
This is a preview roadmap, not a promise.
Current Scope
- Linux/X11 native dictation app
- GTK control panel and AppIndicator tray
- global hotkeys and modifier input mode
- local batch transcription through
faster-whisper - OpenAI-compatible batch STT endpoints
- Riva/NIM realtime STT streaming through WebSocket
- optional rewrite workflows through OpenAI-compatible chat endpoints
- xdotool typing or clipboard paste into the focused window
- Debian package for local Ubuntu/Debian installs
- no hosted Blitztext backend
Next Useful Work
- Capture and add current Linux screenshots for the control panel, settings, benchmark, and About tab.
- Improve first-run setup for STT engines, microphones, and output mode.
- Add a guided realtime STT streaming test that does not type into the active window.
- Add automated tests around config parsing, routing, quality filters, URL handling, and streaming protocol message generation.
- Improve Wayland support with
wtypeorydotoolas alternatives toxdotool. - Add safer clipboard handling and clearer output-mode diagnostics.
- Add a lightweight release checklist for
.debbuilds and source installs. - Document known-good local LLM and STT server configurations.
Not In Scope Yet
- Production support.
- Accounts, sync, teams, or hosted infrastructure.
- Claims that the app is offline or privacy-complete by default.
- App Store distribution.
- A polished one-click consumer release.