Some companies are planning to wait to fix their systems until 2000. This could a huge mistake, says Ed Yourdon. What if it's a public utility, such as a power-generating plant?
But any business faces the same issue. If it waits, shuts down, and can't get back up fast, it will go out of business. Will its customers then sue it? Not if it's dead, but maybe if it's just barely functional. A lawsuit could then kill it.
This is from COMPUTERWORLD (May 25).
* * * * * * * * *
The latest buzzword in the year 2000 industry is"fix-on-failure," a strategy of fixing year 2000 problems after they occur rather than trying to fix them before the Dec. 31, 1999, deadline.
Although such a strategy is better than nothing at all, it's like playing Russian roulette with bullets in five of the six chambers. In particular, it raises the possibility that a critical system may be deliberately allowed to fail, after which the IT "SWAT team" may find that the failure is irreparable. . . .
For the noncritical PC applications-- for example, an office PC used only for word processing-- perhaps the fix-on-failure approach could be regarded as a relatively harmless strategy. But what about using a fix-on-failure strategy for the embedded systems in an electric utility plant? . . .
Thus, before the year 2000 project team announces,"We don't have time to assess this system, and we don't have time to fix it even if we discover that it has a year 2000 problem," it must assess the consequences of the failure.
Obviously, the situation could be quite serious for a process-control system, and we have to hope that the utility executives have thought carefully about the potential consequences of an embedded system failure: If the plant shuts down, how long will it take to bring it up again? . . .
Meanwhile, what about your customers? Do they have a right to know that you've adopted a fix-on-failure strategy? Will they have the patience to wait while you fix your problems, or will they take their business to one of your competitors? And if they suffer a loss because of the (hopefully brief) disruption caused by your fix-on-failure strategy, will they sue you? Will you have a good defense when the opposing lawyer says, "Ladies and gentlemen of the jury, the defendants were worse than negligent: they deliberately decided not to even bother trying to fix their computer systems!"
If not, then you're playing with fire.