Nowadays XML is universally accepted as the standard for structural data representation; XML databases, providing structural querying support, are thus becoming more and more popular. However, XML data changes over time and the task of providing efficient support to queries which also involve temporal aspects goes through the tricky task of time-slicing the input data. In this paper we take up the challenge of providing a native and efficient solution in constructing an XML query processor supporting temporal slicing, thus dealing with non-conventional application requirements while continuing to guarantee good performance in traditional scenarios. Our contributions include a novel temporal indexing scheme relying on relational approaches and a technology supporting the time-slice operator.