Sometimes it's predictable. Sometimes it's not. And sometimes it's both at the same time.

For instance, ten years ago (I'm writing this column in October, 1999) on October 17, 1989, at approximately 5:04pm PDT, at latitude 37 degrees, 02 minutes north, longitude 120 degrees, 53 minutes west, the North American tectonic plate decided to remain stationary while the Pacific plate suffered an oblique-reverse slippage, at a depth of 11.5 miles. In other words, the San Andreas fault let go during the opening minutes of the third game of the World Series in a seismic event that measured 6.9 on the Richter scale, killed 67 people, caused over $10 billion in damage -- and raised the height of the Santa Cruz Mountains by about half an inch.

The exact timing of that quake was entirely unpredictable. Despite the best efforts of geologists and soothsayers alike, earthquake forecasting is still an infant art at best. However, in 1988, the US Geological Survey established the probability of a major seismic event along the San Andreas fault at 30 percent or greater from 1989 through 2019. So, while the precise day and hour of the next large quake couldn't be predicted, that a big one would occur was a foregone conclusion.

By now, it's hard to imagine that there's an adult human being anywhere in the Western world who can't give chapter and verse on the so-called Millennium Bug. It's been so endlessly hyped and dissected by the media that we're all sick to death of it. It's even safe to say that it's the single most-prognosticated disaster in all of history. And, unlike Loma Prieta, we've known for years exactly when it would strike -- right down to the hour, minute and second.

Normally, I devote my January column to predicting the trends and events of the coming year. Sometimes I'm right on the money and sometimes I'm so far off the mark that I'm tempted to chuck my Magic 8 Ball right in the trash.

This year, I'm going to crawl out on a different limb of the same tree -- and you'll be able to tell right away whether I sawed it off behind me. Right here, before your very eyes, I'm going to predict what happened, (and, just as importantly, what didn't happen,) when the calendar at long last rolled over to Y2K.

New Year's Day

So, what didn't happen? First off, the global banking system didn't collapse. That was always the silliest and least credible prophesy of the whole odious crew of professional Millennium Bug doomsayers -- and, yes, I'm talking about "Dr." Gary North here.

The number one goal of the banking industry is to make money -- and they're very focused on that priority. I used to work for one of the biggest banks in America and I know that from personal experience. Allowing a mere software glitch to interfere with that goal was never in the cards. And even if the banking industry had been dumb enough to allow their Y2K issues to remain unaddressed, the Feds (and their global compatriots) weren't about to permit it.

What the hell. I figure all that cash people hoarded probably came in handy for New Year's Eve partying, so the effort wasn't entirely wasted. It was a pretty darned amazing global party, too -- it's going to be awfully hard to top it next year, when the real millennium starts.

The electrical power grid stayed up, too -- at least in Europe and North America it did. As it turns out, contrary to Larry Yourdon's augury, electrical power systems aren't actually all that date-sensitive. So the lights stayed on when the calendar flipped over -- and so did TVS and cell phones and microwave ovens.

That doesn't mean the power companies won't screw up their customers' bills, mind you. The computers that control the grid itself are mostly Unix boxen, while the ones that run the billing and accounting software are still principally mainframes. So don't be surprised if you get a bill for a century's worth of electricity and gas at the end of the month.

Just be grateful you won't have to write the check in the dark. Wearing mittens.

Of course, we're going to have to wait until later in the month before we discover just how many of our service vendors still haven't addressed Y2K problems with their billing and accounting systems. I feel safe in predicting that utility, telephone and credit card companies will turn out to have them under control. After all, they have to turn a profit.

Uncle Sam's Blues

I continue to be less sanguine about government agencies.

It's not that I expected to have to watch out for planes falling out of the sky. I didn't -- at least not in here in North America or in Europe or most of the Pacific Rim. I wasn't particularly worried about an accidental nuclear attack, either -- there are too many fail-safe systems in place for that to happen.

I'm a little concerned about whether my friends in the Friends of the Library will get their Social Security and Medicare checks on time. I know President Clinton has stated that the Social Security Administration is fully Y2K-compliant, and the SSA's own Web site makes the same claim.

I just don't buy their reassurances.

For one thing, the SSA declared itself Y2K-compliant in 1998 by the simple expedient of shifting responsibility for solving its Millennium Bug problems upstairs to the Commerce Department. In my book, ducking a problem doesn't count as a solution to that problem.

So, we'll see.

And, even if the Feds have their ducks in a row -- a dubious proposition at best -- there are still 50 States and a half-dozen Territories and Protectorates that have yet to prove they've got their respective acts together. That's nearly five dozen distinct sets of welfare systems, DMVs, tax authorities and other bureaucracies, none of which is exactly run by geniuses.

I wouldn't bet my worthless Turkish 500 lire piece on any of 'em.

High Anxiety

I also think the gloom-and-doom forecasts of a global stock market meltdown have it completely backwards. Instead of crashing, when the markets open for business on January 3, I think traders will be so relieved about what a non-event the Y2K threat turned out to be that they're going to send indexes all over the world straight through the ceiling.

That exuberance may not last, but speculators who go bottom-fishing during the last week of December are liable to catch their limit of big profits the first week in January. And I expect tech and banking stocks to do particularly well, precisely because those are the ones that are most likely to get hammered as 1999 draws to a close.

In fact, I think computer vendors and their suppliers are liable to have a banner first half, as all the small businesses that discover how much worse than the Y2K rollover issue the February 29 problem turns out to be rush to replace their non-leap-year-compliant systems. Likewise, I'm dead certain that tort lawyers all over the United States are going to have a field day pressing suits on behalf of the small businesses that get bitten by the February 29 bug, too.

Luckily, my friend Mark Perkel, developer of utility software and job control languages for network environments, has put together a legal brief for companies defending against Y2K lawsuits. In it, he makes the case that, for legal purposes, Y2K should be treated as an Act of God.

Mark isn't a lawyer, but he is a programmer who has turned himself into an expert on the law -- especially as it applies to software development and liability issues. His brief is designed to both educate and persuade judges and juries, so it's written in clear, relatively simple and non-technical language.

And it's shareware.

Finally, I think the Internet itself will weather the crisis without any noticeable damage. Some autonomous systems in third world countries may fall off temporarily, but I doubt even that will happen.

It's a self-healing system and most of the traffic flows over routers that run on Unix and Unix-like operating systems. Those have their own impending date with destiny -- but it's not until 2038, when, I think it's safe to assume, they'll have long since been retired and replaced or upgraded to OS versions that don't suffer from the Epoch Bug.

After all, unlike the earthquake that interrupted the third game of the Bay Bridge World Series, we can see it coming from a long way off. Like the Y2K problem, we know the exact second at which it will occur. And, if we're smart, we'll fix it long before 2038 rolls around.