I mean, when the index is created over the text column, instead of just
indexing the text as-is, index the txt2txtidx'd version...?
For two reasons:
1. gist_txtidx_ops create with loss information (for less size), so any
operation with index must be checked
with original txtidx value. The way
Christopher Kings-Lynne writes:
Just out of interest, if someone was going to pay you to hack on Postgres
for 6 months, what would you like to code for 7.4?
Well judging by the hoards on Slashdot, it would appear that
replication is the hot enhancement...
Slashdot | PostgreSQL 7.3 Released
On 2 Dec 2002 at 12:05, Karel Zak wrote:
On Sat, Nov 30, 2002 at 01:11:20PM -0800, Joe Conway wrote:
As I said, this is all very preliminary; comments, suggestions, requests
are all welcome.
Only idea/dream: what implement dblink as virtual schema.
CREATE SCHEMA myschema AS DBLINK TO ...
Native Windows port
Plz don't forget poor victims of Microsoft
!!!
On Sat, Nov 30, 2002 at 01:11:20PM -0800, Joe Conway wrote:
As I said, this is all very preliminary; comments, suggestions, requests
are all welcome.
Only idea/dream: what implement dblink as virtual schema.
CREATE SCHEMA myschema AS DBLINK TO ... some connection options ...;
SELECT *
Am Sonntag, 1. Dezember 2002 06:47 schrieb Tom Lane:
Joe Conway [EMAIL PROTECTED] writes:
Someone asked earlier about how to change a bunch of existing tables int
the PUBLIC schema to some other schema. For grins I tried:
regression=# update pg_class set relnamespace=556829 where relname =
How about giving OLAP (Dimension / Measure) functionality to PG.
Catch all the cricket action. Download
Yahoo! Score tracker
On Fri, Nov 29, 2002 at 10:51:26AM -0800, Christopher Kings-Lynne wrote:
Hi guys,
Just out of interest, if someone was going to pay you to hack on Postgres
for 6 months, what would you like to code for 7.4?
This is interesting discussion..
my wish:
* error codes. It's very
-Original Message-
From: Karel Zak [mailto:[EMAIL PROTECTED]]
Sent: 02 December 2002 11:26
To: Christopher Kings-Lynne
Cc: [EMAIL PROTECTED]
Subject: Re: [HACKERS] 7.4 Wishlist
On Fri, Nov 29, 2002 at 10:51:26AM -0800, Christopher
Kings-Lynne wrote:
Hi guys,
Just out
my wish:
* error codes. It's very interesting that nobody other wants it...
I do :-)
Me too. It is a must in my opinion..
Regards,
Nic.
---(end of broadcast)---
TIP 2: you can get off all lists at once with the unregister command
Joe Conway wrote:
Christopher Kings-Lynne wrote:
possible. We should probably just go with your suggestion. Anything else
beyond the relnamespace and pg_depend entries that need to be dealt
with?
What about sequences for serial columns? What about views or types that
depend on the table?
Tom/Hackers,
Going back a bit, but relevant with 7.3's release...
Tom Lane writes on 03 Sep 2002:
Lee Kindness [EMAIL PROTECTED] writes:
[ original post was regarding the mileage in adding utility
functions to PostgreSQL to cut-out common catalog lookups, thus
making apps less
Hi,
Yet another problem with Turkish encoding. clean_encoding_name()
in src/backend/utils/mb/encnames.c uses tolower() to convert locale
names to lower-case. This causes errors if locale name contains
capital I and current olcale is Turkish. Some examples:
aaa=# \l
List of databases
Fernando Nasser wrote:
Why not just leave the sequence and types in the original schema and
make sure the table refers to them _there_? We just need to make sure
we have schema qualified references to the sequences and types.
Well, the type entry for the relation *is* related to just one
On Fri, 29 Nov 2002, Tom Lane wrote:
Peter Eisentraut [EMAIL PROTECTED] writes:
Has there been a release? It certainly hasn't been announced in the usual
places that I monitor.
Marc claimed he'd put out the announcement on pgsql-announce, but that
copy of the message never arrived here
On Fri, 29 Nov 2002, Vince Vielhaber wrote:
On Fri, 29 Nov 2002, Bruce Momjian wrote:
FYI, Vince, I started reading all my email (using elm) in a special 120
column wide, 38 row xterm. There was just too much detail in those
subjects i was missing.
Doesn't do me much good if too
On Mon, 2 Dec 2002, Marc G. Fournier wrote:
On Fri, 29 Nov 2002, Vince Vielhaber wrote:
On Fri, 29 Nov 2002, Bruce Momjian wrote:
FYI, Vince, I started reading all my email (using elm) in a special 120
column wide, 38 row xterm. There was just too much detail in those
subjects
Karel Zak wrote:
On Sat, Nov 30, 2002 at 01:11:20PM -0800, Joe Conway wrote:
As I said, this is all very preliminary; comments, suggestions, requests
are all welcome.
Only idea/dream: what implement dblink as virtual schema.
CREATE SCHEMA myschema AS DBLINK TO ... some connection options
Hello
In PostgreSQL 7.2/7.1:
template1=# select text(2.000::numeric);
text
--
2
(1 row)
In 7.3:
template1=# select text(2.000::numeric);
text
---
2.000
(1 row)
The text(numeric) function doesn't round numbers. :(
This is bug or feature? :)
I wonder if the sequences created by SERIAL should not be going into a
pg_sequence schema and protected like the toast tables are.
One could still share sequences by explicitly creating them and using a
DEFAULT clause with nextval().
We could even stop printing that annoying NOTICE ;-)
I've been getting reports of one of my test scenarios for libpqxx
failing with postgres 7.3. At the moment I can't reproduce this (I'm
still on 7.2) and I can't find anything pertinent in CVS commit
messages, mailing lists etc. so I'd really appreciate any lucidity from
this list.
My problem
template1=# select text(2.000::numeric);
text
---
2.000
(1 row)
The text(numeric) function doesn't round numbers. :(
This is bug or feature? :)
I'd say feature in that it doesn't reduce the precision of the number.
--
Rod Taylor [EMAIL PROTECTED]
PGP Key:
On Mon, Dec 02, 2002 at 01:35:47PM -0500, Rod Taylor wrote:
template1=# select text(2.000::numeric);
text
---
2.000
(1 row)
The text(numeric) function doesn't round numbers. :(
This is bug or feature? :)
I'd say feature in that it doesn't reduce the precision of the
We could even stop printing that annoying NOTICE ;-)
Agreed with this part :)
--
Rod Taylor [EMAIL PROTECTED]
PGP Key: http://www.rbt.ca/rbtpub.asc
signature.asc
Description: This is a digitally signed message part
On Mon, 2002-12-02 at 10:20, Jeroen T. Vermeulen wrote:
The scenario boils down to: Create a cursor, fetch n rows, move minus 2
billion or so rows, fetch 1 row. That last fetch used to give me the
row I was hoping for (the original first row again), but with 7.3 it
appears to yield nothing.
This feature is missing from 7.3..
(new round function is good idea (e.g. fround(numeric))
double precision | pg_catalog | round | double precision
numeric | pg_catalog | round | numeric
numeric | pg_catalog | round | numeric, integer
Looks
On Mon, Dec 02, 2002 at 02:29:03PM -0500, Rod Taylor wrote:
Seems to work the fine for me:
Puzzling...
Would you mind, if you have time, downloading libpqxx from GBorg and
doing a ./configure; make; make check and telling me if tests 19 38
succeed? I expect them to have identical results,
Mysql is planning on making this work:
SELECT id, @a:=COUNT(*), SUM(sum_col)/@a FROM table_name GROUP BY id.
Do we have anything like it (After a discussion with Tom i figure no).
User variables is nice, especially in these kind of queries.
Nice would be to be able to use them as in C (almost
On Mon, Dec 02, 2002 at 08:23:24PM +0100, Szima Gábor wrote:
OK, but:
template1=# select round('2.001'::numeric);
round
---
2
(1 row)
template1=# select round('2.001'::numeric,2);
round
---
2.00
(1 row)
The good idea (in 7.2):
template1=# select
Hi,
Just compiled the 7.3 branch from source and made some tests using the
JDBC driver coming with it. I did some tests with the henplus
JDBC-shell and noticed some problems quering the database meta data:
o the foreign key name is 'wierd'
---
DatabaseMetaData meta =
Yeah, good point. I think properly dealing with the pg_depends issues will
catch anything of that nature, but what to do with them?
Probably should move dependent type, constraint, index entries to the same
new
namespace. We might want to move related sequences, but I'm not sure we'd
want
to
Good call; they are still default (1 pages) and we are seeing 26000
pages per day being updated:
NOTICE: Removed 102226 tuples in 26002 pages.
CPU 2.02s/1.87u sec elapsed 69.85 sec.
NOTICE: Pages 201819: Changed 26128, Empty 0; Tup 315227: Vac 102226,
Keep
0, UnUsed 393793.
Wow, this sounds great.
Where can I get a copy? Why would anyone use anything else? ;-)
--
Bruce Momjian| http://candle.pha.pa.us
[EMAIL PROTECTED] | (610) 359-1001
+ If your life is a hard drive, | 13 Roberts Road
+ Christ can be your
Mysql is planning on making this work:
SELECT id, @a:=COUNT(*), SUM(sum_col)/@a FROM table_name GROUP BY id.
Do we have anything like it (After a discussion with Tom i figure no).
User variables is nice, especially in these kind of queries.
Well of course they have to make that work -
On Mon, 2 Dec 2002, Henner Zeller wrote:
Hi,
Just compiled the 7.3 branch from source and made some tests using the
JDBC driver coming with it. I did some tests with the henplus
JDBC-shell and noticed some problems quering the database meta data:
o the foreign key name is 'wierd'
Right now ALTER .. ADD PRIMARY KEY insists that the columns must be NOT
NULL first. I'd like to change it so that it automatically creates the
NOT NULL constraint (same as CREATE TABLE). I didn't see anything in
the spec for or against doing this automatically.
I believe the best way to do
Why just restrict them to moving tables? What if someone wants to move a
function or an aggregate to another schema?
What if they want to copy it?
Copying might be tricky, but I'd be happy to help with moving everything
else around. Though I don't think sequences can move (until we can
Christopher Kings-Lynne [EMAIL PROTECTED] wrote:
Mysql is planning on making this work:
SELECT id, @a:=COUNT(*), SUM(sum_col)/@a FROM table_name GROUP BY
id.
Do we have anything like it (After a discussion with Tom i figure
no). User variables is nice, especially in these kind of queries.
I've increased the value to 4 since we have much busier days, and I'll
se what happens.
Hang on - so when vacuum runs it knows exactly how many pages it cleaned up
but if the fsm value is low, it doesn't help? Why don't we just
automaticlaly set the FSM value to the max that vacuum
NOTE: redirecting to hackers
On Mon, 2002-12-02 at 12:50, Joe Conway wrote:
Tom Lane wrote:
For example, something I've heard repeatedly is that people would like to
hide the source code of their SQL or PLxxx functions from users who are
nonetheless allowed to call those functions. A
Yeah, but there is a point about running count(*) one time too many.
Say if i would like to get a prettyprinting query like this:
SELECT id, @a:=COUNT(*), sum_col::text||'/'@a::text FROM table_name;
That would be DAMN expensive doing with a subselect:
SELECT id, sum_col||'/'||(select
Just check out the AlterTableSetNotNull function in tablecmds.c to see a
quick and easy example of checking for NULL values.
Chris
- Original Message -
From: Rod Taylor [EMAIL PROTECTED]
To: PostgreSQL-development [EMAIL PROTECTED]
Sent: Monday, December 02, 2002 12:31 PM
Subject:
Just check out the AlterTableSetNotNull function in tablecmds.c to see a
quick and easy example of checking for NULL values.
Chris
- Original Message -
From: Rod Taylor [EMAIL PROTECTED]
To: PostgreSQL-development [EMAIL PROTECTED]
Sent: Monday, December 02, 2002 12:31 PM
Subject:
It wouldn't be terribly difficult to encrypt prosrc with 3des (or maybe
aes)
using the owner's passwd from pg_shadow. We would need a new bool column
in
pg_proc (proisencrypted?) and some logic in fmgr.c.
Is there sufficient interest to justify the effort?
I think this would be a
Christopher Kings-Lynne [EMAIL PROTECTED] wrote:
Are you sure that postgres evaluates that subselect more than once?
It looks to me like it returns a constant result for every row and
hence it will be evaluated once per statement, not once per row. I'm
no expert tho. Can someone answer this?
Stephan Szabo [EMAIL PROTECTED] wrote:
If you use a scalar subquery, yes, but I think a subselect in from
would help, maybe something like (if you want the total count)
select table_name.id, sum(sum_col)||'/'||t2.count from table_name,
(select count(*) as count from table_name) as t2 group
We need contrib/array in 7.4 and unique name in pg_constraint
Whelp. I'd like to make the constraint name unique too, but how do you
coax everyone into renaming their existing constraints -- especially
when there isn't an ALTER CONSTRAINT ... RENAME type statement?
--
Rod Taylor [EMAIL
On Mon, 2 Dec 2002, Magnus Naeslund(f) wrote:
Stephan Szabo [EMAIL PROTECTED] wrote:
If you use a scalar subquery, yes, but I think a subselect in from
would help, maybe something like (if you want the total count)
select table_name.id, sum(sum_col)||'/'||t2.count from table_name,
Joe Conway wrote:
David Wheeler wrote:
My understanding is that the nul character is legal in a byte sequence,
but if it's not properly escaped, it'll be parsed as the end of the
statement. Unfortunately, I think that it's a very tough problem to solve.
No question wrt '\0' bytes --
I've hacked the spec file from a 7.2.x source RPM to produce a 7.3 source RPM.
I've also created a set of i686 binary RPMs. These are *not* official PGDG
RPMs, and I'm not an RPM expert by any means (so use at your own risk!), but
I've posted them in case anyone is interested. I'll leave them
I want to see:
i) proper resource management a-la Oracle. This would allow a DBA to
limited the amount of time any given user spends in the parser, planner or
executor. It would be limited with a more sophisticated user system,
including things like CREATE USER PROFILE ...
ii) Auditing. Who
i) proper resource management a-la Oracle. This would allow a DBA to
limited the amount of time any given user spends in the parser, planner or
executor. It would be limited with a more sophisticated user system,
including things like CREATE USER PROFILE ...
Hehehe - yeah this would be neat.
Darko Prenosil writes:
OK, here it is. I do not know if this will help to use postgres some more
here in Croatia, but ...
I'll put this in the 7.3 branch, so it will be released when 7.3.1 comes
out.
--
Peter Eisentraut [EMAIL PROTECTED]
---(end of
On Mon, 2 Dec 2002, Christopher Kings-Lynne wrote:
viii) General advocacy, particularly in pushing mainstream IT media
coverage, conferences and university usage -- both for teaching SQL and
for teach database engineering concepts for senior undergrads.
Definitely. How about a resource
I might get together with some of the lecturers I've worked with in
Sydney to give such a document some weight. I must say, the problem is not
a technical one though. I've given talks to 3rd and 4th year students
about PostgreSQL -- technical, conceptual, political talks... you name
it. Out
On Mon, 2 Dec 2002, Christopher Kings-Lynne wrote:
I've given a talk in the 2002 honours lecture series at UWA about Postgres
and some of the things it can do. All of those guys were interested.
Especially since the deptartment does a lot of work in genetic algoriithms.
Excellent. Can you
- Original Message -
From: Shridhar Daithankar [EMAIL PROTECTED]
To: Matthew T. O'Connor [EMAIL PROTECTED]
Sent: Monday, December 02, 2002 11:12 AM
Subject: Re: [HACKERS] Auto Vacuum Daemon (again...)
On 28 Nov 2002 at 3:02, Matthew T. O'Connor wrote:
I went thr. it today and I have
Gavin Sherry wrote:
On Mon, 2 Dec 2002, Christopher Kings-Lynne wrote:
I've given a talk in the 2002 honours lecture series at UWA about Postgres
and some of the things it can do. All of those guys were interested.
Especially since the deptartment does a lot of work in genetic algoriithms.
Magnus Naeslund(f) kirjutas T, 03.12.2002 kell 03:18:
It looks like it (7.2.x):
# time psql genline -c select id from /dev/null
real0m0.694s
user0m0.147s
sys 0m0.025s
# time psql genline -c select id,id||'/'||(select count(*) from )
as x from /dev/null
real
Stephan Szabo [EMAIL PROTECTED] wrote:
On Tue, 3 Dec 2002, Magnus Naeslund(f) wrote:
Now convert this query so that it only evaluates the date_part thing
ONCE:
That's not a good idea as long as t.stamp varies from row to row. ;)
Perhaps once per row, maybe... :)
I give up
Hannu Krosing [EMAIL PROTECTED] wrote:
Magnus Naeslund(f) kirjutas T, 03.12.2002 kell 03:18:
It looks like it (7.2.x):
# time psql genline -c select id from /dev/null
real0m0.694s
user0m0.147s
sys 0m0.025s
# time psql genline -c select id,id||'/'||(select count(*) from
Gavin Sherry wrote:
I want to see:
i) proper resource management a-la Oracle. This would allow a DBA to
limited the amount of time any given user spends in the parser, planner or
executor. It would be limited with a more sophisticated user system,
including things like CREATE USER PROFILE ...
I am working on getting a shrink-wrapped version of PostgreSQL for Windows
Currently it installs a customized version of Cygwin, PostgreSQL 7.2.3,
cygipc, psqlodbc, and pgadminII
I currently have the setup done.
The target audience is not the enterprise, it is aimed at people using
Access
Christopher Kings-Lynne [EMAIL PROTECTED] writes:
Hang on - so when vacuum runs it knows exactly how many pages it cleaned up
but if the fsm value is low, it doesn't help? Why don't we just
automaticlaly set the FSM value to the max that vacuum has to clean
up?
Because we can't resize shared
Magnus Naeslund(f) wrote:
select id, sum(sum_col)||'/'||count(*) from table_name
group by id;
or
select table_name.id, sum(sum_col)||'/'||t2.count from table_name,
(select id, count(*) as count from table_name group by id) as t2 where
table_name.id=t2.id group by
Magnus Naeslund(f) [EMAIL PROTECTED] writes:
Mysql is planning on making this work:
SELECT id, @a:=COUNT(*), SUM(sum_col)/@a FROM table_name GROUP BY id.
We're supposed to spend our time emulating nonstandard features that
don't even exist yet? I think I have better things to do ...
Bruce Momjian [EMAIL PROTECTED] wrote:
Magnus Naeslund(f) wrote:
select id, sum(sum_col)||'/'||count(*) from table_name
group by id;
or
select table_name.id, sum(sum_col)||'/'||t2.count from table_name,
(select id, count(*) as count from table_name group by id) as t2
where
Good!
Now convert this query so that it only evaluates the date_part thing
ONCE:
select t.id, date_part('days',now()-t.stamp) from table_name t where
date_part('days',now()-t.stamp) 20;
Yes, it's a bit of a bugger that one.
I hope you all are kidding me in not seeing the real issue i'm
Magnus Naeslund(f) wrote:
Good!
Now convert this query so that it only evaluates the date_part thing
ONCE:
select t.id, date_part('days',now()-t.stamp) from table_name t where
date_part('days',now()-t.stamp) 20;
I hope you all are kidding me in not seeing the real issue i'm trying to
On Tue, 3 Dec 2002, Magnus Naeslund(f) wrote:
Now convert this query so that it only evaluates the date_part thing
ONCE:
That's not a good idea as long as t.stamp varies from row to row. ;)
Perhaps once per row, maybe... :)
select t.id, date_part('days',now()-t.stamp) from table_name t
On Tue, 3 Dec 2002, Bruce Momjian wrote:
Magnus Naeslund(f) wrote:
Good!
Now convert this query so that it only evaluates the date_part thing
ONCE:
select t.id, date_part('days',now()-t.stamp) from table_name t where
date_part('days',now()-t.stamp) 20;
I hope you all are
71 matches
Mail list logo