11<!--
2- $PostgreSQL: pgsql/doc/src/sgml/release.sgml,v 1.335 2005/05/05 17:59:40 tgl Exp $
2+ $PostgreSQL: pgsql/doc/src/sgml/release.sgml,v 1.336 2005/05/09 00:09:45 tgl Exp $
33-->
44
55<appendix id="release">
@@ -10,7 +10,7 @@ $PostgreSQL: pgsql/doc/src/sgml/release.sgml,v 1.335 2005/05/05 17:59:40 tgl Exp
1010
1111 <note>
1212 <title>Release date</title>
13- <simpara>2005-05-05 </simpara>
13+ <simpara>2005-05-09 </simpara>
1414 </note>
1515
1616 <para>
@@ -72,6 +72,17 @@ misuse</para></listitem>
7272<type>INTERNAL</> function results</para></listitem>
7373<listitem><para>Guard against incorrect second parameter to
7474<function>record_out</></para></listitem>
75+ <listitem><para>Repair ancient race condition that allowed a transaction to be
76+ seen as committed for some purposes (eg SELECT FOR UPDATE) slightly sooner
77+ than for other purposes</para>
78+ <para>This is an extremely serious bug since it could lead to apparent
79+ data inconsistencies being briefly visible to applications.</para></listitem>
80+ <listitem><para>Repair race condition between relation extension and
81+ VACUUM</para>
82+ <para>This could theoretically have caused loss of a page's worth of
83+ freshly-inserted data, although the scenario seems of very low probability.
84+ There are no known cases of it having caused more than an Assert failure.
85+ </para></listitem>
7586<listitem><para>Fix comparisons of <type>TIME WITH TIME ZONE</> values</para>
7687<para>
7788The comparison code was wrong in the case where the
@@ -2684,7 +2695,7 @@ typedefs (Michael)</para></listitem>
26842695
26852696 <note>
26862697 <title>Release date</title>
2687- <simpara>2005-05-05 </simpara>
2698+ <simpara>2005-05-09 </simpara>
26882699 </note>
26892700
26902701 <para>
@@ -2795,6 +2806,17 @@ UPDATE pg_database SET datallowconn = false WHERE datname = 'template0';
27952806misuse</para></listitem>
27962807<listitem><para>Change <filename>contrib/tsearch2</> to avoid unsafe use of
27972808<type>INTERNAL</> function results</para></listitem>
2809+ <listitem><para>Repair ancient race condition that allowed a transaction to be
2810+ seen as committed for some purposes (eg SELECT FOR UPDATE) slightly sooner
2811+ than for other purposes</para>
2812+ <para>This is an extremely serious bug since it could lead to apparent
2813+ data inconsistencies being briefly visible to applications.</para></listitem>
2814+ <listitem><para>Repair race condition between relation extension and
2815+ VACUUM</para>
2816+ <para>This could theoretically have caused loss of a page's worth of
2817+ freshly-inserted data, although the scenario seems of very low probability.
2818+ There are no known cases of it having caused more than an Assert failure.
2819+ </para></listitem>
27982820<listitem><para>Fix comparisons of <type>TIME WITH TIME ZONE</> values</para>
27992821<para>
28002822The comparison code was wrong in the case where the
@@ -5236,7 +5258,7 @@ DROP SCHEMA information_schema CASCADE;
52365258
52375259 <note>
52385260 <title>Release date</title>
5239- <simpara>2005-05-05 </simpara>
5261+ <simpara>2005-05-09 </simpara>
52405262 </note>
52415263
52425264 <para>
@@ -5313,6 +5335,17 @@ UPDATE pg_database SET datallowconn = false WHERE datname = 'template0';
53135335<itemizedlist>
53145336<listitem><para>Change encoding function signature to prevent
53155337misuse</para></listitem>
5338+ <listitem><para>Repair ancient race condition that allowed a transaction to be
5339+ seen as committed for some purposes (eg SELECT FOR UPDATE) slightly sooner
5340+ than for other purposes</para>
5341+ <para>This is an extremely serious bug since it could lead to apparent
5342+ data inconsistencies being briefly visible to applications.</para></listitem>
5343+ <listitem><para>Repair race condition between relation extension and
5344+ VACUUM</para>
5345+ <para>This could theoretically have caused loss of a page's worth of
5346+ freshly-inserted data, although the scenario seems of very low probability.
5347+ There are no known cases of it having caused more than an Assert failure.
5348+ </para></listitem>
53165349<listitem><para>Fix comparisons of <type>TIME WITH TIME ZONE</> values</para>
53175350<para>
53185351The comparison code was wrong in the case where the
@@ -6511,7 +6544,7 @@ operations on bytea columns (Joe)</para></listitem>
65116544
65126545 <note>
65136546 <title>Release date</title>
6514- <simpara>2005-05-05 </simpara>
6547+ <simpara>2005-05-09 </simpara>
65156548 </note>
65166549
65176550 <para>
@@ -6531,6 +6564,17 @@ operations on bytea columns (Joe)</para></listitem>
65316564 <title>Changes</title>
65326565
65336566<itemizedlist>
6567+ <listitem><para>Repair ancient race condition that allowed a transaction to be
6568+ seen as committed for some purposes (eg SELECT FOR UPDATE) slightly sooner
6569+ than for other purposes</para>
6570+ <para>This is an extremely serious bug since it could lead to apparent
6571+ data inconsistencies being briefly visible to applications.</para></listitem>
6572+ <listitem><para>Repair race condition between relation extension and
6573+ VACUUM</para>
6574+ <para>This could theoretically have caused loss of a page's worth of
6575+ freshly-inserted data, although the scenario seems of very low probability.
6576+ There are no known cases of it having caused more than an Assert failure.
6577+ </para></listitem>
65346578<listitem><para>Fix <function>EXTRACT(EPOCH)</> for
65356579<type>TIME WITH TIME ZONE</> values</para></listitem>
65366580<listitem><para>Additional buffer overrun checks in plpgsql
@@ -7133,7 +7177,7 @@ since <productname>PostgreSQL</productname> 7.1.
71337177<listitem><para>Make ALTER TABLE / RENAME COLUMN update column names of indexes (Brent Verner)</para></listitem>
71347178<listitem><para>Fix for ALTER TABLE / ADD CONSTRAINT ... CHECK with inherited tables (Stephan Szabo)</para></listitem>
71357179<listitem><para>ALTER TABLE RENAME update foreign-key trigger arguments correctly (Brent Verner)</para></listitem>
7136- <listitem><para>DROP AGGREGATE and COMMENT ON AGGREGATE nowaccepts an aggtype (Tom)</para></listitem>
7180+ <listitem><para>DROP AGGREGATE and COMMENT ON AGGREGATE nowaccept an aggtype (Tom)</para></listitem>
71377181<listitem><para>Add automatic return type data casting for SQL functions (Tom)</para></listitem>
71387182<listitem><para>Allow GiST indexes to handle NULLs and multikey indexes (Oleg Bartunov, Teodor Sigaev, Tom)</para></listitem>
71397183<listitem><para>Enable partial indexes (Martijn van Oosterhout)</para></listitem>