Since its popularisation by Wing as a general skill set developed by those studying computing, computational thinking has been incorporated in to curriculum worldwide. The idea has also generated a lot of argument, most notably over what it actually is or should be. We review the varying definitions and views, drawing out the agreement. We argue that despite the differences, even between contrasting views, there is a solid core that is agreed on. This still provides a firm foundation for gaining practical benefits from its introduction and from actively teaching it from an early age. We review the research to date based on the varying definitions offered. The differing views have different implications for practice which we also outline. The concept itself is relatively new, however, and given its wide scope, solid research to date is relatively limited. We therefore outline a range of open questions that need answering. Most notably, the question of which definition is the most appropriate to take depends on the answers to open questions. We need more research to move the debate away from personal opinions to being founded on a solid evidence-base.