LONDON -- You can't say that the millennium computer bug industry lacks imagination.

If we breathed a collective sigh of relief to have handled Jan. 1, 2000, we are now told to batten down the hatches in this leap year and watch out for Feb. 29, 2000 and then Oct. 10, 2000.And this after a string of false alarms in 1999.

Bug Cassandras managed to induce hysteria about the possibility of chaos on Jan.1, to such an extent that industry and governments shelled out up to $600 billion to prevent it.

In the event, as midnight came and went and the world ushered in the new millennium, lights stayed on, bank machines still spewed out cash, taps still provided running water and planes stayed in the air.

Money well spent, said the computer software industry.

Others wondered whether panic spending induced by hyperbole might be the explanation.

Now we are asked to believe that another date which is rushing toward us -- Feb. 29, 2000 -- might cause computers to crash. We still can't relax if that passes without incident. Oct. 10, 2000, is the first eight-digit number of the new millennium and might somehow unhinge our computers.

A succession of other warnings last year turned out to be duds.

April 9, 1999 -- the 99th day of the year -- was supposed to upset computer systems. So was Sept. 9, 1999, which could be represented as 9-9-99. In theory this string of nines was a threat to computers because nines were often used by programmers to instruct a computer to shut down, or prepare for maintenance.

Jan. 1, 1999, was a danger because many contracts, insurance policies and loans would reach ahead one year and trigger the millennium bug.

Before the Dow Jones Industrial Average burst through 10,000 earlier this year, some experts said computer programmes might not be able to handle the fifth digit in 10,000. Either the figure generated very little chaos, or companies and organizations did a great job in suppressing bad news.

The popular navigation tool, the satellite-based Global Positioning System, reset its computer clock in August. Although not strictly a Y2K problem, it was heralded as a threat to light planes and yachts. In the event, only a few Japanese taxi drivers using computerised maps got themselves lost.

Feb. 29, 2000. has credibility as a possible problem because it is a one-day-in-400-year event. Because the phases of the moon don't exactly mirror the calendar, every four years an extra day in February brings it into step again. But 1800 and 1900 weren't leap years. The ruling by scientists is that all years divisible by four are leap years, except those divisible by 100. The exception to this is that years divis-

able by 400 are leap years.

Given the opacity of the reasoning, it is not surprising that some computer calendars don't include Feb. 29.

Nick Fitzhugh, Year 2000 teamleader at management consultants Ernst & Young, believes companies would be wise to pay attention to the possibility of problems from the leap year date, not least because the millennium computer bug might be quietly undermining data bases.

"Be vigilant because February 29 is another possible hot spot. It makes sense for companies to increase monitoring of their systems on February 28/29, March first, and the end of the first quarter," Fitzhugh said.

Most experts involved with the millennium computer bug expected that it would not be an explosive, spectacular event. If it had any impact, the use of two digit numbers like 87 or 98 would cause computers to stumble over the zeros in 2000 and gradually degrade before finally crashing. They also warn that the initial euphoria of a glitch-free holiday weekend might not last when corporate, banking and government computers are switched on when New Year holidays are over.

The Gartner Group, a U.S. information technology research company which famously said the global cost of repairing the bug would be from between $300 billion to $600 billion, said fewer than 10 percent of problems would occur during the two weeks surrounding Jan. 1, 2000. Fifty-five percent of problems would hit over the rest of the year.

The leap year date will not cause serious problems, according to Gartner Group analyst Andy Kyte.

"There could be some residual leap year problems because of calculations based on what day of the week comes after February 29, but there won't be any explosion on the day," Kyte said.

February 29, 2000 falls on a Tuesday.

Fons Kuijpers, member of the management group of PA Consulting, doesn't give much credence to the leap year effect.

"One, one, 2000 was the big date. The leap year problem isn't going to be worth worrying about too much. Organisations should be a little more alert than they might usually be. I don't think they will be setting up big command and control centres like we have over the New Year," Kuijpers said.

Margaret Joachim, Year 2000 expert for U.S. computer services company Electronic Data Systems Corp, is looking at other dates for evidence of computer problems and dismisses any possibility of fallout from 10/10/2000.

View Comments

"You should be looking at data processing from the first set of month end figures at the end of January and see if they are markedly different than expected. Then the leap year, first quarter-end and year-end," Joachim said.

"As for 10, 10, 2000, computers fill in the blanks so I can't see that one. I don't think we need to get terribly worried about that," she said.

Gartner Group's Kyte looks back at previous leap years for a guide to its possible danger.

"Some processors failed to work in '96 or '92, but nobody died. The problems caused were so small it wasn't worth fixing," Kyte said.

Join the Conversation
Looking for comments?
Find comments in their new home! Click the buttons at the top or within the article to view them — or use the button below for quick access.