In-Reply-To: <124864.59370.qm@web31806.mail.mud.yahoo.com>
References: <124864.59370.qm@web31806.mail.mud.yahoo.com>
Mime-Version: 1.0 (Apple Message framework v752.2)
Content-Type: text/plain; charset=US-ASCII; delsp=yes; format=flowed
Message-Id: <513B4F8D-D6E6-4B41-BDF8-F1017FE74ECD@seespotcode.net>
Cc: SQL Postgresql List <pgsql-sql@postgresql.org>
Content-Transfer-Encoding: 7bit
From: Michael Glaesemann <grzm@seespotcode.net>
Subject: Re: Temporal Table Relations and Referential Integrity
Date: Mon, 14 May 2007 11:10:02 -0500
To: Richard Broersma Jr <rabroersma@yahoo.com>


On May 7, 2007, at 9:59 , Richard Broersma Jr wrote:

> However, since temporal UPDATEs and DELETEs do not behave in the  
> same way as they do in
> non-temporal tables, is anyone able to successfully use Referential  
> Integrity constraints between
> Parent and Child tables?  Or are custom triggers the only solution  
> to maintain temporal relations
> between Parents and Children?

Currently constraint triggers are the only way I know how to get this  
to work.

> My question deals specifically with the "Current" style of temporal
> relations, where only the currently known attributes and attribute  
> histories are stored.

It'd be helpful if you provided links to what you refer to as the  
Current style. The only reference for managing temporal data in ANSI  
SQL I'm aware of is Snodgrass' "Developing Time-Oriented Database  
Applications in SQL"[1]. Date, Darwen, and Lorentzos' "Temporal Data  
and the Relational Model"[2] is a more general text on the same  
topic, but isn't directly applicable to ANSI SQL.

> As a second question, would anyone know if temporal referential  
> integrity is targeted for future
> inclusion into the SQL standard?

I'm not privy to discussions of the SQL committees :) AIUI, temporal  
support was included in some of drafts for but it did not make it  
into the final standard. Snodgrass' book and Darwen & Date's paper  
"An Overview and Analysis of TSQL2"[3] have some discussion and  
critique of the draft.

> I would be happy to elaborate on what I mean by Temporal  
> Referential-Integrity (RFI) if my
> questions are unclear.

I'm not sure what *you* mean by Temporal Referential-Integrity, but I  
hope the links I've provided help a bit.


Michael Glaesemann
grzm seespotcode net

[1](http://www.cs.arizona.edu/~rts/tdbbook.pdf)
[2](http://www.amazon.com/gp/product/1558608559/)
[3](http://www.dcs.warwick.ac.uk/~hugh/TTM/OnTSQL2.pdf)