[GENERAL] PostGresql USENET Vote

[BARD ROCK]

It is now time to cast your vote as to whether you would like these
mailing lists to become legal USENET newsgroups.

To vote, visit this web page:

http://party.to/bardrock

Fill out the ballot and email it to the address provided.

Thanks.

BARD ROCK

---(end of broadcast)---
TIP 9: the planner will ignore your desire to choose an index scan if your
  joining column's datatypes do not match

[SPAM] Re: [GENERAL] PostGresql USENET Vote

[Michael Glaesemann]

On Nov 13, 2004, at 7:47 PM, BARD ROCK wrote:
It is now time to cast your vote as to whether you would like these
mailing lists to become legal USENET newsgroups.

Don't bother. Nothing to see here. Not Usenet site.
Michael Glaesemann
grzm myrealbox com
---(end of broadcast)---
TIP 5: Have you checked our extensive FAQ?
  http://www.postgresql.org/docs/faqs/FAQ.html

Re: [GENERAL] I spoke with Marc from the postgresql mailing list.

[Richard_D_Levine]

I'm getting more spam also, though not an inordinate amount.  A couple a
week.
The guy from UAE who wants to send me millions, and the v drug.
Didn't happen when I signed on months ago.



 
  Jeff Eckermann
 
  <[EMAIL PROTECTED]To:   Brian {Hamilton 
Kelly} <[EMAIL PROTECTED]>, [EMAIL PROTECTED]  
  om>cc:
 
  Sent by:   Subject:  Re: [GENERAL] I 
spoke with Marc from the postgresql mailing list. 
  [EMAIL PROTECTED] 
   
  tgresql.org   
 

 

 
  11/12/2004 06:30 PM   
 

 

 





--- Brian {Hamilton Kelly} <[EMAIL PROTECTED]> wrote:

> Which only confirms my opinion that he's a fuckwit.

Evidently some of the nastiness out there on Usenet is
propagating into our lists in advance of any change.



__
Do you Yahoo!?
Check out the new Yahoo! Front Page.
www.yahoo.com



---(end of broadcast)---
TIP 7: don't forget to increase your free space map settings




---(end of broadcast)---
TIP 7: don't forget to increase your free space map settings

[GENERAL] Spam notice

[Brian]

Just a heads up folks, someone has harvested emails out of this list, 
or one of the windows users has a virus. Be on the lookout

Full header below.
From [EMAIL PROTECTED] Sat Nov 13 13:16:34 2004
Return-Path: [EMAIL PROTECTED]
Received: from [62.213.193.131] (HELO alibi.mail) by genium.com 
(Stalker SMTP Server 1.8b8) with ESMTP id S.0002169980 for 
<[EMAIL PROTECTED]>; Fri, 12 Nov 2004 14:11:58 -0500
Received: from 50cent (dsl-201-129-146-226.prod-infinitum.com.mx 
[201.129.146.226])
	by alibi.mail (alibi.mail)
	(MDaemon.PRO.v7.2.1.R)
	with ESMTP id md5234957.msg
	for <[EMAIL PROTECTED]>; Fri, 12 Nov 2004 19:59:37 +0100
From: "[EMAIL PROTECTED]" <[EMAIL PROTECTED]>
Subject: Vicodln - no r-x needed - inexpensive
X-Mailer: Microsoft Outlook, Build 10.0.2616
Reply-To: [EMAIL PROTECTED]
Date: Fri, 12 Nov 2004 13:07:30 -0600
Mime-Version: 1.0
Content-Type: text/html; charset=us-ascii
X-Spam-Processed: alibi.mail, Fri, 12 Nov 2004 19:59:37 +0100
	(not processed: message from valid local sender)
X-MDRemoteIP: 201.129.146.226
X-Return-Path: [EMAIL PROTECTED]
X-MDaemon-Deliver-To: [EMAIL PROTECTED]
X-MDAV-Processed: alibi.mail, Fri, 12 Nov 2004 19:59:41 +0100

---(end of broadcast)---
TIP 9: the planner will ignore your desire to choose an index scan if your
 joining column's datatypes do not match

Re: [GENERAL] I spoke with Marc from the postgresql mailing list.

[Klaas]

In article <[EMAIL PROTECTED]>,
 [EMAIL PROTECTED] (Brian {Hamilton Kelly}) wrote:

> newsgroups from ONE server, that of their university/employer/ISP.  Since
> that capability appeared in Netscape, and other browser/newsreaders,
> there has been a proliferation of *really* private newsgroups, such as
> borland.*; even microsoft.* was originally only available from one
> "site".)

Blaming netscape for microsoft.* is rather funny.

-Mike

---(end of broadcast)---
TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]

[GENERAL] List of postgresql rogue groups (was Re: Important Info on comp.databases.postgresql.general)

[Woodchuck Bill]

[EMAIL PROTECTED] wrote in news:[EMAIL PROTECTED]:

>>The other issue is that I would like to add the other postgresql
>>groups for consideration to be included into the big 8.  However there
>>are quite a few of them, and I don't know if all of them deserve to be
>>there. They are all under comp.databases.postgresql.*, and I highly
>>doubt Marc would want to eliminate them.  I will have to include them
>>all and have a heavy postgresql presense in the comp.databases.*
>>hierarchy. 
> 
> I have no advice to offer on this.  How many are there exactly?

21, at least.

comp.databases.postgresql.admin
comp.databases.postgresql.advocacy
comp.databases.postgresql.announce
comp.databases.postgresql.bugs
comp.databases.postgresql.committers
comp.databases.postgresql.docs
comp.databases.postgresql.general
comp.databases.postgresql.hackers
comp.databases.postgresql.interfaces
comp.databases.postgresql.interfaces.jdbc
comp.databases.postgresql.interfaces.odbc
comp.databases.postgresql.interfaces.pgadmin.hackers
comp.databases.postgresql.interfaces.pgadmin.support
comp.databases.postgresql.interfaces.php
comp.databases.postgresql.novice
comp.databases.postgresql.patches
comp.databases.postgresql.performance
comp.databases.postgresql.ports
comp.databases.postgresql.ports.cygwin
comp.databases.postgresql.questions
comp.databases.postgresql.sql

-- 
Bill

---(end of broadcast)---
TIP 5: Have you checked our extensive FAQ?

   http://www.postgresql.org/docs/faqs/FAQ.html

Re: [GENERAL] The Big 9?

[Woodchuck Bill]

Mike Cox <[EMAIL PROTECTED]> wrote in
news:[EMAIL PROTECTED]: 

> Wouldn't a good solution to the "bogus" and rogue groups be a creation
> of a new domain in the big 8?  Suppose there was a rogue.* domain. 
> All the groups that were rogue would be placed there by the usenet
> providers. Therefore those customers who demanded certain rogue groups
> would have them, only they would be moved under rogue.  Say someone
> sets up a rogue group like comp.muffins.  All the usenet providers
> would then just move it to rogue.comp.muffins.
> 
> That way their status is clear to all subscribers, the commercial
> usnet providers would have a well managed big 9, and would not be
> forced to choose between having a well managed hierarchy or doing
> carrying the groups customers demand.

You frighten me. ;-)

-- 
Bill

---(end of broadcast)---
TIP 3: if posting/reading through Usenet, please send an appropriate
  subscribe-nomail command to [EMAIL PROTECTED] so that your
  message can get through to the mailing list cleanly

Re: [GENERAL] RFD: comp.databases.postgresql.general

[Jim Riley]

On Sun, 7 Nov 2004 17:29:28 -0400 (AST), [EMAIL PROTECTED] ("Marc
G. Fournier") wrote:

>On Sun, 7 Nov 2004, Gary L. Burnore wrote:
>
>> The groups aren't listed as moderated. Anyone who wants to post is
>> able to. Those not on the mailing list don't go through.  That's the
>> problem.
>
>As long as the posting gets to the gateway, it gets put into the 
>moderator (me) queue for approval ...

If the group was explicitly marked as moderated, then all posts would
be sent directly to you.  The content of both the mailing list and
newsgroup would be identical, so that whichever a participant chose to
use would be a matter of preferred interface.

-- 
Jim Riley

---(end of broadcast)---
TIP 4: Don't 'kill -9' the postmaster

Re: [GENERAL] RFD: comp.databases.postgresql.general

[Woodchuck Bill]

[EMAIL PROTECTED] ("Marc G. Fournier") wrote in 
news:[EMAIL PROTECTED]:

> Unless its spam, it goes through ... I don't (nor have I ever) refused a 
> post based on content other then spam ... even if its anti-PostgreSQL 
> *shrug*

The problem with the system is that the spam *all* gets posted to Usenet, 
but not the mailing lists. The mailing lists may be moderated, but the 
newsgroups are not. That needs to be changed.

Marc, please stop removing news.groups from your replies.

---(end of broadcast)---
TIP 5: Have you checked our extensive FAQ?

   http://www.postgresql.org/docs/faqs/FAQ.html

Re: [GENERAL] 24x7x365 high-volume ops ideas

[Karim Nassar]

On Sun, 2004-11-07 at 21:16, Christopher Browne wrote:
> None of these systems _directly_ address how apps would get pointed to
> the shifting servers.

> Something needs to be "smart enough" to point apps to the right place;
> that's something to think about...

Seems like it would be pretty easy to be smart in PHP:

function db_connect() {
  $conn = pg_connect("dbname='foo' user='dawg' password='HI!'
  host='master'");
  if (!($conn AND (pg_connection_status($conn) == 0))) {
// problem with master
$conn = pg_connect("dbname='foo' user='dawg' password='HI!'
  host='replica'");
if ($conn AND (pg_connection_status($conn) == 0)) {
  return $conn;
}
  } else {
return $conn;
  }
  return NULL;
}

Whatever client-side language one uses, the technique is the same
(though the coding style might differ :P ), can be used for persistent
connections (eg: with pg_pconnect in PHP), and seems like it could be
extended to any reasonable number of database servers.

What is the problem with this? The only issue I can see is that
"replica" might be behind. Depending on the application, this might not
be bad. If the app MUST have the very most accurate DB, you could remove
the logic that connects to the replica, but then that nullifies this
whole conversation...

\<.




---(end of broadcast)---
TIP 8: explain analyze is your friend

[GENERAL] Change encoding of SQL_ASCII database with multibyte chars?

[leon-postgres]

Hello, there! I recently migrated a medium-sized db from SQL Server to 
postgres. A couple months after the migration, I noticed I was having 
some problems retrieving some data via the JDBC driver. The problem 
turned out to be multi-byte characters (smart quotes, etc) in the 
SQL_ASCII database.

As I understand it, I need to change the encoding to UNICODE in order 
to support non-English characters, smart quotes, etc. The trick is that 
I would like to preserve the small quantity of multi-byte characters 
that are already in the database.

I've tried dumping the db, changing the encoding in the CREATE DATABASE 
command to UNICODE, and reloading the dump, but I get errors like 
"ERROR:  invalid byte sequence for encoding "UNICODE": 0xfd3130 ", 
which makes me afraid that I'm losing data.

Is there any way to re-encode my database that will turn the (possibly 
mangled) multi-byte characters back into their original form?

Thanks for reading
Leon 

---(end of broadcast)---
TIP 4: Don't 'kill -9' the postmaster

Re: [GENERAL] I spoke with Marc from the postgresql mailing list.

[ru . igarashi]

Mike Cox <[EMAIL PROTECTED]> wrote:

>I also offered to let him take over the process as he understands gatewaying
>and the details better than I.  I also offered to let him create the next
>RFD, and hopefully he will agree to do it if he has the time.  He is the
>most connected member of the postgresql team and it would be right for him
>to decide how the groups should be run.

He doesn't have to be the primary proponent.  Many other proposals
have had multiple proponents, with responsibilities distributed in
a variety of ways.  In this case, he could simply be the technical
issues proponent, or topic expert proponent, so to speak.

For example, he might be able to provide you with more charter
info, which would be really helpful given that the groups exist
and that there is probably a decent idea of how the groups seem
to have settled.  Then you can write it up, have him check it,
then you can put it in the next RFD.  

You can also ask him which are the most popular lists, so that
you can add those to the RFD.  That way you can get a good start
on the "official" hierarchy, AND reduce the problem people would
have had in posting to the correct newsgroup for a given postgresql
subtopic.

This way, you can stay on to push the proposal forward, AND the 
current moderator is involved to maintain the continuity with
the mailing lists and gateways, both in "official" capacities.

ru

-- 
My standard proposals rant:
Quality, usefulness, merit, or non-newsgroups popularity of a topic 
is more or less irrelevant in creating a new Big-8 newsgroup.  
Usenet popularity is the primary consideration.

---(end of broadcast)---
TIP 9: the planner will ignore your desire to choose an index scan if your
  joining column's datatypes do not match

Re: [GENERAL] Postresql RFD version 2.0 Help Wanted.

[Brian {Hamilton Kelly}]

On Saturday, in article <[EMAIL PROTECTED]>
 [EMAIL PROTECTED] "Mike Cox" wrote:

> Woodchuck Bill wrote:
> 
> > Mike Cox <[EMAIL PROTECTED]> wrote in
> > news:[EMAIL PROTECTED]:
> > 
> >>  I cannot handle the volume of email that a mailing list would place
> >>  on my
> >> inbox.
> > 
> > Ever heard of a digest version?
> > 
> 
> I don't care. Its too much of a hassle to dig through without being able to
> google groups search it.

Mike makes here a VERY valid point about the mailing list vs newsgroups
controversy: often there is no means to search past articles from the
mailing list unless one maintains one's own complete archive thereof.  In
contrast, posts of articles to *public* newsgroups (which includes the
Big-8, alt.*, and thousands of national and other hierarchies, such as
demon.*) are generally[1] archived by GoogleGroups, and thereby readily
searchable.

Private so-called newsgroups, which are carried on only one (or a handful
of cooperating) server(s), and mailing lists ditto, need to make their
own arrangements for archiving (which they often do) and providing a
search capability (which they often don't).

Newsgroups are much more flexible from the end-users' POV; however, there
is a huge percentage of Internet users that are completely unaware of
the existence of the concept.  In contrast, they perceive using a mailing
list is no different to mailing one person (as, sadly, can often be
observed when mailing list participants use nasty habits picked up from
using Outlook Express in inter-office memoranda).

[1] Except, of course, where the poster uses X-No-Archive: Yes
-- 
Brian {Hamilton Kelly} [EMAIL PROTECTED]
   "I don't use Linux. I prefer to use an OS supported by a large multi-
   national vendor, with a good office suite, excellent network/internet
   software and decent hardware support."

---(end of broadcast)---
TIP 8: explain analyze is your friend

Re: [GENERAL] RFD: comp.databases.postgresql.general

[Woodchuck Bill]

Andrew - Supernews <[EMAIL PROTECTED]> wrote in 
news:[EMAIL PROTECTED]:

>> Marc, please stop removing news.groups from your replies.
> 
> He's posting to the mailing list; he probably can't avoid dropping the
> crosspost.
> 

He can make a nominal effort and post *something* to news.groups.

-- 
Bill

---(end of broadcast)---
TIP 6: Have you searched our list archives?

   http://archives.postgresql.org

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[Woodchuck Bill]

Russ Allbery <[EMAIL PROTECTED]> wrote in 
news:[EMAIL PROTECTED]:

>> The UDP would be aimed at the news server(s) at which the mailing list
>> is being improperly gated.  It is their responsibility to reject
>> improper traffic.  As these same servers would also likely carry the
>> group in question, I have serious doubts that they would remove them
>> without the threat of a UDP hanging over their head.  Whether just the
>> group is blacklisted or the entire server would be the subject of
>> another thread entirely.
> 
> If someone actually seriously tries to do this, I will personally offer
> that news server a feed to break the UDP.

If you're willing to do that, then you should just issue the control 
messages for all 21 groups right now. Why would you want to block others 
from trying to hold a net abuser accountable? For the UDP to be successful, 
it would take more than two proponents. You would really override the 
outcome?

[I'm really not trying to flame Marc by calling him a net abuser, but isn't 
that the category he would fall under, in all seriousness? You are giving 
him the chance to fix his past mistakes. He won't enter the news.groups 
discussion, and he stated that he will not break up *any* of the rogue 
groups if the CFV fails.]

> Now, please try to tone down the level of confrontation and act like
> adults, okay?

It was Marc who set the tone, by claiming that the rogue groups will 
continue to operate as they currently do, regardless of the CFV outcome.  

-- 
Bill

---(end of broadcast)---
TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[Woodchuck Bill]

[EMAIL PROTECTED] (John Stanley) wrote in 
news:[EMAIL PROTECTED]:

> Obviously there is nothing wrong with it. As I seem to recall, one of
> the admins who (routinely?) created bogus groups is now part of the NAN
> moderating team. 

Who would that be?

-- 
Bill

---(end of broadcast)---
TIP 5: Have you checked our extensive FAQ?

   http://www.postgresql.org/docs/faqs/FAQ.html

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[Woodchuck Bill]

[EMAIL PROTECTED] wrote in news:[EMAIL PROTECTED]:

> That's his perogative.  His server, his rules (or whoever's he set the
> groups up on).  We don't have the right to dictate what groups he puts
> on his news server.  If someone else decides to take a feed from him 
> and allow the group on their server, same story, their server, their 
> rules.  That kind of independence is at the foundation of usenet.  
> While I may be displeased that the bogus groups exist, I'm similarly 
> not going to be supportive of moves to dictate what groups he puts
> on his server.

Those groups are propagated to *other* servers, and they confuse lots of 
people into thinking that they are bonafide Big-8 groups. Even Google is 
either confused or careless about the status of those groups. If the NAN 
team announces a reversal of the rec.woodworking.all-ages result in the 
next few days, would you have any problem with the proponents sending out a 
control message anyway? Archiving the rogue group in Google Groups? If 
nothing else, taking no steps toward action sets a bad example, and might 
encourage others to skip the RFD and create more rogue groups.

-- 
Bill

---(end of broadcast)---
TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[ru . igarashi]

Woodchuck Bill <[EMAIL PROTECTED]> wrote:
>Russ Allbery <[EMAIL PROTECTED]> wrote in 
>news:[EMAIL PROTECTED]:

>>> The UDP would be aimed at the news server(s) at which the mailing list
>>> is being improperly gated.  It is their responsibility to reject
>>> improper traffic.  As these same servers would also likely carry the
>>> group in question, I have serious doubts that they would remove them
>>> without the threat of a UDP hanging over their head.  Whether just the
>>> group is blacklisted or the entire server would be the subject of
>>> another thread entirely.
>> 
>> If someone actually seriously tries to do this, I will personally offer
>> that news server a feed to break the UDP.

>If you're willing to do that, then you should just issue the control 
>messages for all 21 groups right now. Why would you want to block others 
>from trying to hold a net abuser accountable? For the UDP to be successful, 
>it would take more than two proponents. You would really override the 
>outcome?

Who's being abused here?  Russ & Co.?  By their own admission, no.
The Big-8?  No, the groups don't exist in the Big-8?  The existing
readers?  No, they can read the group.  The rest of the world?  No
more so than those that don't have groups specifically for their
pet interests, which as far as we are concerned is not sufficient
harm to act upon.  The Big-8 process is partly predicated on only
providing groups for those with sufficient numbers, and since
the latter has not been established, there's no such harm.  Even
then, if the CFV results in a pass, everyone that cares benefits,
and if the CFV results in a failure, the rest of the world doesn't
matter (until they can build up sufficient numbers to pass the
next time).

...
>> Now, please try to tone down the level of confrontation and act like
>> adults, okay?

>It was Marc who set the tone, by claiming that the rogue groups will 
>continue to operate as they currently do, regardless of the CFV outcome.  

That's his perogative.  His server, his rules (or whoever's he set the
groups up on).  We don't have the right to dictate what groups he puts
on his news server.  If someone else decides to take a feed from him 
and allow the group on their server, same story, their server, their 
rules.  That kind of independence is at the foundation of usenet.  
While I may be displeased that the bogus groups exist, I'm similarly 
not going to be supportive of moves to dictate what groups he puts
on his server.

ru

-- 
My standard proposals rant:
Quality, usefulness, merit, or non-newsgroups popularity of a topic 
is more or less irrelevant in creating a new Big-8 newsgroup.  
Usenet popularity is the primary consideration.

---(end of broadcast)---
TIP 6: Have you searched our list archives?

   http://archives.postgresql.org

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[Woodchuck Bill]

[EMAIL PROTECTED] wrote in news:[EMAIL PROTECTED]:

> Who's being abused here?  Russ & Co.?  By their own admission, no.
> The Big-8?  No, the groups don't exist in the Big-8?  The existing
> readers?  No, they can read the group.  The rest of the world?  No
> more so than those that don't have groups specifically for their
> pet interests, which as far as we are concerned is not sufficient
> harm to act upon.  The Big-8 process is partly predicated on only
> providing groups for those with sufficient numbers, and since
> the latter has not been established, there's no such harm.  Even
> then, if the CFV results in a pass, everyone that cares benefits,
> and if the CFV results in a failure, the rest of the world doesn't
> matter (until they can build up sufficient numbers to pass the
> next time).

OK, so you think it is acceptable for anyone to create as many Big-8 rogue 
groups as they like? Some servers will carry the groups, others will not. 
There should be no accountability for someone doing this. There is nothing 
wrong with it. 

-- 
Bill

---(end of broadcast)---
TIP 6: Have you searched our list archives?

   http://archives.postgresql.org

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[Devin L. Ganger]

On 9 Nov 2004 20:36:47 GMT, Woodchuck Bill <[EMAIL PROTECTED]> wrote:

>  OK, so you think it is acceptable for anyone to create as many Big-8 rogue 
>  groups as they like? Some servers will carry the groups, others will not. 
>  There should be no accountability for someone doing this. There is nothing 
>  wrong with it. 
  
It is PERFECTLY acceptable for a server owner to configure their news
server to use whatever groups they wish. Nobody owns the Big 8 namespace;
nobody can force news admins to adhere to a single common version.

Russ et al offer an *advisory* service. That's all the Big-8 really is.
Commonly followed advice, yes, but advice nonetheless.

It would be nice if the admins who were using these feeds configured their
systems to only exchange these groups with other like-minded admins, given
the potential for confusion, but the failure to do so is NOT abuse of the
network.

-- 
Devin L. Ganger <[EMAIL PROTECTED]>
"Aikido is based around the central precept of letting an attack take
its natural course.  You, of course, don't want to impede that natural
flow by being in its way." -- overheard on the PyraMOO

---(end of broadcast)---
TIP 2: you can get off all lists at once with the unregister command
(send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[Marcel Beaudoin]

Woodchuck Bill <[EMAIL PROTECTED]> wrote in
news:[EMAIL PROTECTED]: 

> OK, so you think it is acceptable for anyone to create as many Big-8
> rogue groups as they like? Some servers will carry the groups, others
> will not. There should be no accountability for someone doing this.
> There is nothing wrong with it. 

I think that the question is will the "rogue" groups being created do a 
significant amount of damage to the rest of usenet that a UDP is warranted. 
In this case, recommending a UDP for a set of groups that is, from what I 
can tell, pretty much self-contained, sorta like using a shotgun to open a 
peanut.  It does the job but is way out of scale.
-- 
Marcel

---(end of broadcast)---
TIP 2: you can get off all lists at once with the unregister command
(send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[Brian Edmonds]

Woodchuck Bill <[EMAIL PROTECTED]> writes:
> If the NAN team announces a reversal of the rec.woodworking.all-ages
> result in the next few days, would you have any problem with the
> proponents sending out a control message anyway?

The proponents are entirely welcome to do so, so long as they send it
in their own name(s).  That's the way Usenet is supposed to work.
Operationally their control messages will be ignored at many to most
sites, as those site admins have chosen to follow the NAN Big-8 group
list.  That's also the way Usenet is supposed to work.

If they consulted with me in advance, I would not recommend it, since
so far as I have been able to determine, groups which go through the
NAN process get better distribution than those which do not.  If it
was better, or even equivalent, to be a "rogue" group, then we would
not have these discussions about the scrapbooks group and now the
PostgreSQL groups.

> Archiving the rogue group in Google Groups?

Google is a private site, they can carry whatever groups and whatever
content they wish to.  That is the way Usenet is supposed to work.

> If nothing else, taking no steps toward action sets a bad example,
> and might encourage others to skip the RFD and create more rogue
> groups.

As Russ has pointed out more than once, if so called "rogue" groups
are equally successful as groups that go through the NAN process, this
exposes a problem with the NAN process, not the rogue groups.

Brian.

---(end of broadcast)---
TIP 9: the planner will ignore your desire to choose an index scan if your
  joining column's datatypes do not match

Antw: Re: [GENERAL] trouble with rpmbuild on WBEL3.0/x86_64

[Marcel Gsteiger]

thanks for your reply. My distutils was in /usr/lib64/python2.2 . I installed 
python.i386 (so I had distutils in /usr/lib thereafter).  No change.
I applied the patch, again no change. My error still persists.
I would be happy if I could build postgresql without python support. Is there 
an easy way to accomplish this?
Seems that making x86_64 rpms for postgresql is somewhat tricky...

regards
--marcel

>>> Devrim GUNDUZ <[EMAIL PROTECTED]> 09.11.2004 >>>
-BEGIN PGP SIGNED MESSAGE-
Hash: SHA1


Hi,

On Mon, 1 Nov 2004, Marcel Gsteiger wrote:

> I wanted to install postgreSQL 7.4 on my ia32e box (x86_64 dual xeon PL370G4) 
> running WBEL3.0.  I looked for binary RPMs but did not find any. So I tried to
>
> rpmbuild --rebuild --define 'build9 1' postgresql-7.4.6-2PGDG.src.rpm
>
> but during configure I get the error
>
> 
> checking for python... /usr/bin/python
> checking Python installation directories... /usr/lib/python2.2
> checking how to link an embedded Python application... no
> configure: error: Python Makefile not found
> Fehler: Bad exit status from /var/tmp/rpm-tmp.75919 (%build)
>
> python-devel-2.2.3-5 is installed. On a similar i386 system (same 
> versions of rpms), this error does not happen; here I could build all 
> RPMs successfully (but, of course, not for x86_64).

python-2.2.3-5 should have distuils module installed. Do you have 
/usr/lib/python2.2/distutils ?

Anyway,

http://archives.postgresql.org/pgsql-hackers/2004-09/msg00447.php 

has a patch that fixes your problems. Could you please apply it to your 
postgresql source and retry building rpm?

> On my x86_64 system, rh-postgresql 7.3.6-7 is still installed - I 
> thought I could rebuild newer RPMS before removing the old ones. Should 
> I remove rhdb first?

No. You can still rebuild the RPMS even if you dont't have PostgreSQL 
installed on your server. But you'll need to remove the old binaries 
befora installing the new ones.

Regards,
- --
Devrim GUNDUZ 
devrim~gunduz.org   devrim.gunduz~linux.org.tr
http://www.tdmsoft.com 
http://www.gunduz.org 
-BEGIN PGP SIGNATURE-
Version: GnuPG v1.2.1 (GNU/Linux)

iD8DBQFBj/1Ktl86P3SPfQ4RAlOZAJ9xGK5i+NZ/71dVHmFw1P4nn19KfwCg4I73
q4RUJOwn5HROuPVbgP3ryOU=
=2doM
-END PGP SIGNATURE-


---(end of broadcast)---
TIP 6: Have you searched our list archives?

   http://archives.postgresql.org

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[ru . igarashi]

Woodchuck Bill <[EMAIL PROTECTED]> wrote:
>[EMAIL PROTECTED] wrote in news:[EMAIL PROTECTED]:

>> That's his perogative.  His server, his rules (or whoever's he set the
>> groups up on).  We don't have the right to dictate what groups he puts
>> on his news server.  If someone else decides to take a feed from him 
>> and allow the group on their server, same story, their server, their 
>> rules.  That kind of independence is at the foundation of usenet.  
>> While I may be displeased that the bogus groups exist, I'm similarly 
>> not going to be supportive of moves to dictate what groups he puts
>> on his server.

>Those groups are propagated to *other* servers, and they confuse lots of 
>people into thinking that they are bonafide Big-8 groups. 

I realize that, and that is part of the basis for my objection to
folks creating bogus newsgroups.  But the fact remains that it's
their server, and thus, their rules.

>Even Google is 
>either confused or careless about the status of those groups. 

Google is just one of those services that doesn't give a damn, and 
thus exercising their perogative, just like any number of servers
that wish to collect as many groups as they can.

>If the NAN 
>team announces a reversal of the rec.woodworking.all-ages result in the 
>next few days, would you have any problem with the proponents sending out a 
>control message anyway? 

That's a separate, though related, matter.  If the proponents owned
or had access to a server and decided to create r.w.a-a, that's
their perogative.  If they wish to spread r.w.a-a, well, that turns
out to be their perogative, too.  Anyone is allowed to do this.
It's totally up to rest of the news server admins to honor or reject 
the newgroup.  Again, their server, their rules.  I wouldn't be
happy about it, but I'm not going to force them to do otherwise.

>Archiving the rogue group in Google Groups? 

Google's choice.

>If 
>nothing else, taking no steps toward action sets a bad example, and might 
>encourage others to skip the RFD and create more rogue groups.

As Russ & Co. say, if it comes to that, then 1) the existing process
deserves to be ignored, 2) news admins have voted with their feet
and decided to go with another process, or found a better way.  I
don't like the idea of being bypassed, but they are right.  It's our
job to find ways to make "non-bogus groups" more attractive.  This 
process has always been predicated on voluntary acceptance of the 
results.  I don't think Russ & Co are willing to abandon that 
philosophy, even if it means the (deserved) demise of the Big-8 
process.

ru

-- 
My standard proposals rant:
Quality, usefulness, merit, or non-newsgroups popularity of a topic 
is more or less irrelevant in creating a new Big-8 newsgroup.  
Usenet popularity is the primary consideration.

---(end of broadcast)---
TIP 7: don't forget to increase your free space map settings

[GENERAL] The classic "NEW used in query that is not in a rule" problem again

[John White]

It seems I'm not the first to ask this question but there seem to be
very few answers. I am implementing an audit log facility where INSERT's
or UPDATE's to a number tables get logged to a single table. Each row in
the logging table stores data on one field change in the tables being
logged. My function is available here
http://rafb.net/paste/results/JwQeqj35.html. Basically a SQL string is
generated for each field in the inserted record, this SQL refres to
"new" but when it's executed I get "NEW used in query that is not in a
rule

Here's the bit of SQL that builds the string

(Beware all the single quoting - my eyes have only stopped bleeding)

-- start
FOR recFields in SELECT attname FROM pg_attribute WHERE attrelid =
TG_RELID AND attnum >= 0 loop
auditfieldid =
nextval(\'seq_tbl_auditfields_fld_auditfieldid\'::text);
SQL := \'INSERT INTO tbl_auditfields(fld_auditfieldid,
fld_audittableid, fld_fieldname, fld_newdata) VALUES (\';
SQL := SQL || auditfieldid::text || \', \' ||
audittableid::text || \', \';
SQL := SQL ||  || recFields.attname || 
|| \', new.\' || recFields.attname || \'::text);\';
raise NOTICE ''SQL = %'', SQL;
EXECUTE SQL;
END LOOP;
-- end
 
and here's a sample string that's generated (as reported by the RAISE
NOTICE) and it looks ok.

INSERT INTO tbl_auditfields(fld_auditfieldid, fld_audittableid,
fld_fieldname, fld_newdata) 
VALUES (65, 11, 'fld_uid', new.fld_uid::text);

When I execute this I get the error above. Also the docs for the RENAME
command seem to hint that you can get around this problem by renaming
'new', but rename is broken right now, and is low priority for fixing.

If I could even evaluate the new. outside the SQL and put it's
value in instead, I'd be happy. Any help would be much appreciated (I
really don't want to have to automatically generate a trigger function
for each table).

Cheers,
John.


---(end of broadcast)---
TIP 9: the planner will ignore your desire to choose an index scan if your
  joining column's datatypes do not match

Re: [GENERAL] how to edit a function from psql?

[Karim Nassar]

On Sun, 2004-11-07 at 20:36, Thomas F.O'Connell wrote:
> What version of postgres are you using? In postgresql-7.4.6, I get an  
> empty query buffer when I try what you describe here.

7.4.5

I got it. Apparently the query buffer is the last run query (good or
bad). So, to get my function into the buffer, I would have to cut and
paste it (which I do often, and is what my last post showed). Backslash
functions are not included. I assume you ran no queries before \e, and
got an empty buffer.

Seems like \e is merely a convenient way to deal with long queries. I
always have my editor open, so this isn't much of a win for me.

Still, it might be nice to have a way to get the function into the
buffer without cutting and pasting from the (normally open) editor. 

As a point of interest...

1)  orfs=# select * from person
2)  orfs-# ;
3)  ERROR:  relation "person" does not exist
4)  orfs=# \e
5)  orfs-# select
6)  orfs-# *
7)  orfs-# from
8)  orfs-# person
9)  orfs-# ;
10) ERROR:  syntax error at or near "select" at character 23
11) orfs=# \e
12) orfs=# select * from person;
13) ERROR:  relation "person" does not exist
14) orfs=# \e


At line 4, the query buffer has:
~
select * from person
~

Note that there is no semi-colon.

At line 11, the contents of the query buffer is:
~
select * from person

select
* 
from 
person
~~

And at line 14:
~~
select * from person;
~~


It seems that the query editor just dumps the contents to the command
line. Why the semi-colons from lines 2 and 9 aren't in the query buffers
is a mystery to me.

\<.



---(end of broadcast)---
TIP 3: if posting/reading through Usenet, please send an appropriate
  subscribe-nomail command to [EMAIL PROTECTED] so that your
  message can get through to the mailing list cleanly

Re: [GENERAL] subselect, order by and left join

[Karim Nassar]

On Mon, 2004-11-08 at 05:54, Stephan Szabo wrote:
> > If I have a subselect with an ORDER BY, and I LEFT JOIN the result with the
> > other table, is the order maintained? Or is PostgreSQL free to return the 
> > rows
> > in any order, after the join?
> 
> AFAIK, you have no guarantees as to the output order unless you have
> another order by.  The join may destroy the ordering, so even if you get
> the ordering you want right now, you shouldn't rely on it.

Try http://www.varlena.com/varlena/GeneralBits/35.html

If you look under "Alphanumeric Sorting" about halfway down the page,
you will find the path toward the magic you are looking for.

A particularly nasty working example:

SELECT sort_order, col, code, description, units,
   TO_CHAR(min_value, 'FM999D90') AS min_value, 
   TO_CHAR(max_value, 'FM999D90') AS max_value,
   value AS dv_text, id_datatype_value
 FROM ( SELECT *, (CASE WHEN (SUBSTRING(dv.value FROM '^[0-9\.]{1,3}') IS NOT 
NULL)
THEN (SUBSTRING(dv.value FROM '^[0-9\.]{1,3}')::numeric)
ELSE NULL
END) AS sort_order
 FROM datasheet ds JOIN datasheet_column dc  USING (id_datasheet)
   JOIN datatype dt  USING (id_datatype)
  LEFT JOIN datatype_value dvUSING (id_datatype)
WHERE id_datasheet = '7') END_SORT_FU
ORDER BY col, sort_order, value

\<.



---(end of broadcast)---
TIP 9: the planner will ignore your desire to choose an index scan if your
  joining column's datatypes do not match

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[Joseph Daniel Zukiger]

"Arthur L. Rubin" <[EMAIL PROTECTED]> wrote in message news:<[EMAIL 
PROTECTED]>...
> Rebecca Ore wrote:
> 
> > Since they have traffic and are bidirectional, and since he doesn't have
> > any objections, it looks like creating the groups by fiat would work
> > just fine and avoid a lot of misinformed "but we didn't ask for this"
> > squabbling.  Same as with the scrapbook group.
> 
> It's NOT bidirectional, as I understand it.  Posts to the newsgroup
> go to the mailing list IF the sender is subcribed -- otherwise, they
> go to the list owner for (possible) approval.  If the list owner
> is willing to unmoderate the mailing list, it WOULD be bidirectional.

Marc says he's looking into the setup, to see if he can fix that. I
think his primary problem is that he wants to devote full time to
postgresql development. He's part of the core team, as I recall.

> So -- either the list owner must unmoderate the mailing list,
> or the groups must be moderated. 

A moderationg policy of no SPAM should be sufficient, right?

> I see no reasonable third 
> option.  Under the circumstances, I'm not sure I'd accept
> Marc as moderator on the Usenet side.

Would a group of moderaters to help Marc carry the load be
appropriate? Or would it be preferable to not have Marc moderating
posts from usenet at all? (I'm thinking you mean the former.)

(Apologies for posting through Google.)

JDZ

---(end of broadcast)---
TIP 7: don't forget to increase your free space map settings

[GENERAL] PostGreSQL to Access Updatable recordset

[Sim Zacks]

I have a postgresql backend with an access front end and I am trying to
redefine the recordset of the form to use an ADO recordset. The problem is
that the CursorType always changes to AdOpenStatic, even if I choose
adOpenDynamic. If anyone has any thoughts, please let me know. The form
works great for viewing, but I cannot update or insert any new records.

Below is the code I am using:

   connectstr = "DSN=SAP_PG;uid=postgres"
Set conn = New ADODB.Connection
conn.Open (connectstr)
rs.CursorLocation = adUseClient
rs.Open "select ProductID,ProductName,ProductTypeID,StockTypeID from
Products order by ProductName", conn, adOpenDynamic, adLockOptimistic
Set Me.Recordset = rs
 Me.Requery

The conn object contains the following:
Provider=MSDASQL.1;Extended
Properties="DSN=SAP_PG;DATABASE=sap;SERVER=10.1.1.76;PORT=5432;UID=username;
PWD=password;ReadOnly=0;Protocol=6.4;FakeOidIndex=0;ShowOidColumn=0;RowVersi
oning=1;ShowSystemTables=0;ConnSettings=;Fetch=100;Socket=4096;UnknownSizes=
0;MaxVarcharSize=4094;MaxLongVarcharSize=4094;Debug=0;CommLog=0;Optimizer=1;
Ksqo=1;UseDeclareFetch=0;TextAsLongVarchar=0;UnknownsAsLongVarchar=0;BoolsAs
Char=0;Parse=0;CancelAsFreeStmt=0;ExtraSysTablePrefixes=dd_;;LFConversion=1;
UpdatableCursors=1;DisallowPremature=0;TrueIsMinus1=0;BI=0;ByteaAsLongVarBin
ary=0;UseServerSidePrepare=0"


In answer to your upcoming questions:

1) ProductID is the primary key on the table.
2) I do not want to use linked forms, they go much slower then connecting
directly through the ADO recordset.
3) I put a breakpoint right after the open statement and checked the value
of rs.CursorType and that is when I saw it was adOpenStatic
4) I am using PostGreSQL 8.0beta1
5) ODBC driver 7.03.02.00



---(end of broadcast)---
TIP 5: Have you checked our extensive FAQ?

   http://www.postgresql.org/docs/faqs/FAQ.html

Re: [GENERAL] Visual Designer in linux?

[Gary G. Taylor]

Eric wrote:

> Is there a visual designer (open source) in linux for database?  I would
> like to developp my data model on the computer...
> 
> I see "dia" with uml library but... it won't export to SQL.

GerWin: http://es.gnu.org/~jemarch/cgi-bin/index.pl?Site=GerWin
Design your database as though you were using Microsoft Access (including
relationships), it will then write the necessary code to implement that
design in several different languages, including Postgres.

-- 
Gary G. Taylor * Rialto, CA
gary at donavan dot org / http:// geetee dot donavan dot org
"The two most abundant things in the universe
are hydrogen and stupidity." --Harlan Ellison

---(end of broadcast)---
TIP 7: don't forget to increase your free space map settings

Re: [GENERAL] Functions in postgres

[Sim Zacks]

It would seem that trying to hide code while using 
an open source system is just wrong. 8-)
 
That being said -- You would grant the 
user/group execute rights on the function, but no read rights on the system 
table that contains the function definition, pg_proc. 
How is this user supposed to call the function? 
Does he have a cutomized application or does he have access to an SQL interface, 
such as Psql or PGAdmin? 
 
Finally, I don't think you have to 
apologize for either bad English or for being Spanish, the native English 
speakers (and I only speak for myself here) appreciate the extra effort 
that non-English speaking people put in to allow the project to be fully 
internationalized. You should be proud of being Spanish, because even though 
there were a lot of bad points in Spanish history, Christopher Colombus did 
supposedly discover America, or something similar, and without that we wouldn't 
all be speaking English today.
 
sim
 

  ""Tk421"" <[EMAIL PROTECTED]> wrote in 
  message news:[EMAIL PROTECTED]...
      I've made some functions, an i 
  want these functions to be used by a specified user, but i don't want this 
  user can see the code of the functions. Does somebody knows how can i do 
  this?
   
      Sorry about my english, i'm 
  spanish.
  Víctor Robador Capel
   
  Análisis de Sistemas y Programación
   
  www.creativosdolmen.com

Re: [GENERAL] RFD: comp.databases.postgresql.general

[Woodchuck Bill]

Polarhound <[EMAIL PROTECTED]> wrote in news:kM2dnd_0xq99yw3cRVn-
[EMAIL PROTECTED]:

>> (BTW, since the person responsible for setting up the rogue groups
>> appears to be aware of the discussion to legitimize the groups, why
>> isn't he taking part in it?)
>> 
> 
> That's my whole point.. He's responded in the mailing list to an issue 
> being raised here.  That does him about as much good as wiping his rear 
> with 20 grit sandpaper.

He's being defiant .. Rebel without a CFV! ;-)

-- 
Bill

---(end of broadcast)---
TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[Arthur L. Rubin]

Joseph Daniel Zukiger wrote:
> 
> "Arthur L. Rubin" <[EMAIL PROTECTED]> wrote in message news:<[EMAIL 
> PROTECTED]>...
> > Rebecca Ore wrote:
> >
> > > Since they have traffic and are bidirectional, and since he doesn't have
> > > any objections, it looks like creating the groups by fiat would work
> > > just fine and avoid a lot of misinformed "but we didn't ask for this"
> > > squabbling.  Same as with the scrapbook group.
> >
> > It's NOT bidirectional, as I understand it.  Posts to the newsgroup
> > go to the mailing list IF the sender is subcribed -- otherwise, they
> > go to the list owner for (possible) approval.  If the list owner
> > is willing to unmoderate the mailing list, it WOULD be bidirectional.
> 
> Marc says he's looking into the setup, to see if he can fix that. I
> think his primary problem is that he wants to devote full time to
> postgresql development. He's part of the core team, as I recall.
> 
> > So -- either the list owner must unmoderate the mailing list,
> > or the groups must be moderated.
> 
> A moderationg policy of no SPAM should be sufficient, right?
> 
> > I see no reasonable third
> > option.  Under the circumstances, I'm not sure I'd accept
> > Marc as moderator on the Usenet side.
> 
> Would a group of moderaters to help Marc carry the load be
> appropriate? Or would it be preferable to not have Marc moderating
> posts from usenet at all? (I'm thinking you mean the former.)

Actually, I meant the latter, if (as it appears) he is responsible
for and unrepentant about the lists being gated to Usenet in
the first place.  But it's too early in the process for that to
be my final conclusion.

-- 
This account is subject to a persistent MS Blaster and SWEN attack.
I think I've got the problem resolved, but, if you E-mail me
and it bounces, a second try might work.
However, please reply in newsgroup.

---(end of broadcast)---
TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]

[GENERAL] need simple strategy for universal extension table

[TJ Talluto]

Instead of putting the same 12 columns on every table (these 12 columns
contain info about who created the record and when, for example)... It may
be more efficient to make a new table to hold that data.

This new table would act as a universal extension table.  Instead of having
FKs back to any particular table, it would contain regular keys that point
back to whatever table::record is its source.

I was thinking of using two keys only:

xmOwnerInfo
COL1 PK* oidParentTable
COL2 PK* oidParentTablesRecord
COL3-15 [attribs]

Thoughts?

-- 
TJ Talluto
torpedo51 at yahoo dot com

---(end of broadcast)---
TIP 5: Have you checked our extensive FAQ?

   http://www.postgresql.org/docs/faqs/FAQ.html

Re: [GENERAL] How many secondary databases can I create?

[Dawid Kuroczko]

On Fri, 12 Nov 2004 17:20:59 +0100, Michelle Konzack
<[EMAIL PROTECTED]> wrote:
> Please note, that I do not run seperatly databases.
> I like only to split it per $USER physicaly from the main database.
> 
> I like to have this setup for my new Virtual-Webserver where I have
> for each VHost a local $USER.

Some time ago I talked with a friend working at some ISP, and we were
talking about PostgreSQL's superiority to hmm, other RDBMSs... ;)
Blah, blah, blah.  The question is - how to give many many users
access to PostgreSQL...   I see three approaches, and all of these
have pros and cons.

1) One pgsql per user -- in other words each user runs its own copy of
pgsql -- waste of memory (each user having its own shared mem, etc),
but can enforce quota limits, etc.  Hard to keep all those copies of
pgsql running.

2) one pgsql database per user.  Probably most common. :)  Each user
has full pgsql database at her disposal.  It's a bit hard to manage quotas,
dbsize comes in handy but it involeves creating some sort of "quota
enforcing" daemon...

3) one pgsql database with one schema per user.  Interesting, can very
easily create many "databases", does not give full pgsql power, harder
to manage quotas than previous model...

Hmm, I wonder what are the feelings of people using such models.
I did not use them, so I have no idea.  And I wonder how problematic
is "quota enforcement" there

   Regards.
 Dawid

---(end of broadcast)---
TIP 4: Don't 'kill -9' the postmaster

Re: [GENERAL] List of postgresql rogue groups (was Re: Important Info on comp.databases.postgresql.general)

[Max]

>
> [EMAIL PROTECTED] wrote in news:[EMAIL PROTECTED]:
>
> >>The other issue is that I would like to add the other postgresql
> >>groups for consideration to be included into the big 8.  However there
> >>are quite a few of them, and I don't know if all of them deserve to be
> >>there. They are all under comp.databases.postgresql.*, and I highly
> >>doubt Marc would want to eliminate them.

I think Marc understands by now that these lists won't be eliminated. It's a
parallel process. Let's start with a few and create the other groups down
the road. An RFD with 21 groups has a good chance to be rejected, not
because they are not relevant, just because it's too many at one time. I
assume a second RFD should be 'filed' in 6 months with the remaining groups,
and should be nothing more than a cruise.

Nax



---(end of broadcast)---
TIP 5: Have you checked our extensive FAQ?

   http://www.postgresql.org/docs/faqs/FAQ.html

Re: [GENERAL] The classic "NEW used in query that is not in a rule" problem again

[Alvaro Herrera]

On Tue, Nov 09, 2004 at 11:00:57PM -, John White wrote:
> It seems I'm not the first to ask this question but there seem to be
> very few answers. I am implementing an audit log facility where INSERT's
> or UPDATE's to a number tables get logged to a single table. Each row in
> the logging table stores data on one field change in the tables being
> logged.

This is hard to do with plpgsql, I don't understand why people keeps trying.
It's much easier with PL/Tcl.  An example, if you had these tables

CREATE TABLE usuarios
(usuario_id int,
 nombre text);

CREATE TABLE usuarios_audit
(usuario_id int,
 nombre text,
 op text,
  fecha timestamp with time zone);


You could do something like

CREATE OR REPLACE FUNCTION
audita_usuarios() RETURNS trigger AS '
spi_exec "INSERT INTO usuarios_audit
VALUES ($NEW(usuario_id),
''[ quote $NEW(nombre) ]'',
''[ quote $TG_op ]'',
now())"
return [array get NEW]
' LANGUAGE pltcl;

CREATE TRIGGER audita_usuarios
 BEFORE UPDATE OR INSERT OR DELETE
 ON usuarios FOR EACH ROW
 EXECUTE PROCEDURE audita_usuarios();

Note that you can give parameters to the function, as illustrated
by this other example (doing case folding, but should be trivial
to understand how to modify it):


CREATE TABLE a_table (
column_1text,
column_2text
);

CREATE OR REPLACE FUNCTION minusculas()
RETURNS trigger AS '
foreach key $args {
if {[info exists NEW($key)]} {
  set NEW($key) [string tolower $NEW($key)]
}
}
return [array get NEW]
' LANGUAGE pltcl;

CREATE TRIGGER minusculizar
 BEFORE INSERT OR UPDATE ON a_table
 FOR EACH ROW EXECUTE PROCEDURE
 minusculas('column_1', 'column_2');


You should be able to do whatever you want to do by extending
this examples, much more easily than by using plpgsql.
No wonder your eyes were bleeding.

I think the second example is almost verbatim from Elein Mustain's
excellent General Bits column.

-- 
Alvaro Herrera (<[EMAIL PROTECTED]>)
"Pensar que el espectro que vemos es ilusorio no lo despoja de espanto,
sólo le suma el nuevo terror de la locura" (Perelandra, CSLewis)

---(end of broadcast)---
TIP 3: if posting/reading through Usenet, please send an appropriate
  subscribe-nomail command to [EMAIL PROTECTED] so that your
  message can get through to the mailing list cleanly

Re: [GENERAL] PostGreSQL to Access Updatable recordset

[Alvaro Herrera]

On Tue, Nov 09, 2004 at 01:53:18PM +0200, Sim Zacks wrote:
> I have a postgresql backend with an access front end and I am trying to
> redefine the recordset of the form to use an ADO recordset. The problem is
> that the CursorType always changes to AdOpenStatic, even if I choose
> adOpenDynamic. If anyone has any thoughts, please let me know. The form
> works great for viewing, but I cannot update or insert any new records.

Postgres does not support updatable cursors yet.  Maybe that's why the ADO
driver changes the cursor type from AdOpenDynamic to AdOpenStatic.

-- 
Alvaro Herrera (<[EMAIL PROTECTED]>)
"Nunca se desea ardientemente lo que solo se desea por razón" (F. Alexandre)

---(end of broadcast)---
TIP 6: Have you searched our list archives?

   http://archives.postgresql.org

[GENERAL] Mysterious Death of postmaster (-9)

[Gregory S. Williamson]

Dear peoples,

We had an oddness today with one of of postgres servers (Dell 2 CPU box running 
linux) and postgres 7.4. The server was under heavy load (50+ for a 1 minutes 
spike; about 20 for the 15 minute average) with about 250 connections (we still 
don't understand the heavy load itself).

Looking in the logs I see:
2004-11-13 13:30:28 LOG:  unexpected EOF on client connection
2004-11-13 13:30:40 LOG:  unexpected EOF on client connection
2004-11-13 13:38:28 LOG:  could not send data to client: Broken pipe
2004-11-13 13:42:15 LOG:  server process (PID 30272) was terminated by signal 9
2004-11-13 13:42:16 LOG:  terminating any other active server processes
2004-11-13 13:42:16 WARNING:  terminating connection because of crash of another
 server process

The EOFs are almost certainly Proxool closing connections from the client to 
the database.

The sysad who was on call today swears he didn't send a kill signal (or any 
signal at all) -- suddenly the load dropped off and the server was down. It has 
restarted normally and shows no signs of being worse for the wear (this is 
really a read-only db so data corruption chances are minimal, I think).

Just to rule out any internal chances, is there any way this shutdown could 
have been triggered from within postgres itself ? Can anyone construct any 
scenarios in which Linux, postgres or proxool could have done this without 
human intervention ?

I have looked through manuals and some FAQs and newsgroup discussions and my 
gut feeling is that this can't be from postgres, but I thought I'd ask in the 
chance that I am, as is often the case, Unclear On The Concept.

Thanks for any illumination,

Greg Williamson
DBA
GlobeXplorer LLC 

ps if this is not the right list please let know what might be an appropriate 
one. gracias!

---(end of broadcast)---
TIP 5: Have you checked our extensive FAQ?

   http://www.postgresql.org/docs/faqs/FAQ.html

Re: [GENERAL] Mysterious Death of postmaster (-9)

[Stephan Szabo]

On Sat, 13 Nov 2004, Gregory S. Williamson wrote:

> Looking in the logs I see:
> 2004-11-13 13:30:28 LOG:  unexpected EOF on client connection
> 2004-11-13 13:30:40 LOG:  unexpected EOF on client connection
> 2004-11-13 13:38:28 LOG:  could not send data to client: Broken pipe
> 2004-11-13 13:42:15 LOG:  server process (PID 30272) was terminated by signal 
> 9
> 2004-11-13 13:42:16 LOG:  terminating any other active server processes
> 2004-11-13 13:42:16 WARNING:  terminating connection because of crash of 
> another
>  server process

> Just to rule out any internal chances, is there any way this shutdown
> could have been triggered from within postgres itself ? Can anyone
> construct any scenarios in which Linux, postgres or proxool could have
> done this without human intervention ?

Is it possible that you ran into the out of memory killer?  That's the
most likely thing beyond admin intervention I can think of.


---(end of broadcast)---
TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]

Re: [GENERAL] Mysterious Death of postmaster (-9)

[Alvaro Herrera]

On Sat, Nov 13, 2004 at 02:39:38PM -0800, Gregory S. Williamson wrote:

Gregory,

> We had an oddness today with one of of postgres servers (Dell 2 CPU box
> running linux) and postgres 7.4. The server was under heavy load (50+ for a 1
> minutes spike; about 20 for the 15 minute average) with about 250 connections
> (we still don't understand the heavy load itself).
> 
> Looking in the logs I see:
> 2004-11-13 13:30:28 LOG:  unexpected EOF on client connection
> 2004-11-13 13:30:40 LOG:  unexpected EOF on client connection
> 2004-11-13 13:38:28 LOG:  could not send data to client: Broken pipe
> 2004-11-13 13:42:15 LOG:  server process (PID 30272) was terminated by signal 
> 9

This looks an awful lot like the Linux Out-Of-Memory killer got you.
This happens when the Linux kernel overcommits memory.  There is something
about this on the documentation, and has been discussed in the past
here.  Please see the archives (www.pgsql.ru; look for "OOM killer" and
"linux overcommit").

Luckily it didn't get your postmaster, as has happenned to other
people ...

-- 
Alvaro Herrera (<[EMAIL PROTECTED]>)
"XML!" Exclaimed C++.  "What are you doing here? You're not a programming
language."
"Tell that to the people who use me," said XML.

---(end of broadcast)---
TIP 7: don't forget to increase your free space map settings

Re: [GENERAL] RFD: comp.databases.postgresql.general

[Martijn van Oosterhout]

On Mon, Nov 08, 2004 at 05:28:57PM +, Woodchuck Bill wrote:
> [EMAIL PROTECTED] ("Marc G. Fournier") wrote in 
> news:[EMAIL PROTECTED]:
> 
> > Unless its spam, it goes through ... I don't (nor have I ever) refused a 
> > post based on content other then spam ... even if its anti-PostgreSQL 
> > *shrug*
> 
> The problem with the system is that the spam *all* gets posted to Usenet, 
> but not the mailing lists. The mailing lists may be moderated, but the 
> newsgroups are not. That needs to be changed.
> 
> Marc, please stop removing news.groups from your replies.

I'm on the mailing list and I'm not seeing news.groups in any of the
usual headers. It does sometimes appear in a seperate header
"Newsgroups". Besides, my mailer doesn't post to newsgroups so I'd have
to setup a newsreader just for that. I'd be surprised if I were the
only one...
-- 
Martijn van Oosterhout   <[EMAIL PROTECTED]>   http://svana.org/kleptog/
> Patent. n. Genius is 5% inspiration and 95% perspiration. A patent is a
> tool for doing 5% of the work and then sitting around waiting for someone
> else to do the other 95% so you can sue them.


pgpiDrjB7s8dB.pgp
Description: PGP signature

Re: [GENERAL] Mysterious Death of postmaster (-9)

[Gregory S. Williamson]

Thanks Alvaro and Steven -- this may in fact be what happened as the monitor 
showed that at about that time memory definitely was taxed and showed oddnesses.

I'll read up on this -- thanks very much for the (promising) clue!

Greg W.


-Original Message-
From:   Alvaro Herrera [mailto:[EMAIL PROTECTED]
Sent:   Sat 11/13/2004 3:06 PM
To: Gregory S. Williamson
Cc: [EMAIL PROTECTED]
Subject:Re: [GENERAL] Mysterious Death of postmaster (-9)
On Sat, Nov 13, 2004 at 02:39:38PM -0800, Gregory S. Williamson wrote:

Gregory,

> We had an oddness today with one of of postgres servers (Dell 2 CPU box
> running linux) and postgres 7.4. The server was under heavy load (50+ for a 1
> minutes spike; about 20 for the 15 minute average) with about 250 connections
> (we still don't understand the heavy load itself).
> 
> Looking in the logs I see:
> 2004-11-13 13:30:28 LOG:  unexpected EOF on client connection
> 2004-11-13 13:30:40 LOG:  unexpected EOF on client connection
> 2004-11-13 13:38:28 LOG:  could not send data to client: Broken pipe
> 2004-11-13 13:42:15 LOG:  server process (PID 30272) was terminated by signal 
> 9

This looks an awful lot like the Linux Out-Of-Memory killer got you.
This happens when the Linux kernel overcommits memory.  There is something
about this on the documentation, and has been discussed in the past
here.  Please see the archives (www.pgsql.ru; look for "OOM killer" and
"linux overcommit").

Luckily it didn't get your postmaster, as has happenned to other
people ...

-- 
Alvaro Herrera (<[EMAIL PROTECTED]>)
"XML!" Exclaimed C++.  "What are you doing here? You're not a programming
language."
"Tell that to the people who use me," said XML.




---(end of broadcast)---
TIP 9: the planner will ignore your desire to choose an index scan if your
  joining column's datatypes do not match

Re: [GENERAL] I spoke with Marc from the postgresql mailing list.

[Patrick B Kelly]

 If the process drags on for too long, people will start to get 
annoyed with the process and lose interest. I speak from experience.
This hasn't dragged on too long? People aren't already annoyed? 
Seriously, the usenet people losing interest in this "problem" might be 
the best case scenario.

Patrick B. Kelly
--
  http://patrickbkelly.org
---(end of broadcast)---
TIP 2: you can get off all lists at once with the unregister command
   (send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])

Re: [GENERAL] need simple strategy for universal extension table

[Mike Rylander]

On Tue, 09 Nov 2004 19:20:20 GMT, TJ Talluto <[EMAIL PROTECTED]> wrote:
> This new table would act as a universal extension table.  Instead of having
> FKs back to any particular table, it would contain regular keys that point
> back to whatever table::record is its source.
> 
> I was thinking of using two keys only:
> 
> xmOwnerInfo
> COL1 PK* oidParentTable
> COL2 PK* oidParentTablesRecord
> COL3-15 [attribs]
> 

I am doing something similar to this, though at this point I'm not
using it for storing creator/updator fields.  I need to be able to
group any arbitrary set of rows from multiple tables together and
apply sets of attributes to them.  I have a central 'entity' table
that holds the table name and the value of the 'id' column from that
table.  Then I attach two triggers to each table whos rows I want to
track. I am using tablename and a BIGSERIAL 'id' column because OIDs
on tables

  1) are only a 32 bit INT and I may very well wrap around on them,
whereas a BIGSERIAL is a 64 bit INT and

  2) the fact that OIDs may go away in a future version of PG.

Here's exactly what I'm doing (NOTE: this is for 8.0beta, so you will
have to adjust the quoting on the trigger functions for 7.x):

CREATE SCHEMA func;
CREATE SCHEMA entity;

CREATE TABLE entity.authority_list (
id  BIGSERIAL PRIMARY KEY,
entity  BIGINT,
entity_type text
) WITHOUT OIDS;

CREATE FUNCTION func.add_entity_entry () RETURNS TRIGGER AS $func$
BEGIN
INSERT INTO entity.authority_list (entity,entity_type)
 VALUES (NEW.id,TG_ARGV[0] || '.' || TG_RELNAME);
RETURN NEW;
END;
$func$ LANGUAGE 'plpgsql';

CREATE FUNCTION func.remove_entity_entry () RETURNS TRIGGER AS $func$
BEGIN
DELETE FROM entity.authority_list
WHERE entity = OLD.id AND
entity_type = TG_ARGV[0] || '.' || TG_RELNAME;
RETURN OLD;
END;
$func$ LANGUAGE 'plpgsql';

CREATE TABLE someschema.sometable ( id BIGSERIAL, name TEXT );

CREATE TRIGGER sometable_add_entity_trig
AFTER INSERT ON someschema.sometable
FOR EACH ROW
EXECUTE PROCEDURE func.add_entity_entry(someschema);

CREATE TRIGGER sometable_remove_entity_trig
BEFORE DELETE ON someschema.sometable
FOR EACH ROW
EXECUTE PROCEDURE func.remove_entity_entry(someschema);


I am supplying the schema name to the trigger because the relation
name passed in as TG_RELNAME is the schema unqualified table name and
I have the same table name in several schemas.

Any comments on any of this would be very welcome.

-- 
Mike Rylander
[EMAIL PROTECTED]
GPLS -- PINES Development
Database Developer

---(end of broadcast)---
TIP 4: Don't 'kill -9' the postmaster

Re: [GENERAL] I spoke with Marc from the postgresql mailing

[Gary L. Burnore]

At 07:30 PM 11/13/2004, you wrote:
 If the process drags on for too long, people will start to get annoyed 
with the process and lose interest. I speak from experience.
This hasn't dragged on too long?
Not at all.
People aren't already annoyed?
Not those that matter.
Seriously, the usenet people losing interest in this "problem" might be 
the best case scenario.
Ya know, that's kinda like when a business owner hopes the neigborhood will 
forget about a toxic waste spill.

In this case, it's not going to be forgotten as long as someone's shoving 
your emails up USENet's proverbial ass.  If you REALLY want it over with 
quick, shut off the mail-to-news.


---(end of broadcast)---
TIP 4: Don't 'kill -9' the postmaster

[GENERAL] How the function written on pl/pgsql can be called by the function written on plperl.

[Anatoly Okishev]

How the function written on pl/pgsql can be called 
by the function written on plperl.
 
Error is:
 
error from function: Undefined subroutine 
&Safe::Root2::get_national_code called.

Re: [GENERAL] I'm about to release the next postgresql RFD. Comments wanted.

[Kenneth Downs]

[EMAIL PROTECTED] wrote:

> The new proposal should have all of the lists. I am amazed that an
> unknown interloper is trying to dictate which lists should be voted on.
> I am affiliated with a network of over 60 PostgreSql users/developers,
> and I will e-mail each one of those people a ballot. If your next
> proposal does not contain every single one of the lists, you can expect
> essentially all of those 60 people to vote no. I will campaign
> passionately against your proposal if you leave out even one of the
> lists. Do the right thing if you are sincerely trying to help the list
> subscribers.
> 
> This is my last post to news.groups about this.




-- 
Kenneth Downs
Use first initial plus last name at last name plus literal "fam.net" to
email me

---(end of broadcast)---
TIP 7: don't forget to increase your free space map settings

[GENERAL] Large database

[Alexander Antonakakis]

I would like to ask the more experienced users on Postgres database a 
couple of questions I have on a db I manage with a lot of data. A lot of 
data means something like 15.000.000 rows in a table. I will try to 
describe the tables and what I will have to do on them :)
There is a table that has product data in the form of
Table product:
product_id varchar(8),
product_name text

and
product actions table:
product_id varchar(8),
flow char(1),
who int,
where int,
value float.
I will have to make sql queries in the form "select value from 
product_actions where who='someone' and where='somewhere' and maybe make 
also some calculations on these results. I allready have made some 
indexes on these tables and a view that joins the two of them but I 
would like to ask you people if someone is using such a big db and how 
can I speed up things as much as it is possible on this ... these 
product_actions tables exists for each year from 1988 till 2003 so this 
means a lot of data...

Thanks in Advance

---(end of broadcast)---
TIP 9: the planner will ignore your desire to choose an index scan if your
 joining column's datatypes do not match

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[Joseph Daniel Zukiger]

[EMAIL PROTECTED] wrote in message news:<[EMAIL PROTECTED]>...
> The lists are not bogus and your suggestion is not acceptable. Many of
> the list subscribers have no connection to Usenet. They will be mailed
> the ballots, regardless of what you or anyone else say. They are the
> people who will be directly affected by this.
> 
> Mr. Pauli Vaurio
> ...

Pauli --

"Bogus" is being used here in a technical sense. It is not a judgement
of value.

The groups in question have not conformed to certain rules/customs. In
that sense, their names are "bogus" -- not official, conforming use of
the namespaces the usenet organization has reserved.

The fact that these groups apparently have the most right to the names
in question is a big point in their favor, but the group maintainer
did not go through the customary processes in reserving the names and
requesting listing. So they are not official, thus they are "bogus".

It looks like I'm may have to finally subscribe through my isp and
learn how to configure a newsreader in order to vote in favor, instead
of posting through google all the time. That should speed my access to
usenet up quite a bit. Oh, well.

JouDanZuki

---(end of broadcast)---
TIP 7: don't forget to increase your free space map settings

Re: [GENERAL] I'm about to release the next postgresql RFD. Comments wanted.

[Brian Mailman]

[EMAIL PROTECTED] wrote:
I am affiliated with a network of over 60 PostgreSql users/developers,
and I will e-mail each one of those people a ballot...
You should not do that.
Those ballots will be invalidated since only the official ballots that 
are posted or obtained directly from the votetaker are going to be counted.

B/
---(end of broadcast)---
TIP 7: don't forget to increase your free space map settings

Re: [GENERAL] I'm about to release the next postgresql RFD. Comments wanted.

[Jon Bell]

In article <[EMAIL PROTECTED]>,
Klaas  <[EMAIL PROTECTED]> wrote:
>In article <[EMAIL PROTECTED]>,
> Mike Cox <[EMAIL PROTECTED]> wrote:
>
>> comp.databases.postgresql.general
>
>How about just comp.databases.postgresql or (shudder) 
>comp.databases.postgresql.misc?

The whole point of this proposal is that c.d.p.general *already* exists in 
an unofficial sense on some servers, and that it should be made 
"official."  Creating a new group with a different name will simply 
confuse matters on those servers that already carry c.d.p.general.

-- 
Jon Bell <[EMAIL PROTECTED]> Presbyterian College
Dept. of Physics and Computer ScienceClinton, South Carolina USA

---(end of broadcast)---
TIP 8: explain analyze is your friend

[GENERAL] Value of serial data type after insert.

[Ernest Kim]

I was wondering if there was a way to get the value of serial data
type after an insert.  For example, you have the following table:

create table my_names   (
   id   serial unique,
   name varchar(10)
);

And then you do the following:

insert into my_names values ('Ernie');

How do I get the value of the "id" column after the insert command?  I
want to make sure that I avoid race conditions, for example:

User A: insert into my_names values ('Ernie');
User B: insert into my_names values ('Bob');
User A: select last_value from my_names_id_seq;
User B: select last_value from my_names_id_seq;

Both user A and B end up with the same returned value after the select
command.  

Note this would be coded through PHP if this helps at all.  Thanks.

-Ernie

---(end of broadcast)---
TIP 3: if posting/reading through Usenet, please send an appropriate
  subscribe-nomail command to [EMAIL PROTECTED] so that your
  message can get through to the mailing list cleanly

Re: [GENERAL] These Lists Are Being Cut To FOUR

[Sim Zacks]

I think this is disgusting.
I feel that if someone does not like what Mike is doing the appropriate way
of dealing with it would be to post a rational argument. Not to send an
obviously forged email with his name on the from line.


""Mike Cox"" <[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> Hello. My name is Mike Cox. I am in charge of the changing of these
> postgresql lists. I have decided that we are going to drop most of the
> lists from the vote. We will only be making 4 lists into real Usenet
> newsgroups if we win the election. The rest of the lists are crap and
> they take up too much fucking room. Marc was an asshole for stealing
> BIG8 newsgroup names without voting. He is an irresponsible piece of
> shit scumbag cock-licking anus-eating foreskin-sniffing motherfucking
> faggot. He has an ego the size of Dolly Parton's tits. Be sure to vote
> yes on all four of these lists or every list will be removed.
>
> Your support is most appreciated.
>
> Mike Cox
> [EMAIL PROTECTED]
>
> ---(end of broadcast)---
> TIP 5: Have you checked our extensive FAQ?
>
>http://www.postgresql.org/docs/faqs/FAQ.html
>



---(end of broadcast)---
TIP 8: explain analyze is your friend

Re: [GENERAL] I'm about to release the next postgresql RFD. Comments wanted.

[Arthur L. Rubin]

[EMAIL PROTECTED] wrote:

> This is my last post to news.groups about this.

Good.  With any luck, Marc will block you from the mailing
lists for spamming those lists.

---(end of broadcast)---
TIP 8: explain analyze is your friend

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[Kenneth Downs]

Joseph Daniel Zukiger wrote:

> [EMAIL PROTECTED] wrote in message
> news:<[EMAIL PROTECTED]>...
>> The lists are not bogus and your suggestion is not acceptable. Many of
>> the list subscribers have no connection to Usenet. They will be mailed
>> the ballots, regardless of what you or anyone else say. They are the
>> people who will be directly affected by this.
>> 
>> Mr. Pauli Vaurio
>> ...
> 
> Pauli --
> 
> "Bogus" is being used here in a technical sense. It is not a judgement
> of value.
> 
> The groups in question have not conformed to certain rules/customs. In
> that sense, their names are "bogus" -- not official, conforming use of
> the namespaces the usenet organization has reserved.
> 
> The fact that these groups apparently have the most right to the names
> in question is a big point in their favor, but the group maintainer
> did not go through the customary processes in reserving the names and
> requesting listing. So they are not official, thus they are "bogus".
> 
> It looks like I'm may have to finally subscribe through my isp and
> learn how to configure a newsreader in order to vote in favor, instead
> of posting through google all the time. That should speed my access to
> usenet up quite a bit. Oh, well.
> 
> JouDanZuki

Joe, your access to ng's will speed up *dramatically* when you go through a
news server.  The one Mike Cox recommended in his reply is very reliable.

-- 
Kenneth Downs
Use first initial plus last name at last name plus literal "fam.net" to
email me

---(end of broadcast)---
TIP 4: Don't 'kill -9' the postmaster

Re: [GENERAL] These Lists Are Being Cut To FOUR

[Wayne Brown]

Michael Glaesemann <[EMAIL PROTECTED]> wrote:
> 
> I believe these posts are spoofs. Mike has not posted anything along 
> these lines before, and the headers are different from his usual posts.

I'm sure you're right.  There's been some idiot trolling news.groups
for the past several weeks, impersonating other people and posting
this sort of garbage.  I didn't check the headers on the phony "Mike"
article, but it's probably the same creep impersonating him.  This jerk
has been trying to ruin discussions of several new groups, not just the
comp.databases.postgresql.general discussion.  There even have been forged
votes on some newsgroups and forged emails sent to news.groups regulars,
all before the comp.databases.postgresql.general discussion came along.
There apparently is someone (or maybe several people) trying to disrupt
the newsgroup creation process and I'm afraid the postgresql groups are
getting caught in the crossfire.

-- 
Wayne Brown  (HPCC #1104)  | "When your tail's in a crack, you improvise
[EMAIL PROTECTED]  |  if you're good enough.  Otherwise you give
   |  your pelt to the trapper."
"e^(i*pi) = -1"  -- Euler  |   -- John Myers Myers, "Silverlock"

---(end of broadcast)---
TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]

[GENERAL] ask a question

[suon]

pgsql-general
I am a new user for pgsql.I has a question is, does pgsql(7.5 for 
win32) has a jdbc or a odbc ? If have ,how can i take it ?
thank for your replying.

suon
[EMAIL PROTECTED]
　　2004-11-10

---(end of broadcast)---
TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[Rolf Østvik]

[EMAIL PROTECTED] (Jeff Eckermann) wrote in 
news:[EMAIL PROTECTED]:
> --- "Gary L. Burnore" <[EMAIL PROTECTED]> wrote:
>> Please.  We're trying to help you fix something
>> that's broken. 
> 
> ???  As a longstanding reader of the pgsql-
> mailinglists, (including via news.postgresql.org on
> occasion), all I see is some outsiders trying to help
> us "fix" a problem that does not exist.  And yes, I

I think it's a problem. Not a big one, but still a problem.

> have read most of the messages that have passed by in
> these threads.  After all that, I still don't see the
> benefit.
> 
> Perhaps that is why these conversations have been
> carried on almost totally by people who do not post to
> the pgsql lists.
>  

Or maybe that some of us longstanding (ok, only 4 years for my part) 
readers of pgslq-mailinglist like the proposal and just follow the 
discussion.

By the way, from the header of your post i think that the only audience for 
your post is those who are members of the mailinglist and not so many 
others.

Today everyone can access the news.postgresql.org server and the list is 
carried by some other news providers also. The benefit is that all 
potential new users would have better possibility of using their preffered 
news server to access the groups.
(Well, you may not think that beneficial but i do).

I have seen some discussion about the technicality of gatewaying and of the
filtering of spam which is on post from the news side to the mailinglist 
side. 
Otherwise haven't I seen any disadvantages with getting some of the groups 
"approved".

-- 
Rolf Østvik

---(end of broadcast)---
TIP 5: Have you checked our extensive FAQ?

   http://www.postgresql.org/docs/faqs/FAQ.html

Re: [GENERAL] I'm about to release the next postgresql RFD. Comments wanted.

[Rolf Østvik]

Mike Cox <[EMAIL PROTECTED]> wrote in news:[EMAIL PROTECTED]
berlin.de:

> Obviously, there cannot be 21 postgresql groups in the comp.* hierarchy. 

I agree

> I did a check on news.postgresql.org to see which newsgroups are the most
> popular and also the ones which cover the relevant postgresql topics.  
> Most of the postgresql topics and traffic are represented in these four
> newsgroups:
> 
> comp.databases.postgresql.admin
> comp.databases.postgresql.general
> comp.databases.postgresql.hackers
> comp.databases.postgresql.sql
> 

What about 
comp.databases.postgresql.novice

I think that group could be of general interest and it has a fair amount of 
traffic.

-- 
Rolf Østvik

---(end of broadcast)---
TIP 7: don't forget to increase your free space map settings

Re: [GENERAL] I'm about to release the next postgresql RFD. Comments wanted.

[Klaas]

[EMAIL PROTECTED] (Jon Bell) wrote:

> In article <[EMAIL PROTECTED]>,
> Klaas  <[EMAIL PROTECTED]> wrote:
> >In article <[EMAIL PROTECTED]>,
> > Mike Cox <[EMAIL PROTECTED]> wrote:
> >
> >> comp.databases.postgresql.general
> >
> >How about just comp.databases.postgresql or (shudder) 
> >comp.databases.postgresql.misc?
> 
> The whole point of this proposal is that c.d.p.general *already* exists in 
> an unofficial sense on some servers, and that it should be made 
> "official."  

While that is true, I think part of the point is also to create a 
postgresql big8 newsgroup.

> Creating a new group with a different name will simply 
> confuse matters on those servers that already carry c.d.p.general.

I don't think that is an insurmountable problem if it were actually 
desirable that the name be changed.  Most of the participants on on the 
list, and thus would be unaffected.  The rest use servers that have low 
or automatic standards for group creation, so it wouldn't be too 
difficult to point people to the new group on those servers.

Since people like the current name, though, it is clearly not worth the 
potential confusion .

-Mike

---(end of broadcast)---
TIP 7: don't forget to increase your free space map settings

Re: [GENERAL] I'm about to release the next postgresql RFD. Comments

[Woodchuck Bill]

[EMAIL PROTECTED] ("Marc G. Fournier") wrote in 
news:[EMAIL PROTECTED]:

> As a side note ... if/when the CFV is called and those 4 are 
> approved/rejected, that will not change what is available on 
> news.postgresql.org, it will only improve the propogation of those 4 
> specific groups so that more servers around the world carry them ...

[reply from list owner crossposted to news.groups]

---(end of broadcast)---
TIP 8: explain analyze is your friend

Re: [GENERAL] Value of serial data type after insert.

[Michael Fuhr]

On Wed, Nov 10, 2004 at 02:34:22PM -0500, Ernest Kim wrote:

> I was wondering if there was a way to get the value of serial data
> type after an insert.

Use currval():

http://www.postgresql.org/docs/7.4/static/functions-sequence.html

Race condtions aren't a problem:

"Notice that because this is returning a session-local value, it
gives a predictable answer even if other sessions are executing
nextval meanwhile."

-- 
Michael Fuhr
http://www.fuhr.org/~mfuhr/

---(end of broadcast)---
TIP 6: Have you searched our list archives?

   http://archives.postgresql.org

Re: [GENERAL] need simple strategy for universal extension table

[TJ Talluto]

Since nobody has mentioned any nuances about mapping these keys to system
tables in this particular database, I'll assume this is a good idea for
now.

-- 
TJ Talluto
torpedo51 at yahoo dot com

---(end of broadcast)---
TIP 4: Don't 'kill -9' the postmaster

Re: [GENERAL] I'm about to release the next postgresql RFD. Comments wanted.

[Russ Allbery]

In news.groups, Klaas <[EMAIL PROTECTED]> writes:
> Mike Cox <[EMAIL PROTECTED]> wrote:

>> comp.databases.postgresql.general

> How about just comp.databases.postgresql or (shudder) 
> comp.databases.postgresql.misc?

In this particular situation, I think .general may actually be a better
name than either of those (not to mention matching the name of the mailing
list, as I recall).  I know there's only one other .general group in the
Big Eight, but there are oodles of prior examples in other hierarchies.

-- 
Russ Allbery ([EMAIL PROTECTED]) 

---(end of broadcast)---
TIP 8: explain analyze is your friend

Re: [GENERAL] I'm about to release the next postgresql RFD. Comments wanted.

[Sim Zacks]

BTW, in Outlook Express if you are posting to the news.postgresql.org server
it will not send a message to news.groups. So this message will not get
there unless someone puts it there.

If 4 are official and 16 are unofficial, why would that bother you?
You can still access all of them the same way you do today. 4 of them would
be accessible by more people and if they posted on the wrong group they can
easily be let know which group is appropriate. For example, I posted a bug
on the PGAdmin list, and they told me that it was a plpython bug and I
should post it to the .bugs list.
As for Mike being an " unknown interloper", since when did an open source
forum become proprietary? He is probably a regular reader who is not a
regular poster. He belongs to the postgresql community. Some people
contribute by programming, other people contribute by documentation and he
chose to contribute by helping to propogate the newsgroups. I think that is
a worthy task and he should be thanked for doing this. Everyone always talks
about how much work the core team has to do and I think that any task that
someone else can contribute is a good thing.

What if he recommended a new feature for the database would you say that he
is an interloper and has no right to help determine the feature set? If you
disagree with the proposal then discuss it in a calm fashion without name
calling. If you think it should be better then help fix it. Don't stand on
the sidelines yelling "hey batter batter..."


<[EMAIL PROTECTED]> wrote in message
news:[EMAIL PROTECTED]
> The new proposal should have all of the lists. I am amazed that an
> unknown interloper is trying to dictate which lists should be voted on.
> I am affiliated with a network of over 60 PostgreSql users/developers,
> and I will e-mail each one of those people a ballot. If your next
> proposal does not contain every single one of the lists, you can expect
> essentially all of those 60 people to vote no. I will campaign
> passionately against your proposal if you leave out even one of the
> lists. Do the right thing if you are sincerely trying to help the list
> subscribers.
>
> This is my last post to news.groups about this.
>



---(end of broadcast)---
TIP 8: explain analyze is your friend

Re: [GENERAL] I'm about to release the next postgresql RFD. Comments wanted.

[Klaas]

In article <[EMAIL PROTECTED]>,
 Mike Cox <[EMAIL PROTECTED]> wrote:

> comp.databases.postgresql.general

How about just comp.databases.postgresql or (shudder) 
comp.databases.postgresql.misc?

-Mike

---(end of broadcast)---
TIP 3: if posting/reading through Usenet, please send an appropriate
  subscribe-nomail command to [EMAIL PROTECTED] so that your
  message can get through to the mailing list cleanly

Re: [GENERAL] Large database

[Michael Fuhr]

On Wed, Nov 10, 2004 at 04:10:43PM +0200, Alexander Antonakakis wrote:

> I will have to make sql queries in the form "select value from 
> product_actions where who='someone' and where='somewhere' and maybe make 
> also some calculations on these results. I allready have made some 
> indexes on these tables and a view that joins the two of them but I 
> would like to ask you people if someone is using such a big db and how 
> can I speed up things as much as it is possible on this

Can you give us an example of a query you'd like to speed up?  Please
post the EXPLAIN ANALYZE output for the query as well so we can see
what the planner is doing.

Have you tuned any settings in postgresql.conf?  The following page
has some tuning tips:

http://www.varlena.com/varlena/GeneralBits/Tidbits/perf.html

-- 
Michael Fuhr
http://www.fuhr.org/~mfuhr/

---(end of broadcast)---
TIP 8: explain analyze is your friend

[GENERAL] shrinking physical space used

[Bopolissimus Platypus Jr]

hello all,

i've got a database that takes up 4G of space.  when i run a script that
deletes all rows and then vacuum, the data directory gets down to 
around 3-3.5G.  what i'd like is to get a blank database structure that
really contains no data at all, or any unused space at all.  apparently
that's not what i'm getting now.  is there a way to get this apart from
dump, initdb, restore?  i figure that if i could dump/initdb/restore i'd
probably get a data directory with around 30MB only.

the reason i can't do dump, initdb, restore is:  I'm working with postgresql 
7.1.x (no choice, the client isn't going to be ready to upgrade to 7.4.x or 
better for at least six months or so since they have no regression tests at 
all).

the database i'm working on has problems with pg_dump and restoring a 
dump because of recursive definitions (function selects from a table,
table has a default or constraint referring to the function, neither can be
created when the dump is restored because the other isn't defined yet).  
i'm aware of check_function_bodies in 7.4.x and 8.x, but i can't take 
advantage of that since i can't upgrade. i tried to load the schema and 
data into 7.4.x and 8.x with check_function_bodies=off and then dump 
the data and reload it into 7.1.x but that doesn't work since there is some 
syntax in 7.4 and 8.x that doesn't work in 7.1 ($ instead of ' as 
procedure delimiters, schemas, etc). and anyway, there's still the 
recursive definitions (curious, does 8.x support for dumping the
constraints and defaults at the end of the dump as alter table instead?
how do i turn that on? i thought tom lane had a post that said something
like that was possible, but i haven't worked with 8.x beta long enough
to find where that is).  

at some point i may get desperate and i'll break down and just manually
modify the (very large) schema so that all such recursive definitions are 
removed from the create table statement and moved to alter table 
statements at the bottom of the schema.  i'm resisting that for now though.

thanks for any pointers :) (mainly to how to shrink the space used by
databases, sort of truncating the files so that no space is in there at
all that isn't actually used).

tiger

-- 
Gerald Timothy Quimpo http://bopolissimus.sni.ph
[EMAIL PROTECTED] [EMAIL PROTECTED] [EMAIL PROTECTED]
Public Key: "gpg --keyserver pgp.mit.edu --recv-keys 672F4C78"
 Mene sakhet ur-seveh

---(end of broadcast)---
TIP 2: you can get off all lists at once with the unregister command
(send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])

Re: [GENERAL] How many secondary databases can I create?

[Joshua D. Drake]

Dawid Kuroczko wrote:
On Fri, 12 Nov 2004 17:20:59 +0100, Michelle Konzack
<[EMAIL PROTECTED]> wrote:
 

Please note, that I do not run seperatly databases.
I like only to split it per $USER physicaly from the main database.
I like to have this setup for my new Virtual-Webserver where I have
for each VHost a local $USER.
   

Some time ago I talked with a friend working at some ISP, and we were
talking about PostgreSQL's superiority to hmm, other RDBMSs... ;)
Blah, blah, blah.  The question is - how to give many many users
access to PostgreSQL...   I see three approaches, and all of these
have pros and cons.
1) One pgsql per user -- in other words each user runs its own copy of
pgsql -- waste of memory (each user having its own shared mem, etc),
but can enforce quota limits, etc.  Hard to keep all those copies of
pgsql running.
 

Hello,
We use this version as it allows us to provide the best security
and flexibility to the customer
Sincerely,
Joshua D. Drake

--
Command Prompt, Inc., home of Mammoth PostgreSQL - S/ODBC and S/JDBC
Postgresql support, programming shared hosting and dedicated hosting.
+1-503-667-4564 - [EMAIL PROTECTED] - http://www.commandprompt.com
PostgreSQL Replicator -- production quality replication for PostgreSQL
begin:vcard
fn:Joshua Drake
n:Drake;Joshua
org:Command Prompt, Inc.
adr:;;PO Box 215 ;Cascade Locks;OR;97014;US
email;internet:[EMAIL PROTECTED]
title:Consultant
tel;work:503-667-4564
tel;fax:503-210-0334
x-mozilla-html:FALSE
url:http://www.commandprompt.com
version:2.1
end:vcard


---(end of broadcast)---
TIP 2: you can get off all lists at once with the unregister command
(send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])

Re: [GENERAL] I'm about to release the next postgresql RFD. Comments wanted.

[pvaurio]

The new proposal should have all of the lists. I am amazed that an
unknown interloper is trying to dictate which lists should be voted on.
I am affiliated with a network of over 60 PostgreSql users/developers,
and I will e-mail each one of those people a ballot. If your next
proposal does not contain every single one of the lists, you can expect
essentially all of those 60 people to vote no. I will campaign
passionately against your proposal if you leave out even one of the
lists. Do the right thing if you are sincerely trying to help the list
subscribers.

This is my last post to news.groups about this.


---(end of broadcast)---
TIP 1: subscribe and unsubscribe commands go to [EMAIL PROTECTED]

Re: [GENERAL] I spoke with Marc from the postgresql mailing

[Patrick B Kelly]

Seriously, the usenet people losing interest in this "problem" might 
be the best case scenario.
Ya know, that's kinda like when a business owner hopes the neigborhood 
will forget about a toxic waste spill.

In this case, it's not going to be forgotten as long as someone's 
shoving your emails up USENet's proverbial ass.  If you REALLY want it 
over with quick, shut off the mail-to-news.
Perhaps not the best option but that is an option. The mailing list and 
news.postgresql.org seem to work fine for many people who "matter".

Patrick B. Kelly
--
  http://patrickbkelly.org
---(end of broadcast)---
TIP 4: Don't 'kill -9' the postmaster

Re: [GENERAL] Important Info on comp.databases.postgresql.general

[GreyGeek]

Jeff Eckermann wrote:

> The "problem" that is being solved is, as presented so
> far, too abstract to be convincing, at least for me.

If you say so.  The problem that Mark (Marc?) is trying to solve is that his
ISP doesn't give access to this newsgroup because it is NOT in the Big-8.
The solution will probably not change anything for anyone currently enjoying
access and may be un-noticeable to most if his efforts are successful.

I haven't been able to detect anyone acting as a moderator, and with all the
address spoofing going on I don't know what they could do to eliminate the
spoofing and the childish rants that sometimes flair up.  The volume of
msgs to this group are so great that were it to become moderated and every
post previewed to eliminate bandwidth wasting posts and ego-fights, the
work load would be overwhelming for even two or three moderators. IMO

-- 
--
GreyGeek

---(end of broadcast)---
TIP 7: don't forget to increase your free space map settings

Re: [GENERAL] comp.database.postgresql.*

[Wayne Brown]

Dawid Kuroczko <[EMAIL PROTECTED]> wrote:
> 
> And now, some time ago, PostgreSQL mailing lists were gatewayed
> into Big8 namespace, to help people access them by other means.
> Only it was done "unoffically", i.e. not following standard procedure
> for such things.  It could be thought of as "hooking up" to neighbour's
> cable TV cables, only it does not involve stealing. ;)  Well, I'm sure
> it's easy to find many better comparable situations. ;)
> Anyway, it's a bit kludgy, not following standards.  It would be nice
> however to maybe go to the cable guys and tell them -- connect us
> officialy.  Not because Oracle users did so.  No, it's foolish reason.
> Because it's Right, because it makes us closer to Standards, because
> to be respected means to respect others.  Etc, etc. etc.

Another analogy would be to squatters building houses on public land.
Then, years later, one of them goes to the zoning commission and tries
to get the proper permits issued to give them legitimate use of the land
they're occupying...  but some of the residents want to stay unofficial
because they're afraid outsiders will want to move into the community
if it becomes "legal."

-- 
Wayne Brown  (HPCC #1104)  | "When your tail's in a crack, you improvise
[EMAIL PROTECTED]  |  if you're good enough.  Otherwise you give
   |  your pelt to the trapper."
"e^(i*pi) = -1"  -- Euler  |   -- John Myers Myers, "Silverlock"

---(end of broadcast)---
TIP 9: the planner will ignore your desire to choose an index scan if your
  joining column's datatypes do not match

Re: [GENERAL] I resign as the promoter of the PostgreSQL groups.

[Woodchuck Bill]

[EMAIL PROTECTED] (Kathy Morgan) wrote in
news:[EMAIL PROTECTED]: 

> Max <[EMAIL PROTECTED]> wrote:
> 
>> Yeah, I've seen these forged emails on the mailing list, plus a few
>> other colorfull ones. You should see that people on the list are not
>> completely against this idea, and they are not fooled by those forged
>> emails. 
>> 
>> PLEASE do not give up.
>> 
>> What you are facing is an identity issue with the postgresql users.
> 
> It's also quite likely, as Rebecca Ore said, that the forged emails
> really have nothing to do with the postgresql users.  news.groups and
> its denizens have been under attack for the past several months by a
> troll who, among other things, forges offensive posts to groups
> affected by current proposals.  I think he had taken a short vacation
> and is now back to again disrupt proceedings here by making life
> miserable for people discussing new groups.

Kathy and Rebecca are right. I just looked at the mailing list. The headers 
are the same as those of the usual forger, and at least one person has 
named one of the same suspects[1]. This is a disgrace. 

Mike, please reconsider your position. Your second RFD looks really good, 
and the 5 groups that you chose are right on the money. You are a good 
proponent. You've taken all of the feedback into account, good and bad, and 
implemented the appropriate changes. Don't let this kook thwart your 
efforts.

[1] Message-ID: <[EMAIL PROTECTED]>

http://tinyurl.com/624dy

-- 
Bill

---(end of broadcast)---
TIP 2: you can get off all lists at once with the unregister command
(send "unregister YourEmailAddressHere" to [EMAIL PROTECTED])