== PostgreSQL Weekly News - September 10 2006 ==

== PostgreSQL Weekly News - September 10 2006 ==

The Open Geospatial Consortium has awarded PostGIS a compliance
certificate.
http://postgis.refractions.net/pipermail/postgis-users/2006-September/013058.html

== PostgreSQL Product News ==

pgpool-II-1.0.0 is out, now supporting more than two servers and with
the new pgpoolAdmin tool written in PHP.
http://pgfoundry.org/projects/pgpool/

== PostgreSQL Jobs for September ==

http://archives.postgresql.org/pgsql-jobs/2006-09/threads.php

== PostgreSQL Local ==

Everything this week was global.

== PostgreSQL in the News ==

David Wheeler has published an article on PL/PgSQL at O'Reillynet
http://www.oreillynet.com/pub/a/databases/2006/09/07/plpgsql-batch-updates.html

Planet PostgreSQL: http://www.planetpostgresql.org/

General Bits, Archives and occasional new articles:
http://www.varlena.com/GeneralBits/

== Applied Patches ==

Bruce Momjian committed:

- Michael Glaesemann's patch which adds interval
division/multiplication regression tests.

- A Bruce Momjian/Michael Glaesemann collaboration which fixes an
imprecision from interval rounding of multiplication/division.

- A patch from Jun Kuwamura which updates the Japanese FAQ.

- A patch which updates the setseed() documentation.

- Gregory Stark's patch which doubles the speed of VACUUM on tables
with no indexes by preventing a second scan of such tables in that
case.

- A patch which adds some MSVC build tools.

- Greg Sabino Mullane's patch which fixes the bugs in the information
schema where sequences were not being shown due to the use of
lowercase 's' instead of 'S', and the views were not checking for
table visibility with regards to temporary tables and sequences.

- Simon Riggs' documentation patch which mentions that paremeterized
queries do not work with partial indexes.

- A patch which removes a trailing slash use in ecpg regression
script which caused failures on some platforms.

- A backpatch to 8.1.X fix for to_timestamp() where "PM/AM"
specification was eating too much user input, producing incorrect
results.

- Satoshi Nagayasu's patch which updates the Japanese README for
pgstattuple.

- Add Andrew Dunstan's sample.emacsrc and sample.vimrc to src/tools/

- Added XML documentation to the data types section of the docs.

- A patch which specifies lo_write() to take a _const_ buffer, to
match its documentation.

Tom Lane committed:

- A patch which fixes sslinfo so it builds on Darwin.

- A patch to sslinfo which silences a compiler warning about signed vs
unsigned chars.

- A patch to sslinfo which silences a compiler warning about
incompatible function pointer types.

- A patch which removes pgcrypto functions that had been deprecated
and slated for removal.

- A patch which gets rid of the separate RULE privilege for tables:
now only a table's owner can create or modify rules for the table.
This fixes a bug where ALTER TABLE OWNER would fail to update the
permission checking for associated rules.
Per discussion at
http://archives.postgresql.org/pgsql-hackers/2006-04/msg01138.php

- A patch which makes Gen_fmgrtab.sh locale-proof. Per report from
Marko Kreen and fix suggestion from Peter Eisentraut.

- A patch which removes contrib modules that have been migrated to
pgfoundry:

- A patch by Greg Mullane which fixes information_schema.key_column_usage
to show the correct value of position_in_unique_constraint (column
newly added per SQL2003).

- A patch which changes the sslinfo Makefile to work with VPATH and
pgxs builds.

- A patch which disallows TRUNCATE when there are any pending
after-trigger events for the target relation(s). Per report from
Markus Schiltknecht and subsequent discussion.

- A patch which changes ILIKE to invoke lower() and then do plain LIKE
comparison when working in a multibyte encoding. This fixes the
problems exhibited in bug #1931 and other reports of ILIKE
misbehavior in UTF8 encoding. This change is a stopgap until we
support locales better (ICU?).

- A patch which adds the sslinfo contrib module - information about
current SSL certificate Author: Victor Wagner <vitus(at)cryptocom(dot)ru>

- Itagaki Takahiro's patch which cleans up some leftover problems in
pgstattuple, namely removing unwanted and unportable elog(NOTICE)
report and fixing the install/uninstall sequence.

- Michael Glaesemann's fix to the interval input parser so that
fractional weeks and months are cascaded first to days and only what
is leftover into seconds. This seems to satisfy the principle of
least surprise given the general conversion to three-part interval
values. It was an oversight that these cases weren't dealt with in
8.1.

- Improvements to Susan Ebrecht's UPDATE SET (columnlist) patch. He
wasn't exactly delighted with the state it was committed in or with
the overall approach.

- Patches to fix the PL/Python documents.

- A patch to ecpg which suppress 'unused variable' warnings created by
a recent commit.

- A patch per suggestion by Andrew at Supernews which changes
processing of extended-Query mode so that an unnamed statement that
has parameters is always planned afresh for each Bind command,
treating the parameter values as constants in the planner. This
removes the performance penalty formerly often paid for using
out-of-line parameters. With this definition, the planner can do
constant folding, LIKE optimization, etc.

- Clean up logging for extended-query-protocol operations, as per his
recent proposal. Parameter logging works even for binary-format
parameters, and does not impose overhead when disabled.

- A fix to pg_dump for recent change removing separate RULE privilege,
which lets it recognize when to use ALL for table privileges in 8.2.

- A patch which puts back plan-time check for trying to apply SELECT
FOR UPDATE/SHARE to a relation on the nullable side of an outer
join.

- A change per proposal by Guillaume Smet to the behavior of
log_duration. Rather than being equivalent to setting
log_min_duration_statement to zero, this option now forces logging
of all query durations, but doesn't force logging of query text.
The patch also adds duration logging coverage for fastpath function
calls.

- Germán Méndez (Kronuz) Bravo's patch which replaced
contrib/isbn_issn with contrib/isn, just in time for EAN13, now with
updates.

- Albe Laurenz's patch which fixes linking of OpenLDAP libraries and
adds two macros LDAP_LIBS_FE and LDAP_LIBS_BE for frontend and
backend, respectively.

- A set of patches which renames contains/contained-by operators to @>
and <@, per discussion that agreed these symbols are less easily
confused than the previous versions.

- A cleaner solution to the AIX libpq linking problem, as per an
earlier discussion. This one centralizes assumptions about what
libpq depends on in one place in Makefile.global. Per how
discussion and howecpg does it.

- Remove unhelpful/misleading advice about how to use SPI_saveplan().
Per gripe from Jack Orenstein.

- Fix dblink build for --enable-nls or --enable-openssl on AIX.
Per Chris Browne.

- Add missing documentation for new anyarray-overlap operator.

- Per the above, fix them to work reliably for all arrays. Fix code
to handle NULLs and multidimensional arrays, move it into
arrayfuncs.c. GIN is still restricted to indexing arrays with no null
elements, however.

- Rename contrib contains/contained-by operators to @> and <@, per
discussion.

Teodor Sigaev committed:

- A patch which adds a new contrib module hstore, followed by one
which fixes markup and license.
http://archives.postgresql.org/pgsql-hackers/2006-08/msg01409.php

- A patch which fixes an Intel compiler bug.
http://archives.postgresql.org/pgsql-hackers/2006-08/msg01914.php

Michael Meskes committed:

- Various improvements to ecpg and its regression tests

- A fix to the test case which removes C++ style comments.

Peter Eisentraut committed:

- A patch which adds missing gettext calls for some SSL errors.

Joe Conway committed:

- A Kai Londenberg/Joe Conway patch which adds asynchronous queries
and parallel execution to dblink.

== Rejected Patches (for now) ==

Jie Zhang's bitmap indexes patch has been held back for PostgreSQL
8.3.

Per gripes by Tom Lane, Bruce Momjian reverted the change which caused
FETCH and MOVE to use 64-bit integers.

Per gripe about their unfinished state from author Christopher
Kings-Lynne, Bruce Momjian removed the GIN docs for a re-work.

== Pending Patches ==

Jeremy Drake sent in a patch which adds regression tests for large
objects.

Simon Riggs sent in a patch which documents restartable recovery mode.

PostgreSQL Weekly News is brought to you this week by David Fetter and
Dave Page.