I think the point Quim was trying to make is that our work in Maemo Software is a R&D function. That is 50% R and 50% D. A lot of our work is research at one stage of a project and development at a a later stage when we built a piece of software for a device. Especially, our work in open source is partially separate from the work to built devices. You will see increasingly more and more contributions to upstream projects not only for Fremantle but also Harmattan. To make conclusions from open source contributions about individual products becomes increasingly difficult. Work in open source should not by linked too close to products as long one competes with closed operating systems.