The last five years of user revolts have left Linux desktop users wary of innovation. Too often recently, "innovation" has meant unwanted changes imposed without any consultation by developers upon users. As a result, Linux desktop development has become cautious, avoiding major changes that are visible to users.
One sign of these times is that many users are voicing the opinion that this attitude is a good thing. They talk dismissively of change for change's sake, and regard GNOME 2 with an awe that it never received during its heyday.
However, if unrestrained change is undesirable, such conservatism seems simply its extreme opposite. Even granted that most experiments to improve the desktop will fail, some efforts at innovation seem desirable.
If nothing else, such efforts help to attract and retain developers for a project -- and, at best, they may occasionally come up with features that transform computing for their users, such as KDE's Activities and Folder Views.
Besides, some change is inevitable. Even post-revolt, some innovation persists on all the Linux desktops. Mostly, its long-term goals are poorly defined and sometimes tentative, but as computing changes, a few small innovations continue to find their way on to the screen despite the general lack of encouragement.
Here's what we might expect in the way of changes in the next couple of years on the main desktop environments:
With the 3.8 release, GNOME reached a certain stability. The GNOME-Shell, the center of controversy for two years, is now supplemented by GNOME Shell Extensions, a collection of plugins that, carefully chosen, can almost recreate the GNOME 2 desktop on top of GNOME 3.
Having reached this stage, GNOME seems to be concentrating on new applications for the upcoming 3.10 release. According to Mattias Klasen, these applications include the self-explanatory GNOME Music and GNOME Maps, as well as a viewer for Git repositories, a note-taker, and a re-designed movie player.
Other features for 3.10 include integration of the Flickr photo-sharer and the Zimbra groupware and email server into GNOME.
However, probably the biggest change in the next version of GNOME will be support for Wayland, which many people are expecting will be the replacement for the aging and much-patched X Window system. Preliminary reports suggest that the effort is well under-way, although how much difference casual users will notice seems doubtful.
What happens next with GNOME still seems undecided. At last year's GUADEC, GNOME developers talked about various ways to revitalize the project. A key presentation was Xan Lopez and Juan Jose Sanchez's "A Bright Future for GNOME," which became the inspiration for several breakout sessions.
However, eleven months later, many of the suggestions made in the presentation, such as attention to mobile devices, touch screens, and cloud services, have had little visible effect on GNOME's direction -- at least so far.
A better glimpse of GNOME's future directions may be the copy for Friends of GNOME fund-raising campaign. Although apparently intended only as examples rather than a road map, the copy mentions a number of privacy and security features that might be added to GNOME in the future.
These suggestions include Tor integration for anonymous browsing, increased support for disk encryption, anti-phishing features, and the integration of applications with system-wide privacy settings.
Any work being done on such features is not immediately visible online. However, logically, such features would come after the basic desktop applications due for the 3.10 release.
This year's GUADEC may offer a clearer roadmap, especially if the project decides to go ahead with the suggestion made last year to launch GNOME 4.0 in 2014.
Five years into the 4.x release series, KDE is in the best shape of all the major Linux desktops. It long ago outlived the user revolt touched off by KDE 4.0, and now provides a mature feature set. It is currently in the middle of switching to QT5 and QML, rewriting many aspects of the desktop environment, with 4.11, the next release, intended to be a long-term release that will see no major changes except bug-fixes in the Plasma work-space.