Matt S Trout [Wed, 26 Mar 2008 21:23:16 +0000]
fix pod fail
Michael G Schwern [Wed, 26 Mar 2008 00:38:28 +0000]
CDBICompat::Relationships uses Clone. Make CDBICompat not load without it
so the tests skip if it's not there.
Also improve the message to only show what's missing.
Anders Nor Berle [Tue, 25 Mar 2008 04:03:32 +0000]
Back all savepoint tests out of (mysql|pg).t
There's not a dire need to have them there, the only thing they
would be testing are the ->_svp_* methods on Storage, which are
tested in 98savepoints.t anyway.
I kept the changes in 72pg.t that was neccesary for it to run
to completion though.
Anders Nor Berle [Tue, 25 Mar 2008 02:33:50 +0000]
And document the last changes
Anders Nor Berle [Tue, 25 Mar 2008 02:19:26 +0000]
Various fun things.
DONE:
* We now maintain a stack for savepoints.
* Most savepoint tests are now in a separate test file.
* Arguments to svp_* methods are now optional.
* We throw exceptions instead of warn on many places.
* gphat++
TODO:
* Document all this.
Øystein Torget [Thu, 6 Mar 2008 17:32:12 +0000]
Merge 'storage-ms-access' into 'trunk'
Marc Mims [Tue, 8 Jan 2008 01:06:05 +0000]
Merge 'file_column' into 'trunk'
r15147@agaton (orig r3920): semifor | 2008-01-07 19:06:05 +0000
Restore InflateColumn::File functionality.
Anders Nor Berle [Mon, 24 Mar 2008 10:08:51 +0000]
Merge 'savepoints' into 'trunk'
Anders Nor Berle [Mon, 24 Mar 2008 05:08:51 +0000]
Mention savepoint support in Changes
Anders Nor Berle [Mon, 24 Mar 2008 03:54:07 +0000]
Fix typo.
Anders Nor Berle [Mon, 24 Mar 2008 03:50:43 +0000]
Document the auto_savepoint option for connect_info
Anders Nor Berle [Mon, 24 Mar 2008 03:14:15 +0000]
Initial commit of auto_savepoint + some fixes
Cory G Watson [Sun, 23 Mar 2008 23:52:13 +0000]
Remove unnecessary passing of $self->dbh as a separate param to savepoint methods
Cory G Watson [Sun, 23 Mar 2008 23:42:04 +0000]
Merge 0.09's savepoint branch
Wallace Reis [Sat, 22 Mar 2008 14:28:43 +0000]
pod fixes
Ash Berlin [Wed, 19 Mar 2008 22:31:31 +0000]
Tests for error propogation of multi create errors
Ash Berlin [Mon, 17 Mar 2008 21:56:47 +0000]
Fix test to reflect correct default state for deferrable constraints in SQL::T
Ash Berlin [Mon, 17 Mar 2008 21:46:07 +0000]
Fix is_deferrable and respcet sqltargs properly
Michael G Schwern [Sat, 15 Mar 2008 07:13:01 +0000]
Simplify and speed up the accessor_name_for() code.
Dagfinn Ilmari Mannsåker [Sat, 15 Mar 2008 05:26:05 +0000]
Eliminate remaining uses of $self->dbh_do($self->can(...), ...) now that dbh_do can take a method name
Dagfinn Ilmari Mannsåker [Sat, 15 Mar 2008 05:20:13 +0000]
$dbh->doe doesn't take (Raise|Print)Error attributes, use eval {} instead.
Michael G Schwern [Sat, 15 Mar 2008 05:08:34 +0000]
Missing requires of Test::NoWarnings
Michael G Schwern [Sat, 15 Mar 2008 05:01:47 +0000]
Reduce the number of times $self->_dbh is called inside dbh_do() to speed
up that hot bit of code.
Michael G Schwern [Sat, 15 Mar 2008 05:00:56 +0000]
Forgot to require Test::Warn
Michael G Schwern [Sat, 15 Mar 2008 04:55:30 +0000]
Eliminate expensive calls to can() in some very hot portions of the code by
allowing dbh_do() to take a method name.
$obj->$method_name() is about 50% faster then $obj->can($method_name)->().
Michael G Schwern [Sat, 15 Mar 2008 03:01:57 +0000]
Add missing method to NoObjectIndex.
Michael G Schwern [Fri, 14 Mar 2008 22:50:04 +0000]
A little performance hack to speed up efficiency. Internal checks that the
resultset is defined would make a SQL COUNT happen.
Michael G Schwern [Fri, 14 Mar 2008 03:27:47 +0000]
Better emulation of add_constructor, unfortunately also slower.
Michael G Schwern [Fri, 14 Mar 2008 03:17:12 +0000]
Reorder the accessor_name_for() check to get the more likely one first to
squeeze a little performance out maybe.
Michael G Schwern [Fri, 14 Mar 2008 03:16:36 +0000]
Improve add_constructor() support to handle ORDER BY and LIMIT with newlines in
the SQL
Michael G Schwern [Fri, 14 Mar 2008 02:01:20 +0000]
Move has_a() code out of ColumnCase. That should be the last of it.
Michael G Schwern [Fri, 14 Mar 2008 01:50:05 +0000]
Move code having to do with adding columns out of ColumnCase.
Michael G Schwern [Fri, 14 Mar 2008 01:41:32 +0000]
Move the code to create accessors out of ColumnCase and into ColumnGroups.
This allows one to remove ColumnCase without effecting other features.
Michael G Schwern [Fri, 14 Mar 2008 01:32:02 +0000]
Fix the skip condition.
Michael G Schwern [Thu, 13 Mar 2008 19:06:48 +0000]
Detabify
Michael G Schwern [Thu, 13 Mar 2008 19:04:13 +0000]
Make this test depend on having DeepAbstractSearch installed.
Michael G Schwern [Thu, 13 Mar 2008 19:03:40 +0000]
Fix podcoverage test for the new SQLTransformer.pm
Michael G Schwern [Thu, 13 Mar 2008 18:54:22 +0000]
Document that DeepAbstractSearch will work.
Michael G Schwern [Thu, 13 Mar 2008 18:52:14 +0000]
Make Class::DBI::Plugin::DeepAbstractSearch work.
This required reordering the relationship declarations in the tests to fit
the limitations of CDBICompat. Also fixing the behavior of sth_to_objects()
and transform_sql().
Michael G Schwern [Thu, 13 Mar 2008 18:50:32 +0000]
Test lazy loading with a deleted object.
Michael G Schwern [Thu, 13 Mar 2008 18:49:57 +0000]
Test the behavior of construct() with temp columns.
Michael G Schwern [Wed, 12 Mar 2008 21:29:58 +0000]
Fix lazy loading when the object has been deleted
Matt S Trout [Tue, 11 Mar 2008 20:40:12 +0000]
Merge 'oracle_sequence' into 'trunk'
Ash Berlin [Tue, 11 Mar 2008 15:10:50 +0000]
Make throw_exception in RS a bit more forgiving too
Ash Berlin [Tue, 11 Mar 2008 15:03:34 +0000]
Fix errors from resultset components (and move tests into t/90ensure_class_loaded since its testing same sort of things)
Ash Berlin [Tue, 11 Mar 2008 14:38:10 +0000]
test to make sure the errors in components of resultset classes are reported right.
Ash Berlin [Tue, 11 Mar 2008 14:29:33 +0000]
Fix depend on Time::Piece for cdbi
Matt S Trout [Tue, 11 Mar 2008 03:19:20 +0000]
fix Date::Simple req in t/cdbi-t/04-lazy.t
Matt S Trout [Tue, 11 Mar 2008 03:15:09 +0000]
moved t/cdbi-t/08-inheritcols.t to use CDBICompat
Matt S Trout [Tue, 11 Mar 2008 03:12:33 +0000]
moving t/cdbi-t/hasa_without_loading.t to CDBICompat makes it fail, can't leave it using CDBI so have temporarily skipped the test that fails
Matt S Trout [Tue, 11 Mar 2008 03:03:50 +0000]
Merge 'cdbicompat_integration' into 'trunk'
Michael G Schwern [Tue, 11 Mar 2008 01:18:49 +0000]
Fixed a heisenbug where looking at a column group would cause it to be shared.
Michael G Schwern [Tue, 11 Mar 2008 00:38:09 +0000]
No reason to store an empty hash ref for each column
Marcus Ramberg [Mon, 10 Mar 2008 09:50:47 +0000]
proper skip
Michael G Schwern [Tue, 4 Mar 2008 16:29:45 +0000]
Wrap columns() in an array ref for better is_deeply() diagnostics.
Michael G Schwern [Sun, 24 Feb 2008 17:13:43 +0000]
Make meta_info() 'args' work.
Michael G Schwern [Fri, 7 Mar 2008 16:05:35 +0000]
Fix the skip for DBD::Multi
Michael G Schwern [Fri, 7 Mar 2008 16:04:34 +0000]
Fix the merge with txn_scope_guard() and put the necessary =cut in.
Michael G Schwern [Fri, 7 Mar 2008 15:54:18 +0000]
Merge 'trunk' into 'cdbicompat_integration'
r54995@windhund (orig r4113): ash | 2008-03-01 13:17:54 +0100
Fix versioning test so it works with SQLT 0.09.
r54996@windhund (orig r4114): ash | 2008-03-04 13:06:34 +0100
is_deferrable patch from Debolaz
r55001@windhund (orig r4119): teejay | 2008-03-04 20:29:45 +0100
Applied fixes to replication connect_info method
r55002@windhund (orig r4120): teejay | 2008-03-04 20:31:32 +0100
Applied fixes to replication connect_info method
r55003@windhund (orig r4121): teejay | 2008-03-04 21:18:45 +0100
works but shouldn't
r55004@windhund (orig r4122): teejay | 2008-03-04 21:41:32 +0100
roughly DTRT - please to be making less awful
r55008@windhund (orig r4126): captainL | 2008-03-05 01:07:45 +0100
r3439@luke-mbp (orig r3680): captainL | 2007-08-15 13:56:57 +0100
new branch for ::Versioned enhancements
r3503@luke-mbp (orig r3681): captainL | 2007-08-15 14:29:01 +0100
created ->get_db_version and moved all overridable stuff to do_upgrade
r3955@luke-mbp (orig r3682): captainL | 2007-08-15 23:29:57 +0100
tests are a mess, but Versioned.pm should work now
r3956@luke-mbp (orig r3683): captainL | 2007-08-16 00:45:32 +0100
moved upgrade file reading into upgrade from _on_connect
r3958@luke-mbp (orig r3689): captainL | 2007-08-21 12:56:31 +0100
works well, we now just require a nice way to deploy the SchemaVersions table
r3959@luke-mbp (orig r3692): captainL | 2007-08-21 17:58:17 +0100
determines parser from dbh driver name and gives parser the dbh from schema to connect
r4213@luke-mbp (orig r3831): captainL | 2007-10-23 13:18:13 +0100
fixed versioning test and checked db and schema versions are not equal before upgrading
r4214@luke-mbp (orig r3832): captainL | 2007-10-23 15:08:46 +0100
changed constraint and index generation to be consistent with DB defaults
r4215@luke-mbp (orig r3833): captainL | 2007-10-23 15:40:47 +0100
added entry to Changes
r4283@luke-mbp (orig r3838): captainL | 2007-10-24 21:42:22 +0100
fixed broken regex when reading sql files
r5785@luke-mbp (orig r3891): ash | 2007-11-24 22:17:41 +0000
Change diffing code to use $sqlt_schema. Sort tables in parser
r5786@luke-mbp (orig r3892): captainL | 2007-11-25 16:26:57 +0000
upgrade will only produce a diff between the DB and the DBIC schema if explicitly requested
r5824@luke-mbp (orig r4012): ash | 2008-02-01 19:33:00 +0000
Fallback to SQL->SQL to diff for old producers
r5825@luke-mbp (orig r4014): castaway | 2008-02-01 23:01:26 +0000
Sanitise filename of sqlite backup file
r5830@luke-mbp (orig r4047): captainL | 2008-02-09 15:26:50 +0000
sanified layout of Versioned.pm and documented changes
r6828@luke-mbp (orig r4075): ash | 2008-02-13 13:26:10 +0000
Fix typo
r6831@luke-mbp (orig r4078): captainL | 2008-02-14 00:27:14 +0000
fixed versioned loading split bug
r6846@luke-mbp (orig r4103): captainL | 2008-02-27 15:11:21 +0000
increased sqlt rev dep
r6847@luke-mbp (orig r4104): captainL | 2008-02-27 15:12:12 +0000
fixed behaviour or is_foreign_key_constraint and unique index names
r6848@luke-mbp (orig r4105): captainL | 2008-02-28 10:28:31 +0000
changed versioning table from SchemaVersions to dbix_class_schema_versions with transition ability
r6849@luke-mbp (orig r4106): captainL | 2008-02-28 10:54:28 +0000
hack bugfix for sqlt_type weirdness
r6850@luke-mbp (orig r4107): captainL | 2008-02-28 16:11:44 +0000
cleaned up get_db_version
r6851@luke-mbp (orig r4108): captainL | 2008-02-28 16:41:19 +0000
lowercased column names of versions table
r6852@luke-mbp (orig r4109): captainL | 2008-02-28 16:59:04 +0000
removed startup comment if no action required
r6862@luke-mbp (orig r4123): captainL | 2008-03-04 23:14:23 +0000
improved docs and added env var to skip version checks on connect
r6864@luke-mbp (orig r4125): captainL | 2008-03-04 23:28:21 +0000
manual merge for deferrable changes from trunk
r55009@windhund (orig r4127): captainL | 2008-03-05 11:38:19 +0100
added entry in Changes for versioning branch
r55025@windhund (orig r4143): castaway | 2008-03-07 00:26:08 +0100
Random pod/doc pokage
Jess Robinson [Thu, 6 Mar 2008 23:26:08 +0000]
Random pod/doc pokage
Øystein Torget [Thu, 6 Mar 2008 11:32:12 +0000]
Added the function bind_attributes_by_data_type to cope with the 'Invalid precision value' bug on MEMO columns.
Added documentation for the 'Invalid precision value' bug.
Luke Saunders [Wed, 5 Mar 2008 10:38:19 +0000]
added entry in Changes for versioning branch
Luke Saunders [Wed, 5 Mar 2008 00:07:45 +0000]
Merge 'versioned_enhancements' into 'trunk'
r3439@luke-mbp (orig r3680): captainL | 2007-08-15 13:56:57 +0100
new branch for ::Versioned enhancements
r3503@luke-mbp (orig r3681): captainL | 2007-08-15 14:29:01 +0100
created ->get_db_version and moved all overridable stuff to do_upgrade
r3955@luke-mbp (orig r3682): captainL | 2007-08-15 23:29:57 +0100
tests are a mess, but Versioned.pm should work now
r3956@luke-mbp (orig r3683): captainL | 2007-08-16 00:45:32 +0100
moved upgrade file reading into upgrade from _on_connect
r3958@luke-mbp (orig r3689): captainL | 2007-08-21 12:56:31 +0100
works well, we now just require a nice way to deploy the SchemaVersions table
r3959@luke-mbp (orig r3692): captainL | 2007-08-21 17:58:17 +0100
determines parser from dbh driver name and gives parser the dbh from schema to connect
r4213@luke-mbp (orig r3831): captainL | 2007-10-23 13:18:13 +0100
fixed versioning test and checked db and schema versions are not equal before upgrading
r4214@luke-mbp (orig r3832): captainL | 2007-10-23 15:08:46 +0100
changed constraint and index generation to be consistent with DB defaults
r4215@luke-mbp (orig r3833): captainL | 2007-10-23 15:40:47 +0100
added entry to Changes
r4283@luke-mbp (orig r3838): captainL | 2007-10-24 21:42:22 +0100
fixed broken regex when reading sql files
r5785@luke-mbp (orig r3891): ash | 2007-11-24 22:17:41 +0000
Change diffing code to use $sqlt_schema. Sort tables in parser
r5786@luke-mbp (orig r3892): captainL | 2007-11-25 16:26:57 +0000
upgrade will only produce a diff between the DB and the DBIC schema if explicitly requested
r5824@luke-mbp (orig r4012): ash | 2008-02-01 19:33:00 +0000
Fallback to SQL->SQL to diff for old producers
r5825@luke-mbp (orig r4014): castaway | 2008-02-01 23:01:26 +0000
Sanitise filename of sqlite backup file
r5830@luke-mbp (orig r4047): captainL | 2008-02-09 15:26:50 +0000
sanified layout of Versioned.pm and documented changes
r6828@luke-mbp (orig r4075): ash | 2008-02-13 13:26:10 +0000
Fix typo
r6831@luke-mbp (orig r4078): captainL | 2008-02-14 00:27:14 +0000
fixed versioned loading split bug
r6846@luke-mbp (orig r4103): captainL | 2008-02-27 15:11:21 +0000
increased sqlt rev dep
r6847@luke-mbp (orig r4104): captainL | 2008-02-27 15:12:12 +0000
fixed behaviour or is_foreign_key_constraint and unique index names
r6848@luke-mbp (orig r4105): captainL | 2008-02-28 10:28:31 +0000
changed versioning table from SchemaVersions to dbix_class_schema_versions with transition ability
r6849@luke-mbp (orig r4106): captainL | 2008-02-28 10:54:28 +0000
hack bugfix for sqlt_type weirdness
r6850@luke-mbp (orig r4107): captainL | 2008-02-28 16:11:44 +0000
cleaned up get_db_version
r6851@luke-mbp (orig r4108): captainL | 2008-02-28 16:41:19 +0000
lowercased column names of versions table
r6852@luke-mbp (orig r4109): captainL | 2008-02-28 16:59:04 +0000
removed startup comment if no action required
r6862@luke-mbp (orig r4123): captainL | 2008-03-04 23:14:23 +0000
improved docs and added env var to skip version checks on connect
r6864@luke-mbp (orig r4125): captainL | 2008-03-04 23:28:21 +0000
manual merge for deferrable changes from trunk
Luke Saunders [Tue, 4 Mar 2008 23:28:21 +0000]
manual merge for deferrable changes from trunk
Luke Saunders [Tue, 4 Mar 2008 23:16:41 +0000]
Merge 'trunk' into 'versioned_enhancements'
r6854@luke-mbp (orig r4110): ash | 2008-02-28 19:33:59 +0000
Todo tests for txn_rollback and scope_guard
r6856@luke-mbp (orig r4113): ash | 2008-03-01 12:17:54 +0000
Fix versioning test so it works with SQLT 0.09.
r6857@luke-mbp (orig r4114): ash | 2008-03-04 12:06:34 +0000
is_deferrable patch from Debolaz
r6858@luke-mbp (orig r4119): teejay | 2008-03-04 19:29:45 +0000
Applied fixes to replication connect_info method
r6859@luke-mbp (orig r4120): teejay | 2008-03-04 19:31:32 +0000
Applied fixes to replication connect_info method
r6860@luke-mbp (orig r4121): teejay | 2008-03-04 20:18:45 +0000
works but shouldn't
r6861@luke-mbp (orig r4122): teejay | 2008-03-04 20:41:32 +0000
roughly DTRT - please to be making less awful
Luke Saunders [Tue, 4 Mar 2008 23:14:23 +0000]
improved docs and added env var to skip version checks on connect
Aaron Trevena [Tue, 4 Mar 2008 20:41:32 +0000]
roughly DTRT - please to be making less awful
Aaron Trevena [Tue, 4 Mar 2008 20:18:45 +0000]
works but shouldn't
Aaron Trevena [Tue, 4 Mar 2008 19:31:32 +0000]
Applied fixes to replication connect_info method
Aaron Trevena [Tue, 4 Mar 2008 19:29:45 +0000]
Applied fixes to replication connect_info method
Ash Berlin [Tue, 4 Mar 2008 12:06:34 +0000]
is_deferrable patch from Debolaz
Ash Berlin [Sat, 1 Mar 2008 12:17:54 +0000]
Fix versioning test so it works with SQLT 0.09.
Ash Berlin [Thu, 28 Feb 2008 17:33:59 +0000]
Merge 'trunk' into 'cdbicompat_integration'
r54651@windhund (orig r4110): ash | 2008-02-28 11:33:59 -0800
Todo tests for txn_rollback and scope_guard
Ash Berlin [Thu, 28 Feb 2008 19:33:59 +0000]
Todo tests for txn_rollback and scope_guard
Luke Saunders [Thu, 28 Feb 2008 16:59:04 +0000]
removed startup comment if no action required
Luke Saunders [Thu, 28 Feb 2008 16:41:19 +0000]
lowercased column names of versions table
Luke Saunders [Thu, 28 Feb 2008 16:11:44 +0000]
cleaned up get_db_version
Luke Saunders [Thu, 28 Feb 2008 10:54:28 +0000]
hack bugfix for sqlt_type weirdness
Luke Saunders [Thu, 28 Feb 2008 10:28:31 +0000]
changed versioning table from SchemaVersions to dbix_class_schema_versions with transition ability
Luke Saunders [Wed, 27 Feb 2008 15:12:12 +0000]
fixed behaviour or is_foreign_key_constraint and unique index names
Luke Saunders [Wed, 27 Feb 2008 15:11:21 +0000]
increased sqlt rev dep
Ash Berlin [Sun, 10 Feb 2008 23:31:27 +0000]
Merge 'trunk' into 'versioned_enhancements'
r6807@luke-mbp (orig r4050): ash | 2008-02-10 17:31:27 +0000
Add txn_scope_guard method/object
Michael G Schwern [Sun, 24 Feb 2008 07:45:49 +0000]
Remove the now redudant ObjIndexStubs.
Fix pod coverage.
Michael G Schwern [Sun, 24 Feb 2008 07:43:58 +0000]
Add NoObjectIndex which turns the live object index off and removes all the
hooks in insert and such, but retains stub indexing methods so as not to
break code.
Michael G Schwern [Sun, 24 Feb 2008 07:24:55 +0000]
Fix the DESTROY/discard_changes() infinite recursion at the DBIC level.
Michael G Schwern [Sun, 24 Feb 2008 07:22:57 +0000]
Fix the pod coverage failure.
Michael G Schwern [Sun, 24 Feb 2008 07:20:41 +0000]
Greatly speed up result_source_instance() and all the many things which
use it by eliminating the calls to $self->result_class.
Michael G Schwern [Fri, 22 Feb 2008 03:04:31 +0000]
The infinite loop with sub DESTROY { $_[0]->discard_changes } is a DBIC
thing.
Michael G Schwern [Sun, 24 Feb 2008 09:12:46 +0000]
Ensure that has_many() loads the foreign class.
Michael G Schwern [Thu, 21 Feb 2008 06:34:10 +0000]
Add in a warning if a column is declared as TEMP but it's already declared
real.
Michael G Schwern [Thu, 21 Feb 2008 08:03:47 +0000]
Merge 'trunk' into 'cdbicompat_integration'
r54378@windhund (orig r3954): matthewt | 2008-01-18 05:03:08 -0800
added strict and warnings to HashRefInflator, fixed inflation for empty has_many rels
r54379@windhund (orig r3955): matthewt | 2008-01-20 04:28:55 -0800
made search_rs smarter about when to preserve the cache to fix mm prefetch usage
r54389@windhund (orig r3965): semifor | 2008-01-22 07:13:11 -0800
Added build_datetime_parser method for MSSQL over ODBC.
r54394@windhund (orig r3970): castaway | 2008-01-24 05:19:52 -0800
Oops, fix joining manual to be correct
r54395@windhund (orig r3971): castaway | 2008-01-24 15:22:49 -0800
Version 0.08009
r54396@windhund (orig r3972): castaway | 2008-01-24 15:36:59 -0800
0.08009 released
r54399@windhund (orig r3975): tomboh | 2008-01-25 09:20:38 -0800
Fix a typo and a couple of links.
r54448@windhund (orig r4024): oyse | 2008-02-05 00:42:32 -0800
Added Øystein Torget to the list of contributers
r54474@windhund (orig r4050): ash | 2008-02-10 09:31:27 -0800
Add txn_scope_guard method/object
Michael G Schwern [Thu, 14 Feb 2008 10:23:57 +0000]
Implement cascade => "None"
Michael G Schwern [Thu, 14 Feb 2008 09:57:52 +0000]
cascade delete test from Class-DBI 3.0.17
Michael G Schwern [Thu, 14 Feb 2008 07:43:24 +0000]
Put the stringification back, older versions of DateTime string equality are
still broken. But don't bother to stringify anything that's not a reference.
Luke Saunders [Thu, 14 Feb 2008 00:27:14 +0000]
fixed versioned loading split bug
Michael G Schwern [Wed, 13 Feb 2008 22:10:33 +0000]
Expand the CDBICompat docs to show of it's features, the fact that you're
getting DBIC objects and to be more oriented towards CDBI users.
Michael G Schwern [Wed, 13 Feb 2008 22:10:06 +0000]
Emulate $CDBI::Weaken_Not_Available and CDBI::Plugin::NoCache
Ash Berlin [Wed, 13 Feb 2008 13:26:10 +0000]
Fix typo
Michael G Schwern [Wed, 13 Feb 2008 08:14:34 +0000]
Forgot to set up the relationship between Actor and Film. Turns it out
was working because create() was holding onto the given values but once
we changed it to throw them out it (properly) broke.
Michael G Schwern [Wed, 13 Feb 2008 07:41:26 +0000]
Fix create() in the same way as update() so it throws out the new values and
reloads them from the database on demand.