* [fuzzylite] fix unix build
* update version scheme
* version
* unsupported on UWP
* remove from CI baseline
* version
* fix typo
* version
* windows only
* version
* Add fuzzylite:x64-osx
* Fix typo
* cr
* version
* Fix string replace on static
* version
* Don't rename artifacts
* version
* Trigger some test ports from vcpkg.cmake changes
* Bump test port version
* Revert "Bump test port version"
* Use old toolchain version for parent hashes
* Add ABI hashing note to vcpkg.cmake
* Test vcpkg.cmake with cmake instead vpckg-ci-paraview
PASSING, REMOVE FROM FAIL LIST: intelrdfpmathlib:x64-linux
PASSING, REMOVE FROM FAIL LIST: ocilib:x64-windows-static
PASSING, REMOVE FROM FAIL LIST: skia:arm64-windows
PASSING, REMOVE FROM FAIL LIST: yato:x64-windows-static-md
REGRESSION: argtable3:arm-uwp
No repro. Maybe fixed by https://github.com/microsoft/vcpkg/pull/23655
REGRESSION: infoware:x64-windows-static
```
-- Found Git: C:/agent/externals/git/cmd/git.exe (found version "2.30.2.windows.1")
Cloning into 'pciids'...
fatal: unable to access 'https://github.com/pciutils/pciids/': Could not resolve host: github.com
CMake Error at CMakeLists.txt:111 (message):
Cloning/pulling pciids repository from https://github.com/pciutils/pciids
failed with 128.
The pci.ids file, downloadable from https://github.com/pciutils/pciids or
http://pci-ids.ucw.cz, is required for building infoware, and cloned
automatically from that GitHub repository by default.
To use a local copy, set INFOWARE_PCI_IDS_PATH to its location.
```
Looks like this port needs to be fixed to work with artifact caching.
REGRESSION: libgpg-error:x64-uwp
```
"D:\buildtrees\libgpg-error\x64-uwp-rel\error-1.42-2324ddbc71.clean\SMP\libgpg-error.sln" (Rebuild target) (1) ->
"D:\buildtrees\libgpg-error\x64-uwp-rel\error-1.42-2324ddbc71.clean\SMP\libgpg-error_winrt.vcxproj" (Rebuild target) (2) ->
(_GetPackageFileExtensions target) ->
C:\Program Files\Microsoft Visual Studio\2022\Enterprise\MSBuild\Microsoft\VisualStudio\v17.0\AppxPackage\Microsoft.AppXPackage.Targets(892,25): error MSB4086: A numeric comparison was attempted on "$(TargetPlatformMinVersion)" that evaluates to "" instead of a number, in condition "'$(TargetPlatformMinVersion)' >= '10.0.17200.0'". [D:\buildtrees\libgpg-error\x64-uwp-rel\error-1.42-2324ddbc71.clean\SMP\libgpg-error_winrt.vcxproj]
```
This looks like a VS2022 regression.
REGRESSION: chromium-base:x64-osx
This port is still having the strange behavior where building it once fails but building it again succeeds O_O
REGRESSION: libmikmod:x64-osx
This is emitting a bunch of spew about standard library symbols.
REGRESSION: qtapplicationmanager:x64-windows
Should be fixed by https://github.com/microsoft/vcpkg/pull/23638
* Update to 3.4.2
* Disable ATL for uwp and for arm64
* Fix linking on uwp
* Fix pc file fixup
* Update versions
* Declare dependency on atlmfc.
Co-authored-by: Billy Robert O'Neal III <bion@microsoft.com>
* Update Linux VM CUDA to 11.6.
* Update Intel OneAPI
* Update Visual Studio to 2022, and remove deployment of v140 and v141 compilers.
* Update CUDA to 11.6 on Windows.
* Update pools.
* cix ci.baseline.txt
Co-authored-by: nicole mazzuca <mazzucan@outlook.com>
* Remove pixel from ci baseline
* don't build examples
* [pixel] update to master
* copy license from portfile.cmake
* version
* license
* version
* fix typo
* version
* use version-date
* version
* change version date to today
* version
* disable crt warnings on uwp
* version
* use vcpkg provided libs before system frameworks on macOS, remove now unnecessary mod to giflib
* fix references
* fixes from review
* apply fixes from review - remove warnings too!
* [giflib] add license
* fix references
Co-authored-by: Victor Romero <viromer@microsoft.com>
* add vcpkg-tool-lessmsi
* add vcpkg-tool-python2
* adjust vcpkg_find_acquire_program
* use python2 dependency in qt5-base
* remove python2 from qt5-base
* add python2 to qt5-webengine
* use python3 instead of python2
* add some deps on the new vcpkg-tool-python2
* Apply suggestions from code review
Co-authored-by: nicole mazzuca <83086508+strega-nil-ms@users.noreply.github.com>
* add license field
* fix test detection
* error on failure to find python2
* move logs to correct position.
* add dep to qtwebengine
* adjust pybind cmake config.
* fix shiva
* move python2 to manual-tools
adjust search path
generate a details.cmake for vcpkg_find_acquire_program
* format manifest
* use version-date
* version stuff
* replace version-string
* more version stuff
* fix search path and add supports expression
* version stuff
* shiva version
* remove arm windows from support
* reuse output_path
use lessmsi is a host dep so use the correct path
* version stuff
* Fix x86 regression
* version stuff
* try fixing qtinterfaceframework
* version stuff
Co-authored-by: Alexander Neumann <you@example.com>
Co-authored-by: nicole mazzuca <83086508+strega-nil-ms@users.noreply.github.com>
Co-authored-by: Victor Romero <viromer@microsoft.com>
Co-authored-by: Victor Romero <romerosanchezv@gmail.com>
* [python3] Update to Python 3.10.2.
* Version nonsense
* Update Python tool to 3.10.2.
* [python3] Update version file
* [python3] Update versions again
Co-authored-by: Victor Romero <viromer@microsoft.com>
* Update vcpkg-tool to 2022-03-08.
* Back off required git version to something actually in the build lab.
* Remove hiding of the output of --parent-hashes.
* Update 7zip reference as indicated in https://github.com/microsoft/vcpkg-tool/pull/413
* Add triplet to name of main CI task
* Fix osx bootstrap task name
* Hide output from switch to parent hash
* Bump test port version
* Hardcode osx+linux triplet names
* Improve logging order
* Add error handling to each tool download
* Revert "Bump test port version"
* Don't prefetch 7zip
* Use host tools by default.
* use GLOB_RECURSE and filter the result
* be positiv ;)
* give me debug output
* Revert "give me debug output"
This reverts commit 17737bceda8ed0dbe1a60c2fd831a55e53707dcd.
* remove unnecessary if(IS_DIRECTORY)
Co-authored-by: Alexander Neumann <you@example.com>
* update cmake to 3.22.0-rc3 to verify catalogue compatibility and be ready for final release
* update to cmake 3.22 final release
* [vcpkgTools] update to cmake 3.22.1
* update test port
* update cmake to 3.22.2
and git for windows to 2.35.1.2
* fix missing uuid on osx in ignition-transport.
* modernize
* fix hash
* fix qtbase regression
* try fixing the pc file search
* hidaoi give trace logs
* give me trace output
* bit of pkgconf changes. Need system search paths for pkgconf to work correctly.
* fix underscore
* remove extra x
* revert port changes
* add license to cmake test port
* fix arm regression due to not finding taglib.
* version stuff
Co-authored-by: Stefano Sinigardi <stesinigardi@hotmail.com>
Co-authored-by: Billy Robert O'Neal III <bion@microsoft.com>
Co-authored-by: JackBoosY <yuzaiyang@beyondsoft.com>
Co-authored-by: Alexander Neumann <you@example.com>
* [get_cmake_vars] add -isysroot flag independednt of crosscompiling
Even without crosscompiling it is possible to specify a VCPKG_OSX_SYSROOT or
a SDKROOT this should be popagated to autotools and meson builds as well.
* [meson] add -arch and -isysroot to linker flags in case of osx crosscompile
* [meson] avoid DEPRECATION warning about flags in the [properties] section
* [meson] don't set OSXROOT evironment varibale
* [meson] don't set MACOSX_DEPLOYMENT_TARGET environment variable
cmake appends the -mmacosx-version-min flag c/cxx flags, overriding the
value set via VCPKG_C_FLAGS. By setting the environment variable, the
VCPKG_C_FLAGS value was used for meson builds. Now the same value is
taken for both.
* [meson/make] Add CMAKE_OSX_DEPLOYMENT_TARGET to VCPKG_DETECTED_CMAKE_C_FLAGS
This is done in the same way in CMake internally
* Merge INTERFACE_LINK_LIBRARIES configurations
Direct references to binaries must match the active build type when used.
* Update tmx for cmake_config_fixup
* Update versions
* Prevent repeated fixup of the same file
* Don't change line structure
* Update versions
* CI
* Apply CR suggestion
Co-authored-by: nicole mazzuca <mazzucan@outlook.com>
* Update versions
* Fixup CR suggestion
* Update versions
* Update versions
Co-authored-by: nicole mazzuca <mazzucan@outlook.com>
* Major update to port poppler
* Fixup usage requirements via main pc file
* Add usage based on FindPkgConfig
* Export unofficial cmake config
* uwp is unsupported
* [gdal] Add poppler feature
* [gdal] Update poppler dependency
* [gdal] Add patch for poppler C++17 API
* Disable ENABLE_RELOCATABLE
* Update to 22.02.0 from poppler gitlab
* Update versions
* [skip actions] CI with gdal tools
* Revert "[skip actions] CI with gdal tools"
* Modify feature name for private API
* Update versions
* Add license field to gdal manifest
* Add port name to unofficial namespace
* Rectify poppler version number
* Update versions
* poppler[fontconfig] doesn't build for MSVC
* Update versions
Co-authored-by: Matthias Kuhn <matthias@opengis.ch>
* Change minimum CMake version to 3.7.2.
This moves our oldest supported Linux back up to today minus 5 years, Debian 9 from 2017 comes with 3.7.2.
* Enable more policies.
* [fbthrift, fizz, folly] Fix dependency resolution.
Found by nightly build https://dev.azure.com/vcpkg/public/_build/results?buildId=67402
* wangle was fixed by a compiler update on a patch tuesday.
* this unblocked attempting fbthrift on x64-windows hosts, but it misses a header; added patch already merged upstream adding missing algorithm <include>
* fixed folly attempt to insert the right debug prefix which was broken by vcpkg_fixup_cmake_targets no longer generating _IMPORT_PREFIX.
* disable dynamic build of fbthrift on Windows because it tries to create DLLs with no import libs (because it makes no exports)
* Format and fix version database.
* Back out fizz patches.
* Block dynamic flavors. (The build in question builds an exe and expects all its dependencies to be linked statically and patching the build system to copy over the right DLLs seems out of scope)
* Add libevent patch.
* Also fixup mvfst and add missing CONFIG to patch.
* Remove unused patch.
* Fix version sha
* Some of the Windows runs have started to exceed 24 hours now as well as the MacOS ones as we have added more ports.
* We are no longer trying to constantly "psychic debug" why ports are failing; CI passes somewhat regularly now, and the failures we get seem to be much more deterministic.
* We're still budget constrained on Azure resources (and have been a bit over budget)
To that end, we're going back off nightly CI runs to Monday Wednesday Friday (like osx already is), so we no longer need separate bits for them.
