To: Joe Conway <mail@joeconway.com>
Cc: Josh Berkus <josh@agliodbs.com>, pgsql-advocacy@postgresql.org,
	pgsql-sql@postgresql.org
Subject: Re: [SQL] SQL Spec Compliance Questions 
In-reply-to: <406D9943.8070401@joeconway.com> 
References: <200403302050.12415.josh@agliodbs.com>
	<406D1031.7010603@joeconway.com>
	<200404020838.22698.josh@agliodbs.com>
	<406D9943.8070401@joeconway.com>
Comments: In-reply-to Joe Conway <mail@joeconway.com>
	message dated "Fri, 02 Apr 2004 08:48:03 -0800"
Date: Fri, 02 Apr 2004 12:47:56 -0500
Message-ID: <1129.1080928076@sss.pgh.pa.us>
From: Tom Lane <tgl@sss.pgh.pa.us>

Joe Conway <mail@joeconway.com> writes:
>>> 4.16.2 Referenceable tables, subtables, and supertables
>>> A table BT whose row type is derived from a structured type ST is
>>> called a typed table. Only a base table or a view can be a typed
>>> table. A typed table has columns corresponding, in name and
>>> declared type, to every attribute of ST and one other column REFC
>>> that is the self-referencing column of BT; let REFCN be the

> After re-reading it, I think it is related to (or at least similar to) 
> the work Tom is currently doing to allow composite types as table 
> attributes.

The "structured type" stuff seems closely related, but I do not
understand the business about a "self-referencing column".  I have a
feeling that it might be a mutant version of our notion of inheritance
...

			regards, tom lane