It can take five years or more for a good wine to age properly. A good operating system can take even longer.
This week, while the industry watches Microsoft officially unveil Windows XP, a relative newcomer to the OS market announced a product that could allow enterprise companies a means to circumvent the Windows juggernaut entirely. With nothing but a press release on its Web site, Lindows.com said it is previewing the initial release of its LindowsOS in the fourth quarter. We’re talking about a platform for running Windows applications on Linux.
This is when open source gets interesting. If it delivers on its promises, Lindows would give IT managers the best of both worlds — Windows-style computing at less than US$100. It would also be the culmination of a fascinating exploration of how far code can be cloned without actually ripping off Microsoft.
This project, called Wine, began about eight years ago as a collective effort to implement Windows 3.x and Win32 APIs on the open-standard X Window System and Unix. Written in C language, Wine developers have been rewriting raw source code. In doing so they avoid the emulation route, whereby x86 instructions are translated piece by piece, often slowing down the system’s performance (the name, supposedly, means “Wine Is Not an Emulator, but as an acronym that doesn’t really make sense, does it?).
In theory. Wine would bring open-source Windows to OS/2 and the Mac OS as well, but the excitement and awareness around Linux over the last five years has made it the focal point of the industry’s interest. Canada’s own Corel Corp., for example, was among the most high-profile software firms that were reportedly contributing their programming talents to Wine, back in the days when former Corel chairman Michael Cowpland thought he could bring Linux to the desktop.
Corel Linux, however, met with little success and to a large degree Wine had fallen off the radar — its Web site’s press clippings section more or less stops in 2000. There was a possibility that the Wine project would benefit from the initial verdict against Microsoft when Judge Thomas Penfield Jackson recommended the company release APIs to competitors, but the current appeals process makes that scenario unlikely.
Bringing LindowsOS to market raises the stakes for Wine and similar development efforts considerably. As a quasi-academic project, it was OK for the results to be somewhat scattershot. Some Wine-based versions of Adobe’s Acrobat Reader, for example, didn’t render documents very well. Other basic programs, like Microsoft Word or AutoCAD, didn’t live up to the originals. But once it moves beyond the experimental stages to an actual product, customers will have the same expectations of Lindows.com that they would of Microsoft, or even the various Linux distributions like Red Hat Software and Caldera. It there are many glitches, customers may not have the patience to wait for the product to improve.
The recent licensing pressures Microsoft has put on customers through its Software Assurance program has clearly left some corporate users considering other OSes. A company like Lindows.com would offer a cost-effective, but high-risk, strategy. If it is successful, Lindows could substantially broaden the Linux customer base, but it will need a substantial marketing campaign and some courageous customers who will validate it. The response from the Linux community is also hard to predict. LindowsOS erodes a bit of Linux’s renegade image, and many of those who believe in the platform have some serious issues with Microsoft.
As Linux seeks a foothold beyond the server market, Lindows.com is a pioneer that will serve as the test case for many other open source businesses that follow. It represents our first real taste of Wine, and it had better be easy to swallow.