Update 9.5 release notes through today.

Also do another round of copy-editing, and fix up remaining FIXME items.
This commit is contained in:
Tom Lane 2015-12-15 16:42:18 -05:00
parent 049469e7e7
commit bfc7f5dd5d
1 changed files with 214 additions and 119 deletions

View File

@ -6,8 +6,8 @@
<note>
<title>Release Date</title>
<simpara>2015-??-??</simpara>
<simpara>Current as of 2015-11-07</simpara>
<simpara>2016-??-??</simpara>
<simpara>Current as of 2015-12-15</simpara>
</note>
<sect2>
@ -23,7 +23,7 @@
<listitem>
<para>
Allow <link linkend="sql-on-conflict"><command>INSERTS</></>
Allow <link linkend="sql-on-conflict"><command>INSERT</></>s
that would generate constraint conflicts to be turned into
<command>UPDATE</>s or ignored
</para>
@ -31,7 +31,7 @@
<listitem>
<para>
Add <literal>GROUP BY</> analysis functions <link
Add <literal>GROUP BY</> analysis features <link
linkend="queries-grouping-sets"><literal>GROUPING SETS</></>,
<link linkend="queries-grouping-sets"><literal>CUBE</></> and
<link linkend="queries-grouping-sets"><literal>ROLLUP</></>
@ -181,9 +181,9 @@
2015-05-08 [9a08841] Stephe..: Change default for include_realm to 1
-->
<para>
Change the default value of <link
linkend="gssapi-auth"><varname>include_realm</></> to 1, so that
by default the realm is not removed from a <acronym>GSS</>
Change the default value of the GSSAPI <link
linkend="gssapi-auth"><varname>include_realm</></> parameter to 1, so
that by default the realm is not removed from a <acronym>GSS</>
or <acronym>SSPI</> principal name (Stephen Frost)
</para>
</listitem>
@ -201,18 +201,26 @@
</listitem>
<listitem>
<!--
2015-02-23 [88e9823] Heikki..: Replace checkpoint_segments with min_wal_size a..
2015-03-15 [a0f5954] Andres..: Increase max_wal_size's default from 128MB to 1..
2015-06-29 [d661532] Heikki..: Also trigger restartpoints based on max_wal_siz..
-->
<para>
The configuration parameter <varname>checkpoint_segments</varname> has
been removed. Its place is taken by the new
setting <xref linkend="guc-max-wal-size">. If you had previously
tuned <varname>checkpoint_segments</varname>, the following formula will
give you an approximately equivalent setting:
Replace configuration parameter <varname>checkpoint_segments</>
with <xref linkend="guc-min-wal-size">
and <xref linkend="guc-max-wal-size"> (Heikki Linnakangas)
</para>
<para>
If you previously adjusted <varname>checkpoint_segments</varname>, the
following formula will give you an approximately equivalent setting:
<programlisting>
max_wal_size = (3 * checkpoint_segments) * 16MB
</programlisting>
Note that the default setting for <varname>max_wal_size</varname> is
much higher than the default <varname>checkpoint_segments</varname> used
to be, so setting this might no longer be necessary.
to be, so adjusting it might no longer be necessary.
</para>
</listitem>
@ -222,7 +230,7 @@ max_wal_size = (3 * checkpoint_segments) * 16MB
2015-07-28 [6087d95] Andres..: Remove ssl renegotiation support.
-->
<para>
Disable server configuration
Decommission server configuration
parameter <varname>ssl_renegotiation_limit</>, which was deprecated
in earlier releases (Andres Freund)
</para>
@ -253,8 +261,21 @@ max_wal_size = (3 * checkpoint_segments) * 16MB
2015-03-06 [bb8582a] Peter ..: Remove rolcatupdate
-->
<para>
Remove <link linkend="catalog-pg-authid"><structname>pg_authid</></>'s
<structfield>rolcatupdate</> field, as it had no value (Adam Brightwell)
Remove the <link linkend="catalog-pg-authid"><structname>pg_authid</></>
catalog's <structfield>rolcatupdate</> field, as it had no usefulness
(Adam Brightwell)
</para>
</listitem>
<listitem>
<!--
2015-12-13 [28c3667] Magnus..: Consistently set all fields in pg_stat_replica..
-->
<para>
The <link
linkend="monitoring-stats-views-table"><structname>pg_stat_replication</></link>
system view's <structfield>sent</> field is now NULL, not zero, when
it has no valid value (Magnus Hagander)
</para>
</listitem>
@ -264,9 +285,13 @@ max_wal_size = (3 * checkpoint_segments) * 16MB
-->
<para>
Allow <type>json</> and <type>jsonb</> array extraction operators to
accept negative subscripts, which count from the end of JSON arrays.
accept negative subscripts, which count from the end of JSON arrays
(Peter Geoghegan, Andrew Dunstan)
</para>
<para>
Previously, these operators returned <literal>NULL</> for negative
subscripts. (Peter Geoghegan, Andrew Dunstan)
subscripts.
</para>
</listitem>
@ -322,10 +347,10 @@ max_wal_size = (3 * checkpoint_segments) * 16MB
</para>
<para>
Previously, a common table expression was required to return a
large number of rows ordered by bounding-box distance, and the
result then had to be filtered further with a more accurate
non-bounding-box distance calculation.
Previously, to exploit such an index a subquery had to be used to
select a large number of rows ordered by bounding-box distance, and
the result then had to be filtered further with a more accurate
distance calculation.
</para>
</listitem>
@ -352,9 +377,11 @@ Add GUC and storage parameter to set the maximum size of GIN pending list.
</para>
<para>
Previously the list size was controlled
by <xref linkend="guc-work-mem">. This can also be set as an index
storage parameter.
This value can also be set on a per-index basis as an index storage
parameter. Previously the pending-list size was controlled
by <xref linkend="guc-work-mem">, which was awkward because
appropriate values for <varname>work_mem</> are often much too large
for this purpose.
</para>
</listitem>
@ -701,7 +728,6 @@ Add GUC and storage parameter to set the maximum size of GIN pending list.
2015-02-23 [88e9823] Heikki..: Replace checkpoint_segments with min_wal_size a..
2015-03-15 [a0f5954] Andres..: Increase max_wal_size's default from 128MB to 1..
2015-06-29 [d661532] Heikki..: Also trigger restartpoints based on max_wal_siz..
FIXME: Add docs about restartpoint behaviour change
-->
<para>
Replace configuration parameter <varname>checkpoint_segments</>
@ -710,10 +736,14 @@ FIXME: Add docs about restartpoint behaviour change
</para>
<para>
This allows the allocation of a large number of <acronym>WAL</> files
without keeping them if they are not needed. Thus the default
for <xref linkend="guc-max-wal-size"> has been increased
to <literal>1GB</>.
This change allows the allocation of a large number of <acronym>WAL</>
files without keeping them after they are no longer needed.
Therefore the default for <varname>max_wal_size</> has been set
to <literal>1GB</>, much larger than the old default
for <varname>checkpoint_segments</>.
Also note that standby servers perform restartpoints to try to limit
their WAL space consumption to <varname>max_wal_size</>; previously
they did not pay any attention to <varname>checkpoint_segments</>.
</para>
</listitem>
@ -750,8 +780,9 @@ FIXME: Add docs about restartpoint behaviour change
2015-04-03 [a75fb9b] Alvaro..: Have autovacuum workers listen to SIGHUP, too
-->
<para>
Allow running <link linkend="autovacuum">autovacuum workers</>
to respond to configuration parameter changes (Michael Paquier)
Allow <link linkend="autovacuum">autovacuum workers</>
to respond to configuration parameter changes during a run
(Michael Paquier)
</para>
</listitem>
@ -993,7 +1024,7 @@ FIXME: Add docs about restartpoint behaviour change
-->
<para>
Allow <filename>recovery.conf</>'s <link
linkend="primary-conninfo"><varname>primary_conninfo</></> to
linkend="primary-conninfo"><varname>primary_conninfo</></> setting to
use connection <acronym>URI</>s, e.g. <literal>postgres://</>
(Alexander Shulgin)
</para>
@ -1014,7 +1045,7 @@ FIXME: Add docs about restartpoint behaviour change
2015-05-08 [2c8f483] Andres..: Represent columns requiring insert and update p..
-->
<para>
Allow <link linkend="sql-on-conflict"><command>INSERTS</></>
Allow <link linkend="sql-on-conflict"><command>INSERT</></>s
that would generate constraint conflicts to be turned into
<command>UPDATE</>s or ignored (Peter Geoghegan, Heikki
Linnakangas, Andres Freund)
@ -1032,7 +1063,7 @@ FIXME: Add docs about restartpoint behaviour change
2015-05-16 [f3d3118] Andres..: Support GROUPING SETS, CUBE and ROLLUP.
-->
<para>
Add <literal>GROUP BY</> analysis functions <link
Add <literal>GROUP BY</> analysis features <link
linkend="queries-grouping-sets"><literal>GROUPING SETS</></>,
<link linkend="queries-grouping-sets"><literal>CUBE</></> and
<link linkend="queries-grouping-sets"><literal>ROLLUP</></>
@ -1205,8 +1236,9 @@ FIXME: Add docs about restartpoint behaviour change
</para>
<para>
This controls viewing and adding/modifying rows via new commands
<link linkend="SQL-CREATEPOLICY"><command>CREATE</></>/<link
This feature allows row-by-row control over which users can add,
modify, or even see rows in a table. This is controlled by new
commands <link linkend="SQL-CREATEPOLICY"><command>CREATE</></>/<link
linkend="SQL-ALTERPOLICY"><command>ALTER</></>/<link
linkend="SQL-DROPPOLICY"><command>DROP POLICY</></> and <link
linkend="SQL-ALTERTABLE"><command>ALTER TABLE ... ENABLE/DISABLE
@ -1221,7 +1253,7 @@ FIXME: Add docs about restartpoint behaviour change
<para>
Allow changing of the <acronym>WAL</acronym>
logging status of a table after creation with <link
linkend="SQL-ALTERTABLE"><command>ALTER TABLE .. SET LOGGED /
linkend="SQL-ALTERTABLE"><command>ALTER TABLE ... SET LOGGED /
UNLOGGED</></> (Fabr&iacute;zio de Royes Mello)
</para>
</listitem>
@ -1277,7 +1309,7 @@ FIXME: Add docs about restartpoint behaviour change
2015-06-21 [ad89a5d] Alvaro..: Add transforms to pg_get_object_address and fr..
-->
<para>
Allow comments on <link linkend="SQL-CREATEDOMAIN">domain
Support comments on <link linkend="SQL-CREATEDOMAIN">domain
constraints</> (&Aacute;lvaro Herrera)
</para>
</listitem>
@ -1297,14 +1329,14 @@ FIXME: Add docs about restartpoint behaviour change
2015-05-11 [fa26424] Stephe..: Allow LOCK TABLE .. ROW EXCLUSIVE MODE with IN..
-->
<para>
Allow <link linkend="SQL-LOCK"><command>LOCK TABLE .. ROW EXCLUSIVE
Allow <link linkend="SQL-LOCK"><command>LOCK TABLE ... ROW EXCLUSIVE
MODE</></> for those with <command>INSERT</> privileges (Stephen
Frost)
</para>
<para>
Previously this required <command>UPDATE</>, <command>DELETE</>, or
<command>TRUNCATE</> privileges.
Previously this command required <command>UPDATE</>, <command>DELETE</>,
or <command>TRUNCATE</> privileges.
</para>
</listitem>
@ -1336,10 +1368,10 @@ FIXME: Add docs about restartpoint behaviour change
</para>
<para>
This allows these database settings to be
changed more easily than modifying the <link
This allows these per-database settings to be
changed without manually modifying the <link
linkend="catalog-pg-database"><structname>pg_database</></>
system table.
system catalog.
</para>
</listitem>
@ -1368,21 +1400,6 @@ FIXME: Add docs about restartpoint behaviour change
<listitem>
<!--
2015-03-22 [cb1ca4d] Tom Lane: Allow foreign tables to participate in inherita..
-->
<para>
Allow foreign tables to participate in inheritance (Shigeru Hanada,
Etsuro Fujita)
</para>
<para>
This also allows foreign tables to mark check constraints as not
valid, and to set storage and <type>OID</> characteristics.
</para>
</listitem>
<listitem>
<!--
2014-12-17 [fc2ac1f] Tom Lane: Allow CHECK constraints to be placed on foreign..
-->
<para>
@ -1391,7 +1408,7 @@ FIXME: Add docs about restartpoint behaviour change
</para>
<para>
These constraints are assumed to be enforced on the remote server,
Such constraints are assumed to be enforced on the remote server,
and are not enforced locally. However, they are assumed to hold for
purposes of query optimization, such
as <link linkend="ddl-partitioning-constraint-exclusion">constraint
@ -1401,6 +1418,23 @@ FIXME: Add docs about restartpoint behaviour change
<listitem>
<!--
2015-03-22 [cb1ca4d] Tom Lane: Allow foreign tables to participate in inherita..
-->
<para>
Allow foreign tables to participate in inheritance (Shigeru Hanada,
Etsuro Fujita)
</para>
<para>
To let this work naturally, foreign tables are now allowed to have
check constraints marked as not valid, and to set storage
and <type>OID</> characteristics, even though these operations are
effectively no-ops for a foreign table.
</para>
</listitem>
<listitem>
<!--
2015-05-01 [e7cb7ee] Robert..: Allow FDWs and custom scan providers to replace..
2015-05-10 [1a8a4e5] Tom Lane: Code review for foreign/custom join pushdown pa..
-->
@ -1422,12 +1456,18 @@ FIXME: Add docs about restartpoint behaviour change
<listitem>
<!--
2015-05-11 [b488c58] Alvaro..: Allow on-the-fly capture of DDL event details
FIXME: Better description?
-->
<para>
Add a set-returning function <link
linkend="pg-event-trigger-ddl-command-end-functions"><function>pg_event_trigger_ddl_commands()</></>, which returns <acronym>DDL</> activity
associated with event triggers (&Aacute;lvaro Herrera)
Whenever a <literal>ddl_command_end</> event trigger is installed,
capture details of <acronym>DDL</> activity for it to inspect
(&Aacute;lvaro Herrera)
</para>
<para>
This information is available through a set-returning function <link
linkend="pg-event-trigger-ddl-command-end-functions"><function>pg_event_trigger_ddl_commands()</></link>,
or by inspection of C data structures if that function doesn't
provide enough detail.
</para>
</listitem>
@ -1488,7 +1528,7 @@ FIXME: Better description?
-->
<para>
Allow the <link linkend="datatype-xml"><type>xml</></> data type
to accept empty or all-whitespace values (Peter Eisentraut)
to accept empty or all-whitespace content values (Peter Eisentraut)
</para>
<para>
@ -1510,17 +1550,19 @@ FIXME: Better description?
<listitem>
<!--
2014-10-18 [417f924] Bruce ..: interval: tighten precision specification
FIXME: Add more specifics?
-->
<para>
Tighten syntax of <link
linkend="datatype-interval-input"><type>interval</></> precision
specifications (Bruce Momjian)
Disallow non-SQL-standard syntax for <link
linkend="datatype-interval-input"><type>interval</></link> with
both precision and field specifications (Bruce Momjian)
</para>
<para>
Only allow interval precision to be specified after the
<literal>INTERVAL</> keyword if no units are specified.
Per the standard, such type specifications should be written as,
for example, <literal>INTERVAL MINUTE TO SECOND(2)</>.
<productname>PostgreSQL</> formerly allowed this to be written as
<literal>INTERVAL(2) MINUTE TO SECOND</>, but it must now be
written in the standard way.
</para>
</listitem>
@ -1545,8 +1587,8 @@ FIXME: Add more specifics?
Add data
types <link linkend="datatype-oid-table"><type>regrole</></link>
and <link linkend="datatype-oid-table"><type>regnamespace</></link>
that simplify pretty-printing the <type>OID</> of a role or
namespace (Kyotaro Horiguchi)
that simplify entering and pretty-printing the <type>OID</> of a role
or namespace (Kyotaro Horiguchi)
</para>
</listitem>
@ -1563,7 +1605,7 @@ FIXME: Add more specifics?
2015-05-31 [37def42] Andrew..: Rename jsonb_replace to jsonb_set and allow it ..
-->
<para>
Add <type>JSONB</> functions <link
Add <type>jsonb</> functions <link
linkend="functions-json-processing-table"><function>jsonb_set()</></>
and <link
linkend="functions-json-processing-table"><function>jsonb_pretty()</></>
@ -1576,12 +1618,7 @@ FIXME: Add more specifics?
2014-12-12 [7e354ab] Andrew..: Add several generator functions for jsonb that ..
-->
<para>
Add several generator functions for <type>jsonb</> that already
existed for <type>json</> (Andrew Dunstan)
</para>
<para>
These are <link
Add <type>jsonb</> generator functions <link
linkend="functions-json-creation-table"><function>to_jsonb()</></>,
<link
linkend="functions-json-creation-table"><function>jsonb_object()</></>,
@ -1592,7 +1629,12 @@ FIXME: Add more specifics?
<link
linkend="functions-aggregate-table"><function>jsonb_agg()</></>,
and <link
linkend="functions-aggregate-table"><function>jsonb_object_agg()</></>.
linkend="functions-aggregate-table"><function>jsonb_object_agg()</></>
(Andrew Dunstan)
</para>
<para>
Equivalent functions already existed for type <type>json</>.
</para>
</listitem>
@ -1638,7 +1680,7 @@ FIXME: Add more specifics?
linkend="functions-json-processing-table"><function>json_strip_nulls()</></>
and <link
linkend="functions-json-processing-table"><function>jsonb_strip_nulls()</></>
functions to remove <type>json</> null values from documents
functions to remove JSON null values from documents
(Andrew Dunstan)
</para>
</listitem>
@ -1694,7 +1736,7 @@ FIXME: Add more specifics?
2014-12-15 [4520ba6] Heikki..: Add point <-> polygon distance operator.
-->
<para>
Add <link
Add a <link
linkend="gist-builtin-opclasses-table"><type>point</>-to-<type>polygon</></>
distance operator (&lt;-&gt;) (Alexander Korotkov)
</para>
@ -1705,7 +1747,7 @@ FIXME: Add more specifics?
2014-08-27 [8167a38] Jeff D..: Allow multibyte characters as escape in SIMILA..
-->
<para>
Allow multibyte characters as escape in <link
Allow multibyte characters as escapes in <link
linkend="functions-similarto-regexp"><literal>SIMILAR TO</></>
and <link linkend="functions-string-sql"><literal>SUBSTRING</></>
(Jeff Davis)
@ -1804,8 +1846,8 @@ FIXME: Add more specifics?
</para>
<para>
Now, role membership is sufficient; previously only the same
role could perform these operations.
Previously, only the specific role owning the target session could
perform these operations; now membership in that role is sufficient.
</para>
</listitem>
@ -1951,7 +1993,7 @@ FIXME: Add more specifics?
2015-03-25 [a4847fc] Tom Lane: Add an ASSERT statement in plpgsql.
-->
<para>
Add <link linkend="plpgsql-statements-assert"><command>ASSERT</></>
Add an <link linkend="plpgsql-statements-assert"><command>ASSERT</></>
statement in <application>PL/pgSQL</> (Pavel Stehule)
</para>
</listitem>
@ -2054,6 +2096,17 @@ FIXME: Add more specifics?
<listitem>
<!--
2015-11-12 [5094da9] Alvaro..: vacuumdb: don't prompt for passwords over and ..
-->
<para>
In <link linkend="APP-VACUUMDB"><application>vacuumdb</></>, do not
prompt for the same password repeatedly when multiple connections
are necessary (Haribabu Kommi, Michael Paquier)
</para>
</listitem>
<listitem>
<!--
2015-05-15 [458a077] Fujii ..: Support &#8211;&#8211;verbose option in reindexdb.
-->
<para>
@ -2107,13 +2160,13 @@ FIXME: Add more specifics?
2014-07-10 [5b214c5] Fujii ..: Add new ECHO mode 'errors' that displays only ..
-->
<para>
Add <literal>errors</> mode to <application>psql</>'s
<varname>ECHO</> variable to display only failed commands
Add an <literal>errors</> mode that displays only failed commands
to <application>psql</>'s <varname>ECHO</> variable
(Pavel Stehule)
</para>
<para>
This can also be enabled with the <application>psql</>
This behavior can also be selected with <application>psql</>'s
<option>-b</> option.
</para>
</listitem>
@ -2168,6 +2221,22 @@ FIXME: Add more specifics?
<listitem>
<!--
2015-12-03 [07338cb] Tom Lane: Clean up some psql issues around handling of t..
2015-12-08 [e90371d] Tom Lane: Make failure to open psql log-file fatal.
-->
<para>
<application>psql</> now fails if the file specified by
an <option>--output</> or <option>--log-file</> switch cannot be
written (Tom Lane, Daniel Verite)
</para>
<para>
Previously, it effectively ignored the switch in such cases.
</para>
</listitem>
<listitem>
<!--
2014-07-12 [bd40951] Andres..: Minimal psql tab completion support for SET se..
-->
<para>
@ -2207,10 +2276,12 @@ FIXME: Add more specifics?
</para>
<para>
<literal>variables</> shows <application>psql</>'s special variables
and <literal>options</> shows command-line options.
<command>\? commands</> is the default output. This help
information can also be output via <literal>--help=section</>.
<literal>\? variables</> shows <application>psql</>'s special
variables and <literal>\? options</> shows the command-line options.
<command>\? commands</> shows the meta-commands, which is the
traditional output and remains the default. These help displays
can also be obtained with the command-line
option <literal>--help=<replaceable>section</></literal>.
</para>
</listitem>
@ -2245,7 +2316,7 @@ FIXME: Add more specifics?
<para>
Also prevent <option>--echo-hidden</> from echoing
<command>\watch</> queries.
<command>\watch</> queries, since that is generally unwanted.
</para>
</listitem>
@ -2254,8 +2325,8 @@ FIXME: Add more specifics?
2014-11-22 [eca2b9b] Andrew..: Rework echo_hidden for \sf and \ef from commit ..
-->
<para>
Allow <application>psql</>'s <literal>\sf</> and <literal>\ef</>
to honor <envar>ECHO_HIDDEN</> (Andrew Dunstan)
Make <application>psql</>'s <literal>\sf</> and <literal>\ef</>
commands honor <envar>ECHO_HIDDEN</> (Andrew Dunstan)
</para>
</listitem>
@ -2310,10 +2381,26 @@ FIXME: Add more specifics?
<listitem>
<!--
2015-11-21 [5f5e68b] Tom Lane: Adopt the GNU convention for handling tar-arch..
-->
<para>
Support table sizes exceeding 8GB in tar archive format (Tom Lane)
</para>
<para>
The POSIX standard for tar format does not allow elements of a tar
archive to exceed 8GB, but most modern implementations of tar
support an extension that does allow it. Use the extension format
when necessary, rather than failing.
</para>
</listitem>
<listitem>
<!--
2014-07-07 [7700597] Tom Lane: In pg_dump, show server and pg_dump versions w..
-->
<para>
Always have <application>pg_dump</> print server and
Make <application>pg_dump</> always print the server and
<application>pg_dump</> versions (Jing Wang)
</para>
@ -2348,8 +2435,8 @@ FIXME: Add more specifics?
2014-08-25 [ebe30ad] Bruce ..: pg_ctl, pg_upgrade: allow multiple -o/-O opti..
-->
<para>
Allow multiple <application>pg_ctl</> <option>-o</> options to
be appended (Bruce Momjian)
Support multiple <application>pg_ctl</> <option>-o</> options,
concatenating their values (Bruce Momjian)
</para>
</listitem>
@ -2368,6 +2455,23 @@ FIXME: Add more specifics?
</para>
</listitem>
<listitem>
<!--
2015-11-08 [bdb42ba] Noah M..: Don't connect() to a wildcard address in test_..
-->
<para>
If the server's listen address is set to a wildcard value
(<literal>0.0.0.0</> in IPv4 or <literal>::</> in IPv6), connect via
the loopback address rather than trying to use the wildcard address
literally (Kondo Yuta)
</para>
<para>
This fix primarily affects Windows, since on other platforms
<application>pg_ctl</> will prefer to use a Unix-domain socket.
</para>
</listitem>
</itemizedlist>
</sect4>
@ -2398,8 +2502,9 @@ FIXME: Add more specifics?
2014-08-25 [ebe30ad] Bruce ..: pg_ctl, pg_upgrade: allow multiple -o/-O optio..
-->
<para>
Allow multiple <application>pg_upgrade</>
<option>-o</>/<option>-O</> options to be appended (Bruce Momjian)
Support multiple <application>pg_upgrade</>
<option>-o</>/<option>-O</> options,
concatenating their values (Bruce Momjian)
</para>
</listitem>
@ -2415,17 +2520,6 @@ FIXME: Add more specifics?
<listitem>
<!--
2015-03-18 [417f78a] Bruce ..: pg_upgrade: document use of rsync for slave up..
-->
<para>
Document the use of <application>rsync</> for standby server
upgrades using <application>pg_upgrade</> (Stephen Frost,
Bruce Momjian)
</para>
</listitem>
<listitem>
<!--
2014-08-25 [2209b39] Bruce ..: pg_upgrade: remove support for 8.3 old clusters
-->
<para>
@ -2480,7 +2574,7 @@ FIXME: Add more specifics?
</para>
<para>
This is controlled by new <option>--latency-limit</> option.
This is controlled by a new <option>--latency-limit</> option.
</para>
</listitem>
@ -2497,11 +2591,12 @@ FIXME: Add more specifics?
<listitem>
<!--
2015-03-02 [878fdcb] Robert..: pgbench: Add a real expression syntax to \set
FIXME: Improve description, link
-->
<para>
Allow <application>pgbench</>'s <command>\set</> command to handle
multi-operator expressions (Robert Haas, Fabien Coelho)
arithmetic expressions containing more than one operator, and add
<literal>%</> (modulo) to the set of operators it supports
(Robert Haas, Fabien Coelho)
</para>
</listitem>
@ -2698,8 +2793,8 @@ FIXME: Improve description, link
-->
<para>
Allow higher-precision time stamp resolution on <systemitem
class="osname">Windows 8</> or <systemitem class="osname">Windows
Server 2012</> and later Windows systems (Craig Ringer)
class="osname">Windows 8</>, <systemitem class="osname">Windows
Server 2012</>, and later Windows systems (Craig Ringer)
</para>
</listitem>