Rant Mode Equals One: The Numbers Don't Add Up

Is it More Expensive to Continue Funding the Monopoly? Yes!

Is it impractical for the government to even consider the break-up of Microsoft? Yes, according to Stan Liebowitz, an economist who claims that the cost of such a break-up would likely be in the order of $30,000,000,000. To wit:

The study, authored by Professor Stan Liebowitz of the University of Texas at Dallas and presented today by the Association for Competitive Technology (ACT), specifically addresses recent papers that claim there would be no costs associated with a Windows breakup. Professor Liebowitz' paper uses real-world, fact-based examples from the high-tech marketplace to support his conclusion that breaking Windows into competing versions will result in consumers seeing fewer choices and higher prices for software.

Gee, isn't that special. I think I'd like to ask Dr. Liebowitz to define the words "fact-based" and "real-world" for me. Maybe it will help explain the holes in his arguments. Let's look closely: more competition means fewer choices? Last I checked adding one to another number results in a higher number, not a lower one. When there are more choices for a product, the prices usually drop, not rise.

But this is just the beginning of the clueless fuzzy-logic train:

"People are living in a fool's paradise if they think that the Windows standard will remain intact if the product is divided among multiple companies," Dr. Liebowitz said. "Just look at what has happened with the UNIX and Linux operating systems where competing versions are nowhere near compatible. This example provides absolute and directly relevant evidence of what we would see in a fragmented Windows marketplace," he added. The professor's findings were debated today at a conference that explored remedy options in the Department of Justice's antitrust case against Microsoft.

First, there's the fragmentation thing. I've stated before that Linux is not fragmented. To even jokingly point at it and say that it's got the same diseases that proprietary Unix suffered a long time ago is utterly baseless, outstripping reason, logic, facts or even speculation. All evidence to date shows that, if anything, Linux is helping UNIX become more unified.

Just look at the recent announcements by SCO to open-source their UNIX code. Since the fragmentation battles of the UNIX era, things have gotten downright cozy in the standards area. There exist literally hundreds of open-source products that compile across many open-source Linux and BSD distributions, as well as proprietary UNIX platforms. Some of these products even work on Microsoft Windows platforms.

It's real innovation, and the fact that over half of the Web is powered by these innovative products tells us that it's a good thing.

The word "standards" really, truly applies here. It isn't being used loosely, as in its application above: "Windows standard." Where's that "standard" written? Who can obtain a copy? Can a third-party developer of Windows applications know that anything will remain usable from one release of Windows to the next? No; hence, it is not a real "standard."

If we want to point to things that truly are becoming more fragmented, what about Windows itself? Windows 2000's API (Application Programming Interface) is different from Windows NT's, which is different from Windows 98's and even Windows CE's. There's speculation as well that the Windows 64-bit API will also be another (truly) fragmented mess.

So what we have as "reason" from this economist is inverse logic. He is wrong about Linux, but it's worse than that. The very disease that he diagnoses in proprietary UNIX--and wrongly in Linux as well--is actually being harbored by the subject he's trying to defend: Windows.

And somebody needs to clue this guy in on what a standard really is. Windows, at most, is a de facto standard in the loosest sense. It's a standard that was imposed, not agreed upon. It's a standard that has just recently changed, thanks to Microsoft's release of yet another incompatible version of Windows--Windows 2000.

I wonder what the cost to businesses is going to be now that they get to repurchase Windows technology, rejigged as Windows 2000, a product with only a few more features than its predecessor, and yet unable to run large quantities of current, off-the-shelf software. The features that are new aren't ones that Microsoft innovated, they're ones that they were obliged to include, under threat of watching their market be taken by rivals who were already delivering that very technology.