Ian Lance Taylor wrote:
>
> You're right. The mechanism used to preserve multiple versions of
> heap tuples could be extended to index tuples as well.
>
> Based on the heap tuple implementation, this would require adding two
> transaction ID's and a few flags to each index tuple. That's not
> > Also, more work would be required for every update. Right now an
> > update requires a B-tree insert for each index. With this change,
> > every update would require an additional B-tree lookup and write for
> > each index. That would require on average a bit less than one
> > additional bl
Richard Lynch writes:
> If I'm reading "man pg_passwd" correctly, I can create a standard
> Un*x passwd file and use that with "password" in pg_hba.conf
Correct.
> However, the current installation seems to be using "crypt", with no
> passwd file, and with unencrypted passwords in the pg_shadow
On Wed, 21 Feb 2001, Tom Lane wrote:
> Antonis Antoniou <[EMAIL PROTECTED]> writes:
> > I would like to know what is causing this error. The machine has 1GB of
> > memory and I use PostgreSQL 7.1beta4.
> > The same function on PostgreSQL 7.0.2 works ok. Could someone explain
> > me what th
Hello All,
I have installed the latest version of PostgreSQL 7.0.3 and all seems to
be fine with it.
My problem is that I have used www-sql in the past with MySQL which
works fine, but since I have just migrated to the newest version of
PostgreSQL, I have tried a recompile --with-database-pgsql
BTW what is show below was done on
a different machine where postgres is
installed in a user directory..
Just to rule out confusion I had earlier
I deleted bin lib include in the user
home directory and reinstall it under
/usr/local/pgsql and the problem still
remains namely I have to supply extra
Peter Eisentraut <[EMAIL PROTECTED]> writes:
> Is it good practice to scribble around in data type internals? Why not
> text *
> hello()
> {
> return textin("hello world");
> }
Well, he could do that if he wanted, but that doesn't teach him anything
about how to prepare his own datatype,
Thank you. Look what I get..
--
$ pg_ctl start -o "-F -S 2048"
postmaster successfully started up.
$ usage: /home/newsreader/pgsql/bin/postmaster [options]
-B nbufsset number of shared buffers
-D datadir set data directory
-S silent mode (disa
How can I use sequences in a stored procedure written with pl/pgsql?
Thanks in advance!
Renaud THONNART
Bruce Momjian writes:
> > Bruce Momjian <[EMAIL PROTECTED]> writes:
> >
> > > No, we have no ability to randomly pick rows to use for
> > > estimating statistics. Should we have this ability?
> >
> > That would be really slick, especially given the fact that VACUUM
> > runs much faster t
On Wednesday 21 February 2001 01:19, Joseph wrote:
> I republish my question for I had no answer, and this is a serious
> problem to me... I've used explain, vacuum, indexes... and so on, few
> nested requests...
Do not use ReiserFS on production machines. There are still enormous amount
of bugs
Hello All!
I used MySQL for a long time. There is SET() conception.
I can define SET('one','two','three') and use that type:
CREATE TABLE "test" (
"a" SET('one','two','three')
};
And then:
SELECT * FROM test WHERE a = 'one';
or
SELECT * FROM test WHERE a like 'one,three';
How can
On Wed, 21 Feb 2001, Martijn van Oosterhout wrote:
> On Tue, Feb 20, 2001 at 05:02:22PM -0800, Stephan Szabo wrote:
> >
> > IIRC, There's something which is effectively :
> > estimated rows = *
> > I think fraction defaults to (is always?) 1/10 for the standard
> > index type. That's where the
13 matches
Mail list logo
