Thursday 24 July 2014

Always In Vogue- Model Based Testing



The business of software development is complicated and challenging to say the least! The biggest challenge faced by entrepreneurs owning ISVs, like me, irrespective of whether the business is big or small, is the “time to market” factor. Intrinsic to the development process itself is the process of testing which is vital to ensure quality of the end product and consumer satisfaction. One factor common to all forms of testing is being aware, with complete clarity, of what needs to be accomplished by the end product. Model based testing or MBT is also no different when it comes to this factor. First a mental model is formed in the minds of the testers before further steps can be taken and the actual testing process can be executed. Simplistic though this sounds, it is not a trivial process at all because modern systems have complex interfaces and even more complicated functionalities. Moreover the software is deployed within vast operating systems and among a myriad of other applications, so the number of factors affecting the outcome of the end product can be mind boggling. Now imagine forming a mental working model with all these factors included!

Model basedtesting was not only used effectively in the past but is also currently in use in the present and will not go out of fashion in the future simply because of its versatile and all encompassing nature. The entire system and what the specific software being developed will do within the system is viewed as a single model or a series of smaller models and the testing procedure is largely based on these models. Thus MBT is both basic in nature and complicated too and it lends itself with great versatility to any scenario. There are no software models which function as a “one size fits all” and every situation demands a unique model. This is why testers following MBT should be thoroughly knowledgeable about the systems and the functionality of the end product and also the end users. This will enable them to choose the right models for executing the testing process efficiently and getting the most accurate results.

No comments:

Post a Comment