Visit Your Local PBS Station PBS Home PBS Home Programs A-Z TV Schedules Watch Video Donate Shop PBS Search PBS
Y2K: The Winter of Our Disconnect?

 

[next] Home

BOB: What do you know about Y2K? The mysterious date problem that may confuse some computers when their internal clocks strike the year 2000. What do you think will happen? In the face of uncertainty, people have questions about what America is facing. We set out to ask some ourselves.

BOB: Hi, I'm Bob Cringely. Maybe, like me, you get frustrated by some of the advantages of modern life made possible by computers. Take new cars, for example. This one probably can't be fixed here. We'll have to tow it into town. It's yet another symptom of what created the Y2K crisis -- technology run amok.

ACT ONE -- WHAT IS GOING TO HAPPEN IN JANUARY? SHOULD WE BE WORRIED?

LEN the DRIVER: I can't fix these things in the field like I used to. All I can do basically is tow you into town.

BOB: There's no doubt; the computer has revolutionized the world. But for all it benefits, this wondrous new technology has some rough edges.

BOB: All the new cars are like this. You plug them into a computer to find out what's wrong with em. And then you have to have some way to tell what this computer tells you.

BOB: Where many people see computers as mankind's greatest invention, others, like Forrest, see them in a different light. It's not surprising that people are distrustful of what Y2K may have in store.

BOB: How's this going to affect you?

FORREST: My plan is to retire and get out of this business.

(Bob OUTSIDE THE SHOP)

Bob: My best guess about what's really likely to happen with Y2K is significantly different from what's been shown already on television. There are extremists on both sides, but what's really likely to happen is more like a normal curve with a few very bad things, a few very good things and lots of medium-sized things in the middle. Most people don't really even understand the problem.

(Bob WANDERS ABOUT TOWN)

BOB: Here I was in St. Helens, Oregon waiting for a part for my car. What, I wondered must the people here think about Y2K? The media has been hyping it so big for so long, surely, in this small town, I could come in touch with what Americans really feel about it.

(Bob ENTERS CITY HALL)

BOB: There's been a lot of talk about the problems communities face insuring that their basic services will continue to operate as needed. Were city officials here prepared for what might happen?

BRIAN LITTLE: There only two signals -- two stop lights in town -- a small town, but , ah, they're so old they're not reliant on any type of chip technology. It's 45 seconds for green one way, 15 seconds for green the other way so those systems work fine.

BOB: It sounded to me like the town was in good shape. They'd checked out and OK'd their water system and the sewage plant and they'd met with the companies who supplied the city's other vital services to insure they were compliant as well.

BRIAN LITTLE: The Red Cross probably had the best advice. It's plan for this like you would plan for any other potential disaster. People need to be prepared for that. You need to be prepared for 72 hours of being without basic necessities, so if you have your kits at home and everything like that and something unforeseen should happen, should carry you through.

BOB: Its time for a Cringely Crash Course in computer history.

BOB: How did we get here? What brought us to this confusion? Well, the simple answer is to blame it on a lack of computer memory. In the early days of computing, it made technical and financial sense to store data information in a shorthand form.

Until the late 1960s, mainframe computers were built with memory cards like this one that were expensive and didn't have much capacity.

This is a memory card from 1953. It hold 1K -- 1,024 bits of information. Cost about a dollar a bit. That's a thousand 1952 dollars. So if you were computerizing a warehouse or a bank, the fewer bits you could use to store a part number or a date, the better.

BOB: If we dig still deeper, beyond the technical explanations, to find out how we really got here we could probably start with a name: GRACE HOPPER, inventor of the COBOL software language and coiner of the term "bug." Grace, who has been called, the Mother of the Computer, was a Navy Officer who traveled the world lecturing about her favorite topic: the FUTURE of computing.

BOB: What Grace and her cohorts were guilty of is not a particular line of code but a particular code of behavior. They built a system in the 1950s that pretty much exists today. This system is to blame for Y2K.

BOB: The mainframes of their day weren't built for only two digits; they accepted the software that was written for them -- in COBOL. And where that programming language erred was in leaving the choice of integers up to the programmers. And it was THEY who made the decision to go with two digits. As a consequence, today, when "99" turns to "00" many computers MAY think the year is "1900."

Compounding the problem created by what programmers call "memory constrained environments" was the business practice that developed in the computer industry whereby each new software upgrade was kept compatible with the old code.

While this may have resulted in happy customers, it also led to the preservation of the two-digit year.

RICH SEIDNER: The problem is that, you know, code goes away, but the programs don't. The programs live forever it seems, you know. The backward compatibility is one of the keys of the technology industry. You can't -- a program -- if a product stops working every couple of years and the customer has to replace it, that's kind of -- that's ridiculous and expensive. So that means that all the old programs, once they're running and doing really good work, they just keep running. And they don't -- they don't ever go away.

DALE WAY: People didn't believe the software was going to last so long. They didn't understand the longevity of software; of how once you make an investment of, in some cases, millions of dollars, into something, and then once you've built a database and data structures that it relates to, it's very hard to throw that away. And as soon as you start looking at the data, you realize that the interconnections between the data are the complex things. And this is how Y2K came into the picture. Because if you change the way you represent a year in a database, that has impacts all over to all the software that talks to the data, that accesses the data, that subsequently processes that data and hands it on to someone else.

Most people fixate on the clock issues because computers have clocks...But data and databases shared by systems all over the world's a much more abstract and difficult concept for people to intuitively grasp.

(Bob in front of St. Helen's CITY HALL)

BOB: Before you get crazy wondering how we got into this mess, keep in mind that there are a lot of thing not to worry about: like clock chips in devices. This town's two traffic signals, 911 system, sewer and water delivery.

These are basic services that have provoked some of the most radical fears. But their failure modes when it comes to dates are not year dependent. Nothing is going to happen. And even if it did, once we're past the date window and the systems are reset, they should work fine from then on.

BOB: The important thing to remember is that these physical devices are designed by engineers and not programmers. Engineers know that things will fail and that failures can hurt people, so they try to anticipate problems and prevent them. Programmers generally do not. This is a VITAL philosophical difference.

BOB: The big difference between engineers and programmers is that engineers work on physical systems that have to follow the laws of science while programmers are building their programs inside the memory or processor of the computer.

(Bob beside the TRAIN)

BOB: programmer can design a locomotive that's the size of a butterfly or goes the speed of light, and when that locomotive hits the wall, explodes, he just shuts the computer down and turns it back on. When a real locomotive hits the wall, it's very different.

(PILOT and CO-PILOT IN THE COCKPIT of a PASSENGER PLANE [The Magnetic Monster ] CLIP)

INTERCOM: Flight 17, return to the airfield at once, your engines might quit at any moment.

PILOT: Quit at any moment...what does that mean?

(The engines quit)

BOB: Speaking of big hardware: If there's one thing that's really focused public attention on Y2K, it's the question of how it might affect the safety of air travel. Plane manufacturers and the Federal Aviation Administration have issued assurances that airplane safety and air traffic control systems are completely OK and there won't be problems.

(Bob in the pilot's chair of a 777 FLIGHT SIMULATOR, cruising at 20,000 feet)

BOB: We decided to check this out for ourselves so we went to Boeing the airplane company and they assured us that Y2K would not affect their aircraft's performance. To prove it, they had devised a test that rolled the clocks on their planes forward to fool them into thinking it was New Year's Eve, 1999. And they let us go along to see for ourselves.

(The aircraft's CLOCK RUNS PAST 00 INTO THE NEW YEAR and NOTHING HAPPENS so BOB DOES A "VICTORY ROLL.")

BOB: Do you think the passengers noticed?

INSTRUCTOR: I don't think so. That was pretty smooth.

BOB: So we've heard all the words of reassurance, airplanes won't fall out of the sky, the air traffic control system will continue to work, GPS satellites will continue to orbit. But what if they're wrong? What if December 31st it all just stops working? Well, then pilots will have to look out the window to figure out where they are. Doesn't sound so bad to me.

BOB: 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. A few thousand things that bug us for years, though! Now that's a problem.

(Bob outside the NEW YORK STOCK EXCHANGE)

BOB: The real and enduring problems are likely to be with the world financial system and in big software. In the case of big software -- mission-critical applications at outfits like here at the New York Stock Exchange have been running for years on mainframe computers sometimes long after they are no longer even needed.

JOHN PANCHERY: This happens in every firm. You receive a report in, every morning, and it's called XYZ trade blotter. The report goes on someone's desk. You say to that person, "Uh, what's that report for? Who gets it?" And the answer might be, "Well, I'm not sure, but I get it every day."

BOB: Some applications are old, they have been tweaked by so many hands over the years and complexified until nobody at all understands ANYTHING. Software is the key. The extent to which something has software the more dangerous it is.

DALE WAY: The core problem of Y2K, these missing digits in the century place, is -- is sort of the reason we have a problem. But the reason we have a crisis is because of the complexity of this software infrastructure that we have allowed to evolve more or less un -- unwatched.

These systems are too big to change all at once, and they're too complicated for us to change them in pieces. Every time we try to pull out a piece and put in a more modern piece, we've got blood vessels and nerves dangling. We just don't know enough to -- to graft them in smoothly. And so we keep trying and things don't work right. And it's a very difficult thing to modernize these and that's because we don't understand them at a very de -- at the level of detail necessary to be able to transform them in a meaningful way.

BILL JOY: One of the big ways that flaws get introduced into programs is people are manually managing who is using which word of memory, and because of mistakes, two parts of the program that didn't intend to use the same memory are - and now you've got like a wormhole.

So if you have a building and you have a thousand bricks, if one brick is defective, it doesn't make one of the other bricks fail. Ah. That...that would be a goofy concept. But in software if you have a bunch of say, C code and there's a bug in one part of the program, it can induce a flaw in an unrelated part of the program because the programs are coupled together. The wormhole is this accidentally shared bits of memory. And so that means I can never trust any part of my program.

(BOB in WASHINGTON on the CAPITAL GROUNDS)

BOB: And speaking of big software, what about the federal government? Well, 40 federal programs have already admitted that they won't be making their Y2K deadlines, including food stamps and Medicaid which won't be ready until December, leaving hardly any time for testing. And what about Social Security? Will those checks go out on time? What are the chances of real problems?

JOHN KOSKINEN: The federal govt. is going to be, I think, 100 percent ready. The OMB quarterly reports to the public and the Congress have shown that 92 percent of the mission critical systems of the govt met the President's March 31, '99, deadline, a deadline earlier than any other industry.

Nationally, we think that beyond the federal government the critical infrastructure systems are going to hold. There's no evidence we'll have any failure of the power grids. The telecommunications systems are going to work. The air transport and other transportation systems are going to work. The banking system actually has always been ahead of eveybody else. We think that there will not be a banking problem.

If you move into the second tier in terms of state, the states run for us ten major federal programs -- unemployment insurance, Medicaid, food stamps, uh, income maintenance programs...all of them say they're going to be done, but a number of them are going to be done late in the year. And what we're working with them is saying they have to now -- and we're in the process of doing that -- satisfy us that they have appropriate contingency and backup plans.

Some of it will be paper processing; there will be eligibility determinations made on the spot rather than electronically. It will be more cumbersome, less efficient, more costly for the states, perhaps to run, but we are going to make sure that in every one of those states that can't deliver on its responsibilities, that there's a backup plan so that the beneficiaries get the benefits to which they're entitled to.

ACT TWO -- WHY ARE PEOPLE AFRAID OF Y2K? WHY SHOULDN"T THEY BE?

(Bob surrounded by survival paraphernalia at an ARMY SURPLUS STORE)

BOB: Why are people so afraid of computers at the end of the century? Y2K mania is why. Survivalists have generally been looking for excuses to pitch their tents and hide from the world. They're gratefully leaping on Y2K and heading for the back country.

SCENE OF A FAMILY BUYING STAPLES FROM A SMALL STORE AND LOADING THEM INTO A TRAILER CLIP

[FATHER-to store owner]: We're going into the hills to rough it for awhile.

[OWNER]: For awhile? -- for the rest of your lives it looks like.

[FATHER]: Maybe.

BOB: So who are these Y2K extremists anyway -- the ones stockpiling canned goods and heading for the hills? Well, many of them are religious fundamentalists or survivalists...

(Bob at ARMY SURPLUS STORE)

BOB: And much of the literature is religious too. It's hardly surprising considering Y2K also ushers in the new millenium.

BOB: Michael Dowd, who was once a minister, now calls himself a community organizer and devotes much of his time to advising the public on Y2K issues from an urban perspective. He's stocked his basement with supplies to tide himself -- and his neighbors -- over should the worst occur.

BOB: So, Michael, this is your...your basement and Y2K headquarters.

MICHAEL DOWN: Yeah, Y2K central...

BOB: DOWD has enough food and supplies down here to last eight adults for about three months. He doesn't think the Y2K problem will persist quite that long, but the information he's gathered from literature and videos and the internet has him concerned enough that he's turned almost evangelical on the subject.

MICHAEL DOWD: I actually personally believe we're going to go into at least a five-year recession and it could even, ah, turn out to be a depression. There may only be a 30% chance. Maybe it's only a 10% chance.

It just makes sense in my mind to be prepared and help others be prepared for whatever level they can be prepared. Stock up what you can. I mean don't hoard, don't take more than is fair, but at this point, um, find ways of being assistance to others who are less...less fortunate or more vulnerable. Those who would be impacted more quickly and, ah...and, ah, more painfully if Y2K does turn out to be a big deal.

BOB: So I'm buying a new house. Is this a mistake?

MICHAEL: Well, I'm not a financial expert but...

BOB: Michael Dowd's point-of-view is mild compared to some others we talked with who see little safety in numbers. JOE, for instance, is a truck driver who fears Y2K is a government plot and he's thinking of trucking for the hills.

BOB: ....are you prepared for any eventuality?

JOE: Oh, yeah, because I have my mind right. I mean I think about it so it's not going to jump up and bite me, you know...

BOB: Where Michael Dowd is content to hunker down in the inner city and wait for things to blow over, Joe is preparing for the long haul and the possibility that Y2K may shut the economy down completely and that people will panic.

JOE: ...they're going to be rushing out with the money they just took out of the bank...to get food and...and other supplies which is going to cause a drain on those markets, which expands. Do you see what I'm saying? It's just a big ball that gets bigger and bigger and bigger.

And unfortunately, the media, um--all the national media--has hyped Y2K to such an extent that if nothing happens, nothing whatsoever, we're still going to have trouble. Because people perceive there will be a problem.

BOB: What if nothing happens?

JOE: Great. I'm hopin' nothing happens, but I'm an optimist. I hope nothin' happens. I...I believe it's going to be a ripple, but on the other hand, I'm kind of pragmatic, you know...So, from that standpoint, I prepare for the absolute worst.

BOB: Are ya packin'?

JOE: (laughs) Well, that... Yeah, sorta. I mean I'm not gonna... If it looks like it's going to break, I...I am not stayin' in town. That's for sure. I'm goin' away. Because, you know, the large metropolitan areas are going to be a mess if panic hits, you know.

(OUTSIDE, Joe shows Bob the SURVIVAL PACK he keeps in his truck.)

BOB: Joe's main concern is human nature; he's prepared but others aren't. And there's the rub. If Y2K sends things into a tailspin, other people may want what he has. So he has to stay light -- and mobile.

JOE: I've got three changes of clothes in here and they're not this kind of clothing. They're, ah, more of, ah,...Well, they're military dungarees type things...

(Bob WATCHES JOE DRIVE AWAY in his truck.)

JOE: I can keep all my good supplies in here. And...if something happens I'm in a place where I need to bug out right now, I'm not worried because I can survive on what's inside here.

(Bob walking in St Helens. He enters a COMPUTER STORE and chats with OWNER KEITH LOCKE)

BOB: I notice you've got your free Y2K check coupon here.

KEITH LOCKE: Uh-huh.

BOB: So Do people bring in a lot of these?

KEITH: There's been a few brought them in. That's kind of a service I provide. I just give them a free checkup and whatnot and they can see. And then if it's not, I can give them some options and tell them whether it's really important or not.

BOB: Do you find people who are coming in here really worried?

KEITH LOCKE: There are a few people, but it's like everything else, you know. I mean, you always have your extremists that are concerned about this or that, you know, and the world's going to end and all this kind of stuff.

BOB: So the world's not going to end?

KEITH LOCKE: I don't think so, but you never know (they laugh). Could end tomorrow.

(MOVIE CLIP: A MAN is subjected to an ATOMIC BLAST [The Amazing Colossal Man])

BOB: Much of what we hear from alarmists has to do with big infrastructure. Nuclear safety, for instance, is one hot button. Another involves the delivery of electricity and concerns for the stability of the power grid.

(ELECTRIC PLANT EXPLODES and a CITY GOES DARK [The Magnetic Monster])

(BOB at a HYDRO-ELECTRIC DAM)

BOB: This is a hydro-electric power plant where electricity is generated then sent on to consumers. I'm curious about the role computers play in what goes on here. Is there anything to be worried about?

(INSIDE THE PLANT)

BOB: Despite the size of this powerhouse and all it's sophistication, it turns out that the process of actually generating electricity doesn't involve the date at all.

When I finally located a computer what I found was a system that had been purchased in the 70s that was being used simply for data acquisition. And it could care less about the year 2000.

BOB: So, the infrastructure that delivers your electricity generally isn't date dependent -- except for the accounting system used to bill for it. In each case of failure we have to differentiate between productive activity and the accounting return loop used to pay for it. Generally we lose track of the money, not the delivery.

DALE WAY: We can look at systems as two basic types...large scale systems...those that are applied to production activities that have to do with assembly lines and delivering services. Things that have to do with the delivery of the product or service that the organizations involved in. And then we have sort of a backward flowing accounting systems that sort of handles the invoicing and the billing and the money flow back. And those two systems are -- usually have a distinct character.

The production systems are -- are usually developed with much more discipline and much more focused objectives. And those systems, I believe are going to be very solid. They're -- I don't think we're going to have a problem with power and I don't think we're going to have a power with phones -- a problem with phones.

The accounting systems because they're usually based on something much older -- because that was the first use of computers -- IBM came of age in accounting -- the accounting systems usually have an older core which has been much modified. And -- and the accounting systems have a great deal more complexity in them because their -- their objectives have changed more over time.

Those systems are not quite as coherent as the, as the production systems.

So those are where the real problems are most difficult to find and are probably going to be the most troublesome..

(MOVIE CLIP: A MILITARY COUNTDOWN AND LAUNCH [Day the Sky Exploded])

BOB: Many people are worried about how Y2K might affect computers that control our national defense.

(MOVIE CLIP: CROWDS running in panic [Day the Sky Exploded])

BOB: Others are afraid we might experience shortages of vital materials such as gas and oil products.

(NEWS ARCHIVE FOOTAGE: 1973 gas lines)

BOB: Is there a chance that Y2K problems in foreign countries could choke off our petroleum supplies and leave us stranded in gas lines as happened in 1973 when the country was forced to ration gasoline?

JIM SULLIVAN: We have looked at our supply chain critically. Much of our facilities in West Africa, for example, are more or less self-sufficient and they can operate pretty much independent of whatever the infrastructure problems may be.

As far as our transportation shipping fleet, we've done a very thorough job. We have 36, plus or minus, tankers we've inspected, we've tested. We've fixed anything that needed fixing. There were some minor fixes that needed attention. And I'm sure that our shipping company is going to be fully ready.

I think 99 percent of the public has confidence in our industry, has confidence in the utilities, has confidence in the infrastructure that we have, because we depend on it and it operates day after day after day, through earthquakes and hurricans and storms and so on and so forth.

We prepare for this type of thing on a regular basis. This got us a little more focused on computer programs and embedded systems and contingency planning, and, as I said, I think it helped us, because we've done away with a lot of obsolete computer programs that nobody was using. They had links all here and there and we've streamlined that. We've upgraded a lot of our control systems because it was a good opportunity to upgrade them, not just for Y2K, but to get them up to state of the art. It's been a good exercise.

BOB: And have you done any contingency planning for a possible comet impact

SULLIVAN: Comet?

(FILM CLIP: COMET zooms by and EXPLODES)

( Bob at a NUCLEAR PLANT)

BOB: There will continue to be questions about big infrastructure like this nuclear power plant. We've talked to lots of responsible peple who say that their industries are ready for Y2K; our power grid, our sources of oil and gas, our national defense are just fine. But we'll never know for sure until the date comes.

Until then, there's uncertainty. Uncertainty leads to fear and fear leads to consultants.

BOB: Concerns about Y2K are fueled by people like Andy Kyte, whose prediction, in 1997, that Y2K would cost $300-600 billion worldwide has become gospel. Where did he get that figure anyway and what's the real number? How much of this number is stuff people wanted to spend anyhow and now they get to throw it under the unquestionable Y2K category?

ANDY KYTE: A large amount of what organizations have had to spend on the year 2000 is fixing the drains, it's things that you have to do but you don't feel very good about. If they hadn't spent money fixing the drains then they could of done something to make themselves more competitive, more attractive to deliver better services to their customer, so the losers have actually been the efficiencies that were their to begins there has been an opportunity cost and it hasn't been new money. It has been a diversification of existing funds.

(Bob prowls ST HELENS and enters a THRIFT SHOP)

BOB: The most important thing to remember is that life will continue. In a Y2K crisis, your grocery will always accept your check and your bank will definitely pay the grocery. They don't want to lose you and thousands of others as their customers.

(Bob brouses the shop)

BOB: Take this thrift shop, for instance. The only thing that might affect business here, is if the credit card machine goes down but then Kay, who owns the place, doesn't do much of that business anyway. If there's a problem, she'll just tell customers to write a check or pay cash. Or come back later. Still, she's got a few concerns.

(Bob AT COUNTER with shop owner KAYE FRIDAY)

KAYE FRIDAY: I do worry a little bit about public power, public water and that kind of thing.

BOB: Really.

KAYE: If they're computer driven, whether we'll have that. But other than that, I'm not too worried about it.

(Bob exits the shop)

BOB: In the short term we'll probably do crude deals: '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. The world will move on.

BOB: Take AT&T. This is their network operations center. AT&T sends out 80 million invoices a month. Could they keep the phone system running even if they weren't being paid? Of course they could.

JOHN PASQUA: The basic telephony of the communication networks, to a great extent, is not affected by Y2K.

Dial tone will be available, calls will be initiated and completed successfully. However, the operational support systems, the OA -- the OAM systems, operational, administrative and maintenance systems, can be affected by Y2K and that's why we've addressed all of those systems.

The networks bend, but they don't break. If excessive congestion occurs, we have flow control mechanisms built into our network architecture that can cause some fast busies, but won't compromise the integrity of the network and servicing the vast majority of its users.

BOB: So I was hoping that you would you forget to send me my cable TV bill. That's unlikely?

JOHN PASQUA: That's not our intention.

(FILM CLIP: A DARK ALLEY LEADS to a MENACING SHADOW descending stairs into a BASEMENT)

BOB: As the institutions and companies that put computers in place in the 1950s grew and merged and bought each other out, so too, their computing systems grew and were interconnected.

And as management fell in love with each new latest and greatest technology, what did they do with the old non-sexy core stuff -- the application software that actually ran the business?

(BOB is revealed as he TURNS ON A LIGHT BULB)

BOB: They left it here. In the basement.

(ARCHIVE: ant colony)

BOB: Someone estimated that in the last fifty years there have been five million computer programmers; the unsung maintenance workers of the information technology industry. Most of them never even went to school to learn computers. They just learned it on the job. They were the ones who built this core software.

(Bob in BASEMENT)

BOB: These programmers were so busy with their individual tasks that none of them saw the whole picture. They could have used four digits for the date but didn't. How were they to know? Their bosses were supposed to make that decision, but didn't.

And now its 50 years later, and no one understands what those programmers actually did. But what they did is immense; trillions of lines of code stacked 50 years deep. And this core stuff -- this code -- is still changing, constantly, until it no longer resembles what it started out to be.

(FILM CLIP: A BLOB invades a bowling alley [The Blob])

BOB: Its an organic living thing.

RICH SEIDNER: By the time you reach even thousands of parts, let alone millions and tens of millions of parts -- which is what we now have in the computers and in the, um, in the programs we write and use -- these things can never be...the computational complexity is beyond many lifetimes.

Let me give you an example. Microsoft has an operating system commonly used in business, called Windows NT. And the new version coming out of Windows NT has roughly 25, 30 million lines of code. Now, I -- if a person, a programmer were to stay up 24 hours a day, reading a line of code every second...it would take an entire year just to simply write -- read the code 24 hours a day, 365 days let alone to understand the consequences of the interaction of all that code.

DALE WAY: How did these systems come to be? We -- the general public has the idea that they were somehow designed or some genius sat down and figured out how it was all going to go together and then built it and -- like a bridge -- like you would do a bridge.

But they didn't just design those systems. They designed a system to do a thing. And then, first of all, they found out it didn't do what they thought it was going to do so they had to fix it. Then they found out it didn't do a whole bunch of things they wanted it to do, so they had to add to it.

Then the situation -- the requirements change because the competitive environment changed, the regulations changed, and so they had to change it yet again. And this has been going on for 30, 40, 50 years in some places. So systems were not -- systems that run our core functions and core businesses were not really designed. They were more maintained into existence by these processes of constant modifications.

(BOB pushes a SHOPPING CART down the AISLE of a St Helen's MARKET)

BOB: You probably don't realize it but you're already being inconvenienced by computers, and have been for some time. Say you come here to the store to buy your favorite bleach -- and there isn't any. Why is that?

There's a good chance it was because of a computer. Perhaps the computer wasn't notified that the bleach was out of stock, or it wasn't running when it should have put the order in. These things happen all the time but we just don't notice.

ANDY KYTE: The year 2000 isn't occurring in a healthy population. Computer systems around the world have got errors in them. Everyone who deals with them knows this.

So what you've got is a population where computers have problems but they live with them. And then they get additional year 2000 problems. What the reality is that most of the can cope with the additional problems but there are going to be some who are going to suffer. What we are going to find is that the unhealthy ones in the population that get killed off.

BOB: This is Fort Detrick in Maryland, a level 4 containment facility where America keeps its most dangerous germs. There is stuff in here you don't even want to know about. Experts here work on ways to isolate, contain and eradicate these micro-organisms. Their goal is to prevent epidemics.

Y2K is like an epidemic too. Like HIV, it attacks the immune system and leaves the underlying organism.

ANDY KYTE: There are very few people who use computers daily on a daily or weekly basis who can claim that they have never experienced an error. Now, what year 2000 problems do is they cause an increase in the number of errors. But they don't put errors where none ever existed before.

Now what this means is that where organizations have had a requirement for systems which are very reliable which have very small numbers of errors, for instance, major banks and financial institutions like insurance companies, they already had excellent systems and they had the processes in place and the skills in place to be able to identify any remaining year 2000 problems and to seek to eliminate them.

But there are lots of organizations around the world that quite honestly live with very poor quality information technology systems. So they have lots of bugs everyday and they are going to get some year 2000 bugs in amongst them. And some of those bugs will hurt them. But it's wrong to think that the year 2000 bugs will represent a terminal condition. They don't -- they are not significantly different from other areas that affect computer systems on a day to day basis.

(Bob at FT DITRICK)

BOB: Some business function failures lead to business problems and organizational failures. Say one airline's reservation computer system goes down and that airline can't sell any tickets. But an organizational failure doesn't have to mean an industry failure. One airline's problem doesn't affect all airlines. And there are many points along the path where this can be stopped.

In this way, Y2K is exactly like an epidemic -- only in this problem it's goods and services that are affected and business processes that need to be quarantined.

DALE WAY: You don't exactly know where the origin of that problem is. It might not be right where you see it. It might have come from another system. It might have come from another sytem in another company, or in another country, if -- if you're part of a long international chain of information flow.

The closer you get to large, old, interconnected software, the longer duration the problems are likely to be, the harder they are to find, the more random they're going to be in terms of when they let themselves be seen. And that's why that part of the problem is going to go on for many months, if not years, after January 1st, 2000.

ANDY KYTE: Now organizations run the risk of small incremental changes -- number of errors happening one after another -- not realizing it and this is why the year 2000 can be so dangerous for some organizations. If they are waiting for a pyrotechnic event, if they are waiting for a spectacular sign in the sky that isn't what is going to hurt them. What is going to hurt them is going to be a death of a thousand cuts. It's the number of small things that go wrong.

(Bob in WASHINGTON)

BOB: No matter how confusing the situation, the failure of a computer application for any complex, crazy reason does not, in itself, create a problem. An accounting system may fail but that doesn't keep us from selling electricity or pumping gas. Computers break all the time yet we still rely on them - don't we?

ACT 3 -- WHO IS TO BLAME? -- WHY DIDN'T WE FIX THIS BEFORE?

(Bob in St Helens)

BOB: If this was a banking crisis we'd be looking for white collar criminals. If this was a health epidemic, we'd be looking for terrorists or at least people with dirty hands or dirty consciences. Yet for some reason, we don't seem to be looking for responsible parties for Y2K.

Most of the Y2K legislation, in fact, is intended to indemnify various groups and head off lawsuits. Is this American? The main reasons why we aren't looking for culprits is that most involved parties don't see it as being in their interest to assign blame. The media don't understand the situation well enough to even guess who is to blame.

BOB: The Y2K frenzy is the logical result of the information technology industry that spawned it: A giant game of promises, fear, uncertainty, and doubt backed by bad decisions.

BOB: Part of the problem is complexity. We have designed and built and become dependent on systems that are too complex even to understand. Part of the reason we did this is because the systems worked pretty well. But even more so its because we are lazy and it's easier to patch a system than to replace it.

(BOB wanders through a COMPUTER MUSEUM)

BOB: Software development hasn't evolved a wit since three decades ago when hard drives were as big as plows. Change in the information technology sector is glacial and it is often easier on the career of a CIO to not do something than to do something, especially since he might be retired before the real problem appears.

I read recently that the percentage of big IT projects that meet their deadlines and budgets and actually do what they are supposed to do is eight percent. That's a 92 percent failure rate. At least half of the big information technology projects are never finished at all. No wonder we'd rather fix than replace.

(Bob wanders WALL STREET)

BOB: The real cause of the Y2K problem was Wall Street and the lack of professionalism. Up through the 1970's, information technology led to productivity increases in business. It made businesses more competitive.

But, back then, the computers were simple, the applications were simple, the software was simple, and, most important, the big bosses never really saw those mainframe computers that were locked away in back rooms and in basements.

Then, in the 1980's, the personal computer introduced confusion.

(Bob on the steps of the FEDERAL BANK)

BOB: Computers on desktops became visible cost centers. Wall Street began to view information technology as a cost. Something to cut. Over time, all the people who understood the applications and could fix them were offered early retirement.

Now, this wasn't a problem immediately because if you don't change anything, computers can run by themselves for a very long time.

Information technology departments came to be judged on their size and budget, rather than on the value they contributed to the company. To stay visible to the big boss, they began embracing every new high tech fad, while rarely extending or updating their existing software investments.

(Bob in COMPUTER MUSEUM with an OLD PUNCHCARD MACHINE)

BOB: Now the end of the century is looming and it won't go away. We have to do something about this stuff, but what? That's where the Y2K industry is showing that it isn't any different than what came before.

Remediation -- that's fixing Y2K problems -- is being approached as a software maintenance issue: We find the bugs, we fix them, and then we test the fix. If more bugs are created, we fix those and start the testing cycle over again.

(Bob in COMPUTER MUSEUM)

BOB: What's wrong with that? Well, it's not the only way to solve the problem. Another thing we could do is just throw everythng away and start over. Here's a dirty little secret from the world of computers.

This is an IBM computer that's at least 45 years old and definitely not Y2K compliant. In fact, it's not even a computer in the sense that we know them today. There's no programming language. This is a program card. You write the programs by putting these jumper wires in these thousands of holes. The only way to make this Y2K compliant is to take the computer out of service. And that's what they did -- last year!

The reason we fix Y2K problems like we do is because of Grace Hopper's idea that the program is the system, that code is everything. The company that originally bought this IBM 1401 mainframe computer in the 1950s didn't do so to have a piece of iron humming in the back room. They wanted to run business applications written in the COBOL language. Taking any other approach to Y2K risks dethroning software from its position as the center of the IT universe.

IT values software. Users value data. Making software sacred forces us to solve Y2K in this horribly expensive and labor-intensive system that actually creates more problems than it solves.

RICH SEIDNER: One of the first things that we did was to start writing more programs to find the old problems in the old code...not seeing the irony in the fact that we're using the very technology -- nobody knows that these -- that these programs are perfect. Nobody knows what they're missing. Nobody in the world. Anybody who says that they know, they're a fool or they're a liar. Nobody can know this.

(Xerox commercial "Brother Dominic" Way)

I'm reminded of the great Xerox commercial ... where this cute little monk, Brother Dominic, approaches his superior with this wonderfully scribed page of manuscript. And he gives it to the senior monk and the monk looks at it and says, "This is very good, Brother Dominic. Now make me 500 more. And he gives it back to Dominic and Dominic goes (mouths "500 more") and he's just amazed. How's he going to do 500 more?

You can't do things that are vast in size and scope the same way you did them when you only had one to do. It requires a whole different set of tools and a whole different set of thought ... thought structures.

There's a number of different approaches that are -- that are so novel that the software gods are resistant to because it's not software centric.

(Bob walks through the storage array of a FEDERAL GOVT. MAINFRAME)

BOB: Is there a better solution? Yes. We simply tell these U.S. Treasury computers they are still living in 1972 where the days of the week and dates are all the same as in 2000 and where Y2K is 28 years away. We keep the mainframes in 1972 and we run all inputs and outputs through a PC to fix the dates.

This technique of course causes the system to run slower, so we'll need to upgrade hardware, not software, to get back the speed. Thus, we've swiftly solved our Y2K problems without having to even touch the old code.

Its a simple solution but hardly ever used. Why? Because the software experts say its inelegant. What they really mean is, it takes power away from the gods of software. The folks who caused Y2K in the first place.

DALE WAY: My sense is that most of the really important systems will work most of the time. Uh, it will be disruptions. It will -- the computer will be down a lot more often than we're used to , although everyone is familiar with, "the computer is down," "I can't process your ticket," or, "I can't, uh, handle your luggage right now." But there'll be more of that.

And what that will result in is fundamentally a slow down of how things happen. The velocity of the movement of informa -- information will -- will, in net, slow down.

JOHN KOSKINEN: There will be some glitches, as I say -- some people relying on particular developing countries that may have difficulties. Some international trade around the edges that may create problems for individual companies or even individual areas, if you're depending upon a particular, uh, set of countries in southeast Asia that have difficulties. But as a general matter -- and we've worked at this with the Council of Economic Advisers and the National Economic Commission, we don't see any indication that there's going to be any noticeable impact on the economy generally, but that doesn't mean that in someplace you're going to call -- won't call a company and find out that their holding pattern is longer because they've got some -- some slowdown. But we don't think that's going to be an economic factor for people to have to deal with.

(Bob back at the St Helens REPAIR SHOP)

BOB: So where does this put us? What can we expect from Y2K? The best I can figure it is, the big stuff is going to be generally OK. Airplanes won't drop out of the sky, our gasoline supply won't be cut off, the power grid will stay energized, our nuclear plants will stay safe, and national defense won't be compromised.

The big question is, how are the agencies and suppliers and sub-contractors connected to all those big outfits going to handle it all?

ANDY KYTE: The reality is that nobody can fix the year 2000 problem as a whole but any individual instance of the year 2000 problem can be fixed. Remember that these computer systems are being managed everyday. They have errors in them and they are being fixed everyday, so nobody can say I can fix any year 2000 problem. That would simply be ridiculous. But for every computer system in the world there is somebody who can fix the year 2000 problems on that particular system. And those people exist because they have been doing the work.

But, in general, what we are going to see is an overall reduction in efficiency lasting perhaps a few weeks. But the reduction in efficiency might only be two or three percent reduction in efficiency there are huge tracks of the world where people simply won't know anything has happened in regards to year 2000. There will be small pockets where they're badly effected.

Think of it like a weather system. Remember 1998. If I said to you it was a bad winter you'd say, "well, I don't know." If you lived in Montreal you'd know. They had terrible ice storms, they lost power, it was a very difficult winter. Well the year 2000 is going to be a bit like that. Some people are going to say that nothing much happened. And other people are going to say well lit was pretty tough you know but we survived it. But those people are very much going to be in the minority.

RICH SEIDNER: The best lesson we can learn from this is that in the future, we're putting all of our lives and all of our governmental and insitutional systems increasingly dependant on a variety of technologies that we don't understand what they can do and how they can go wrong. And so that's the lesson, which is that understand we're increasing the risk.

(Bob slides behind the wheel)

BOB: My only specific Y2K recommendation: If you are accustomed to buying your essentials (groceries, gas, medications) with a credit card, bring your checkbook to the store, just in case. The real danger here isn't technology but our emotional reaction to uncertainy.

(ARCHIVE: FDR surrounded by crowd)

BOB: We've been here before.

(ARCHIVE: FDR Speech FDR "Let me assert my firm belief that the only thing we have to fear is fear itself. Nameless, unreasoning, unjustified terror.")

( Bob in his CAR)

BOB: Don't give in to fear. Use Y2K as an opportunity to bring some discipline in your life. In the end, I guess we'll all have to just buckle down and deal with it. Just buckle down and enjoy the new millenium.

[next] Home


UPDATE | Millennial Mania | Truth or Dire Consequences
Double Digit Debacle | Survivalists and Social Security
Home | Radio Library | Y2K, Cringely

Copyright 1999 Oregon Public Broadcasting and PBS Online

Home PBS Online Home