Scott Jenson, a veteran UX designer who spent years at Apple and Google, delivered a talk at Akademy 2026 arguing that the open-source desktop has spent two decades copying macOS and Windows — and that strategy has expired. Apple shifted focus to iPhone and iPad, letting the Mac languish. Microsoft is busy inserting ads, pushing OneDrive, and deploying the privacy-invasive Recall feature. Neither company is doing meaningful desktop UX research anymore, which means the Linux community's long-running strategy of waiting and borrowing has hit a dead end. Jenson's core thesis is simple: the WIMP paradigm (windows, icons, menus, pointer) stabilized in the early 2000s and has barely moved since. Rapid innovation in the 1980s and 1990s gave way to a long plateau. People are doing vastly more with desktops today, but the interaction model hasn't kept pace. Mobile won the consumer and social-media markets, but it did not win productivity — people still work on desktops with keyboards and large screens. That fact, Jenson argues, is underappreciated and under-invested. He illustrated the gap with micro-level examples from his own career. While working on the Macintosh Finder, he pushed for mid-string ellipsis in truncated file names — a detail now considered a signature Mac polish element. He also described how macOS handles click-and-drag between windows: selection happens on mouse-down, but the window raises only on mouse-up, enabling seamless cross-window file dragging. Most Linux desktops raise the window on mouse-down, breaking the workflow. A KDE developer saw Jenson discuss this on Mastodon two years ago, agreed it was a good idea, and shipped the fix within hours. That anecdote is the strongest case for the open-source model: zero institutional friction between a good idea and its implementation. But Jenson was careful to frame the structural deficit. Open-source projects are chronically understaffed. GNOME has conducted some user research, and KDE has its human-interface guidelines, but no one funds or conducts regular UX research and development for the Linux desktop overall. The exception-driven model cannot substitute for systematic design work. Jenson categorized the three main objections he encounters: 'mobile won, desktops are legacy,' 'WIMP is a solved problem,' and 'don't touch my stuff.' He rejected the first two — productivity remains desktop-bound, and the WIMP model is a plateau, not a ceiling. The third he acknowledged as legitimate emotional resistance, but argued the world is changing and the desktop must grow into new capabilities gently rather than freezing indefinitely. His talk also broke UX into layers that technologists frequently conflate: style (pixels, color, icons), structure (navigation, information architecture), strategy (target user, priorities, the discipline of saying no), and what he called 'stuff' — the low-level platform constraints that shape what's possible. Forty years of desktop evolution have deposited historical cruft at every layer, and most open-source contributors focus on style while structure and strategy go unaddressed. The implicit question hanging over the talk is funding. Jenson now works with Mastodon and Home Assistant, squeezing better UX into open-source projects on the margins. His Ubuntu Summit talk reached half a million YouTube views, proving demand exists. But demand without resources is just a wishlist. The generative potential here is real — the zero-friction KDE anecdote proves it — but the systemic underinvestment in UX research means most of that potential sits dormant, waiting for someone to fund the work Apple and Microsoft have abandoned.
