Some things are just never as good the second time around. But whereas the Y2K bug promised to deliver digital Armageddon and delivered very little, few people have got excited about the latest ‘computer clock’ problem – and that could make its impact much worse.
The latest date and time bug is the result of America changing its daylight saving time (DST) – one of the many provisions brought about by its new Energy Policy Act. Normally DST is scheduled to start on the first Sunday in April, but from 2007, it begins one month earlier on the second Sunday in March – and ends on the first Sunday in November, instead of the last Sunday in October.
In effect, the time difference for the most of March between London and New York, for example, will be six hours, instead of five.
Unlike the Millennium Bug, however, old Cobol and Fortran programmers will not be pulled out of retirement to change lines of old code as most applications simply take their time from the underlying operating system. In many cases a patch is all that is required, although calendar dates in a few email systems will have to be changed manually.
But perhaps having learnt a hard lesson from the Y2K days – which eventually turned into the costliest non-event in history – analysts are sounding a more cautious and prudent note this time around.
This is a minor problem compared to past issues like Y2K or the Euro conversion, says analyst group Gartner. Nevertheless, “significant business damage and liabilities could occur from applications performing their processing at the incorrect time if organisations do nothing,” they warn.
Time and date problems are a perennial issue for many computer systems and applications. Microsoft’s Excel spreadsheet package, for example, incorrectly has the year 1900 as a leap year, although this has been fixed in the latest release. Similarly, older Unix systems will experience a ‘Year 2038’ problem when its 32-bit time integer runs out of space.