Message-ID: <3BFBE06D.3010103@home.com>
Date: Wed, 21 Nov 2001 09:12:13 -0800
From: Joe Conway <joseph.conway@home.com>
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US;
	rv:0.9.6) Gecko/20011116
MIME-Version: 1.0
To: Ezra Epstein <akee@speakeasy.net>
CC: pgsql-docs@postgresql.org
Subject: Re: Oracle -> PostgreSQL ==>  RAW -> ???
References: <200111211532.fALFW4721515@spidey.speakeasy.net>
Content-Type: text/plain; charset=us-ascii; format=flowed
Content-Transfer-Encoding: 7bit

Ezra Epstein wrote:

 > Hi,
 >
 > I'm moving a database from Oracle to PostgreSQL.  I use RAW (as in
 > BINARY, as in BYTE(x) where x < 20) for quite a few columns that
 > hold application generated keys, etc.
 >
 > How do I represent a RAW in PostgreSQL.  I'm too much of a newbie to
 >  know how to create my own type.  BIT seems to be wrong.  And BYTEA
 >  can't be it either as there are all sorts of restrictions on using
 >  BYTEAs for keys. Seems OIDs are for BLOBs and this are not large
 > binary objects...
 >
 > HELP!!!  Answers, pointers, tips, doc refs, anything.

Most of the BYTEA restrictions are gone in 7.2 (about to go into 3rd 
beta release). Specifically, you can now create btree indexes on BYTEA 
columns. There's documentation too, in CVS and starting with beta3. Look 
for "Binary Strings" under the section on datatypes and functions.

Hope this helps,

Joe