Off the Beat: Bruce Byfield's Blog

OK, it's finally happening. For a couple of years, the standard for desktops has been edging towards mobile devices with every new netbook interface. Now, with Ubuntu's decision to replace GNOME with its own Unity desktop, the shift away from the workstation standard has actually arrived.

But while the decision may be in keeping with the times, it's still every bit the "risky step" that Mark Shuttleworth described it as when he made the announcement earlier this week.

In making the announcement, Shuttleworth focused on technical issues. The GNOME project, he said, had made some design decisons that were not in keeping with what Ubuntu wanted to do. He also mentioned the greater ease of supporting one desktop, rather than two.

These comments have been duly reported. But what has not widely discussed are the implications of the decisions for Ubuntu's future growth and its relationship with the larger free software community.

The risk begins with the fact that Unity is still very much a work in progress. The current version of Unity has elements that need to be standardized, and is designed with assumptions about how the desktop is used that hold true for only some users.

No doubt Unity will improve by the time it becomes Ubuntu's default desktop, but a poll on the Ubuntu forums whose results are roughly equally divided between Unity, the forthcoming GNOME Shell, and "Anything that works" suggests that, even with improvements, Unity will have to struggle for acceptance.

Jockeying for market advantages

As the reception of KDE 4.0 and the lukewarm enthusiasm for the GNOME Shell demonstrate, any major interface change may meet with resistance. There seems no reason to think Unity will be an exception. For this reason,you can safely assume that Ubuntu and Canonical, its commercial arm, have some compelling reasons (read: business reasons) to take the risk in the first place.

Perhaps, as some pundits have suggested, the somewhat minimalist Unity is intended for the promised future of cloud computing, in which the desktop is just the launcher for the browser. The only trouble with that, however, is that one innovation rarely replaces older ones completely; cloud computing is likely to continue to grow, but it is likely to co-exist with traditional computing for some years.

Similarly, if the distro is willing to risk annoying existing users -- users who have already endured a barrage of changes in the sometimes dubious name of usability -- then perhaps another calculation is that Unity will appeal to the new users whom Shuttleworth expects to be flocking to Ubuntu in the next five years. But will those users, most of whom will have experience with Windows or OS X, appreciate Unity's ease of use, or regard it as a step down from what they are used to?

Probably the most compelling reason is ease of maintenance as Canonical and Ubuntu expand. Not only is one desktop easier to maintain than two on a workstation or netbook, as Shuttleworth suggests, but it is far easier to maintain on tablets and mobile devices, too -- assuming that Canonical should ever expand into them.

KDE has already shown a realization of this simple truth with Plasma Netbook, which uses most of the same code as standard KDE except for the desktop shell. However, perhaps Ubuntu's plan is to take the approach one step further, and make the code completely identical across all hardward platforms. If so, then the increased coding and testing efficiency is self-evident.

Yet the nagging question remains: Do users want or expect the same desktop everywhere? What may seem efficient on an eight centimeter phone screen can seem cramped and restrictive on a workstation.

Like Shuttleworth said, it's risky. Yet if Canonical is going to manage what no one else has ever managed to do and make a living from a distribution, some assumptions have to be made and some risks taken. What is interesting about Ubuntu standardizing on Unity is that the decision strongly suggests what assumptions Ubuntu is currently making.

Redefining relationships

Just as importantly, with the decision to focus on Unity, Ubuntu now has the usability issues upon which it is focusing completely under its control. This is a position that Ubuntu has been moving towards for several years now, as a review of Shuttleworth's blog reveals .

A few years ago, Shuttleworth was attempting to get the larger free software projects to coordinate release schedules -- a move that would have greatly aided commercial distributions like Ubuntu, but done little for most projects.

When that effort was greeted with only mild interest, Shuttleworth next attempted to rally free software around usability issues. That call to common interest aroused interest, but GNOME in particular moved more slowly and in different directions than Ubuntu. Consequently, Ubuntu began its own development, modifying such desktop elements as notifications and title bar buttons on its own, and introducing centralized indicators for social networking and sound.

These modifications of GNOME left Ubuntu in an anomalous position. With the release of GNOME 3.0, Ubuntu would have been faced with a choice of abandoning or overhauling its interface code, or of continuing to develop an offshoot of GNOME 2.0 -- either of which would have been embarrassing and harmful to its reputation. Now, by focusing on Unity, Ubuntu can announce its own new desktop at about the same time that GNOME 3.0 is released.

Just as importantly, with Unity, Ubuntu has control of its significant code. Not that Unity is not free software -- it is. But, practically speaking, its Ubuntu volunteers and Canonical employees are apt to be the majority its contributors. Both are more likely to accept Canonical's directives than outsiders. In this way, Ubuntu can continue to express its allegiance to free software while not having to manoeuver through the usual give and take found in the community. In effect, it can now make decisions more like a proprietary company can.

Waiting for the dice to fall

How successful Unity will become remains to be seen. Given Ubuntu's popularity, probably it cannot help but succeed to a large extent. That inevitability, more than anything else, may make the risks associated with Unity acceptable. Eight months from now, we will probably see distributions derived from Ubuntu happily using Unity, rather than the GNOME 2 or 3 series of releases.

Yet if Unity is unlikely to fail completely, it may not completely succeed, either. Instructions about how to switch from Unity to standard GNOME are sure to be widely available, and we may never have a sense of how many people use them. Perhaps, too, using Unity as default will increase the popularity of Kubuntu and Xubuntu, Ubuntu's in-family alternatives.

But these, too, are part of the risk: that, despite its name, Unity simply becomes another choice among the distinctly un-unified GNOME-based desktops available.

No matter how you interpret the gamble, it's not a small one, and seeing how it plays out promises to make 2011 an interesting year.

Comments