Assume ownership of the task - Management should realize the significance of the project and hold senior staff accountable.

Create an awarness within the company of the critical nature of the project - At this point, the staff should be aware of the implications of the conversion's success; steps should be taken to ensure that the remaining employees are aware of its significance.

Set up a task force - Appoint representatives who have the authority to provide resources, and demand cooperation.

Plan strategically - This is a crucial step - what happens, for example, if specific elements of systems fail or become unreliable? What must be done to identify, prioritize, and then begin work on the various date-dependent systems?

Create high-level planning - Carve out smaller tasks for teams to complete the details.

Don't wait for planning to be completed in detail before starting the conversion - Some detail planning can be completed once the conversion process is underway.

"Year 2000: no COBOL, no mainframe, no problem, right? Wrong. Place your bet correctly, on the year 2000 roulette wheel." - Joe Celko & Jackie Celko, Byte magazine, July 1997

The Year 2000 problem can be broken down into four separate but related problems ("Year 2000 - Threat or Menace?" DBMS magazine, July , 1997):

• The Odometer Problem
• The Millenium Problem
• The Weird Dates Problem
• The Leap Year Problem

You need to test year 2000 changes. Testing is crucial to ensure success in either disaster recovery or year 2000 code conversion, according to experts in the former field.

"It's about getting people not under your control to prepare and participate in testing and opearting viable systems." Allan Graham, Senior VP, Operations, Comdisco - AS/400 Systems Management, December 1997

"Companies will spend between $300 and $600 billion between now and the end of the century to address date change problem." - James Dukart, Washington News Bureau, in The Night Before Chrismas HP Professional, November 1997