Computer experts from Oak Brook, Ill., to Osaka, Japan, plan to spend the next 2-1/2 years in a crusade to mop up the effects of the Millennium Bug. The war could cost $600 billion.
The bug is either a gift from the past or a hangover waiting to happen, depending on whether one is an exterminator or just another company with bugs waiting to hatch at 12:01 a.m. on Jan. 1, 2000.
Exterminators, in fact, have become a huge industry. Roughly 800 computer executives and assorted experts from throughout the country met in Chicago recently in what amounts to a trade show for a new industry blossoming as the clock ticks steadily toward the end-of-the-century rollover.
The Millennium Bug refers to warnings of potential meltdowns everywhere from microwave ovens to mainframe computers as machines programmed to keep time using only the last two digits of the year become confused when the 1900s give way to the 2000s.
Fixing the bug before it bites may be a short-lived industry, but it certainly is a booming one.
The size of the Chicago conference itself is a sign that computer programming companies specializing in the seemingly simple task of replacing two-digit years with four-digit numbers should enter the next millennium considerably richer.
Growing numbers of corporate executives are beginning to realize the extent of their potential problem and the dwindling number of months left to fix it, said sponsors of the recent Year 2000 Chicago show.
In many ways, those at the Chicago session were hired guns, companies that tend to sign on as consultants to corporations and ferret out bugs.
While major computer consulting companies such as Chicago-based Andersen Consulting or Ernst and Young work internally to fix Year 2000 problems, much of the work in corporate circles is being farmed out to enterprises that have decided to specialize in the issue.
One of the companies catering to millennium clientele is SPR Inc., which was founded in 1973 as a mainframe service provider. Already, more than a quarter of the Oak Brook-based company’s $32 million in annual sales comes from contracts to correct the date problem in old software. But even that is small potatoes compared to the potential market as the great number-flipping inches nearer, said SPR chief executive Rob Figliulo.
“There is a huge pent-up need out there,” he said.
Most of the software that SPR and others are fixing was written in a now antiquated programming language called COBOL that used the 2-digit method of calendar notation to save memory on early mainframe computers.
To fix one of these programs, experts must first perform an operation called decompiling, to restore it from the 0s and 1s called machine code into English words and numbers that can be read and corrected.
Once the glitches are found, the exterminators must repair the damage and then test the fixed program to make sure that no new gremlins are introduced while fixing the old ones.
This complicated and laborious process is costly, but less costly than it would be to replace all the affected computers and software with new, bug-free systems, experts say.
Ironically, many of those now reaping rewards for recoding systems are the same people who helped create the problem in the first place.
In fact, the veterans who wrote mainframe software with the millennium problem 20 and even 30 years ago now can argue persuasively that they are uniquely suited to fix it today.
Thomas DePasquale, senior vice president for data warehousing and Year 2000 services at Oakbrook Terrace, Ill.-based Platinum Technology Inc., confessed in an interview that when he started to fix the bug for one of Platinum’s corporate clients, he decompiled one offending program and saw that he himself had written it as a young programmer.
Now Platinum expects that up to 20 percent of its revenues over the next three years will come from fixing the problem, also known as Y2K (for the metric equivalent of 2000), for blue-ribbon corporate clients, according to DePasquale.
He emphasized that the problem arose because there had been no choice but to use the shortened dates in the pioneering days of the computer revolution.
“You need to remember,” said DePasquale, “that one of today’s desktop computers is far more powerful than the mainframes we had to work with when this was done.”
Archaic mainframe tongues like COBOL, long derided by computer scientists who believed that workstations and PCs would slay the old dinosaurs, are highly valued now that so many aging mainframes are due for surgery.
As a result, firms like SPR have poured resources into training fluent COBOL speakers. SPR’s Figliulo is snatching up humanities-trained college graduates who never ran afoul of the anti-mainframe bias and now hope their computer careers outlast the Millennium Bug cleanup.
At the same time, college programs that focus on mainframes are thriving. A beneficiary is DePaul University’s Computer Career Program, which is one of the few places still churning out COBOL specialists, according to Helmut Epp, dean of the School of Computer Science and Telecommunications.
MEMO: See related story under the headline: For bugs in Spokane, who ya gonna call?