Skip to content

LINUX.md

Latest commit

 

History

History
55 lines (44 loc) · 1.77 KB

LINUX.md

File metadata and controls

55 lines (44 loc) · 1.77 KB
Raw

LocalText2Voice on Linux

The application core is cross-platform (PySide6). On Windows the installer bundles everything; on Linux you run from source. Tested on Arch/CachyOS (KDE Plasma, Wayland) with Python 3.12.

Requirements

  • Python 3.10+ with virtual-environment support. On Debian/Ubuntu install python3-venv; on Arch this is included with python.
  • ffmpeg available in PATH (sudo pacman -S ffmpeg / sudo apt install ffmpeg). The bundled ffmpeg/ folder is only needed on Windows; on Linux the app finds the system binary automatically (see app/utils/ffmpeg_utils.py).
  • For local TTS engines: they are downloaded by the app itself (Settings -> Engine management). Piper and Kokoro work on CPU; Chatterbox / Qwen3 TTS / OmniVoice can use an NVIDIA GPU.

Run

./run_dev.sh          # creates .venv and installs deps on first run
# or manually:
python3 -m venv .venv
.venv/bin/pip install -r requirements.txt
QT_QPA_PLATFORM=wayland .venv/bin/python main.py   # or omit for X11

Desktop integration (optional)

mkdir -p ~/.local/share/applications
cat > ~/.local/share/applications/localtext2voice.desktop <<EOF
[Desktop Entry]
Type=Application
Name=LocalText2Voice
Exec=$(pwd)/run_dev.sh
Icon=$(pwd)/assets/logotipo.png
Categories=Audio;Utility;
EOF

Tests

QT_QPA_PLATFORM=offscreen .venv/bin/python -m pytest -q

Known platform notes

  • Auto-update flow (app/core/update_manager.py) is Windows-only and is skipped on other platforms; update with git pull.
  • Writable data lives in $XDG_DATA_HOME/LocalText2Voice (~/.local/share/LocalText2Voice) via app/utils/paths.py:app_data_root().
  • If Qt multimedia warnings about FFmpeg appear on startup, they are informational - playback still uses the system FFmpeg.