The Once and Never King

The Linux desktop market is about eight years behind schedule

We are about eight years behind schedule.


Around the millennial epoch I helped SuSE whelp the SuSE Linux Enterprise Desktop (SLED), a product that is till on the Novell price list. Growing frustration with Microsoft viruses, continued vendor lock-in and the inherent lower cost of an Open Source desktop were hailed as the beginning of the end of Microsoft’s dominance.

Yeah. I didn’t believe it either.

SuSE arguably had and still has the best alternative to a Microsoft desktop, and considering that it is not Vista, perhaps the best desktop available. But like many good technology solutions it never took the market by storm despite continued frustration with Vista, continued vendor lock-in and the inherent lower cost of an Open Source. Market dynamics and realities stopped SLED’s march.

“The reality is that (the Linux desktop) is a slow, gradual, unstoppable growth,” said my buddy Jeremy White at CodeWeavers, a crew that knows more about Linux desktops than Torvalds. “Emphasis on ‘slow’. Growth is steady, but measured in a fraction of a percent and mainly overseas.”

The reason Linux desktops have never been a big hit with businesses is because no segment of the market sees enough advantage. Technically challenged small businesses have enough to deal with, and can’t even migrate away from Windows 98.

Midsized firms have the technical skills required, but the cost savings do not justify it. With discounts, the combined Microsoft operating system and common office applications cost about $200 a seat. If we assume a business with 500 employees and a four year hard/software life cycle, that is an annual expense of $25,000. This savings would be entirely wiped out by the switch costs and requisite hair plugs for the IT staffers who ripped out their own follicles during the migration effort.


Enterprises simply have too much inertia. They have invested heavily over the years in management infrastructure and IT expertise with Windows. Conversion to a Linux desktop might be profitable, but the scope of the project (not to mention technology territoriality) scares away IT staffs with urgent issues, like installing more network bandwidth so play Gears of War on company time is more enjoyable.

Whatever promise Linux desktops had in the early days were hampered by a premature rush to deploy. One former Novell product manager told me “There were a million dirty little secrets of failure.” Many of the big deployments heralded in the trade press never happened. He noted that lawyers panned early editions of Open Office for the simple lack of strike-through fonts. Forget retraining users when the basic tools were unusable.

But the Linux desktop market is growing, and will (eventually) create a significant dent in Microsoft’s territory … someday. In Asia and in developing countries, Linux is becoming the de facto OS for all computers including desktops. This includes foreign enterprises. Assume that their desktop administrators make $15,000 a year. It is cheaper to have their administrators and support staffs learn GIMP and train all employees than it is to buy copies of Photoshop. Disparity in economics drive differential rates of adoption.

This regionalized momentum will not directly affect industrialized countries, though over time we will have to deal with incoming documents in native Open Office formats. In industrialized countries Linux desktops will be found in the enterprise, but only on techie desks (but then again it was the techies who brought Linux servers into the enterprise, and we see where that led). There will be niche uses, including virtualization controllers and some thin clients.

Yet Linux adoption will grow mainly from consumers. And, yes, I did take my medication this morning.

The tech industry inverted in the new millennia. It used to be that computer technology was designed for business first, and that technology later filtered down to consumers. Now consumers are leading in many fields (laptops, cell phones, etc.) and innovations for consumers later reach the enterprise. How many enterprises are providing Phones or their equivalents to employees? How many will five years from now?


This is where netbooks and cell phones come into play. Devices people use the most set their expectations for usability. Though in their nascent stage, netbooks (which are primarily Linux based) are becoming the tool of choice for many households who now do everything on the web. They use Google Docs, Quicken Online and keep family photos on Flickr. The old concept of a laptop, much less a desktop, is gone. And Linux/Firefox has become the OS/UI with which they are most accustomed.

Then there are cell phones. Android is easing past its growing pains and ready for broader adoption. Palm is resurrecting itself with the Linux-based Pre. Motorola, HTC, Samsung all build Linux handsets. Android and the Pre show Linux can be cooler than Apple. People use their smart phones more than their netbooks. Thus usability levels of expectations are being set by smaller devices that are used more frequently, and over which Microsoft does not have the same leverage as with desktops.

I won’t predict that Linux will rule the end user space (and if I knew this for sure, I would keep it between me and my stock broker). But like life, Linux finds a way. It is the original computer virus, able to infiltrate every hardware platform. It’s only limitations appear to be the imagination of developers … who are smart enough not to tie ribbons around applications.