As firms rail against "buggy" code, testing is growth area
What is it?
Testing used to be the bit that got left out when software projects
overran. The search for 2000 date-change problems turned testing
into big business, and it received a further boost when a number of
large financial institutions suffered the embarrassment of having
their online account services crash because they had not been
load-tested for scalability.
Software testing is now in the news again, as the implications of
Microsoft's decision to release monthly batches of patches sinks
in. A change to any part of any system can have an unforeseen
impact on any other part. This is multiplied when, as happened with
Microsoft last month, a service pack includes 14 different patches.
The only way of ensuring there are no unintended consequences is
thorough regression testing, which essentially means putting the
whole system through all its paces before putting it back
online.
Where did it originate?
Testing as an independent discipline, with packaged tools and
external verification and validation services, began in the
mid-1960s. Automated tools, a humane alternative to the slog of
working through printouts to find and fix bugs, became widely
available in the early 1990s, but as Y2K revealed, a lot of
companies did not invest.
What is it for?
Functional testing makes sure the software does what it is supposed
to, and reliability testing makes sure the software does not fall
over as it is doing it.
The British Standards Institute defines testing as "the process of
exercising software to verify that it satisfies specified
requirements and to detect errors".
What makes it special?
Ovum, which publishes reviews of all the major testing suites,
said, "The cost of fixing an error increases exponentially as
development proceeds and the error becomes more ingrained in
subsequent work. If an error is present in a delivered application,
the consequential loss amounts to many times the cost of actually
repairing the software."
There has been a growing backlash from groups such as The Corporate
IT Forum against buggy software. Many companies have deferred their
software purchases for a year or two until new releases have had
the bugs thrashed out of them. As software suppliers are losing
revenue, more thorough testing could restore customer trust.
Where is it used?
Testing should be built into the development cycle from the outset
and continued for as long as systems are operational.
How difficult is it to master?
According to risk management company Vizuri, 70% of software
testers are not worthy of the title "professional" because they are
not only non-accredited, but also under-qualified and lacking in
industry experience.
What systems does it run on?
Test suites need to be able to handle all the different
environments your application needs to interact with.
Not many people know that . . .
About 50% of testers are women, and that number is expected to grow
to 75% by 2006.
What is coming up?
Automated testing tools for web services.
Training
You could learn a single testing suite, but in recent years there
have been a lot of mergers and acquisitions among testing tool
suppliers and some tools have been discontinued as product sets are
consolidated.
You are probably better advised to take a generic course from the
BCS, the University of Sheffield or other supplier-independent
sources.
www1.bcs.org.ukwww.shef.ac.ukRates of pay
For junior testers, rates can be as low as £15,000, although
£23,000 to £26,000 is more typical. Senior testers and team leaders
can get £35,000 to £43,000.