On Mon, Jul 13, 2009 at 4:17 PM, David Golden <xda...@gmail.com> wrote:
> On Mon, Jul 13, 2009 at 1:53 PM, Oliver Paukstadt <c...@sourcentral.org>wrote: > >> Maybe David from CPAN::Reporter is able to explain what happens ;-) >> Excerpt from logfile attached: >> >> The system recognizes it is a duplicate "NA" after the perl Makefile.PL >> does not create any output. Nevertheless it tries to start make test >> after that and FAILs which results in the FAIL reports. >> > > I can replicate it on my perl 5.6.2. > > I think it might be a CPAN bug -- it shouldn't try to test if configuration > failed. But it tries and then fails and CPAN::Reporter sees that as a FAIL. > > When I get a chance, I'll poke around and see if I can figure it out. > I take it back. I can't replicate it. CPAN.pm is correctly refusing to run "make test". Oliver -- (a) are you able to replicate it? (b) are you reusing build directories? (c) Are you testing with multiple copies of perl? Otherwise, I'm stumped, though I note that DBIx::Class uses autoinstall and I always get suspicious when autoinstall is involved. -- David
