Column: Apocalypse Soon

It was the biggest, best-attended year 2000 conference yet, but it still wasn't enough. "If everyone who should be working on the problem were here, we'd be in Yankee Stadium," said Leon Kappelman, co-chair of the Society for Information Management's (SIM) Y2K Working Group and the Y2K conference's chairman.

More than 1,500 people gathered in New York City in March for SPG's Year 2000 Conference and Expo. The presenters included Kappelman, economist Ed Yardeni and tireless gong-banger Peter de Jager. Each one underscored Kappelman's comment: After more than a year of awareness-raising, too few companies and government agencies are paying enough attention to their Y2K problems.

At the same time, new facets of Y2K continue to emerge, like problems with embedded chips and PCs, IT's lack of experience with contingency planning and widespread concern about how governments and the utilities industry are handling their own Y2K issues.

For these reasons, there was a surreal mood to the conference. Attendees felt that the scope of their projects is expanding as the time frame to complete them is shrinking, and yet it seems that the rest of the world regards Y2K as either a low priority or a simple fix that's not of concern to nontechnical types. Yardeni, the Deutsche Morgan Grenfell economist who has speculated about the possibility of a global recession sparked by Y2K, captured the atmosphere best in his keynote: "I don't know about you, but [Y2K] has been an X Files experience for me. I can't tell if I'm being paranoid, or if the rest of the world is just oblivious. Should I be taking heavy doses of Prozac? Check into the Betty Ford Clinic for Delusional Economists?"

Yardeni's line got a laugh -- but it was a very sympathetic one. This audience knew how he felt. It was not an optimistic group.

Embedded Systems Dangers

Two sessions directly addressed the issue of Y2K bugs etched into embedded systems, but questions about the risks they pose were sprinkled throughout the conference. In a three-hour workshop on the topic, David Hall, a senior consultant at Cara Corp. in Chicago, said that identifying all the embedded systems throughout an enterprise is difficult, and testing them for compliance is nearly impossible.

"There's been approximately [US]$10 billion in microprocessors manufactured and sold since 1991," Hall said. "And only 10 percent of those have gone into PCs. Finding and fixing the other 90 percent is going to be a pretty big challenge."

Hall recommended that Y2K managers get test data from vendors and then rigorously verify it. Jay Abshier, Y2K manager at Texaco Inc. in Bellaire, Texas, supported that approach. He said that his company had received assurances from an embedded chip's manufacturer that it was compliant, but after Abshier's team performed an assessment of their systems, they found some systems didn't recognize Feb. 29, 2000, as a valid date. "Doing that kind of testing isn't cheap," Abshier said. "And it's something that has to be in your budget for 1999."

Separately, de Jager, in a keynote address, said that he is considering relinquishing responsibility for his Project Damocles effort. (He has since terminated the project altogether.) De Jager started the project late last year to attempt to bring more attention -- and accountability -- to the embedded systems issue. But now he's worried that by serving as a conduit for complaints about noncompliant embedded systems he'll be subpoenaed countless times in post-Y2K litigation. "Like everyone else here, I want to avoid the courtroom," he said.

PC Risks

Why haven't PCs been considered part of the Y2K equation? Speaker Karl W. Feilder, CEO and president of Greenwich Mean Time Inc. in Chichester, England, a provider of Y2K PC tools that researches the impact of Y2K on desktop systems, has a few theories. First, he says that few companies are selling tools to assess and fix Y2K problems on PCs relative to the number of vendors focusing on mainframes. Second, there's a psychological issue. "Most senior decision makers grew up with mainframes," Feilder says. "They think people don't do anything useful on a PC. And that's fundamentally wrong." Third, most IT organizations exercise less control over PCs than over other systems. "I spoke with a Baby Bell the other day that thought it had 50,000 PCs," Feilder says, nursing a café latte in the lobby bar of the New York Hilton. "They actually had 75,000."

Feilder pointed out that the Y2K problem on PCs doesn't end with the BIOS chip. "That's only 1 percent of the problem," he says. Feilder describes the PC problem as having five layers: hardware, operating systems, software programs, data and data sharing. He recommends addressing each of those layers the same way one would handle the Y2K project for a mainframe system: assessment, triage, conversion and testing.

And companies that allow employees to telecommute have to be concerned not only with PCs in the workplace but with PCs at home. Can organizations afford not to prepare those PCs for the new millennium? "You have to do it," says Feilder. "Otherwise, you wouldn't want to let them connect to your network. It could be seen as legally negligent."

Uncertainty About Utilities

One of the dominant worries at the conference was how the nation's power and telecommunications infrastructure will weather the transition from 1999 to 2000. While most companies take electricity and telecom service for granted, several presenters brought up troubling examples of how even isolated failures can wreak havoc on the business world. Among the anecdotes were discussions of last winter's ice storm in Canada and northern New England, as well as the lengthy power outage in Auckland, New Zealand.

"In my opinion, there is a 100 percent certainty that we will see blackouts as a result of Y2K if current trends continue," said Hall. "How long and how large will they be? I don't know." Hall, a member of the SIM Working Group, also pointed out that power plants of all types are highly dependent on security and safety infrastructures that could fail. "If the local 911 system doesn't work, a nuclear plant can't operate," he observed, noting that regulations require that staff be able to summon firetrucks and ambulances in an emergency.

The only representative of the utilities sector to speak was Cliff DeAnna, Y2K program director for AT&T Corp. in Warren, New Jersey. He noted that AT&T's Y2K project started in October 1996 and that the company spent approximately $100 million on it last year. But he offered only boilerplate assurances that the company's Y2K project had buy-in from top management and that AT&T customers wouldn't suffer any network outages. When asked by an audience member what the company's biggest mistake was in dealing with Y2K, DeAnna responded, "I would've started earlier."

My lunch companion one day offered a similarly disconcerting message. He was a high-ranking engineer at a Southwestern utility that operates one of the country's largest nuclear generators. "The industry has had meetings where we've discussed Y2K," he said between bites of salad. "But the resources just aren't there to handle it correctly. And that's kind of scary."

Others wondered how their companies could endure lengthy power or phone outages. "How do you create a contingency plan for having no power and no telephone?" asked one frustrated attendee during a workshop on embedded systems. The speaker could offer no good answer. On the other hand, another attendee pointed out, "If there's no power on Jan. 1, then no one has a year 2000 problem, do they?"

The Universal Scapegoat

Government-bashing is emerging as a popular pastime in the Y2K community. From start to finish at the SPG conference, nearly every workshop and keynote included at least one dig at federal, state or local government. One well-worn chestnut: "[Insert name of agency here] has announced that it will have its year 2000 problem licked. By 2020." (According to a source at IBM Corp., this joke is actually rooted in reality.)Speakers also repeatedly questioned the sensibility of the FAA's strategy of fixing and replacing its air traffic control systems in tandem, bemoaned the resignation of top IT execs at the U.S. Department of Defense and the IRS, and wondered about the priorities of the new Y2K czar, John Koskinen, the former deputy director for management at the Office of Management and Budget whom U.S. President Clinton appointed in February.

Focusing on the shortcomings of governmental Y2K efforts has its practical side, of course: Most companies rely on services from government entities or exchange data with them. But it also seemed to serve a stress-reducing (or confidence-building) purpose. When compared with government agencies, almost any company's Y2K project looks good.

Yardeni made the Feds the butt of so many punch lines that his keynote address might have been more appropriate for a Borscht Belt audience. His talk even included a fleeting (and dead-on) Jackie Mason impression. "What is contingency planning for the FAA? Binoculars? Pin maps? Ham radio?" Yardeni asked, eliciting laughs.

"On Feb. 2, the president announced the formation of a Millennium Council," he said later. "Its job is to organize all the parties. It wasn't until Feb. 4 that the White House announced the Y2K Conversion Council and named Koskinen the head of it. And Koskinen, believe it or not, was on vacation and didn't start until March 9. That's how serious this problem is to the federal government.

"[Koskinen] also announced that he doesn't want a big staff," Yardeni continued, pausing for a moment. "If there's one time you should have a big staff, it's now. Bring in Joseph Stalin. Give these guys the authority to break kneecaps if they have to."

Petition for a Name Change

Perhaps the most sobering thought to take shape throughout the course of the conference was that the year 2000 problem is actually inaccurately named. For many companies, bugs related to the date change are already appearing. Lawsuits are being filed. And the intensity will only increase when Jan. 1, 1999, arrives, when many systems that look forward a year will encounter problems. Then there's the infamous 9/9/99, often used as an end-of-file date, and the latter half of 1999, when many companies start their fiscal year 2000. "It's scary," said de Jager, who addressed the issue in his keynote. "I scare myself sometimes."

But Kappelman offered a more positive spin. He emphasized that the Y2K project has the potential to make (or break) careers and increase IT's profile within the enterprise. Wearing a patriotic stars-and-stripes necktie, Kappelman told his audience, "We are the generals. We are the field marshals. We have to think not just about the mainframes but [about] the PCs, the embedded systems, the facilities. This is a big promotion. It's a chance to save the world. To paraphrase Winston Churchill, if the IS profession lasts for a thousand years, hopefully we will be able to look back and say, 'This was our finest hour.'"(Scott Kirsner, a Boston-based writer, is working on a series of articles for CIO on the Y2K problem. He can be reached at Kirsner@worldnet.att.net.)