sequel 4.36.0 → 5.61.0

data/doc/release_notes/3.6.0.txt

@@ -1,366 +0,0 @@
-New Features
-------------
-
-* Dataset#filter and related methods now accept a string with named
-  placeholders, and a hash with placeholder values:
-
-    ds.filter('copies_sold > :sales', :sales=>500000)
-
-  Sequel's general support for this syntax is nicer:
-
-    ds.filter{copies_sold > 500000}
-
-  But named placeholder support can make it easier to port code
-  from other database libraries.  Also, it works much better than
-  the ? placeholder support if you have a long SQL statement:
-
-    DB['SELECT :n FROM t WHERE p > :q AND p < :r', :n=>1,:q=>2,:r=>3]
-
-  Sequel doesn't subsitute values that don't appear in the hash:
-
-    ds.where('price < :p AND id in :ids', :p=>100)
-    # WHERE (price < 100 AND id in :ids)
-
-  This makes it easier to spot missed placeholders, and avoids issues
-  with PostgreSQL's :: casting syntax or : inside string literals.
-
-* The Model add_ association method now accepts a hash and creates
-  a new associated model object associated to the receiver:
-
-    Artist[:name=>'YJM'].add_album(:name=>'RF')
-
-* The Model remove_ association method now accepts a primary key
-  and removes the associated model object from the association. For
-  models using composite primary keys, an array of primary key values
-  can be used.  Example:
-
-    Artist[:name=>'YJM'].remove_album(1) # regular pk
-    Artist[:name=>'YJM'].remove_album([2, 3]) # composite pk
-
-* Dataset#bind was added, allowing you to bind values before calling
-  Dataset#call.  This is more consistent with Sequel's general
-  approach where queries can be built in any order.
-
-* The native postgres adapter now has Dataset#use_cursor, which
-  allows you to process huge datasets without keeping all records in
-  memory.  The default number of rows per cursor fetch is 1000, but
-  that can be modified:
-
-    DB[:huge_table].use_cursor.each{|r| p r}
-    DB[:huge_table].use_cursor(:rows_per_fetch=>10000).each{|r| p r}
-
-  This probably won't work with prepared statements or
-  bound variables.
-
-* The nested_attributes plugin now adds newly created objects to the
-  cached association array immediately, even though the changes
-  are not persisted to the database until after the object is saved.
-  The reasoning for this is that otherwise there is no way to access
-  the newly created associated objects before the save, and no way
-  to access them at all if validation fails.
-  
-  This makes the nested_attributes plugin much easier to use, since
-  now you can just iterate over the cached association array when
-  building the form.  If validation fails, it will have the newly
-  created failed objects in the array, so you can easily display the
-  form as the user entered it for them to make changes.
-
-  This change doesn't affect many_to_one associations, since those
-  don't have a cached association array.  This also does not affect
-  updating existing records, since those are already in the cached
-  array.
-
-* You can now easily override the default options used in the
-  validation_helpers plugin (the recommended validation plugin).
-  Options can be overridden at a global level:
-
-    Sequel::Plugins::ValidationHelpers::DEFAULT_OPTIONS[:format].
-      merge!(:message=>"incorrect format", :allow_missing=>true)
-
-  Options can also be overridden on a per-class level:
-
-    class Album < Sequel::Model
-      plugin :validation_helpers
-      DEFAULT_VALIDATION_OPTIONS = {
-       :format=>{:message=>"incorrect format", :allow_missing=>true}}
-    
-      private
-    
-      def default_validation_helpers_options(type)
-        super.merge(DEFAULT_VALIDATION_OPTIONS[type] || {})
-      end
-    end
-
-* You can now use a proc instead of a string for the
-  validation_helpers :message option.  This should allow much
-  easier internationalization support.  If a proc is given, Sequel
-  calls it to get the format string to use.  Whether the proc should
-  take an argument depends on whether the associated validation
-  method takes an argument before the array of columns to validate,
-  and the argument provided is what is passed to the proc.  The
-  exception to this is the validates_not_string method, which doesn't
-  take an argument, but does pass one to the proc (a symbol with the
-  schema type of the column).
-
-  Combined with the above default option support, full
-  internationalization support for the validation_helpers plugin
-  should be fairly easy.
-
-* The nested_attributes plugin now accepts a :fields option that
-  specifies the fields that are allowed.  If specified, the
-  plugin will use set_only instead of set when mass assigning
-  attributes.  Without this, the only way to control which fields
-  are allowed is to set allowed/restricted attributes at a class
-  level in the associated class.
-
-* Associations now accept a :distinct option which uses the SQL
-  DISTINCT clause.  This can be used instead of :uniq for
-  many_to_many and many_through_many associations to handle
-  the uniqueness in the database instead of in ruby.  It can
-  also be useful for one_to_many associations to models that
-  don't have primary keys.
-
-* The caching plugin now accepts an :ignore_exceptions option that
-  allows it to work with memcached (which raises exceptions instead
-  of returning nil for missing records).
-
-* Sequel now emulates JOIN USING poorly using JOIN ON for databases
-  that don't support JOIN USING (MSSQL and H2).  This isn't
-  guaranteed to work for all queries, since USING and ON have
-  different semantics, but should work in most cases.
-
-* The MSSQL shared adapter now supports insert_select, for faster
-  model object creation.  If for some reason you need to disable it,
-  you can use disable_insert_output.
-
-* Model#modified! has been added which explicitly marks the object
-  as modified.  So even if no column values have been modified,
-  calling save_changes/update will still run through the regular
-  save process and call all before and after save/update hooks.
-
-* Model#marshallable! has been added which removes unmarshallable
-  attributes from the object.  Previously, you couldn't marshal
-  a saved model object because it contained a dataset with a
-  singleton class.  Custom _dump and _load methods could be used
-  instead, but this approach is easier to implement.
-
-* Dataset#literal_other now calls sql_literal on the object with
-  the current dataset instance, if the object responds to it.
-  This makes it easier to support the literalization of arbitrary
-  objects.
-
-  Note that if the object is a subclass of a class handled by
-  an existing dataset literalization method, you cannot use this
-  method.  You have to override the specific Dataset#literal_* method
-  in that case.
-
-* Model#save_changes now accepts an option hash that is passed to
-  save:
-
-    album.save_changes(:validate=>false)
-  
-* A bunch of Dataset#*_join methods have been added, for specific
-  join types:
-
-  * cross_join
-  * natural_join
-  * full_join
-  * left_join
-  * right_join
-  * natural_full_join
-  * natural_left_join
-  * natural_right_join
-
-  Previously, you had to use join_table(:cross, ...) to use a CROSS
-  JOIN.
-
-* You can now create clustered indexes on Microsoft SQL Server using
-  the :clustered option.
-
-* AssociationReflection#associated_object_keys has been added,
-  specifying the keys in the associated model object that are related
-  to this association. 
-
-* Sequel::SQL::SQLArray#to_a was added.
-
-Other Improvements
-------------------
-
-* Constant lookup in virtual row blocks now works correctly in ruby
-  1.9.  Virtual row blocks are based on BasicObject on ruby 1.9,
-  which doesn't allow referencing objects in the top level scope. So
-  the following code would cause an error on 1.9:
-
-    DB[:bonds].filter{maturity_date > Time.now}
-
-  Sequel now uses a Sequel::BasicObject class on 1.9 with a
-  const_missing that looks up constants in Object, which allows the
-  above code to work.
-
-* Sequel no longer attempts to load associated objects when
-  one of the key fields in the current table is NULL.  This fixes
-  the behavior when the :primary_key option for the association
-  is used to point to a non-primary key.
-
-  A consequence of this change is that attempting to load a
-  *_to_many association for a new model object now returns
-  an empty array instead of raising an exception.  This has its
-  own advantage of allowing the same association viewing code
-  to work on both new and existing objects.  Previously, you had
-  to actively avoid calling the association method on new objects,
-  or Sequel would raise an exception.
-
-* Dataset aggreate methods (sum/avg/min/max/range/interval) now
-  work correctly with limited, grouped, or compound datasets.
-  Previously, count worked with them, but other aggregate methods
-  did not.  These methods now use a subquery if called on a
-  limited, grouped or compound dataset.
-
-* It is no longer required to have an existing GROUP BY clause
-  to use a HAVING clause (except on SQLite, which doesn't permit
-  it).  Sequel has always had this limitation, but it's not required
-  by the SQL standard, and there are valid reasons to use HAVING
-  without GROUP BY.
-
-* Sequel will now emulate support for databases that don't support
-  multiple column IN/NOT IN syntax, such as MSSQL and SQLite:
-
-    ds.filter([:col1, :col2]=>[[1, 2], [3, 4]].sql_array)
-    # default:  WHERE (col1, col2) IN ((1, 2), (3, 4))
-    # emulated: WHERE (((col1 = 1) AND (col2 = 2)) OR
-    #                  ((col1 = 3) AND (col2 = 4)))
-
-  This is necessary for eager loading associated objects for models
-  with composite primary keys.
-
-* Sequel now emulates :column.ilike('blah%') for case insensitive
-  searches on MSSQL and H2.  MSSQL is case insensitive by default,
-  so it is the same as like.  H2 is case sensitive, so Sequel
-  uses a case insensitive cast there.
-
-* The nested_attributes plugin no longer allows modification of
-  keys related to the association.  This fixes a possible security
-  issue with the plugin, where a user could associate the nested
-  record to a different record.  For example:
-
-    Artist.one_to_many :albums
-    Artist.plugin :nested_attributes
-    Artist.nested_attributes :albums
-    artist = Artist.create
-    artist2 = Artist.create
-    album = Album.create
-    artist.add_album(album)
-    artist.albums_attributes = [{:id=>album.id,
-     :artist_id=>artist2.id}]
-    artist.save
-
-* The one_to_many remove_* association method now makes sure that the
-  object to be removed is currently associated to this object.
-  Before, the method could be abused to disassociate the object from
-  whatever object it was associated to.
-
-* Model add_ and remove_ association methods now check that the passed
-  object is of the correct class.
-  
-* Calling the add_* association method no longer adds the record
-  to the cached association array if the object is already in the
-  array.  Previously, Sequel did this for reciprocal associations,
-  but not for regular associations.
-
-  This makes the most sense for one_to_many associations, since
-  those can only be associated to the object once.  For many_to_many
-  associations, if you want an option to disable the behavior, please
-  bring it up on the Sequel mailing list.
-
-* An array with a string and placeholders that is passed to
-  Dataset#filter is no longer modified.  Previously:
-
-    options = ["name like ?", "%dog%"]
-    DB[:players].where(options)
-    options # => ["%dog%"]
-
-* Getting the most recently inserted autoincremented primary key
-  is now optimized when connecting to MySQL via JDBC.
-
-* Model.inherited now calls Class.inherited.
-
-* The MSSQL shared adapter once again works on ruby 1.9.  It was
-  broken in 3.5.0 due to minor syntax issues.
-
-* The force_encoding plugin now handles refreshing an existing
-  object, either explicitly or implicitly when new objects are
-  created.
-
-  To use the force_encoding plugin with the identity_map plugin, the
-  identity_map plugin should be loaded first.
-
-* Using nil as a bound variable now works on PostgreSQL.  Before,
-  Sequel would incorrectly use "" instead of NULL, since it
-  transformed all objects to strings before binding them.  Sequel
-  now binds the objects directly.
-
-* The Amalgalite adapter is now significantly faster, especially for
-  code that modifies the schema or submits arbitrary SQL statements
-  using Database <<, run, or execute_ddl.
-
-* Model#save_changes is now used when updating existing associated
-  objects in the nested_attributes plugin.  This should be
-  significantly faster for the common case of submitting a complex
-  form with nested objects without making modifications.
-
-* You can now prepare insert statements that take multiple arguments,
-  such as insert(1, 2, 3) and insert(columns, values).
-
-* Dataset#group_and_count now supports aliased columns.
-
-* Adding indexes to tables outside the default schema now works.
-
-* Eager graphing now works better with models that use aliased tables.
-
-* Sequel now correctly parses the column schema information for tables
-  in a non-default schema on Microsoft SQL Server.
-
-* changed_columns is now cleared for when saving new model objects
-  for adapters that support insert_select, such as PostgreSQL.
-
-* Dataset#replace on MySQL now works correctly when default values
-  are used.
-
-* Dataset#lock on PostgreSQL now works correctly.
-
-* Dataset#explain now works correctly on SQLite, and works using
-  any adapter.  It also works correctly on Amalgalite.
-
-* The JDBC adapter now handles binding Time arguments correctly when
-  using prepared statements.
-
-* Model add_ and remove_ association methods now have more
-  descriptive exception messages.
-
-* Dataset#simple_select_all? now ignores options that don't affect
-  the SQL, such as :server.
-
-* Dataset#window in the PostgreSQL adapter now respects existing
-  named windows.
-
-* Sequel now better handles a failure to begin a new transaction.
-
-* The dataset code was split into some additional files for improved
-  readability.
-
-* Many documentation improvements were made.
-
-Backwards Compatibility
------------------------
-
-* Model::Errors no longer uses a default proc, but emulates one in the
-  [] method.  This is unlikely to have a negative affect unless you
-  are calling a method on it that doesn't call [] (maybe using it in
-  a C extension?).
-
-* Model#table_name now only provides the alias if an aliased table is
-  used.
-
-* The Sequel::Dataset::STOCK_COUNT_OPTS constant has been removed.
-
-* Dataset#lock on PostgreSQL now returns nil instead of a dataset.

data/doc/release_notes/3.7.0.txt

@@ -1,179 +0,0 @@
-New Features
-------------
-
-* Sequel now has support for deleting and updating joined datasets
-  on MySQL and PostgreSQL.  Previously, Sequel only supported this to
-  a limited extent on Microsoft SQL Server, and support there has been
-  improved as well.  
-  
-  This allows you to do:
-  
-    DB.create_table!(:a){Integer :a; Integer :d}
-    DB.create_table!(:b){Integer :b; Integer :e}
-    DB.create_table!(:c){Integer :c; Integer :f}
-    
-    # Insert some rows
-    
-    ds = DB.from(:a, :b).
-      join(:c, :c=>:e.identifier).
-      where(:d=>:b)
-    ds.where(:f=>6).update(:a => 10)
-    ds.where(:f=>5).delete
-      
-  Which will set the a column to 10 for all rows in table a, where
-  an associated row in table c (through table b) has a value of 6 for
-  column f.  It will delete rows from table a where an associated row
-  in table c (through table b) has a value of 5 for column f.
-
-  Sequel assumes the that first FROM table is the table being
-  updated/deleted.  MySQL and Microsoft SQL Server do not require
-  multiple FROM tables, but PostgreSQL does.
-
-* Dataset #select_map, #select_order_map, and #select_hash
-  convenience methods were added for quickly creating arrays and
-  hashes from a dataset.
-  
-  select_map and select_order_map both return arrays of values for the
-  column specified.  The column can be specified either via an argument
-  or a block, similar to Dataset#get.  Both accept any valid objects as
-  arguments.
-
-  select_hash returns a hash.  It requires two symbol arguments, but
-  can handle implicit qualifiers or aliases in the symbols.
-
-  Neither of these methods offer any new functionality, they just cut
-  down on the number of required key strokes:
-
-    select_map(:column)       # select(:column).map(:column)
-    select_order_map(:column) # select(:column).order(:column).
-                              #  map(:column)
-    select_hash(:key_column, :value_column)
-    # select(:key_column, :value_column).
-    #  to_hash(:key_column, :value_column)
-
-* The NULL, NOTNULL, TRUE, SQLTRUE, FALSE, and SQLFALSE constants
-  were added to Sequel::SQL::Constants.  This allows you to do:
-  
-    include Sequel::SQL::Constants
-    DB[:table].where(:a=>'1', :b=>NOTNULL)
-    
-  Previously, the shortest way to do this was:
-
-    DB[:table].where(:a=>'1').exclude(:b=>nil)
-    
-  It may make the code more descriptive:
-
-    DB[:table].where(:b=>NULL)
-    # compared to
-    DB[:table].where(:b=>nil)
-
-  This gives the option to use SQL terminology instead of ruby
-  terminology.
-
-  The other advantage of using the constants it that they handle
-  operators and methods like other Sequel::SQL objects:
-
-    NULL & SQLFALSE # BooleanExpression => "(NULL AND FALSE)"
-    nil & false     # false
-
-    NULL + :a       # NumericExpression => "(NULL + a)"
-    nil + :a        # raises NoMethodError
-    NULL.sql_string + :a  # StringExpression => "(NULL || a)"
-    NULL.as(:b)     # AliasedExpression => "NULL AS b"
-
-  For complex systems that want to represent SQL boolean objects
-  in ruby (where you don't know exactly how they'll be used), using
-  the constants is recommended.
-
-  In order not to be too verbose, including Sequel::SQL::Constants
-  is recommended.  It's not done by default, but you can still
-  reference the constants under the main Sequel module by default
-  (e.g. Sequel::NULL).
-  
-* The validates_unique method in the validation_helpers plugin now
-  supports an :only_if_modified option, which should speed up the
-  common case where the unique attribute is not modified for an
-  existing record.  It's not on by default, since it's possible the
-  database could be changed between retrieving the model object and
-  updating it.
-  
-* The Dataset #union, #intersect, and #except methods now accept an
-  :alias option which is used as the alias for the returned dataset.
-  
-    DB[:table].union(DB[:old_table], :alias=>:table)
-
-* Model#destroy now supports a :transaction option, similar to
-  Model#save.
-
-* The shared Oracle adapter now supports Dataset#sequence for
-  returning autogenerated primary key values on insert from a
-  related sequence.
-
-  This makes Oracle work correctly when using models, with
-  something like the following:
-  
-    class Album < Sequel::Model
-      set_dataset dataset.sequence(:seq_albums_id)
-    end
-    
-  You currently need to call Dataset#sequence in every model
-  class where the underlying table uses a sequence to generate
-  primary key values.
-
-Other Improvements
-------------------
-
-* In Model #save and #destroy when using transactions and when
-  raise_on_save_failure is false, ensure that transactions are rolled
-  back if a before hook returns false.
-
-* Dataset#group_and_count now handles arguments other than Symbols.
-  A previous change to the method raised an exception if a Symbol was
-  not provided.  It also handles AliasedExpressions natively, so the
-  following works correctly:
-  
-    DB[:table].group_and_count(:column.as(:alias))
-
-* Sequel no longer uses native autoreconnection in the mysql adapter.
-  Native autoreconnection has problems with prepared statements,
-  where a new native connection is used behind Sequel's back, so
-  Sequel thinks the prepared statement has already been defined on
-  the connection, when it fact it hasn't.  Any other changes that
-  affect the state of the connection will be lost when native
-  autoreconnection is used as well.
-
-  Sequel's connection pool already handles reconnection if it detects
-  a disconnection.  This commit also adds an additional exception
-  message to recognize as a disconnect.  If there other exception
-  messages related to disconnects, please post them on the Sequel
-  mailing list.
-
-* The schema_dumper plugin now specifies the :type option for primary
-  key if it isn't Integer.
-
-* On PostgreSQL, the bigserial type is used if :type=>Bignum is
-  given as an option to primary key.  This makes it operate more
-  similarly to other adapters that support autoincrementing 64-bit
-  integer primary keys.
-  
-* The native mysql adapter will now attempt to load options in the
-  [client] section of the my.cnf file.
-
-* The rake spec tasks for the project now work correctly with RSpec
-  1.2.9.
-
-Backwards Compatibility
------------------------
-
-* Dataset::GET_ERROR_MSG and Dataset::MAP_ERROR_MSG constants were
-  removed.  Both were replaced with Dataset::ARG_BLOCK_ERROR_MSG.
-  
-* The behavior of the Model#save_failure private instance method was
-  modified.  It now always raises an exception, and validation
-  failures no longer call it.
-
-* The internals of how autogenerated primary key metadata is stored
-  when creating tables on PostgreSQL has been modified.
-
-* The native MySQL adapter no longer sets the OPT_LOCAL_INFILE option
-  to "client" on the native connection.

data/doc/release_notes/3.8.0.txt

@@ -1,151 +0,0 @@
-New Features
-------------
-
-* Dataset#each_server was added, allowing you to run the same query
-  (most likely insert/update/delete) on all shards.  This is useful
-  if you have a sharded database but have lookup tables that should
-  be identical on all shards.  It works by yielding copies of the
-  current dataset that are tied to each server/shard:
-  
-    DB[:table].filter(:id=>1).each_server do |ds|
-      ds.update(:name=>'foo')
-    end
-
-* Database#each_server was added, allowing you to run schema
-  modification methods on all shards.  It works by yielding a
-  new Sequel::Database object for each shard, that will connect to
-  only that shard:
-  
-    DB.each_server do |db|
-      db.create_table(:t){Integer :num}
-    end
-    
-* You can now add and remove servers/shards from the connection
-  pool while Sequel is running:
-  
-    DB.add_servers(:shard1=>{:host=>'s1'}, :shard2=>{:host=>'s2'})
-    DB.remove_servers(:shard1, :shard2)
-    
-* When you attempt to disconnect from a server that has connections
-  currently in use, Sequel will now schedule those connections to
-  be disconnected when they are returned to the pool.  Previously,
-  Sequel disconnected available connections, but ignored connections
-  currently in use, so it wasn't possible to guarantee complete
-  disconnection from the server.  Even with this new feature, you can
-  only guarantee eventual disconnection, since disconnection of
-  connections in use happens asynchronously.
-  
-* Database#disconnect now accepts a :servers option specifying the
-  server(s) from which to disconnect.  This should be a symbol or
-  array of symbols representing servers/shards.  Only those specified
-  will be disconnected:
-  
-    DB.disconnect(:servers=>[:shard1, :shard2])
-
-* A validates_type validation was added to the validation_helpers
-  plugin.  It allows you to check that a given column contains
-  the correct type.  I can be helpful if you are also using the
-  serialization plugin to store serialized ruby objects, by making
-  sure that the objects are of the correct type (e.g. Hash):
-  
-    def validate
-      validates_type(Hash, :options)
-    end
-
-* Sequel::SQL::Expression#== is now supported for all expressions:
-
-    :column.qualify(:table).cast(:type) == \
-      :column.qualify(:table).cast(:type)
-    # => true
-    :column.qualify(:table).cast(:type) == \
-      :other_column.qualify(:table).cast(:type)
-    # => false
-
-* When using the generic File type to create blob columns on
-  MySQL, you can specify the specific database type by using the
-  :size option (with :tiny, :medium, and :long values recognized):
-  
-    DB.create_table(:docs){File :body, :size=>:long} # longblob
-
-* The mysql adapter will now default to using mysqlplus, falling
-  back to use mysql.  mysqlplus is significantly better for threaded
-  code because queries do not block the entire interpreter.
-
-* The JDBC adapter is now able to detect certain types of disconnect
-  errors.
-  
-* ConnectionPool.servers and Database.servers were added, which
-  return an array of symbols specifying the servers/shards in use.
-
-Other Improvements
-------------------
-
-* The single-threaded connection pool now raises
-  DatabaseConnectionErrors if unable to connect, so it now operates
-  more similarly to the default connection pool.
-
-* The single-threaded connection pool now operates more similar
-  to the default connection pool when given a nonexistent server.
-
-* PGErrors are now correctly converted to DatabaseErrors in the
-  postgres adapter when preparing statements or executing prepared
-  statements.
-  
-* DatabaseDisconnectErrors are now raised correctly in the postgres
-  adapter if the connection status is not OK after a query raises an
-  error.
-
-* In the mysql adapter, multiple statements in a single query should
-  now be handled correctly in the all cases, not just when using
-  Dataset#each.  So you can now submit multiple queries in a single
-  string to Database#run.
-
-* Model object creation on Microsoft SQL Server 2000 once again
-  works correctly.  Previously, an optimization was used that was
-  only supported on 2005+.
-
-* Backslashes are no longer doubled inside string literals when
-  connecting to Microsoft SQL Server.
-  
-* The ORDER clause now correctly comes after the HAVING clause on
-  Microsoft SQL Server.
-  
-* Sequel now checks that there is an active transaction before
-  rolling back transactions on Microsoft SQL Server, since
-  there are cases where Microsoft SQL Server will roll back
-  transactions implicitly.
-  
-* Blobs are now handled correctly when connecting to H2.
-
-* 64-bit integers are now handled correctly in JDBC prepared
-  statements.
-
-* In the boolean_readers plugin, correctly handle columns not in
-  the db_schema, and don't raise an error if the model's columns
-  can't be determined.
-  
-* In the identity_map plugin, remove instances from the cache if they
-  are deleted or destroyed.
-
-Backwards Compatibility
------------------------
-
-* Dataset::FROM_SELF_KEEP_OPTS was merged into
-  Dataset::NON_SQL_OPTIONS.  While used in different places, they
-  were used for the same purpose, and entries missing from one should
-  have been included in the other.
-
-* The connection pool internals changed substantially.  Now,
-  ConnectionPool #allocated and #available_connections will return
-  nil instead of an array or hash if they are called with a
-  nonexistent server.  These are generally only used internally,
-  though they are part of the public API.  #created_count and #size
-  still return the size of the :default server when called with a
-  nonexistent server, though.
-
-* The meta_eval and metaclass private methods were removed from
-  Sequel::MetaProgramming (only the meta_def public method remains).
-  If you want these methods, use the metaid gem.
-
-* The irregular ox->oxen pluralization rule was removed from the
-  default inflections, as it screws up the more common box->boxes.