A Year Ago: Microsoft one of the good guys?

First published: Wed, 22 Apr 1998 13:26:29 GMT
Written by Guy Kewney, Contributor

Microsoft's attitude to its own Year 2000 failings, frankly, stinks. It has taken exposure by independent consultants to persuade the company to admit that its own Web sites and documentation have falsely claimed there was no problem when we enter the next century.

That said, it is simply wrong to castigate Microsoft as if that corporation might be any worse than any other IT outfit around. For once, they're one of the semi-good guys. At least they have now started to come clean.

Quite simply, the corporation which has accurately assessed its IT risk from the century changeover, does not exist. Nobody knows what will actually happen over the next eighteen months, which data will turn out to be corrupt, which machines will turn out to be wonky.

What is frightening, however, is the number of companies which, like Microsoft, took an arrogant attitude to the problem. "We're a great organisation," was the message, "we don't make stupid mistakes, so there won't be a problem." For the last two years or up to last October at least, Microsoft took exactly this attitude and now its being brought to task for its, shall we say, oversight.

But at least Microsoft has this excuse: any year 2000 failings in its code will become apparent. There are millions of users who will ensure the company knows. No credit to Microsoft for pretending there wasn't a problem; even less for not bothering to check. But they have this justification: they can argue they don't need to because their users will shout if a problem shows up.

And they will, and they have; and it's not a pretty sight. But the real problem lies in data validation.

There are companies whose very survival depends on getting dates right: commodity brokers, futures merchants, insurance companies, and loan sharks. Incredibly, they don't have a standard strategy for passing dates between them. Indeed, it's hard to see how they could have, since most of them don't yet admit they have a problem with dates.

These companies have no products in warehouses. Their data is all they have to justify their existence; if the data goes bad, they die. When these companies die, because they have no stock to trade, they will take their clients down with them. Yet even today, they are entering dates in eight digit fields, but the operators are only typing in the year as two digits, and software is expanding it to four. It's working, because some of the time it's obvious. But when they start getting dodgy dates from their own clients and find they don't have any way of validating them, it will all go horribly pear-shaped.

This stupidity is far from uncommon, but it's merely stupidity, not evil. The evil exists. A good example is a firm of lawyers which is touting for business by saying: "Your IT company is lying when they say they need to charge money for support. All today's software actually works perfectly, and doesn't need support. Don't pay them."

Another example of evil is the firm of "consultants" which tells Y2K-aware companies not to try fixing their year 2000 problems. "Just wait till the software falls over, and then sue the IT firm for millions. It's their fault."

The faults inside Microsoft code for year 2000 have not yet all been exposed; and when it comes to the Euro, there'll be a new crop. But at least Microsoft is now admitting that it has a problem. The real problem is the thousands of people who think there's no need to worry about it, because they'll sort it out if it turns out to be necessary.

It's true, some of these people are ordinary users who have gone to Microsoft's Web site in the past, and searched for year 2000 problems, and have been reassured by utterly false statements there suggesting that their Microsoft programs are just fine. At the same time, Microsoft's duplicity in this matter is almost an industry standard, and most, not just some, of the other software vendors are equally guilty.

Editorial standards