Front page | perl.perl5.porters |
Postings from July 2009
Dear helpful people, On Thu, Jul 16, 2009 at 02:00:17PM -0500, David Favor wrote: > From latest Perl::Critic... > > t/01_config.t ................................. 216/217 > # Failed test 'vague -single-policy' > # at t/01_config.t line 498. > # 'Bizarre copy of HASH at > /root/.cpanplus/5.10.1/build/Perl-Critic-1.098/blib/lib/Perl/Critic/Config.pm line 366. > # ' > # doesn't match '(?msx-i:matched [ ] multiple [ ] policies)' > # Looks like you failed 1 test of 217. > t/01_config.t ................................. Dubious, test returned 1 > (wstat 256, 0x100) > Failed 1/217 subtests > t/01_config_bad_perlcriticrc.t ................ ok > > This started with the last few levels of 5.10.0 and persists > in 5.10.1 as of the GitLive-maint-5.10-1589-gf10445d level. This is a regression. Unaddressed it will block the release of 5.10.1 Would anyone like to help David, Dave and 5.10.1 by using git bisect to find the change that caused this failure? There's instructions on how to run git bisect on perl in perlrepository.pod: http://perl5.git.perl.org/perl.git/blob/HEAD:/pod/perlrepository.pod#l534 I suspect for this case it will require the test perl to be installed, and Perl::Critic to be build from it, so you'd need to add code to automate that to the sample script given in perlrepository.pod Doing this would be very helpful as it spreads the workload around, getting more bugs found and fixed in parallel, reducing the lead time to 5.10.1-RC1 Nicholas ClarkThread Previous | Thread Next