Message-ID: <3E01ADF1.5050003@klaster.net>
Date: Thu, 19 Dec 2002 12:30:57 +0100
From: Tomasz Myrta <jasiek@klaster.net>
User-Agent: Mozilla/5.0 (Windows; U; Win 9x 4.90; en-US;
	rv:1.2b) Gecko/20021016
MIME-Version: 1.0
To: Gary Stainburn <gary.stainburn@ringways.co.uk>
Cc: pgsql-sql@postgresql.org
Subject: Re: references table(multiple columns go here)
References: <200212181625.11316.gary.stainburn@ringways.co.uk>
	<200212190928.19736.gary.stainburn@ringways.co.uk>
	<3E019829.204@klaster.net>
	<200212191030.11661.gary.stainburn@ringways.co.uk>
In-Reply-To: <200212181625.11316.gary.stainburn@ringways.co.uk>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

Gary Stainburn wrote:

> On Thursday 19 Dec 2002 9:58 am, Tomasz Myrta wrote:
>
> >Gary Stainburn wrote:
> >
> >>That did the trick.  However, I now have another problem with the
> >>constraint
> >>complaining about there not being an index to refer to.  However,
> >>there is.

The error is inside declaration of table "ranks.
You can't create two similiar foreign keys: one based on field (rrank) 
and second one based on fields (rdid,rrank).
You have to change:
jrank		int4 not null references ranks(rrank),	-- needs sorting
to
jrank		int4 not null,	-- needs sorting


> This is probably because of my background in as a COBOL programmer where
> having multiple fields of the same name can cause problems (especially 
> with
> MF Cobol which only partially supports it) as well as early (read 
> early 80's)
> database experince where it wasn't allowed.   Also, I find it usefull 
> because
> I know immediately which table a field has come from.

As you wish.

>
>
> Why does it make joins easier to use the same name for fields?

If you create queries like this, you get rid of duplicates.
select *
from
jobtypes
join departments using (did)

If you are afraid of duplicates, you can always use an alias:
select
  ranks.rank_id as rid,
...

If you want, here is my minimal version of your tables:

create table depts (
  dept_id	int4 primary key,
  ...
};

create table ranks (
rank_id     	int4 default nextval('ranks_rid_seq') primary key,
dept_id		int4 references depts, -- department
rank		int4 not null,		-- departmental rank
rdesc		character varying(40)	-- Rank Description
);

create table jobtypes (
jobtype_id     	int4 default nextval('jobs_jid_seq') primary key,
rank_id		int4 references ranks(rank_id),
jdesc		character varying(40)	-- job description
);

Tomasz Myrta