Open source projects and ecosystems can be studied due to the public availability of their data. The main reasons for studying this data is to collect operationalizable metrics that can be used for the improvement of the project or ecosystem. We can for instance use these metrics to do prediction, study adoption rates, and perform scenario modeling. Presently, in literature, the reigning health factors that are acknowledged are Robustness, Productivity, Niche creation. It is also common to look at ecosystem health from two dimensions: the partner/network level versus the system/project level. Each dimension provides a unique perspective on open source health and enables improvement in a different manner: one focuses on the activity within the platform, whereas the other focuses on the activity outside of it. Typically, in open source ecosystem health research the metrics are characterized along several axes: they are evaluated for availability, collectability, generalizability, comparability, user friendliness, etc. Examples of metrics are interactions between developers, clones, branches, and numbers of commits. We also find that metrics that are typically easy to collect are not very meaningful. Also, the need arises for a meaningful compact subset of metrics, instead of throwing the kitchen sink at evaluation projects. Also, we suspect that “typical” developer behaviors can be extracted from the correlations between different metrics. Finally, we find that the goal-question-metric approach is insufficiently employed in the study of the health of ecosystems. One of the bigger challenges in assessing ecosystem health is the myriad of perspectives on ecosystems. For instance, we can look at network health versus economic health. Furthermore, ecosystems themselves are made up of ecosystems, and we need to establish beforehand what the best manner is of decomposing an ecosystem.