Year 2000 Needs More Government Attention By Michael Aisenberg
Recent weeks have seen the rekindling of concern about the year 2000 readiness posture of the U.S. government's enormous information technology infrastructure. Both the Office of Management and Budget and the House of Representatives Government Oversight Committee have raised urgent concerns with agencies' capacity to respond to the requirements of becoming ready for the arrival of FY'00.
The vendor community is rapidly dividing itself into a range of approaches to working with our government customers. Among agencies, such as the Social Security Administration, where alarm bells went off early, there have been long-standing efforts at addressing date-handling issues and assuring continued mission fulfillment by using planned programs of remediating systems.
Newly finalized Federal Acquisition Regulations on year 2000 published in the Federal Register Aug. 22 require agencies to acquire only year 2000 compliant information technology in the future.
Unfortunately, the inevitability of the millennium's arrival doesn't seem to have uniformly drawn the attention of federal budget managers when it comes to maintaining the date-handling performance of the installed base.
As both Rep. Steve Horn's recent report card and OMB's third quarter progress report detail, there are likely to be huge shortfalls of readiness among federal agencies.
The painful fact of the matter is that time and resources are both too short to allow for the tedious process of line-by-line remediation of all existing application code. Some estimates say that 30 percent of the job cannot or will not get done in time.
Years ago, the Gartner Group, perhaps sounding a bit like Chicken Little, cited the enormity of the year 2000 effort and suggested the notion of triage. In this context, "triage" means cut obsolete applications, put off work on applications that are not critical, prioritize the remainder and get busy now.
OMB now estimates a $3.8 billion tab for federal year 2000 fixes. The Department of Defense alone estimates $1.45 billion. Yet these numbers have both nearly doubled in 12 months, while year 2000 program completions have tallied at a snail's pace. The cost estimates can only be expected to rise, even as the clocks tick down.
Many agencies have indeed entered into at least a pro forma triage exercise. But it is apparent from anyone dealing with chief information officers that many are not convinced that trouble lurks around the turn of the calendar, nor that it is a politically correct career move to demand the funds for a timely completion of the year 2000 program.
When DoD publicly declares that more than 20 percent of its 18,000 systems are mission-critical, yet that it has only begun remediation on half and will likely miss the start of the federal fiscal year 2000 on Oct. 1, 1999 for many of these, taxpayers surely have reason for concern.
Is there an answer? Without access to unlimited budgets and a great deal more time for completing year 2000 fixes, for some systems, there may not be a solution.
It would have been nice if the President's Commission on Critical Infrastructure Protection had started two years earlier, with year 2000 high on its agenda. Then the administration, Congress and the vendor community would have a common plan. But it didn't, and we don't.
Some vendors, including my company, have been attempting to work with agencies on affordable, deliverable alternative strategies for achieving year 2000 readiness.
Foremost among these is platform migration. In this environment, migration involves identifying those mission-critical applications for which there is enough time, as well as mission, budget and system life-cycle reasons to consider substantial or even complete redeployment of the system hardware, operating system and applications onto current, year 2000-ready technology.
In a DoD apparently tiring of its flirtation with a proprietary Common Operating Environment, the marketplace nudge toward Defense Messaging System on a commercial operating system platform (such as Windows NT) gives a long-term justification for migration. For any other agency, especially those already facing system replacements in the next several years, migrating now and solving your year 2000 as a free benefit may make sense.
Migration is not a year 2000 planning panacea. Without time and resources, it is no better than the use of windowing remediations now, field expansion to four-digit year fields at some point in the future, and prayer. But for mission-critical applications, it may well be the planning priority that CIOs must recognize if they are to minimize the number of systems on the triage list tagged "kill" or "defer."
With auditing firms reporting up to 50 percent error rates in windowing remediations already completed, there is a compelling need to promptly revisit work already done, plan for work not yet done and provide resources for considering alternative strategies.
Michael Aisenberg is director of strategy and policy for Digital Equipment Corp. in Washington. He is also chairman of the ITI Year 2000 Task Force and a member of the ITAA Year 2000 Legal Committee.
©1997, TechNews. All rights reserved.