| Subject: | tests fail when run in parallel |
Looks like the tests are stomping on each other. They should their own temp
directory, e.g. from File::Temp::tempdir, instead of using the same one, so
tests do not fail when run under HARNESS_OPTIONS=j9 --
t/03_bad_dependency.t ... ok
t/09_regression.t ....... ok
# Failed test 'Found test_one.t'
# at t/06_multifile.t line 59.
t/01_compile.t .......... ok
t/04_file_handler.t ..... ok
t/07_add_directory.t .... ok
# Failed test 'Found test_three.t'
# at t/06_multifile.t line 65.
# Failed test 'Found test_four.t'
# at t/06_multifile.t line 67.
# Failed test 'Found test_four.t'
# at t/06_multifile.t line 68.
# Failed test 'Found manifest file'
# at t/06_multifile.t line 73.
t/08_add_all.t .......... ok
# Failed test 'Found manifest file'
# at t/06_multifile.t line 74.
# Looks like you failed 6 tests of 23.
t/06_multifile.t ........
Dubious, test returned 6 (wstat 1536, 0x600)
Failed 6/23 subtests
t/05_check_count.t ...... ok
t/02_main.t ............. ok
t/10_legacy_extract.t ... ok
t/11_legacy_script.t .... ok
t/12_content_handler.t .. ok
t/13_api.t .............. ok