Skip to content

CHANGELOG.md

Latest commit

 

History

History
124 lines (102 loc) · 4.93 KB

CHANGELOG.md

File metadata and controls

124 lines (102 loc) · 4.93 KB
Raw

Changelog

All notable changes to Fragforce Reads will be documented in this file.

The format is based on Keep a Changelog, and this project adheres to Semantic Versioning.

1.0.3 - 2026-07-01

Fixed

  • Release workflow now uses fragforce-automation app token instead of GITHUB_TOKEN. Events triggered by GITHUB_TOKEN will not create new workflow runs to prevent recursive workflows, which prevented the prod image workflow from triggering on release events.

1.0.2 - 2026-07-01

Fixed

  • Production image workflow now triggers on release publication in addition to tag pushes

1.0.1 - 2026-07-01

Fixed

  • Release workflow now correctly extracts changelog content for single/first version entries
  • Release workflow now tags the correct commit (main merge commit) instead of dev branch head
  • Dev branch now automatically syncs after main merges to prevent divergence

1.0.0 - 2026-07-01

Initial MVP release of Fragforce Reads - an audio playback service for the VTO Book Reading project.

Added

Core Recording Features

  • Browser-based audio recording workflow with mic access checks
  • Pre-flight checklist for copyright compliance before recording
  • Upload progress bar with user-friendly error handling
  • Automatic audio remux pipeline for seeking support with ffmpeg
  • Recording status tracking (pending, processing, ready, failed)
  • Narrator dashboard showing available books and personal recordings
  • Recording detail page with playback and flag-for-review functionality
  • Profile page for narrators to update name and email
  • Re-record capability for flagged recordings

QR Code System

  • Auto-generated QR codes with short codes and passwords
  • QR code labels with book info, narrator name, and password
  • Printable QR sheet view for admins
  • Short URL redirect (/q/<code>) to playback pages
  • Case-insensitive password entry with rate limiting

Playback & Security

  • Public playback page with password protection
  • Range request support for audio seeking
  • License expiry enforcement at playback time
  • Session-based password unlocking (7-day sessions)
  • Rate limiting on password attempts (5 attempts, 5-minute lockout)
  • Narrator attribution visible before password entry
  • Duration display in mm:ss format

Registration & Authentication

  • Event code registration with rate limiting
  • Invite link registration system
  • Passphrase login with rate limiting
  • Case-insensitive login
  • Logout endpoint
  • Welcome/login flow at /login/

Admin Features

  • Book management with max narrator limits
  • Narrator management with passphrase generation
  • Recording admin with status filters and retry action
  • QR code management with admin links
  • Event code and invite link management

Infrastructure & Deployment

  • Docker containerization (dev and prod variants)
  • PostgreSQL 18.3 support
  • GitHub Actions CI/CD (lint, test, coverage, SonarCloud)
  • Docker image builds for dev and prod
  • Whitenoise static file serving
  • SSL security settings with SECURE_SSL flag
  • Health check endpoint (/healthz/) for internal networks
  • Persistent database connections (CONN_MAX_AGE=600)
  • Non-root container execution
  • Media file storage and serving

Testing & Quality

  • Comprehensive test suite (146+ tests)
  • Coverage reporting
  • SonarCloud quality gate integration
  • Per-app test package structure
  • CODEOWNERS for CI/build config protection

Security

  • Rate limiting on login, event registration, and playback passwords
  • Upload size validation (100MB limit)
  • Recording duration bounds validation (max 3600s)
  • Attestation text length validation (max 5000 chars)
  • HTTP method restrictions on all views
  • CSRF protection
  • Session security with 7-day expiry
  • File handle cleanup in range-request serving
  • Healthz endpoint restricted to internal networks only

Changed

  • Simplified QR password format from 4alphanumeric-word-word to word-word-2digits
  • Narrator attribution now visible before password entry
  • Audio files served through view with auth checks (no direct static serving)
  • Recovered stuck recordings on app startup (gunicorn/runserver only)

Fixed

  • N+1 query optimization in dashboard view
  • File handle leak in range-request audio serving
  • Max narrators check when set to 0
  • Visited link color on buttons
  • Audio URL routing conflicts
  • Admin retry action queryset re-evaluation
  • Remux claim race condition with SELECT FOR UPDATE SKIP LOCKED
  • Signal registration and file cleanup on recording delete
  • HTML accessibility and CSS deprecation warnings
  • Cognitive complexity in views (SonarCloud)