in reply to Re: WWW-Mechanize-Firefox install issue
in thread WWW-Mechanize-Firefox install issue

> go ahead and skip those tests

This isn't a good piece of advice. Without knowing the reason of the failure, this can break modules that worked previously. Moreover, newbies usually don't backup their PERL5LIB before the installation, so there might be no easy way back.

($q=q:Sq=~/;[c](.)(.)/;chr(-||-|5+lengthSq)`"S|oS2"`map{chr |+ord }map{substrSq`S_+|`|}3E|-|`7**2-3:)=~y+S|`+$1,++print+eval$q,q,a,

Replies are listed 'Best First'.
Re^3: WWW-Mechanize-Firefox install issue
by Anonymous Monk on Feb 24, 2017 at 08:00 UTC

    This isn't a good piece of advice. Without knowing the reason of the failure, this can break modules that worked previously. Moreover, newbies usually don't backup their PERL5LIB before the installation, so there might be no easy way back.

    Prove it? Disengage pedantic autopilot and look up, those aren't trees, thats a forest!

    Newbies aren't aware you can install modules with failing tests, or skip testing altogether -- an option that is always available to test

    Skip the tests or force the install and install the modules today now with no waiting

    Then see if that breaks anything in your existing codebase

    See if firefox mechanize will work now

    IPC::Run is pure-perl, easy to install, uninstall, reinstall, revert ... backup not required

    Look at the failing test, the single solitary failing test ... trying to close an already destroyed filehandle ... thats nothing to worry about ... and certainly nothing for any kind of non author to fuss about

    Also look at the documentation, esp the caveats for windows

    Bad advice? No thats just uninformed criticism

      > uninformed criticism

      I've spent days fixing our users' setup after they force-installed a library, they often didn't know what they were doing, so they installed newer versions of libraries into their system Perl and broke everything. Finding the reason why the test failed is often much cheaper than that, but YMMV.

      ($q=q:Sq=~/;[c](.)(.)/;chr(-||-|5+lengthSq)`"S|oS2"`map{chr |+ord }map{substrSq`S_+|`|}3E|-|`7**2-3:)=~y+S|`+$1,++print+eval$q,q,a,

        > uninformed criticism

        I've spent days fixing our users' setup after they force-installed a library, they often didn't know what they were doing, so they installed newer versions of libraries into their system Perl and broke everything. Finding the reason why the test failed is often much cheaper than that, but YMMV.

        ++:)++ thats some nice truth from the trenches (truism?) ... not relevant to IPC-Run-0.94/t/run.t

        See Ignoring test failures is stupid, most minor test failures are failures of the test suite not of the module