On 9 December 2012 17:04, Guillaume Rousse <[email protected]> wrote: >> Yours does nothing: >> t/pod-coverage.t ................................................ >> skipped: Author test, set $ENV{TEST_AUTHOR} to a true value to run >> t/pod-spelling.t ................................................ >> skipped: Author test, set $ENV{TEST_AUTHOR} to a true value to run >> t/pod-syntax.t .................................................. >> skipped: Author test, set $ENV{TEST_AUTHOR} to a true value to run > > They work if you set the variable, as explicitely reported. That's perfectly > useless to run those tests anywhere else as on developper machine, and > especially during package build. And that's a standard practice in perl > community.
No, I want pod errors to be found prior submitting them. And I don't want to manually run several tests. Before all I had to do was to run "make test" and it reported me any errors before commiting. Now it doesn't anymore. > And your claim are quite unfair, given than the previous test didn't > prevented trivial syntax errors to be commited. Yes it did. I provided you the "make test" log proving it. Yours are not run anymore when running "make test". What I did fail to mention is that they don't even work with that varible: [root@localhost t]# TEST_AUTHOR=test perl pod-coverage.t You said to run 0 tests at pod-coverage.t line 21. You don't even have tried them, did you? > Feel free to revert those changes, this kind of discussion isn't worth the > pain. Well, if that mean less quality testsuite and if you refuse to answer reviewing, I'll eventually revert those or at least put back the old working pod test.
