- When sending incremental report sync updates, extraction_stats array was empty
- This happened because extraction stats were only included if they were new/updated
- But when new bets are placed on matches with existing extraction stats, those stats weren't included
- Now extraction stats are included for ALL matches referenced by bets being synced
- This ensures server receives match information for all matches that have bets placed
- Matches without extraction stats (not completed) get minimal entries with default values

- Collect unique match IDs from bets being synced
- Query and include match data for all matches referenced by bets
- Add 'matches' array to report sync payload with full match details
- Fixes issue where extraction_stats was empty and matches were not sent with bets

- Changed payout calculation from ExtractionStatsModel.total_redistributed to sum of winning bet win_amount
- Now correctly calculates payout by summing win_amount for all winning bet details in the selected timeframe
- Uses is_bet_detail_winning() function to determine winning bets
- Fixes issue where payout showed 0 instead of actual winning amounts

- Added match_number field to extraction stats data sent during incremental sync
- This allows proper correlation of extraction stats with match results on server side
- Fixes issue where extractions_stats field had no relative matches with results

- Add joinedload for bet_details relationship when querying bets
- This ensures bet_details are loaded for ALL bets, not just synced ones
- Fixes total_payout showing 0 when there are winning bets
- Summary is calculated from ALL bets (synced and not synced) as required for server reports

- Fix _calculate_summary() to check for result == 'win' (correct enum value)
- Previously checked for result == 'won' which doesn't exist in BetResult enum
- Add safety check for bet_details to handle lazy loading
- Fixes total payout showing 0 when there are winning bets

- Pass 'bets' and 'stats' (all items in date range) to _calculate_summary()
- Previously passed 'bets_to_sync' and 'stats_to_sync' (only new/changed items)
- Summary now correctly reflects total payin, payout, and net profit for entire date range
- Fixes net payout showing 0 when only incremental data is being synced

- Fix _calculate_summary() to sum win_amount from bet details where result is 'won'
- Previously used total_redistributed which is amount redistributed to players, not payout to winners
- This fixes total_payout showing as 0 when there are winning bets
- Net profit calculation now works correctly

- Add complete logic for checking timestamp mismatch
- Ensure all code paths return tuple (needs_recovery, is_full_resync)
- Fixes 'cannot unpack non-iterable bool object' error

- Check if rustdesk_id is configured before including client_id in report data
- Use rustdesk_id value as client_id when configured
- Set client_id to None if rustdesk_id is not set
- Prevents sending machine-generated client_id when rustdesk_id is not configured

- Modify needs_recovery() to return tuple (needs_recovery, is_full_resync)
- Skip recover_local_tracking() when full resync is needed
- This prevents creating new sync tracking record that would cause incremental sync
- Ensures full sync is actually performed when server requests it

- Check if rustdesk_id is configured before sending client_id parameter
- Use rustdesk_id value as client_id when configured
- Do not send client_id parameter if rustdesk_id is not set
- Fixes 400 error when server requires client_id only when configured

- Make client_id parameter optional in query_server_last_sync()
- Auto-generate client_id if not provided using _get_client_id()
- Add debug logging for client_id parameter
- Fixes 400 error when server requires client_id parameter

- Add detection of server full resync indication (needs_full_resync flag or null last_sync_id)
- Add _clear_local_tracking() method to force full sync
- Modify needs_recovery() to check for full resync conditions
- Modify collect_report_data() to force full sync when needed
- Add comprehensive test suite for full resync scenarios
- Add implementation documentation

Handles cases where server indicates full resync is needed:
- Server responds with needs_full_resync=true
- Server responds with null last_sync_id
- Server returns 404 (no record for client)

- Add cap_percent field to MatchModel database schema (default: 70.0)
- Save cap_percent when match completes in games_thread.py
- Include cap_percent in reports sync request payload
- Add Migration_041_AddCapPercentToMatches for database migration

This ensures historical tracking of the CAP percentage configured at the moment each match ends, allowing for accurate analysis and reporting.

- Update version from 1.0.12 to 1.0.13 in main.py, build.py, mbetterclient/web_dashboard/app.py, mbetterclient/config/settings.py, mbetterclient/__init__.py
- Update user agent from 1.0r12 to 1.0r13 in mbetterclient/api_client/client.py and test_reports_sync_fix.py
- Add time filtering feature to reports page with start/end time controls
- Update daily-summary, match-reports, and download-excel API endpoints to support time filtering
- Update CHANGELOG.md with version 1.0.13 entry

- Updated version strings in main.py, build.py, mbetterclient/__init__.py, mbetterclient/web_dashboard/app.py
- Updated user agent from MbetterClient/1.0r12 to MbetterClient/1.0r13 in mbetterclient/config/settings.py
- Updated user agent in test_reports_sync_fix.py
- Added changelog entry for version 1.0.13

- Add accumulated_shortfall field to MatchModel to store historical shortfall at match completion
- Create Migration_040_AddAccumulatedShortfallToMatches for database schema update
- Update extraction flow in games_thread.py to store accumulated shortfall in match records
- Update sync report logic in client.py to use match's stored accumulated shortfall value
- Add test_last_sync_query.py to verify implementation
- Add CLIENT_SYNC_MINIMAL_PROMPT.md with API specifications

This ensures accurate reporting of cap compensation balance as it existed at the time each match was completed, rather than using the current global value.

- Add ReportsSyncQueueModel database model for persistent sync queue storage
- Update ReportsSyncResponseHandler to use database instead of JSON files
- Add transaction safety with rollback support for queue operations
- Implement automatic cleanup of completed items
- Add database indexes for efficient querying
- Create comprehensive test suite for database-based queue (test_reports_sync_db.py)
- Update documentation to reflect database implementation

This change improves performance, data integrity, and scalability of the
reports synchronization feature by migrating from JSON file storage to SQLite database.

- Modified video input to not loop intro/match/result videos (only black screen loops)
- Added automatic transition to black screen when non-looping videos end
- Restart HLS process during video switches for clean segment generation
- Added robust error handling for connection resets during transitions
- Increased HLS playlist size and disabled caching for smoother transitions
- Fixed video switching logic to ensure proper HLS content updates

- Update version numbers across all files (main.py, __init__.py, settings.py, build.py, web_dashboard)
- Add comprehensive changelog entry for 1.0.11 documenting the venue timezone system implementation
- Document the fix for UTC midnight day change issues and venue timezone configuration

- Add timezone_utils.py with system timezone auto-detection
- Configure venue timezone with system timezone as default fallback
- Update games_thread.py to use venue timezone for day calculations
- Update web dashboard API routes to use venue timezone for date filtering
- Add venue timezone configuration to web UI
- Fix day boundary issues where UTC midnight didn't match venue local time
- Ensure fixtures and bets are correctly associated with venue days

This resolves the problem where operations stored in UTC caused incorrect
day change calculations when midnight UTC didn't correspond to actual
midnight in the venue's timezone.

 * print now is automatically on bet insertion

- Modified mbetter_latest.md5.txt to contain two lines:
  client:md5sum_of_MbetterClient_binary
  zip:md5sum_of_mbetterc_latest.zip
- Provides integrity verification for both the executable and distribution package

- After successful binary creation, create mbetterc_latest.zip containing the executable
- Generate MD5 checksum of the zip file as mbetter_latest.md5.txt
- Both files are created in the dist directory for easy distribution

- Changed build version from 1.2.11 to 1.0.10 in build.py to match application version
- Updated AI.PROMPT with instructions for updating build version in addition to runtime version
- Ensures consistent versioning across application runtime, build system, and packages

- Updated version numbers across all files: main.py, mbetter_discovery.py, mbetterclient/__init__.py, mbetterclient/config/settings.py, mbetterclient/web_dashboard/app.py
- Changed 1.0.9 to 1.0.10 and 1.0r9 to 1.0r10
- Added AI.PROMPT at project root with version update instructions for future use

- Add runtime CSS loading detection for Bootstrap and Font Awesome
- Download local copies of Bootstrap 5.3.0 and Font Awesome 6.0.0
- Modify base.html to check CDN loading and fallback to local versions
- Add AI.PROMPT file with offline compatibility requirements
- Ensure web interface works when internet connection is down