8 years ago, I made this diagram tongue-in-cheek to show my colleagues that it was possible to write academic papers with Pandoc Markdown.
Since then, there’s been a lot of improvements to the Pandoc Markdown workflow for writing papers. Pandoc went from 1.15.2 to 3.1.9, adding an extraordinary number of features. Of note, panflute was created as a better and more fun way to write Python Pandoc filters. And Pandoc embedded lua filters to allow writing pandoc filters directly in lua to prevent any marshalling of data.
I am continuously amazed and deeply impressed by the remarkable feat of software engineering that Pandoc represents, a sentiment echoed by everyone I know who has incorporated Pandoc into their workflow.
Collaborative academic writing often gravitates towards using Microsoft Word with OneDrive support. WYSIWYG is a powerful paradigm and provides immediate ease of use. Word’s robust commenting and review tools are second to none. Outside of Word, OverLeaf or ShareLaTeX seem to have the most mindshare. I reckon when people choose \(\LaTeX\) , they tend to do so because it is a safe bet. As a markup language, it is older than I am. That means there are resources online to address pretty much any problem you will come across1 , there are templates everywhere for all kinds of publications, and text editor support is high quality.
I still believe Pandoc is a phenomenal piece of software for writing academic papers (or any technical content for that matter). Being able to version control through git, and to produce outputs in diverse formats like PDF, Word, and HTML, adds immense value for me. But is it really going to be easier for everyone I’m collaborating with?