GROWING UP

Dick Pountain/23 November 1998/Idealog 52

Just for a while last week I was actually persuaded to believed that I'm working in the most irresponsible industry in the history of humankind. What caused this momentary crisis was a spate of newpaper headlines about the Y2K problem - in particular how the government has decided it can't afford to scrap the Territorial Army after all, because they will be needed to patrol the streets and prevent us from eating one another during the general breakdown of civilisation that is supposed to happen on January 1st, 2000. I imagined a Hieronymus Bosch landscape, under lurid red twilight, dotted with the huge silvery burning beacons of crashed Boeing 747s; acrid smoke drifts across the hellish scene as a million Pop Tarts are incinerated by malfunctioning toasters; wild-eyed mobs race from one defunct CashPoint machine to another, deprived of Chicken Tikka Masala for almost 24 hours; many, already suffering from horrendous hangovers, are made crazy with sleep deprivation because their radio/clock/alarms went off one hundred years too soon.                             

And all this because a bunch of irresponsible programmers tried to save a couple of bytes on storing their dates, to save their bloated capitalist employers a few measly pence. At least, this is the picture that the popular media are trying to build up as we head toward Y2K. Of course those consultants who will trouser most of the £500 billion some pundits reckon will be spent on Y2K cures, are in no hurry to counter this view either. But is the computer industry really so unusually irresponsible?

I'm not the best person to try this line on, because I was brought up in a small mining village called Hasland, quite close to Bolsover in Derbyshire. I remember my first bicycle and riding around the green fields that surrounded the nearby Staveley Chemicals plant. That plant is now chopped up into little pieces and buried under those same fields, the scene of the world's first and biggest dioxin pollution disaster, decades before Seveso. As it happens I went on to train as a chemist, so my knowledge of that industry is fairly well founded: certainly well enough to remember the grotesquely-named Love Canal in the USA, where an entire housing estate discovered it had been build over drums of toxic goo. Let's not even talk about Agent Orange or Bhopal. Or how about agri-business, which may just possibly have infected 75% of the population with a 20-year incubation brain-sponge disease (they don't know yet), and is busy spraying food crops (and suburban housing estates) with pesticides that the army would have happily snapped up for nerve gas 50 years ago. 

So on balance, no I don't think the computer industry is outstandingly irresponsible, by the standards that these things ought to be judged. Unlike those chemical and agri-business disasters, there's no evidence that two-digit dates were adopted for purely penny-pinching reasons. The lapse occurred in the infancy of the technology, when memory and mass storage were terribly scarce resources. Negligent yes, but who is going to cast the first stone at those proto-programmers from the 1950s, trying to shoe-horn a banking application onto a machine with less memory than my PalmPilot, and for whom the millenium, if it ever entered their heads at all, seemed a lifetime away.  

Which (poor-taste joking aside) is not to say that Y2K is not a real problem, nor that something should not have been done to catch it earlier. The question is what? I've for many years been a fan and advocate of Professor Niklaus Wirth's Pascal/Modula/Oberon languages, and so it would be easy enough for me to take a "told you so" line, that if we had taken strong data-typing more seriously years ago this could never have happened. I won't though, because it wouldn't be very helpful; the market chose Cobol, and C and dBase II and all the rest, and it is pointless to imagine otherwise.

Which brings me to the crux of the matter. Who can we expect to impose the sort of standards on the industry that will ensure there are no more Y2Ks lurking? Some might look to the government, but government attempts at regulation of the computer industry don't have an impressive track record. Despite government endorsement for the SSADM design methodology, and the plethora of memorably-named standards like BS5750, IEC65A and ISO9000-3, the only programmers who are ever likely to encounter any of these are working on government contracts, and then the standards are often regarded as just a hurdle to be scrambled over. In truth, governments and standards bodies are too slow-moving to regulate a fleet-footed industry like computing, which has always been riven by an internal struggle between standardisation and innovation. The standards that matter are all de facto ones: for example MS-DOS and then Windows have provided the "standards" that allowed software to grow into a multi-billion dollar industry.

My own guess is that it's the people that need to change, not the industry standards. Programming needs to become a proper profession, with proper professional qualifications, and that doesn't mean single-vendor schemes, like MSCE which is mainly intended to tie people into Microsoft's development tools for life. It means a certificate from an industry-wide certification body, like the accounts and lawyers have, except that, such is the pace of development, you'd have to be periodically re-examined to keep this certificate. If that means the end of self-taught programmers (and I'm one myself) then so be it, because as the world gets more and more wired, programming is getting to be too important to leave to gifted amateurs.