Pushing complexity down the stack
Gregory Todd Williams, Kjetil Kjernsmo · 2014
Abstract. We believe that there is great potential for performance im-provements in making high-level query features visible to low-level data sources. In our work in the PerlRDF and SPARQLKit systems, we have tried to apply this thinking while maintaining flexibility and extensibil-ity. We discuss two major approaches we have taken – growing low-level APIs to handle more complex query operations, and allowing low-level frameworks to participate in query planning – and briefly examine some benefits and challenges to their use. 1