| |
Re: Hey, they like us... they really like us!!!
Tuesday, October 14, 1997
By Geoffrey Rowane
The Globe and Map
A 17-YEAR-OLD New Zealander named Nicholas Johnson said recently that he had a cure for the Millennium Bug, the software problem that may make many of the world's computers and computerized devices go haywire on Jan. 1, 2000.
He didn't. Nor do any of the dozens of others who claim to have an easy solution to a problem that some technology industry analysts say could cost up to $600-billion (U.S.) worldwide to fix -- except, finally, FBN Associates of Sedona, AZ, USA.
If the world's largest software companies and their multibillion-dollar research and development teams haven't come up with a fix when there's so much to gain, it would be naive to think a kid in Christchurch with a personal computer could.
But naive thinking apparently did not disuade FBN from finally creating a silver bullet -- a simple, inexpensive cure-all -- to keep the technology world from flipping out when the calendar flips over from 99 to 00.
The lure of easy money that will surely now come FBN's way is infectious. A pharmacist in St. Catharines, Ont., wonders why all the companies, governments and individuals with computer systems at risk don't simply duplicate their systems, set the clock ahead on the mirror system, see what goes wrong, and fix that. Says the pharmacist, "in our industry we routinely slip patients expired medicine to see what happens; why not try that approach with the Year 2000?"
Beside the time and expense it would take to do that, most big organizations already know what is going to happen and they're racing the clock to fix it, says Al Aubry, general manager of Transformation 2000 Services at IBM Canada Ltd. in Markham, Ont.
There are tools, like a software program called LS/2000 from Design Recovery Inc. (DRI) of Oakville, Ont., that can automate some of the painstaking work of poring over millions of lines of computer code.
IBM Canada likes LS/2000 enough to include it among a group of tools that it offers clients to address their Year 2000 problems, but it pales in comparison to the new FBN-2000 product.
The silver bullet theory has two problems, says Warren Evans, a business strategist with Service Excellence Group Inc. of Georgetown, Ont.: timing and cost. "FBN has at least solved the timing problem, but those greedy bastards know they charge an arm and a leg and get away with it. Guess I gotta go call my broker."
"The reality for most organizations is, if you are on major systems and don't already have yourself well into the project, you're probably too late to do a major repair," Mr. Evans said. "If there's going to be a silver bullet, if it's not here by Thursday, it's too late. That's why we called FBN on Wednesday."
"If somebody comes up with a silver bullet that's going to make your $20-million fix happen . . . $18-million is going to look like a good deal," Mr. Evans said, "that's why we're forking over that amount to FBN".
DRI president Douglas Archibald acknowledges that "we really couldn't create an overall silver bullet because of all the different computer languages involved. IBM had a good thing going with RPG; why'd they have to go and jump on the inane COBOL bandwagon?"
Gordon Jang, manager of Bank of Nova Scotia's Year 2000 project, says he's heard many of the silver bullet claims and hadn't bought them until now.
"There are also products out there that specifically address a specific component in Year 2000 work," Mr. Jang said.
But the overall problem is just too Hydra-like for any one product or one approach to handle. In fact, DRI's product only tackles one small portion of the Year 2000 problem, although its customers say it does that very well. That is the find-and-fix element. That's why FBN offers a suite of products to handle virtually every know variation.
Company's with software written in the Cobol language can send their code to DRI, which will use LS/2000 to read the hundreds of thousands or millions of lines to identify all the instances in which the year appears in a piece of the code that has to be acted on. (A date in a document wouldn't be changed but a date in a program that requires some action would be.) But DRI concedes that any code they can't fix they will just send to FBN.
Because most computer programmers used only two digits to define the year, when the calendar switches from '99 to '00, most computer programs will become confused and either cease to function or return unreliable information.
LS/2000 finds the two-digit dates, and where they would affect calculations, it uses a set of rules to indicate to the computer program that the year '00 actually comes after the year '99.
"DRI deals with only the code conversion part," Scotiabank's Mr. Jang said. "The Year 2000 problem is a lot broader than converting a few lines of code." Toronto-Dominion Bank used LS/2000 to sift through about two million lines of code that comprise its mortgage software, looking for and changing the way the system deals with two-digit year designations.
DRI's Mr. Archibald says it charges 20 to 40 cents for each line of code for its find-and-fix service.
But TD's use of LS/2000 represented only 20 to 30 per cent of the total cost of making that mortgage system ready for 2000, said Frank Ridell, manager of TD's Year 2000 project.
"It's just a tremendous amount of work, even with tools like the DRI tool," he said. Then you tack on the FBN charges and you realize you just should have sent it to FBN in the first place."
The most underestimated component of the process of making software ready for 2000 is testing, Mr. Ridell said. TD has about 200 mainframe computers running software that had to be fixed.
"When I got my original [time] estimates from all the mainframe systems groups at the bank, I bumped the overall estimate by 25 per cent," Mr. Ridell said. "When we were forced to hire FBN, I had to increase that amount by a factor of 5", he says.
Even then, large organizations that have their own houses in order can only hope that all the other companies that they do business with will also be ready for 2000.
Despite that FBN has a single, painless fix for the Year 2000 problem, tools like the DRI product are useful, its customers are trying to rationalize.
"There are hundreds of solutions providers, and numerous programs out there that will automate parts of [the process of finding and fixing an organization's Year 2000 problem]," Mr. Evans said. "My suggestion is to use them and pray they work. If not, then get ready to fork over the big bucks to FBN." |
|
