Path: utzoo!attcan!uunet!husc6!mailrus!ames!lll-lcc!unisoft!mtxinu!sybase!jeffl
From: jeffl@sybase.sybase.com (Jeff Lichtman)
Newsgroups: comp.databases
Subject: Re: Referential Integrity
Message-ID: <2481@sybase.sybase.com>
Date: 24 Dec 88 20:03:45 GMT
References: <560@bloom.UUCP>
Organization: Sybase, Inc., Emeryville, CA
Lines: 21

> Why don't any of the RDBMS packages we all know and love REALLY support
> referential integrity??

Sybase supports referential integrity, although the syntax is not anything
like what you showed.  With Sybase, you can put a trigger on a table which
will be executed whenever the table is updated.  Within triggers, one can
tell which columns have been updated and select the newly inserted and
deleted rows.  Here is an example of a cascading delete trigger:

	create trigger student_delete on student for delete as
		delete enrollment from enrollment, deleted
			where enrollment.student_id = deleted.student_id

As I said, the syntax isn't the same.  There main advantage to this approach
is that the integrity conditions and actions can be arbitrarily complex.
That is, you're not limited to simple existence checks or simple cascade and
block actions.
-- 
Jeff Lichtman at Sybase
{mtxinu,pacbell}!sybase!jeffl
"Saints should always be judged guilty until they are proved innocent..."