Information Technology Reference
In-Depth Information
The way computers store images can change to suit the needs
of users. Pictures stored in a small file save space but may
look fuzzy or blurred; those stored in an intermediatesized
file may display well on a computer monitor, but appear
grainy or fuzzy when printed. Pictures stored in a largesized
format may display nicely on paper, but may consume con
siderable space and require much time to move between ma
chines (e.g., download over the Internet).
As these examples suggest, computer applications often must
find an appropriate balance between speed and accuracy and be
tween size and usability when they store data. The basic questions
they face involve how much accuracy is really required, how quickly
results are needed, and how much flexibility is appropriate. In sim
ple applications, the best tradeoff solutions may be obvious, or any
choice may work well. In other cases, however, data management
decisions can have a significant impact on the application's effective
ness and the user's satisfaction. The important decisions often de
pend upon matters of data representation, so a basic understanding
of representation issues can have a direct effect on our applications.
Before I address your detailed questions about data storage, we
need a common understanding of some of the technical details in
volved in data representation.
The simplest and cheapest electrical circuits involve elementary
on/off switches. When you flick a light switch on or off, electricity
flows or it does not; a voltage is present or it is not. If we interpret
the presence of voltage as the digit 1 and the lack of voltage as the
digit 0, then having a light on can be represented as a 1, and having
it off, a 0. In mathematics, numbers formed out of only the digits 0
and 1 are called
binary numbers
. The reasonably natural relation
ship between circuitry and binary numbers has made binary num
bers an effective way to store data in computers.
To expand upon this idea somewhat, some early computers were
actually based on the decimal system (which is based on the 10 digits,
(0 through 9). In such a system, one might think of a 9volt charge as
corresponding to the digit 9, an 8volt charge to the digit 8, and so forth,
with a 0volt charge representing the digit 0. Although a decimal system
