<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN" "http://www.w3.org/TR/html4/loose.dtd">
<HTML
><HEAD
><TITLE
>Release 7.4</TITLE
><META
NAME="GENERATOR"
CONTENT="Modular DocBook HTML Stylesheet Version 1.79"><LINK
REV="MADE"
HREF="mailto:pgsql-docs@postgresql.org"><LINK
REL="HOME"
TITLE="PostgreSQL 9.3.9 Documentation"
HREF="index.html"><LINK
REL="UP"
TITLE="Release Notes"
HREF="release.html"><LINK
REL="PREVIOUS"
TITLE="Release 7.4.1"
HREF="release-7-4-1.html"><LINK
REL="NEXT"
TITLE="Release 7.3.21"
HREF="release-7-3-21.html"><LINK
REL="STYLESHEET"
TYPE="text/css"
HREF="stylesheet.css"><META
HTTP-EQUIV="Content-Type"
CONTENT="text/html; charset=ISO-8859-1"><META
NAME="creation"
CONTENT="2015-06-13T20:07:22"></HEAD
><BODY
CLASS="SECT1"
><DIV
CLASS="NAVHEADER"
><TABLE
SUMMARY="Header navigation table"
WIDTH="100%"
BORDER="0"
CELLPADDING="0"
CELLSPACING="0"
><TR
><TH
COLSPAN="5"
ALIGN="center"
VALIGN="bottom"
><A
HREF="index.html"
>PostgreSQL 9.3.9 Documentation</A
></TH
></TR
><TR
><TD
WIDTH="10%"
ALIGN="left"
VALIGN="top"
><A
TITLE="Release 7.4.1"
HREF="release-7-4-1.html"
ACCESSKEY="P"
>Prev</A
></TD
><TD
WIDTH="10%"
ALIGN="left"
VALIGN="top"
><A
HREF="release.html"
ACCESSKEY="U"
>Up</A
></TD
><TD
WIDTH="60%"
ALIGN="center"
VALIGN="bottom"
>Appendix E. Release Notes</TD
><TD
WIDTH="20%"
ALIGN="right"
VALIGN="top"
><A
TITLE="Release 7.3.21"
HREF="release-7-3-21.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
></TABLE
><HR
ALIGN="LEFT"
WIDTH="100%"></DIV
><DIV
CLASS="SECT1"
><H1
CLASS="SECT1"
><A
NAME="RELEASE-7-4"
>E.219. Release 7.4</A
></H1
><DIV
CLASS="NOTE"
><BLOCKQUOTE
CLASS="NOTE"
><P
><B
>Release Date: </B
>2003-11-17</P
></BLOCKQUOTE
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="AEN145729"
>E.219.1. Overview</A
></H2
><P
> Major changes in this release:
</P
><P
></P
><DIV
CLASS="VARIABLELIST"
><DL
><DT
><TT
CLASS="LITERAL"
>IN</TT
> / <TT
CLASS="LITERAL"
>NOT IN</TT
> subqueries are
now much more efficient</DT
><DD
><P
> In previous releases, <TT
CLASS="LITERAL"
>IN</TT
>/<TT
CLASS="LITERAL"
>NOT
IN</TT
> subqueries were joined to the upper query by
sequentially scanning the subquery looking for a match. The
7.4 code uses the same sophisticated techniques used by
ordinary joins and so is much faster. An
<TT
CLASS="LITERAL"
>IN</TT
> will now usually be as fast as or faster
than an equivalent <TT
CLASS="LITERAL"
>EXISTS</TT
> subquery; this
reverses the conventional wisdom that applied to previous
releases.
</P
></DD
><DT
>Improved <TT
CLASS="LITERAL"
>GROUP BY</TT
> processing by using hash buckets</DT
><DD
><P
> In previous releases, rows to be grouped had to be sorted
first. The 7.4 code can do <TT
CLASS="LITERAL"
>GROUP BY</TT
>
without sorting, by accumulating results into a hash table
with one entry per group. It will still use the sort
technique, however, if the hash table is estimated to be too
large to fit in <TT
CLASS="VARNAME"
>sort_mem</TT
>.
</P
></DD
><DT
>New multikey hash join capability</DT
><DD
><P
> In previous releases, hash joins could only occur on single
keys. This release allows multicolumn hash joins.
</P
></DD
><DT
>Queries using the explicit <TT
CLASS="LITERAL"
>JOIN</TT
> syntax are
now better optimized</DT
><DD
><P
> Prior releases evaluated queries using the explicit
<TT
CLASS="LITERAL"
>JOIN</TT
> syntax only in the order implied by
the syntax. 7.4 allows full optimization of these queries,
meaning the optimizer considers all possible join orderings
and chooses the most efficient. Outer joins, however, must
still follow the declared ordering.
</P
></DD
><DT
>Faster and more powerful regular expression code</DT
><DD
><P
> The entire regular expression module has been replaced with a
new version by Henry Spencer, originally written for Tcl. The
code greatly improves performance and supports several flavors
of regular expressions.
</P
></DD
><DT
>Function-inlining for simple SQL functions</DT
><DD
><P
> Simple SQL functions can now be inlined by including their SQL
in the main query. This improves performance by eliminating
per-call overhead. That means simple SQL functions now
behave like macros.
</P
></DD
><DT
>Full support for IPv6 connections and IPv6 address data types</DT
><DD
><P
> Previous releases allowed only IPv4 connections, and the IP
data types only supported IPv4 addresses. This release adds
full IPv6 support in both of these areas.
</P
></DD
><DT
>Major improvements in SSL performance and reliability</DT
><DD
><P
> Several people very familiar with the SSL API have overhauled
our SSL code to improve SSL key negotiation and error
recovery.
</P
></DD
><DT
>Make free space map efficiently reuse empty index pages,
and other free space management improvements</DT
><DD
><P
> In previous releases, B-tree index pages that were left empty
because of deleted rows could only be reused by rows with
index values similar to the rows originally indexed on that
page. In 7.4, <TT
CLASS="COMMAND"
>VACUUM</TT
> records empty index
pages and allows them to be reused for any future index rows.
</P
></DD
><DT
>SQL-standard information schema</DT
><DD
><P
> The information schema provides a standardized and stable way
to access information about the schema objects defined in a
database.
</P
></DD
><DT
>Cursors conform more closely to the SQL standard</DT
><DD
><P
> The commands <TT
CLASS="COMMAND"
>FETCH</TT
> and
<TT
CLASS="COMMAND"
>MOVE</TT
> have been overhauled to conform more
closely to the SQL standard.
</P
></DD
><DT
>Cursors can exist outside transactions</DT
><DD
><P
> These cursors are also called holdable cursors.
</P
></DD
><DT
>New client-to-server protocol</DT
><DD
><P
> The new protocol adds error codes, more status information,
faster startup, better support for binary data transmission,
parameter values separated from SQL commands, prepared
statements available at the protocol level, and cleaner
recovery from <TT
CLASS="COMMAND"
>COPY</TT
> failures. The older
protocol is still supported by both server and clients.
</P
></DD
><DT
><SPAN
CLASS="APPLICATION"
>libpq</SPAN
> and
<SPAN
CLASS="APPLICATION"
>ECPG</SPAN
> applications are now fully
thread-safe</DT
><DD
><P
> While previous <SPAN
CLASS="APPLICATION"
>libpq</SPAN
> releases
already supported threads, this release improves thread safety
by fixing some non-thread-safe code that was used during
database connection startup. The <TT
CLASS="COMMAND"
>configure</TT
>
option <TT
CLASS="OPTION"
>--enable-thread-safety</TT
> must be used to
enable this feature.
</P
></DD
><DT
>New version of full-text indexing</DT
><DD
><P
> A new full-text indexing suite is available in
<TT
CLASS="FILENAME"
>contrib/tsearch2</TT
>.
</P
></DD
><DT
>New autovacuum tool</DT
><DD
><P
> The new autovacuum tool in
<TT
CLASS="FILENAME"
>contrib/autovacuum</TT
> monitors the database
statistics tables for
<TT
CLASS="COMMAND"
>INSERT</TT
>/<TT
CLASS="COMMAND"
>UPDATE</TT
>/<TT
CLASS="COMMAND"
>DELETE</TT
>
activity and automatically vacuums tables when needed.
</P
></DD
><DT
>Array handling has been improved and moved into the server core</DT
><DD
><P
> Many array limitations have been removed, and arrays behave
more like fully-supported data types.
</P
></DD
></DL
></DIV
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="AEN145826"
>E.219.2. Migration to Version 7.4</A
></H2
><P
> A dump/restore using <SPAN
CLASS="APPLICATION"
>pg_dump</SPAN
> is
required for those wishing to migrate data from any previous
release.
</P
><P
> Observe the following incompatibilities:
</P
><P
></P
><UL
><LI
><P
> The server-side autocommit setting was removed and
reimplemented in client applications and languages.
Server-side autocommit was causing too many problems with
languages and applications that wanted to control their own
autocommit behavior, so autocommit was removed from the server
and added to individual client APIs as appropriate.
</P
></LI
><LI
><P
> Error message wording has changed substantially in this
release. Significant effort was invested to make the messages
more consistent and user-oriented. If your applications try to
detect different error conditions by parsing the error message,
you are strongly encouraged to use the new error code facility instead.
</P
></LI
><LI
><P
> Inner joins using the explicit <TT
CLASS="LITERAL"
>JOIN</TT
> syntax
might behave differently because they are now better
optimized.
</P
></LI
><LI
><P
> A number of server configuration parameters have been renamed
for clarity, primarily those related to
logging.
</P
></LI
><LI
><P
> <TT
CLASS="LITERAL"
>FETCH 0</TT
> or <TT
CLASS="LITERAL"
>MOVE 0</TT
> now
does nothing. In prior releases, <TT
CLASS="LITERAL"
>FETCH 0</TT
>
would fetch all remaining rows, and <TT
CLASS="LITERAL"
>MOVE 0</TT
>
would move to the end of the cursor.
</P
></LI
><LI
><P
> <TT
CLASS="COMMAND"
>FETCH</TT
> and <TT
CLASS="COMMAND"
>MOVE</TT
> now return
the actual number of rows fetched/moved, or zero if at the
beginning/end of the cursor. Prior releases would return the
row count passed to the command, not the number of rows
actually fetched or moved.
</P
></LI
><LI
><P
> <TT
CLASS="COMMAND"
>COPY</TT
> now can process files that use
carriage-return or carriage-return/line-feed end-of-line
sequences. Literal carriage-returns and line-feeds are no
longer accepted in data values; use <TT
CLASS="LITERAL"
>\r</TT
> and
<TT
CLASS="LITERAL"
>\n</TT
> instead.
</P
></LI
><LI
><P
> Trailing spaces are now trimmed when converting from type
<TT
CLASS="TYPE"
>char(<TT
CLASS="REPLACEABLE"
><I
>n</I
></TT
>)</TT
> to
<TT
CLASS="TYPE"
>varchar(<TT
CLASS="REPLACEABLE"
><I
>n</I
></TT
>)</TT
> or <TT
CLASS="TYPE"
>text</TT
>.
This is what most people always expected to happen anyway.
</P
></LI
><LI
><P
> The data type <TT
CLASS="TYPE"
>float(<TT
CLASS="REPLACEABLE"
><I
>p</I
></TT
>)</TT
> now
measures <TT
CLASS="REPLACEABLE"
><I
>p</I
></TT
> in binary digits, not decimal
digits. The new behavior follows the SQL standard.
</P
></LI
><LI
><P
> Ambiguous date values now must match the ordering specified by
the <TT
CLASS="VARNAME"
>datestyle</TT
> setting. In prior releases, a
date specification of <TT
CLASS="LITERAL"
>10/20/03</TT
> was interpreted as a
date in October even if <TT
CLASS="VARNAME"
>datestyle</TT
> specified that
the day should be first. 7.4 will throw an error if a date
specification is invalid for the current setting of
<TT
CLASS="VARNAME"
>datestyle</TT
>.
</P
></LI
><LI
><P
> The functions <CODE
CLASS="FUNCTION"
>oidrand</CODE
>,
<CODE
CLASS="FUNCTION"
>oidsrand</CODE
>, and
<CODE
CLASS="FUNCTION"
>userfntest</CODE
> have been removed. These
functions were determined to be no longer useful.
</P
></LI
><LI
><P
> String literals specifying time-varying date/time values, such
as <TT
CLASS="LITERAL"
>'now'</TT
> or <TT
CLASS="LITERAL"
>'today'</TT
> will
no longer work as expected in column default expressions; they
now cause the time of the table creation to be the default, not
the time of the insertion. Functions such as
<CODE
CLASS="FUNCTION"
>now()</CODE
>, <CODE
CLASS="FUNCTION"
>current_timestamp</CODE
>, or
<CODE
CLASS="FUNCTION"
>current_date</CODE
> should be used instead.
</P
><P
> In previous releases, there was special code so that strings
such as <TT
CLASS="LITERAL"
>'now'</TT
> were interpreted at
<TT
CLASS="COMMAND"
>INSERT</TT
> time and not at table creation time, but
this work around didn't cover all cases. Release 7.4 now
requires that defaults be defined properly using functions such
as <CODE
CLASS="FUNCTION"
>now()</CODE
> or <CODE
CLASS="FUNCTION"
>current_timestamp</CODE
>. These
will work in all situations.
</P
></LI
><LI
><P
> The dollar sign (<TT
CLASS="LITERAL"
>$</TT
>) is no longer allowed in
operator names. It can instead be a non-first character in
identifiers. This was done to improve compatibility with other
database systems, and to avoid syntax problems when parameter
placeholders (<TT
CLASS="LITERAL"
>$<TT
CLASS="REPLACEABLE"
><I
>n</I
></TT
></TT
>) are written
adjacent to operators.
</P
></LI
></UL
></DIV
><DIV
CLASS="SECT2"
><H2
CLASS="SECT2"
><A
NAME="AEN145896"
>E.219.3. Changes</A
></H2
><P
> Below you will find a detailed account of the changes between
release 7.4 and the previous major release.
</P
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN145899"
>E.219.3.1. Server Operation Changes</A
></H3
><P
></P
><UL
><LI
><P
> Allow IPv6 server connections (Nigel Kukard, Johan Jordaan,
Bruce, Tom, Kurt Roeckx, Andrew Dunstan)
</P
></LI
><LI
><P
> Fix SSL to handle errors cleanly (Nathan Mueller)
</P
><P
> In prior releases, certain SSL API error reports were not
handled correctly. This release fixes those problems.
</P
></LI
><LI
><P
> SSL protocol security and performance improvements (Sean Chittenden)
</P
><P
> SSL key renegotiation was happening too frequently, causing poor
SSL performance. Also, initial key handling was improved.
</P
></LI
><LI
><P
> Print lock information when a deadlock is detected (Tom)
</P
><P
> This allows easier debugging of deadlock situations.
</P
></LI
><LI
><P
> Update <TT
CLASS="FILENAME"
>/tmp</TT
> socket modification times
regularly to avoid their removal (Tom)
</P
><P
> This should help prevent <TT
CLASS="FILENAME"
>/tmp</TT
> directory
cleaner administration scripts from removing server socket
files.
</P
></LI
><LI
><P
>Enable PAM for Mac OS X (Aaron Hillegass)</P
></LI
><LI
><P
>Make B-tree indexes fully WAL-safe (Tom)</P
><P
> In prior releases, under certain rare cases, a server crash
could cause B-tree indexes to become corrupt. This release
removes those last few rare cases.
</P
></LI
><LI
><P
>Allow B-tree index compaction and empty page reuse (Tom)</P
></LI
><LI
><P
> Fix inconsistent index lookups during split of first root page (Tom)
</P
><P
> In prior releases, when a single-page index split into two
pages, there was a brief period when another database session
could miss seeing an index entry. This release fixes that rare
failure case.
</P
></LI
><LI
><P
>Improve free space map allocation logic (Tom)</P
></LI
><LI
><P
>Preserve free space information between server restarts (Tom)</P
><P
> In prior releases, the free space map was not saved when the
postmaster was stopped, so newly started servers had no free
space information. This release saves the free space map, and
reloads it when the server is restarted.
</P
></LI
><LI
><P
>Add start time to <TT
CLASS="LITERAL"
>pg_stat_activity</TT
> (Neil)</P
></LI
><LI
><P
>New code to detect corrupt disk pages; erase with <TT
CLASS="VARNAME"
>zero_damaged_pages</TT
> (Tom)</P
></LI
><LI
><P
>New client/server protocol: faster, no username length limit, allow clean exit from <TT
CLASS="COMMAND"
>COPY</TT
> (Tom)</P
></LI
><LI
><P
>Add transaction status, table ID, column ID to client/server protocol (Tom)</P
></LI
><LI
><P
>Add binary I/O to client/server protocol (Tom)</P
></LI
><LI
><P
>Remove autocommit server setting; move to client applications (Tom)</P
></LI
><LI
><P
>New error message wording, error codes, and three levels of error detail (Tom, Joe, Peter)</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN145950"
>E.219.3.2. Performance Improvements</A
></H3
><P
></P
><UL
><LI
><P
>Add hashing for <TT
CLASS="LITERAL"
>GROUP BY</TT
> aggregates (Tom)</P
></LI
><LI
><P
>Make nested-loop joins be smarter about multicolumn indexes (Tom)</P
></LI
><LI
><P
>Allow multikey hash joins (Tom)</P
></LI
><LI
><P
>Improve constant folding (Tom)</P
></LI
><LI
><P
>Add ability to inline simple SQL functions (Tom)</P
></LI
><LI
><P
>Reduce memory usage for queries using complex functions (Tom)</P
><P
> In prior releases, functions returning allocated memory would
not free it until the query completed. This release allows the
freeing of function-allocated memory when the function call
completes, reducing the total memory used by functions.
</P
></LI
><LI
><P
>Improve GEQO optimizer performance (Tom)</P
><P
> This release fixes several inefficiencies in the way the GEQO optimizer
manages potential query paths.
</P
></LI
><LI
><P
> Allow <TT
CLASS="LITERAL"
>IN</TT
>/<TT
CLASS="LITERAL"
>NOT IN</TT
> to be handled via hash
tables (Tom)
</P
></LI
><LI
><P
> Improve <TT
CLASS="LITERAL"
>NOT IN (<TT
CLASS="REPLACEABLE"
><I
>subquery</I
></TT
>)</TT
>
performance (Tom)
</P
></LI
><LI
><P
> Allow most <TT
CLASS="LITERAL"
>IN</TT
> subqueries to be processed as
joins (Tom)
</P
></LI
><LI
><P
> Pattern matching operations can use indexes regardless of
locale (Peter)
</P
><P
> There is no way for non-ASCII locales to use the standard
indexes for <TT
CLASS="LITERAL"
>LIKE</TT
> comparisons. This release
adds a way to create a special index for
<TT
CLASS="LITERAL"
>LIKE</TT
>.
</P
></LI
><LI
><P
>Allow the postmaster to preload libraries using <TT
CLASS="VARNAME"
>preload_libraries</TT
> (Joe)</P
><P
> For shared libraries that require a long time to load, this
option is available so the library can be preloaded in the
postmaster and inherited by all database sessions.
</P
></LI
><LI
><P
> Improve optimizer cost computations, particularly for subqueries (Tom)
</P
></LI
><LI
><P
> Avoid sort when subquery <TT
CLASS="LITERAL"
>ORDER BY</TT
> matches upper query (Tom)
</P
></LI
><LI
><P
> Deduce that <TT
CLASS="LITERAL"
>WHERE a.x = b.y AND b.y = 42</TT
> also
means <TT
CLASS="LITERAL"
>a.x = 42</TT
> (Tom)
</P
></LI
><LI
><P
> Allow hash/merge joins on complex joins (Tom)
</P
></LI
><LI
><P
> Allow hash joins for more data types (Tom)
</P
></LI
><LI
><P
> Allow join optimization of explicit inner joins, disable with
<TT
CLASS="VARNAME"
>join_collapse_limit</TT
> (Tom)
</P
></LI
><LI
><P
> Add parameter <TT
CLASS="VARNAME"
>from_collapse_limit</TT
> to control
conversion of subqueries to joins (Tom)
</P
></LI
><LI
><P
> Use faster and more powerful regular expression code from Tcl
(Henry Spencer, Tom)
</P
></LI
><LI
><P
> Use bit-mapped relation sets in the optimizer (Tom)
</P
></LI
><LI
><P
>Improve connection startup time (Tom)</P
><P
> The new client/server protocol requires fewer network packets to
start a database session.
</P
></LI
><LI
><P
> Improve trigger/constraint performance (Stephan)
</P
></LI
><LI
><P
> Improve speed of <TT
CLASS="LITERAL"
>col IN (const, const, const, ...)</TT
> (Tom)
</P
></LI
><LI
><P
> Fix hash indexes which were broken in rare cases (Tom)
</P
></LI
><LI
><P
>Improve hash index concurrency and speed (Tom)</P
><P
> Prior releases suffered from poor hash index performance,
particularly for high concurrency situations. This release fixes
that, and the development group is interested in reports
comparing B-tree and hash index performance.
</P
></LI
><LI
><P
>Align shared buffers on 32-byte boundary for copy speed improvement (Manfred Spraul)</P
><P
> Certain CPU's perform faster data copies when addresses are
32-byte aligned.
</P
></LI
><LI
><P
>Data type <TT
CLASS="TYPE"
>numeric</TT
> reimplemented for better performance (Tom)</P
><P
> <TT
CLASS="TYPE"
>numeric</TT
> used to be stored in base 100. The new code
uses base 10000, for significantly better performance.
</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN146034"
>E.219.3.3. Server Configuration Changes</A
></H3
><P
></P
><UL
><LI
><P
>Rename server parameter <TT
CLASS="VARNAME"
>server_min_messages</TT
> to <TT
CLASS="VARNAME"
>log_min_messages</TT
> (Bruce)</P
><P
> This was done so most parameters that control the server logs
begin with <TT
CLASS="LITERAL"
>log_</TT
>.
</P
></LI
><LI
><P
>Rename <TT
CLASS="VARNAME"
>show_*_stats</TT
> to <TT
CLASS="VARNAME"
>log_*_stats</TT
> (Bruce)</P
></LI
><LI
><P
>Rename <TT
CLASS="VARNAME"
>show_source_port</TT
> to <TT
CLASS="VARNAME"
>log_source_port</TT
> (Bruce)</P
></LI
><LI
><P
>Rename <TT
CLASS="VARNAME"
>hostname_lookup</TT
> to <TT
CLASS="VARNAME"
>log_hostname</TT
> (Bruce)</P
></LI
><LI
><P
>Add <TT
CLASS="VARNAME"
>checkpoint_warning</TT
> to warn of excessive checkpointing (Bruce)</P
><P
> In prior releases, it was difficult to determine if checkpoint
was happening too frequently. This feature adds a warning to the
server logs when excessive checkpointing happens.
</P
></LI
><LI
><P
>New read-only server parameters for localization (Tom)</P
></LI
><LI
><P
> Change debug server log messages to output as <TT
CLASS="LITERAL"
>DEBUG</TT
>
rather than <TT
CLASS="LITERAL"
>LOG</TT
> (Bruce)
</P
></LI
><LI
><P
>Prevent server log variables from being turned off by non-superusers (Bruce)</P
><P
> This is a security feature so non-superusers cannot disable
logging that was enabled by the administrator.
</P
></LI
><LI
><P
> <TT
CLASS="VARNAME"
>log_min_messages</TT
>/<TT
CLASS="VARNAME"
>client_min_messages</TT
> now
controls <TT
CLASS="VARNAME"
>debug_*</TT
> output (Bruce)
</P
><P
> This centralizes client debug information so all debug output
can be sent to either the client or server logs.
</P
></LI
><LI
><P
>Add Mac OS X Rendezvous server support (Chris Campbell)</P
><P
> This allows Mac OS X hosts to query the network for available
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> servers.
</P
></LI
><LI
><P
> Add ability to print only slow statements using
<TT
CLASS="VARNAME"
>log_min_duration_statement</TT
>
(Christopher)
</P
><P
> This is an often requested debugging feature that allows
administrators to see only slow queries in their server logs.
</P
></LI
><LI
><P
>Allow <TT
CLASS="FILENAME"
>pg_hba.conf</TT
> to accept netmasks in CIDR format (Andrew Dunstan)</P
><P
> This allows administrators to merge the host IP address and
netmask fields into a single CIDR field in <TT
CLASS="FILENAME"
>pg_hba.conf</TT
>.
</P
></LI
><LI
><P
>New read-only parameter <TT
CLASS="VARNAME"
>is_superuser</TT
> (Tom)</P
></LI
><LI
><P
>New parameter <TT
CLASS="VARNAME"
>log_error_verbosity</TT
> to control error detail (Tom)</P
><P
> This works with the new error reporting feature to supply
additional error information like hints, file names and line
numbers.
</P
></LI
><LI
><P
><TT
CLASS="LITERAL"
>postgres --describe-config</TT
> now dumps server config variables (Aizaz Ahmed, Peter)</P
><P
> This option is useful for administration tools that need to know
the configuration variable names and their minimums, maximums,
defaults, and descriptions.
</P
></LI
><LI
><P
> Add new columns in <TT
CLASS="LITERAL"
>pg_settings</TT
>:
<TT
CLASS="LITERAL"
>context</TT
>, <TT
CLASS="LITERAL"
>type</TT
>, <TT
CLASS="LITERAL"
>source</TT
>,
<TT
CLASS="LITERAL"
>min_val</TT
>, <TT
CLASS="LITERAL"
>max_val</TT
> (Joe)
</P
></LI
><LI
><P
> Make default <TT
CLASS="VARNAME"
>shared_buffers</TT
> 1000 and
<TT
CLASS="VARNAME"
>max_connections</TT
> 100, if possible (Tom)
</P
><P
> Prior versions defaulted to 64 shared buffers so <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
>
would start on even very old systems. This release tests the
amount of shared memory allowed by the platform and selects more
reasonable default values if possible. Of course, users are
still encouraged to evaluate their resource load and size
<TT
CLASS="VARNAME"
>shared_buffers</TT
> accordingly.
</P
></LI
><LI
><P
> New <TT
CLASS="FILENAME"
>pg_hba.conf</TT
> record type
<TT
CLASS="LITERAL"
>hostnossl</TT
> to prevent SSL connections (Jon
Jensen)
</P
><P
> In prior releases, there was no way to prevent SSL connections
if both the client and server supported SSL. This option allows
that capability.
</P
></LI
><LI
><P
> Remove parameter <TT
CLASS="VARNAME"
>geqo_random_seed</TT
>
(Tom)
</P
></LI
><LI
><P
> Add server parameter <TT
CLASS="VARNAME"
>regex_flavor</TT
> to control regular expression processing (Tom)
</P
></LI
><LI
><P
> Make <TT
CLASS="COMMAND"
>pg_ctl</TT
> better handle nonstandard ports (Greg)
</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN146127"
>E.219.3.4. Query Changes</A
></H3
><P
></P
><UL
><LI
><P
>New SQL-standard information schema (Peter)</P
></LI
><LI
><P
>Add read-only transactions (Peter)</P
></LI
><LI
><P
>Print key name and value in foreign-key violation messages (Dmitry Tkach)</P
></LI
><LI
><P
>Allow users to see their own queries in <TT
CLASS="LITERAL"
>pg_stat_activity</TT
> (Kevin Brown)</P
><P
> In prior releases, only the superuser could see query strings
using <TT
CLASS="LITERAL"
>pg_stat_activity</TT
>. Now ordinary users
can see their own query strings.
</P
></LI
><LI
><P
>Fix aggregates in subqueries to match SQL standard (Tom)</P
><P
> The SQL standard says that an aggregate function appearing
within a nested subquery belongs to the outer query if its
argument contains only outer-query variables. Prior
<SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> releases did not handle
this fine point correctly.
</P
></LI
><LI
><P
>Add option to prevent auto-addition of tables referenced in query (Nigel J. Andrews)</P
><P
> By default, tables mentioned in the query are automatically
added to the <TT
CLASS="LITERAL"
>FROM</TT
> clause if they are not already
there. This is compatible with historic
<SPAN
CLASS="PRODUCTNAME"
>POSTGRES</SPAN
> behavior but is contrary to
the SQL standard. This option allows selecting
standard-compatible behavior.
</P
></LI
><LI
><P
>Allow <TT
CLASS="LITERAL"
>UPDATE ... SET col = DEFAULT</TT
> (Rod)</P
><P
> This allows <TT
CLASS="COMMAND"
>UPDATE</TT
> to set a column to its
declared default value.
</P
></LI
><LI
><P
>Allow expressions to be used in <TT
CLASS="LITERAL"
>LIMIT</TT
>/<TT
CLASS="LITERAL"
>OFFSET</TT
> (Tom)</P
><P
> In prior releases, <TT
CLASS="LITERAL"
>LIMIT</TT
>/<TT
CLASS="LITERAL"
>OFFSET</TT
> could
only use constants, not expressions.
</P
></LI
><LI
><P
>Implement <TT
CLASS="LITERAL"
>CREATE TABLE AS EXECUTE</TT
> (Neil, Peter)</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN146165"
>E.219.3.5. Object Manipulation Changes</A
></H3
><P
></P
><UL
><LI
><P
>Make <TT
CLASS="COMMAND"
>CREATE SEQUENCE</TT
> grammar more conforming to SQL:2003 (Neil)</P
></LI
><LI
><P
>Add statement-level triggers (Neil)</P
><P
> While this allows a trigger to fire at the end of a statement,
it does not allow the trigger to access all rows modified by the
statement. This capability is planned for a future release.
</P
></LI
><LI
><P
>Add check constraints for domains (Rod)</P
><P
> This greatly increases the usefulness of domains by allowing
them to use check constraints.
</P
></LI
><LI
><P
>Add <TT
CLASS="COMMAND"
>ALTER DOMAIN</TT
> (Rod)</P
><P
> This allows manipulation of existing domains.
</P
></LI
><LI
><P
>Fix several zero-column table bugs (Tom)</P
><P
> <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> supports zero-column tables. This fixes various bugs
that occur when using such tables.
</P
></LI
><LI
><P
>Have <TT
CLASS="LITERAL"
>ALTER TABLE ... ADD PRIMARY KEY</TT
> add not-null constraint (Rod)</P
><P
> In prior releases, <TT
CLASS="LITERAL"
>ALTER TABLE ... ADD
PRIMARY</TT
> would add a unique index, but not a not-null
constraint. That is fixed in this release.
</P
></LI
><LI
><P
>Add <TT
CLASS="LITERAL"
>ALTER TABLE ... WITHOUT OIDS</TT
> (Rod)</P
><P
> This allows control over whether new and updated rows will have
an OID column. This is most useful for saving storage space.
</P
></LI
><LI
><P
> Add <TT
CLASS="LITERAL"
>ALTER SEQUENCE</TT
> to modify minimum, maximum,
increment, cache, cycle values (Rod)
</P
></LI
><LI
><P
>Add <TT
CLASS="LITERAL"
>ALTER TABLE ... CLUSTER ON</TT
> (Alvaro Herrera)</P
><P
> This command is used by <TT
CLASS="COMMAND"
>pg_dump</TT
> to record the
cluster column for each table previously clustered. This
information is used by database-wide cluster to cluster all
previously clustered tables.
</P
></LI
><LI
><P
>Improve automatic type casting for domains (Rod, Tom)</P
></LI
><LI
><P
>Allow dollar signs in identifiers, except as first character (Tom)</P
></LI
><LI
><P
>Disallow dollar signs in operator names, so <TT
CLASS="LITERAL"
>x=$1</TT
> works (Tom)</P
></LI
><LI
><P
> Allow copying table schema using <TT
CLASS="LITERAL"
>LIKE
<TT
CLASS="REPLACEABLE"
><I
>subtable</I
></TT
></TT
>, also SQL:2003
feature <TT
CLASS="LITERAL"
>INCLUDING DEFAULTS</TT
> (Rod)
</P
></LI
><LI
><P
> Add <TT
CLASS="LITERAL"
>WITH GRANT OPTION</TT
> clause to
<TT
CLASS="COMMAND"
>GRANT</TT
> (Peter)
</P
><P
> This enabled <TT
CLASS="COMMAND"
>GRANT</TT
> to give other users the
ability to grant privileges on an object.
</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN146220"
>E.219.3.6. Utility Command Changes</A
></H3
><P
></P
><UL
><LI
><P
>Add <TT
CLASS="LITERAL"
>ON COMMIT</TT
> clause to <TT
CLASS="COMMAND"
>CREATE TABLE</TT
> for temporary tables (Gavin)</P
><P
> This adds the ability for a table to be dropped or all rows
deleted on transaction commit.
</P
></LI
><LI
><P
>Allow cursors outside transactions using <TT
CLASS="LITERAL"
>WITH HOLD</TT
> (Neil)</P
><P
> In previous releases, cursors were removed at the end of the
transaction that created them. Cursors can now be created with
the <TT
CLASS="LITERAL"
>WITH HOLD</TT
> option, which allows them to
continue to be accessed after the creating transaction has
committed.
</P
></LI
><LI
><P
><TT
CLASS="LITERAL"
>FETCH 0</TT
> and <TT
CLASS="LITERAL"
>MOVE 0 </TT
> now do nothing (Bruce)</P
><P
> In previous releases, <TT
CLASS="LITERAL"
>FETCH 0</TT
> fetched all
remaining rows, and <TT
CLASS="LITERAL"
>MOVE 0</TT
> moved to the end
of the cursor.
</P
></LI
><LI
><P
> Cause <TT
CLASS="COMMAND"
>FETCH</TT
> and <TT
CLASS="COMMAND"
>MOVE</TT
> to
return the number of rows fetched/moved, or zero if at the
beginning/end of cursor, per SQL standard (Bruce)
</P
><P
> In prior releases, the row count returned by
<TT
CLASS="COMMAND"
>FETCH</TT
> and <TT
CLASS="COMMAND"
>MOVE</TT
> did not
accurately reflect the number of rows processed.
</P
></LI
><LI
><P
>Properly handle <TT
CLASS="LITERAL"
>SCROLL</TT
> with cursors, or
report an error (Neil)</P
><P
> Allowing random access (both forward and backward scrolling) to
some kinds of queries cannot be done without some additional
work. If <TT
CLASS="LITERAL"
>SCROLL</TT
> is specified when the cursor
is created, this additional work will be performed. Furthermore,
if the cursor has been created with <TT
CLASS="LITERAL"
>NO SCROLL</TT
>,
no random access is allowed.
</P
></LI
><LI
><P
> Implement SQL-compatible options <TT
CLASS="LITERAL"
>FIRST</TT
>,
<TT
CLASS="LITERAL"
>LAST</TT
>, <TT
CLASS="LITERAL"
>ABSOLUTE <TT
CLASS="REPLACEABLE"
><I
>n</I
></TT
></TT
>,
<TT
CLASS="LITERAL"
>RELATIVE <TT
CLASS="REPLACEABLE"
><I
>n</I
></TT
></TT
> for
<TT
CLASS="COMMAND"
>FETCH</TT
> and <TT
CLASS="COMMAND"
>MOVE</TT
> (Tom)
</P
></LI
><LI
><P
>Allow <TT
CLASS="COMMAND"
>EXPLAIN</TT
> on <TT
CLASS="COMMAND"
>DECLARE CURSOR</TT
> (Tom)</P
></LI
><LI
><P
>Allow <TT
CLASS="COMMAND"
>CLUSTER</TT
> to use index marked as pre-clustered by default (Alvaro Herrera)</P
></LI
><LI
><P
>Allow <TT
CLASS="COMMAND"
>CLUSTER</TT
> to cluster all tables (Alvaro Herrera)</P
><P
> This allows all previously clustered tables in a database to be
reclustered with a single command.
</P
></LI
><LI
><P
>Prevent <TT
CLASS="COMMAND"
>CLUSTER</TT
> on partial indexes (Tom)</P
></LI
><LI
><P
>Allow DOS and Mac line-endings in <TT
CLASS="COMMAND"
>COPY</TT
> files (Bruce)</P
></LI
><LI
><P
> Disallow literal carriage return as a data value,
backslash-carriage-return and <TT
CLASS="LITERAL"
>\r</TT
> are still allowed
(Bruce)
</P
></LI
><LI
><P
><TT
CLASS="COMMAND"
>COPY</TT
> changes (binary, <TT
CLASS="LITERAL"
>\.</TT
>) (Tom)</P
></LI
><LI
><P
>Recover from <TT
CLASS="COMMAND"
>COPY</TT
> failure cleanly (Tom)</P
></LI
><LI
><P
>Prevent possible memory leaks in <TT
CLASS="COMMAND"
>COPY</TT
> (Tom)</P
></LI
><LI
><P
>Make <TT
CLASS="COMMAND"
>TRUNCATE</TT
> transaction-safe (Rod)</P
><P
> <TT
CLASS="COMMAND"
>TRUNCATE</TT
> can now be used inside a
transaction. If the transaction aborts, the changes made by the
<TT
CLASS="COMMAND"
>TRUNCATE</TT
> are automatically rolled back.
</P
></LI
><LI
><P
> Allow prepare/bind of utility commands like
<TT
CLASS="COMMAND"
>FETCH</TT
> and <TT
CLASS="COMMAND"
>EXPLAIN</TT
> (Tom)
</P
></LI
><LI
><P
>Add <TT
CLASS="COMMAND"
>EXPLAIN EXECUTE</TT
> (Neil)</P
></LI
><LI
><P
>Improve <TT
CLASS="COMMAND"
>VACUUM</TT
> performance on indexes by reducing WAL traffic (Tom)</P
></LI
><LI
><P
>Functional indexes have been generalized into indexes on expressions (Tom)</P
><P
> In prior releases, functional indexes only supported a simple
function applied to one or more column names. This release
allows any type of scalar expression.
</P
></LI
><LI
><P
> Have <TT
CLASS="COMMAND"
>SHOW TRANSACTION ISOLATION</TT
> match input
to <TT
CLASS="COMMAND"
>SET TRANSACTION ISOLATION</TT
>
(Tom)
</P
></LI
><LI
><P
> Have <TT
CLASS="COMMAND"
>COMMENT ON DATABASE</TT
> on nonlocal
database generate a warning, rather than an error (Rod)
</P
><P
> Database comments are stored in database-local tables so
comments on a database have to be stored in each database.
</P
></LI
><LI
><P
> Improve reliability of <TT
CLASS="COMMAND"
>LISTEN</TT
>/<TT
CLASS="COMMAND"
>NOTIFY</TT
> (Tom)
</P
></LI
><LI
><P
>Allow <TT
CLASS="COMMAND"
>REINDEX</TT
> to reliably reindex nonshared system catalog indexes (Tom)</P
><P
> This allows system tables to be reindexed without the
requirement of a standalone session, which was necessary in
previous releases. The only tables that now require a standalone
session for reindexing are the global system tables
<TT
CLASS="LITERAL"
>pg_database</TT
>, <TT
CLASS="LITERAL"
>pg_shadow</TT
>, and
<TT
CLASS="LITERAL"
>pg_group</TT
>.
</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN146331"
>E.219.3.7. Data Type and Function Changes</A
></H3
><P
></P
><UL
><LI
><P
> New server parameter <TT
CLASS="VARNAME"
>extra_float_digits</TT
> to
control precision display of floating-point numbers (Pedro
Ferreira, Tom)
</P
><P
> This controls output precision which was causing regression
testing problems.
</P
></LI
><LI
><P
>Allow <TT
CLASS="LITERAL"
>+1300</TT
> as a numeric time-zone specifier, for FJST (Tom)</P
></LI
><LI
><P
> Remove rarely used functions <CODE
CLASS="FUNCTION"
>oidrand</CODE
>,
<CODE
CLASS="FUNCTION"
>oidsrand</CODE
>, and <CODE
CLASS="FUNCTION"
>userfntest</CODE
> functions
(Neil)
</P
></LI
><LI
><P
>Add <CODE
CLASS="FUNCTION"
>md5()</CODE
> function to main server, already in <TT
CLASS="FILENAME"
>contrib/pgcrypto</TT
> (Joe)</P
><P
> An MD5 function was frequently requested. For more complex
encryption capabilities, use
<TT
CLASS="FILENAME"
>contrib/pgcrypto</TT
>.
</P
></LI
><LI
><P
>Increase date range of <TT
CLASS="TYPE"
>timestamp</TT
> (John Cochran)</P
></LI
><LI
><P
> Change <TT
CLASS="LITERAL"
>EXTRACT(EPOCH FROM timestamp)</TT
> so
<TT
CLASS="TYPE"
>timestamp without time zone</TT
> is assumed to be in
local time, not GMT (Tom)
</P
></LI
><LI
><P
>Trap division by zero in case the operating system doesn't prevent it (Tom)</P
></LI
><LI
><P
>Change the <TT
CLASS="TYPE"
>numeric</TT
> data type internally to base 10000 (Tom)</P
></LI
><LI
><P
>New <CODE
CLASS="FUNCTION"
>hostmask()</CODE
> function (Greg Wickham)</P
></LI
><LI
><P
>Fixes for <CODE
CLASS="FUNCTION"
>to_char()</CODE
> and <CODE
CLASS="FUNCTION"
>to_timestamp()</CODE
> (Karel)</P
></LI
><LI
><P
> Allow functions that can take any argument data type and return
any data type, using <TT
CLASS="TYPE"
>anyelement</TT
> and
<TT
CLASS="TYPE"
>anyarray</TT
> (Joe)
</P
><P
> This allows the creation of functions that can work with any
data type.
</P
></LI
><LI
><P
> Arrays can now be specified as <TT
CLASS="LITERAL"
>ARRAY[1,2,3]</TT
>,
<TT
CLASS="LITERAL"
>ARRAY[['a','b'],['c','d']]</TT
>, or
<TT
CLASS="LITERAL"
>ARRAY[ARRAY[ARRAY[2]]]</TT
> (Joe)
</P
></LI
><LI
><P
> Allow proper comparisons for arrays, including <TT
CLASS="LITERAL"
>ORDER
BY</TT
> and <TT
CLASS="LITERAL"
>DISTINCT</TT
> support
(Joe)
</P
></LI
><LI
><P
>Allow indexes on array columns (Joe)</P
></LI
><LI
><P
>Allow array concatenation with <TT
CLASS="LITERAL"
>||</TT
> (Joe)</P
></LI
><LI
><P
> Allow <TT
CLASS="LITERAL"
>WHERE</TT
> qualification
<TT
CLASS="LITERAL"
><TT
CLASS="REPLACEABLE"
><I
>expr</I
></TT
> <TT
CLASS="REPLACEABLE"
><I
>op</I
></TT
> ANY/SOME/ALL
(<TT
CLASS="REPLACEABLE"
><I
>array_expr</I
></TT
>)</TT
> (Joe)
</P
><P
> This allows arrays to behave like a list of values, for purposes
like <TT
CLASS="LITERAL"
>SELECT * FROM tab WHERE col IN
(array_val)</TT
>.
</P
></LI
><LI
><P
> New array functions <CODE
CLASS="FUNCTION"
>array_append</CODE
>,
<CODE
CLASS="FUNCTION"
>array_cat</CODE
>, <CODE
CLASS="FUNCTION"
>array_lower</CODE
>,
<CODE
CLASS="FUNCTION"
>array_prepend</CODE
>, <CODE
CLASS="FUNCTION"
>array_to_string</CODE
>,
<CODE
CLASS="FUNCTION"
>array_upper</CODE
>, <CODE
CLASS="FUNCTION"
>string_to_array</CODE
> (Joe)
</P
></LI
><LI
><P
>Allow user defined aggregates to use polymorphic functions (Joe)</P
></LI
><LI
><P
>Allow assignments to empty arrays (Joe)</P
></LI
><LI
><P
> Allow 60 in seconds fields of <TT
CLASS="TYPE"
>time</TT
>,
<TT
CLASS="TYPE"
>timestamp</TT
>, and <TT
CLASS="TYPE"
>interval</TT
> input values
(Tom)
</P
><P
> Sixty-second values are needed for leap seconds.
</P
></LI
><LI
><P
>Allow <TT
CLASS="TYPE"
>cidr</TT
> data type to be cast to <TT
CLASS="TYPE"
>text</TT
> (Tom)</P
></LI
><LI
><P
>Disallow invalid time zone names in SET TIMEZONE</P
></LI
><LI
><P
> Trim trailing spaces when <TT
CLASS="TYPE"
>char</TT
> is cast to
<TT
CLASS="TYPE"
>varchar</TT
> or <TT
CLASS="TYPE"
>text</TT
> (Tom)
</P
></LI
><LI
><P
> Make <TT
CLASS="TYPE"
>float(<TT
CLASS="REPLACEABLE"
><I
>p</I
></TT
>)</TT
> measure the precision
<TT
CLASS="REPLACEABLE"
><I
>p</I
></TT
> in binary digits, not decimal digits
(Tom)
</P
></LI
><LI
><P
>Add IPv6 support to the <TT
CLASS="TYPE"
>inet</TT
> and <TT
CLASS="TYPE"
>cidr</TT
> data types (Michael Graff)</P
></LI
><LI
><P
>Add <CODE
CLASS="FUNCTION"
>family()</CODE
> function to report whether address is IPv4 or IPv6 (Michael Graff)</P
></LI
><LI
><P
> Have <TT
CLASS="LITERAL"
>SHOW datestyle</TT
> generate output similar
to that used by <TT
CLASS="LITERAL"
>SET datestyle</TT
> (Tom)
</P
></LI
><LI
><P
> Make <TT
CLASS="LITERAL"
>EXTRACT(TIMEZONE)</TT
> and <TT
CLASS="LITERAL"
>SET/SHOW
TIME ZONE</TT
> follow the SQL convention for the sign of
time zone offsets, i.e., positive is east from UTC (Tom)
</P
></LI
><LI
><P
>Fix <TT
CLASS="LITERAL"
>date_trunc('quarter', ...)</TT
> (Böjthe Zoltán)</P
><P
> Prior releases returned an incorrect value for this function call.
</P
></LI
><LI
><P
>Make <CODE
CLASS="FUNCTION"
>initcap()</CODE
> more compatible with Oracle (Mike Nolan)</P
><P
> <CODE
CLASS="FUNCTION"
>initcap()</CODE
> now uppercases a letter appearing
after any non-alphanumeric character, rather than only after
whitespace.
</P
></LI
><LI
><P
>Allow only <TT
CLASS="VARNAME"
>datestyle</TT
> field order for date values not in ISO-8601 format (Greg)</P
></LI
><LI
><P
> Add new <TT
CLASS="VARNAME"
>datestyle</TT
> values <TT
CLASS="LITERAL"
>MDY</TT
>,
<TT
CLASS="LITERAL"
>DMY</TT
>, and <TT
CLASS="LITERAL"
>YMD</TT
> to set input field order;
honor <TT
CLASS="LITERAL"
>US</TT
> and <TT
CLASS="LITERAL"
>European</TT
> for backward
compatibility (Tom)
</P
></LI
><LI
><P
> String literals like <TT
CLASS="LITERAL"
>'now'</TT
> or
<TT
CLASS="LITERAL"
>'today'</TT
> will no longer work as a column
default. Use functions such as <CODE
CLASS="FUNCTION"
>now()</CODE
>,
<CODE
CLASS="FUNCTION"
>current_timestamp</CODE
> instead. (change
required for prepared statements) (Tom)
</P
></LI
><LI
><P
>Treat NaN as larger than any other value in <CODE
CLASS="FUNCTION"
>min()</CODE
>/<CODE
CLASS="FUNCTION"
>max()</CODE
> (Tom)</P
><P
> NaN was already sorted after ordinary numeric values for most
purposes, but <CODE
CLASS="FUNCTION"
>min()</CODE
> and <CODE
CLASS="FUNCTION"
>max()</CODE
> didn't
get this right.
</P
></LI
><LI
><P
>Prevent interval from suppressing <TT
CLASS="LITERAL"
>:00</TT
>
seconds display</P
></LI
><LI
><P
> New functions <CODE
CLASS="FUNCTION"
>pg_get_triggerdef(prettyprint)</CODE
>
and <CODE
CLASS="FUNCTION"
>pg_conversion_is_visible()</CODE
> (Christopher)
</P
></LI
><LI
><P
>Allow time to be specified as <TT
CLASS="LITERAL"
>040506</TT
> or <TT
CLASS="LITERAL"
>0405</TT
> (Tom)</P
></LI
><LI
><P
> Input date order must now be <TT
CLASS="LITERAL"
>YYYY-MM-DD</TT
> (with 4-digit year) or
match <TT
CLASS="VARNAME"
>datestyle</TT
>
</P
></LI
><LI
><P
> Make <CODE
CLASS="FUNCTION"
>pg_get_constraintdef</CODE
> support
unique, primary-key, and check constraints (Christopher)
</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN146500"
>E.219.3.8. Server-Side Language Changes</A
></H3
><P
></P
><UL
><LI
><P
> Prevent PL/pgSQL crash when <TT
CLASS="LITERAL"
>RETURN NEXT</TT
> is
used on a zero-row record variable (Tom)
</P
></LI
><LI
><P
> Make PL/Python's <CODE
CLASS="FUNCTION"
>spi_execute</CODE
> interface
handle null values properly (Andrew Bosma)
</P
></LI
><LI
><P
>Allow PL/pgSQL to declare variables of composite types without <TT
CLASS="LITERAL"
>%ROWTYPE</TT
> (Tom)</P
></LI
><LI
><P
>Fix PL/Python's <CODE
CLASS="FUNCTION"
>_quote()</CODE
> function to handle big integers</P
></LI
><LI
><P
>Make PL/Python an untrusted language, now called <TT
CLASS="LITERAL"
>plpythonu</TT
> (Kevin Jacobs, Tom)</P
><P
> The Python language no longer supports a restricted execution
environment, so the trusted version of PL/Python was removed. If
this situation changes, a version of PL/Python that can be used
by non-superusers will be readded.
</P
></LI
><LI
><P
>Allow polymorphic PL/pgSQL functions (Joe, Tom)</P
></LI
><LI
><P
>Allow polymorphic SQL functions (Joe)</P
></LI
><LI
><P
> Improved compiled function caching mechanism in PL/pgSQL with
full support for polymorphism (Joe)
</P
></LI
><LI
><P
> Add new parameter <TT
CLASS="LITERAL"
>$0</TT
> in PL/pgSQL representing the
function's actual return type (Joe)
</P
></LI
><LI
><P
> Allow PL/Tcl and PL/Python to use the same trigger on multiple tables (Tom)
</P
></LI
><LI
><P
> Fixed PL/Tcl's <CODE
CLASS="FUNCTION"
>spi_prepare</CODE
> to accept fully
qualified type names in the parameter type list
(Jan)
</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN146533"
>E.219.3.9. psql Changes</A
></H3
><P
></P
><UL
><LI
><P
>Add <TT
CLASS="LITERAL"
>\pset pager always</TT
> to always use pager (Greg)</P
><P
> This forces the pager to be used even if the number of rows is
less than the screen height. This is valuable for rows that
wrap across several screen rows.
</P
></LI
><LI
><P
>Improve tab completion (Rod, Ross Reedstrom, Ian Barwick)</P
></LI
><LI
><P
>Reorder <TT
CLASS="LITERAL"
>\?</TT
> help into groupings (Harald Armin Massa, Bruce)</P
></LI
><LI
><P
>Add backslash commands for listing schemas, casts, and conversions (Christopher)</P
></LI
><LI
><P
> <TT
CLASS="COMMAND"
>\encoding</TT
> now changes based on the server parameter
<TT
CLASS="VARNAME"
>client_encoding</TT
> (Tom)
</P
><P
> In previous versions, <TT
CLASS="COMMAND"
>\encoding</TT
> was not aware
of encoding changes made using <TT
CLASS="LITERAL"
>SET
client_encoding</TT
>.
</P
></LI
><LI
><P
>Save editor buffer into readline history (Ross)</P
><P
> When <TT
CLASS="COMMAND"
>\e</TT
> is used to edit a query, the result is saved
in the readline history for retrieval using the up arrow.
</P
></LI
><LI
><P
>Improve <TT
CLASS="COMMAND"
>\d</TT
> display (Christopher)</P
></LI
><LI
><P
>Enhance HTML mode to be more standards-conforming (Greg)</P
></LI
><LI
><P
>New <TT
CLASS="COMMAND"
>\set AUTOCOMMIT off</TT
> capability (Tom)</P
><P
> This takes the place of the removed server parameter <TT
CLASS="VARNAME"
>autocommit</TT
>.
</P
></LI
><LI
><P
>New <TT
CLASS="COMMAND"
>\set VERBOSITY</TT
> to control error detail (Tom)</P
><P
> This controls the new error reporting details.
</P
></LI
><LI
><P
>New prompt escape sequence <TT
CLASS="LITERAL"
>%x</TT
> to show transaction status (Tom)</P
></LI
><LI
><P
>Long options for <SPAN
CLASS="APPLICATION"
>psql</SPAN
> are now available on all platforms</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN146578"
>E.219.3.10. pg_dump Changes</A
></H3
><P
></P
><UL
><LI
><P
>Multiple pg_dump fixes, including tar format and large objects</P
></LI
><LI
><P
>Allow pg_dump to dump specific schemas (Neil)</P
></LI
><LI
><P
>Make pg_dump preserve column storage characteristics (Christopher)</P
><P
> This preserves <TT
CLASS="LITERAL"
>ALTER TABLE ... SET STORAGE</TT
> information.
</P
></LI
><LI
><P
>Make pg_dump preserve <TT
CLASS="COMMAND"
>CLUSTER</TT
> characteristics (Christopher)</P
></LI
><LI
><P
> Have pg_dumpall use <TT
CLASS="COMMAND"
>GRANT</TT
>/<TT
CLASS="COMMAND"
>REVOKE</TT
> to dump database-level privileges (Tom)
</P
></LI
><LI
><P
> Allow pg_dumpall to support the options <TT
CLASS="OPTION"
>-a</TT
>,
<TT
CLASS="OPTION"
>-s</TT
>, <TT
CLASS="OPTION"
>-x</TT
> of pg_dump (Tom)
</P
></LI
><LI
><P
>Prevent pg_dump from lowercasing identifiers specified on the command line (Tom)</P
></LI
><LI
><P
> pg_dump options <TT
CLASS="OPTION"
>--use-set-session-authorization</TT
>
and <TT
CLASS="OPTION"
>--no-reconnect</TT
> now do nothing, all dumps
use <TT
CLASS="COMMAND"
>SET SESSION AUTHORIZATION</TT
>
</P
><P
> pg_dump no longer reconnects to switch users, but instead always
uses <TT
CLASS="COMMAND"
>SET SESSION AUTHORIZATION</TT
>. This will
reduce password prompting during restores.
</P
></LI
><LI
><P
>Long options for <SPAN
CLASS="APPLICATION"
>pg_dump</SPAN
> are now available on all platforms</P
><P
> <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> now includes its own
long-option processing routines.
</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN146615"
>E.219.3.11. libpq Changes</A
></H3
><P
></P
><UL
><LI
><P
> Add function <CODE
CLASS="FUNCTION"
>PQfreemem</CODE
> for freeing memory on
Windows, suggested for <TT
CLASS="COMMAND"
>NOTIFY</TT
> (Bruce)
</P
><P
> Windows requires that memory allocated in a library be freed by
a function in the same library, hence
<CODE
CLASS="FUNCTION"
>free()</CODE
> doesn't work for freeing memory
allocated by libpq. <CODE
CLASS="FUNCTION"
>PQfreemem</CODE
> is the proper
way to free libpq memory, especially on Windows, and is
recommended for other platforms as well.
</P
></LI
><LI
><P
>Document service capability, and add sample file (Bruce)</P
><P
> This allows clients to look up connection information in a
central file on the client machine.
</P
></LI
><LI
><P
> Make <CODE
CLASS="FUNCTION"
>PQsetdbLogin</CODE
> have the same defaults as
<CODE
CLASS="FUNCTION"
>PQconnectdb</CODE
> (Tom)
</P
></LI
><LI
><P
>Allow libpq to cleanly fail when result sets are too large (Tom)</P
></LI
><LI
><P
> Improve performance of function <CODE
CLASS="FUNCTION"
>PQunescapeBytea</CODE
> (Ben Lamb)
</P
></LI
><LI
><P
> Allow thread-safe libpq with <TT
CLASS="FILENAME"
>configure</TT
>
option <TT
CLASS="OPTION"
>--enable-thread-safety</TT
> (Lee Kindness,
Philip Yarra)
</P
></LI
><LI
><P
> Allow function <CODE
CLASS="FUNCTION"
>pqInternalNotice</CODE
> to accept a
format string and arguments instead of just a preformatted
message (Tom, Sean Chittenden)
</P
></LI
><LI
><P
> Control SSL negotiation with <TT
CLASS="LITERAL"
>sslmode</TT
> values
<TT
CLASS="LITERAL"
>disable</TT
>, <TT
CLASS="LITERAL"
>allow</TT
>,
<TT
CLASS="LITERAL"
>prefer</TT
>, and <TT
CLASS="LITERAL"
>require</TT
> (Jon
Jensen)
</P
></LI
><LI
><P
>Allow new error codes and levels of text (Tom)</P
></LI
><LI
><P
>Allow access to the underlying table and column of a query result (Tom)</P
><P
> This is helpful for query-builder applications that want to know
the underlying table and column names associated with a specific
result set.
</P
></LI
><LI
><P
>Allow access to the current transaction status (Tom)</P
></LI
><LI
><P
>Add ability to pass binary data directly to the server (Tom)</P
></LI
><LI
><P
> Add function <CODE
CLASS="FUNCTION"
>PQexecPrepared</CODE
> and
<CODE
CLASS="FUNCTION"
>PQsendQueryPrepared</CODE
> functions which perform
bind/execute of previously prepared statements (Tom)
</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN146664"
>E.219.3.12. JDBC Changes</A
></H3
><P
></P
><UL
><LI
><P
>Allow <CODE
CLASS="FUNCTION"
>setNull</CODE
> on updateable result sets</P
></LI
><LI
><P
>Allow <CODE
CLASS="FUNCTION"
>executeBatch</CODE
> on a prepared statement (Barry)</P
></LI
><LI
><P
>Support SSL connections (Barry)</P
></LI
><LI
><P
>Handle schema names in result sets (Paul Sorenson)</P
></LI
><LI
><P
>Add refcursor support (Nic Ferrier)</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN146679"
>E.219.3.13. Miscellaneous Interface Changes</A
></H3
><P
></P
><UL
><LI
><P
>Prevent possible memory leak or core dump during libpgtcl shutdown (Tom)</P
></LI
><LI
><P
>Add Informix compatibility to ECPG (Michael)</P
><P
> This allows ECPG to process embedded C programs that were
written using certain Informix extensions.
</P
></LI
><LI
><P
>Add type <TT
CLASS="TYPE"
>decimal</TT
> to ECPG that is fixed length, for Informix (Michael)</P
></LI
><LI
><P
> Allow thread-safe embedded SQL programs with
<TT
CLASS="FILENAME"
>configure</TT
> option
<TT
CLASS="OPTION"
>--enable-thread-safety</TT
> (Lee Kindness, Bruce)
</P
><P
> This allows multiple threads to access the database at the same
time.
</P
></LI
><LI
><P
>Moved Python client PyGreSQL to <A
HREF="http://www.pygresql.org"
TARGET="_top"
>http://www.pygresql.org</A
> (Marc)</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN146698"
>E.219.3.14. Source Code Changes</A
></H3
><P
></P
><UL
><LI
><P
>Prevent need for separate platform geometry regression result files (Tom)</P
></LI
><LI
><P
>Improved PPC locking primitive (Reinhard Max)</P
></LI
><LI
><P
>New function <CODE
CLASS="FUNCTION"
>palloc0</CODE
> to allocate and clear memory (Bruce)</P
></LI
><LI
><P
>Fix locking code for s390x CPU (64-bit) (Tom)</P
></LI
><LI
><P
>Allow OpenBSD to use local ident credentials (William Ahern)</P
></LI
><LI
><P
>Make query plan trees read-only to executor (Tom)</P
></LI
><LI
><P
>Add Darwin startup scripts (David Wheeler)</P
></LI
><LI
><P
>Allow libpq to compile with Borland C++ compiler (Lester Godwin, Karl Waclawek)</P
></LI
><LI
><P
>Use our own version of <CODE
CLASS="FUNCTION"
>getopt_long()</CODE
> if needed (Peter)</P
></LI
><LI
><P
>Convert administration scripts to C (Peter)</P
></LI
><LI
><P
> Bison >= 1.85 is now required to build the <SPAN
CLASS="PRODUCTNAME"
>PostgreSQL</SPAN
> grammar, if building from CVS</P
></LI
><LI
><P
>Merge documentation into one book (Peter)</P
></LI
><LI
><P
>Add Windows compatibility functions (Bruce)</P
></LI
><LI
><P
>Allow client interfaces to compile under MinGW (Bruce)</P
></LI
><LI
><P
>New <CODE
CLASS="FUNCTION"
>ereport()</CODE
> function for error reporting (Tom)</P
></LI
><LI
><P
>Support Intel compiler on Linux (Peter)</P
></LI
><LI
><P
>Improve Linux startup scripts (Slawomir Sudnik, Darko Prenosil)</P
></LI
><LI
><P
>Add support for AMD Opteron and Itanium (Jeffrey W. Baker, Bruce)</P
></LI
><LI
><P
>Remove <TT
CLASS="OPTION"
>--enable-recode</TT
> option from <TT
CLASS="COMMAND"
>configure</TT
></P
><P
> This was no longer needed now that we have <TT
CLASS="COMMAND"
>CREATE CONVERSION</TT
>.
</P
></LI
><LI
><P
>Generate a compile error if spinlock code is not found (Bruce)</P
><P
> Platforms without spinlock code will now fail to compile, rather
than silently using semaphores. This failure can be disabled
with a new <TT
CLASS="COMMAND"
>configure</TT
> option.
</P
></LI
></UL
></DIV
><DIV
CLASS="SECT3"
><H3
CLASS="SECT3"
><A
NAME="AEN146751"
>E.219.3.15. Contrib Changes</A
></H3
><P
></P
><UL
><LI
><P
>Change dbmirror license to BSD</P
></LI
><LI
><P
>Improve earthdistance (Bruno Wolff III)</P
></LI
><LI
><P
>Portability improvements to pgcrypto (Marko Kreen)</P
></LI
><LI
><P
>Prevent crash in xml (John Gray, Michael Richards)</P
></LI
><LI
><P
>Update oracle</P
></LI
><LI
><P
>Update mysql</P
></LI
><LI
><P
>Update cube (Bruno Wolff III)</P
></LI
><LI
><P
>Update earthdistance to use cube (Bruno Wolff III)</P
></LI
><LI
><P
>Update btree_gist (Oleg)</P
></LI
><LI
><P
>New tsearch2 full-text search module (Oleg, Teodor)</P
></LI
><LI
><P
>Add hash-based crosstab function to tablefuncs (Joe)</P
></LI
><LI
><P
>Add serial column to order <CODE
CLASS="FUNCTION"
>connectby()</CODE
> siblings in tablefuncs (Nabil Sayegh,Joe)</P
></LI
><LI
><P
>Add named persistent connections to dblink (Shridhar Daithanka)</P
></LI
><LI
><P
>New pg_autovacuum allows automatic <TT
CLASS="COMMAND"
>VACUUM</TT
> (Matthew T. O'Connor)</P
></LI
><LI
><P
>Make pgbench honor environment variables <TT
CLASS="ENVAR"
>PGHOST</TT
>, <TT
CLASS="ENVAR"
>PGPORT</TT
>, <TT
CLASS="ENVAR"
>PGUSER</TT
> (Tatsuo)</P
></LI
><LI
><P
>Improve intarray (Teodor Sigaev)</P
></LI
><LI
><P
>Improve pgstattuple (Rod)</P
></LI
><LI
><P
>Fix bug in <CODE
CLASS="FUNCTION"
>metaphone()</CODE
> in fuzzystrmatch</P
></LI
><LI
><P
>Improve adddepend (Rod)</P
></LI
><LI
><P
>Update spi/timetravel (Böjthe Zoltán)</P
></LI
><LI
><P
>Fix dbase <TT
CLASS="OPTION"
>-s</TT
> option and improve non-ASCII handling (Thomas Behr, Márcio Smiderle)</P
></LI
><LI
><P
>Remove array module because features now included by default (Joe)</P
></LI
></UL
></DIV
></DIV
></DIV
><DIV
CLASS="NAVFOOTER"
><HR
ALIGN="LEFT"
WIDTH="100%"><TABLE
SUMMARY="Footer navigation table"
WIDTH="100%"
BORDER="0"
CELLPADDING="0"
CELLSPACING="0"
><TR
><TD
WIDTH="33%"
ALIGN="left"
VALIGN="top"
><A
HREF="release-7-4-1.html"
ACCESSKEY="P"
>Prev</A
></TD
><TD
WIDTH="34%"
ALIGN="center"
VALIGN="top"
><A
HREF="index.html"
ACCESSKEY="H"
>Home</A
></TD
><TD
WIDTH="33%"
ALIGN="right"
VALIGN="top"
><A
HREF="release-7-3-21.html"
ACCESSKEY="N"
>Next</A
></TD
></TR
><TR
><TD
WIDTH="33%"
ALIGN="left"
VALIGN="top"
>Release 7.4.1</TD
><TD
WIDTH="34%"
ALIGN="center"
VALIGN="top"
><A
HREF="release.html"
ACCESSKEY="U"
>Up</A
></TD
><TD
WIDTH="33%"
ALIGN="right"
VALIGN="top"
>Release 7.3.21</TD
></TR
></TABLE
></DIV
></BODY
></HTML
>
