67 lines
2.3 KiB
Markdown
67 lines
2.3 KiB
Markdown
# Daily Digest Podcast - Implementation Complete
|
|
|
|
**Task:** Research and build a digital podcast solution for the Daily Digest blog
|
|
**Status:** ✅ Complete (Ready for Review)
|
|
**Date:** 2026-02-23
|
|
|
|
## What Was Delivered
|
|
|
|
### 1. Research Findings
|
|
- **TTS Providers**: OpenAI TTS ($2-4/month, best quality) vs Piper (free, local) vs macOS say (free, basic)
|
|
- **Audio Storage**: Supabase Storage (free tier 1GB)
|
|
- **RSS Feed**: Next.js API route with iTunes extensions
|
|
- **Integration**: Async TTS generation triggered after digest post
|
|
|
|
### 2. Implementation
|
|
|
|
**Core Services:**
|
|
- `src/lib/tts.ts` - Multi-provider TTS abstraction (OpenAI, Piper, macOS)
|
|
- `src/lib/storage.ts` - Supabase Storage audio upload/management
|
|
- `src/lib/podcast.ts` - RSS generation utilities
|
|
|
|
**API Endpoints:**
|
|
- `GET /api/podcast/rss` - RSS 2.0 feed for podcast apps
|
|
- `POST /api/digest` - Updated to trigger async TTS generation
|
|
|
|
**UI Pages:**
|
|
- `/podcast` - Web audio player with episode listing
|
|
- Updated blog post view with inline audio player
|
|
- Added podcast link to header navigation
|
|
|
|
**Scripts:**
|
|
- `npm run generate-tts -- <post_id>` - Generate audio for specific post
|
|
- `npm run generate-tts:all` - Batch generate for all missing posts
|
|
|
|
**Documentation:**
|
|
- `PODCAST_SETUP.md` - Complete setup guide
|
|
- `PODCAST_ARCHITECTURE.md` - Architecture overview
|
|
- `PODCAST_SUMMARY.md` - Implementation summary
|
|
|
|
### 3. Key Features
|
|
- ✅ Automatic TTS when new digest posted
|
|
- ✅ RSS feed compatible with Apple Podcasts, Spotify
|
|
- ✅ Web audio player on podcast page
|
|
- ✅ Multiple TTS provider options
|
|
- ✅ Free tier coverage (Supabase storage + Piper TTS = $0)
|
|
- ✅ Async processing (doesn't block post creation)
|
|
|
|
### 4. URLs
|
|
- RSS Feed: `https://blog-backup-two.vercel.app/api/podcast/rss`
|
|
- Podcast Page: `https://blog-backup-two.vercel.app/podcast`
|
|
|
|
### 5. Next Steps for Deployment
|
|
1. Run SQL to add audio columns to database
|
|
2. Add `ENABLE_TTS=true` and `OPENAI_API_KEY` to Vercel env vars
|
|
3. Deploy: `npm run build && vercel --prod`
|
|
4. Submit RSS to Apple Podcasts, Spotify
|
|
|
|
### 6. Files Modified/Created
|
|
- 13 files changed, 1792 insertions
|
|
- Committed to main branch
|
|
|
|
## Cost Analysis
|
|
- OpenAI TTS: ~$2-4/month (optional)
|
|
- Piper/macOS TTS: $0
|
|
- Supabase Storage: $0 (within free tier)
|
|
- Total: $0-4/month depending on TTS provider
|