Cosma is a visualization tool for knowledge workers. It helps create, use and share non-linear scientific documentation. It works by reading plain text files interconnected with [[wiki links]] and rendering them as an interactive network of index cards inside a standalone HTML file.

This talk will present the design choices of Cosma, which uses existing ideas but in a peculiar way:

• it includes graph visualization but primarily for network synthesis rather than analysis;
• it creates contextualized backlinks based on wiki links but also based on citations (bibliographic references);
• it is a command-line program but most of its features are located within the files it exports.

These choices reflect the core idea of the project (which is as much an ambition as it is a set of constraints): leveraging innovation from the “tools for thought” space (easy hypertext authoring, interactive publications) while addressing the needs of creative, investigative knowledge work (owning robust data, backing claims with sources).

Project website, examples, documentation: cosma.arthurperret.fr

Project repository: github.com/graphlab-fr/cosma

Read the blog post for this presentation : www.arthurperret.fr/blog/2024-02-03-cosma-at-fosdem-24.html