hi Christian Lohmaier, if I understand this correctly, then we can actually remove the pkgconfig from the cygwin dependencies, because it cannot be used. I have added the pkgconfig download in the hard way, look https://wiki.documentfoundation.org/Development/BuildingOnWindows Thanks for the explanation now I know what it is needed for.
> Then you might also have used the wsl-as-helper approach as opposed to > using cygwin. I wanted to use wsl but have (host) Win10 there Hyper-V in that it runs with (guest) Win11 the buildsystem. But because it needs the nested (in hyper-v) virtualization the host must also be Win11 > Junit 4.10 should still include hamcrest, so if you're already using > the separate hamcrest you could/should also be using newer junit > probably... am now a little confused should be using of junit 4.13 with hamcrest 2.2? I can build and it takes 30 min for the complete build > MSYSTEM is only set in git-bash (i.e. in the wsl-as-helper method), > not in the default cygwin terminal, so unless you manually set > gb_COLOR to true (something non-empty), it should not use colored > output to begin with. In cygwin i never using the cygwin-git, it is faster with Git for Windows, but the bulid is in cygwin and the confuse output it is the same without color but the autogen.sh output is correct. Juergen ________________________________________ Von: Christian Lohmaier <lohma...@googlemail.com> Gesendet: Mittwoch, 2. Juli 2025 13:07 An: Juergen Funk Cc: libreoffice@lists.freedesktop.org Betreff: Re: Trouble with new lo-build Hi Jürgen, *, On Mon, Jun 30, 2025 at 11:14 PM Juergen Funk <j-f...@outlook.de> wrote: > > i have install Win11 (in Hyper-V), Then you might also have used the wsl-as-helper approach as opposed to using cygwin. > for install the LibreOffice build i have use this docu > https://wiki.documentfoundation.org/Development/BuildingOnWindows > and use the hard way (not with LODE). > my autogen.input is > --with-ant-home=/cygdrive/d/src/lo/sources/java/apache-ant-1.9.5 > --with-junit=/cygdrive/d/src/lo/sources/java/junit-4.10.jar > --with-hamcrest=/cygdrive/d/src/lo/sources/java//hamcrest-2.2.jar > […] Junit 4.10 should still include hamcrest, so if you're already using the separate hamcrest you could/should also be using newer junit probably... > In this case i get a lot of errors: > > checking for pkgconf-2.4.3.exe... no > configure: error: > A windows version of pkgconf is required to build harfbuzz. > Add PKG_CONFIG=/path/to/pkgconf-2.4.3.exe to autogen.input or put > it in PATH > Error running configure at ./autogen.sh line 333. > > and Well, if configure is failing, the build is known to not work, so any further errors are expected. > [meson configure of harfbuzz] > Found pkg-config: YES (D:/dev/cygwin64/bin/pkgconf-2.4.3.exe) 2.4.3 > Found CMake: C:\Program Files\Microsoft Visual > Studio\2022\Community\Common7\IDE\CommonExtensions\Microsoft\CMake\CMake\bin\cmake.EXE > (3.31.6) > Run-time dependency graphite2 found: NO (tried pkgconfig and cmake) > > meson.build:147:16: ERROR: Dependency "graphite2" not found, tried pkgconfig > and cmake That looks like you were trying to use a cygwin/unix version of pkgconf. That won't work. configure does look for the versioned executable to prevent accidental mixups with cygwin-provided pkgconf or pkg-config. > After that I have compare the hard way with LODE and found a lot of diffs. > I have adapted, testing and improve the docu of > https://wiki.documentfoundation.org/Development/BuildingOnWindows Well, wouldn't call that "lot of diffs" and furthermore: Lode's cygwin install script is not authoritative due to the chicken-and-egg issue. You first have to install cygwin to clone. Also by looking at the changes it is clear that there is personal bias by including emacs and nano for example. And while I don't have hard feelings on adding those, I have a problem with removing vim from the list in return. All CI bots have been set up by using the cygwin package list from the wiki, not from the file in lode. The changes in the lode package list are absolutely unrelated to you ignoring configure's warning (or "fixing" the warning the wrong way). "A windows version of pkgconf is required", so using a cygwin version won't do. > But I have not download/install the meson 1.7.2, You don't need to provide meson anymore. Too few people read configure's output/explicitly disrespected it so meson is now provided as an external, downloaded and extracted during the build. > above have get the error in meason, after my adaption is all fine, for what > need meson? Well obviously it was not fine. it requires a windows version so it can use windows-style paths. meson might claim to be working nicely in cross-compilation szenarios, but that only holds true in very small preconditions. It is just a lot easier to please meson by using an appropriate pkgconf than to patch fundamentals of meson to work within cygwin for building for windows. > A other question or hint (only a wrong output) > > When start ./autogen.sh the ouptut is okay and right > but > when start make the output is mixed up (with color or without): > […] That didn't make it across properly in the mail, but also looks like an explicit override or not using the default cygwin terminal: solenv/gbuild/Output.mk: # default to color output, if interactive ifeq ($(origin gb_COLOR),undefined) ifneq ($(MAKE_TERMOUT),) # Cygwin mintty has issues where gb_Output_error is swallowed # but git-bash/wsl-as-helper seems fine with it ifneq ($(OS)$(MSYSTEM),WNT) gb_COLOR=$(true) endif endif endif MSYSTEM is only set in git-bash (i.e. in the wsl-as-helper method), not in the default cygwin terminal, so unless you manually set gb_COLOR to true (something non-empty), it should not use colored output to begin with. ciao Christian