Also see: Why I Switched to GNOME
Five years ago, GNOME was the main contender for the Linux desktop. It battled KDE, and, more often than not, came out on top. Today, it is down -- if far from out -- to the extent that any observer has to ask: Can GNOME ever regain its former predominance?
True, GNOME technology still dominates the desktop, with Cinnamon, GNOME, MATE, and Unity all using GNOME-based applications and utilities. However, the last few years have not been kind to the former giant.
First, the early releases of GNOME 3 were different enough that many users deserted it after a quick glance, turning to Xfce and Linux MInt's Cinnamon and MATE -- neither of which would probably exist otherwise. In the Linux Journal's Readers' Choice Awards of 2013, GNOME was the choice of only 14%. It did even worse in the 2013 LinuxQuestions.org Members Choice Awards, with only 10%. By themselves, these are hardly definitive numbers, but their consistency is enough to make them ominous.
Nor are these the only signs of decline. In 2013, GNOME's annual report recorded a shortfall of $78,965, about two-thirds of which was due to a decline in corporate sponsorship.
Almost the only sign of encouragement is that, eleven months after first considering whether to replace GNOME with Xfce as the default desktop environment, Debian decided to stay with GNOME. This decision has been widely -- and erroneously -- reported as Debian switching back to GNOME, but it amounts to more of a rearguard action than a victory, considering that no switch had ever been made. That this decision should be hailed as a victory for GNOME only emphasizes how hungry for good news GNOME die-hards have become.
None of these developments is definitive by itself. Yet their cumulative effect is a portrait of a project in slow decline.
Hearts and Minds
If there was any justice in the world, much of this decline would not be continuing. Just as Linux itself is often judged by the way it was years ago, so GNOME is judged by the first releases in the GNOME 3 series.
However, GNOME 3.14, which was released in September 2014, is hardly recognizable as the source of the original complaints. By itself, GNOME's promotion of extensions is enough to answer most of the complaints about the interface, giving users a wide array of choices.
In addition, the latest GNOME release shows a concern with design and usability unmatched by any other desktop environment, especially in the details. Recent releases have also shown a growing interest in security and accessibility.
Admittedly, GNOME may still need to be tweaked after installation to satisfy users, but that is hardly unique among Linux desktop environments. If anything, such customization is usually seen as a benefit among Linux users.
Some GNOME supporters see GNOME's technical side as proof of recovery. A review of GNOME 3.14 went so far as to run under the title "How GNOME 3.14 is winning back disillusioned Linux users."
However, while the article praised GNOME's recent developments, it offered few proofs that users were being won over again.
Moreover, the replies to the article suggested that users are not being won over. One commenter wrote that the recent improvements were "too little too late," and the only agreement in the comments came from GNOME volunteers.
Otherwise, the most enthusiasm that could be mustered came from one or two comments that suggested that, while GNOME might be fine for others, it wasn't for them.
To me, these comments sound as though those who moved away from GNOME, having found acceptable replacements, have little interest in another upheaval on their desktop. GNOME today has its points of appreciation, but it lacks any killer apps that make it more compelling than any other desktop environment, especially since GNOME apps are available on other desktops without the GNOME shell.
The real problem seems to be less technical than social. GNOME might have weathered the hostile early reactions to GNOME 3 if they were purely technical. After all, faced with a similar revolt a few years earlier, KDE managed to recover.