If a requirement cannot be fulfilled with an existing piece of software, one approach is to initiate a custom development. However, several reasons may lead to the decision of extending an existing open source: Existing open source software may already cover some of the required functionality: It can be cheaper to start from there instead of building everything “from scratch”. It may be that there is enough funding for an initial development project, but not for maintaining the resulting product during the coming years. Universities usually prefer open source to closed source because this minimizes their risk of a dependency (e.g. on a supplier or their NREN), and they often have a preference for open source because they know this model from own projects and open research. It may be advantageous to combine the efforts of several parties – even internationally – without having to set up huge, risky contracts and projects. This session will look at the challenges around all of these approaches.