On Sat, Sep 03, 2005 at 09:29:20AM +0530, Mavinakuli, Prasanna (STSD) wrote:
> here is the sample code which works fine and prints column
> lengths in HP-UX PA machines and breaks in HP-UX IA machines.
The code you posted works for me with PostgreSQL 8.0.3 on FreeBSD
4.11-STABLE/i386 and Solaris
I assumed as much. Now's the time for me to optimize
so I'd rather know and make optimizations accordingly,
than step blindly. Thanks for the reply. As always,
your a big help.
--- Richard Huxton wrote:
> Matthew Peter wrote:
> > same size VARCHAR or INT IX faster? i assume INT.
> The
> > reason
On Fri, Sep 02, 2005 at 07:16:04PM -0400, Hrishikesh Deshmukh wrote:
> What i meant was when the third component of tables are ready and i can
> create tables any time what happens to relationships when i am building the
> third component tables? How smooth the process will be?
What relationshi
HI All,
What i meant was when the third component of tables are ready and i can
create tables any time what happens to relationships when i am building
the third component tables? How smooth the process will be?
Regards,
Hrishi
On 9/2/05, Michael Fuhr <[EMAIL PROTECTED]> wrote:
On Fri, Sep 02,
On Fri, 2 Sep 2005, Tom Lane wrote:
> "Sergey E. Koposov" <[EMAIL PROTECTED]> writes:
> > I'm interested in simplifying those queries and introducing the operator
> > doing something like this:
>
> > my_operator(bigint x, bigint[] arr)
> > checking the condition:
> > ((x>arr[1]) AND (xarr[3]
On Fri, Sep 02, 2005 at 05:28:11PM -0400, Hrishikesh Deshmukh wrote:
> We are designing a database, the database has three major components (each
> component with 4-8 tables)
> two components are ready, is it possible to 'overlay' the third one later?
> or
> Do i have to drop all the components (
Hrishikesh Deshmukh <[EMAIL PROTECTED]> writes:
> What i am saying is some tables are existing, could i later add a bunch of
> tables at a later date without dropping the entire database?
Sure, you can do CREATE TABLE at any time.
-Doug
---(end of broadcast)-
Hi All,
A simple question to the list.
We are designing a database, the database has three major components (each component with 4-8 tables)
two components are ready, is it possible to 'overlay' the third one later? or
Do i have to drop all the components (along with data!!!) and then add the thir
On Fri, Sep 02, 2005 at 10:39:14PM +0200, Poul Møller Hansen wrote:
> >
> >See the has_table_privilege() function.
>
> Thanks, but how does it work ?
>
> select has_table_privelege('public.mytable', 'select');
> ERROR: function has_table_privelege("unknown", "unknown") does not exist
> HINT: No
On Fri, Sep 02, 2005 at 12:17:09PM -0600, Cristian Prieto wrote:
Hey Cristian,
> Right now it works just with one field and I guess it is working well, but
> I am very worried about the performance of SPI_execute() and SPI_execp(). I
> read in the Developer FAQ something about accessing the dat
See the has_table_privilege() function.
Thanks, but how does it work ?
select has_table_privelege('public.mytable', 'select');
ERROR: function has_table_privelege("unknown", "unknown") does not exist
HINT: No function matches the given name and argument types. You may
need to add explicit t
"Sergey E. Koposov" <[EMAIL PROTECTED]> writes:
> I'm interested in simplifying those queries and introducing the operator
> doing something like this:
> my_operator(bigint x, bigint[] arr)
> checking the condition:
> ((x>arr[1]) AND (xarr[3]) AND (x
Hello All,
I have a following question:
I'm using some set of queries like:
SELECT * FROM
(SELECT my_function(ra, dec, 0.001) AS ipix1, ra1, dec1
FROM table1) AS jtable1, table2
WHERE table2.ipix>=ipix1[1] AND table2.ipix<=ipix1[2] );
ipix is bigint column, on which the Btree
First at all, thanks a lot for your help with my trouble, it was very
helpfull...
This is my complete code:
#include "postgres.h"
#include "fmgr.h"
#include "executor/spi.h"
#include
PG_FUNCTION_INFO_V1(myspi);
Datum
myspi(PG_FUNCTION_ARGS)
{
int ret;
bool isnull;
bytea *val;
void *plan;
O
On Fri, 2005-09-02 at 09:51 -0700, Bob Pawley wrote:
> regarding the setting up of tables
> ...
> I could create a separate table for each category of elements ... The
> challenge is ... a distinct possibility of the user inserting an
> element that we haven’t considered. The above approach would
I have a question to ask regarding the setting up of tables.
I am developing Postgresql for an engineering application. This
application will have an initial user succeeded by multiple users in multiple
engineering disciplines downstream.
Level One
What I would like to do is to ha
On Fri, Sep 02, 2005 at 05:31:54PM +0200, Bohdan Linda wrote:
> Hello,
>
> I have encountered on (for me) wierd thing. When dropping an user, the
> database will not forget his permissions. After his recreation he has the
> original permissions.
Known problem, partially fixed in 8.1. (At least
Hello,
I have encountered on (for me) wierd thing. When dropping an user, the
database will not forget his permissions. After his recreation he has the
original permissions.
I use an approach of dropping all users when recreating the database
environment and user recreation to avoid any unwanted
I looked at the documentation before I ever posted anything. I didn't know how
EXECUTE works and I actually thought the UPDATE statement itself would set
the variable.
Thanks for the reply.
On Friday 02 September 2005 10:47 am, Tom Lane saith:
> Terry Lee Tucker <[EMAIL PROTECTED]> writes:
> >
marcelo Cortez wrote:
hi Richard ,folks
Question: is normal this locks ?
which is the reason of this?
As I said - I don't see any locks that could cause you problems. Can you
provide the error message showing locks timing out?
--
Richard Huxton
Archonet Ltd
"Mavinakuli, Prasanna (STSD)" <[EMAIL PROTECTED]> writes:
> When I try to insert the data of size 40764 bytes(data type bytea) the
> call PQexecparams won't return and when I kill that one it gives
> The log:"LOG: unexpected EOF within message length word"
> Or "Log:Incomplete message transfer fro
Terry Lee Tucker <[EMAIL PROTECTED]> writes:
> Will the FOUND variable be set, when using EXECUTE, as it would be with a
> normal UPDATE statement?
The documentation is pretty specific about which plpgsql statements set
FOUND, and I don't see EXECUTE in that list ...
http://developer.postgresql.o
hi Richard ,folks
Question: is normal this locks ?
which is the reason of this?
thanks for your time
best regards
MDC
--- Richard Huxton escribió:
> marcelo Cortez wrote:
> > Richard
> >
> > Sorry for a delay
> >
> > question the select * from pg_stat_activity
> > thows
Apparently, the FOUND variable is set to false regardless of the outcome of
the dynamic update statement. I placed a call to GET DIAGNOSTICS var =
ROW_COUNT. It returned 1. Question is, was that the result of a direct
update previously coded in the same trigger, or is that the result of the
d
On Fri, Sep 02, 2005 at 09:30:31AM +0530, Mavinakuli, Prasanna (STSD) wrote:
> When I try to insert the data of size 40764 bytes(data type bytea) the
> call PQexecparams won't return and when I kill that one it gives
> The log:"LOG: unexpected EOF within message length word"
> Or "Log:Incomplete m
I haven't tried it. It's a complicated trigger function with updates to other
tables that are NOT dynamic in nature. If the EXECUTE statement doesn't set
the FOUND variable, then I will be reading the result from a previous
operation. I thought maybe somebody would know this already.
On Friday
On Fri, Sep 02, 2005 at 08:51:41AM -0400, Terry Lee Tucker wrote:
>
> Will the FOUND variable be set, when using EXECUTE, as it would be with a
> normal UPDATE statement?
What happened when you tried it?
--
Michael Fuhr
---(end of broadcast)---
T
Hi,
When I try to insert the data of size 40764 bytes(data type bytea) the
call PQexecparams won't return and when I kill that one it gives
The log:"LOG: unexpected EOF within message length word"
Or "Log:Incomplete message transfer from client".
If I try to insert data of size 40760 bytes (data
Greetings,
I have a question regarding the use of the FOUND variable within a plpgsql
function. I have a trigger fuction which executes a dynamic update on a
different table with the EXECUTE statement. Here's the question:
Will the FOUND variable be set, when using EXECUTE, as it would be with
Hi list/Michael,
Sorry I forgot "reply to all"
It proved to be a problem with the permissions on the table and view! So, the
error that was reported was completely different from the actual error. I do
not know how this can happen, but by making a direct connection to the
database within Zope,
in PgAdmin, running wrong insert command causes error like
ERROR: insert or update on table "dok" violates foreign key constraint
"dok_klient_fkey"
DETAIL: Key (klient)=(gg ) is not present in table "klient".
Running same command through ODBC driver returns only first line.
How to re
Matthew Peter wrote:
same size VARCHAR or INT IX faster? i assume INT. The
reason I ask is I was wondering what (if any) is the
avg delay from one over the other? And benefit of one
over the other? Thanks.
If you want numbers, use INT. If you want text use a VARCHAR.
It's probably difficult t
On Thu, Sep 01, 2005 at 02:10:04PM -0600, Cristian Prieto wrote:
> Hello, I've been reading a little the ECPG (Embedded SQL in C) and the doc is
> (I guess) very clear, but I cannot find any examples in the documentation,
> any idea where to get examples? inside the pgsql source code?
Yes, there
33 matches
Mail list logo
