Plasma's fifth release series is an improvement over the fourth, with many core components rewritten and speeded up. However, like the recent work on most other desktops, it shows few signs of any rethinking.
The array of innovations that marked the fourth series, making the desktop more flexible and customizable, is missing from the 5.2 beta. Instead, it adds administrative and configuration tools, a few new widgets, and a handful of minor improvements, such as the ability to undo previous changes to the desktop.
The greatest change is that the new version of Plasma shows the efforts of the Visual Design Group that has been working for the last year on improving KDE's often under-organized dialogues. Unlike GNOME, which might be described as humanistic in typographical terms, KDE has opted for a geometrical design, resulting in a distinctly different look. Otherwise, Plasma 5 might more accurately be described as Plasma 4, Part 2.
Ubuntu's Unity goes where Canonical Software goes. Unity started as an attempt to upgrade the aesthetics of the Linux desktop. In the last couple of years, however, its releases have offered little immediate to the desktop user.
Instead, the development of Unity has been largely invisible to users. Its development has been focusing on porting the desktop to mobile devices, in keeping with Canonical's plans to release a phone and a tablet.
More recently, Canonical has concentrated on cloud services, improvements to virtualization via Driver, and readying for smart devices. Sooner or later, administrative tools for these purposes are likely to be added to Unity, extending the traditional desktop in several directions.
However, while undeniably innovative, these directions all depend on Canonical establishing niches in highly competitive markets, which is by no means assured. Should Canonical fail, then any tools added to Unity will be white elephants rather than meaningful contributions to the free desktop.
Current plans for the Linux desktop are decidedly more modest than those of the last era. But perhaps that is what both users and developers want -- to avoid a repetition of the user revolts, and get on with their work as usual.
At any rate, the changes coming for all the major desktops may be enough without complicating events with new options. All the major desktops will be facing the transitions from the X Window System to Wayland, and from Init to Systemd in the next couple of years, and Systemd is already the center of controversy. With these challenges looming, who has the time or desire to take on more?
In fact, under these conditions, perhaps the fact that most changes will be minor, invisible, or aesthetic should not be surprising. Perhaps the surprise is that we are seeing any improvements at all.