murtaza-nasir/speakr v0.7.1-alpha

v0.7.1 - Stability & Audio Improvements

on GitHub

What's New

PostgreSQL Support

• Added psycopg2-binary driver for PostgreSQL database option
• Configuration examples updated with PostgreSQL connection strings

Contributed by sakowicz

Audio Download Button

• Explicit download button next to audio player
• Works on mobile browsers where native audio controls hide the download option

Audio Conversion Pipeline Refactor

• Improved format detection using ffprobe instead of file extensions
• Better codec support with automatic conversion for unsupported formats
• Unsupported codecs are automatically converted before transcription

Contributed by JadedBlueEyes

Bug Fixes

Job Queue Race Condition Fix

• Fixed issue where multiple workers could claim the same job simultaneously
• This caused spurious "FFmpeg not found" errors when running with multiple workers
• Now uses atomic database UPDATE to ensure only one worker claims each job

Thanks to Daabramov for detailed debugging that identified this issue

Upgrading

Docker (Recommended):

docker compose pull
docker compose up -d

Manual:

git pull
pip install -r requirements.txt
python src/app.py

Full Changelog: v0.7.0...v0.7.1