Early in the beta, three testers asked me the same thing: "Can you add a bell every 25 minutes?" The Pomodoro technique. Classic.
The three of us sat down one Monday morning and argued about it for three weeks. Here's the short version.
For the bell
It works. Plenty of people who struggle to hold a session find a scaffold in 25-minute cycles. It isn't controversial — it's a proven method.
And it's opt-in. If you don't want the bell, you don't turn it on.
Against the bell
Three objections came up.
First: a bell is a notification. A notification in a time-tracking tool is an interruption. We'd already published a manifesto against interruptions. It would contradict ourselves.
Second: opt-in in theory, default in practice. Users accept defaults. If we set the bell to ON by default, we impose a rhythm. If we set it to OFF by default, we add a feature almost nobody will use.
Third: 25-minute sessions fit some kinds of work and not others. A developer entering a debugging loop at minute 22 will be broken by the bell. A writer working longhand has no use for an alarm.
What we decided
No bell. Not in V1. Probably not in V2.
Instead, we put a small passive visual indicator — a discreet green light in the app window. You look at it if you want. It isn't looking at you.
Silence is expensive
Counter-intuitively, adding nothing takes time. We could have coded the bell in an afternoon. Three weeks of debate cost more. But we gained something: a concrete example we can cite every time a user asks for a feature that interrupts.
Silence isn't an oversight. It's a choice we defend.
— Patrick