1. Identify the inputs. What specific input events are you going to consider within a given model program? Assign a name to each event and describe it. For example, name specific input messages belonging to specific input domains, a fixed sequence of inputs. It is impossible to consider all possible inputs in the model, so do not use too many input events.
2. Set the input encoding. You have chosen the input encoding that will be used in your model. It can correspond to the same encoding used by programmers, and may not correspond. Do not include in the model such functions that are supported by the software, but which you are not going to test. For example, if the operating system itself converts the cursor positions and mouse clicks into a menu item selection, then you do not need to include this in your model, consider that your input comes to you already encoded (for example, the operating system). You should check the input encoding only if it is part of the program implementation.
3. Identify the state. States are often created as a product of multipliers. If so, identify the multipliers and take into account that each combination of multipliers will correspond to its non recurring state and that the product grows very quickly and you may need to use nested models. Make a list of identifiable states and name them using a system. The resulting states should be counted in tens, but not thousands.
4. Set the encoding of states. If the developers implement a software project using a design pattern like machine with a finite number of states then the program can contain the procedure for encoding the current state and the state counter. If this is the case, then the state encoding procedure must be tested, but this can be very difficult in the context of pure behavioral testing. If the program was designed without using a finite state machine, state encoding is one of the tasks of your model and may not be contained as part of the program. However, you should check the correctness of the encoding in your model.
5. Identify the output events. Only the simplest models have single output events, for example, corresponding to a single character output. Much more often, output events consist of a sequence of actions. Identify these sequences. In nested models, the output event is the activation of the next lower level model in the model hierarchy. Assign a name to each output event, regardless of whether the event is a single event or a sequence of actions.
Website testing service is widely used to check how good a certain e-commerce resource is and what can be done to improve it. If a website makes bad first impression, its owner may begin losing potential sales and the business will be hurt. It depends on you if want to win website visitors or lose them, so contact us if the quality matters much to you.
Comments are closed.
