 | |
| Vol. 17, No. 8 | Oct. 23, 1997 |
| |
| Vol. 17, No. 8 | Oct. 23, 1997 |
Many moons ago, when computing was in its infancy, when systems were longer and wider than a conference table and taller than an average freshman, when memory and disk space were expensive, computer programmers fretted how to save space in their datasets.
"Let us abbreviate the year," said the programmers.
"Let it be done, for it shall save our budget on this project," spake the managers.
"Let it be so," said the CEOs.
And so it came to pass that birth dates, dates of deposits and withdrawals, transaction registers, digital watches and other devices and data records all expressed years as two-digit numbers.
Twenty-five years ago, people in the data processing industry were concerned with the efficiency of storing information, barely glancing ahead to the end of the century. It was a common practice to store dates using six characters instead of eight: For example, Nov. 22, 1973, would be stored as "731122" instead of "19731122."
In a personnel data file that contained each employee's birth date, date of hire, date of most recent changes of address, marital status, health insurance, job title, salary and date of separation, a programmer could save at least 16 bytes per employee by abbreviating the year.
The space savings may seem trivial to us today, but in the '70s and early '80s, disk space and memory were very expensive. If you could save even 64K worth of space, you were saving money and making a file that was much easier for your organization's computer to process.
Now, however, this cost-saving strategy has led to a balloon payment due before midnight on Jan. 1, 2000. Organizations across the globe are investing staff time in rooting out all these abbreviations in their stored information.
If the year is abbreviated as two digits, computer programs will act as if "00" means 1900. In other words, without intervention, some software may treat people born in 1971 as if they were minus 71 years old instead of 29, and all records, including those maintained by lending and financial instuitutions would be affected as well.
It is highly unlikely that there will be more than minor interruptions to systems across the U.S. because most organizations, like the University of Delaware, have been working on the problem for some time, identifying computer applications and data files that need patching before the big day, making the changes and then testing the new applications and data files.
"The University is making every effort to be assured that our computing systems and those of organizations upon whom we rely for electronic information are all Year 2000-compliant," Vice President Susan Foster, information technologies (IT), said. "We plan to have a fully compliant and tested computing environment by January 1999."
"We are ahead of the curve in terms of the work we've done to get ready," Carl Jacobson, IT/Management Information Services, said. "We don't have all of the University business systems converted yet, but we are well on our way."
While most of the conversion work on the UD campus has been happening behind the scenes, there will be some things that University students, faculty and staff will need to do to be sure that their own computers and datasets are ready for the year 2000.
"We are evaluating utilities that will help faculty and researchers determine which of their programs on the central UNIX systems will be affected," Dick Sacher, IT/User Services, said.
"We anticipate having testing procedures in place for most desktop and laptop computers in the spring of 1998," Ron Nichols, IT/Technology Solutions Center, said.
According to Nichols, most Macintosh computers should be fine until Feb. 29, 2020. However, the next release of the Macintosh operating system will take care of that bug. "We'll have an announcement for campus Windows users in the spring," he said.
-Richard Gordon
