Desktop Linux: By Rikki McGinty
Are we there yet?
Chicago, 4/20: The second and final day of Comdex's Linux Global Summit opened yet again with show organizers having underestimated attendee interest in the open source operating system. The flock of penguin-loving geeks had to be herded from the scheduled 75-seat basement conference room to a larger, more prominent room, where various members of the nonprofit advocacy group Linux International presided over a lively discussion on the direction of Linux. While the spirit of the summit was overwhelmingly celebratory and self-congratulating, the panelists all agreed that Linux had a long way to go. As Larry Augustin, CEO of leading Linux systems vendor VA Research, pointed out, "Linux is not appropriate for everyone everywhere."
Not your mother's OS While a whopping 26 percent of Internet service providers use Linux servers, for example, and countless corporate MIS departments run at least part of their operations on Linux (often unbeknownst to their department heads), the nascent operating system has yet to make inroads on the desktop. And though the various Linux-in-a-box retailers are working on enabling a kinder, gentler install, the OS is unlikely to gain real end-user market share until "I can give a Linux CD to my mom, and she can use it," said Jeff Carr of LinuxPowerPC.
The other hurdle in short-term Linux adoption is the lack of desktop applications and cross-platform software support. Users are understandably reluctant to run Linux - no matter how customizable and reliable - if this means compatibility problems with platforms or programs used by colleagues and clients. Some solutions have been emerging to address this issue, such as the StarOffice and Applixware office suite, which resemble and are compatible with MS Office 97.
Beware the hype Considering the relative rawness of Linux for the consumer or professional desktop, it's perhaps surprising that the OS has received such immense mainstream attention. And while Linux developers and users may be expected to rejoice in the spotlight, it is making them nervous. "We're not ready yet," said Dave Sifry, CTO of fledgling support service LinuxCare. "There's too much hype. Linux is great for supercomputing, Web servers, Beowulf clustering - but not for the desktop." During his Monday presentation, Linus Torvalds also confessed to being concerned about "the hype." "Linux is more than hype," he said. And in an interesting twist he added, "Linux does work despite the hype."
It wouldn't be the first time that a technology was considered a failure merely because the expectations generated by press attention couldn't be met at the time. Moreover, the motivations behind hyping a technology are often not consistent with those pushing the developers themselves. As Linux Journal senior editor Doc Searls pointed out, mainstream and trade press are likely more interested in stories than facts. Referring to preponderance of headlines over the "browser wars" between Microsoft and Netscape, Searls - whose pet project is the Cluetrain -- said "markets are not wars. Markets are conversations." In these conversations, he explained, there should be an exchange of ideas and a recognition of what one party has to offer the other, what one can learn from another.
Innovation, not duplication And in fact, the Linux community seeks to "innovate, not duplicate," according to S.u.S.E. president Marc Torres. Developing for Linux should involve creating new and better functionality and usability, not emulating existing systems in an attempt to beat competitors at their own game.
With a global team of developers who consider programming a form of art and do it for love, not money, the field is wide open for new and creative improvements in desktop computing. "We've created a business model where you're rewarded for making [the product] better," summarized VA Research's Augustin. "Linux holds the most potential for the next generation of systems." ø
Rikki McGinty is the editor of open source IT.