<note>
<title>Release Date</title>
- <simpara>2013-XX-XX, CURRENT AS OF 2013-08-16</simpara>
+ <simpara>2013-09-09</simpara>
</note>
<sect2>
<title>Overview</title>
<para>
- Major enhancements include:
+ Major enhancements in <productname>PostgreSQL</> 9.3 include:
</para>
<!-- This list duplicates items below, but without authors or details-->
<listitem>
<para>
- Many <type>JSON</> improvements, including the addition of <link
- linkend="functions-json">operators and functions</link> to extract
- values from <type>JSON</> data strings
+ Add many features for the <type>JSON</> data type,
+ including <link linkend="functions-json">operators and functions</link>
+ to extract elements from <type>JSON</> values
</para>
</listitem>
<listitem>
<para>
- Add a <productname>Postgres</> <link linkend="postgres-fdw">foreign
- data wrapper</link> contrib module
+ Add a <link linkend="postgres-fdw"><productname>Postgres</> foreign
+ data wrapper</link> to allow access to
+ other <productname>Postgres</> servers
</para>
</listitem>
<listitem>
<para>
- Allow a streaming replication standby to <link
- linkend="protocol-replication">follow a timeline switch</link>,
- and faster failover
+ Prevent non-key-field row updates from blocking foreign key checks
</para>
</listitem>
<listitem>
<para>
- Dramatically reduce System V <link linkend="sysvipc">shared
+ Greatly reduce System V <link linkend="sysvipc">shared
memory</link> requirements
</para>
</listitem>
- <listitem>
- <para>
- Prevent non-key-field row updates from locking foreign key rows
- </para>
- </listitem>
-
</itemizedlist>
<para>
<para>
Users who have set <link
linkend="guc-work-mem"><varname>work_mem</></link> based on the
- previous behavior should revisit that setting.
+ previous behavior may need to revisit that setting.
</para>
</listitem>
<listitem>
<para>
- Throw an error if expiring tuple is again updated or deleted (Kevin Grittner)
- DETAILS?
+ Throw an error if a tuple to be updated or deleted has already been
+ updated or deleted by a <literal>BEFORE</> trigger (Kevin Grittner)
+ </para>
+
+ <para>
+ Formerly, the originally-intended update was silently skipped,
+ resulting in logical inconsistency since the trigger might have
+ propagated data to other places based on the intended update.
+ Now an error is thrown to prevent the inconsistent results from being
+ committed. If this change affects your application, the best solution
+ is usually to move the data-propagation actions to
+ an <literal>AFTER</> trigger.
+ </para>
+
+ <para>
+ This error will also be thrown if a query invokes a volatile function
+ that modifies rows that are later modified by the query itself.
+ Such cases likewise previously resulted in silently skipping updates.
</para>
</listitem>
<listitem>
<para>
- Change <link linkend="SQL-CREATETABLE"><literal>ON UPDATE
+ Change multicolumn <link linkend="SQL-CREATETABLE"><literal>ON UPDATE
SET NULL/SET DEFAULT</></link> foreign key actions to affect
- all referenced columns, not just those referenced in the
+ all columns of the constraint, not just those changed in the
<command>UPDATE</> (Tom Lane)
</para>
<para>
- Previously only columns referenced in the <command>UPDATE</> were
- set to null or <literal>DEFAULT</>.
+ Previously, we would set only those referencing columns that
+ correspond to referenced columns that were changed by
+ the <command>UPDATE</>. This was what was required by SQL-92,
+ but more recent editions of the SQL standard specify the new behavior.
</para>
</listitem>
<listitem>
<para>
- Internally store default foreign key matches (non-<literal>FULL</>,
- non-<literal>PARTIAL</>) as <quote>simple</> (Tom Lane)
+ Force cached plans to be replanned if the <link
+ linkend="guc-search-path"><varname>search_path</></link> changes
+ (Tom Lane)
</para>
<para>
- These were previously stored as "<unspecified>".
- This changes the value stored in system column <link
- linkend="catalog-pg-constraint"><structname>pg_constraint.confmatchtype</></link>.
+ Previously, cached plans already generated in the current session were
+ not redone if the query was re-executed with a
+ new <varname>search_path</> setting, resulting in surprising behavior.
</para>
</listitem>
<listitem>
<para>
- Store <link linkend="wal"><acronym>WAL</></link> in a continuous
- stream, rather than skipping the last 16MB segment every 4GB
- (Heikki Linnakangas)
+ Fix <link
+ linkend="functions-formatting-table"><function>to_number()</></link>
+ to properly handle a period used as a thousands separator (Tom Lane)
</para>
<para>
- Previously, <acronym>WAL</> files with names ending in <literal>FF</>
- were not used. If you have <acronym>WAL</> backup or restore scripts
- that took that skipping into account, they will need to be adjusted.
+ Previously, a period was considered to be a decimal point even when
+ the locale says it isn't and the <literal>D</> format code is used to
+ specify use of the locale-specific decimal point. This resulted in
+ wrong answers if <literal>FM</> format was also used.
</para>
</listitem>
<listitem>
<para>
- Allow <link
- linkend="functions-formatting-table"><function>to_char()</></link>
- to properly handle <literal>D</> (locale-specific decimal point) and
- <literal>FM</> (fill mode) specifications in locales where a
- period is a group separator and not a decimal point (Tom Lane)
+ Fix <literal>STRICT</> non-set-returning functions that have
+ set-returning functions in their arguments to properly return null
+ rows (Tom Lane)
</para>
<para>
- Previously, a period group separator would be misinterpreted as
- a decimal point in such locales.
+ A null value passed to the strict function should result in a null
+ output, but instead, that output row was suppressed entirely.
</para>
</listitem>
<listitem>
<para>
- Fix <literal>STRICT</> non-set-returning functions that take
- set-returning functions as arguments to properly return null
- rows (Tom Lane)
+ Store <link linkend="wal"><acronym>WAL</></link> in a continuous
+ stream, rather than skipping the last 16MB segment every 4GB
+ (Heikki Linnakangas)
</para>
<para>
- Previously, rows with null values were suppressed.
+ Previously, <acronym>WAL</> files with names ending in <literal>FF</>
+ were not used because of this skipping. If you have <acronym>WAL</>
+ backup or restore scripts that took this behavior into account, they
+ will need to be adjusted.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ In <link
+ linkend="catalog-pg-constraint"><structname>pg_constraint.confmatchtype</></link>,
+ store the default foreign key match type (non-<literal>FULL</>,
+ non-<literal>PARTIAL</>) as <literal>s</> for <quote>simple</>
+ (Tom Lane)
+ </para>
+
+ <para>
+ Previously this case was represented by <literal>u</>
+ for <quote>unspecified</>.
</para>
</listitem>
<listitem>
<para>
- Prevent non-key-field row updates from locking foreign key rows
+ Prevent non-key-field row updates from blocking foreign key checks
(Álvaro Herrera, Noah Misch, Andres Freund, Alexander
- Shulgin, Marti Raudsepp)
+ Shulgin, Marti Raudsepp, Alexander Shulgin)
</para>
<para>
- This improves concurrency and reduces the probability of deadlocks.
- <command>UPDATE</>s on non-key columns use the new <command>SELECT
- FOR NO KEY UPDATE</> lock type, and foreign key checks use the
- new <command>SELECT FOR KEY SHARE</> lock mode.
+ This change improves concurrency and reduces the probability of
+ deadlocks when updating tables involved in a foreign-key constraint.
+ <command>UPDATE</>s that do not change any columns referenced in a
+ foreign key now take the new <literal>NO KEY UPDATE</> lock mode on
+ the row, while foreign key checks use the new <literal>KEY SHARE</>
+ lock mode, which does not conflict with <literal>NO KEY UPDATE</>.
+ So there is no blocking unless a foreign-key column is changed.
</para>
</listitem>
<listitem>
<para>
Add configuration variable <link
- linkend="guc-lock-timeout"><varname>lock_timeout</></link> to limit
- lock wait duration (Zoltán Böszörményi)
- </para>
- </listitem>
-
- <listitem>
- <para>
- Add cache of local locks (Jeff Janes)
- </para>
-
- <para>
- This speeds lock release at statement completion in
- transactions that hold many locks; it is particularly useful
- for <application>pg_dump</> and the restoration of such dumps.
+ linkend="guc-lock-timeout"><varname>lock_timeout</></link> to
+ allow limiting how long a session will wait to acquire any one lock
+ (Zoltán Böszörményi)
</para>
</listitem>
<listitem>
<para>
- Add <link linkend="rangetypes-indexing"><type>SP-GiST</></link>
+ Add <link linkend="rangetypes-indexing"><acronym>SP-GiST</></link>
support for range data types (Alexander Korotkov)
</para>
</listitem>
<listitem>
<para>
- Allow unlogged <link linkend="SPGiST"><type>GiST</></link> indexes
- (Jeevan Chalke)
+ Allow <link linkend="GiST"><acronym>GiST</></link> indexes to be
+ unlogged (Jeevan Chalke)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Improve performance of <acronym>GiST</> index insertion by randomizing
+ the choice of which page to descend to when there are multiple equally
+ good alternatives (Heikki Linnakangas)
</para>
</listitem>
<listitem>
<para>
- Improve concurrency of hash indexes (Robert Haas)
+ Improve concurrency of hash index operations (Robert Haas)
</para>
</listitem>
<listitem>
<para>
- Collect and use histograms for <link linkend="rangetypes">range
- types</link> (Alexander Korotkov)
+ Collect and use histograms of upper and lower bounds, as well as range
+ lengths, for <link linkend="rangetypes">range types</link>
+ (Alexander Korotkov)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Improve optimizer's cost estimation for index access (Tom Lane)
</para>
</listitem>
<listitem>
<para>
- Reduce optimizer overhead by discarding plans with unneeded cheaper
- startup costs (Tom Lane)
+ Improve optimizer's hash table size estimate for
+ doing <literal>DISTINCT</> via hash aggregation (Tom Lane)
</para>
</listitem>
<listitem>
<para>
- Improve optimizer cost estimation for index access (Tom Lane)
+ Suppress no-op Result and Limit plan nodes
+ (Kyotaro Horiguchi, Amit Kapila, Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Reduce optimizer overhead by not keeping plans on the basis of cheap
+ startup cost when the optimizer only cares about total cost overall
+ (Tom Lane)
</para>
</listitem>
<listitem>
<para>
Add <link linkend="SQL-COPY"><command>COPY FREEZE</></link>
- option to avoid the overhead of marking tuples as committed later
+ option to avoid the overhead of marking tuples as frozen later
(Simon Riggs, Jeff Davis)
</para>
</listitem>
<listitem>
<para>
- Improve grouping of sessions waiting for <link
+ Improve synchronization of sessions waiting for <link
linkend="guc-commit-delay"><varname>commit_delay</></link>
(Peter Geoghegan)
</para>
<para>
- This improves the usefulness and behavior of
- <varname>commit_delay</>.
+ This greatly improves the usefulness of <varname>commit_delay</>.
</para>
</listitem>
<listitem>
<para>
- Improve performance for transactions creating, rebuilding, or
- dropping many relations (Jeff Janes, Tomas Vondra)
+ Improve performance of the <link
+ linkend="SQL-CREATETABLE"><command>CREATE TEMPORARY TABLE ... ON
+ COMMIT DELETE ROWS</></link> option by not truncating such temporary
+ tables in transactions that haven't touched any temporary tables
+ (Heikki Linnakangas)
</para>
</listitem>
<listitem>
<para>
- Improve performance of the <link
- linkend="SQL-CREATETABLE"><command>CREATE TEMPORARY TABLE ... ON
- COMMIT DELETE ROWS</></link> clause by only issuing delete if
- the temporary table was accessed (Heikki Linnakangas)
+ Make vacuum recheck visibility after it has removed expired tuples
+ (Pavan Deolasee)
+ </para>
+
+ <para>
+ This increases the chance of a page being marked as all-visible.
</para>
</listitem>
<listitem>
<para>
- Have vacuum recheck visibility after it has removed expired tuples
- (Pavan Deolasee)
+ Add per-resource-owner lock caches (Jeff Janes)
</para>
<para>
- This increases the chance of a page being marked as all-visible.
+ This speeds up lock bookkeeping at statement completion in
+ multi-statement transactions that hold many locks; it is particularly
+ useful for <application>pg_dump</>.
</para>
</listitem>
<listitem>
<para>
- Split the <link
- linkend="guc-stats-temp-directory"><filename>pg_stat_tmp</></link>
- statistics file into per-database and global files (Tomas Vondra)
+ Avoid scanning the entire relation cache at commit of a transaction
+ that creates a new relation (Jeff Janes)
+ </para>
+
+ <para>
+ This speeds up sessions that create many tables in successive
+ small transactions, such as a <application>pg_restore</> run.
</para>
+ </listitem>
+ <listitem>
<para>
- This reduces the I/O overhead for statistics tracking.
+ Improve performance of transactions that drop many relations
+ (Tomas Vondra)
</para>
</listitem>
<listitem>
<para>
- Allow <link
- linkend="functions-admin-signal-table"><function>pg_terminate_backend()</></link>
- to terminate other backends with the same role (Dan Farina)
+ Split the <link linkend="monitoring-stats">statistics collector's</link>
+ data file into separate global and per-database files (Tomas Vondra)
+ </para>
+
+ <para>
+ This reduces the I/O required for statistics tracking.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Fix the statistics collector to operate properly in cases where the
+ system clock goes backwards (Tom Lane)
</para>
<para>
- Previously, only superusers could terminate other sessions.
+ Previously, statistics collection would stop until the time again
+ reached the latest time previously recorded.
</para>
</listitem>
<listitem>
<para>
- Allow the <link
- linkend="runtime-config-statistics-collector">statistics
- collector</link> to operate properly in cases where the system
- clock goes backwards (Tom Lane)
+ Emit an informative message to postmaster standard error when we
+ are about to stop logging there
+ (Tom Lane)
</para>
<para>
- Previously statistics collection would stop until the time again
- reached the previously-stored latest time.
+ This should help reduce user confusion about where to look for log
+ output in common configurations that log to standard error only during
+ postmaster startup.
</para>
</listitem>
<itemizedlist>
+ <listitem>
+ <para>
+ When an authentication failure occurs, log the relevant
+ <link linkend="auth-pg-hba-conf"><filename>pg_hba.conf</></link>
+ line, to ease debugging of unintended failures
+ (Magnus Hagander)
+ </para>
+ </listitem>
+
<listitem>
<para>
Improve <link linkend="auth-ldap"><acronym>LDAP</></link> error
<listitem>
<para>
- Add support for <acronym>LDAP</> authentication to be specified
- in <acronym>URL</> format (Peter Eisentraut)
+ Add support for specifying <acronym>LDAP</> authentication parameters
+ in <acronym>URL</> format, per RFC 4516 (Peter Eisentraut)
</para>
</listitem>
</para>
<para>
- It is assumed <literal>DEFAULT</> is more appropriate cipher set.
+ This should yield a more appropriate SSL cipher set.
</para>
</listitem>
</para>
<para>
- This is similar to how <link
- linkend="auth-pg-hba-conf"><filename>pg_hba.conf</></link>
- is processed.
+ This is similar to how <filename>pg_hba.conf</> is processed.
</para>
</listitem>
<listitem>
<para>
- Dramatically reduce System V <link linkend="sysvipc">shared
+ Greatly reduce System V <link linkend="sysvipc">shared
memory</link> requirements (Robert Haas)
</para>
<para>
- Instead, on Unix-like systems, <function>mmap()</> is used for
- shared memory. For most users, this will eliminate the need to
- adjust kernel parameters for shared memory.
+ On Unix-like systems, <function>mmap()</> is now used for most
+ of <productname>PostgreSQL</>'s shared memory. For most users, this
+ will eliminate any need to adjust kernel parameters for shared memory.
</para>
</listitem>
<listitem>
<para>
Remove the <link linkend="guc-external-pid-file">external
- <acronym>PID</> file</link> on postmaster exit (Peter Eisentraut)
+ <acronym>PID</> file</link>, if any, on postmaster exit
+ (Peter Eisentraut)
</para>
</listitem>
</para>
<para>
- This allows streaming standbys to feed from newly-promoted slaves.
- Previously slaves required access to a <acronym>WAL</> archive directory to
- accomplish this.
+ This allows streaming standby servers to receive WAL data from a slave
+ newly promoted to master status. Previously, other standbys would
+ require a resync to begin following the new master.
</para>
</listitem>
</para>
<para>
- This information is useful for determining
the <acronym>WAL</>
- files needed for restore.
+ This information is useful for determining
which <acronym>WAL</>
+ files are needed for restore.
</para>
</listitem>
<listitem>
<para>
- Have <link
+ Make <link
linkend="app-pgbasebackup"><application>pg_basebackup</></link>
<option>--write-recovery-conf</> output a
- minimal <filename>recovery.conf</> (Zoltán
+ minimal <filename>recovery.conf</> file (Zoltán
Böszörményi, Magnus Hagander)
</para>
<para>
Add <link
linkend="guc-wal-receiver-timeout"><varname>wal_receiver_timeout</></link>
- parameter to control the <acronym>WAL</> receiver timeout
+ parameter to control the <acronym>WAL</> receiver's timeout
(Amit Kapila)
</para>
</para>
</listitem>
- </itemizedlist>
-
- <sect4>
- <title><link linkend="wal">Write-Ahead Log</link>
- (<acronym>WAL</>)</title>
-
- <itemizedlist>
-
<listitem>
<para>
- Change the <acronym>WAL</> record format to allow splitting the record header
- across pages (Heikki Linnakangas)
+ Change the <link linkend="wal"><acronym>WAL</></link> record format to
+ allow splitting the record header across pages (Heikki Linnakangas)
</para>
<para>
</itemizedlist>
- </sect4>
-
</sect3>
<sect3>
Add support for piping <link
linkend="SQL-COPY"><command>COPY</></link> and <link
linkend="APP-PSQL"><application>psql</></link> <command>\copy</>
- to/from an external program (Etsuro Fujita)
+ data to/from an external program (Etsuro Fujita)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Allow a multirow <link
+ linkend="SQL-VALUES"><literal>VALUES</></link> clause in a rule
+ to reference <literal>OLD</>/<literal>NEW</> (Tom Lane)
</para>
</listitem>
<para>
This allows server-side functions written in event-enabled
- languages, e.g. C, PL/pgSQL, to be called when DDL commands
- are run.
+ languages to be called when DDL commands are run.
</para>
</listitem>
</para>
</listitem>
- <listitem>
- <para>
- Allow a multirow <link
- linkend="SQL-VALUES"><literal>VALUES</></link> clause in a rule
- to reference <literal>OLD</>/<literal>NEW</> (Tom Lane)
- </para>
- </listitem>
-
<listitem>
<para>
Add <link linkend="SQL-CREATESCHEMA"><command>CREATE SCHEMA ... IF
<listitem>
<para>
- Have <link linkend="SQL-REASSIGN-OWNED"><command>REASSIGN
+ Make <link linkend="SQL-REASSIGN-OWNED"><command>REASSIGN
OWNED</></link> also change ownership of shared objects
(Álvaro Herrera)
</para>
</listitem>
- </itemizedlist>
-
- <sect4>
- <title><link linkend="SQL-CREATETABLE"><command>CREATE TABLE</></link></title>
-
- <itemizedlist>
+ <listitem>
+ <para>
+ Make <link linkend="sql-createaggregate"><command>CREATE
+ AGGREGATE</></link> complain if the given initial value string is not
+ valid input for the transition datatype (Tom Lane)
+ </para>
+ </listitem>
<listitem>
<para>
- Suppress messages about implicit index and sequence creation
+ Suppress <link linkend="SQL-CREATETABLE"><command>CREATE
+ TABLE</></link>'s messages about implicit index and sequence creation
(Robert Haas)
</para>
</para>
</listitem>
- </itemizedlist>
-
- </sect4>
-
- <sect4>
- <title>Constraints</title>
-
- <itemizedlist>
-
<listitem>
<para>
Provide clients with <link
</para>
<para>
- This allows clients to retrieve table, column, data type, or constraint
- name error details. Previously such information had to be extracted from
- error strings. Client library support is required to access these
- fields.
+ This allows clients to retrieve table, column, data type, or
+ constraint name error details. Previously such information had to be
+ extracted from error strings. Client library support is required to
+ access these fields.
</para>
</listitem>
</itemizedlist>
- </sect4>
-
<sect4>
<title><command>ALTER</></title>
<listitem>
<para>
Add <link linkend="SQL-ALTERROLE"><command>ALTER ROLE ALL
- SET</></link> to add settings to all users (Peter Eisentraut)
+ SET</></link> to establish settings for all users (Peter Eisentraut)
</para>
<para>
<listitem>
<para>
- Improve view/rule printing code to handle cases where referenced
- tables are renamed, or columns are renamed, added, or dropped
- (Tom Lane)
+ Add <link linkend="SQL-CREATEVIEW"><command>CREATE RECURSIVE
+ VIEW</></link> syntax (Peter Eisentraut)
</para>
<para>
- Table and column renamings can produce cases where, if we merely
- substitute the new name into the original text of a rule or view, the
- result is ambiguous. This patch fixes the rule-dumping code to insert
- table and column aliases if needed to preserve the original semantics.
+ Internally this is translated into <command>CREATE VIEW ... WITH
+ RECURSIVE ...</>.
</para>
</listitem>
<listitem>
<para>
- Add <link linkend="SQL-CREATEVIEW"><command>CREATE RECURSIVE
- VIEW</></link> syntax (Peter Eisentraut)
+ Improve view/rule printing code to handle cases where referenced
+ tables are renamed, or columns are renamed, added, or dropped
+ (Tom Lane)
</para>
<para>
- Internally this is translated into <command>CREATE VIEW ... WITH
- RECURSIVE ...</>.
+ Table and column renamings can produce cases where, if we merely
+ substitute the new name into the original text of a rule or view, the
+ result is ambiguous. This change fixes the rule-dumping code to insert
+ manufactured table and column aliases when needed to preserve the
+ original semantics.
</para>
</listitem>
<listitem>
<para>
- Increase the maximum length of <link linkend="lo-open">large
+ Increase the maximum size of <link linkend="largeObjects">large
objects</link> from 2GB to 4TB (Nozomi Anzai, Yugo Nagata)
</para>
<para>
- This change includes new libpq and server-side 64-bit-capable
- large object access functions.
+ This change includes adding 64-bit-capable large object access
+ functions, both in the server and in libpq.
</para>
</listitem>
<listitem>
<para>
Allow text <link linkend="datatype-timezones">timezone
- designations</link>, e.g. <quote>America/Chicago</> when using
- the <acronym>ISO</> <quote>T</> <type>timestamptz</type> format (Bruce Momjian)
+ designations</link>, e.g. <quote>America/Chicago</>, in the
+ <quote>T</> field of <acronym>ISO</>-format <type>timestamptz</type>
+ input (Bruce Momjian)
</para>
</listitem>
<listitem>
<para>
Add <link linkend="functions-json">operators and functions</link>
- to extract values from <type>JSON</> data strings (Andrew Dunstan)
+ to extract elements from <type>JSON</> values (Andrew Dunstan)
</para>
</listitem>
<listitem>
<para>
- Allow <type>JSON</> data strings to be <link
+ Allow <type>JSON</> values to be <link
linkend="functions-json-table">converted into records</link>
(Andrew Dunstan)
</para>
<listitem>
<para>
- Add <link linkend="functions-json-table">functions</link>
- to convert values, records, and hstore data to <type>JSON</>
- (Andrew Dunstan)
+ Add <link linkend="functions-json-table">functions</link> to convert
+ scalars, records, and <type>hstore</> values to <type>JSON</> (Andrew
+ Dunstan)
</para>
</listitem>
<para>
Improve <link
linkend="functions-string-format"><function>format()</></link>
- to handle field width and left/right alignment (Pavel Stehule)
+ to provide field width and left/right alignment options (Pavel Stehule)
</para>
</listitem>
<listitem>
<para>
- Have <link
+ Make <link
linkend="functions-formatting-table"><function>to_char()</></link>,
<link
linkend="functions-formatting-table"><function>to_date()</></link>,
and <link
linkend="functions-formatting-table"><function>to_timestamp()</></link>
- properly handle negative century designations (<literal>CC</>)
+ handle negative (BC) century values properly
(Bruce Momjian)
</para>
<para>
Previously the behavior was either wrong or inconsistent
- with positive/<acronym>AD</> handling, e.g. format mask
+ with positive/<acronym>AD</> handling, e.g. with the format mask
<quote>IYYY-IW-DY</>.
</para>
</listitem>
<listitem>
<para>
- Have <link
+ Make <link
linkend="functions-formatting-table"><function>to_date()</></link>
and <link
linkend="functions-formatting-table"><function>to_timestamp()</></link>
<para>
Cause <link
linkend="functions-info-catalog-table"><function>pg_get_viewdef()</></link>
- to start a new line by default after each <link
- linkend="SQL-SELECT"><command>SELECT</></link> target list entry and
- <literal>FROM</> entry (Marko Tiikkaja)
+ to start a new line by default after each <literal>SELECT</> target
+ list entry and <literal>FROM</> entry (Marko Tiikkaja)
</para>
<para>
</para>
</listitem>
- <listitem>
- <para>
- Force cached plans to be replanned if the <link
- linkend="guc-search-path"><varname>search_path</></link> changes
- (Tom Lane)
- </para>
-
- <para>
- Previously cached plans already generated in the current session
- ignored <varname>search_path</> changes.
- </para>
- </listitem>
-
</itemizedlist>
</sect3>
<sect3>
<title>Server-Side Languages</title>
- <itemizedlist>
-
- <listitem>
- <para>
- Allow <link linkend="spi-spi-execute"><acronym>SPI</></link>
- functions to access the number of rows processed by
- <link linkend="SQL-COPY"><command>COPY</></link> (Pavel Stehule)
- </para>
- </listitem>
-
- </itemizedlist>
-
<sect4>
<title><link linkend="plpgsql">PL/pgSQL</link> Server-Side Language</title>
</para>
<para>
- The command is <link
+ A <command>COPY</> executed in a PL/pgSQL function now updates the
+ value retrieved by <link
linkend="plpgsql-statements-diagnostics"><command>GET DIAGNOSTICS
x = ROW_COUNT</></link>.
</para>
<listitem>
<para>
- Allow greater flexibility in where keywords can be used in PL/pgSQL (Tom Lane)
+ Allow unreserved keywords to be used as identifiers everywhere in
+ PL/pgSQL (Tom Lane)
+ </para>
+
+ <para>
+ In certain places in the PL/pgSQL grammar, keywords had to be quoted
+ to be used as identifiers, even if they were nominally unreserved.
</para>
</listitem>
</sect3>
+ <sect3>
+ <title>Server Programming Interface (<link linkend="spi">SPI</link>)</title>
+
+ <itemizedlist>
+
+ <listitem>
+ <para>
+ Prevent leakage of <acronym>SPI</> tuple tables during subtransaction
+ abort (Tom Lane)
+ </para>
+
+ <para>
+ At the end of any failed subtransaction, the core SPI code now
+ releases any SPI tuple tables that were created during that
+ subtransaction. This avoids the need for SPI-using code to keep track
+ of such tuple tables and release them manually in error-recovery code.
+ Failure to do so caused a number of transaction-lifespan memory leakage
+ issues in PL/pgSQL and perhaps other SPI clients. <link
+ linkend="spi-spi-freetupletable"><function>SPI_freetuptable()</></link>
+ now protects itself against multiple freeing requests, so any existing
+ code that did take care to clean up shouldn't be broken by this change.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Allow <acronym>SPI</> functions to access the number of rows processed
+ by <link linkend="SQL-COPY"><command>COPY</></link> (Pavel Stehule)
+ </para>
+ </listitem>
+
+ </itemizedlist>
+
+ </sect3>
+
<sect3>
<title>Client Applications</title>
</para>
<para>
- This is similar to the <link
- linkend="APP-PGDUMP"><application>pg_dump</></link> <option>--table</>
- option.
+ This is similar to the way <link
+ linkend="APP-PGDUMP"><application>pg_dump</></link>'s
+ <option>--table</> option works.
</para>
</listitem>
linkend="app-pgbasebackup"><application>pg_basebackup</></link>, and
<link
linkend="app-pgreceivexlog"><application>pg_receivexlog</></link>
- to specify the connection string (Amit Kapila)
+ to allow specifying a connection string (Amit Kapila)
</para>
</listitem>
<listitem>
<para>
Adjust function cost settings so <application>psql</> tab
- completion and pattern searching is more efficient (Tom Lane)
+ completion and pattern searching are more efficient (Tom Lane)
</para>
</listitem>
<listitem>
<para>
- Improve <application>psql</> tab completion coverage (Jeff Janes,
- Peter Eisentraut)
+ Improve <application>psql</>'s tab completion coverage (Jeff Janes,
+ Dean Rasheed, Peter Eisentraut, Magnus Hagander)
</para>
</listitem>
</para>
<para>
- The warning when connecting to a newer server was retained.
+ A warning is still issued when connecting to a server of a newer major
+ version than <application>psql</>'s.
</para>
</listitem>
</itemizedlist>
<sect5>
- <title><link linkend="R2-APP-PSQL-4">Backslash Commands</link></title>
+ <title><link linkend="APP-PSQL-meta-commands">Backslash Commands</link></title>
<itemizedlist>
<listitem>
<para>
- Add <application>psql</> <command>\watch</> command to repeatedly
- execute commands (Will Leinweber)
+ Add <application>psql</> command <command>\watch</> to repeatedly
+ execute a SQL command (Will Leinweber)
</para>
</listitem>
<listitem>
<para>
- Add <quote>Security</> label to <application>psql</> <command>\df+</>
- output (Jon Erdman)
+ Add <quote>Security</> column to <application>psql</>'s
+ <command>\df+</> output (Jon Erdman)
</para>
</listitem>
<listitem>
<para>
- Allow <application>psql</> <command>\l</> to accept a database
+ Allow <application>psql</> command <command>\l</> to accept a database
name pattern (Peter Eisentraut)
</para>
</listitem>
<listitem>
<para>
- Properly reset state if the SQL command executed with
- <application>psql</>'s <quote>\g file</> fails (Tom Lane)
+ Properly reset state after failure of a SQL command executed with
+ <application>psql</>'s <literal>\g</> <replaceable>file</>
+ (Tom Lane)
</para>
<para>
<listitem>
<para>
- In <application>psql</> tuples-only and expanded modes, no longer
- output <quote>(No rows)</> (Peter Eisentraut)
+ In <application>psql</>'s tuples-only and expanded output modes, no
+ longer emit <quote>(No rows)</> for zero rows (Peter Eisentraut)
</para>
</listitem>
<listitem>
<para>
- In <application>psql</>, no longer print an empty line for
- unaligned, expanded output for zero rows (Peter Eisentraut)
+ In <application>psql</>'s unaligned, expanded output mode, no longer
+ print an empty line for zero rows (Peter Eisentraut)
</para>
</listitem>
<listitem>
<para>
-
Have <application>pg_dump</> output functions in a more predictable
+
Make <application>pg_dump</> output functions in a more predictable
order (Joel Jacobson)
</para>
</listitem>
<listitem>
<para>
- Fix tar files emitted by <application>pg_dump</> and <link
- linkend="app-pgbasebackup"><application>pg_basebackup</></link>
+ Fix tar files emitted by <application>pg_dump</>
to be <acronym>POSIX</> conformant (Brian Weaver, Tom Lane)
</para>
</listitem>
<listitem>
<para>
- Have initdb fsync the newly created data directory (Jeff Davis)
+ Make initdb fsync the newly created data directory (Jeff Davis)
</para>
<para>
- This can be disabled by using <option>--nosync</>.
+ This insures data integrity in event of a system crash shortly after
+ initdb. This can be disabled by using <option>--nosync</>.
</para>
</listitem>
<listitem>
<para>
- Have initdb issue a warning about placing the data directory at the
+ Make initdb issue a warning about placing the data directory at the
top of a file system mount point (Bruce Momjian)
</para>
</listitem>
<listitem>
<para>
- Add an embedded list interface (Andres Freund)
- </para>
- </listitem>
-
- <listitem>
- <para>
- Add infrastructure to better support plug-in <link
+ Add infrastructure to allow plug-in <link
linkend="bgworker">background worker processes</link>
(Álvaro Herrera)
</para>
</para>
<para>
- This allows libpgport to be used solely for porting code.
+ This allows libpgport to be used solely for portability-related code.
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Add support for list links embedded in larger structs (Andres Freund)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Use <literal>SA_RESTART</> for all signals,
+ including <literal>SIGALRM</> (Tom Lane)
</para>
</listitem>
<listitem>
<para>
- Standardize on naming of client-side memory allocation functions (Tom Lane)
+ Ensure that the correct text domain is used when
+ translating <function>errcontext()</> messages
+ (Heikki Linnakangas)
</para>
</listitem>
<listitem>
<para>
- Add compiler designations to indicate some <function>ereport()</>
+ Standardize naming of client-side memory allocation functions (Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Provide support for <quote>static assertions</> that will fail at
+ compile time if some compile-time-constant condition is not met
+ (Andres Freund, Tom Lane)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Support <function>Assert()</> in client-side code (Andrew Dunstan)
+ </para>
+ </listitem>
+
+ <listitem>
+ <para>
+ Add decoration to inform the C compiler that some <function>ereport()</>
and <function>elog()</> calls do not return (Peter Eisentraut,
Andres Freund, Tom Lane, Heikki Linnakangas)
</para>
<listitem>
<para>
Remove <link linkend="configure">configure</link> flag
- <option>--disable-shared</>, as it is no longer used (Bruce
- Momjian)
+ <option>--disable-shared</>, as it is no longer supported
+ (Bruce Momjian)
</para>
</listitem>
<listitem>
<para>
- Add Emacs macro to match <productname>PostgreSQL</> perltidy
- formatting (Peter Eisentraut)
+ Provide Emacs macro to set Perl formatting to
+ match <productname>PostgreSQL</>'s perltidy settings (Peter Eisentraut)
</para>
</listitem>
<listitem>
<para>
- Run tool to check the keyword list when the backend grammar is
+ Run tool to check the keyword list whenever the backend grammar is
changed (Tom Lane)
</para>
</listitem>
<listitem>
<para>
- Centralize <application>flex</> and <application>bison</>
- <application>make</> rules (Peter Eisentraut)
+ Change the way <literal>UESCAPE</> is lexed, to significantly reduce
+ the size of the lexer tables (Heikki Linnakangas)
</para>
+ </listitem>
+ <listitem>
<para>
- This is useful for <application>pgxs</> authors.
+ Centralize <application>flex</> and <application>bison</>
+ <application>make</> rules (Peter Eisentraut)
</para>
- </listitem>
- <listitem>
<para>
- Support <function>Assert()</> in client-side code (Andrew Dunstan)
+ This is useful for <application>pgxs</> authors.
</para>
</listitem>
<listitem>
<para>
- Change many internal backend functions to return <type>OID</>s
+ Change many internal backend functions to return object <type>OID</>s
rather than void (Dimitri Fontaine)
</para>
<para>
Add function <link
linkend="functions-info-catalog-table"><function>pg_identify_object()</></link>
- to dump an object in machine-readable format (Álvaro
- Herrera)
+ to produce a machine-readable description of a database object
+ (Álvaro Herrera)
</para>
</listitem>
<listitem>
<para>
- Improve ability to detect official timezone abbreviation changes
+ Provide a tool to help detect timezone abbreviation changes when
+ updating the <filename>src/timezone/data</> files
(Tom Lane)
</para>
</listitem>
<listitem>
<para>
- Add <application>pkg-config</> support <application>libpq</>
+ Add <application>pkg-config</> support
for <application>libpq</>
and <application>ecpg</> libraries (Peter Eisentraut)
</para>
</listitem>
<listitem>
<para>
- Add a <productname>Postgres</> <link linkend="postgres-fdw">foreign
- data wrapper</link> contrib module (Shigeru Hanada)
+ Add a <link linkend="postgres-fdw"><productname>Postgres</> foreign
+ data wrapper</link> contrib module to allow access to
+ other <productname>Postgres</> servers (Shigeru Hanada)
</para>
<para>
<listitem>
<para>
- Improve <link linkend="pgtrgm"><productname>pg_trgm</></link>
+ Improve <link linkend="pgtrgm"><productname>pg_trgm</></link>'s
handling of multibyte characters (Tom Lane)
</para>
+
+ <para>
+ On a platform that does not have the wcstombs() or towlower() library
+ functions, this could result in an incompatible change in the contents
+ of <productname>pg_trgm</> indexes for non-ASCII data. In such cases,
+ <command>REINDEX</> those indexes to ensure correct search results.
+ </para>
</listitem>
<listitem>
<para>
- Add <link linkend="pgstattuple">pgstattuple function</link> to report the
- size of the <type>GIN</> pending index insertion list (Fujii Masao)
+ Add a <link linkend="pgstattuple">pgstattuple</link> function to report
+ the size of the pending-insertions list of a <acronym>GIN</> index
+ (Fujii Masao)
</para>
</listitem>
<listitem>
<para>
-
Have <link linkend="oid2name"><application>oid2name</></link>,
+
Make <link linkend="oid2name"><application>oid2name</></link>,
<link linkend="pgbench"><application>pgbench</></link>, and
<link linkend="vacuumlo"><application>vacuumlo</></link> set
<varname>fallback_application_name</> (Amit Kapila)
<listitem>
<para>
- Improve <link linkend="dblink">dblink</link> option validator
- (Tom Lane)
+ Create a dedicated foreign data wrapper, with its own option validator
+ function, for <link linkend="dblink">dblink</link> (Shigeru Hanada)
</para>
<para>
- Details?
+ When using this FDW to define the target of a <application>dblink</>
+ connection, instead of using a hard-wired list of connection options,
+ the underlying <application>libpq</> library is consulted to see what
+ connection options it supports.
</para>
</listitem>
<listitem>
<para>
- Allow <application>pg_upgrade</> <option>--jobs</> to do
- parallelism (Bruce Momjian, Andrew Dunstan)
+ Allow <application>pg_upgrade</> to do dumps and restores in
+ parallel (Bruce Momjian, Andrew Dunstan)
</para>
<para>
This allows parallel schema dump/restore of databases, as well as
- parallel copy/link of data files per tablespace.
+ parallel copy/link of data files per tablespace. Use the
+ <option>--jobs</> option to specify the level of parallelism.
</para>
</listitem>
<listitem>
<para>
-
Have <application>pg_upgrade</> create Unix-domain sockets in
+
Make <application>pg_upgrade</> create Unix-domain sockets in
the current directory (Bruce Momjian, Tom Lane)
</para>
<listitem>
<para>
-
Have <application>pg_upgrade</> <option>--check</> mode properly
+
Make <application>pg_upgrade</> <option>--check</> mode properly
detect the location of non-default socket directories (Bruce
Momjian, Tom Lane)
</para>
<listitem>
<para>
- Increase <application>pg_upgrade</> logging content by showing
- executed command (Álvaro Herrera)
+ Improve <application>pg_upgrade</>'s logs by showing
+ executed commands (Álvaro Herrera)
</para>
</listitem>
<listitem>
<para>
- Allow <application>pgbench</> to use a larger scale factor
+ Allow <application>pgbench</> to use much larger scale factors,
+ by changing relevant columns from <type>integer</> to <type>bigint</>
+ when the requested scale factor exceeds 20000
(Greg Smith)
</para>
</listitem>
<listitem>
<para>
Improve <link linkend="tutorial-window"><literal>WINDOW</>
- function</link> documentation (Bruce Momjian, Tom Lane)
+ function</link> documentation (Bruce Momjian, Florian Pflug)
</para>
</listitem>
<listitem>
<para>
- Add <link linkend="docguide">instructions</link> for setting
+ Add <link linkend="docguide-toolsets">instructions</link> for setting
up the documentation tool chain on Mac <productname>OS X</>
(Peter Eisentraut)
</para>
projects / postgresql / commitdiff