Linux Format logo

Future desktops

After reading Andrew’s excellent Roundup on alternative desktops (p30), I’m not sure how I feel about the way desktops are going. I’m still surprised, for example, that both Gnome and KDE developers made such massive changes to their desktops, when for many years the old versions had worked brilliantly. KDE 4.9 is stable, but it still takes a lot of effort to make the environment your own. And despite Microsoft staking some of its future on it, I don’t like Gnome’s homogenous touch interface. I accept that tablets and smartphones are being used more for serious work, but the desktop is an ‘interface’, and until the interfaces that connect the user to the computer are the same, I think it makes little sense to make the ‘user interface’ similar, especially when there are far more pressing problems.

One such problem is how little input mechanisms have changed. Developers are building the visual aspects of a touch interface, but they’re not considering the touch input itself. There seems to be no standardised gesture support, for instance, even though the back-end technology exists, and nothing to create a series of multi-touch gestures that should work across all desktops, distros and hardware. Surely this should come before design.

A second problem is resolution. I recently saw Linux running on a super-high resolution display. These are going to become increasingly common, and yet the Linux desktop is not prepared for this shift. Fonts sometimes scale, sometimes don’t. Window decorations are often too small, and nearly all icon sets are rendered at far too low a resolution to be usable at 220 dpi. Perhaps we need a shift in priority?

Your comments


I think it is important to forget the desktop paradigm. This was introduced in the early 1980s as part of Digital Research's attempts to attract business users to its software.

At the time the Xerox Parc paradigm was quite different and much closer to KDE4: an interface on which the user could place whatever they wanted. So you drew windows and started applications within them - not unlike the widgets in KDE4. Though the MS RT interface draws on Gnome for its look and feel, the idea of having applications running inside small windows on the screen is closer to Xerox Parc than Gnome.

The problem with KDE4 is that it is still a hybrid - it has a window manager for traditional desktop applications because people haven't worked out/got used to running such applications in a widget - which they may do with the MS RT interface.

But KDE4 did take the far-sighted decision to move to SVG which makes it much easier to handle higher resolution screens and, though the window manager still lacks some functionality which certain users need, it does automatically adjust to the available monitors. So unlike fellow users using non-KDE desktops, I don't have to open a dialogue to show a presentation at a strange location because KWin simply adjusts everything automatically.

In the late 1970s there was a shift from the teletype interface to the monitor interface; now we have the addition of tablet and mobile phone interfaces. We need to think in terms of interfaces rather than the desktop - which is just one of many examples of an interface paradigm.