June 1999 Issue

TechCenter

This is the debut of a monthly feature of The Trusted Professional that will look at technology issues that impact professional industries. Members are invited to submit questions, suggestions, and requests for review. Member input will direct the focus of this column. Ideas for articles need not fall under any particular level of complexity--the column will cover both basic and advanced topics. To submit suggestions, or for more information about this column, please contact NYSSCPA Manager of Technology and Information Systems Kevin Lewis at (212) 719-8340, (800) 633-6320, or klewis@nysscpa.org.

Y2K: What Really Causes the Problem?

By Kevin Lewis

There are two common Year 2000, or Y2K, problems that members of the computer industry frequently need to repair: two-digit date storage and leap year calculation.

In the past, software used dates with only two digits representing the year. These older programs operated on the principle that the first two digits of the year were always 1 and 9--only the remaining two digits were subject to change. Programmers used this shortened date to save space and money at a time when the cost of hardware memory was prohibitive. As a result, software designed in this way will suffer disruptions in programs that are date sensitive.

For example, if someone took out a bank loan in 1995 with interest calculated on a yearly basis, there will be reconciliation problems if the computer reads the year 2000 as 1900. Instead of five years of accrued interest the computer will calculate ­95 years of interest.

Second, a problem is derived from the way a program calculates the leap year of 2000. Contrary to grammar school teaching, a leap year does not always occur every four years. The complete rules for calculating whether a year is in fact a leap year are threefold:

1) A leap year qualifies as such if it can be divided evenly by the number four (i.e., 2000 ÷ 4 = 500);

2) However, if the same year is evenly divisible by 100, it is not considered a leap year (i.e., 2000 ÷ 100 = 20);

3) Unless, the same year can be divided evenly by 400 (i.e., 2000 ÷ 400 = 50).

Because Rule 3 applies to the year 2000, it is a leap year. The programming problem arises because old computer code frequently took into account Rules 1 and 2 only. These programs will not recognize 2000 as a leap year. This omission will cause all dates following February 29, 2000, to be off by one day.

The vast scope of the Y2K problem may make the remainder of this century a very difficult time for businesses. Every part of our "electronic lives" must be scrutinized to ensure that it will not be adversely affected by the new millennium. This does not just include computers and software but also phone systems, security systems, and even appliances in office kitchens.