Y2k; Millennium Bug Essays - Calendars, Software Bugs, COBOL

Y2k; Millennium Bug Essays - Calendars, Software Bugs, COBOL Y2k; Millennium Bug The Millennial sun will first rise over human civilization in the independent republic of Kiribati, a group of some thirty low lying coral islands in the Pacific Ocean that straddle the equator and the International Date Line, halfway between Hawaii and Australia. This long awaited sunrise marks the dawn of the year 2000, and quite possibly, the onset of unheralded disruptions in life as we know it in many parts of the globe. Kiribatis 81,000 Micronesians may observe nothing different about this dawn; they only received TV in 1989. However, for those who live in a world that relies on satellites, air, rail and ground transportation, manufacturing plants, electricity, heat, telephones, or TV, when the calendar clicks from 99 to 00, we will experience a true millennial shift. As the sun moves westward on January 1, 2000, as the date shifts silently within millions of computerized systems, we will begin to experience our computer-dependent world in an entirely new way. We will finally see the extent of the networked and interdependent processes we have created. At the stroke of midnight, the new millennium heralds the greatest challenge to modern society that we have yet to face as a planetary community. I am describing the year 2000 problem, known as Y2K (K signifying 1000.) Nicknamed at first The Millennial Bug, increasing sensitivity to the magnitude of the impending crisis has escalated it to The Millennial Bomb. The problem begins as a simple technical error. Large mainframe computers more than ten years old were not programmed to handle a four digit year. Sitting here now, on the threshold of the year 2000, it seems incomprehensible that computer programmers and microchip designers didn't plan for it. But when these billions of lines of computer code were being written, computer memory was very expensive. Remember when a computer only had 16 kilobytes of RAM? To save storage space, most programmers allocated only two digits to a year. 1993 is 93 in data files, 1917 is 17. These two-digit dates exist on millions of files used as input to millions of applications. Programmers did whatever was required to get a product up and working; no one even thought about standards. The same thing happened in the production of microchips as recently as three years ago. Microprocessors and other integrated circuits are often just sophisticated calculators that count and do math. They count many things: fractions of seconds, days, inches, pounds, degrees, lumens, etc. Many chips that had a time function designed into them were only structured for this century. And when the date goes from '99 to '00 both they and the legacy software that has not been fixed will think it is still the 20th century not 2000, but 1900 Y2K Date calculations affect far more millions of systems than those that deal with inventories, interest rates, or insurance policies. Every major aspect of our modern infrastructure has systems and equipment that rely on such calculations to perform their functions. We are dependent on computerized systems that contain date functions to effectively manage defense, transportation, power generation, manufacturing, telecommunications, finance, government, education, healthcare, and more. The list is longer, but the picture is pretty clear. We have created a world whose efficient functioning in all but the poorest and remotest areas is dependent on computers. It doesnt matter whether you personally use a computer, or that most people around the world dont even have telephones. The worlds economic and political infrastructures rely on computers. And not isolated computers. We have created dense networks of reliance around the globe. We are networked together for economic and political p urposes. Whatever happens in one part of the network has an impact on other parts of the network. We have created not only a computer-dependent society, but, also an interdependent planet. We already have had frequent experiences with how fragile these systems are, how failure's cascade through a networked system. While each of these systems relies on millions of lines of code that detail the required processing, they handle their routines in serial fashion. Any next step depends on the preceding step. This serial nature makes systems, no matter their size, vulnerable to even the slightest problem anywhere in the system. In