From: Peter Rabbitson <ribasushi@cpan.org>
Date: Sun, 18 Nov 2012 20:17:09 +0000 (+0100)
Subject: Back out frivolous loading of threads.pm
X-Git-Url: http://git.shadowcat.co.uk/gitweb/gitweb.cgi?a=commitdiff_plain;h=dfd722efa2fb42fce57a64e0bcc818a900e85a92;p=dbsrgits%2FDBIx-Class-Historic.git

Back out frivolous loading of threads.pm

Not only was this an unsuccesful attempt to mask out a bug introduced
in Moo's inlining of Devel::GlobalDestruction (since fixed), but it
broke a test under older win32 perls.
Revert 8bab2062, bump Moo dep
---

diff --git a/Changes b/Changes
index 88feed4..a5c8fe7 100644
--- a/Changes
+++ b/Changes
@@ -3,6 +3,7 @@ Revision history for DBIx::Class
     * Fixes
         - More robust behavior under heavily threaded environments - make
           sure we do not have refaddr reuse in the global storage registry
+        - Fix failing test on 5.8 under Win32 (RT#81114)
 
 0.08204 2012-11-08
     * New Features / Changes
diff --git a/Makefile.PL b/Makefile.PL
index 87f3a40..6e1ed0d 100644
--- a/Makefile.PL
+++ b/Makefile.PL
@@ -42,9 +42,6 @@ my $runtime_requires = {
   # FIXME - temporary, needs throwing out for something more efficient
   'Data::Compare'            => '1.22',
 
-  # Moo does not yet depend on this higher version
-  'strictures'               => '1.003001',
-
   # DBI itself should be capable of installation and execution in pure-perl
   # mode. However it has never been tested yet, so consider XS for the time
   # being
@@ -64,7 +61,7 @@ my $runtime_requires = {
   'Data::Page'               => '2.00',
   'Devel::GlobalDestruction' => '0',
   'Hash::Merge'              => '0.12',
-  'Moo'                      => '0.009100',
+  'Moo'                      => '1.000006',
   'MRO::Compat'              => '0.09',
   'Module::Find'             => '0.06',
   'namespace::clean'         => '0.20',
diff --git a/TODO_SHORTTERM b/TODO_SHORTTERM
index ca4fc83..d7247f7 100644
--- a/TODO_SHORTTERM
+++ b/TODO_SHORTTERM
@@ -1,6 +1,3 @@
-* 8bab2062 preloading threads.pm is absolute bullshit - remove and find out
-the real problem (might very well be a Moo bug). But seriously - preloading
-threads.pm in a BEGIN block?!?!?!
 * 50e6c52e reenable the test and investigate. A blatant skip is fucking
 bullshit
 * a0361822 reenable the win32 test - if it no longer fails for anyone - how
diff --git a/t/storage/txn.t b/t/storage/txn.t
index 41df4d5..09260f0 100644
--- a/t/storage/txn.t
+++ b/t/storage/txn.t
@@ -1,10 +1,6 @@
 use strict;
 use warnings;
 
-BEGIN {
-  require threads if $^O eq 'MSWin32'; # preload due to fork errors
-}
-
 use Test::More;
 use Test::Warn;
 use Test::Exception;