@@ -1,3 +1,49 @@
1+Version 0.3.0.1 (2013-03-26)
2+ * A large chunk of the documentation inside the FromField module had
3+ silently failed to render in Haddock.
4+
5+Version 0.3.0.0 (2013-03-25)
6+ * Added support for PostgreSQL's Array Types. Thanks to Jason Dusek
7+ for his work on this feature.
8+
9+ * Added a brand new TypeInfo system that gives FromField instances
10+ convenient and efficient access to the pg_type metatable. This
11+ replaced the older typename cache, and was neccesary to properly
12+ support postgres array types. Thanks to Bas van Dijk for his
13+ work on this feature.
14+
15+ * Changed the type of the `fromField` and `fromRow` methods to allow
16+ a restricted set of IO actions, and stopped pre-calculating the type
17+ name of every column. As a result, the type of the `typename`
18+ operator changed from `Field -> ByteString` to
19+ `Field -> Conversion ByteString`, where Conversion is the new monad
20+ that conversion computations run inside.
21+
22+ * Improved the documentation of the FromField module.
23+
24+ * Added the Database.PostgreSQL.Simple.Errors module, which offers
25+ some predicates and functions for interpreting SqlError values, courtesy
26+ of Leonid Onokhov.
27+
28+ * Added a the name of a column and the associated table's object identifier
29+ to ResultError exceptions, courtesy of Jeff Chu.
30+
31+ * Moved most of the more detailed transaction operators into the
32+ Database.PostgreSQL.Simple.Transaction module.
33+
34+ * Changed withTransactionModeRetry to accept a predicate of which
35+ SqlErrors to retry, due to the fact that serialization errors can
36+ sometimes manifest themselves as constraint violations. Thanks
37+ to Oliver Charles for pointing this out and implementing the change.
38+
39+ * Added simple tests of the fold operator, thanks to Joey Adams.
40+
41+ * Added simple tests of the array conversion code.
42+
43+ * Added recognition of -- comments in the quasiquoter, which are
44+ now stripped out.
45+
46+
147Version 0.2.4.1 (2012-08-29)
248 * Fixed the documentation of `In`. Thanks to rekado and dstcruz for
349 pointing this out.