Building myself a liability

Rebeca Sarai · May 10, 2021 · (Updated: May 29, 2022)




I have been tracking different aspects of my life for almost two years now. It started after the purchase of a smartwatch, after a few months, the tracking started to leak to other aspects of my life besides the usual heart rate, sleep, exercise, and stress provided by the smartwatch. Progressively, I changed my note-taking tools to be more active and free of paywalls, used other tools with seamless exports to track moods and habits, and even started to use a time tracking app while working on personal projects. That’s how I can tell you that last year I spent more than 164 hours practicing the piano. I can also say that May was the month that I most practiced and October was the month that I least practiced.

Inspired by the work of @karlicoss, I found myself thinking about ways to track more aspects of my life and how to make this data more useful. The interest was there, I was just not yet sold on the utility of it. After a year of tracking my productive time, the insight I have is that it does not matter to have the data if you can not draw connections from it. I need to be able to query the data, I need replicable queries for the other types of data, and more importantly, I need to be able to see the big picture. It is already a lot of work to be able to gather your data in a single place (away from the mega-companies) and here I include the processing and the formatting to a friendly language-specific data type; besides the extra layer of processing relationships manually from different types and formats was a burden that I was not willing to take.

Nevertheless, an unknown connection came to the surface, something already mentioned by @karlicoss but that I missed completely. The usability comes with a timeline view called lifelogging, a way to see all your personal data in one place, search, filter, and see the context of events. The most exciting and advanced tool is not an open-source tool, developed by Andrew Louis, and it’s called Memex, highly recommend watching his talk and reading his blog posts.

Leaving the utility issue behind me, I decided to put myself through the exercise of building such an interface for myself. The intention is to build a local timeline with no connection with the cloud to search and analyze all of my personal data. A few projects already try to accomplish what I want here but first, they don’t quite fit my workflow, second what would be the fun of getting something ready? This post starts this process, excited to see how long this project will endure.

A detailed list of similar projects can be found here and below are the ones I stumble with:






Did I make a mistake? Please consider sending a pull request.