Veteran user-experience designer Scott Jenson used KDE's Akademy 2026 conference to press open-source desktop developers to make experimentation a central part of their work. His argument was not that familiar windows, icons, menus and pointer controls should disappear, but that the desktop has changed too little while the range of tasks people perform on computers has expanded.
Jenson has worked on interfaces at Apple and Google and now contributes design work to projects including Mastodon and Home Assistant. He told the conference that open-source software often lacks sustained UX research because projects are understaffed. That limitation, he argued, has encouraged Linux desktops to follow patterns established by commercial operating systems instead of developing and testing their own alternatives.
One example showed how small interaction decisions can have broad effects. On macOS, selecting an item in a background window begins when the mouse button is pressed, but that window is not raised until the button is released. This lets a user drag a file between overlapping windows without the source covering the destination. Jenson said he previously raised the differing Linux behavior publicly and a KDE developer implemented a change within hours. He presented that response as evidence that open development can turn a focused design observation into working software quickly.
His larger concern was that depending on Apple and Microsoft for desktop research is no longer a reliable strategy. Jenson said commercial vendors increasingly prioritize ecosystem integration, cloud services and other business goals that do not necessarily improve basic desktop work. Linux projects can still preserve conventional workflows for users who want them, he said, while creating optional experiments that test different ways to organize applications, files and tasks.
The proposals include using contextual information about a user's activity to reduce repetitive setup and exploring interfaces that expose relationships between work rather than treating every program as an isolated window. Jenson also emphasized prototyping: ideas can be placed in front of users, observed and discarded when they do not help. That approach treats design as an iterative engineering discipline rather than a one-time visual overhaul.
The talk highlights a recurring tension for community projects. Stability and familiarity make software accessible, but they can also discourage trials whose value is uncertain. Open-source teams rarely have the research budgets of large platform companies, yet they can invite public testing and improve prototypes in the open. Jenson's case to KDE was that this structure is a strength if developers deliberately use it: keep dependable desktops available, create room for experiments, and measure whether new interactions solve real problems before making them defaults.
