Dave Rolsky [Wed, 15 Jul 2009 14:36:55 +0000]
More code tidying
Dave Rolsky [Wed, 15 Jul 2009 14:35:21 +0000]
Fix a typo in Changes
Dave Rolsky [Wed, 15 Jul 2009 14:35:07 +0000]
tidy new code from Goro
gfx [Wed, 15 Jul 2009 10:52:04 +0000]
various tweaks, and refactor _instantiate_module not to use eval STRING
gfx [Wed, 15 Jul 2009 10:16:05 +0000]
no get_method_map in get_all_method_names
gfx [Wed, 15 Jul 2009 08:39:01 +0000]
Fix bench/foo.pl
gfx [Wed, 15 Jul 2009 08:37:12 +0000]
Profile script for KiokuDB
Shawn M Moore [Wed, 15 Jul 2009 07:32:49 +0000]
Note about JPA's sponsorship and the importance of testing for back-compat
Shawn M Moore [Wed, 15 Jul 2009 07:29:44 +0000]
Changes entry for gfx/topic/reduce-inline-constructor
Shawn M Moore [Wed, 15 Jul 2009 07:28:54 +0000]
Merge branch 'topic/reduce-inline-constructor' of git://github.com/gfx/class-mop
Shawn M Moore [Wed, 15 Jul 2009 07:21:19 +0000]
Changes entry for gfx/topic/no-get_method_map
Shawn M Moore [Wed, 15 Jul 2009 07:21:06 +0000]
Merge branch 'topic/no-get_method_map' of git://github.com/gfx/class-mop
gfx [Wed, 15 Jul 2009 06:13:52 +0000]
Improve profiling script. Type "perl bench/profile.pl" and nytprof-$branch will be created.
gfx [Wed, 15 Jul 2009 04:17:56 +0000]
Extend bench/loading-benchmark.pl
Usage: perl loading-benchmark.pl [count] [module]
gfx [Wed, 15 Jul 2009 04:14:00 +0000]
Move __class_to_pmfile() in load_first_existing_class() to a correct place
gfx [Sun, 12 Jul 2009 07:04:38 +0000]
Implement an idea of reducing inline constructors in basic metaclasses
put all the inlined constructors _new() in modules directly.
gfx [Sun, 12 Jul 2009 04:41:30 +0000]
add_attribute tweaks
gfx [Sun, 12 Jul 2009 04:32:46 +0000]
Merge branch 'master' of gitmo@git.moose.perl.org:Class-MOP
gfx [Sun, 12 Jul 2009 04:31:21 +0000]
Cache init_arg() in _generate_slot_initializer()
gfx [Sun, 12 Jul 2009 03:55:08 +0000]
no get_method_map in get_method_list
gfx [Sun, 12 Jul 2009 03:54:02 +0000]
Remove method object generation stuff, which is no longer required
Jesse Luehrs [Sat, 11 Jul 2009 22:10:09 +0000]
this syntax apparently hasn't worked in a while
gfx [Fri, 10 Jul 2009 06:31:49 +0000]
refactoring no-get_method_map with package symmbol APIs
gfx [Fri, 10 Jul 2009 03:26:44 +0000]
Improve get_method/has_method/add_method not to use get_method_map.
Normal method objects, or implicit methods, are no longer created until required.
bench/loading-benchmark.pl says this makes loading time 10% faster.
Dave Rolsky [Tue, 7 Jul 2009 20:46:01 +0000]
Revert "give unique names to method modifiers"
This reverts commit
3a642c024cf3e964eedddffa171cc501bac9265d.
Dave Rolsky [Tue, 7 Jul 2009 20:45:54 +0000]
Revert "really try to make method modifier names unique"
This reverts commit
4884b0cd5dd83114eba79aa05372cb0d52309f80.
Dave Rolsky [Tue, 7 Jul 2009 18:07:57 +0000]
really try to make method modifier names unique
Dave Rolsky [Tue, 7 Jul 2009 18:05:48 +0000]
give unique names to method modifiers
Dave Rolsky [Mon, 6 Jul 2009 18:51:24 +0000]
apply patch from Goro Fuji for anon class mem leak
Dave Rolsky [Mon, 6 Jul 2009 14:27:24 +0000]
Add changes for mem leak fixes
Dave Rolsky [Mon, 6 Jul 2009 14:25:49 +0000]
Applied patch from Goro Fuji to fix leaks in creating an anonymous
class.
Dave Rolsky [Mon, 6 Jul 2009 14:23:45 +0000]
Add test count
Dave Rolsky [Mon, 6 Jul 2009 14:23:23 +0000]
Add test count
Dave Rolsky [Mon, 6 Jul 2009 14:15:18 +0000]
no Win32 line endings, code formatting tweaks
gfx [Mon, 6 Jul 2009 07:20:16 +0000]
add bench/oading-benchmark.pl
gfx [Mon, 6 Jul 2009 07:19:31 +0000]
add bench/loading-profile.pl
gfx [Mon, 6 Jul 2009 05:42:10 +0000]
Removed "use B" in Class::MOP::Package
Dave Rolsky [Fri, 3 Jul 2009 15:32:22 +0000]
These tests all had tabs, so I just ran them through perltidy
Dave Rolsky [Fri, 3 Jul 2009 15:30:55 +0000]
add a notabs test
Dave Rolsky [Fri, 3 Jul 2009 15:22:51 +0000]
Fix release date
Dave Rolsky [Fri, 3 Jul 2009 02:24:26 +0000]
bump version to 0.89
Dave Rolsky [Fri, 3 Jul 2009 02:23:51 +0000]
This release breaks with last Moose
Dave Rolsky [Fri, 3 Jul 2009 02:23:29 +0000]
Add release date
Dave Rolsky [Fri, 3 Jul 2009 02:22:48 +0000]
Add changes for 0.89
Dave Rolsky [Fri, 3 Jul 2009 02:20:37 +0000]
Merge branch 'master' into immutable_as_real_trait
Dave Rolsky [Fri, 3 Jul 2009 02:20:22 +0000]
please do not use Moose in CMOP tests!
Dave Rolsky [Fri, 3 Jul 2009 02:13:09 +0000]
Remove some cruft left over from earlier experimentation
Dave Rolsky [Wed, 1 Jul 2009 21:42:48 +0000]
next person who uses Moose in the CMOP tests gets a beat down
Dave Rolsky [Wed, 1 Jul 2009 17:57:09 +0000]
Merge branch 'master' into immutable_as_real_trait
Dave Rolsky [Wed, 1 Jul 2009 17:55:29 +0000]
Changes tweaks
Dave Rolsky [Wed, 1 Jul 2009 17:55:10 +0000]
This test does pass (but we'll keep it since apparently it was broken
at some point).
Dave Rolsky [Wed, 1 Jul 2009 17:45:28 +0000]
Reverse sense of test to check for success, not failure
Dave Rolsky [Wed, 1 Jul 2009 17:44:36 +0000]
Add some parens to is()
Dave Rolsky [Wed, 1 Jul 2009 17:44:25 +0000]
perltidy this test a bit
Dave Rolsky [Wed, 1 Jul 2009 17:43:07 +0000]
Merge branch 'master' into immutable_as_real_trait
Dave Rolsky [Wed, 1 Jul 2009 17:42:11 +0000]
renumber test
Dave Rolsky [Wed, 1 Jul 2009 17:41:37 +0000]
Improve changes for ash's fix
Tomas Doran (t0m [Wed, 1 Jul 2009 17:35:26 +0000]
Test as requested by autarch
Chris Prather [Wed, 1 Jul 2009 14:32:47 +0000]
apply patch from ash on #moose
Dave Rolsky [Tue, 30 Jun 2009 20:10:11 +0000]
Fixes to accomodate a metaclass which is an anon subclass of something
like Moose::Meta::Class. This happens when you make your own metaclass
using Moose and then load modules which apply roles to the metaclass.
Also, when making the new metaclass immutable, don't inline the
constructor.
Dave Rolsky [Tue, 30 Jun 2009 20:01:16 +0000]
The immutable class's metaclass should be an object of the mutable
class's metaclass.
Wrap _immutable_metaclass in the trait for the sake of the tests (not
sure if the test is valid, but whatever).
Dave Rolsky [Tue, 30 Jun 2009 19:49:08 +0000]
What the heck, make the new immutable subclass immutable itself. All
tests pass, and it just feels nice.
Dave Rolsky [Tue, 30 Jun 2009 19:45:35 +0000]
almost all tests passing!
Dave Rolsky [Tue, 30 Jun 2009 19:40:56 +0000]
Hacking on making immutable a real(ish) trait so it doesn't show up in
the inheritance hierarchy at all.
Dave Rolsky [Tue, 30 Jun 2009 19:38:25 +0000]
Attempting to call ->make_immutable on a class instead of an object is
just wrong, even if it accidentally works.
Florian Ragwitz [Tue, 30 Jun 2009 18:18:51 +0000]
Ignore temporary files created by xsubpp/eumm.
Florian Ragwitz [Tue, 30 Jun 2009 18:17:52 +0000]
There can only be one PPCODE section.
Dave Rolsky [Tue, 30 Jun 2009 18:03:06 +0000]
De-tabify
Dave Rolsky [Tue, 30 Jun 2009 17:28:41 +0000]
Tweak comment for clarity
Dave Rolsky [Tue, 30 Jun 2009 16:10:47 +0000]
There's no point making the new metaclass for immutable metaclasses
immutable itself. We will never construct an object (we just rebless),
so there's no constructor to inline, and it's relatively uncommon to
need to introspect the new class.
Dave Rolsky [Tue, 30 Jun 2009 04:09:06 +0000]
Avoid a warning when the inherited method has no uninlined body (which
is weird but seems to happen in some bizarro cases).
Florian Ragwitz [Tue, 30 Jun 2009 18:00:47 +0000]
Return an empty method map instead of undef for non-existant stashes.
Dave Rolsky [Tue, 30 Jun 2009 04:05:03 +0000]
Improve error message for metaclass incompat errors
Shawn M Moore [Thu, 25 Jun 2009 21:48:46 +0000]
Update CMOP comments
Cory Watson [Thu, 25 Jun 2009 19:30:34 +0000]
Make inline_slot_access use double quotes to be consistent (found by
me, blessed by Sartak)
Jesse Luehrs [Thu, 25 Jun 2009 15:35:20 +0000]
doc inline_rebless_instance_structure
Jesse Luehrs [Thu, 25 Jun 2009 15:29:24 +0000]
update Changes
Jesse Luehrs [Thu, 25 Jun 2009 15:27:51 +0000]
add a few more tests for inline instance methods
Jesse Luehrs [Thu, 25 Jun 2009 15:15:20 +0000]
add inline version of rebless_instance_structure
Dave Rolsky [Tue, 23 Jun 2009 19:17:03 +0000]
Conflicts with the last Moose release
Dave Rolsky [Tue, 23 Jun 2009 19:15:04 +0000]
bump version to 0.88
Dave Rolsky [Tue, 23 Jun 2009 19:14:53 +0000]
Add release date
Jesse Luehrs [Tue, 23 Jun 2009 05:36:07 +0000]
update Changes
Jesse Luehrs [Tue, 23 Jun 2009 05:24:09 +0000]
fix up the 000_load test to handle the immutable trait changes
Jesse Luehrs [Tue, 23 Jun 2009 05:23:27 +0000]
need to initialize a moose meta if called from Moose::Meta::Class
Jesse Luehrs [Tue, 23 Jun 2009 05:22:49 +0000]
make CMOP::Immutable::Trait stop lying about its metaclass
Class::MOP::Class is an immutable class, so its metaclass really does
need to be the immutable version of itself
(Class::MOP::Class::Immutable::Class::MOP::Class). Hiding this just
causes confusing issues. Class::MOP::Class::Immutable::Class::MOP::Class
isa Class::MOP::Class, so this shouldn't actually affect the behavior of
any code (that wasn't doing things wrong to begin with).
Dave Rolsky [Tue, 23 Jun 2009 04:28:08 +0000]
Class::MOP::Class::Immutable::Trait can not really be made immutable (it lies when you call ->meta on it)
Dave Rolsky [Mon, 22 Jun 2009 13:28:37 +0000]
Remove references to Moose
Dave Rolsky [Sun, 21 Jun 2009 20:10:04 +0000]
Add an explicit return
Dave Rolsky [Sun, 21 Jun 2009 20:00:15 +0000]
Add release date
Jesse Luehrs [Mon, 22 Jun 2009 05:34:47 +0000]
update Changes
Jesse Luehrs [Mon, 22 Jun 2009 02:08:05 +0000]
document the __INSTANCE__ thing being pushed back to CMOP
Jesse Luehrs [Sun, 21 Jun 2009 20:33:30 +0000]
push the __INSTANCE__ parameter back into cmop
Dave Rolsky [Sun, 21 Jun 2009 16:29:06 +0000]
Update changes
Dave Rolsky [Sun, 21 Jun 2009 16:18:32 +0000]
bump version to 0.87
Dave Rolsky [Sun, 21 Jun 2009 16:18:05 +0000]
spelling and dictionary fixes
Dave Rolsky [Sun, 21 Jun 2009 16:14:09 +0000]
Make sure we always local-ize $@ and $SIG{__DIE__} for code evals.
This required some changes to the CMOP::Method::Inlined API which
propogated out to anything that calls ->_eval_closure.
Dave Rolsky [Sun, 21 Jun 2009 15:57:49 +0000]
Make sure to local-ize $@ and $SIG{__DIE__} for each eval, and rethrow
$@ as needed.
Dave Rolsky [Sun, 21 Jun 2009 15:54:27 +0000]
Remove a useless eval block
Dave Rolsky [Sun, 21 Jun 2009 15:53:03 +0000]
The immutable_transformer method is gone