Skip to content

Wöchentlicher PostgreSQL Newsletter - 26. April 2009


Der Originalartikel befindet sich unter:
http://www.postgresql.org/community/weeklynews/pwn20090426


== Wöchentlicher PostgreSQL Newsletter - 26. April 2009 ==

Die Spanisch-sprechende PostgreSQL Community hat jetzt ihre eigene
Webseite. Grüße!
http://www.postgresql-es.org/

== PostgreSQL Produkt Neuigkeiten ==

DBD::Pg 2.13.1, ein Perl Connector für PostgreSQL, ist erschienen.
http://search.cpan.org/dist/DBD-Pg/

pgFouine 1.1, ein PostgreSQL Log Processor und Report Generator, ist
erschienen.
http://pgfouine.projects.postgresql.org/

pgpool-II 2.2.1, ein Connection Pooler und mehr, ist erschienen.
http://pgfoundry.org/projects/pgpool/

Slony-ctl 1.0, ein Satz Skripte zum Erstellen und Verwalten von
Slony Clustern, ist erschienen.
http://pgfoundry.org/projects/slony1-ctl/

== PostgreSQL 8.4 Feature der Woche ==

pg_stat_functions: dieser neue Systemview zeigt, wie oft eine Funktion
aufgerufen und wieviel Zeit für die Ausführung benötigt wurde.

== PostgreSQL Tipp der Woche ==

PostgreSQL unterstützt Subselects in SELECT, FROM und WHERE Klauseln.
Denke darüber nach, wo du den Subselect einfügst, damit er vernünftig
ausgeführt wird und die gewünschten Ergebnisse liefert.

== PostgreSQL Jobs im April ==

http://archives.postgresql.org/pgsql-jobs/2009-04/threads.php

== PostgreSQL Lokal ==

PgDay Florianopolis wird am 22. Mai stattfinden. Kontaktiere Dickson S.
Guedes (guedes AT guedesoft DOT net) um teilzunehmen oder einen Vortrag
einzureichen.
http://www.postgresql.org.br/eventos/pgday/sc

Die Percona Performance Konferenz wird im Santa Clara Convention Center
in Santa Clara, Kalifornien, USA, stattfinden.
http://conferences.percona.com/

Der CfP für den PgDay in Sao Paulo am 24. April ist eröffnet.
Kontaktiere marins DOT consultoria AT gmail DOT com oder marcelojscosta
AT gmail DOT com wenn du teilnehmen möchtest.

PostgreSQL Conference U.S. (JDCon) wird einen PgDay auf dem Linuxfest
Nordwest am 25. und 26. April veranstalten. Der Call for Papers ist
unter:
http://www.postgresqlconference.org/

Es werden weiterhin PgDays am 29. April in Porto Velho, RO und am 20.
April in Ji-Parana, RO stattfinden. Kontaktiere Luis Fernando Bueno:
proflfbueno AT gmail DOT com wenn du teilnehmen möchtest.

Michael Renner wird einen Workshop über PostgreSQL Replikation auf
der Netways OSDC 2009 in Nürnbern, Deutschland geben.
http://www.netways.de/english/osdc/y2009/programm/w/michael_renner_postgresql_repliziert_ein_ueberblick/

PGCon 2009 wird vom 21. bis 22. Mai 2009 in Ottawa an der Universität
von Ottawa stattfinden. Dem gehen zwei Tage mit Tutorials vom 19. bis
20. Mai 2009 voraus.
http://www.pgcon.org/2009/

PgDay Florianopolis wird am 22. Mai stattfinden. Kontaktiere Dickson S.
Guedes (guedes AT guedesoft DOT net) um teilzunehmen oder einen Vortrag
einzureichen.
http://www.postgresql.org.br/eventos/pgday/sc

Termin vormerken: pgDay San Jose, am Sonntag dem 19. Juli unmittelbar
vor der OSCON. CfP und mehr Informationen bald.

PGCon Brasilien wird am 23.-24. Oktober 2009 auf dem Unicamp in
Campinas, Sao Paulo, stattfinden.

== PostgreSQL in den News ==

Planet PostgreSQL: http://planet.postgresql.org/

Dieser wöchentliche PostgreSQL Newsletter wurde erstellt von David
Fetter und Josh Berkus.

Sende Neuigkeiten und Ankündigungen bis Sonntag, 15 Uhr Pazifischer
Zeit. Bitte sende englische Beiträge an david@fetter.org, deutsche an
pwn@pgug.de, italienische an pwn@itpug.org.



== Angewandte Patches ==

Magnus Hagander committed:

- In pgsql/src/tools/msvc/Install.pm, read nls.nk files to determine
  the proper name of NLS catalogs to install.  Hiroshi Inoue, with
  minor modifications by me.

- Move gettext encoding names into encnames.c, so we only have one
  place to update.  Per discussion.

- Remove sslverify parameter again, replacing it with two new sslmode
  values: "verify-ca" and "verify-full".  Since "prefer" remains the
  default, this will make certificate validation off by default, which
  should lead to less upgrade issues.

- In pgsql/doc/src/sgml/release.sgml, remove sslverify parameter from
  release notes, mention that sslmode is used instead.

- In pgsql/doc/src/sgml/libpq.sgml, more clearly document what the
  different sslmode options mean, both the new and the old ones.
  Consistently talk about certificate verification, and not
  validation.

Bruce Momjian committed:

- Add agg/normal/trigger/window flags for psql \df and in \df output.
  David Fetter.

- Merge psql \df options into a single \? line, and update docs.

- Improve psql \df error handling.  David Fetter.

- In pgsql/src/bin/psql/help.c, use brackets in psql \df \? help.

- In pgsql/doc/src/sgml/runtime.sgml, improve server spoofing wording,
  per request from Magnus Hagander.

- Remove HELIOS Software GmbH name and copyright from AIX dynloader
  files, per approval from Helmut Tschemernjak, President.  Only back
  branches;  files removed from CVS HEAD.

Tom Lane committed:

- In pgsql/src/backend/port/dynloader/osf.c, fix obsolete
  cross-reference (this file isn't called alpha.c anymore)

- Remove the long-obsolete homebrew dl*() functions for AIX, in favor
  of just using the system functions all the time.  (These files are
  now just copies of the osf.* files.)  The homebrew functions were
  not getting used anyway on AIX versions that have dlopen(), that is
  4.3 and up, so they are not needed on any AIX that is even remotely
  supported by the vendor anymore.  We'd have probably left them here
  anyway, except some questions were raised about the copyright.

- Change the default value of max_prepared_transactions to zero, and
  add documentation warnings against setting it nonzero unless active
  use of prepared transactions is intended and a suitable transaction
  manager has been installed.  This should help to prevent the type of
  scenario we've seen several times now where a prepared transaction
  is forgotten and eventually causes severe maintenance problems (or
  even anti-wraparound shutdown).  The only real reason we had the
  default be nonzero in the first place was to support regression
  testing of the feature.  To still be able to do that, tweak
  pg_regress to force a nonzero value during "make check".  Since we
  cannot force a nonzero value in "make installcheck", add a variant
  regression test "expected" file that shows the results that will be
  obtained when max_prepared_transactions is zero.  Also, extend the
  HINT messages for transaction wraparound warnings to mention the
  possibility that old prepared transactions are causing the problem.
  All per today's discussion.

- In pgsql/src/backend/commands/dbcommands.c, don't use the result of
  strcmp as if it were a boolean.  A service of your local coding
  style police.

- In pgsql/src/backend/libpq/ip.c, suppress some 'variable may be used
  uninitialized' warnings from gcc 4.4.

- In pgsql/src/backend/parser/parse_func.c, fix some more 'variable
  may be used uninitialized' warnings from gcc 4.4.

- Fix the handling of sub-SELECTs appearing in the arguments of an
  outer-level aggregate function.  By definition, such a sub-SELECT
  cannot reference any variables of query levels between itself and
  the aggregate's semantic level (else the aggregate would've been
  assigned to that lower level instead).  So the correct, most
  efficient implementation is to treat the sub-SELECT as being a
  sub-select of that outer query level, not the level the aggregate
  syntactically appears in.  Not doing so also confuses the heck out
  of our parameter-passing logic, as illustrated in bug report from
  Daniel Grace.  Fortunately, we were already copying the whole Aggref
  expression up to the outer query level, so all that's needed is to
  delay SS_process_sublinks processing of the sub-SELECT until control
  returns to the outer level.  This has been broken since we
  introduced spec-compliant treatment of outer aggregates in 7.4; so
  patch all the way back.

- Update citext expected output to exactly match the real output,
  rather than having some whitespace discrepancy.  Although whitespace
  is supposed to be ignored in our regression tests, for some reason
  buildfarm member spoonbill doesn't like it.

- In pgsql/src/bin/psql/copy.c, remove Windows-specific definition of
  S_ISDIR().  This should not be here; if there are any Windows
  configurations where port/win32.h fails to provide the macro, it
  should be fixed in the latter file not here.

- In pgsql/doc/src/sgml/release.sgml, some more work on the 8.4
  release notes.  Document a few changes that Bruce apparently found
  uninteresting, and do minor wordsmithing on a number of the existing
  entries.

Heikki Linnakangas committed:

- In pgsql/src/backend/access/transam/xlog.c, after archive recovery,
  mark the last WAL segment from the parent timeline ready for
  archival. It was marked at the next checkpoint anyway, but waiting
  for the next checkpoint is an unnecessary delay.  Fujii Masao.

- varstr_cmp and any comparison function that piggybacks on it can
  return any negative or positive number, not just -1 or 1. Fix
  comment on varstr_cmp and citext test case accordingly.  As pointed
  out by Zdenek Kotala, and buildfarm member gothic moth.

- In pgsql/doc/src/sgml/maintenance.sgml, improve the documentation on
  8.4 visibility map related VACUUM changes.  Explain how
  vacuum_freeze_table_age should be tuned, and how it relates to the
  other settings. Mention that vacuum_freeze_table_age also affects
  when autovacuum scans the whole table.

- In pgsql/doc/src/sgml/storage.sgml, mention that tables have a
  visibility map fork alongside the main fork and FSM.

== Abgelehnte Patches (bis jetzt) ==

No one was disappointed this week :-)

== Eingesandte Patches ==

Pavel Stehule sent in another revision of his patch to add
transformation hooks into the parser.

Fujii Masao sent in two more revisions of the stats_temp_directory
patch.

Hiroshi Inoue sent in a patch which adds some encoding conversion
logic on Windows.

Fujii Masao sent in another revision of the pg_standby trigger patch.

Tom Lane sent in a patch to fix a performance regression in hash
joins.

Robert Haas sent in an experimental patch to change NTUP_PER_BUCKET
into a GUC called hash_load, and add EXPLAIN support to show the
number of buckets and batches.

Brendan Jurd sent in an updated version of Pavel Stehule's patch to
add scientific notation to to_char().

Hannu Valtonen sent in a patch to add support for integer and text
arrays to PL/PythonU.

  • Twitter
  • Bookmark Wöchentlicher PostgreSQL Newsletter - 26. April 2009 at del.icio.us
  • Facebook
  • Google Bookmarks
  • FriendFeed
  • Digg Wöchentlicher PostgreSQL Newsletter - 26. April 2009
  • Bloglines Wöchentlicher PostgreSQL Newsletter - 26. April 2009
  • Technorati Wöchentlicher PostgreSQL Newsletter - 26. April 2009
  • Fark this: Wöchentlicher PostgreSQL Newsletter - 26. April 2009
  • Bookmark Wöchentlicher PostgreSQL Newsletter - 26. April 2009 at YahooMyWeb
  • Bookmark Wöchentlicher PostgreSQL Newsletter - 26. April 2009 at Furl.net
  • Bookmark Wöchentlicher PostgreSQL Newsletter - 26. April 2009 at reddit.com
  • Bookmark Wöchentlicher PostgreSQL Newsletter - 26. April 2009 at blinklist.com
  • Bookmark Wöchentlicher PostgreSQL Newsletter - 26. April 2009 at Spurl.net
  • Bookmark Wöchentlicher PostgreSQL Newsletter - 26. April 2009 at Simpy.com
  • Bookmark Wöchentlicher PostgreSQL Newsletter - 26. April 2009 at blogmarks
  • Bookmark Wöchentlicher PostgreSQL Newsletter - 26. April 2009 with wists
  • wong it!
  • Bookmark using any bookmark manager!
  • Stumble It!
  • Identi.ca

Trackbacks

No Trackbacks

Comments

Display comments as Linear | Threaded

chrislay on :

COMMENT_DELETED
Comments ()

Add Comment

Enclosing asterisks marks text as bold (*word*), underscore are made via _word_.
E-Mail addresses will not be displayed and will only be used for E-Mail notifications.
To leave a comment you must approve it via e-mail, which will be sent to your address after submission.
Form options