Simple clear advice in plain English

The art of databases

Mark Whitehorn demystifies the storage and manipulation of data and shows how even the most disorganised among us can learn to love databases

The relational model
The relational model was the work of Edgar Codd (known as Ted), a British mathematician working for IBM. Dissatisfied with the state of data handling, he wrote several papers describing relational database theory and later set out his rules for how any relational database management system (RDBMS) should be built. These are known as Codd’s 12 Rules.

Just to keep us on our toes, there are 13 of them, numbered in best techno-geek style from zero to 12. (Slip Ted Codd and his rules into any conversation about databases and you win many brownie points). These rules remain the fundamental tenets to which any RDBMS must adhere.

Concepts
The theory underlying the relational model is mathematical ­ – set theory to be precise ­ – but that need not worry us because the software we use is at such a level of abstraction from the maths that we can effectively ignore it. But this rigorous mathematical basis stands us in good stead: it means that if we store data appropriately in a relational database we can guarantee that any correctly posed question will receive the right answer.

A bit of history
Adopting Codd’s ideas, IBM developed a prototype RDBMS called System R (R for relational). This matured into DB2, which was launched in 1982. However, the first RDBMS was released in 1979 by the Oracle Corporation: it was called, imaginatively, Oracle.

Of course, in 1981 the PC burst into our lives and client-server systems (where a database is stored centrally for access by many simultaneous users) also started to appear.

Since then three development threads have been running. On the mainframe various systems fought it out and IBM’s DB2 became the most commonly used. On the server, Oracle and Microsoft’s SQL Server are still slugging it out. On the PC platform a clutch of early DBMSs based on flat-file databases (see more below) appeared.

Ashton Tate’s dBase became top dog, with Borland’s Paradox second and others (R:Base, Dataflex, Dataease, for example) bringing up the rear. However, none of these made much of a stab at adhering to the relational model.

Article tags

Reader Comments

The art of databases - future articles?

I found the article informative and interesting. I have long wanted to learn how to construct a database, have you ever published a simple intro to building a database? Or is it your intention to publish one in the future?

Posted by John P Chopin, 23 Jan 2009

   

Add your comment

All fields must be completed. Your email address will not be displayed or used to send marketing messages.

All messages will be checked by moderators before appearing on the site.

See our Privacy Policy for more information.

Related articles

Troubleshoot and repair your laptop

Keeping your notebook in tip-top condition isn’t as difficult as it first seems, as Jason d’Allison discovers

Max out your Mac Mini

Got an out-of-date Mac Mini? Don’t let it go to waste ­ – turn it into a streaming audio server instead, says Chris Wiles

Slash your computing costs

There’s no reason your computer habit has to hit you in the wallet during the credit crunch. Joe Cassels points out some basic, but timely, cost-cutting measures

Question & Answer

Q.Why are some of the keys on my keyboard doing strange...

> Read the answer

Q.Is my phone’s Bluetooth any use?

> Read the answer

Q.Can I switch boot drives so that I can work on older...

> Read the answer

Best deals on the web

img

Samsung RV520-A07

£359.98- Buy it now

img

Acer Aspire 5750G (LX.RXP02.019)

£399.99- Buy it now

img

Apple MacBook Pro (MD313B/A)

£904.37- Buy it now

Latest issue & subscription deals

Poll

Are you concerned about viruses that target mobile phones?

Jargon Buster

Computing terms explained in plain English

Restore point

A Windows backup of system files and settings.

Great shopping deals from Computeractive