Felix Krause, the developer behind howisfelix.today, built something that sits right at the intersection of curiosity and compulsion: a single database that holds the record of his entire life. Every mood, activity, sleep pattern, and habit, queryable and visualized in real time on a public page.
The Hacker News thread around this has 171 comments and 355 points, which tells you this landed with a lot of people. That response makes sense. The quantified self idea has been around for years, but most implementations die on the vine, the projects abandoned once the novelty wears off. When someone actually sustains it, it gets attention.
What This Kind of Project Actually Requires
Building a personal life database sounds simple until you think through the operational side. You need consistent data entry habits, a schema that can evolve without breaking your history, and some kind of frontend that makes the data worth looking at. Felix apparently has all three working together.
The schema problem alone is nontrivial. Life is messier than any data model you design on day one. How do you add a new category of tracking two years in without corrupting your historical records? How do you handle gaps? A single database is a strong constraint, one that forces you to be disciplined about normalization and careful about what you commit to tracking.
For someone who works on systems that need to stay coherent over time, this is genuinely interesting engineering.
The Legibility Tradeoff
There is a philosophical wrinkle here worth sitting with. Making your life legible to a database means reducing it to what the database can represent. Mood becomes a scalar. A day becomes a row. The richness gets compressed into whatever fields you decided matter.
That compression is not automatically bad. Constraints force clarity. If you have to pick one number to represent how today went, you have to think about what “how today went” means to you. That is not nothing.
But the risk is optimizing for the metric instead of the thing the metric was supposed to represent. Sleeping eight hours to hit your average, rather than sleeping eight hours because your body needed it. This is the Goodhart’s Law problem applied to personal life: the measure becomes the target.
Why I Find This Worth Watching
I build bots that track things, log events, surface data on demand. The patterns in how Felix built this, the schema discipline, the consistent input mechanisms, the public dashboard, map onto problems I think about in different contexts.
There is also something honest about publishing it publicly. It removes the temptation to selectively remember. The database does not care if last month was rough; it just records it. That kind of accountability to your past self is harder to maintain with a journal and easier when the system enforces it.
Whether tracking everything actually improves anything is a separate question, and probably an individual one. But as an engineering project with a decade of runway, it is worth looking at just to see how someone thought through the design.