Abstract: The idea that source code for computer software be accessible to anyone has gained increasing popularity among software developers, fueling the rapid growth of the open source software (OSS) movement. Platforms for OSS development currently host incredibly valuable projects like the Linux kernel, TensorFlow and various blockchain software projects. We study the role that star software developers play in OSS innovation, treating innovative software as a kind of knowledge that is created and successfully commercialized. Drawing on the literature on scientific innovation and contrasting the knowledge domains of science and software, we hypothesize that stars play a crucial role in guiding the direction of innovation in OSS development, but need not actively participate in its commercialization. However, unlike scientists, whose professional activities like patenting behavior, research funding, and employers are well documented in public databases, there is no equivalent source of data on software developers. In this study, we aim to a) build such a database in the OSS context by joining together information about star developers, including their software output, locations, employers, educations and influence on other developers, and b) analyze it to test our hypotheses on the role of star developers in OSS innovation.