Message-ID: <3DC62BB2.72FC66C8@technovell.com>
Date: Mon, 04 Nov 2002 09:11:30 +0100
From: Terry Yapt <pgsql@technovell.com>
MIME-Version: 1.0
To: pgsql-sql@postgresql.org
Subject: Comparing Numeric and Double Precision (float8)..
Content-Type: text/plain; charset=us-ascii
Content-Transfer-Encoding: 7bit

Hello all,

i DON'T know what is the proper forum to throw this question and I must
to insist in this "feature".  Sorry.

I have a lot of tables from Oracle 8i Databases with a lot of columns
with numeric(x,0) definition.

Ok.. I am traslating my oracle tables to PostgreSQL tables.  But I am
having a serious problem with my client aplications.

When I compare a numeric(x,0) field with a float8 field I have an error
on PostgreSQL what I didn't have with Oracle.  I mean:

CREATE test (one numeric(2,0));

SELECT * FROM test WHERE one = 1.00000; 

This runs fine on my Oracle Systems.. but I have problems with my
PostgreSQL system.  I have tried to create an operator to workaround
this inconvenience:

numeric '=' float8
     with CREATE OPERATOR command and calling to a function to return a
boolean.

Ok.. great.. It is running now.  But when it runs.. I have another
problems comparing numeric with integers and so on.  So I must to DROP
OPERATOR......

I don't understand what is the problem and what options I have to
workaround it (without re-write a lot of client applications).

I have a lot of code I don't want to modify.  The question is:

Why we cannot compare numeric with double precision ?  And why Oracle or
SQL can do it without problems ?

Thanks a lot.