At least on my system, when I test the built cygport through cygport itself, I get lots of test fails due to:
/mnt/share/packages/cygport.noarch/src/rpm-style/lib/src_fetch.cygpart: line 153: git_fetch: command not found When I run the tests manually, that doesn't happen. I've verified this error occurs because I build cygport from git, so that function is already defined by the outermost cygport (and declared unmutable). It seems that the test driver should start a fresh cygport instance to decouple the two shell environments, but I didn't really get why the environment bleeds into the test in this undesirable way to start with. The rest of the test fail s is either me not having installed some stuff that would be needed or file names that differ only in case, like so: --8<---------------cut here---------------start------------->8--- @@ -2,5 +2,5 @@ /usr/i686-w64-mingw32/sys-root/mingw/include/crypt.h /usr/i686-w64-mingw32/sys-root/mingw/lib/libcrypt.a /usr/i686-w64-mingw32/sys-root/mingw/lib/libcrypt.dll.a -/usr/share/doc/mingw64-i686-crypt/COPYRIGHT +/usr/share/doc/mingw64-i686-crypt/Copyright /usr/share/doc/mingw64-i686-crypt/crypt.README --8<---------------cut here---------------end--------------->8--- It seems that the expected outputs need updating or else the comparison should be done case-insensitively by default? The test-driver script already has provisions for that, so these fails vanish if I export CASE_INSENSITIVE=1 in the cygport file, but I guess something more targeted would be better. Regards, Achim. -- +<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+ Wavetables for the Waldorf Blofeld: http://Synth.Stromeko.net/Downloads.html#BlofeldUserWavetables
