Goodbye Roam Research, Hello Obsidian

[ til productivity ] · 2 min read

I was bored over the weekend and migrated my notes from Roam Research to Obsidian. It was easier than expected and took under an hour. Here are the steps I took.

First, I selectively followed this guide to download my notes as a zip file (steps 1) and create an Obsidian vault. Then, I downloaded the images in my notes into /assets (step 11). A few notes were in folders because they had / in the title; I cleaned these up by hand.

As part of format conversion (step 4), I incorrectly enabled “Roam Research tag fixer”. As a result, my #tag and #[[tag]] were converted to [[tag]]. Fortunately, my tags always came after tags: and were on the first line of each file. Thus, I wrote some basic regex to fix it.

Next, I set up obsidian-git for syncing across devices. Installation on Mac and mobile was straightforward. The first sync took seconds and subsequent syncs were under a second.

Syncing images was trickier as git doesn’t do well with large commits. For my 1k+ existing images (mostly screenshots of talks and papers), I iteratively committed those that were less than 1,024kb in size. Images greater than 1,024kb were shrunk—by converting image format from png to jpg—before committed. Syncing the images to GitHub took about 15 minutes. New images will be synced via obsidian-git.

A few community plug-ins I use:

Others have also recommended their favorite plug-ins on this tweet.

I’m loving Obsidian so far. It feels snappier than web-based Roam during startup and while using it. It’s also more customizable. I don’t foresee it hindering my Zettelkasten workflow. While I have a forever-free Roam graph (I was an early adopter) and there’s no push factor, I’m going to stick with Obsidian for a bit and see how it goes.

Are you an Obsidian user as well? What features or plug-ins do you recommend?


Share on:

Browse related tags: [ til productivity ]

Join 4,300+ readers getting updates on data science, ML systems, & career.

    Welcome gift: A 5-day email course on How to be an Effective Data Scientist 🚀