Information Technology Reference
In-Depth Information
In particular, the correction of an error in one part of a pro
gram sometimes creates a new error in another part of a large
program. Thus, it is sometimes better not to correct a known
error in a program, because the fix of that problem might
have a reasonable chance of creating a new and unanticipated
error elsewhere—and we have no way of knowing ahead of
time if that new error will be critical, major, or minor.
From another perspective, major new releases of operating sys
tems or application software often contain a large number of new
features and capabilities. Also, factors such as those described ear
lier often result in a relatively high number of errors in these major
new releases. The features may be nice, but at the cost of reliability
(errors). For example, the software may malfunction or crash often.
After encountering such problems for some time, experienced
users often wait to obtain major new releases for awhile—until a
few rounds of corrections and adjustments come out. The later, cor
rected versions still may have some errors, but such problems typi
cally are less significant than those found in the major new releases.
When a new version of a program debuts, therefore, take a critical
look at whether its new features are worth your investment in both
time and money. What advertisers claim as a significant advance
may or may not actually have the effect you want, and new software
may contain some errors you would prefer to avoid.
Greater Speed: As with our discussion of features, we need to
consider context while evaluating the significance of speed. Of
course, computers can be helpful only if they process data in a
timely manner. As individuals, we are annoyed if we must wait
awhile at a computer screen while processing occurs—particularly
if we cannot do anything else during that period. On the other
hand, we are unlikely to notice if processing takes milliseconds
rather than microseconds, or if one operation takes some time
while we are doing something else. For example, we may not
think about how long printing takes to finish if we have gone on
to other tasks, so lowering printing time from 15 seconds to 10
seconds may not matter. Certainly, we would like to get answers
quickly, but often our thinking and typing times are really the
slow steps. In such cases, faster processing may arise more from
our having a good breakfast than from getting faster computing
equipment!
