Add _routes.json configuration to properly handle direct /r/ and /a/ paths
on Cloudflare Pages deployments. This ensures that client-side routes are
served correctly instead of returning 404 errors.
- Highlights now include p tag referencing original article author
- Allows authors to discover highlights of their work
- Follows NIP-84 best practices for highlight attribution
- Add eslint-disable comments for intentionally unused _settings parameters
- Parameters kept for API compatibility with existing code
- All linter and type checks now pass
- Move refresh button from top bar to end of bookmarks list
- Show relative time of last fetch next to refresh button
- Add 'Explore' button (fa-newspaper icon) to top bar that links to /explore
- Track lastFetchTime in useBookmarksData hook
- Better UX with explore more prominent and refresh less intrusive
- Replace div+onClick with Link components
- Enable CMD+click to open articles in new tabs
- Preserve SPA navigation for normal clicks
- Better UX with standard browser link behavior
- Display 'Connecting' with spinner for first 4 seconds after page load
- Give relays time to establish connections before showing 'Offline'
- Immediately switch to normal state once any relay connects
- Better UX - most refreshes aren't actually offline, just connecting
- Remove complex Cache API management with blob URLs and metadata
- useImageCache now simply returns the URL (Service Worker handles caching)
- imageCacheService reduced to just stats and clear functions
- Service Worker automatically caches all images on fetch
- Much simpler, DRY code that 'just works' for offline mode
- Stats now read directly from Cache API instead of localStorage metadata
- Implement Service Worker to intercept and cache image requests
- Service Worker persists across hard reloads unlike Cache API alone
- Simplify useImageCache hook to work with Service Worker
- Images now work offline even after hard reload
- Service Worker handles transparent cache-first serving for images
- Clean up stale metadata when Cache API doesn't have cached data
- Handle online/offline state properly in image loading
- Show original URL when online, blob URL from cache when offline
- Prevent cache misses when browser clears Cache API on hard reload
- Create exploreService to fetch kind:30023 events from followed users
- Add BlogPostCard component for displaying blog post previews
- Add Explore page component with grid layout
- Add /explore route to App.tsx (not linked in navigation yet)
- Add responsive CSS styles for explore page and blog post cards
- Clicking blog post cards navigates to article view
BREAKING CHANGE: Image cache now uses Cache API instead of localStorage
Benefits:
- Support for actual 210MB cache size (localStorage limited to 5-10MB)
- Store native Response objects (no base64 overhead)
- Asynchronous, non-blocking operations
- Better suited for large binary blobs like images
- Can handle hundreds of MB to several GB
Changes:
- Rewrite imageCacheService to use Cache API for image storage
- Keep metadata in localStorage for LRU tracking (small footprint)
- Update useImageCache hook to handle async Cache API
- Add blob URL cleanup to prevent memory leaks
- Update clearImageCache to async function
The cache now works as advertised and won't hit quota limits.
