Pulling the plug on Roamflow

TLDR: I will stop working on Roamflow and focus my attention on Markway.

If you follow me on Twitter, you know I'm fond of Roam Research. And for the past 9 months I have worked on a new project called Roamflow. With this post, I want to announce I will stop working on Roamflow.

We (Leo and I) started working on the project to enable non-tech to customize Roam Research in a way that best fits their workflow. We (Leo and I) started working on the project to enable non-tech to customize Roam Research in a way that best fits their workflow. Because until today, this is only possible by loading custom CSS and Javascript directly in Roam Research.

After some conversations with members of the community, we were highly motivated to make Roamflow a reality. We planned to launch a small beta before Christmas, to gather early feedback from users. But as we both work full-time, we couldn't achieve what we initially set out to do. It was just too much work.

Looking back we should have scoped the project better. In the beginning, we planned to launch a theme editor, plugin library, quick settings and custom features (like a dark mode). At some point, we realised that this was too much. So we focused our attention on the theme editor.

Here we actually made great progress. Not only regarding the development, but we also created a dynamic design system for Roam Research. We thought long about how we could balance usability and customizability. We ended up by writing a dynamic CSS theme, where all the color values for UI elements were mapped to ten shades of a primary, secondary and neutral color (So 30 colors in total). The user could pick those three colors and the app was basically doing the rest. As a result, every theme was beautifult by default. No matter what the user chose.

Besides the colors, the user could pick a custom border radius and freely pick any Google Font. With those settings, it was possible to quickly create multiple themes for different scenarios (e.g writing sessions, creative work)

But the more hours we put into the project, the more it was growing in size. This project is a perfect example of "The devil is in the details". The work on the dynamic CSS theme was especially tedious. Especially, because Roam Research was changing quickly and the existing code was chaotic.

The project dragged on, without us moving any closer to the beta launch. At the same time, I was completely neglecting my web highlighter extension with more than 6000 active users. I didn't touch any code in months. And on top of that feedback from users was piling up.

At one point, I had 400 feedback emails in my inbox and wasn't replying to anyone. This felt bad, especially because some users were really passionate about the extension. I never planned that Roamflow would take up so much of my attention.

So after many months of working on the project, we will pull the plug on Roamflow. Even if we launch it, it wouldn't be sustainable because the time investment is just too big.

I thought long about that decision, but believe it's also the right one for the community. By stopping with Roamflow, I can focus again on the Highlighter. Bringing new features, talking with users and making the app more accessible for other apps like Obsidian, Logseq and Notion.

But the first thing I will do is to rebrand from Roam Highlighter to Markway and to bring the extension to Firefox and Safari soon. If you want to follow along, I've put up a small landing page here.

PS: If you have ideas for the future of Roamflow, let me know. The best way to do so is to write me directly on Twitter