Paul Ford's TinyLittleOffice argues the browser is now more than enough for rich apps

An office suite that lives in a URL

Last week Paul Ford, the writer and consultant behind ftrain.com, launched tinylittleoffice.com, a small, deliberately old-fashioned office suite with a gimmick at its core: documents are stored in the URL itself rather than on a server. In an essay titled "The Web Is More Than Enough," Ford reports that the project found a far bigger audience than expected. Traffic overwhelmed his free Cloudflare plan, and once he added a payment card, he says a couple of hundred thousand people came to look. The Register covered the launch, and the suite briefly reached Product Hunt's top ten. Reaction split along familiar lines: Hacker News commenters dismissed it as pointless, while GitHub users opened feature requests. Ford says the project will keep being supported and improved in small steps.

The claim underneath the launch

The essay's real subject is the platform. Ford's argument is that the browser has been powerful application software for a long time, but the cost of reaching that power now appears to fall almost daily. By his estimate, he put roughly twice the effort into the office suite that he would spend writing a long newsletter post — and even after the traffic spike, Cloudflare's bill came to essentially nothing.

To illustrate what sits inside a modern browser, Ford lists capabilities many developers know by name but few have explored deeply: the Web Crypto API for security work, WebAssembly for near-native speed, WebGL for 2D and 3D graphics, the Web Audio API for sound, PDF.js for handling PDFs, and HTML's ability to represent most of what a document can do on a computer. TinyLittleOffice itself shipped in 21 languages, chosen to match the most common languages of New York City — which is why Creole and Yiddish made the list — with accessibility infrastructure included.

These APIs have existed for years, Ford notes, but each one is enormous, and getting good at Web Audio leaves little time for Web Crypto or WebGL. His reading is that the browser's mega-libraries were built for large teams — organizations like Google shipping software securely to billions — with individual developers as an afterthought. That balance, he argues, is shifting.

A second data point

Ford also points at Simon Willison's Battle Drones, a 3D multiplayer browser game hosted on GitHub. Willison is not a games developer, Ford emphasizes, yet the game loads and plays directly in the browser, supports multiplayer, and works offline. According to Willison's own notes, cited in the essay, an approach that carries multiplayer data over audio appears to be a first for a real-time strategy game, and he sees a path for quickly built, AI-assisted games to be taken seriously. For Ford, it is more evidence that capabilities once reserved for large studios now fit inside a personal project.

Consequences for the software economy

Ford draws out two implications. The first is competitive: when a working idea can be tested from a web server for effectively no money after a couple of weeks, the toll app stores charge starts to look much less worth paying. The second concerns labor. He is sharply skeptical of large AI companies' loftier predictions, but he thinks an economy in which perhaps a hundred times as many people build software for each other would remake the current tech industry. He also notes an uncomfortable parallel: software development increasingly feels like writing to him, and writers are famously paid little. Perpetual devaluation would be a new experience for software workers, even as clients who need real problems solved find themselves in an unusually capable moment.

Why it matters

For web developers, the essay works as a corrective to platform fatigue. The browser is often treated as a messy compilation target; Ford frames it as a broad, nearly free application runtime whose entry price keeps dropping. TinyLittleOffice is the proof by construction: a quirky side project, built with roughly newsletter-grade effort and served for about $0, reached hundreds of thousands of people, in 21 languages, with no app store in sight. Storing documents in URLs may or may not catch on, but the economics Ford describes — free distribution, instant availability, offline capability, no gatekeeper — strengthen the case for the web as the default target for new software, and put fresh pressure on intermediaries whose value depended on distribution being expensive.