At 01:40 PM 12/16/2005 -0500, Tom Lane wrote:
Nobody's said anything about giving up locale-sensitive sorting. The
question is about locale-sensitive equality: does it really make sense
that 'tty' = 'tyty'? Would your answer change in the context
'/dev/tty' = '/dev/tyty'? Are you willing to *
wrote:
> Hello!
>
> Please, help. I need to get complete list of 'client_encoding'
> values supported by the PostgreSQL Server. Can I somehow get these
> values using select query or function call?
Uh, we have a pg_conversion table, but I don't know how to tell which
are av
On Fri, Dec 16, 2005 at 07:14:42PM -0500, Bruce Momjian wrote:
> Actually we allow negative offsets and limits. Seems we should disallow
> them. Yes, it would be interesting if they did offsets relative to the
> end of the result set.
Either interesting or a great way to introduce bugs... AFAIK
CSN wrote:
> I was playing around with negative offsets:
>
> select * from table1 order by col1 offset -5 limit 25;
> select * from table1 order by col1 offset -25 limit
> 25;
> select * from table1 order by col1 offset -250 limit
> 25;
>
> They all return the same resultset (offset 0). Is
> ther
The way I handle this is to version the entire schema and have scripts
that know how to upgrade from one version to another. If you think about
it, you really want/need everything in the database to be designed to
run together anyway. I've yet to find a case where I'd want some of the
stuff in the
On Fri, Dec 16, 2005 at 11:25:22PM +0100, Martijn van Oosterhout wrote:
> On Fri, Dec 16, 2005 at 03:18:04PM -0600, Jim C. Nasby wrote:
> > BTW, if you wanted a more integrated solution, you could build a custom
> > type that would store the timestamp info. There's a good chance such
> > code would
hi all,
We installed a first version (1.0.0.1) of our schema. then came a few
patches we had for a few stored procedures and tables (1.0.0.2). Then
even more (1.0.0.3) (1.0.0.4). Some chose to upgrade to version 1.0.0.3
and stick to it, while some others chose to upgrade to 1.0.0.4.
Now when i ha
On Fri, Dec 16, 2005 at 03:18:04PM -0600, Jim C. Nasby wrote:
> BTW, if you wanted a more integrated solution, you could build a custom
> type that would store the timestamp info. There's a good chance such
> code would end up being what's used in the server as the official type,
> too.
Well, I wr
[EMAIL PROTECTED] (Madison Kelly) writes:
>In this case I can't predict what a given install's postgresql
> will be used for (outside of my program) because it is meant for
> general distribution (it's a backup program). This obviously makes
> things a lot more complicated. :p
No, it oughtn't.
Tom Lane <[EMAIL PROTECTED]> writes:
> The other approach we could take is to define text comparison as
> yielding equality only for bitwise-equal strings. If strcoll() returns
> zero then ask strcmp() for a second opinion.
Fwiw this has come up before on this list and it was discovered this
Please forgive the empty reply preceeding this note.
Tom Lane <[EMAIL PROTECTED]> writes:
> Jerry Sievers <[EMAIL PROTECTED]> writes:
> > Hi. I am curious how/why the Pg 8.1.1 build on Solaris
> > SunOS $hostname 5.9 Generic sun4u sparc SUNW,Ultra-250
> > chose LATIN9 charset as default.
>
"Andrus" <[EMAIL PROTECTED]> writes:
>> Has anyone used OpenOffice Base? Just a thought. Or Rekall - it's
>> a bit immature, but it might do what you want. The dreaded MS
>> Access can do what you describe in about 4 minutes...
>
> Postgres lacks easy GUI frontend with report generation capabili
On Fri, Dec 16, 2005 at 08:49:32PM +, Carlos Benkendorf wrote:
> I would like to say "take longer than" with "bad" statements...
> but I´m only logging statements that take longer than 700 ms
> and DECLARE statements are not logged of course.
>
> I think the only way in my case is logging
BTW, if you wanted a more integrated solution, you could build a custom
type that would store the timestamp info. There's a good chance such
code would end up being what's used in the server as the official type,
too.
On Fri, Dec 16, 2005 at 08:57:04AM -0800, Andrew B. Young wrote:
> I just realiz
Well that kills two birds. I completely didn't understand
the difference between SETOF and just RECORD return types.
That fixed it up, and for the record here is what the function looks
like now.
CREATE TYPE myrec AS (
id int
);
CREATE OR REPLACE FUNCTION test(x int) RETURNS SETOF myrec
On Fri, Dec 16, 2005 at 03:23:07PM -0400, Juan Garcés Bustamante wrote:
> Hola
Guten Abend !
> Estoy trabajando con Postgres 8.0.3 en Ubuntu.
Hm, ich benutze Debian, mit PostgreSQL 7.4 in einem Cluster.
Wie das bei Ubuntu so funktioniert, weiß ich nicht so
richtig.
> Necesito realizar transaccio
Tom Lane wrote:
Madison Kelly <[EMAIL PROTECTED]> writes:
Oh shoot, I really wasn't very verbose, was I? Sorry about that.
[ default pg_hba.conf with only "ident" lines ]
Ah, that explains your question about whether passwords were good for
anything at all. With this pg_hba.conf they aren't
Juan Garcés Bustamante wrote:
Hola
Estoy trabajando con Postgres 8.0.3 en Ubuntu.
Necesito realizar transacciones anidadas, pero no logro que se aborten
transacciones intermedias al abortarse una superior.
The main language spoken here is english. If you ask your questions in english,
you'll h
Sorry, I would like to say "take longer than" with "bad" statements... but I´m only logging statements that take longer than 700 ms and DECLARE statements are not logged of course. I think the only way in my case is logging all statements... Am I right? Or is there another way? Thanks
On Fri, Dec 16, 2005 at 03:23:07PM -0400, Juan Garcs Bustamante wrote:
> Estoy trabajando con Postgres 8.0.3 en Ubuntu.
Jaime Casanova has already mentioned this list is in English and
that questions in Spanish should go to pgsql-es-ayuda.
> Necesito realizar transacciones anidadas, pero no logro
On Fri, Dec 16, 2005 at 03:30:01PM -0500, Tom Lane wrote:
> "Niblett, David A" <[EMAIL PROTECTED]> writes:
> > Is there no way in Postgres that I can simply not return
> > anything so I show zero rows?
>
> Make the function return SETOF myrec not just myrec. Then you
> can return zero or one (or
"Niblett, David A" <[EMAIL PROTECTED]> writes:
> Is there no way in Postgres that I can simply not return
> anything so I show zero rows?
Make the function return SETOF myrec not just myrec. Then you
can return zero or one (or more) myrec's.
regards, tom lane
---
Jerry Sievers <[EMAIL PROTECTED]> writes:
> Hi. I am curious how/why the Pg 8.1.1 build on Solaris
> SunOS $hostname 5.9 Generic sun4u sparc SUNW,Ultra-250
> chose LATIN9 charset as default.
It probably deduced that from the locale setting (LANG or LC_ALL)
that initdb was run in. Try "sho
Madison Kelly <[EMAIL PROTECTED]> writes:
> Oh shoot, I really wasn't very verbose, was I? Sorry about that.
> [ default pg_hba.conf with only "ident" lines ]
Ah, that explains your question about whether passwords were good for
anything at all. With this pg_hba.conf they aren't --- the server wi
Tom Lane wrote:
Madison Kelly <[EMAIL PROTECTED]> writes:
May I ask then? What *is* considered "best practices" for securing a
database in PostgreSQL? Assuming I leave the 'pg_hba.conf' file at it's
default values, is there any real point to having a password on a
postgresql user account?
I've got an issue with a plpgsql function and I have not
been able to find any reference to this situation in searching.
I hope that someone can point me in the correct direction.
I'm running v8.0.3 and assuming a test set up of:
-
CREATE TYPE myrec AS (
id int
);
""dfx"" <[EMAIL PROTECTED]> wrote
> Dear Sirs,
> it exist some tool to convert a SQL script generated from MS SQL 7 to the
> syntax
> compatible with Postgresql v. 8.x.x?
In general, I bet no - since some functionalities in SQL Server is different
from ours, say TableSpace.
Regards,
Qingqing
Hi. I am curious how/why the Pg 8.1.1 build on Solaris
SunOS $hostname 5.9 Generic sun4u sparc SUNW,Ultra-250
chose LATIN9 charset as default. Of course we can override this and
might if there's a good reason move to LATIN9 as default.
Afraid though of having anything go flakey when
Martijn van Oosterhout wrote:
On Fri, Dec 16, 2005 at 02:09:52PM -0500, Madison Kelly wrote:
May I ask then? What *is* considered "best practices" for securing a
database in PostgreSQL? Assuming I leave the 'pg_hba.conf' file at it's
default values, is there any real point to having a password
Madison Kelly <[EMAIL PROTECTED]> writes:
> May I ask then? What *is* considered "best practices" for securing a
> database in PostgreSQL? Assuming I leave the 'pg_hba.conf' file at it's
> default values, is there any real point to having a password on a
> postgresql user account?
Well, if ther
> Has anyone used OpenOffice Base? Just a thought. Or Rekall - it's a bit
> immature, but it might do what you want. The dreaded MS Access
> can do what you describe in about 4 minutes...
Postgres lacks easy GUI frontend with report generation capabilities like
Access.
To fix this pgAdmin shou
On Fri, Dec 16, 2005 at 02:09:52PM -0500, Madison Kelly wrote:
> May I ask then? What *is* considered "best practices" for securing a
> database in PostgreSQL? Assuming I leave the 'pg_hba.conf' file at it's
> default values, is there any real point to having a password on a
> postgresql user ac
On Fri, 2005-12-16 at 13:11, Florian G. Pflug wrote:
> Andrus wrote:
> > Florian,
> >
> > your reply is not is not in news.postgresql.org server.
> > How to receive all messages using news.postgresql.org server ? It is tedious
> > to check the newsgroup archives using web interface also.
>
> Bad,
Florian,
>> your reply is not is not in news.postgresql.org server.
>
> Bad, but I guess there is nothing I can do about that... I'm subscriber
> of the mailinglist, not of the newsgroup... Sorry..
Your messages are digitally signed.
Maybe this prevents them from distributing using news protocol
Hola
Estoy trabajando con Postgres 8.0.3 en Ubuntu.
Necesito realizar transacciones anidadas, pero no logro que se aborten
transacciones intermedias al abortarse una superior.
Ejemplo:
BEGIN WORK;
INSERT INTO mitabla VALUES (1);
BEGIN TRANSACTION;
INSERT INTO mitabla VALUES (2);
I
Andrus wrote:
Florian,
your reply is not is not in news.postgresql.org server.
How to receive all messages using news.postgresql.org server ? It is tedious
to check the newsgroup archives using web interface also.
Bad, but I guess there is nothing I can do about that... I'm subscriber
of the m
Tom Lane wrote:
Madison Kelly <[EMAIL PROTECTED]> writes:
I want to find a way to let the user set the password on the new
database and have postgres actually ask for it without editing the
default 'pg_hba.conf' file, if at all possible.
There is no such animal as a "database password" in
Madison Kelly <[EMAIL PROTECTED]> writes:
> I want to find a way to let the user set the password on the new
> database and have postgres actually ask for it without editing the
> default 'pg_hba.conf' file, if at all possible.
There is no such animal as a "database password" in PG. There are
On Fri, Dec 16, 2005 at 06:28:03PM +, Carlos Benkendorf wrote:
> I´m trying to discover some bad SQL statements in our environment
> but the logs are not helping too much... they say we´ve some very
> bad fetch statements but I don´t know how to discover the
> corresponding SELECT statement...
Martijn van Oosterhout writes:
> On Fri, Dec 16, 2005 at 01:06:58PM -0500, Tom Lane wrote:
>> Ah. So we could redefine hashtext() to return the hash of the strxfrm
>> value. Slow, but a lot better than giving up hash join and hash
>> aggregation altogether...
> Not to put too fine a point on it
Hi all,
I am working on an installer for my program that creates a postgres
database and user (the installer is written in perl and runs as 'root').
I want to find a way to let the user set the password on the new
database and have postgres actually ask for it without editing the
default '
Martijn van Oosterhout writes:
> FWIW, here's some links to Microsoft and MySQL dealing with the same
> issue, so we're not alone here. Hungarian seems to be a complex
> language to sort, but it seems that glibc is right in this case.
The mysql bug link has a fairly detailed description, but it d
Hi, I´m trying to discover some bad SQL statements in our environment but the logs are not helping too much... they say we´ve some very bad fetch statements but I don´t know how to discover the corresponding SELECT statement... How can I find them? Some example: <10.0.1.122(3821) 5072 200
On Fri, 16 Dec 2005 11:59:48 -0600, Csaba Nagy <[EMAIL PROTECTED]> wrote:
On Fri, 2005-12-16 at 18:52, Tom Lane wrote:
Csaba Nagy <[EMAIL PROTECTED]> writes:
> ... So "tyty" and "tty" could be arguably both taken as double "ty",
> except that the official form is "tty"... but from a pronunciati
OK, thanks meneer!
Ale
-Original Message-
From: Martijn van Oosterhout [mailto:[EMAIL PROTECTED]
Sent: Thursday, December 15, 2005 11:35 AM
To: Ale Raza
Cc: pgsql-general@postgresql.org
Subject: Re: [GENERAL] Debug user lib for user define data type
On Wed, Dec 14, 2005 at 11:52:19AM -0
On Fri, Dec 16, 2005 at 01:06:58PM -0500, Tom Lane wrote:
> Ah. So we could redefine hashtext() to return the hash of the strxfrm
> value. Slow, but a lot better than giving up hash join and hash
> aggregation altogether...
Not to put too fine a point on it, but either you want locale-sensetive
On Fri, Dec 16, 2005 at 06:54:15PM +0100, Martijn van Oosterhout wrote:
> That's true, in the sense that unconverted strings are not hashable.
> This is what strxfrm was created for, to return the sorting key for a
> string. A quick C program demonstrates that indeed in that locale these
> two stri
Martijn van Oosterhout writes:
> On Fri, Dec 16, 2005 at 12:12:08PM -0500, Tom Lane wrote:
>> Perhaps the fast-path check is a bad idea, but fixing this is not just
>> a matter of removing that. If we subscribe to strcoll's worldview then
>> we have to conclude that *text strings are not hashable
On Fri, 2005-12-16 at 18:52, Tom Lane wrote:
> Csaba Nagy <[EMAIL PROTECTED]> writes:
> > ... So "tyty" and "tty" could be arguably both taken as double "ty",
> > except that the official form is "tty"... but from a pronunciation point
> > of view they ARE equivalent in hungarian.
>
> That's fair
On Fri, Dec 16, 2005 at 12:12:08PM -0500, Tom Lane wrote:
> Perhaps the fast-path check is a bad idea, but fixing this is not just
> a matter of removing that. If we subscribe to strcoll's worldview then
> we have to conclude that *text strings are not hashable*, because
> strings that should be "
Csaba Nagy <[EMAIL PROTECTED]> writes:
> ... So "tyty" and "tty" could be arguably both taken as double "ty",
> except that the official form is "tty"... but from a pronunciation point
> of view they ARE equivalent in hungarian.
That's fair enough, but the question is should they be taken as
equiv
Michael Fuhr <[EMAIL PROTECTED]> writes:
> I don't know if this is related or not, but did the following issue
> from a year ago ever get resolved? We were wondering then if the
> Hungarian locale on some platforms might be causing problems.
AFAIR we never did figure that one out. I wasn't able
On Fri, 2005-12-16 at 18:12, Tom Lane wrote:
[snip]
> I don't know anything about hu_HU comparison rules, but it appears that
> strcoll() believes that these two strings should be considered equal.
> Is that sane?
It is sane in a way, as the "ty" combination is pronounced together as a
single con
On Fri, Dec 16, 2005 at 12:12:08PM -0500, Tom Lane wrote:
> Thanks. After digging through it, I can exhibit the problem: in hu_HU
> locale and LATIN2 encoding, on a Linux box (Fedora Core 4, but probably
> all glibc installations will do this), I get
I don't know if this is related or not, but di
Mage <[EMAIL PROTECTED]> writes:
> Tom, I can send you the data I dumped and you can try it out with same
> settings. I trust you.
Thanks. After digging through it, I can exhibit the problem: in hu_HU
locale and LATIN2 encoding, on a Linux box (Fedora Core 4, but probably
all glibc installations
I just realized that the type "timestamp with time zone" does NOT store
the time zone!
(And I just found this support list to look for help.)
I am developing an application that stores power (watts) readings from
meters
located around the world (abridged)--
power_meter_id integer NOT NULL,
Has anyone used OpenOffice Base? Just a thought. Or Rekall - it's a
bit immature, but it might do what you want. The dreaded MS Access
can do what you describe in about 4 minutes...
EE
> I'm in progress of selecting development platform.
pgAdmin3 is written in C++, using the cross-plattform
Florian,
your reply is not is not in news.postgresql.org server.
How to receive all messages using news.postgresql.org server ? It is tedious
to check the newsgroup archives using web interface also.
> > I'm in progress of selecting development platform.
>pgAdmin3 is written in C++, using the cr
Dear Sirs,
it exist some tool to convert a SQL script
generated from MS SQL 7 to the syntax compatible with Postgresql v.
8.x.x?
Best regards.
Domenico
59 matches
Mail list logo
