Open knowledge has diverse forms — including open-source software, scientific research, and collaborative data commons — that are all built upon intricate webs of dependencies. As vividly demonstrated in cascading OSS failures caused by overlooked maintenance of dependencies (e.g., the npm left-pad incident), sustainable open knowledge production requires ongoing care and investment in its foundational contributions. This talk explores "knowledge dependency graphs" as a conceptual tool, generalizing from software dependency graphs, to better understand public goods coordination problems across diverse open knowledge domains. By examining how decentralized dependency funding protocols have met these challenges within software ecosystems, we explore their potential, and limitations, as part of an iterative approach toward sustainably supporting critical dependencies throughout the broader landscape of open knowledge.