Dave Rolsky [Sat, 6 Jun 2009 19:47:52 +0000]
Make this a TODO test since it's easily worked around. Load Moose.pm before loading Moose::Meta::Class.
Dave Rolsky [Sat, 6 Jun 2009 19:43:20 +0000]
Formatting
Dave Rolsky [Sat, 6 Jun 2009 19:42:50 +0000]
Formatting
Dave Rolsky [Sat, 6 Jun 2009 19:41:44 +0000]
rename this test
Dave Rolsky [Sat, 6 Jun 2009 19:38:37 +0000]
Minor formatting tweaks in Changes
Dave Rolsky [Sat, 6 Jun 2009 19:37:24 +0000]
bump version to 0.80
Dave Rolsky [Sat, 6 Jun 2009 19:37:10 +0000]
Add release date
Shawn M Moore [Thu, 4 Jun 2009 06:28:09 +0000]
Attribute the change to alias
Shawn M Moore [Thu, 4 Jun 2009 06:26:34 +0000]
We do have doc now!
Shawn M Moore [Thu, 4 Jun 2009 05:15:48 +0000]
Improve the error message of a role using "extends"
Dave Rolsky [Wed, 3 Jun 2009 19:42:23 +0000]
Wording and formatting tweaks for Changes
Jesse Luehrs [Tue, 2 Jun 2009 23:41:10 +0000]
metarole is past the scary experimental stage, i think
Shawn M Moore [Tue, 2 Jun 2009 04:06:26 +0000]
Excluding a conflicting method is a certainly valid resolution (Ovid++)
Guillermo Roditi [Mon, 1 Jun 2009 21:29:26 +0000]
changelog
Guillermo Roditi [Mon, 1 Jun 2009 21:25:34 +0000]
Add test for infinite recursion when creating an anon class with Moose::Object
as the superclass
Shawn M Moore [Sun, 31 May 2009 05:03:01 +0000]
Save the curses for comments O:)
Shawn M Moore [Sun, 31 May 2009 04:56:01 +0000]
Moose::Manual::Delta isn't for our internal test changes, silly yuval
Shawn M Moore [Sun, 31 May 2009 04:53:36 +0000]
Changes entry for today's branch :)
Shawn M Moore [Sun, 31 May 2009 02:57:10 +0000]
Use class/type information instead of a useless is_conflict method
Shawn M Moore [Sun, 31 May 2009 02:53:32 +0000]
More doc for Method::Conflicting
Shawn M Moore [Sun, 31 May 2009 02:49:31 +0000]
Merge branch 'master' into role-methods
Conflicts:
lib/Moose/Meta/Role.pm
Shawn M Moore [Sun, 31 May 2009 02:42:25 +0000]
Make roles required
Shawn M Moore [Sun, 31 May 2009 02:37:22 +0000]
Add Ovid's test case
Shawn M Moore [Sun, 31 May 2009 02:27:39 +0000]
Move Method::Conflicted to Conflicting
Shawn M Moore [Sun, 31 May 2009 02:27:15 +0000]
Begin s/conflicted/conflicting/
Shawn M Moore [Sun, 31 May 2009 02:19:54 +0000]
Make Conflicted immutable
Shawn M Moore [Sun, 31 May 2009 02:13:53 +0000]
Much better error message for role conflicts
Shawn M Moore [Sun, 31 May 2009 02:06:22 +0000]
Use role names for consistency with ToRole
Shawn M Moore [Sun, 31 May 2009 02:03:34 +0000]
Include roles in the role summation add_conflicted_method
Shawn M Moore [Sun, 31 May 2009 01:59:02 +0000]
Basic conflict message
Shawn M Moore [Sat, 30 May 2009 22:43:14 +0000]
Method::Required->name is required
Shawn M Moore [Sat, 30 May 2009 22:43:01 +0000]
Role::Method->is_conflict
Shawn M Moore [Sat, 30 May 2009 22:41:45 +0000]
Use add_conflicted_method instead of add_required_methods where
appropriate
Shawn M Moore [Sat, 30 May 2009 22:41:29 +0000]
Role->add_conflicted_method
Shawn M Moore [Sat, 30 May 2009 22:27:44 +0000]
Add conflicted method metaclass
Shawn M Moore [Sat, 30 May 2009 22:08:42 +0000]
Use required method objects in ToClass, not names
Shawn M Moore [Sat, 30 May 2009 22:05:12 +0000]
Use required_method objects, not names, in ToRole
Shawn M Moore [Sat, 30 May 2009 22:04:17 +0000]
Use objects, not names, in RoleSummation
Shawn M Moore [Sat, 30 May 2009 22:00:34 +0000]
Fix role delegation
Shawn M Moore [Sat, 30 May 2009 22:00:22 +0000]
get_required_method_list now returns objects not names
Shawn M Moore [Sat, 30 May 2009 19:51:58 +0000]
Add get_values
Shawn M Moore [Sat, 30 May 2009 19:49:59 +0000]
Rename internal get_list to get_keys for clarity
Shawn M Moore [Sat, 30 May 2009 19:41:20 +0000]
Grammar fix
Shawn M Moore [Sat, 30 May 2009 19:34:34 +0000]
required_methods is now a hashref of Method::Requireds not Strs
Shawn M Moore [Sat, 30 May 2009 19:34:19 +0000]
Typo fix
Shawn M Moore [Sat, 30 May 2009 19:23:10 +0000]
No need to explicitly uniq, since assigning to a hash does that already
Shawn M Moore [Sat, 30 May 2009 19:19:05 +0000]
Use blessed not ref
Shawn M Moore [Sat, 30 May 2009 19:06:16 +0000]
Have add_required_methods inflate the required methods
Shawn M Moore [Sat, 30 May 2009 19:06:07 +0000]
Documentation
Shawn M Moore [Sat, 30 May 2009 19:05:34 +0000]
We need to say new -> _new? what's the better way of doing this?
Shawn M Moore [Sat, 30 May 2009 18:51:07 +0000]
Add required_method_metaclass attribute
Shawn M Moore [Sat, 30 May 2009 18:46:47 +0000]
Make Role::Method::Required its own entity, since it is not actually a
subclass of Role::Method
Jesse Luehrs [Sat, 30 May 2009 18:44:52 +0000]
grammar fix
Shawn M Moore [Sat, 30 May 2009 18:25:30 +0000]
Mark possessives with an apostrophe!
Shawn M Moore [Sat, 30 May 2009 18:24:38 +0000]
Fix missing ) in pod
Shawn M Moore [Sat, 30 May 2009 18:15:59 +0000]
Test fixes for lazy build no longer setting required
Jesse Luehrs [Sat, 30 May 2009 06:17:15 +0000]
pass 0, not 1, to inline_constructor to silence warnings
Shawn M Moore [Sat, 30 May 2009 06:05:58 +0000]
subname is computerese
Shawn M Moore [Sat, 30 May 2009 06:02:10 +0000]
Remove punctuation from an error message
Jesse Luehrs [Sat, 30 May 2009 04:17:46 +0000]
add todo test for immutable metaclasses with traits
hakim [Fri, 29 May 2009 21:42:30 +0000]
Fix documentation to point to correct superclass
Marc Mims [Tue, 26 May 2009 18:29:44 +0000]
Moved to t/100_bugs per stevan
Marc Mims [Tue, 26 May 2009 18:24:55 +0000]
Test around 'isa'
Marc Mims [Tue, 26 May 2009 01:51:26 +0000]
TODO tests for UNIVERSAL methods
Dave Rolsky [Mon, 25 May 2009 16:59:49 +0000]
Rephrase required docs
Chris Prather [Mon, 25 May 2009 16:51:38 +0000]
remove required => 1 from lazy_build
Chris Prather [Mon, 25 May 2009 16:50:05 +0000]
correct the documentation for required
Yuval Kogman [Sat, 23 May 2009 21:13:42 +0000]
Update test suite for subname changes in Class::MOP
Yuval Kogman [Sat, 23 May 2009 20:45:05 +0000]
Inline DESTROY even without DEMOLISH methods
This prevents Moose::Object::DESTROY from calling
Moose::Object::DEMOLISHALL only to find that there are no DEMOLISH
methods at every object destruction
Dave Rolsky [Fri, 22 May 2009 14:09:01 +0000]
Clarify why tests are not passing
Dave Rolsky [Fri, 22 May 2009 14:06:31 +0000]
Fix test count
Dave Rolsky [Fri, 22 May 2009 14:05:56 +0000]
Add an additional test to make sure that the original meta method
object is being preserved.
Dave Rolsky [Fri, 22 May 2009 14:02:18 +0000]
More tidying of the test file
Dave Rolsky [Fri, 22 May 2009 14:01:41 +0000]
Tidy tests
Dave Rolsky [Fri, 22 May 2009 14:01:07 +0000]
Separate packages and remove useless do
Shawn M Moore [Fri, 22 May 2009 13:57:55 +0000]
No need to document an internals change
Shawn M Moore [Fri, 22 May 2009 03:39:36 +0000]
Rename _make_wrapper to _curry_wrapper and fiddle its interface
Shawn M Moore [Fri, 22 May 2009 03:25:09 +0000]
Use class_of instead of CMOPC->initialize
Shawn M Moore [Wed, 20 May 2009 22:00:42 +0000]
TODOify the failing test I added
Chris Prather [Tue, 19 May 2009 16:01:37 +0000]
document FAQ change in Changes
Chris Prather [Tue, 19 May 2009 15:48:20 +0000]
add a FAQ about the coercion change in 0.76
Shawn M Moore [Mon, 18 May 2009 06:12:04 +0000]
When your ->meta method does not return a Moose::Meta::Class, the error message gave the wrong output
Shawn M Moore [Sat, 16 May 2009 15:02:31 +0000]
Add failing test for anonymous method metaclass persistency
Shawn M Moore [Sat, 16 May 2009 10:12:44 +0000]
If we M'Ex/also a module that isn't loaded, hint that to the user
michaelr [Wed, 13 May 2009 19:23:45 +0000]
added check for an even number of options to 'has'
Dave Rolsky [Wed, 13 May 2009 17:11:02 +0000]
bump version to 0.79
Dave Rolsky [Wed, 13 May 2009 17:10:52 +0000]
Changes for 0.79
Dave Rolsky [Wed, 13 May 2009 16:08:47 +0000]
At least try to get the inheritance chain from a metaclass object
before falling back to mro::get_linear_isa.
Dave Rolsky [Wed, 13 May 2009 16:04:39 +0000]
More global destruction fixes.
The existing global destruction tests were not happy on Win32. This is
based on a patch from Robert Krimen, but is greatly simplified by
making sure the only test output we provide is from the .t file,
rather than mixing it between the .t and the helper script.
This also changes how we loop through the ancestors for a class in
DEMOLISH. I was seeing failures on 5.8.8 that happened when the
metaclass object had been garbage collected and we tried to make a new
one. Making a new one failed because the meta-attribute objects for
Moose::Meta::Class itself had also been garbage collected!
This is a hack to avoid the meta-API entirely during destruction.
Dave Rolsky [Tue, 12 May 2009 20:16:39 +0000]
Missed a test that was badly testing for \n
Dave Rolsky [Tue, 12 May 2009 20:16:29 +0000]
Add note about test fixes
Dave Rolsky [Tue, 12 May 2009 20:12:24 +0000]
bump version to 0.78
Dave Rolsky [Tue, 12 May 2009 20:09:47 +0000]
Add release date
Dave Rolsky [Tue, 12 May 2009 20:07:33 +0000]
Fix tests that tested for literal newlines in warnings, as this failed
on Win32.
Shawn M Moore [Tue, 12 May 2009 20:05:44 +0000]
Test::Output IS optional
Dave Rolsky [Tue, 12 May 2009 14:08:15 +0000]
Oops, CMOP depends on Test::Exception 0.27, not 0.22
Dave Rolsky [Tue, 12 May 2009 14:06:08 +0000]
Depend on the latest Test::Exception (we just bumped the dep in CMOP,
might as well do it here too)
Dave Rolsky [Sun, 10 May 2009 22:26:57 +0000]
remove_metaclass_by_name only takes a single argument
Tomas Doran (t0m [Sun, 10 May 2009 16:49:02 +0000]
Tweak latter instances of the test to remove the metaclass instances for the classes which did not initially (at the time the sub class with the test in says 'extends') have them - therefore the latter tests still test the same issue
Shawn M Moore [Sun, 10 May 2009 16:41:00 +0000]
Factor out Role's before/after/around into a helper function
I would use Moose::Util::add_method_modifier but that supports
regular expressions for the method name, which roles do not.