Short term plans and expected results

• Optimization for XQuery In the short term, we plan to define a formal framework to study the optimization problem for XQuery extended with updates (XQuery-U). We plan to study the independence problem, that is the property that allows two XQuery-U expressions to be evaluated in any order with no effect on the result. This will be the basic property needed to extend standard optimization techniques to this new context.
• Spatial logics for SSD. TQL uses logic to describe queries, but the same logic can also be used to describe SSD types. A query is well-typed when it has the possibility to retrieve data, hence a query is well-typed when its conjunction with the input type is satisfiable. In the short term, we plan to extend TQL system with the ability to define types as sentences in a sublogic of its full logic, and with the ability to prove the satisfiability of the conjunction of a query with such sentences. The problem is undecidable in general, but can be tackled through heuristic techniques.
• A peer-to-peer distributed system for managing SSD. In the short term, we plan to finish the implementation of the XPeer p2p system, and to measure its performance, in order to demonstrate its ability to achieve high scalability.