JuliaCon 2025

We’re excited to share the story behind the new native Julia engine for the technical publishing system Quarto, developed by PumasAI in collaboration with the Quarto project. Released right after last year's JuliaCon, this engine, based on the QuartoNotebookRunner.jl package, makes it easier than ever to author, render, and publish computational documents with Julia.

Previously, Julia’s integration with Quarto relied on the IJulia package, requiring Python, Jupyter, and a carefully configured Julia environment. This setup often caused frustration due to its complexity and dependency management. The new native Julia engine eliminates these issues. With just Julia and Quarto installed, users can now start creating and rendering documents without additional setup. On first use, Quarto manages the installation of the required Julia packages automatically, providing a frictionless experience.

This engine also opens up exciting new possibilities for Julia users. The new expandable cell mechanism, a unique feature of our engine, lets users programmatically generate sections of their document which previously required hardcoding markdown code, unlocking more modular and easily reproducible workflows. Native support for R code cells, powered by RCall, allows effortless integration of Julia and R in a single document. (A proof of concept for a similar mechanism for Python using PythonCall also exists.)

In this talk, I will explain the design and architecture of QuartoNotebookRunner.jl, discuss how it interfaces with quarto-cli, and show how it can be extended with custom Julia packages. I’ll also highlight opportunities for further feature development now that Julia has its own native engine, bringing it closer to parity with R’s rich feature set in Quarto.

Whether you’re a scientist, data analyst, or developer, this session will provide a comprehensive overview of how the native Julia engine simplifies workflows and expands the possibilities for document creation with Julia and Quarto.