Re: The Theoretical Foundations of the Relational Model
Date: 25 Jun 2002 11:51:29 +0200
Message-ID: <3d183d21$1_at_news.uia.ac.be>
In article <3d17a990.31505041_at_news.verizon.net>, JRStern <JXSternChangeX2R_at_gte.net> wrote:
>
> Come on, you know that. There just seems to be something about the very
> suggestion of adding order as a property of views and tables, that causes
> reflex reactions of fear and loathing in otherwise sane and sober
> relational database experts.
There seems to be a slight misunderstanding here about what it is exactly that we are arguing about. You seem to think I claim that the RM is under all circumstances the best data model for all users and all applications. However, I don't claim that (although it comes close) and I don't think that this is the claim that the relational model makes.
To understand that you have to keep in mind that the above claim is about the *external* schemas as defined in the ANSI/SPARC model. It is not the data model for the external schemas that makes a database a relational database; a relational database with an OO wrapper is still a relational database. What makes a database a relational database is the use of the RM for its *conceptual* schema, and for these the RM claims that it is indeed the best data model.
The fact that the RM is based on logic and is therefore a well-understood, simple, powerful and universal way to represent knowledge is only one part of the support for this claim. The other part comes from the fact that it allows you to define a unifying view over all the data in the external schemas that is not biased towards certain external schemas and also not biased towards certain internal schemas. It is this that becomes threatened when you start introducing ordered tables and this is why non-logical issues also become important for choosing the best logical data model.
- Jan Hidders