Problems and solutions to the modern desktop (Make tmux the OS)

Scott Jenson's presentation, "Are we really going to use the same Desktop UX forever?" argues that Apple and Microsoft are unlikely to innovate further in the desktop OS space, and it is up to the community to decide the future of the desktop. Jenson suggests that many long-standing issues with desktop UX are essentially workarounds for hardware that is no longer present, and questions why we continue to accept these tradeoffs.

He proposes a concept called "tmux is the OS," which aims to take the exceptional experience of tmux—a scrollable, session-persisted, detachable, task-oriented system—and make it accessible to the average user. Jenson identifies three main problems to tackle: dynamic screen real estate, varying window management preferences, and the shifting nature of software and application storage.

Jenson notes that today, a significant portion of our software runs within browser tabs, which behave differently than traditional local applications. To address this, the desktop OS should treat a browser just like a normal application. Furthermore, the concept of the filesystem is becoming less relevant, as notes, texts, and team communications are increasingly stored within application databases rather than local folders.

To explore potential solutions, Jenson references the 1986 paper by Henderson, Austin, and Card titled "Rooms: the use of multiple virtual workspaces to reduce space contention in a window-based graphical user interface." This research suggests that managing windows more effectively by preloading the next set of windows before they are requested could significantly reduce the cognitive load on users. This concept aligns with the idea of "rooms" as virtual workspaces that minimize the user's memory faulting.

Jenson also mentions the paper "No Task Left Behind? Examining the Nature of Fragmented Work" and its confirmation that modern window management systems do not cater to the way people actually work, which often involves juggling multiple tasks simultaneously. In contrast, WindowScape—a task-oriented window manager—addresses the issue of window arrangement by creating photographs of the desktop layout, allowing users to return to specific snapshots easily.

This approach solves the problem of windows needing to belong to a single group, as users intuitively understand that multiple photos can exist of the same object.