<!-- $PostgreSQL: pgsql/doc/src/sgml/release-8.4.sgml,v 1.3 2009/05/11 09:11:41 mha Exp $ -->

<!-- $PostgreSQL: pgsql/doc/src/sgml/release-8.4.sgml,v 1.4 2009/05/12 23:43:50 tgl Exp $ -->

<!-- See header comment in release.sgml about typical markup -->

<sect1 id="release-8-4">

<title>Release 8.4</title>

<note>

<title>Release date</title>

<simpara>2009-??-??, ITEMS CURRENT AS OF 2009-04-08</simpara>

<simpara>2009-??-??, ITEMS CURRENT AS OF 2009-05-11</simpara>

</note>

<sect2>

</para>

</listitem>

<listitem>

<para>

Change default setting for <literal>max_prepared_transactions</> to

zero (previously it was 5) (Tom)

</para>

</listitem>

<listitem>

<para>

Make <literal>debug_print_parse</>, <literal>debug_print_rewritten</>,

<title>Authentication and security</title>

<itemizedlist>

<listitem>

<para>

Report appropriate error message for combination of <literal>MD5</>

authentication and <varname>db_user_namespace</> enabled (Bruce)

</para>

</listitem>

<listitem>

<para>

Remove support for the (insecure) <literal>crypt</> authentication method

commercial CAs.

</para>

</listitem>

<listitem>

<para>

Report appropriate error message for combination of <literal>MD5</>

authentication and <varname>db_user_namespace</> enabled (Bruce)

</para>

</listitem>

</itemizedlist>

</sect4>

<listitem>

<para>

so that errors are reported immediately (Magnus)

syntax (Magnus)

</para>

<para>

Previously, most errors in the file wouldn't be detected until clients

tried to connect, so an erroneous file could render the system

unusable. With the new behavior, if an error is detected during

reload then the bad file is rejected and the postmaster continues

to use its old copy.

</para>

</listitem>

<listitem>

<para>

Show all parsing errors in <filename>pg_hba.conf</> instead of

aborting after the first one (Selena Deckelmann)

This makes incompatible changes to the <literal>ldap</>,

<literal>pam</> and <literal>ident</> authentication methods. All

<filename>pg_hba.conf</> entries with these methods need to be

rewritten using the new format.

</para>

</listitem>

</para>

</listitem>

<listitem>

<para>

Change all authentication options to use <literal>name=value</>

syntax (Magnus)

</para>

<para>

This makes incompatible changes to the <literal>ldap</>,

<literal>pam</> and <literal>ident</> authentication methods. All

<filename>pg_hba.conf</> entries with these methods need to be

rewritten using the new format.

</para>

</listitem>

<listitem>

<para>

Allow a usermap parameter for all external authentication methods

</para>

</listitem>

<listitem>

<para>

Parse <filename>pg_hba.conf</> fully when it is loaded,

so that errors are reported immediately (Magnus)

</para>

<para>

Previously, most errors in the file wouldn't be detected until clients

tried to connect, so an erroneous file could render the system

unusable. With the new behavior, if an error is detected during

reload then the bad file is rejected and the postmaster continues

to use its old copy.

</para>

</listitem>

<listitem>

<para>

Show all parsing errors in <filename>pg_hba.conf</> instead of

aborting after the first one (Selena Deckelmann)

</para>

</listitem>

<listitem>

<para>

Support <literal>ident</> authentication over Unix-domain sockets

</para>

</listitem>

<listitem>

<para>

Reject <literal>\000</> in string literals and <command>COPY</> data

(Tom)

</para>

<para>

Previously, this was accepted but had the effect of terminating

the string contents.

</para>

</listitem>

<listitem>

<para>

Improve the parser's ability to report error locations (Tom)

<para>

Improve checks that the database encoding, collation

(<varname>LC_COLLATE</>), and character classes

(<varname>LC_CTYPE</>) match (Heikki)

(<varname>LC_CTYPE</>) match (Heikki, Tom)

</para>

<para>

Note in particular that a new database's encoding and locale

settings can be changed only when copying from <literal>template0</>.

This prevents possibly copying data that doesn't match the settings.

</para>

</listitem>

</para>

</listitem>

<listitem>

<para>

Make <command>EXIT</> without a label always exit the innermost

loop (Tom)

</para>

<para>

Formerly, if there were a <literal>BEGIN</> block more closely nested

than any loop, it would exit that block instead. The new behavior

matches Oracle(TM) and is also what was previously stated by our own

documentation.

</para>

</listitem>

<listitem>

<para>

Make processing of string literals and nested block comments

match the main SQL parser's processing (Tom)

</para>

</listitem>

<listitem>

<para>

Avoid memory leakage when the same function is called at varying

</para>

</listitem>

<listitem>

<para>

Add a function type column to <command>\df</>'s output, and add

options to list only selected types of functions (David Fetter)

</para>

</listitem>

<listitem>

<para>

Make <command>\df</> not hide functions that take or return

<listitem>

<para>

Make Kerberos use the same method to determine the username of the

client as all other authentication methods (Magnus)

Make Kerberosconnectionsuse the same method to determine the

username of theclient as all other authentication methods (Magnus)

</para>

<para>

and the name of the server when making <acronym>SSL</>

connections. If a root certificate is not available to use for

verification, <acronym>SSL</> connections will fail. The

<literal>sslmode</> parameter is used to enable the certificate

verification and set the level.

</para>

<para>

<literal>sslmode</> parameter is used to enable certificate

verification and set the level of checking.

The default is still not to do any verification, allowing connections

to SSLenabled servers without requiring a root certificate on the

to SSL-enabled servers without requiring a root certificate on the

client.

</para>

</listitem>

</para>

</listitem>

<listitem>

<para>

Recover better if dynamically-loaded code executes <function>exit()</>

(Tom)

</para>

</listitem>

<listitem>

<para>

Add a hook to let plug-ins monitor the executor (Itagaki

</para>

</listitem>

<listitem>

<para>

Make <filename>contrib/pgbench</> use table names

<structname>pgbench_accounts</>, <structname>pgbench_branches</>,

<structname>pgbench_history</>, and <structname>pgbench_tellers</>,

rather than just <structname>accounts</>, <structname>branches</>,

<structname>history</>, and <structname>tellers</> (Tom)

</para>

<para>

This is to reduce the risk of accidentally destroying real data

by running <application>pgbench</>.

</para>

</listitem>

<listitem>

<para>

Fix <filename>contrib/pgstattuple</> to handle tables and