Reactions to Ubuntu are rarely balanced. Too often, people love or hate it so extravagantly that the opinions negate themselves. Often, the reactions are so extreme that a fair assessment of the popular distribution is difficult. Add some genuine mis-steps, and the assessment seems almost impossible.
So exactly what does Ubuntu contribute to free software? The answer, I think, is different from what those in either extreme seem to believe.
Looking at the praise for Ubuntu seems largely a waste of time. Scanning through "Why do you use Ubuntu?" a page created by the Canonical design team to gather feedback, what is likely to strike anyone familiar with free software is how often Ubuntu is praised for aspects of either free software as a whole, or of the default GNOME desktop.
The same is true to a lesser extent of Mark Shuttleworth's justifications for Ubuntu , with its stories of how the distribution has positively affected users' lives. The praise does not belong to Ubuntu alone, and most distributions could show similar stories if they chose.
By contrast, the negative views are more thoughtful, but ultimately just as limited.
Some, like the since-recanted attack by ex-Fedora community developer Greg DeKoenigsberg, seem openly partisan. The same is true of complaints from Debian developers, who often seem miffed that Ubuntu's success is built upon their work (as if anything was wrong with that by free software standards).
In other cases, such as the denouncements of Ubuntu because it ships with Mono, the distribution seems unfairly singled out for doing exactly what plenty of other distros are doing. But, of course, other distros are not as popular, so they tend to be overlooked.
Nor is assessment any easier for the fact that a core of valid criticism exists. Increasingly, Ubuntu and its commercial arm Canonical tend to develop internally, rather than with existing projects. Whether this tendency is caused by rejection of Ubuntu contributions or by Canonical's impatience with free software conventions is debatable, but hardly matters. Either way leaves the impression that Ubuntu/Canonical isolates itself from the larger community, and fails to contribute as much as such a large distribution should.
Then, too, Ubuntu has made some genuine mis-steps, such as Shuttleworth's slowness to apologize for sexism in a public speech, or its Contributors Agreement, which is so restrictive that free software developers and advocates such as Aaron Seigo feel unable to sign it.
Such problems seem the inevitable result of a conflict between a company pushing for profitability and the non-commercial standards of much of the free software community. However, if given too much weight, they can easily tempt observers into simply denouncing Ubuntu -- and that would be as unfair as giving it too much credit and praising it uncritically.
Beyond the static
Despite all this background noise, I think that Ubuntu's main contributions to free software are reasonably clear and consistent. Ubuntu is neither the freeloader it is sometimes accused of being, nor simply the savior that is introducing millions to free software. Rather, both directly and indirectly, Ubuntu has been a source of new ideas that has brought the community's attention to matters that it has sometimes overlooked.
Specifically, for the last six years, Ubuntu has been the desktop and usability advocate among free software projects.
Ubuntu is not the first to urge improvements in the desktop. The DotCom era saw an upswing in desktop development at the turn of the last millennium, and GNOME had developed its Human Interface Guidelines several years before Ubuntu began.p> All the same, since 2004, Canonical has been the main commercial advocate of the Linux desktop. Although both Red Hat (despite its claims) and Novell contribute regularly to the desktop, neither has pegged its success to the desktop the way that Canonical has.
Moreover, while Red Hat's and Novell's contributions have been practical, Canonical/Ubuntu has been almost alone in its efforts to try to articulate a vision of the desktop and what it should be.
This emphasis may not be practical. Part of Red Hat's and Novell's reluctance to focus solely on the desktop is based on their own failure to sell distributions commercially, and, in the last year, Canonical has appeared increasingly desperate in it efforts to build a desktop-centered business. But, in the absence of strong rivals, for better or worse Ubuntu has become a main influence on thought about the free desktop.
It helps, too, that in former space-tourist Mark Shuttleworth, Canonical/Ubuntu has a famous representative with some charisma and considerable intellect. Shuttleworth has frequently been accused of egotism, but, to the extent that the charge is true, he is reminiscent of other free and open source software leaders such as Richard Stallman or Linus Torvalds.
Although often criticized, his is a style of leadership that the community understands, and helps to ensure that he has an audience -- especially in the absence of any rival advocates for the desktop.
At any rate, Ubuntu has been an advocate for improvement in the desktop from the beginning. If its name and its original slogan, "Linux for Human Beings," made that position very clear, so, too did its enhancements and tinkering. However, its contributions have been obscured because they have been released one or two at a time. Also, of course, few users both to go back and see what was done in earlier releases.
However, if my memories are correct, Ubuntu can claim a long string of firsts on the desktop. I believe that it was first major distribution to place the menu at the top left, where the eye falls on it first, instead of imitating Windows and placing it on the bottom left. Ubuntu was certainly the first to make tools for switching to multiple keyboards part of the standard installation -- or to add the fonts needed to use different layouts.
More recently, Ubuntu has incorporated more detailed help into the shell and into desktop applications. It has also introduced tools for managing software sources and updates as well as (in a triumph of pragmatism over ideology) restricted drivers. Once you start tallying, the list of Ubuntu's innovations quickly becomes a long one.
Just as important, behind these practical improvement has been Shuttleworth's constant discussions of interfaces and usability. You might think that Shuttleworth's challenge a couple of years ago to equal or outstrip Apple tinged was by jingoism or that his enthusiasm for Ubuntu's latest color scheme makes too much out of too little -- and, at times, I would have to agree with you.
Yet the point is that, by talking about usability and interfaces, then backing up the talk with concrete options, Shuttleworth and Ubuntu have made the free software community aware of these issues in a way that it had not been before.
Admittedly, some of the ideas implemented by Ubuntu have been more successful than others. The verdict on the shift of the title bar buttons to the left is still undecided.
Nor has GNOME accepted Ubuntu's modifications of the notification system -- although, significantly, Ubuntu's efforts seem to have encouraged GNOME to do its own overhaul on notifications.
Even KDE has been working on its notifications -- which, considering that Ubuntu uses GNOME as its default desktop and pays the most attention to it, shows just how far-reaching Shuttleworth's and Ubuntu's influence has been.
Setting the agenda
Whether Canonical and Ubuntu will continue to contribute to the free desktop in the same way is uncertain. The change in default colors early in 2010 seems to have symbolized a partial change in goals as well, a move away from innovation and towards ways of increasing profitability. You might even observe a sense of desperation in this change, as Canonical struggles to succeed with a desktop-centered business plan that has worked for no other company.
Yet even if Canonical/Ubuntu should switch directions or fail financially, that might not matter except sentimentally. Ubuntu has already made the desktop and usability part of the free software agenda in a way that they simply were not before -- and that may be a contribution more significant that the number of patches it contributes to the kernel or to GNOME.