The Workspace as a Calm Place
Not every tool needs to be an infinite canvas. The argument for focused software that knows what it isn't.
You've probably seen the Notion setup. A hundred pages, color-coded by project. Task databases with seventeen properties. A bookmarks table, a reading list, a daily journal template with linked databases. The person who built it spent two weeks getting it right. By March they had stopped updating it. By May they had started fresh in Apple Notes.
The tools I trust most tend to do one thing.
Not because simplicity is a virtue in the abstract. There's a practical asymmetry at work: a tool that does one thing can be really good at it. It can be fast, because it isn't loading features you're not using. It can be opinionated, because there's a single clear purpose to be opinionated toward. And it can be calm — which is a word I've started taking more seriously as a design quality.
The everything-app as aspiration
The consolidation dream keeps recurring because the logic is appealing. Fewer tools means fewer switching costs, fewer places to look, less overhead. In 1978, Bell Labs programmer Doug McIlroy wrote what would become one of the most cited pieces of software advice: "Write programs that do one thing and do it well." He was describing Unix tools. The insight was about composability, not consolidation — small programs are more powerful than large ones because they can be combined.
That principle didn't stop the everything-app from becoming a persistent fantasy. WeChat in China remains the canonical example: one app for messaging, payments, shopping, flight booking, gaming. Elon Musk bought Twitter in 2022 and announced, without much irony, that he intended to build the Western equivalent. Notion became the productivity world's version: notes, databases, wikis, projects, AI assistant, calendar integration, automations. The 2024 feature list is extensive.
The result is a tool that is capable of anything and optimized for nothing. Performance degrades as the feature surface expands. Setups that felt elegant in January become graveyards by June. The promise of one place for everything quietly becomes the problem of everything being slightly worse than a specialized alternative.
What calm actually requires
Amber Case has been thinking about this longer than most. In her 2015 book Calm Technology, she formalized an idea that the researchers Mark Weiser and John Seely Brown first developed at Xerox PARC: that the best technology "informs but doesn't demand our focus or attention." In 2024, she founded the Calm Tech Institute to create measurable standards for attention-aware product design.
Her core principle is blunt: "the right amount of technology is the minimum needed to solve the problem." It sounds obvious. It is almost never what actually gets built.
The structural reason is clear. Software products grow because growth is how you retain users and justify investment. Each new feature is a reason not to switch. An everything-app is maximally defensible from a product perspective: every piece of your workflow you store there becomes a switching cost. The incentives are strong and point in one direction.
But the cost of that expansion accrues to the user. A tool that wants to handle everything asks you to become the person who decides how it handles everything. You become the systems administrator of your own productivity stack. Most people do not want this job, and they make their feelings known by stopping to use the tool.
The shape of a good workspace
The thing that distinguishes tools I've kept from ones I've abandoned is not how many problems they solve. It's that the tools I keep have edges. They know what they're not.
A calendar that doesn't want to be a task manager. A code editor that doesn't manage your files. A knowledge base that stores notes and documents, not email, not spreadsheets, not video. The edge is where the tool stops asking for attention and lets you redirect it elsewhere.
Calm, as a design quality, is not the same as minimal. A tool can look sparse and still require constant reconfiguration. Calm means something more specific: the tool does its job, stays in the background, and doesn't continuously expand the surface area of decisions it wants you to make.
Building software with real constraints is harder than building without them. It means saying no — to obvious features, to plausible extensions, to the gravitational pull of "well, why not?" It's easier to add a calendar integration than to think clearly about whether the calendar belongs here at all.
But the software that comes out of that discipline has a shape. You understand what it's for. And there's something to trust in a tool that isn't trying to become everything, because you can see exactly where it stops.
Asgeir Albretsen is the founder of Harbor.