RE: Re: can't access table properties in 1_1_20

Hi Fred,

I've prefixed commens below with ##

--------------Replied message--------------
Date: Tue, 21 Aug 2001 17:30:46 -0500
From: fred(at)ontosys(dot)com
Sender: fred(at)ontosys(dot)com
To: dpage(at)vale-housing(dot)co(dot)uk
Reply-To: fred(at)ontosys(dot)com
Subject: Re: [pgadmin-support] can't access table properties in
1_1_20
Message-ID: <20010821173046(dot)A3343(at)ontosys(dot)com>
In-Reply-To: <200108212112(dot)f7LLClN08098(at)mailgate(dot)vale-
housing.co.uk>; from dpage(at)vale-housing(dot)co(dot)uk on Tue, Aug
21, 2001 at 10:13:00PM +0000

On Tue, Aug 21, 2001 at 10:13:00PM +0000, dpage(at)vale-
housing.co.uk wrote:
> - Does this happen with every table? What about pg_class
> for example?

It's happened on every table that I've tried so far -- maybe ten or
so
-- in several databases, but all on the same server (same
postmaster). The problem occurs on pg_class just like the
others.

> - Does the table have any checks, foreign keys etc. on it.

No.

> - If you switch on the definition pane, what sql is generated
> for the table?

I'll attach the SQL for pg_class, and the stuff that appeared in the
LogView window when I went to inspect pg_class.

After looking at the SQL a bit, I wonder if my pg_user data is
confusing pgAdmin. I've got one row with usename = 'Fred' and
another
with usename = 'fred' -- the values differ only in upper vs lower
case. I created that duplicate entry because sometimes I login to
NT
as "fred" and sometimes as "Fred". NT doesn't care about the
difference, but PostgreSQL does, so I created user entries to
cover
both cases. Might that confuse pgAdmin?

AHA! I just dropped one of the duplicate user names, and now I
don't
get the error message when inspecting properties of any table.
That
must be it.

## Hmm, I think I can guess what's causing that. I'll look into
it tomorrow.

> As for the pg_ objects, yes they are obsolete now. pgAdmin
> (original) has an option on the Advanced menu to remove
> them all a bit more easily.

Cool, I'll try that once I give up the old pgAdmin for good.

> I'd also like to hear any comments you may have about the
> new program...

It looks pretty good so far.

It seems slower than the old pgAdmin, but I haven't actually
timed
it. I suspect that the new version does more network traffic,
and I'm
limited by a 128K ISDN link.

## The pgSchema library that does the work builts it's object
model on demand. Once it's all built it should be very fast.
You can speed things up though by switching off the
definition pane which can cause low level parts of the model
to be built quite early in the session. Also, running in full
debug logging mode is *very* slow.

The inability to manipulate tables, because I'm blocked by
the problem
I reported, is a real show-stopper. I have been able to fill data
into tables though.

## I'll sort that problem for you - in the meantime you should
be ok with only 1 Fred!

Speaking of that, it would be handy to have a mode for entering a
bunch of data into many rows at once, much as one would do
with
Excel. Clicking the Add button and doing individual inserts
works,
but it feels slow. It's better than psql for my purposes though,
particularly for updating.

## I will be building new versions of the import/migration
wizards soon - the import wizard should help with bulk
loading from csv files etc. If you have any other thoughts
about the features you require, please let me know.

hth, regards Dave.

--
Fred Yankowski fred(at)OntoSys(dot)com tel:
+1.630.879.1312
Principal Consultant www.OntoSys.com fax:
+1.630.879.1370
OntoSys, Inc 38W242 Deerpath Rd, Batavia, IL 60510,
USA

--
Contents of Definition window:

- Table: pg_class
CREATE TABLE "pg_class" (
"relname" name,
"reltype" oid,
"relowner" int4,
"relam" oid,
"relfilenode" oid,
"relpages" int4,
"reltuples" int4,
"reltoastrelid" oid,
"reltoastidxid" oid,
"relhasindex" bool,
"relisshared" bool,
"relkind" char(1),
"relnatts" int2,
"relchecks" int2,
"reltriggers" int2,
"relukeys" int2,
"relfkeys" int2,
"relrefs" int2,
"relhaspkey" bool,
"relhasrules" bool,
"relhassubclass" bool,
"relacl" _aclitem,
CONSTRAINT "pg_class_oid_index" UNIQUE ("oid"),
CONSTRAINT "pg_class_relname_index" UNIQUE
("relname")
);

--
Contents of LogView window:

8/21/01 5:07:27 PM - SQL (batavians): SELECT a.oid,
a.attname, a.attnum, t.typname, CASE WHEN ((a.attlen = -1)
AND ((a.atttypmod)::int4 = (-1)::int4)) THEN (0)::int4 ELSE
CASE WHEN a.attlen = -1 THEN CASE WHEN ((t.typname =
'bpchar') OR (t.typname = 'char') OR (t.typname = 'varchar'))
THEN (a.atttypmod -4)::int4 ELSE (a.atttypmod)::int4 END
ELSE (a.attlen)::int4 END END AS length, a.attnotnull,
(SELECT adsrc FROM pg_attrdef d WHERE d.adrelid =
a.attrelid AND d.adnum = a.attnum) AS default, (SELECT
indisprimary FROM pg_index i, pg_class ic, pg_attribute ia
WHERE i.indrelid = a.attrelid AND i.indexrelid = ic.oid AND
ic.oid = ia.attrelid AND ia.attname = a.attname LIMIT 1) AS
primarykey FROM pg_attribute a, pg_type t WHERE
a.atttypid = t.oid AND attrelid = 1259 ORDER BY attnum
8/21/01 5:07:27 PM - SQL (batavians): SELECT i.oid,
i.relname, x.indisunique, x.indisprimary,
pg_get_indexdef(i.oid) AS definition FROM pg_index x,
pg_class i WHERE i.oid = x.indexrelid AND indrelid = 1259
ORDER BY relname
8/21/01 5:07:27 PM - SQL (batavians): SELECT attname FROM
pg_attribute WHERE attrelid = 17118 ORDER BY attnum
8/21/01 5:07:27 PM - SQL (batavians): SELECT attname FROM
pg_attribute WHERE attrelid = 17121 ORDER BY attnum
8/21/01 5:07:27 PM - SQL (batavians): SELECT oid, rcname,
rcsrc FROM pg_relcheck WHERE rcrelid = 1259 ORDER BY
rcname
8/21/01 5:07:27 PM - SQL (batavians): SELECT oid, tgrelid,
tgconstrname, tgnargs, tgargs, tgdeferrable, tginitdeferred FROM
pg_trigger WHERE tgisconstraint = TRUE AND tgtype = 21 AND
tgrelid = 1259 ORDER BY tgconstrname
8/21/01 5:07:28 PM - SQL (batavians): SELECT c.relname
FROM pg_class c, pg_inherits i WHERE c.oid = i.inhparent AND
i.inhrelid = 1259 ORDER BY inhseqno
8/21/01 5:08:05 PM - SQL (template1): SELECT * FROM
pg_user ORDER BY usename
8/21/01 5:08:05 PM - Error in pgAdmin II:frmTable.Initialise: 457 -
This key is already associated with an element of this collection