|
Software testing is kind of difficult rather a tricky way to improve the quality of the
software, even in some cases Software testing become an infinite process as errors cannot
be completely removed. 1991, California; disaster happened by changing three lines of code
in signaling system which proves that fixing problem can create more complex issue even some
time failure of the system. Testing techniques it self can have flaws. For example if we closely study we will find not even single definition which can prove any relevance between the software quality and Code Coverage or Branch Coverage. Human testing which includes inspection, walkthroughs and reviews are alternatives to conventional testing methods. Advocate inspection is a kind of cheap way of doing the unit testing. Results confirmed that process of reading code stepwise abstraction is as less affective as on-line functional and structure testing in terms number of cost of faults observed. Official method of proving the accuracy of the software is the area which requires more study and research but still that method cannot beat the complexity barrier. It can be useful way for comparatively simple software's but not for the complex ones. Here a question arise, if testing is not improving software quality or reliability or even causing more trouble then why do we do this. This whole scenario is like old car manufacturing process, where we use to build cars first and then removes the defects and problems, which makes the defect free. |
| Sponsored Links: |
| << back |