The Penguin's Place

In my last post, I said that I expected Apple to have at least 10% market share by January 2008, and possibly 30% or more by 2011. Barring any paradigm-shattering advancements due to Kurzweil's nasty little Law of Accelerating Returns, I stand by that prediction. Obviously, to gain that kind of market share they're going to have to earn it from the current leader, Microsoft Windows, and I think the coming year will show they've got what it takes to do that.

However, someone asked me where Linux fits into this market share figure. I'm a Linux user, and love the little operating system that could, but I just don't think it has what it takes to break 10%-15% desktop market share in the next 5 years.

Why? Because Linux comes in too many flavors.

For example, on Skype's download page, there are options for Linux, OS X, and Windows. Click OS X, you get one simple download. Click Windows, you get one simple download. Click Linux, you get--well damn, you get six different selections. The average user doesn't even know what version of their operating system they're running, much less whether they need the dynamic or the static binary.

Or what the hell a binary is. There's too much confusion, too much fracturing. Desktop distributions like Ubuntu will ameliorate this problem, but not enough. The whole point of Linux is choice. Choice of technologies, of architectures, of platforms. This is a strength on the server, but it's hell on the desktop. Unless the entire Linux community commits to stringent standards on software installation, package management, and a big fat "et cetera," Linux will flat-out fail to deliver a compelling desktop experience.

And if they do deliver, they will have done Linux a disservice by creating a monolithic "official Linux" environment. It is in Linux's best interest to not succeed on the desktop, because of the inherent qualities of a successful desktop OS.

Computers are becoming more and more a required commodity to get our work done. For computers to properly serve that purpose, the interfaces need to get out of the way, and allow us to do what we need. For the non-technical person, environments like OS X and Windows (even though Windows sucks, no zealotry intended) will be the platform of choice for whom computing is a means to an end, not the end itself.

On the desktop, Linux is an end, not a means. Programmers will use it, and so will moderately technical people who have special needs. But that won't ever equate to higher market share, especially since most programmers make software for non-Linux platforms, and thus use said non-Linux platforms.