Here's how Linux gets to the desktop

Nearly 200 years ago Lewis and Clark arrived in Portland, Oregon on their cross-country expedition to explore the West. It was a journey that captured the spirit of a young and inspired New World that thrived on trailblazing new frontiers. Attending the O'Reilly Open Source Convention (OSCON) in Portland gave me the sense that the open-source movement is opening a new frontier, and at the same time shedding some of its more idealistic trappings.

The new open-source frontier is now moving toward the front lines of the enterprise. We are in an era of network-centric computing, and Microsoft's hegemony and pricing structures are being seriously questioned. As I wrote earlier, the well articulated benefits of open standards and open-source computing are pushing IT organisations to consider Linux, Apache, and other alternatives to proprietary back-end platforms. That trend has been widely documented, promoted and practiced by the likes of IBM, HP, and others. IT managers are figuring out on their own that Linux, MySQL, Jboss, and other open-source distributions can deliver real value.

On the other hand, desktop Linux environments -- such as KDE, Gnome, Mozilla and OpenOffice -- have been far slower to blaze a trail into the enterprise. You could point to Microsoft's dominance on the desktop as a deterrent, but that's not the only barrier.

In a keynote at OSCON yesterday, Mitch Kapor of the Open Software Applications Foundation (OSAF), offered his predictions for the future of desktop Linux and introduced a market study that assesses the current market for an alternative desktop to Windows.

Kapor and the OSAF certainly have an agenda to promote Linux on the desktop, but the report provides an honest appraisal. Kapor said that the immediate challenge for desktop Linux is to prove itself among transactional workers -- such as those working in call centres and help desks. The study gave 2004 as the year in which desktop Linux would advance significantly on that frontier. General purpose, knowledge workers, would not jump on the desktop Linux bandwagon until 2007 at the earliest.

Kapor said he would not be surprised to see 10 percent of global desktops running Linux in the near future. That's a good bet.

He cautioned that gaining desktop Linux users beyond the techie crowd would require much improved Microsoft Office compatibility as well as fit and finish in the desktop environments and applications. OpenOffice, for example, still lacks the polish necessary to convert the mass of Office users, despite the cost advantages. Kapor said the OSAF is funding extensive testing for Microsoft Excel compatibility to help on that front.

Kapor also said that the hardware abstraction layer and integration among various open-source desktop environments should get the attention of the developer community. Installing Linux on a laptop can be a nightmare, and device driver support remains inadequate.

The lack of a broad base of applications hampers the desktop Linux effort. However, Kapor doesn't believe that Linux will be adopted because of a killer app. "Like the (Linux) server, it starts at the edge with less-than-mission-critical applications and moves toward the centre." Nonetheless, the OSAF is hoping to catch a wave with its an email/personal information manager, code-named Chandler. The software, due for a 1.0 release in 2004, will be free and support Windows, Macintosh and Linux platforms.

Kapor offered a few observations based on the current trends. Microsoft will cut prices in response to open source, and the public sector and geographies outside the US will lead in the adoption of Linux. We have already seen Microsoft cutting pricing to gain deals with governments outside the US, and the city of Munich, Germany recently standardised on desktop Linux. Kapor said the non-US governments were more cost conscious and willing to mandate software standards. The OSAF announced a new Web site -- opensector.org -- to assist government agency decision makers.

Kapor issued report cards for desktop Linux hardware and software, and the various communities involved in the ecosystem. (View the report cards on this page.) For hardware and software, Windows connectivity received the highest grade and peripheral device support the worst. The desktop development platform also needs some significant improvements. In general, the lack of consistency evidenced in the report card is a hindrance for desktop Linux reaching into new markets.

Among the ecosystem communities, the platforms (e.g., KDE, Gnome and Mozilla) received high marks along with the Linux distributions, consortia, and the developer community. Where desktop Linux has problems is among the independent software developers (ISVs), original equipment manufacturers (OEMs) and the distribution channel.

Based on the OSAF report, it's clear that desktop Linux has made some good progress, but has a long way to go to dislodge Windows or surpass the quality of the Macintosh environment. It will take a focused group of individuals, a modern day Corps of Discovery, to harness the community energy to deliver a Linux desktop that does more than change the economics of computing; that greatly advances the software's usability.