Why did your customer want a “scientific” method to validate requirement instead of any other (possibly non scientific method)?

can you share one potential defect type that is hypothesized  given  (a) requirement, domain, technology, cleanliness criteria?

I hope you are not making the /mistake/ of counting the requirements

Shrini

Ashok said – an entity like “Test” or “test case” can be decomposed of two parts -

1. Conditions that govern the behavior – notion of scenarios that come out
2. Stimuli to excite the system so that it displays some specific behavior – Inputs supplied to the system and their nature.

When we say countability – it is not I am somehow generating scenarios or test cases – but I model the system’s behavior and understand the conditions (binary or 3 values, n value) for specific behavior – hence I take number of conditions, number of values. Then I apply various techniques for generating permutations and combinations such decision tables, state machines etc to arrive at a set of meaningful  and provable scenarios.

For every scenario, as per this definition is -

There certain no of inputs – each of these take certain values and seek an optimum combination (some max and min) – this helps to comprehend the notion of countability objectively.

But the original question remains.

Are test cases countable?
Given an application, its behavior, conditions and inputs – it is possible arrive at A NUMBER of test cases that would be considered as OPTIMUM?

I know this is a tough question – but I think you folks have cracked it. Can you explain with an example?

Shrini