In January, Minnesota turns into a winter wonderland. “Why would anyone want to live in a land like this?” we wonder, gazing out our windows at the pools of liquid nitrogen dotting the landscape.

Under these cryogenic conditions, some Minnesotans head north to go camping (honest!). Those of us who, less hardy or more sensible, view chess as an ideal winter sport spend our Januaries stoking the fire, pondering existential mysteries, or predicting the future: It will be cold again tomorrow.

Before making other new forecasts, let’s update old ones, starting with a prediction made way back on August 12th, 1996. Some excerpts:

“You can’t ignore the Web, and so, probably for the first time, you have to start thinking about serving your company’s [Real Paying] Customers. That will change everything.” “… You’re going to start working in marketing time … in months. Sometimes weeks. That means a whole different way of designing and building systems.” “… you’re going to justify your existence based on how well you help the company attract new customers, retain the customers it has, and encourage every customer to do more business with you.” “From here on in, you’re face to face with real customers. And that really does change everything.”

Not bad for 1996. I’ll claim victory. But never mind that. The riskier forecast, first made in January, 1998, scheduled Microsoft’s implosion for 2000. I’ll claim a provisional victory here as well.

Microsoft’s two big revenue generators, Windows and Office, have performed lackadaisically. Windows 2000 has taken the marketplace by drizzle, not storm; likewise Office 2000. Windows Me, the current end-of-the-DOS/Windows product, is a ride-along on new PCs — upgrade purchasers are rare as spotted owls.

Was it Microsoft or the economy? 2000 was tougher than 1999 for selling anything. On the other hand, 2000 was supposed to be the year of new investment in IT, after 1999’s laser focus on Y2K remediation. As pointed out here last February, though, spending was on new business applications, not more infrastructure costs, and especially not optional ones like upgrades like Windows 2000 and Office 2000.

Microsoft’s financial performance was poor. Superficially, Microsoft’s 2000 growth was a solid if disappointing 21% (1999’s figure was 73%). Factor out investment income, though, and growth was a paltry 11%. And Microsoft has issued earnings warnings for its next-quarter results.

Will next year be any better? Almost certainly, Dubyah will either cripple the antitrust suit or kill it outright. The DoJ will be MIA. (Related prediction: No new antitrust suits of any kind unless a Bush campaign contributor benefits, and free reign on mergers and acquisitions unless a Bush crony is put at risk.)

The bad news? It won’t help. Microsoft already owns the desktop. Whistler will end up like Cairo, Windows Me-Too will happen but will be an even less interesting upgrade than its predecessor, and the next version of Office will be boring. There’s no source for growth. The data center won’t help: Microsoft has squandered its credibility while facing intense competition from Linux, Sun, Oracle, and IBM. The only gains will come from the ever-hapless Novell.

In a bizarre twist, Microsoft’s only hope for major growth is the PocketPC, thanks to Palm having grown so complacent that no platform innovation is even envisioned. The PocketPC isn’t enough to rescue 2001, though.

It looks like a tough year for the Redmond contingent.

It’s 2001. We’re supposed to have a self-supporting space station, a permanent outpost on the moon, and a psychotic computer running a manned mission to the moons of Jupiter. Had we not lost interest, it might have happened. Instead, our interest in cyberspace exceeds our interest in outer space. Am I the only one who thinks something is fundamentally wrong?

Maybe not. According to research performed by an organization called Virtual Society, reported in The Register (www.register.com), there are 28 million ex-web surfers in the United States alone — actual reality appears to be making a comeback.

Quite a few pundits have expounded on the ultimate importance of the Internet, some ascribing to it a level of significance unparalleled since the invention of the printing press. Last year, I challenged that position as being highly exaggerated. It appears I’m not alone.

Yes, with the advent of a new year, it’s time to review old predictions and make new ones. Let’s start with the review:

Prediction: Rather than the gratitude they deserve, the hard-working programmers who prevented a Y2K disaster will mostly receive unemployment.

Result: Pretty much as predicted. Ignorant, self-appointed voices of outrage declaimed the whole Y2K situation as a hoax. Meanwhile the staffing strategy known as “rolling layoffs” — firing programmers with “old” skills while simultaneously recruiting others with needed new ones — has increased in popularity.

Prediction: PCs won’t be replaced by fat network technologies. They also won’t gain much in new functionality, nor will the user interface improve dramatically. They will, however, increase in reliability.

Result: While PC sales are off this year, that’s probably an indicator of economic jitters coupled with market saturation, not abandonment of the platform. Meanwhile, Windows 2000 is “an order of magnitude” more reliable than NT (translation: It has one more to go before it’s as reliable as its competitors). And Apple is, finally, nearly ready to release Mac OS X, which, due to its BSD UNIX core, will be far more stable than prior Mac OS releases.

Prediction: Java won’t turn into the dominant language for all application development. It will find a niche as the mid-tier language of choice in n-tier OO/client/server development projects, and will continue to be used to extend browser functionality. Its performance deficiencies compared to compiled languages will continue to constrain its value.

Result: On target thus far. Java is popular in the mid-tier, nobody has yet released a significant commercial application written in pure Java, and Java continues to be slower than compiled C++. Microsoft’s impending (translation: indefinite release date) of C# and HP’s jump onto the Microsoft.NET bandwagon may further impede Java’s progress. (On the other hand, I’ve predicted failure for the Microsoft.NET scheme — but it doesn’t have to succeed to hamper Java in the marketplace.)

Prediction: Within two years, Java will either be turned over to a standards body or Sun’s Java allies will start to de-emphasize it.

Result: Neither has happened yet. One year to go.

Prediction: A long-shot, that Macromedia will turn its Flash/Shockwave/Director combination into an application development platform.

Result: I have seen an actual application (a sales force automation package) written with Macromedia tools. If nothing else, I’m not wrong yet.

Prediction: Linux will be very successful in as a server technology; far less successful on the desktop.

Result: Linux adoption as a server platform is growing faster than Windows 2000. Linux’s desktop marketshare is still miniscule.

We’ll continue reviewing past predictions next week.