Testing context-aware software systems: unchain the context, set it free!

Background: In the era of digitalization, context awareness has become more important to allow software systems adaptation. Therefore, the quality assurance of such systems must consider the variation of context. However, there is a lack of software quality technologies doing so, which can increase the risk of failure of these systems.

Objective/Aim: To evidence recent advances regarding the testing of context-aware software systems (CASS), focusing on the abstraction that context should freely vary during test execution as it does in production environments,

Method: Based on knowledge acquired with quasi-Systematic Literature Reviews, we evaluate current testing approaches and discuss benefits and limitations of applying the observation above about CASS.

Results: It was not possible to observe any software testing technology supporting the unconstrained variation of context during testing a CASS. The practitioners, meanwhile, can use three main evidence-based strategies to test CASS: to assure functional correctness before turning to testing context-aware requirements; to design test cases to target context variables and to take advantage of automatic testing tools as much as possible.

Conclusions: To allow context to vary during testing freely, seems plausible. It provides a new frame of thought to enable the design of novel technologies to improve our capacity of testing CASS.