The Necessity of Software Testing
Unless you're a software tester yourself, it can sometimes be hard to see the value inherent in software testing. Programmers have a certain level of complacency towards the quality of their code, as they know that software can run through various pathways even if there's an error in multiple aspects of it. Business people can look at the time and resources that testing eats up and dismiss it as a needless expense. But the real picture of the necessity of testing is very different.
When you break testing down, and de-mystify it a little, it's easy to see why it's important. A piece of software is made up of coding, which is a language a computer reads to tell it to perform particular tasks. When there's a problem with the code, the computer gets confused. Testers look into the code by doing white box and black box testing. Black box means making the piece of software perform actions for them, and seeing how well it does it. White box means actually looking at the code and tracking where errors have been written.
All of this testing gives you a picture of the flaws in the system. The amount of flaws varies with each product, but it's commonplace for testers to be handed a 'finished' system that has ten errors for every hundred lines of code. Once the problems have been discovered they can be corrected, and that means that you're much less likely to get users complaining about the performance of your software.
Ignoring this, and hoping for the best, is a complete lottery. Once you've released a product with a glitch users will spread the word quickly, and whether you take the product back in or improve it with patches, the damage is already done. It's not a very modern approach to patch things later, and users are much less likely to forgive than they were in the 90s and 00s.
It saves an awful lot of trouble to get things right first time these days. Testing is something that extends the release date of a product, and costs money, but it's better to see it as an essential part of the process of making software than some kind of extra feature, otherwise your products will consistently fall below the standard.
When you break testing down, and de-mystify it a little, it's easy to see why it's important. A piece of software is made up of coding, which is a language a computer reads to tell it to perform particular tasks. When there's a problem with the code, the computer gets confused. Testers look into the code by doing white box and black box testing. Black box means making the piece of software perform actions for them, and seeing how well it does it. White box means actually looking at the code and tracking where errors have been written.
All of this testing gives you a picture of the flaws in the system. The amount of flaws varies with each product, but it's commonplace for testers to be handed a 'finished' system that has ten errors for every hundred lines of code. Once the problems have been discovered they can be corrected, and that means that you're much less likely to get users complaining about the performance of your software.
Ignoring this, and hoping for the best, is a complete lottery. Once you've released a product with a glitch users will spread the word quickly, and whether you take the product back in or improve it with patches, the damage is already done. It's not a very modern approach to patch things later, and users are much less likely to forgive than they were in the 90s and 00s.
It saves an awful lot of trouble to get things right first time these days. Testing is something that extends the release date of a product, and costs money, but it's better to see it as an essential part of the process of making software than some kind of extra feature, otherwise your products will consistently fall below the standard.
You are receiving this because you signed up for it on 2011-10-03 from IP 203.109.66.181
To fine-tune your selection of which articles to receive, just login here
using your username:
To unsubscribe please use the following link:
Unsubscribe
No comments:
Post a Comment