If you had described the modern computing landscape to a developer in 2008, the part they would have struggled with is not the AI or the phones or the screens that fold. It would have been the boredom of the operating system. The OS used to be where the interesting fights happened — Mac versus Windows versus the early years of Linux on the desktop. People had strong opinions. People wrote essays. People switched, with ceremony.
Nobody switches operating systems anymore. We change browsers. We change tabs. The OS is the thing the browser runs on, and the browser is the thing the work happens in. This is not a small reframing — it is the central platform shift of the last fifteen years, and it has consequences that the desktop-software trade press is still slow to acknowledge.
How we got here
The shift was not sudden. It took three overlapping waves to get the browser from 'document viewer' to 'runtime for everything.'
The first wave was Google Maps in 2005 and Gmail not long after — the moment people realized a browser tab could feel like an application, not a webpage. The second was the cloud revolution of the early 2010s, when storage and compute moved off the desktop and the file system started to feel quaint. The third, the one we are still in, is the consolidation of every productivity workflow into a small set of multiplayer web apps that assume the network the way old software assumed the disk.
By the time WebAssembly matured enough to ship Figma's renderer in 2017, the conversation was already over. Figma was not the first design tool to exist in a browser, but it was the first to be obviously better there than anywhere else. The thing that 'could not run in a browser' ran in a browser, and ran in a browser well enough to win the category outright in five years.
What the browser swallowed
A short, incomplete inventory of categories the browser owns now that it did not own in 2015:
- Office suites. Google Docs is the default for most teams under fifty people. Microsoft 365 in the browser is the default for most teams over fifty.
- Design. Figma, Canva, and a long tail of niche tools have made desktop design software a holdout product, not a default.
- Code editing. VS Code in the browser, Codesandbox, Replit, Stackblitz, GitHub Codespaces — for entire categories of work the local editor is now optional.
- Music production. BandLab, Soundtrap, AudioMass — not quite Logic, but enough for the long tail of casual production.
- Video editing. Descript, Capcut Web, Veed — still a notch behind native for high-end work, but for podcasts, social cuts, and screen recordings, the browser version is now the right tool.
- Whiteboards and diagrams. Miro, FigJam, tldraw, Excalidraw, Lucid. The whole category was born in the browser.
- Project management. Linear, Asana, Notion, ClickUp. Even Jira, the holdout that defined desktop client memories for a generation of engineers, is functionally a web app now.
What the OS still owns
Three categories still meaningfully belong to native software in 2026. They are worth naming because they are also the categories where the browser is making the slowest, hardest progress.
1. Pro media work
Final Cut, DaVinci Resolve, Logic, Ableton, Adobe's premium tools, Blender, ZBrush. The reasons are technical and getting weaker. GPU access is improving in WebGPU. Real-time audio latency, the killer issue for music production, is finally usable in the browser as of 2025. But the entrenched workflows are deep, the plugin ecosystems are decades old, and 'good enough on the web' is not the same as 'as good as a tool I have used for fifteen years.' The browser will probably win these too, but slowly. Maybe by 2030.
2. Games
AAA games have not meaningfully moved to the browser, and likely will not for the foreseeable future. Streaming services are a partial workaround. WebGPU is closing the rendering gap. The real obstacles are distribution (Steam's grip on the player community) and the install-once, play-many ergonomics that the browser does not naturally support. Smaller, web-native games are flourishing — itch.io's web games tab is one of the most interesting corners of the internet right now — but the AAA category remains native.
3. Local-first systems
Anything that needs to work without a network, manage strict offline state, or interact deeply with the filesystem still benefits from being native. PWAs have closed a lot of this gap. The File System Access API exists. But for the small set of users whose workflow is 'I open this app on a plane and need it to behave like the network does not exist,' native is still the safer bet.
Why this keeps happening
The browser wins for boring reasons, and they compound.
- 1Distribution is free. There is no app store gatekeeper, no review queue, no thirty-percent tax, no separate Mac and Windows and Linux build pipelines. A web app ships when the engineer clicks deploy.
- 2Updates are continuous. Users always run the latest version. The 'are you on the right build?' bug class disappears entirely.
- 3Collaboration is native to the medium. Multiplayer is the default mode of a web app, not a feature that gets bolted on in version four.
- 4Cross-platform is the default. The same product runs on a Chromebook, a Steam Deck, a phone, an iPad, a Linux desktop, and the receptionist's locked-down Windows machine. Native software has to be ported. Web software is portable by construction.
- 5The cost to try is zero. No install. No commitment. The funnel from 'I read about this app' to 'I am using it' is one click. This single fact has reshaped how products grow.
Any one of these is a meaningful advantage. The compound effect is overwhelming. A native company has to be better in some specific, defensible way that the browser cannot match. Most cannot articulate what that way is.
What this means for makers
If you are starting a new productivity tool in 2026, the answer is almost certainly 'build it in a browser.' Not 'build it as a web app with a desktop wrapper.' Build it for the browser, with the assumption that the browser is the most important platform you will ever target. The desktop wrapper, if you want one, comes later and is mostly there to satisfy the small slice of users who insist on a dock icon.
The exception, still, is the three categories above — pro media, games, local-first. If you are building in those, the calculation is different and the native option is real. For everything else, the question is not 'native or web' anymore. It is 'how good can the web version be.'
What this means for users
The practical takeaway is that the time you spend choosing software has gotten dramatically cheaper. The penalty for picking the wrong tool is smaller — switch the tab, sign up for the alternative, move your data. Tools should be evaluated on what they do, not on the fact that you have to install them. The install button is no longer the commitment it once was, because nothing about your computer changes when you press it.
More provocatively: the operating system you run no longer determines what you can do. It determines which browser performs best, which fonts render most pleasantly, and which battery life you have. Those are not nothing. They are also not the categories the old OS holy wars were fought over.
This is, in part, why dotstore exists. The catalog is a catalog of browser-native software because that is where the interesting work has moved. The OS has become the thing that gets out of the way. The tab is the thing you are actually using. Treating the tab as the platform — naming the apps, ranking them, comparing them, building taste around them — is a small bet that the shift is real and worth taking seriously.
The browser ate the operating system. Most software made the migration. The catalog above is the part that survived the journey.
Tags