cpan-testers-discuss  

Re: PM post on Makefile.PL FAIL for missing libraries

David Golden
Thu, 27 Sep 2007 03:10:44 -0700

On 9/27/07, Barbie <[EMAIL PROTECTED]> wrote:
> On Thu, Sep 27, 2007 at 05:26:17PM +1000, Sisyphus wrote:
> >
> > The big question for the author of the PM post (who I happen to know quite
> > well) is "Having had the 'Makefile.PL' process establish that the build
> > environment is inadequate, how does one then have the process terminate in
> > such a way that an NA is reported ?"
>
> Just to be clear, it shouldn't be reported at all. It's not an NA report
> as with the right librar(y|ies) it would work with that Perl/platform.
>
> The problem is alerting the testing system in a way that it knows that
> the reason the testing has been stopped is because of a missing
> prerequisite. At the moment we don't have that.

At least for CPAN -- I think this is as simple as just *exiting* from
a Makefile.PL with an exit value of 0.  CPAN::Reporter considers that
"success" and does not issue a FAIL report, but then CPAN.pm notices
that no Makefile was created and refuses to continue into make or test
phases.

I just released CPAN-Test-Dummy-Perl5-Make-PLExit to CPAN to confirm
this behavior.  if CPANPLUS has similar behavior to
CPAN/CPAN::Reporter, then when the smokers get it, I shouldn't see any
test reports for it at all.

Regards,
David