Programmer Ed Yourdon (TIME BOMB 2000) says that it's not enough to fix only mission-critical systems. Organizatiins need to test non-critical systems now to create contingency plans.
A system not critical to a business may be critical to a consumer. Not to fix a noncritical system is to risk alienating customers.
In short, y2k repair projects need priorities and contoingency plans. Problem: there is little evidence that most organizations have a y2k repair project, let alone a structureed one that involves noncritical systems.
This is from COMPUTERWORLD (April 20).
* * * * * *
In all of the current year 2000 discussions about prioritization, I fear that many IT managers are overlooking an unpleasant by-product of their triage efforts: the near-certain failure of their noncritical systems.
Those systems may not be critical, but many are still important. Unless conscious plans are made for fixing or shutting down all of those orphan systems, chaos will ensue.
Consider, for example, the latest "report card" from U.S. Rep. Stephen Horn's year 2000 oversight committee: In the process of giving the federal government a "D-minus" grade for its year 2000 efforts, the California Republican predicts that only 67% of the government's systems will be repaired by Jan. 1, 2000. But that 67% reflects only mission-critical systems. The noncritical systems don't even show up in reports from the Office of Management and Budget. . . .
Organizations might still be able to carry out their primary functions if a noncritical system fails, but the company could still be hurt badly. . . .
So why not shut down the noncritical systems now, while there is still time to do so in an orderly, controlled fashion? The reaction from most organizations is likely to be, "No, they're too important. Even if these systems aren't mission-critical, they are important and profitable. We won't be bankrupted if they fail, but it will cost us a lot of revenue. So we intend to keep running them right up until New Year's Eve 1999. And there might be a miracle. Maybe we'll find that we do have the resources to fix these systems at the last moment." Unfortunately, that is highly optimistic. If the organization hasn't allocated resources to the noncritical systems by now, the chances of a last-minute miracle are slim. . . .
It's naive to expect that these systems will die gracefully on their own Jan. 1, 2000. Many will continue running but will spew out incorrect output and begin corrupting enterprise databases. Even if a noncompliant system halts abruptly at the stroke of midnight Dec. 31, 1999, it will leave end users and customers stranded without any advance warning. After all, a system that's noncritical to your organization may be very mission-critical to some of your external suppliers, vendors or customers. You probably don't want them to abandon you in 1998 just because you've decided to shut down those systems at the end of 1999.
On the other hand, if you don't notify them at all, you may be vulnerable to litigation if a post-year 2000 lawsuit uncovers the fact that your organization relegated System X to the noncritical priority level and it becomes apparent that you had no credible plan for repairing System X. In contrast, if you formally notify your vendors, suppliers and customers between now and, say, Jan. 1, 1999, that you have decided to shut down System X on Dec. 31, 1999, then you have a much better chance of avoiding a lawsuit.