There's a big demand for new software at the moment. Almost everyone seems to have a smartphone these days, and aside from the integral software that makes android and iOS tick, there is a multitude of app-based software hitting the market. This is on top of the existing market for personal computers. But with demand for software comes an increased danger of products not coming up to scratch, and that means that software testing has to be more intense than ever.
Seeing the scale of errors in a piece of software when it turns up on the desk of the testers puts this necessity into perspective. There can be one mistake for every ten lines of code. A tenth of all code being flawed is an unassailable blot on the product's effectiveness, and it seems remarkable that it can get as far as testing to spot this, but that's just the way things are. It's not creative programmers' faults - they have to churn out new code as fast as they can in most instances - but if it goes unchecked it can lead to the product failing.
You don't have to look far to find instances of just this happening. Bugs that make the news tend to be fairly dramatic, such as a dive warning being initiated in an Airbus during a transatlantic flight. But it's the minor commercial mistakes that can ruin a company. The Galaxy Nexus bug on the ice cream sandwich device meant lots of retailers didn't stock it just when it was supposed to be taking a big slice of the Christmas smartphone market. Errors, however minor, look bad, and retailers don't want to be associated with them.
Software testing takes things through from the physical coding of the software, right up to user reception, so it also gives a good indication of how the market will respond to the device before it's put into production. This is another way that it gives companies an edge in the marketplace.
You can't cover absolutely every eventuality, but software testing can make the chances of a product working well once it's released much higher. It is as essential as taste-testing a new recipe before giving it to other people.
Seeing the scale of errors in a piece of software when it turns up on the desk of the testers puts this necessity into perspective. There can be one mistake for every ten lines of code. A tenth of all code being flawed is an unassailable blot on the product's effectiveness, and it seems remarkable that it can get as far as testing to spot this, but that's just the way things are. It's not creative programmers' faults - they have to churn out new code as fast as they can in most instances - but if it goes unchecked it can lead to the product failing.
You don't have to look far to find instances of just this happening. Bugs that make the news tend to be fairly dramatic, such as a dive warning being initiated in an Airbus during a transatlantic flight. But it's the minor commercial mistakes that can ruin a company. The Galaxy Nexus bug on the ice cream sandwich device meant lots of retailers didn't stock it just when it was supposed to be taking a big slice of the Christmas smartphone market. Errors, however minor, look bad, and retailers don't want to be associated with them.
Software testing takes things through from the physical coding of the software, right up to user reception, so it also gives a good indication of how the market will respond to the device before it's put into production. This is another way that it gives companies an edge in the marketplace.
You can't cover absolutely every eventuality, but software testing can make the chances of a product working well once it's released much higher. It is as essential as taste-testing a new recipe before giving it to other people.
0 comments
Post a Comment