In the context of Open Science, almost every ‘traditional’ research activity and output has been affected and transformed by means of web based technology. New forms of research output have emerged, among them software as an important means and method for data driven science. But how can software be treated as scholarly work, and how can it be integrated into a digital research infrastructure? The paper depicts software development related to Open Science and points out some future directions for software to become part of a sustainable research infrastructure.