But some time in the last few years, that goal was reached, so quietly that exactly when is impossible to say. Some might still quibble over a feature or two, but the competitiveness of the free desktop is strong enough that equaling rivals no longer seems a major concern.
Instead, the Linux desktop has become the testing ground for exactly what the future of computing will be. It is not the only place where such testing is taking place, but the speed of development and the frequency of releases for GNOME, KDE, and other alternatives often places the free desktop at the center of experiment and innovation.
In fact, there are at least seven issues with which various versions of the free desktop are struggling. Many of these have not been formally announced, but, watching developments over several years, they become obvious:
1) Fine-tuning usability
These efforts have been criticized for being made in the distribution, rather than contributed to the GNOME project. However, Shuttleworth's efforts have got people thinking, and both GNOME and KDE have been paying more attention to such matters than they probably would have otherwise.
Regardless of whether Shuttleworth's efforts will be lost with the switch to GNOME 3 -- which seems likely -- they do seem to have pointed out several areas that needed closer attention.
2) The shape of a perfect menu
The classic menu unfolds across the desktop, one level at a time. However, desktop designers have been unsatisfied with this structure for some years -- presumably because it takes up space.
On Windows, one effort to reduce the size of menus has been to display by default only the most recently used menu items. So far as I know, nobody has copied this tactic on the free desktop, but another tactic borrowed from Windows and OS X is to confine the menu to a fixed size on the screen.
In the KDE 4, series, for instance, the default menu divides entries into several categories, and displays only one menu level at a time. KDE 4.x also offers Lancelot, which freezes the menu size, but displays several sub-levels.
The GNOME 2 series has stayed with the classic menu, but, the last time I looked at GNOME 3, it was opting for moving the menu off the panel and making it a permanent part of the desktop. Other offshoots of GNOME, such as the Unity desktop do the same thing, most likely influence by mobile devices.
Yet another alternative, as typified by Krunner and GNOME-Do, is a floating palette in which users can search for an application to run. However, this approach requires a knowledge of exactly what applications are available, and is largely for expert users.
So far, none of these alternatives seems completely satisfactory. Perhaps the menu is overdue to be retired in favor of some new organizational structure?
3) The netbook challenge
Small-screened and originally for light computing, netbooks are a design challenge. On netbook desktops like KDE's Plasma Netbook, designers have tried to rethink the desktop for the smaller screen and simpler use cases.
So far, however, few projects seem to be considering that modern netbooks no longer have minimal memory, and are being used at least part of the time for normal computing. To some extent, developers are still designing for the last generation of netbooks.
Still another issue is whether netbook desktops are a separate code base or not. KDE has solved this problem by making the Plasma Netbook simply one of several interfaces possible in KDE 4. In many cases, though, netbook desktops seem to be different enough that they need to be maintained separately -- and inefficiently -- from the standard desktops they are based upon.
4) The influence of mobile devices
Today, developers can no longer assume that the desktop or even the laptop is the main computers that users interact with. Instead, user's main computers are likely to be music players, phones, or other mobile devices.
With their small screens, interfaces for mobile devices have restrictions that do not apply even on a netbook. Yet the question remains: if mobile devices are what people are most familiar with, to what extent should their other computers borrow from mobile interfaces?
So far, the most favored answer is that netbooks should borrow heavily from mobile devices, and other computers shouldn't. However, some mobile-like features do seem to be creeping into some interfaces, such as the live links in Amarok that you can drop a track on to perform an action.
Even more importantly, the upcoming GNOME 3 seems to change entire screens in a way that is more reminiscent of a mobile device than a traditional workstation.
It may be that the influence of mobile devices will continue, and that the workstation of the near future will include a desktop that looks far more like that of a phone than like the desktops we are using now.
5) Local vs. Cloud Apps
Chrome OS, Google's upcoming operating system, presumably started as an effort to integrate the desktop with Google's online applications. However, it was quickly followed by similar efforts, such as the Jolicloud distribution.
So far, distributions that rely on applications that are not installed locally are a minority, and people are investigating them because of their novelty as much as for their usefulness. When you are using free software, cloud-based computing is less compelling, because you can install what you need on any computer without worrying about licensing -- and you may be disinclined to use online apps that do not have a free license.
Still, the possibility remains that cloud computing will become an increasingly important part of the free desktop.
Alternatively, instead of merging the desktop with the browser, perhaps other free desktops will go the route of KDE, and scatter online resources through the desktop. If so, the outcome may be a hybrid of local and online services.
6) Anticipating future needs
More than any other free desktop, KDE is trying to anticipate new ways that users might want to use the desktop. The KDE 4 series includes a number of special effects, some of them eye-candy, but many practical aids for convenience or accessibility. It also includes hot spots on the edges of the screen, and the remote sharing of widgets. At the same time, it has added geo-location to file attributes, and widgets for the social desktop that add easy access to online resources directly to the desktop.
Both GNOME and KDE are also moving towards various means of swapping icon sets, to make the desktop more task-oriented and less application-oriented. Aaron Seigo, the KDE developer, even envisions icon sets that change automatically depending on user's locations.
Whether users want all these tools, or will use them, may be questionable. However, more than anything else, such experiments show just how far the free desktop has come from the days when it was trying to catch up with Windows and OS X.
7) User or development focus?
All these issues raise a question unique to the free desktop: Are these experiments for the benefit of the users, or the developers?
To those who do proprietary development, this question would hardly need to be asked. Officially, at least, proprietary development is supposed to be done for the users -- although it is easy to be cynical about this claim.
However, the question arises naturally in free software because, to attract and retain developers, a project needs to offer the chance to do new and interesting coding. Given this circumstance, desktops like GNOME or KDE cannot afford to stand still.
The problem is that, today, free desktops can add little to basic functionality. Any experiments have to be tentative, and risk leaving users behind, or making them actively hostile -- especially since a growing body of new users react to free software companies exactly as they once did to commercial companies.
And, since these new users are not coders, telling them to get involved in the development process (the traditional answer to those who complain about the state of free software projects) is not a satisfactory answer.
The result is that, somewhere in the middle of all these technical concerns, free desktops and distributions have the social problem of trying to find the right balance between pleasing developers and users. Some observers already believe that the reception of KDE 4.0 was partly the result of too little attention to users, and worry that the same imbalance may be behind GNOME 3.0
Which direction to take?
Behind these issues are some larger ones about the general direction that the free desktop should take. In both GNOME and KDE, the mainstream assumption seems to be that the desktop will continue to grow, adding feature to feature until, at some point, a rewriting or a general cleaning of the code becomes necessary.
Yet that direction is being challenged. For some, the current state of the free desktop is advanced enough that little is needed. Presumably, incremental changes will happen, but not major ones -- which is presumably one reason why some distributions and users continue to prefer the KDE 3 series of releases over KDE 4. A similar reaction may occur when GNOME 3.0 is released.
Moreover, to judge by the popularity of Ubuntu variants that use desktops such as Xfce or LXDE, for some the free desktop as defined by GNOME or KDE is already too bloated. For such users, the issue is not features, but reducing the memory overhead as far as possible.
None of these alternatives is likely to overwhelm the others -- not even mainstream GNOME and KDE. But, like the other issues, if you value choice (and who doesn't?), these possible directions make this an interesting time to be a user of a free desktop.
ALSO SEE: Top Ten Apps That Make Linux Fun To Use