Post-Y2K, problems could be stored up for the future

The year 2000 could be leaving a legacy of problems for future IT staff, even though the massive effort put in during the late...

The year 2000 could be leaving a legacy of problems for future IT staff, even though the massive effort put in during the late 1990s apparently kept any crises to a minimum, according to a BCS specialist, writes John Kavanagh.

But the number of companies that suffered will never be known, because few if any owned up, says John Ivinson, reflecting on the issue a year on. Ivinson, an independent consultant, worked for the Government's Action 2000 programme and chaired the BCS working group that produced the society's best-selling guides. He is the BCS vice-president for professional and public affairs.

"Naturally, there were problems, but the cases listed on various Web sites generally use the word 'reports' or 'alleged', leaving people with the possibility - no stronger - that the problems were not in fact caused by the bug," Ivinson says.

"Few, if any, reports were volunteered by private enterprise. For a major company to admit it experienced problems creates all sorts of reputation management issues."

Ivinson also wonders whether the end of 1999 brought any more problems than in previous years. There are no statistics to make such comparisons.

He continues, "Were there any benefits from doing Year 2000 remediation work? As far as I am aware no formal analysis was conducted to answer this point. However, anecdotal evidence suggests enterprises that ran Year 2000 projects experienced very few problems. Some felt they benefited by gaining more efficient use of their systems, better understanding of the role of IT in their business, and understanding of continuity planning."

No one knows the total cost, Ivinson says - but he believes the problem is not over yet. "I believe we are leaving a legacy of problems for future generations," he says. "Many different techniques were used to get around the issue - and virtually none will have been documented. Future programmers will come across code that they will fail to recognise as being a bug fix, and may be confused. Systems will crash because of year calculations based on fixes such as 'if less than 50, assume 20xx; if greater than 49, assume 19xx', which will not be fully understood."

He adds, "There still remain systems that have not yet been put through all their date routines. For example, in 2005 we will see children born in 2000 being processed by primary school systems. Those same children in about 2017 will be processed by university admission systems.

"If they are still in operation, will they work, or will the systems think they are dealing with 1900 or 2100? Who knows?"

A paper by John Ivinson on these issues is at

Year 2000 legacy

Sites reporting incidents include:n




  • Read more on IT legislation and regulation