The art of databases

Databases –­ they’re everywhere! Computerisation has made them much more efficient and has raised their profile to the point where almost everybody has heard of them. The high profile is partly because we now have a compendium term ­ – database ­ – for modern data storage.

We’ve always had databases but in the pre-IT world they came in a multitude of guises, as ledger books kept by businesses, card indices in libraries, filing cabinets of medical records and publications such as Who’s Who (for name droppers) and Wisden (for cricketers).

Databases are cool, clever and infinitely variable and a well-designed relational database is a joy to behold: it stores data efficiently, can be queried to retrieve any or all of its contents, and is user-friendly (we kid you not: a well-designed one really is user-friendly).

They can ensure a doctor has all the information about a patient to hand before a consultation, assist in managing your money, help find a book you’ve been looking for since 1974, book tickets to the other side of the world from your living room and even make tea. OK, perhaps not that last one, but you get the idea.

Why were databases invented?
In the beginning, pre-electronic computing was the province of major governments that wished to process data for all sorts of reasons, including counting votes. In 1888 Herman Hollerith invented just such a machine: he gave his name to Hollerith cards, the punched cards used by early computers, and his company, after a merger, became IBM.

After IBM launched the first mainframe computer in 1952, people started to get to grips with computers and use them seriously in the commercial world. But at that time nobody had put much thought into how we organise, store and manipulate data in an electronic database and the first attempts were pretty horrible. During the 1960s our understanding of the problems and solutions increased rapidly and the concept of a ‘database model’ evolved.

A database model is a description of a strategy for handling data: the network and hierarchical models arose during this period. In the 1970s the relational model emerged (read more later), and it quickly gained popularity and has yet to be seriously challenged: it is by far the most commonly used today.