== PostgreSQL Weekly News - January 03 2010 ==
RPM packages for 8.5alpha3 are out, thanks to Devrim's hard work. Test!
== PostgreSQL Product News ==
Benetl 3.3, an ETL tool for files using PostgreSQL, released.
Dataware Quick Application Suite 1.1, a RAD tool which supports
Npgsql 2.0.8, a .NET data provider for PostgreSQL, released.
psqlODBC 08.04.0200, an ODBC driver for PostgreSQL, released.
== PostgreSQL 8.5 Feature of the Week ==
In psql, \d now shows how many inherited tables a parent has, and \d+
== PostgreSQL Jobs for January ==
== PostgreSQL Local ==
The New York Capital District PUG's first meeting will be January 7,
Stefan Keller will be teach a course called, "Introduction to
PostGIS/PostgreSQL" at the University of Applied Sciences Rapperswil
(HSR) on January 14, 2010 in Zurich, Switzerland. Details on this and
other upcoming events below.
Seeking booth staffers for Open Day Linux Conf AU in Wellington, NZ
January 23, 2010, 11-2pm, Contact selena -at- postgresql.org to
The Call for Papers for East is open until January 30, 2010. Details at:
PGDay-Cuba 2010 will be held 4-6 February 2010, in La Havana at the
University of Science Informatic. Attendees from Cuba, the Dominican
Republic and Venezuela are invited.
FOSDEM 2010 will be in Brussels, Belgium on February 6-7, 2010.
The German PostgreSQL User Group has a stand at the Open Source Park
at Cebit 2010 from March 02 to March 06, 2010 in Hannover, Germany.
Chemnitzer Linuxtage will be in Chemnitz, Germany on March 13-14, 2010.
PgCon 2010 will be held May 20-21 2010 in Ottawa with tutorials before
on the 18th and 19th. The RfP is open!
The Call for Proposals for OSCON is open. OSCON will take place in
Portland, Oregon July 19-23, 2010.
== PostgreSQL in the News ==
Planet PostgreSQL: http://planet.postgresql.org/
PostgreSQL Weekly News is brought to you this week by David Fetter
Submit news and announcements by Sunday at 3:00pm Pacific time.
Please send English language ones to email@example.com, German language
to firstname.lastname@example.org, Italian language to email@example.com.
== Reviews ==
Robert Haas reviewed Pavel Stehule's patch to enable single- and
double-quoting behavior for psql variables.
Robert Haas reviewed ITAGAKI Takahiro's patch to add buffer usage
columns to contrib/pg_stat_statements.
Robert Haas reviewed the patch implementing red-black tree for GIN by
Jaime Casanova reviewed Robert Haas's patch to add per-tablespace
random_page_cost and seq_page_cost settings.
== Applied Patches ==
Bruce Momjian committed:
- In pgsql/src/backend/port/win32/mingwcompat.c, remove non-ascii
characters from source code.
- Add PGDLLIMPORT for binary_upgrade global variables so shared object
libraries can access them.
- Remove PGDLLIMPORT used for binary upgrade; must be on the externs,
Peter Eisentraut committed:
- In pgsql/doc/src/sgml/plpgsql.sgml, small wording improvement and
clarification in PL/pgSQL trigger documentation
- In pgsql/doc/src/sgml/Makefile, revert makefile refactoring (version
1.123) because it doesn't work when building several files at once
(e.g., gmake postgres-A4.pdf postgres-US.pdf).
- In pgsql/src/backend/catalog/sql_features.txt, update SQL
conformance: search conditions on triggers are supported
- Fill in information schema column for trigger WHEN condition.
- Add information_schema.triggered_update_columns. This reflects the
recently added support for triggers on columns.
- In pgsql/src/backend/catalog/sql_features.txt, update SQL features
list for aggregate ORDER BY support
- In pgsql/src/backend/catalog/sql_features.txt, update SQL features
Heikki Linnakangas committed:
- Previous fix for temporary file management broke returning a set
from PL/pgSQL function within an exception handler. Make sure we
use the right resource owner when we create the tuplestore to hold
returned tuples. Simplify tuplestore API so that the caller doesn't
need to be in the right memory context when calling tuplestore_put*
functions. tuplestore.c automatically switches to the memory
context used when the tuplestore was created. Tuplesort was already
modified like this earlier. This patch also removes the now useless
MemoryContextSwitch calls from callers. Report by Aleksei on
pgsql-bugs on December 22, 2009. Backpatch to 8.1, like the
previous patch that broke this.
- In, pgsql/contrib/tablefunc/tablefunc.c, oops, previous backpatch
- In pgsql/contrib/tablefunc/tablefunc.c, remove a now unused local
Tom Lane committed:
- Add the ability to store inheritance-tree statistics in
pg_statistic, and teach ANALYZE to compute such stats for tables
that have subclasses. Per my proposal of yesterday. autovacuum
still needs to be taught about running ANALYZE on parent tables when
their subclasses change, but the feature is useful even without
- Add an index on pg_inherits.inhparent, and use it to avoid seqscans
in find_inheritance_children(). This is a complete no-op in
databases without any inheritance. In databases where there are
just a few entries in pg_inherits, it could conceivably be a small
loss. However, in databases with many inheritance parents, it can
be a big win.
- Set errno to zero before invoking SSL_read or SSL_write. It appears
that at least in some Windows versions, these functions are capable
of returning a failure indication without setting errno. That puts
us into an infinite loop if the previous value happened to be EINTR.
Per report from Brendan Hill. Back-patch to 8.2. We could take it
further back, but since this is only known to be an issue on Windows
and we don't support Windows before 8.2, it does not seem worth the
- Revise pgstat's tracking of tuple changes to improve the reliability
of decisions about when to auto-analyze. The previous code depended
on n_live_tuples + n_dead_tuples - last_anl_tuples, where all three
of these numbers could be bad estimates from ANALYZE itself. Even
worse, in the presence of a steady flow of HOT updates and matching
HOT-tuple reclamations, auto-analyze might never trigger at all,
even if all three numbers are exactly right, because n_dead_tuples
could hold steady. To fix, replace last_anl_tuples with an
accurately tracked count of the total number of committed tuple
inserts + updates + deletes since the last ANALYZE on the table.
This can still be compared to the same threshold as before, but it's
much more trustworthy than the old computation. Tracking this
requires one more intra-transaction counter per modified table
within backends, but no additional memory space in the stats
collector. There probably isn't any measurable speed difference; if
anything it might be a bit faster than before, since I was able to
eliminate some per-tuple arithmetic operations in favor of adding
sums once per (sub)transaction. Also, simplify the logic around
pgstat vacuum and analyze reporting messages by not trying to fold
VACUUM ANALYZE into a single pgstat message. The original thought
behind this patch was to allow scheduling of analyzes on parent
tables by artificially inflating their changes_since_analyze count.
I've left that for a separate patch since this change seems to stand
on its own merit.
- In pgsql/src/backend/commands/analyze.c, dept of second thoughts:
recursive case in ANALYZE shouldn't emit a pgstats message. This
might need to be done differently later, but with the current logic
that's what should happen.
- Redefine Datum as uintptr_t, instead of unsigned long. This is more
in keeping with modern practice, and is a first step towards porting
to Win64 (which has sizeof(pointer) > sizeof(long)). Tsutomu
Yamada, Magnus Hagander, Tom Lane.
- In pgsql/src/backend/storage/ipc/procarray.c, add missing 'static'
- In pgsql/src/backend/storage/ipc/standby.c, suppress compiler
warning (pid_t isn't int everywhere).
- In pgsql/src/backend/libpq/auth.c, suppress compiler warning, per
buildfarm member narwhal.
- Support "x IS NOT NULL" clauses as indexscan conditions. This turns
out to be just a minor extension of the previous patch that made "x
IS NULL" indexable, because we can treat the IS NOT NULL condition
as if it were "x < NULL" or "x > NULL" (depending on the index's
NULLS FIRST/LAST option), just like IS NULL is treated like "x =
NULL". Aside from any possible usefulness in its own right, this is
an important improvement for index-optimized MAX/MIN aggregates: it
is now reliably possible to get a column's min or max value cheaply,
even when there are a lot of nulls cluttering the interesting end of
- Add an "argisrow" field to NullTest nodes, following a plan made way
back in 8.2beta but never carried out. This avoids repetitive tests
of whether the argument is of scalar or composite type. Also, be a
bit more paranoid about composite arguments in some places where we
previously weren't checking.
- check_exclusion_constraint didn't actually work correctly for index
expressions: FormIndexDatum requires the estate's scantuple to
already point at the tuple the values are supposedly being extracted
from. Adjust test case so that this type of confusion will be
exposed. Per report from hubert depesz lubaczewski.
- In pgsql/src/backend/utils/adt/regexp.c, fix similar_escape() to
convert parentheses to non-capturing style. This is needed to avoid
unwanted interference with SUBSTRING behavior, as per bug #5257 from
Roman Kononov. Also, add some basic intelligence about character
classes (bracket expressions) since we now have several behaviors
that aren't appropriate inside a character class. As with the
previous patch in this area, I'm reluctant to back-patch since it
might affect applications that are relying on the prior behavior.
- In pgsql/src/bin/psql/tab-complete.c, add missing
schema-qualification in tab completion query.
- In pgsql/src/backend/access/nbtree/nbtutils.c, dept of second
thoughts: my first cut at supporting "x IS NOT NULL" btree
indexscans would do the wrong thing if index_rescan() was called
with a NULL instead of a new set of scankeys and the index was DESC
order, because sk_strategy would not get flipped a second time. I
think that those provisions for a NULL argument are dead code now as
far as the core backend goes, but possibly somebody somewhere is
still using it. In any case, this refactoring seems clearer, and
it's definitely shorter.
Robert Haas committed:
- Reject invalid input in int2vectorin. Since the int2vector type is
intended only for internal use, this patch doesn't worry about
prettifying the error messages, which has the fringe benefit of
avoiding creating additional translatable strings. For a type
intended to be used by end-users, we would want to do better, but
the approach taken here seems like the correct trade-off for this
case. Caleb Welton.
Heikki Linnakangas committed:
- In pgsql/src/backend/access/transam/xlog.c, reset minRecoveryPoint
at checkpoints, so that we don't uselessly update it in the control
file at crash recovery following an archive recovery. Per Fujii
Masao and subsequent discussion.
Magnus Hagander committed:
- In pgsql/src/tools/msvc/Solution.pm, exclude part of the product
name string that is localized in Japanese versions of MSVC when
detecting MSVC version. Hiroshi Inoue.
- In pgsql/src/port/win32env.c, make the win32 putenv() override
update *all* present versions of the MSVCRxx runtime, not just the
current + Visual Studio 6 (MSVCRT). Clearly there can be an almost
unlimited number of runtimes loaded at the same time. Per report
from Hiroshi Inoue.
- In pgsql/src/tools/msvc/clean.bat, delete solution cache file on
clean, if it exists.
- Detect a 64-bit build environment on Windows, and generate the
appropriate project files. Based on the work of Tsutomu Yamada, but
- In pgsql/src/backend/utils/mmgr/aset.c, silence compiler warning on
64-bit windows build
- In pgsql/src/port/open.c, fix cast for _open_osfhandle(). Tsutomu
- Fix one more cast for _open_osfhandle(). Tsutomu Yamada.
- Support 64-bit shared memory when building on 64-bit Windows.
- In pgsql/src/include/pg_config.h.win32, set proper sizes for size_t
and void* on 64-bit Windows builds. Tsutomu Yamada.
- In pgsql/src/tools/msvc/Project.pm, silence compiler warning about
size of size_t being larger than the result variable it's stored in.
We know this can never happen. Per discussion.
- In pgsql/src/include/port/win32.h, make ssize_t 64-bit on Win64, for
compatibility with for example plpython.
Andrew Dunstan committed:
- Enable examples to compile on Windows. Patch from Hiroshi Saito.
ITAGAKI Takahiro committed:
- In pgsql/src/backend/port/win32_shmem.c, silence compiler warning
about printf format for HANDLE.
== Rejected Patches (for now) ==
No one was disappointed this week :-)
== Pending Patches ==
ITAGAKI Takahiro sent in a patch to make pg_bench error out better on
bad variable names.
Robert Haas sent in another revision of the patch to enable setting
random_page_cost and seq_page_cost at the tablespace level.
Zoltan Boszormenyi sent in a patch to add row counts to SELECT INTO
and CREATE TABLE AS [query].
Andreas Freund sent in a patch to delay fsync in database creation,
speeding up that operation.
Mark Cave-Ayland sent in two revisions of a patch to add subdirectory
support to DATA and DOCS in PGXS.
Hitoshi Harada sent in a patch to move the declaration of IntArray
from c.h to execQual.c, the only place it is currently used.
Pavel Stehule sent in a patch to add quote_ident and quote_literal,
using :"foo" and :'foo' respectively, to psql's variables.
Tatsuo Ishii sent in two patches to correct an issue where pgpool was
crashing PostgreSQL under certain conditions.
Heikki Linnakangas sent in a patch in response to Fujii Masao's patch
for streaming replication.
Teodor Sigaev sent in another revision of the k-nearest-neighbor
Teodor Sigaev sent in an updated patch to add point operations to
GiST. Robert Haas reviewed and revised this patch.
Hitoshi Harada sent in a patch to add ROWS|RANGE n PRECEDING|FOLLOWING
to windowing functions.
Simon Riggs sent in a patch which expands pg_cancel_backend()'s
functionality to allow it to cancel and idle transaction.
Gurjeet Singh sent in a patch to re-allow SET ROLE in SECURITY DEFINER
Simon Riggs sent in a patch to implement recovery conflict signalling
using SIGUSR1 multiplexing, then uses a SessionCancelPending mode
similar to Joachim Wieland's TransactionCancelPending patch.
John Naylor sent in another revision of the patch to refactor BKI.
Magnus Hagander sent in a patch to fix an issue in win64 where
socket definitions are not declared correctly.
KaiGai Kohei sent in another revision of the patch to disallow ALTER
COLUMN...RENAME TO... in the case where a column name is inherited
from multiple relations.
Magnus Hagander sent in a patch to change to config.pl processing in
the msvc build environment.
David Fetter sent in a patch to add tab completion for DO blocks in
ITAGAKI Takahiro sent in another revision of the patch to check
correctness of variable names in pgbench.