Subject: | Tests sometimes fail with perl-5.18.1-3.mga4 |
Hi J. J.,
the tests sometimes (and quite often) fail on perl-5.18.1-3.mga4 on Mageia Linux 4/Cauldron x86-64. I suspect it is caused by the hash keys order randomisation in perl-5.18.x:
shlomif@telaviv1:~/Algorithm-MasterMind-v0.4.5$ make test
PERL_DL_NONLAZY=1 /usr/bin/perl5.18.1 "-MExtUtils::Command::MM" "-MTest::Harness" "-e" "undef *Test::Harness::Switches; test_harness(0, 'blib/lib', 'blib/arch')" t/*.t
t/00.load.t ............. 1/1 # Testing Algorithm::MasterMind v0.4.5
t/00.load.t ............. ok
t/00_functions.t ........ 1/? # Testing Algorithm::Mastermind v0.4.5, Perl 5.018001, /usr/bin/perl5.18.1
t/00_functions.t ........ ok
t/00_sequential.t ....... 1/? # This might take a while while it finds the code FAFA
t/00_sequential.t ....... ok
t/01_random.t ........... 1/? # This might take a while while it finds the code EAFC
t/01_random.t ........... ok
t/02_eda.t .............. 1/? # This might take a while while it finds the code EAFC
# This might take a while while it finds the code EAFC
# This might take a while while it finds the code EAFC
# This might take a while while it finds the code EAFC
t/02_eda.t .............. ok
t/03_sequential_alt.t ... 1/? # This might take a while while it finds the code ADCB
t/03_sequential_alt.t ... ok
t/04_consistent_set.t ... 1/?
# Failed test 'Consistent creation'
# at t/04_consistent_set.t line 48.
# Structures begin differing at:
# $got->[0]{_alphabet}[0] = 'B'
# $expected->[0]{_alphabet}[0] = 'C'
# Looks like you failed 1 test of 142.
t/04_consistent_set.t ... Dubious, test returned 1 (wstat 256, 0x100)
Failed 1/142 subtests
t/04_partitions.t ....... # This could take a while
t/04_partitions.t ....... ok
t/05_partition_most.t ... 1/? # This might take a while while it finds the code ADCB
^CMakefile:973: recipe for target 'test_dynamic' failed
make: *** [test_dynamic] Interrupt
shlomif@telaviv1:~/Algorithm-MasterMind-v0.4.5$
grepping for "keys" in Algorithm-MasterMind found many cases where they are used without being sorted. Please look into fixing it.
Regards,
-- Shlomi Fish