Toward a woven literature: Open-source infrastructure for networked scientific publishing

Scientific publishing must evolve to reflect the computational nature of modern research. While code, data, and analysis increasingly shape scientific findings, current publishing formats often flatten these elements into static narratives. In response, a new model is emerging: the woven literature. Inspired by principles of literate programming, this approach integrates code, data, narrative, and interactive environments to enable reproducibility and transparency within the article itself. This review explores the technical foundations, cultural barriers, and infrastructural innovations shaping this transition. Drawing on case studies from the NeuroLibre platform, it illustrates how open-source tools can support next-generation publications that are executable, sustainable, and adaptable across disciplines. The discussion highlights how workflows of varying complexity can be modularized and surfaced through reproducibility-focused platforms, while also addressing limitations imposed by current incentive structures and legacy systems. By embedding computation directly within the scholarly article, woven literature transforms the role of publishing from passive documentation into an active and verifiable part of the research process, allowing readers to engage with and extend the work itself.