[catagits/Catalyst-Runtime.git] / TODO

TBD pre 5.8000_04 dev release:

  - Go through everything which used to use CAF, and ensure that it now uses MX::E::CAF + tests, 
    extending t/caf_backcompat.t and replacing t/custom_request.t (t0m)

  - Simple test for NEXT compat in core (t0m).

  - Looks like stash is not available during prepare_path when running under
    Catalyst::Test.. Run MojoMojo's t/01app.t and it will explode with
    "Can't use an undefined value as a HASH reference" - test case for core
    + fix maybe (marcus)

---

  - Make the skipped test at the bottom of t/aggregate/live_engine_request_uri.t
    pass / not be skipped. (From what C::P::SmartURI used to do)

    This can be fixed by the following patch:

Index: lib/Catalyst/Request.pm
===================================================================
--- lib/Catalyst/Request.pm (revision 8709)
+++ lib/Catalyst/Request.pm (working copy)
@@ -96,7 +96,7 @@
   lazy => 1,
   default => sub {
     my $self = shift;
-    return $self->path if $self->uri;
+    return $self->{path} if $self->{uri};
   },
 );

    But I'd like a 2nd opinion from someone who knows core better than me
    about if that is the correct fix.. (t0m / Caelum)

  - Common engine test failures, look into and get tests into core.

  - Catalyst-Plugin-Authorization-ACL, Can't locate object method "tree" via package "Catalyst::Dispatcher", fix the plugin as tree was never a public method.

  - Catalyst-Model-Akismet, incorrectly inherits from Catalyst::Component::InstancePerContext, should compose as a role, fix the module.

  - CatalystX-Imports, Class::MOP doesn't consider anon subs in the symbol table as methods, tests + fix?

  - Catalyst::Plugin::Cache::Curried
  - Catalyst::Controller::WrapCGI
    - Both import 'meta' into controller package - mst thinks this can be fixed?.
      but @marcus> and the generated accessors even call $_[0]->meta, so I don't think we can remove it. 
      . Add a test for this in core.
      . Fix core to use CMOP to look up metaclass, rather than ->meta method, and
        namespace::clean the meta method out, fix generated accessors to do the same?

  - MooseX::Emulate::Class::Accessor::Fast
    - Tests for uses of C::A::F from Catalyst repos. - t0m
    - New release once these are done.

  - Update Test suite to not assume MyApp ISA Controller
    - After that set up attr handlers that will output helpful error messages when you do it as well as how to fix it. (done already?)
  
  - Comments marked /Moose TODO/i in the code
  
  - Eliminate all instances of $instance->{$key}

  - Catalyst-Log-Log4perl - deep recursion in the test suite, investigate

  - Profiling vs 5.70 and optimisation as needed.

  - http://lists.scsys.co.uk/pipermail/catalyst-dev/2008-November/001546.html
    - patch to list, andyg to look at?
Commit	Line	Data
dae2b0fa	1	TBD pre 5.8000_04 dev release:
2105b371	2
241edc9b	3	- Go through everything which used to use CAF, and ensure that it now uses MX::E::CAF + tests,
241edc9b	4	extending t/caf_backcompat.t and replacing t/custom_request.t (t0m)
30a50b86	5
dae2b0fa	6	- Simple test for NEXT compat in core (t0m).
	7
	8	- Looks like stash is not available during prepare_path when running under
	9	Catalyst::Test.. Run MojoMojo's t/01app.t and it will explode with
	10	"Can't use an undefined value as a HASH reference" - test case for core
	11	+ fix maybe (marcus)
	12
	13	---
	14
	15	- Make the skipped test at the bottom of t/aggregate/live_engine_request_uri.t
	16	pass / not be skipped. (From what C::P::SmartURI used to do)
	17
	18	This can be fixed by the following patch:
	19
	20	Index: lib/Catalyst/Request.pm
	21	===================================================================
	22	--- lib/Catalyst/Request.pm (revision 8709)
	23	+++ lib/Catalyst/Request.pm (working copy)
	24	@@ -96,7 +96,7 @@
	25	lazy => 1,
	26	default => sub {
	27	my $self = shift;
	28	- return $self->path if $self->uri;
	29	+ return $self->{path} if $self->{uri};
	30	},
	31	);
	32
	33	But I'd like a 2nd opinion from someone who knows core better than me
	34	about if that is the correct fix.. (t0m / Caelum)
2105b371	35
	36	- Common engine test failures, look into and get tests into core.
	37
	38	- Catalyst-Plugin-Authorization-ACL, Can't locate object method "tree" via package "Catalyst::Dispatcher", fix the plugin as tree was never a public method.
	39
	40	- Catalyst-Model-Akismet, incorrectly inherits from Catalyst::Component::InstancePerContext, should compose as a role, fix the module.
	41
2105b371	42	- CatalystX-Imports, Class::MOP doesn't consider anon subs in the symbol table as methods, tests + fix?
2105b371	43
8610de43	44	- Catalyst::Plugin::Cache::Curried
	45	- Catalyst::Controller::WrapCGI
	46	- Both import 'meta' into controller package - mst thinks this can be fixed?.
	47	but @marcus> and the generated accessors even call $_[0]->meta, so I don't think we can remove it.
	48	. Add a test for this in core.
	49	. Fix core to use CMOP to look up metaclass, rather than ->meta method, and
	50	namespace::clean the meta method out, fix generated accessors to do the same?
	51
dae2b0fa	52	- MooseX::Emulate::Class::Accessor::Fast
	53	- Tests for uses of C::A::F from Catalyst repos. - t0m
	54	- New release once these are done.
8610de43	55
2105b371	56	- Update Test suite to not assume MyApp ISA Controller
	57	- After that set up attr handlers that will output helpful error messages when you do it as well as how to fix it. (done already?)
	58
	59	- Comments marked /Moose TODO/i in the code
	60
	61	- Eliminate all instances of $instance->{$key}
	62
	63	- Catalyst-Log-Log4perl - deep recursion in the test suite, investigate
	64
	65	- Profiling vs 5.70 and optimisation as needed.
	66
8610de43	67	- http://lists.scsys.co.uk/pipermail/catalyst-dev/2008-November/001546.html
	68	- patch to list, andyg to look at?
	69