DIGIKAM DEEP-LEARNING MODEL FILES REPOSITORY ============================================ digiKam uses huge deep-learning pre-trained model files to process items. These heacy files are not includes in source code tarball and are loaded in demand at run-time. Files are hosted in these sub-directories: - aestheticdetector/ For the aesthetic contents detector. - aitools/ For the image processing AI based tools. - autotags/ For the auto-tags engines. - facesengine/ For the face management. - llm/ For the native language search engine. DIGIKAM EXIFTOOL ROLLING RELEASE REPOSITORY =========================================== ExifTool project do not host a repository of archives to the rolling release of the public tarball. To allows to build the digiKam bundles tested with a specific version of ExifTool the exiftool/ subdirectory host the current validated tarballs. Updating file is done with the update_exiftool.sh script: https://invent.kde.org/graphics/digikam/-/blob/master/project/scripts/update_exiftool.sh?ref_type=heads DIGIKAM CONTINUOUS INTEGRATION REPORTS ====================================== In the reports/ sub-directory are hosted the archives generated by the static codes analyzers runs over the whole code of digiKam. Scripts used to generated these reports are hosted here: https://invent.kde.org/graphics/digikam/-/tree/master/project/reports?ref_type=heads DIGIKAM API DOCUMENTATION ========================= The source code API documentation used by the project are hosted in the api/ sub-directory. Two files generated with Doxygen are provided: * a compressed archive with the HTML version usable in the web browser. * a PDF version. DIGIKAM WEEKLY PRE-VERSION BUNDLES REPOSITORY FOR TESTING ========================================================= The bundle files provided in this repository are built weekly and can be used to test new features quickly. Do not use these versions in production, unless you know what are you doing. Linux universal AppImage bundles: --------------------------------- * digiKam-version-timestamp-Qt6-x86-64.appimage => for Linux 64 bits. * digiKam-version-timestamp-Qt6-x86-64-debug.appimage => for Linux 64 bits with debug symbols. These bundles are compatible with all Linux distributions which support Fuse file system. It must be runnable on all distributions older than 3/4 years. Make file executable and run it as well. Nothing is installed and do not require root rights. See here for details : http://askubuntu.com/questions/774490/what-is-an-appimage-how-do-i-install-it Bundle executable can take arguments to run: without option : run digiKam. : run Showfoto instead digiKam. : run digiKam and wait at startup to be attached to a remote GDB instance (only suitable with AppImage compiled with debug symbols). : run Showfoto and wait at startup to be attached to a remote GDB instance (only suitable with AppImage compiled with debug symbols). : show these information. To run the AppImage debug version into GDB, see the instructions here: https://www.digikam.org/contribute/#appimage-bundle Windows bundle installers: -------------------------- * digiKam-version-timestamp-Qt6-win64.exe => for Windows 64 bits. * digiKam-version-timestamp-Qt6-win64-debug.exe => for Windows 64 bits with debug symbols. These bundles are built with Microsoft VCPKG environnement. They do not require extra Microsoft run-time libraries and are compatible with Windows >= 10. Windows Portable archives: -------------------------- * digiKam-version-timestamp-Qt6-win64.tar.xz => for Windows 64 bits. * digiKam-version-timestamp-Qt6-win64-debug.tar.xz => for Windows 64 bits with debug symbols. Self archives of Windows installer contents to make Portable versions of digiKam. This do not requires admin rights to install digiKam in your Windows home directory as well. MacOS bundle package: --------------------- * digiKam-version-timestamp-Qt6-MacOS-arch.pkg => for MacOS 64 bits. * digiKam-version-timestamp-Qt6-MacOS-arch-debug.pkg => for MacOS 64 bits with debug symbols. Arm64 annoted bundles are built with Macports for Silicon computer with backward compatibility to macOS 11.3. Intel annoted bundles are built with Macports for Intel computers with backward compatibility to macOS 11.3. They can be used on Apple Silicon based computers within Rosetta 2 emulator, but it's highly recommended to use native arm64 bundles instead. Notes : ------- * To know the file SHA256 check-sums, use this kind of url: https://files.kde.org/digikam/_bundle_file_name_.sha256 They can be used to verify if bundles files downloaded are corrupted or not. * The bundle build-timestamp included in file name follow this ISO scheme : YYYYMMTDDhhmmss, with Y : year M : month D : day T : separator h : hour m : minute s : seconds * Bundles name with '-debug' suffix are compiled with full debug symbols. File sizes are more heavy but it's possible to run application in debugger and to report dysfunctions to developpers with a backtrace to hack and fix quickly the problems. * The bundles continuous deployement log files compressed with gzip are stored in build.logs/ sub-directory.