It's been bugging me for weeks, ever since I read the announcement from LinuxWorld Conference and Expo that Microsoft Corp. is coming to the show. No, the announcement didn't mean stealth attendance by solo Microsoft employees or contractors. The show has had those in the past. It meant Microsoft will have a booth. My questions are what do they hope to accomplish and how will they do it? What would motivate Microsoft to participate in an event celebrating a cancer-causing, communist-inspired, anti-American miscreant of an operating system called Linux?

I queried Microsoft's crack PR firm's "Rapid Response Team" immediately after reading the announcement for answers to those questions. I didn't even receive an acknowledgement to the questions. A day or two later I read an obviously Microsoft-approved news account with an official and entirely forgettable explanation. Something about showing off embedded XP. Not exactly what you would expect to see at a Linux show.

The questions took on a life of their own. They grew from their humble beginning that focused only on why Microsoft would venture deep into enemy territory to participate in a Linux trade show to a larger scenario: What is Microsoft's long-term strategy for dealing with Linux and free/open source software?

Here are the short versions of answers I concocted. Microsoft wants what it has always wanted: power and money. They hope to get it the same way they have always gotten it, by extending and embracing the technology. Remember, Microsoft has never been satisfied to simply make a profit. Its management wants to control the markets that they play in.

The longer answers are speculative. Take them for the workings of an overactive imagination fueled by an overheated speculator. Each part requires a leap of faith. I have no hard evidence to reveal, and only suspicions based on many years of watching Microsoft.

Waking up with Microsoft on my mind

I woke up one morning to the realization that Microsoft was going to announce a port of Microsoft Office to Linux. Now that would really be an icebreaker at LWCE. Such a port has been the subject of speculation for some time, so much so that Microsoft has actually denied it, thus giving the notion even more credibility.

The office suite is definitely the sweet spot for Linux applications these days. Everyone and their sister are offering one. Perhaps because it is the No. 1 factor slowing the migration of desktops from Windows to Linux. Hancom, Gobe, K-Office, StarOffice, the list goes on. Forgive me for not itemizing them all, but you get the idea. This is the Office Age for Linux. If there is one part of the Linux application market that is likely to start showing real profits any time soon, it is this sector.

Such a move could put Microsoft in a position to grab some of the Linux pie. However, that's not a real big pie in terms of dollars. I doubt that a slice of it would ever capture much attention in Redmond. They have bigger fish to fry.

The port of Microsoft Office to Macintosh's (BSD-based) Mac OS X has given Microsoft a big leg up on a port to Linux. It wouldn't be the degree of difficult that holds Microsoft back from taking a seat at the table. Nevertheless, Microsoft certainly wouldn't want to encourage that migration in any case.

To see why Microsoft would bother to do such a port requires you to look a move or two ahead. The lack of an immediate, substantial payback and making it easier for customers to move off Windows may not be the overriding concerns.

Microsoft has always done best when it has cloned the efforts of others. If Microsoft's engineers have a "cut along the dotted lines" template to base their work on, they don't do half bad. From the Windows GUI to compression algorithms to browsers, if Microsoft has done well with a technology, it was not its invention. Microsoft "borrowed" the look and feel of Apple Computer Inc.'s GUI when it created Windows 3. I believe Microsoft is now returning to the scene of the crime. As Apple has done with BSD, so Microsoft will do with Linux.

But unlike Apple, I am not talking about Microsoft using the Linux kernel for Windows. I am talking about Microsoft making a drop-in GUI replacement for X Window System. A proprietary GUI. A closed GUI. A GUI with clout. An extend and embrace GUI. Get the picture?

Even among the Linux faithful, many are unhappy with the storied and hallowed X Window system. Some believe it is the primary reason Linux has not grown as quickly on the desktop as it has in the server room. If Microsoft were to develop a GUI to sit atop the Linux kernel it would be a very big deal in more ways than one.

For one thing, it would be a big seller. There are many people sitting on the fence, wanting to try Linux on their desktops but lack the nerve. With Microsoft in the picture, they would get off the fence. Suddenly, lots of dollars could be made. The smell of money would attract more developers to write or port applications to Linux.

What's the problem with that? New users, new opportunities to make money. It's all good, right? Well, not exactly. Microsoft would control a choke point on the Linux application market. Microsoft has a great fondness for choke points. That's why their domination on the desktop is so strong and so complete.

Microsoft Office and a Microsoft GUI for Linux would be good short-term, but extremely dangerous for the long-term viability of the free/open source software movements. Microsoft would gain tremendous influence over Linux and perhaps be able to play the distributions against each other in exactly the same way they have played Dell against Compaq and the other OEMs. On the other hand, perhaps there will simply be a Microsoft Linux distribution.

My advice to you, if you are attending LWCE next month, is to be wary of Microsoft. No matter how friendly they might seem, no matter what "gifts" they might bring, remember who they are and how they operate. Remember Kerberos, and how smoothly they mingled with that community for a couple of years. Right up to the day they went back on their promises of an open implementation of Kerberos in Windows.