Commentary/Dilip D'Souza

Year 2000: Forward To The Past

I magine that, in a short-sighted attempt to economise, some vital part in every 1970 model of India's very own Ambassador car was made of cardboard. Nobody will keep the car until the part crumbles, the company may have reasoned then, so let's use the cardboard and cut some costs. Only, over a quarter of a century later, thousands of those 1970 models are still roaming the streets. In 1997, the cardboard parts are finally wearing so thin that driving is a hazard. Accidents are happening, owners are angry. What does the car company do?

It recalls all 1970 cars to replace the cardboard with the steel part they should have had to begin with. Because thousands of cars need to go through this little operation, there is a substantial amount of money involved. The problem is, this being an old model that is not produced any more, there are very few people still around who know how to open it up and do the replacement. Given Indian genius at keeping cars going, I know you'll say this is hardly likely. But remember, I'm asking you to imagine.

So in this imaginary land, those few mechanics are suddenly in great demand, the skills they thought they would never need again suddenly greatly valued. But that's only one side of the story. Recognising the profits to be made from this cardboard caper, all kinds of institutes spring up almost overnight. Almost overnight too, and for fat fees, they train mechanics to do this one specific replacement. The newly-trained mechanics go out and offer the world their services at even fatter fees.

In a few months, all 1970 Ambassadors have had their operation done. All's well.

Or is it? The job complete, all the mechanics are suddenly out of work, saddled with this one somewhat silly skill they will never need again, good for almost nothing else.

Maybe you will recognise the very real situation that I am trying to draw an analogy to here: the Year 2000, or Y2K problem, the Millennium Bug. In the software world, I probably don't need to tell you, Y2K is big news, big money. Stories abound about how companies across the West are shelving all other projects and flinging money at Y2K. How much money? There are estimates that Y2K will be a $1.5 trillion business over the next three years. To put it in some perspective, that's 150 times India's defence budget this year.

So what exactly is this Y2K problem, apart from its vaguely and appealingly hi-tech sounding label? Back in the mists of prehistory when computer memory was expensive -- the 60s and 70s -- programmers wrote thousands of programmes that, among other things, kept track of dates. Looking for ways to save memory, they hit upon a neat trick: they used two digits instead of four to represent the year in dates. So "56" stood for "1956", "07" for "1907", and so on. In particular, "00" stood for "1900".

Unnecessarily stingy? Given the times, not at all. Suppose that the TLA company has 10,000 employees. TLA maintains records of employees's birth dates and the date they joined the company. That's two years, and so eight digits, for each employee; across the company, 80,000 digits to be stored. If one little trick could reduce that to 40,000, in a time when memory was expensive, that trick must have seemed very clever indeed. Innumerable programmes and vast databases were designed this way.

All very well. Except that the year 2000 is coming up. Suppose TLA hires you that year. Your record in the company's database will indicate that you joined in the year "00": 1900. I'll leave you to dream up the consequent headaches.

How do you handle "00"? That's Y2K for you.

Finding a solution is not easy. Those programmes and databases were written in computer languages that are largely out of use today. Much as car technology has moved on since the 1970 Ambassador, computers and their languages have evolved. Today, memory is cheap and there's no need to be stingy with space any more. But also today, few people are left who still know the old languages, who can update the old programmes to understand "00" correctly. Besides, there is such a lot of data around that updating it all to recognise four digit years is a fantastically large problem. It is also acquiring some urgency, because The Year 2000 is not very far away.

So the few programmers who know those old languages are suddenly in great demand, the skills they thought they would never need again suddenly greatly valued. Recognising the profits to be made, there are institutes popping up every day to train starry-eyed youths in this one specific task: fixing the Millennium Bug. Indian companies are falling over themselves to sign Y2K contracts with Western firms. Those firms are happy to pay less to get the job done by Indians than their countrymen would charge.

In Hyderabad, which seems to have taken to Y2K in a big way, a new Y2K institute opens every third day. Every month, they churn out some 3,600 of these new-age mechanics. Fresh from training, pockets lighter by the outlandish fees they have had to pay, many of them board the first available flight to the West.

Going abroad, and particularly to the USA, is Y2K's big attraction. That craze has had a curious side effect. Y2K has sown the seeds of a minor social mutation, now apparent in Hyderabad, but perhaps soon to spread across the nation. Yes, the fix-it job is worth thousands of dollars and a foreign trip, but those are by the way. Those programmers of the 60s and 70s could hardly have imagined that their programmes would still be around to need upgrading today, true. But far less could they have imagined that their little trick would, in the late '90s, significantly upgrade the marriage prospects of thousands of young men in Hyderabad.

Indeed, this is Y2K's real fallout. Back from his bug-fixing jaunt in the US, the Y2K mechanic can now command over twice the dowry he might otherwise have been worth. In some cases, that is a rise from 1.5 million to 2 million rupees to 3 million to 4 million rupees, just for having been abroad. The Asian Age reported recently that even those sent back because they were incompetent ask unabashedly for a fatter dowry: they did visit the USA, that promised land, did they not?

It's bad enough that Y2K is unwittingly feeding a greedy social evil. Flush with dowry money, I expect these youths are not overly worried about future career prospects. But consider what happens on January 1, 2000, if not earlier. Presumably, by then everyone now afflicted with the Millennium Bug will have got it fixed. Suddenly, the floods of Y2K experts will be out of work, saddled with this one somewhat silly skill they will never need again, good for almost nothing else.

This is the truth about Y2K. It sounds fascinatingly futuristic -- which mention of computers is not dressed up like that? which mention of 2000 AD is not? -- it holds out the prospect of a trip to foreign lands, it must seem like a windfall to thousands of people. But it is just a routine fix-it job, more trivial than most. Nothing more, nothing less. Learning to do that job imparts no skills that might prepare you for a long, productive career. It makes you valuable for no more than three short years. After that, you are worth nothing.

In a lot of ways, the Indian software industry grew up on being cheap, because it was willing to take on mundane software jobs. It spawned thousands of profitable training institutes whose graduates, despite the hype, never learn the skills they really need for the complex demands of software development.

As the industry matures, much of that is changing.

But along comes the Millennium Bug. Once again, being cheap is our selling point. Once again, quick but short-sighted training is big business. Thousands are wildly excited about doing one trivial, even foolish, job.

The worst thing about Y2K is the false picture it draws, right from its name, of a brave new world of technology. Even if they have dowry on their minds, Y2K mechanics deserve better.

RELATED FEATURE: Indian software industry loses $ 14 million every day

Tell us what you think of this column