1- <!-- $PostgreSQL: pgsql/doc/src/sgml/release.sgml,v 1.525 2007/10/1816:47:58 momjian Exp $ -->
1+ <!-- $PostgreSQL: pgsql/doc/src/sgml/release.sgml,v 1.526 2007/10/1823:05:06 momjian Exp $ -->
22<!--
33
44Typical markup:
@@ -66,8 +66,8 @@ do it for earlier branch release files.
6666
6767 <listitem>
6868 <para>
69- Support for the SQL/XML standard, including new operators and a
70- new <type>XML</type>builtin type
69+ Support for the SQL/XML standard, including new operators and
70+ an <type>XML</type>data type
7171 </para>
7272 </listitem>
7373
@@ -79,8 +79,7 @@ do it for earlier branch release files.
7979
8080 <listitem>
8181 <para>
82- Universally Unique Identifier (<type>UUID</>) data type, similar
83- to that defined by RFC 4122
82+ Universally Unique Identifier (<type>UUID</>) data type
8483 </para>
8584 </listitem>
8685
@@ -171,28 +170,27 @@ do it for earlier branch release files.
171170
172171 <listitem>
173172 <para>
174- <quote>Distributed</> checkpoints prevent I/O spikes during
175- checkpoints
173+ Distributed checkpoints prevent I/O spikes during checkpoints
176174 </para>
177175 </listitem>
178176
179177 <listitem>
180178 <para>
181- Heap-Only Tuples (<acronym>HOT</>)improves <command>UPDATE</>
182- spaceusage
179+ Heap-Only Tuples (<acronym>HOT</>)accelerate <command>UPDATE</>
180+ spacereuse
183181 </para>
184182 </listitem>
185183
186184 <listitem>
187185 <para>
188- Just-in-time background writer strategyto improve disk write
186+ Just-in-time background writer strategyimproves disk write
189187 efficiency
190188 </para>
191189 </listitem>
192190
193191 <listitem>
194192 <para>
195- Reduction ofboth per-field and per-row storage requirements
193+ Reduction of per-field and per-row storage requirements
196194 </para>
197195 </listitem>
198196
@@ -218,7 +216,8 @@ do it for earlier branch release files.
218216
219217 <listitem>
220218 <para>
221- Use pseudo-transaction ids in read-only transactions
219+ Reduce need for vacuum by using pseudo-transaction ids in
220+ read-only transactions
222221 </para>
223222 </listitem>
224223
@@ -507,7 +506,7 @@ do it for earlier branch release files.
507506
508507 <listitem>
509508 <para>
510- <quote> Distributed</> checkpoints prevent I/O spikes during
509+ Distributed checkpoints prevent I/O spikes during
511510 checkpoints (Itagaki Takahiro and Heikki Linnakangas)
512511 </para>
513512
@@ -522,8 +521,8 @@ do it for earlier branch release files.
522521
523522 <listitem>
524523 <para>
525- Heap-Only Tuples (<acronym>HOT</>)improves <command>UPDATE</>
526- spaceusage (Pavan Deolasee, with ideas from many others)
524+ Heap-Only Tuples (<acronym>HOT</>)accelerate <command>UPDATE</>
525+ spacereuse (Pavan Deolasee, with ideas from many others)
527526 </para>
528527
529528 <para>
@@ -540,7 +539,7 @@ do it for earlier branch release files.
540539
541540 <listitem>
542541 <para>
543- Just-in-time background writer strategyto improve disk write
542+ Just-in-time background writer strategyimproves disk write
544543 efficiency (Greg Smith, Itagaki Takahiro)
545544 </para>
546545
@@ -551,8 +550,8 @@ do it for earlier branch release files.
551550
552551 <listitem>
553552 <para>
554- Reduction ofboth per-field and per-row storage requirements
555- (Greg Stark)
553+ Reduction of per-field and per-row storage requirements (Greg
554+ Stark)
556555 </para>
557556
558557 <para>
@@ -565,21 +564,23 @@ do it for earlier branch release files.
565564
566565 <listitem>
567566 <para>
568- Use pseudo-transaction ids in read-only transactions (Florian Pflug)
567+ Reduce need for vacuum by using pseudo-transaction ids in
568+ read-only transactions (Florian Pflug)
569569 </para>
570570
571571 <para>
572- This reduces transaction overhead for read-only transactions,
573- and reduces the necessity of vacuuming for transaction id
574- wrap-around. Other transaction performance improvements were
575- also made that should improve concurrency.
572+ Pseudo-transaction ids do not increment the global transaction
573+ counter. Therefore, they do not add to the need for vacuum to
574+ read all database rows to prevent problems with transaction id
575+ wrap-around. Other transaction performance improvements were also
576+ made that should improve concurrency.
576577 </para>
577578 </listitem>
578579
579580 <listitem>
580581 <para>
581582 Create a dedicated <acronym>WAL</> writer process to off-load
582- WAL-writing work from backends (Simon)
583+ work from backends (Simon)
583584 </para>
584585 </listitem>
585586
@@ -619,12 +620,6 @@ do it for earlier branch release files.
619620 </para>
620621 </listitem>
621622
622- <listitem>
623- <para>
624- Reduce overhead of populating the statistics tables. (Tom)
625- </para>
626- </listitem>
627-
628623 <listitem>
629624 <para>
630625 Allow <literal>ORDER BY ... LIMIT</> to be done without sorting
@@ -637,6 +632,12 @@ do it for earlier branch release files.
637632 </para>
638633 </listitem>
639634
635+ <listitem>
636+ <para>
637+ Reduce overhead of populating the statistics tables (Tom)
638+ </para>
639+ </listitem>
640+
640641 <listitem>
641642 <para>
642643 Improve hash join performance for cases with many NULLs (Tom)
@@ -1327,8 +1328,8 @@ do it for earlier branch release files.
13271328
13281329 <listitem>
13291330 <para>
1330- Support for the SQL/XML standard, including new operators anda
1331- new <type>XML</type>builtin type (Nikolay Samokhvalov, Peter)
1331+ Support for the SQL/XML standard, including new operators andan
1332+ <type>XML</type>data type (Nikolay Samokhvalov, Peter)
13321333 </para>
13331334 </listitem>
13341335
@@ -1346,8 +1347,12 @@ do it for earlier branch release files.
13461347
13471348 <listitem>
13481349 <para>
1349- Universally Unique Identifier (<type>UUID</>) data type, similar
1350- to that defined by RFC 4122 (Gevik Babakhani, Neil)
1350+ Universally Unique Identifier (<type>UUID</>) data type (Gevik
1351+ Babakhani, Neil)
1352+ </para>
1353+
1354+ <para>
1355+ This closely matches <acronym>RFC</> 4122.
13511356 </para>
13521357 </listitem>
13531358