informa
/
4 min read
Commentary

Linux Yesterday And Linux Tomorrow, But Never Linux Today

What would you say if I told you that plans to deploy Linux in corporations aren't accelerating, but are tapering off? That's what one study suggests. When asked if they were planning to deploy Linux in the current year, more than 90% of CIOs said no. That's up from 87% of the past year, and 60% of the year before that. So is Linux finally reaching its saturation point?

What would you say if I told you that plans to deploy Linux in corporations aren't accelerating, but are tapering off? That's what one study suggests. When asked if they were planning to deploy Linux in the current year, more than 90% of CIOs said no. That's up from 87% of the past year, and 60% of the year before that. So is Linux finally reaching its saturation point?

My colleague John Soat has his own take on the issue, and wonders aloud what the big hang-up is. But I don't think this is a sign that Linux has been hung out to dry -- if only because, for starters, one survey does not a future for a product make. But there are other things that come to mind, because the Linux of today is not the Linux of even a month ago, and it's definitely not the Linux of tomorrow. That's something inherent to the way Linux is -- its plasticity.

For one, the tapering-off probably reflects saturation with Linux as it currently is, in the distributions and versions we've come to know well over the last couple of years. It doesn't say anything about what Linux could turn into in two years, or five, or 10. By that time, Linux could be completely unrecognizable, and in a good way. I'm not just talking about what GUI or other add-ons might be supplied with a distribution, but how the Linux kernel is being used as the starting point for other things which don't bear much of a resemblance to the Linux that most folks might know.

If people say no to Linux right now, that doesn't really say much, because in a year or two they could be most enthusiastically saying yes to something that's Linux-based and doesn't remind them of anything from Novell or Red Hat. Case in point: When Linux is used as the basis for an appliance or a handheld device, rather than a desktop or even server OS, the Linux-ness of the whole thing is essentially invisible.

Keep in mind, this sort of mutation also has happened with Windows itself, and the Mac as well. Neither bears much resemblance to its earlier incarnations, either visually or architecturally, but the changes have been almost entirely for the better. "Linux" (which one?) is a term that is as plastic as "Windows" (which one?), if not more so. That might explain why in the same survey, "most" (their word) of the respondents plan to deploy Windows Vista sometime in 2008, after Vista SP1 is out.

To be scrupulously fair, there are no details about to what extent they're going to deploy Vista, just that they're going to do it in some form. I should also say that Linux's plasticity is often used as a criticism of its very nature, and sometimes a very deserving one: When you have a hundred distributions to choose from, too much choice is sometimes worse than none at all.

Still, when people say "We have no plans to deploy Linux," ask yourself: Which Linux, and in what form? I say this only because Linux has a funny way of sneaking up on people when they least expect it.

(Footnote. One other thing that is clear from the survey is something people have known for a long time: garden-variety Unix (barring maybe Solaris) is falling by the wayside in favor of Linux. I imagine this has been accelerating all the more after SCO's graphic humiliation in the courts by IBM ... which closely follows the humiliation in the marketplace they took for years before that. I guess one beating wasn't enough?)