|Think about Loose Coupling|
cpan test redux...better, but...by perl-diddler (Hermit)
|on Nov 10, 2013 at 21:08 UTC||Need Help??|
perl-diddler has asked for the
wisdom of the Perl Monks concerning the following question:
The latest test reports on CPAN show up at
In there, only 3 people had failures:
Except for the 5.8 version of perl, which might have some
Specifically, these lines that seemingly conflict:
This was also a problem in 1.1.1, but there it was near 100%, across the board.
I also had a "use_ok(P)" and use_ok(mem) at the beginning
So now I'm getting 'mem' being both present and non-present
I did contact the tester with all the fails and asked if
Anyway -- just thought I'd throw this out, since this
They said it looked like the program under test (being called from the test program), did not have the correct PERL5LIB set in the environment. It seems test programs are expected to copy %INC to PERL5LIB? That seems an odd requirement for a test script???
I.e. a Test harness driven by some script, that calls or invokes program under test with various inputs, and checks for correct output. At least that's how I'm used to testing, since you don't want your scripts to be in the same program as the program under test so the scripts won't be accidently mangled by the program under test. Is that not correct procedure? *wimper*?