This is a report on differences between the output of 'make test' and 'prove' which are, well, different from those reported by Stevan Little in a thread beginning on Sept 4.
While teaching myself how to use 'prove' tonight, I decided to try it out on a Perl core module which I have also been using as a test case for my understanding of Devel::Cover. The core module in question is Time::Local, version 1.10. When I called 'make test' on this module, I got the following results: [Time-Local-1.10 562]$ make test PERL_DL_NONLAZY=1 /usr/local/bin/perl "-MExtUtils::Command::MM" "-e" "test_harness(0, 'blib/lib', 'blib/arch')" t/*.t t/Local....ok All tests successful. Running this with TEST_VERBOSE=1 indicated that the number of tests run was 102. However, when I called 'prove t', I got different results: [Time-Local-1.10 563]$ prove t t/Local....Cannot handle date (7, 14, 3, 19, 0, 2038) at t/Local.t line 105 t/Local....dubious Test returned status 255 (wstat 65280, 0xff00) DIED. FAILED tests 101-102 Failed 2/102 tests, 98.04% okay Failed Test Stat Wstat Total Fail Failed List of Failed ------------------------------------------------------------------------------- t/Local.t 255 65280 102 4 3.92% 101-102 Failed 1/1 test scripts, 0.00% okay. 2/102 subtests failed, 98.04% okay. As best as I can tell, these are the tests in t/Local.t which failed: ok(sprintf('%x', timegm(gmtime(0x7fffffff))), sprintf('%x', 0x7fffffff), # line 105 '0x7fffffff round trip through gmtime then timegm'); ok(sprintf('%x', timelocal(localtime(0x7fffffff))), sprintf('%x', 0x7fffffff), '0x7fffffff round trip through localtime then timelocal'); Any ideas? Jim Keenan ===== Affiliations: Perl Seminar NY / New York Perlmongers / Toronto Perlmongers __________________________________ Do you Yahoo!? Check out the new Yahoo! Front Page. www.yahoo.com