Files
TimeTracker/APPLY_FIXES_NOW.md
Dries Peeters 20824dbcb1 feat: Add customizable Kanban board columns and enhance CSRF configuration
This commit introduces a comprehensive Kanban board customization system and
improves CSRF token configuration for Docker deployments.

## Major Features

### 1. Customizable Kanban Board Columns
Add complete kanban column customization system allowing users to define
custom workflow states beyond the default columns.

**New Components:**
- Add KanbanColumn model with full CRUD operations (app/models/kanban_column.py)
- Add kanban routes blueprint with admin endpoints (app/routes/kanban.py)
- Add kanban column management templates (app/templates/kanban/)
- Add migration 019 for kanban_columns table (migrations/)

**Features:**
- Create unlimited custom columns with unique keys, labels, icons, and colors
- Drag-and-drop column reordering with position persistence
- Toggle column visibility without deletion
- Protected system columns (todo, in_progress, done) prevent accidental deletion
- Complete state marking for columns that should mark tasks as done
- Real-time updates via SocketIO broadcasts when columns change
- Font Awesome icon support (5000+ icons)
- Bootstrap color scheme integration
- Comprehensive validation and error handling

**Integration:**
- Update Task model to work with dynamic column statuses (app/models/task.py)
- Update task routes to use kanban column API (app/routes/tasks.py)
- Update project routes to fetch active columns (app/routes/projects.py)
- Add kanban column management links to base template (app/templates/base.html)
- Update kanban board templates to render dynamic columns (app/templates/tasks/)
- Add cache prevention headers to force fresh column data

**API Endpoints:**
- GET /api/kanban/columns - Fetch all active columns
- POST /api/kanban/columns/reorder - Reorder columns
- GET /kanban/columns - Column management interface (admin only)
- POST /kanban/columns/create - Create new column (admin only)
- POST /kanban/columns/<id>/edit - Edit column (admin only)
- POST /kanban/columns/<id>/delete - Delete column (admin only)
- POST /kanban/columns/<id>/toggle - Toggle column visibility (admin only)

### 2. Enhanced CSRF Configuration
Improve CSRF token configuration and documentation for Docker deployments.

**Configuration Updates:**
- Add WTF_CSRF_ENABLED environment variable to all docker-compose files
- Add WTF_CSRF_TIME_LIMIT environment variable with 1-hour default
- Update app/config.py to read CSRF settings from environment
- Add SECRET_KEY validation in app/__init__.py to prevent production deployment
  with default keys

**Docker Compose Updates:**
- docker-compose.yml: CSRF enabled by default for security testing
- docker-compose.remote.yml: CSRF always enabled in production
- docker-compose.remote-dev.yml: CSRF enabled with production-like settings
- docker-compose.local-test.yml: CSRF can be disabled for local testing
- Add helpful comments explaining each CSRF-related environment variable
- Update env.example with CSRF configuration examples

**Verification Scripts:**
- Add scripts/verify_csrf_config.sh for Unix systems
- Add scripts/verify_csrf_config.bat for Windows systems
- Scripts check SECRET_KEY, CSRF_ENABLED, and CSRF_TIME_LIMIT settings

### 3. Database Initialization Improvements
- Update app/__init__.py to run pending migrations on startup
- Add automatic kanban column initialization after migrations
- Improve error handling and logging during database setup

### 4. Configuration Management
- Update app/config.py with new CSRF and kanban-related settings
- Add environment variable parsing with sensible defaults
- Improve configuration validation and error messages

## Documentation

### New Documentation Files
- CUSTOM_KANBAN_README.md: Quick start guide for kanban customization
- KANBAN_CUSTOMIZATION.md: Detailed technical documentation
- IMPLEMENTATION_SUMMARY.md: Implementation details and architecture
- KANBAN_AUTO_REFRESH_COMPLETE.md: Real-time update system documentation
- KANBAN_REFRESH_FINAL_FIX.md: Cache and refresh troubleshooting
- KANBAN_REFRESH_SOLUTION.md: Technical solution for data freshness
- docs/CSRF_CONFIGURATION.md: Comprehensive CSRF setup guide
- CSRF_DOCKER_CONFIGURATION_SUMMARY.md: Docker-specific CSRF setup
- CSRF_TROUBLESHOOTING.md: Common CSRF issues and solutions
- APPLY_KANBAN_MIGRATION.md: Migration application guide
- APPLY_FIXES_NOW.md: Quick fix reference
- DEBUG_KANBAN_COLUMNS.md: Debugging guide
- DIAGNOSIS_STEPS.md: System diagnosis procedures
- BROWSER_CACHE_FIX.md: Browser cache troubleshooting
- FORCE_NO_CACHE_FIX.md: Cache prevention solutions
- SESSION_CLOSE_ERROR_FIX.md: Session handling fixes
- QUICK_FIX.md: Quick reference for common fixes

### Updated Documentation
- README.md: Add kanban customization feature description
- Update project documentation with new features

## Testing

### New Test Files
- test_kanban_refresh.py: Test kanban column refresh functionality

## Technical Details

**Database Changes:**
- New table: kanban_columns with 11 columns
- Indexes on: key, position
- Default data: 4 system columns (todo, in_progress, review, done)
- Support for both SQLite (development) and PostgreSQL (production)

**Real-Time Updates:**
- SocketIO events: 'kanban_columns_updated' with action type
- Automatic page refresh when columns are created/updated/deleted/reordered
- Prevents stale data by expiring SQLAlchemy caches after changes

**Security:**
- Admin-only access to column management
- CSRF protection on all column mutation endpoints
- API endpoints exempt from CSRF (use JSON and other auth mechanisms)
- System column protection prevents data integrity issues
- Validation prevents deletion of columns with active tasks

**Performance:**
- Efficient querying with position-based ordering
- Cached column data with cache invalidation on changes
- No-cache headers on API responses to prevent stale data
- Optimized database indexes for fast lookups

## Breaking Changes

None. This is a fully backward-compatible addition.

Existing workflows continue to work with the default columns.
Custom columns are opt-in via the admin interface.

## Migration Notes

1. Run migration 019 to create kanban_columns table
2. Default columns are initialized automatically on first run
3. No data migration needed for existing tasks
4. Existing task statuses map to new column keys

## Environment Variables

New environment variables (all optional with defaults):
- WTF_CSRF_ENABLED: Enable/disable CSRF protection (default: true)
- WTF_CSRF_TIME_LIMIT: CSRF token expiration in seconds (default: 3600)
- SECRET_KEY: Required in production, must be cryptographically secure

See env.example for complete configuration reference.

## Deployment Notes
2025-10-11 19:56:45 +02:00

2.8 KiB

Apply These Changes NOW

Step 1: Restart the Application

The files have been updated with aggressive cache clearing. Now restart:

docker-compose restart app

Wait 10 seconds for restart, then proceed.

Step 2: Test Creating a Column

  1. Go to: http://your-domain/kanban/columns
  2. Click "Add Column"
  3. Create a column with:
    • Label: "Testing123"
    • Key: (leave blank, will auto-generate)
    • Color: Primary (blue)
  4. Click "Create Column"

Expected: You should see "Testing123" in the list

Step 3: Check the Kanban Board

  1. Open new tab
  2. Go to: http://your-domain/tasks
  3. Look at the kanban board

Expected: "Testing123" column should appear on the board

Step 4: If Still Not Working

Run these diagnostic commands:

# 1. Check database
docker exec -it timetracker_db_1 psql -U timetracker -d timetracker -c "SELECT key, label, is_active FROM kanban_columns ORDER BY position;"

# 2. Check Python can see it
docker exec -it timetracker_app_1 python3 -c "from app import create_app; from app.models import KanbanColumn; app = create_app(); app.app_context().push(); cols = KanbanColumn.get_active_columns(); print(f'Found {len(cols)} columns:'); [print(f'  - {c.label}') for c in cols]"

# 3. Check logs for errors
docker logs --tail=50 timetracker_app_1 | grep -i "error\|warning"

What Changed

I've added db.session.expire_all() before EVERY query for kanban columns:

  • In /tasks route
  • In /tasks/my-tasks route
  • In /projects/<id> route
  • In /kanban/columns route
  • In /api/kanban/columns endpoint
  • After every column modification

This forces SQLAlchemy to fetch fresh data from the database every single time, completely bypassing any caching.

Performance Note

This adds a tiny bit of overhead (< 1ms per request) but ensures fresh data always.

If STILL Not Working After Restart

Then the issue is one of these:

Issue 1: Changes Not Saving to Database

Check step 4.1 above. If your column isn't in the database, there's a form/validation issue.

Issue 2: Browser Caching

Press Ctrl+Shift+R (hard refresh) after creating column.

Issue 3: Multiple Database Instances

Unlikely, but check if you have multiple postgres containers:

docker ps | grep postgres

Should only show ONE container.

Issue 4: Permission Issues

Check Docker logs:

docker logs timetracker_app_1 2>&1 | tail -100

Look for "Permission denied" or "Access denied" errors.

Report Back

After restart and testing, tell me:

  1. Do you see your column in the database? (Step 4.1)
  2. Does Python see it? (Step 4.2)
  3. Do you see it on /kanban/columns page?
  4. Do you see it on /tasks page?
  5. Any errors in logs? (Step 4.3)

With these answers, I can pinpoint the exact issue!