Sanmill v4.0.1

• Fix bugs of Perfect AI.

Sanmill v4.0.0

• Introduced support for Perfect AI, enhancing challenge and engagement levels. Acknowledgments to Gábor Gévay and Gábor Danner.
• Enabled the transfer of PGN files from phones to PCs.
• Updated to Material3 design.

Sanmill v3.8.0

Release Notes

Localization Updates

• Synced Brazilian Portuguese localization.
• Changed App name for Arabic audience. More details can be found here.
• For the Albanian audience, the game Nine Men's Morris will now be called "Mulli", in line with the theme of the game.
• Changed the Finnish App name to Mylly. More information can be found on Finnish Wikipedia.

CI/CD Improvements

• Upgraded Flutter to v3.13.8 and v3.13.6.
• Updated qt-on-windows.yml.
• Upgraded jurplel/install-qt-action to v3.

Snap Fixes and Enhancements

• Modified ci-upgrade-flutter-version.sh in Snap.
• Changed the base from core18 to core22, resolving runtime issues with GLIBC versions.
• Added comment for how to install snapd in snap.sh.

Artificial Intelligence

• Adjusted the AI's search depth for White's 5th and 6th moves, addressing issue #744. The adjustment alters the outcome in certain sequences.

Flutter App

• Updated pubspec.yaml to pin package versions for increased stability.
• Upgraded package version in pubspec.yaml.
• Changed flutter_email_sender to v6.0.2, adding Java 1.8 compile options and Kotlin JVM target in build.gradle.

Qt App

• Resolved issues in human player draw repetition judgment.
• Added install-qt-env.sh.

Miscellaneous

• Fixed issue with detection of repeated game positions in CPP.
• Improved code structure and resolved Clang build warnings in build phase.

Sanmill v3.7.0

• The Qt application now formally introduces support for integrating the Perfect Database.
  Note: Mobile applications are not yet equipped to support loading the Perfect Database.
  For access to the Perfect Database feature, kindly opt for the PC version of the software.

• Enhanced language translations across multiple locales.

Sanmill v3.5.0

New Features

• Implemented the MCTS algorithm, similar to the one used by AlphaGo, for a more intuitive gameplay experience.
• Added the MCTS to the Algorithms menu, allowing users to select and explore different game algorithms.

Known Issues

• The MCTS algorithm is still in the early development stage and may not exhibit strong performance.
• Gameplay with the MCTS algorithm focuses more on intuition rather than precise calculations.

We encourage you to download and try out version 3.5.x, and kindly share your feedback with us. Your valuable input will help us continuously improve and enhance the game experience. Thank you for your ongoing support!

Sanmill v3.4.2

Support some Chinese rules.

Sanmill v3.3.3

Fixed several issues with locales.

Sanmill v3.3.0

Fully support iOS/Windows/Linux.

Sanmill v3.0.0

We're excited to announce that the new version now fully supports the Linux Desktop environment for the first time!

Some bugs in the Windows Desktop version have also been fixed.

Sanmill v2.0.20

Fix tutorial.