Abstract
In this paper we study the idea of XML benchmarking from various points of view. We describe classical approaches to testing efficiency and versatility of XML processing methods, we provide an overview of the best-known representatives of particular types of methods and, in particular, we focus on their current limitations and possible improvements.