- Fixed wssshc to act as SSH client instead of SSH server
- wssshc now connects to target SSH server (localhost:22) when receiving tunnel_request
- Removed incorrect SSH server listening code that was binding to port 22
- Implemented proper SSH client connection forwarding
- Fixed 'Address already in use' error by removing port 22 binding conflict
- Corrected tunnel architecture: wssshc → SSH server, not SSH server → wssshc
- Updated handle_tunnel_request to establish SSH client connection to target
- Removed accept_ssh_connection function (not needed for client mode)
- Fixed bidirectional forwarding between WebSocket and SSH client connection
- Resolved socket lifecycle issues by using correct connection direction
- Eliminated port binding conflicts with system SSH server
- Implemented proper SSH protocol flow for client-side tunneling

- Added detailed error reporting when socket becomes invalid
- Mark tunnel as inactive when SSH connection is broken
- Enhanced debugging for socket closure detection
- Added errno reporting for connection errors
- Improved error messages to distinguish between different failure modes
- Added proper tunnel state management when connections fail
- Better handling of SSH client disconnections during handshake
- Added mutex protection for tunnel state updates during errors

- Added fcntl() check to validate socket before sending data
- Enhanced error reporting with socket descriptor and errno details
- Added detailed send operation logging for troubleshooting
- Improved socket state validation to prevent bad file descriptor errors
- Added errno reporting for better error diagnosis
- Enhanced debugging output for socket operations
- Added socket validity checks using file descriptor flags
- Improved error messages with more context information

- Added proper seeding of random number generator with time and PID
- Prevents duplicate request IDs across program runs
- Uses srand() with combination of current time and process ID
- Ensures each tunnel session gets a truly unique request ID
- Added thread-safe seeding with static flag to prevent reseeding
- Fixed potential security issue with predictable request IDs
- Improved randomness quality for tunnel identification

- Added proper mutex protection for active_tunnel access
- Fixed race condition between main thread and forwarding thread
- Protected SSL connection access with mutex locks
- Fixed socket setting with proper mutex synchronization
- Ensured thread-safe access to tunnel state variables
- Prevented concurrent access to shared tunnel resources
- Added proper SSL connection lifecycle management
- Fixed potential data races in WebSocket message handling

- Added check for valid target socket before attempting to send data
- Prevents 'Bad file descriptor' errors when local socket isn't ready yet
- Added debug message when target socket is not available
- Ensures tunnel_data messages are ignored until SSH connection is established
- Fixes race condition between WebSocket message arrival and SSH client connection

- Improved handle_tunnel_data() with better hex validation and decoding
- Added bounds checking for hex data length validation
- Enhanced error handling for hex decoding failures
- Added debug logging for hex decoding issues
- Improved socket send error handling with detailed error messages
- Added bounds checking for hex encoding in forwarding functions
- Fixed potential buffer overflow issues with large hex data
- Enhanced robustness for handling large SSH protocol packets

- Modified handle_tunnel_request() to connect to local SSH server (localhost:22)
- Added forward_ws_to_local() function for WebSocket to local SSH forwarding
- Updated handle_tunnel_data() to properly route data to correct socket
- wssshc now establishes bidirectional forwarding between WebSocket and local SSH server
- Fixed tunnel blocking issue by implementing proper data flow in both directions
- Added thread for handling data from local SSH server back to WebSocket

- Added send_registration_message() function to websocket library
- Updated wssshc.c to use send_registration_message() instead of send_json_message()
- Registration message now sends: {"type":"register","client_id":"<id>","password":"<password>"}
- Instead of: {"type":"register","client_id":"<id>","request_id":"<password>"}
- Server can now properly authenticate clients with passwords

- Added patterns for new shared library build artifacts (*.a, *.so)
- Added patterns for compiled library executables (wssshlib, websocket, wssh_ssl, tunnel)
- Enabled logos/ directory exclusion (generated assets)
- Added comprehensive Debian packaging artifact patterns
- Added build info file patterns (*_*.buildinfo)
- Ensures clean git status by excluding all build-generated files

- Removed incorrect 'closed' attribute check on websockets Server object
- Added proper exception handling for server shutdown
- Fixed 'Server' object has no attribute 'closed' error during shutdown
- Server now shuts down cleanly without throwing exceptions

- Modified wssshd.py to accept both 'id' and 'client_id' fields for registration
- Added fallback logic: client_id = data.get('client_id') or data.get('id')
- This allows the C client to send 'client_id' while maintaining backward compatibility
- Fixes the KeyError: 'id' issue when C client tries to register

- Extracted ~1500+ lines of duplicate code into shared libraries:
  * wssshlib.h/.c - Shared utilities (config, flags, ports, IDs)
  * websocket.h/.c - WebSocket functions (handshake, framing, messaging)
  * wssh_ssl.h/.c - SSL/TLS setup functions
  * tunnel.h/.c - Tunnel management (setup, reconnection, data handling)

- Fixed critical port option behavior:
  * -p/-P options now correctly specify wssshd server port
  * Removed misleading documentation about options being passed through
  * Updated help text and man pages to reflect correct behavior

- Cleaned up help documentation:
  * Removed duplicate 'Options:' sections from wsssh and wsscp
  * Consolidated all options into single, clean sections

- Updated build system:
  * Modified configure.sh and Makefile for new library structure
  * Proper linking of wsssh and wsscp against shared libraries

- Updated documentation:
  * CHANGELOG.md - Added version 1.4.6 with comprehensive details
  * README.md - Updated architecture and project structure
  * DOCUMENTATION.md - Reflected new library structure
  * Man pages - Corrected port option examples
  * TODO.md - Updated completion status

- Maintained backward compatibility while improving maintainability

- Delete wssshc.py, wsssh.py, wsscp.py files
- Update build.sh to remove pyinstaller commands and output messages
- Update Debian packaging to exclude Python script installations
- Update documentation to reflect C-only implementations
- Update CHANGELOG.md with version 1.4.5 entry
- Focus on C implementations for better performance and minimal dependencies

- Fix navbar brand link to point to dashboard home page
- Add back button to terminal page for easy navigation to dashboard
- Improve user experience with better navigation between pages
- Maintain consistent navigation throughout the web interface

- Updated CHANGELOG.md with new features and fixes for v1.4.4
- Updated README.md with version 1.4.4 information and new features
- Updated DOCUMENTATION.md with enhanced feature descriptions
- Updated TODO.md to reflect completed tasks and new version
- Updated Debian changelog files for both wsssh-server and wsssh-tools packages
- Added comprehensive documentation for:
  * Dynamic terminal sizing and resize support
  * Force echo mode implementation
  * Enhanced logging with logrotate integration
  * Terminal dimension fixes and improvements
  * Backend API enhancements for terminal management

- Calculate terminal dimensions after initial fit to container
- Store initial dimensions for use during connection
- Use calculated dimensions for both startup and resize events
- Ensure proper terminal sizing for applications like top
- Handle browser window resize to update remote terminal dynamically

- Modify wssshd.py to run 'stty echo' before launching wsssh
- Use shell command wrapper to execute stty echo && wsssh
- Ensure echo mode is forced before SSH connection is established
- Remove JavaScript-based stty echo attempts
- Fix terminal echo issue by setting echo mode at the right time

- Remove JavaScript-based echo handling and password mode detection
- Send all keystrokes directly to server without local echo manipulation
- Let SSH server handle echo control through terminal protocol
- Proper terminal behavior where server controls what gets displayed
- No more manual echo logic, respects SSH server's echo settings

- Add passwordMode flag to detect password prompts from SSH server
- Disable local echo when password prompts are detected (Password:, password:, etc.)
- Re-enable echo after password is submitted (on Enter key or new prompt)
- Reset passwordMode on disconnect for clean state
- Maintain security by not displaying passwords in plain text
- Preserve normal command echo for non-password input

- Add proper xterm.js configuration with disableStdin: false and other settings
- Implement local echo handling for printable characters, backspace, and enter
- Ensure terminal input is visible to user while maintaining server communication
- Fix input handling to properly display typed commands in the web interface

- Update .terminal-container CSS to use calc(100vh - 200px) for responsive height
- Remove fixed padding and height constraints
- Add xterm-addon-fit for proper terminal resizing
- Update terminal initialization to use fit addon for dynamic sizing
- Ensure terminal fills entire container width and height
- Add window resize event handler for responsive terminal sizing

- Add logging section to README.md with logrotate configuration details
- Add comprehensive Logging Configuration section to DOCUMENTATION.md
- Update CHANGELOG.md with version 1.4.4 entry for logging features
- Update table of contents in DOCUMENTATION.md to include logging section

Features documented:
- Automatic log rotation with logrotate
- Log file locations and management
- Manual log management commands
- Troubleshooting with logs
- Log analysis techniques

Use start-stop-daemon --output option to properly redirect both stdout and stderr
to /var/log/wssshd/wssshd.log while preserving correct PID storage in the PID file

- Add wssshd.log and wssshc.log to respective logrotate configurations
- Modify init scripts to redirect stdout and stderr to log files
- Ensure proper log rotation with weekly rotation, compression, and HUP signals

- Create PID file in start case with correct background process PID
- Remove PID file creation from main() function to avoid empty files
- Validate background process is running before reporting success
- Clean up PID file if process fails to start
- Fix both wssshd and wssshc watchdogs

- Capture PID of background main() process with MAIN_PID=
- Store correct PID in watchdog PID file instead of main() function PID
- Prevent PID file from containing exited process PID
- Fix both wssshd and wssshc watchdogs
- Watchdog PID file now contains PID of actual running process

- Separate PID file reading from process checking
- Prevent command substitution failures in kill command
- Add proper error handling for PID file operations
- Fix false 'watchdog failed' reports due to bash syntax issues
- Fixed both wssshd and wssshc init scripts

- Add detailed logging to main() monitoring loops
- Track loop iterations and daemon status checks
- Log sleep intervals and monitoring status
- Help diagnose why watchdog exits prematurely
- Fixed both wssshd and wssshc watchdogs

- Stop existing watchdog before starting new one
- Use pgrep to detect running watchdog processes
- Prevent 'watchdog already running' conflicts
- Ensure clean watchdog startup in init scripts
- Fixed both wssshd and wssshc init scripts

- Removed duplicate function definitions causing bash arithmetic errors
- Kept the more complete check_restart_limits() function with awk timestamp parsing
- Fixed both wssshd and wssshc watchdog scripts
- Eliminates 'integer expression expected' errors

- Init scripts now start wssshd/wssshc daemons directly
- Wait for daemons to be fully running before starting watchdog
- Watchdog only monitors existing daemons, doesn't start them
- Watchdog exits if daemon restart fails
- Proper startup sequence: daemon → verify running → watchdog
- Fixed both wssshd and wssshc init scripts and watchdogs

- Watchdog now waits 3 seconds after starting daemon before checking status
- Prevents false detection of daemon failure during startup
- Gives daemon time to fully initialize and create PID file
- Fixed both wssshd and wssshc watchdogs
- Improves reliability of daemon startup detection

- Use setsid to properly daemonize main() function
- setsid bash -c 'main' & creates independent process group
- Prevents main() from being killed when parent script exits
- Fixed both wssshd and wssshc watchdogs
- Watchdog processes should now persist correctly

- Watchdog now properly daemonizes by running main() in background
- Removed complex PID file waiting logic that was causing issues
- Simplified start case to just run main() and check for PID file
- Fixed both wssshd and wssshc watchdogs
- Watchdog processes should now persist after script exits

- Stop functions now kill daemon processes regardless of PID file existence
- Added fallback pkill mechanism for processes without PID files
- Uses graceful TERM signal followed by force KILL signal
- Fixed both wssshd and wssshc init scripts
- Ensures complete process cleanup in all scenarios