The so-called Millennium Bug: Is it really a threat?

By Zatni Arbi

JAKARTA (JP): My three-year-old Pentium 90 MHz PC has been sitting here on the desk right behind me for quite some time.

I haven't had a chance to re-install the programs I used to run on it since the last time its hard disk was refomatted. So it is the perfect test bed for the so-called Millennium Bug. It is surely an appropriate sacrifice for the digital demon.

I turn the PC on, and when the CMOS screen appears I hit the Del key. I change the system calendar to December 31, 1999, and I set the system clock to 23:57:20. Then I save the new CMOS setting, sit back and watch as Windows 95 loads. Within three minutes, according to many computer experts, something will happen. A disaster. A calamity. A devastating catastrophe that will bring the entire universe down and send our planet spiraling in bits and pieces into nothingness.

Or, according to a more realistic scenario, just as we toast the turn of the millennium, the lights will go out. On the streets, cars will crash into each other because they've suddenly gone bananas. The computers inside their engines will get confused and refuse to function. In the midst of the ensuing chaos, the elevators will be stuck, and in the sky man-made communications satellites will collide. Explosions will occur in nuclear power plants, and nuclear weapons will be launched without warning. They are all operated by computers.

Are these two realistic scenarios for what we're going to experience in exactly two years and 177 days? Or are they just a hype, a ploy by computer experts and business entrepreneurs who want to capitalize on public fear?

Are experts and entrepreneurs trying to scare the public and create hysteria? Wouldn't they benefit tremendously from the resulting panic? If not, are these concerns well-founded?

Today, less than three years before the predicted doomsday, we are beginning to read warnings in the media about these disasters. The intensity of the news and public discussion is escalating, and even skeptical people like myself are forced to look into the matter.

Oh, I almost forgot. The clock has continued ticking. According to this PC, it's now already 00:04 of January 1, 2000.

Nothing has happened to my PC.

The hard disk is still spinning, Windows 95 is still running. I install the latest WordPerfect Suite that I have just received from Corel Corp, then I run WordPerfect 8.0. I click on the "Insert" menu, and then I click on "Date and time". Lo and behold, WP 8.0 automatically types in "January 1, 2000", for me.

But, three points should be kept in mind here. Firstly, although it is, at least, three years old, this PC of mine belongs to a newer generation. As Jeffrey Sukardi, senior manager of BCA's Technology Planning Bureau, reminded me, older generations of PCs -- the 8088s, 80286s, 80386s and 80486s -- may be unable to handle the year 2000. In experiments such as this one, they will revert back to the year 1980. And a lot of them are still in use in this country today.

Secondly, new software programs, such as WordPerfect Suite 8.0 and Microsoft Office 97, are already equipped with program routines that either converts the date 00 to 2000 or read the entire four digits of the year from the system calendar. Applications that were written many, many years ago are not capable of doing it.

Thirdly, my PC is just one tiny speck in the computer macrocosm. When these experts remind us that our life is now almost totally dependent on computers, they are right on the mark. My test PC is a lonely machine with no responsibility. Millions of other computers, on the contrary, may contain data on sick patients, flight schedules, security codes and control programs for nuclear reactors. If my PC crashes as the system clock shows the digits 00 for the year 2000, only my own life and nobody else's will be affected. On the other hand, if the systems in banks, insurance companies or government agencies crash, the lives of so many people may never be the same again.

What has to be done, as Doug Hall from Australia told me, is that computers with critical applications that serve critical functions around the world should be tested to ensure that they are Millennium-proof.

Hall belongs to one of the companies that can provide the testing service, but his group is certainly not the only one. The Millennium Bug, according to the Gartner Group, a U.S.-based IT research organization, will create US$600 billion of business worldwide. It should not be a surprise if throngs of computer experts and entrepreneurs have jumped to cash in on this tremendous opportunity.

Hall says that in some test runs an exact copy of the actual system will be created on an independent machine. After the copy is verified to work in precisely the same way as the original, the tester will adjust the computer clock and calendar just like I did. They will then assess the impact as the computer reads the last two digits 00 of its system calendar. Problem spots will be identified, and adjustments will be made.

Actually, the culprit is the application systems. Because data storage space was so expensive back in the early days of computerization, programmers had to crimp on the amount of data their programs could handle. That was why they chose to use only the last two digits of the calendar to indicate the year. So, for example, the date "July 3, 1996" would be recorded as "7/03/96", whereas "November 23, 1926" would be recorded as "11/23/26". When the first day of the year 2000 arrives, the system calendar will read "01/01/00". These application programs, particularly the ones written in COBOL or the first generation of database management system (DBMS), will be unable to determine whether these two digits mean 2000 or 1900 -- or even the year 0, for that matter. Confusion will reign, and computations that involve dates will be affected. Retirement plans, savings interest and similar applications may be turned upside down and the loss could amount to trillions of rupiah.

Software

In answer to my question, Hall explained that there are already off-the-shelf software patches that can be used to fix certain applications, but no real panacea is available so far. Therefore, large and critical systems still have to be tested and fixed individually.

Luckily, many of our private institutions have been quietly addressing the problem. At Bank BCA, for instance, engineers started looking for and fixing the problem two years ago. In my telephone conversation with him, Jeffrey told me that they had formed a team to work with IBM to identify every potential trap in their systems. The scope of the job is impressive, though. They even checked that their modems were 2000-compliant!

Unfortunately, as Jeffrey confided to me, and as many of you also know, not everybody is taking this potential time bomb seriously enough. In many computerized institutions, and particularly in government-owned enterprises, not enough is being done to make sure their applications will not crash on Day One of the new millennium. The same is true with a lot of private enterprises, including banks, manufactures and service-based industries.

Folks, we should not let ourselves become overwhelmed by the panic that unscrupulous entrepreneurs are trying to create. Yet we do have to start spreading awareness of this problem and have the responsible parties do something about it. Some of the scenarios we've been reading may be exaggerated, but nonetheless we should begin a concerted effort to ask those who offer public services to have their systems thoroughly tested and made completely Year 2000-ready. We should require them to make sure that no disaster will happen to their computers -- and our lives -- after Dec. 31, 1999. Simply because the potential is there.