Visit Your Local PBS Station PBS Home PBS Home Programs A-Z TV Schedules Watch Video Donate Shop PBS Search PBS
I, Cringely - The Survival of the Nerdiest with Robert X. Cringely
Search I,Cringely:

The Pulpit
The Pulpit

<< [ I'm From the Government, Trust Me ]   |  Y2K: The Winter of Our Disconnect?  |   [ Everybody's a Critic ] >>

Weekly Column

Y2K: The Winter of Our Disconnect?: Not If We Stay Calm and Don't Do Anything Stupid

Status: [CLOSED]
By Robert X. Cringely
bob@cringely.com

About once a year in this space I get to shamelessly, promote my own work. Next week, PBS will air my one-hour special, "Y2K: The Winter of Our Disconnect." In most places, it will be shown Tuesday night. For those who can't see the show or don't want to, here's my take on what to expect come January and why. It's written not for nerds but for normal folks. It was written, literally, for my Mom.

After 20 years of ignoring the problem, five years of trying to fix it, and two depressing and alarmist segments on "60 Minutes" (with a third to come), we are down to the wire on Y2K, the so-called millennium bug. In these last months and weeks of 1999, it is too late to do anything more about Y2K than to make our personal preparations. How will Y2K affect our lives and our fortunes, our economy and our public safety, our stock portfolios and our children? The media coverage to date has for the most part offered no help at all on any of these questions, preferring instead to express fears and to document the hoarding of canned goods. We deserve better.

For the three people in America who still don't know about Y2K, the term refers to a problem that many computers will have in determining dates after the new year. The goof goes back to the 1950s, when, in order to save computer punch-card holes, programmers left off the numerals "1" and "9" and 1953 became just "53." The obvious problem comes in 2000, which many computers may interpret as 1900. Uncorrected, this ambiguity will mess with the minds of computers at the phone company and the bank, and we may find ourselves owing (or receiving) 99 years worth of back interest. Worse still, there is the fear that any computer that pays attention to dates will go berserk and simply freeze in some logical paradox worthy of an old "Star Trek" episode.

How could this have happened? The people who made the mistakes those many years ago never expected their programs to still be running 50 years later. They underestimated both our innate laziness and the extent to which we would make ourselves dependent on automation. They assumed, too, that programs would get easier to write over time and that their poor efforts, no matter how well they worked, would be quickly superseded. Didn't happen. Software is hard to write, good software almost impossibly so, and when it didn't break, we didn't fix it. Alas, we also didn't save any of the original manuals or programmer comments, so much of the problem with Y2K is that our economy is dependent on programs we no longer even understand.

What's the likely scenario for January 2000? The short version is that Y2K will probably not be a big deal for most of us. Extremist expectations exist on both ends, but what's likely is a normal curve with a few really bad events, a few really good events, and a lot of medium-sized events.

Fortunately, there are plenty of things not to worry about, especially clock chips in devices. Most clock chips don't care about the date at all but rather are concerned with "how long since the last event happened?" Traffic lights (a common focus of anxiety) don't care about dates, just days of the week, and should be unaffected. Elevator controllers might care about the time of day, but generally not the date, except for reporting capabilities. And their failure modes — at least for dates — generally allow them to keep running.

The important thing to remember here is that physical devices are designed by engineers, not programmers. So these physical devices are designed to cope safely with failure, unlike computer programs, which just bomb. Engineers know that things will fail and that failures can hurt people. Engineers try to anticipate problems and prevent them. Programmers do not. They just re-boot the computer. This is a vital philosophical difference.

Let's talk about severity and duration. We could have a ton of things go wrong — millions of things — but have it all over in a single day. Would that be a problem? Not really. But what if we have a few thousand things that bug us for years? Now, that's a problem.

The real and enduring problems will be with the world financial system and in big software. In the case of big software — mission-critical applications that have been running for years on mainframe computers — the applications are old. They have been tweaked by many hands and complexified over the years until nobody at all understands anything. Software is the key: The greater the extent to which something has software, the more dangerous it is.

Accounting systems will be especially vulnerable, because they are dependent on time stamps. This is going to be a mess — but one we can deal with. In the short term, we'll probably make crude arrangements: "Let me pay you what I paid you the same time last year (or last month), and we'll sort out the details later." Sure, people will take advantage of this, but are we going to let the world grind to a halt because we're off plus or minus 10 percent? No. We'll move on. Y2K might actually make us talk to each other.

Now, just because an application breaks doesn't necessarily mean it is a problem. Failure is not, in itself, a problem. In each case of failure, we have to differentiate between productive activities and the accounting return loop through which we pay for them. It's the money that might get screwed up, not the delivery. AT&T sends out 80 million invoices per month. Could AT&T keep the phone system running even if it wasn't being paid? Of course it could — and would. The electrical grid has no date dependence. Your power is unlikely to be turned off.

If these predictions sound bad, take some comfort in knowing that the United States will be among the least affected parts of the world. Western financial markets and most big businesses are likely to greet the new year just fine. The problems that do happen are likely to be localized or to be in other parts of the globe, specifically South America, Africa, Eastern Europe, and the Indian subcontinent. But since this is a global economy in which financial jitters in Brazil and Russia can easily knock a few hundred points off the Dow, we should expect that there will be some impact.

Certainly some individual companies are likely to lose their luster for a while because of Y2K, specifically the outfits that have done so well from preparing for it. U.S. companies have accelerated their purchases of PCs as one response to Y2K, so Dell, Compaq, and the other big suppliers to corporate America are likely to see orders drop in the new year. So, too, computer-services outfits like IBM and Electronic Data Systems may see their business drop for a while. If Y2K problems are localized in the third world, there will also likely be trouble for companies that have large exposures in those regions.

What matters is how we respond to these troubles. Now we are in the emotional part of Y2K, the really dangerous territory. There is no reason at all that Y2K has to have any significant effect on the world economy, unless we expect it to. If everyone expects a depression and pulls money from the market, we'll have a depression. If, instead, we view Y2K as a buying opportunity, it will be business as usual. Fortunately, recent opinion polls show that most consumers and investors aren't particularly concerned about Y2K. Let's hope it stays that way.

Then who the heck are those Y2K extremists, the ones who are stockpiling canned goods? Most of them are religious fundamentalists and survivalists. The former seem quite certain that God is on the decimal system. Until very recently, Rev. Jerry Falwell was making lots of money selling Y2K video sermons predicting an apocalyptic January. Meanwhile, the survivalists have generally been looking for excuses to hide from the world. They are gratefully leaping on this excuse to build cabins and purchase generators. Their fear, interestingly enough, isn't Y2K itself, but rather the reaction to Y2K of the rest of us, the unprepared. The survivalists choose to be apart because they expect us to want their stuff and to come for it with weapons. I am not making this up.

Here is an important insight about both these groups. I have met with many of these people and what becomes clear is that the outcome they say they fear is really the outcome they want. The religious right seems to want some kind of cleansing of our society and see Y2K as part of that. The survivalists just want to be right. And both groups have done a very good job of making many perfectly normal people nervous enough about Y2K to consider taking some not very well-advised actions like cashing-in their savings and stocks. This will only get worse as the year winds down.

The rest of us just need to do a few simple things: get your financial records in order. That's it. No heading for the hills, withdrawing your savings, or stockpiling food. There will be some problems, but probably nothing more than you'd get from a bad storm. Sleep in. Savor the new century.

Comments from the Tribe

Status: [CLOSED] read all comments (0)