When the IT industry's PR and marketing maestros continue to hype a concept using the same buzz words, it's a fair bet that a cover up is underway aimed at papering over the cracks in the technologies we use. "Plug and play" was supposed to make make adding an expansion card a simple matter. And while we no longer have to deal with DIP switches, jumpers, and headers associated with older PC bus architectures, when it comes to adding certain kinds of expansion cards, many users today still prefer the more realistic description of "plug and pray." Further, when a bunch of differing technologies forced to work together are typically described as "seamless," the term optimistically contends that the technologies have been so elegantly integrated that they appear to work as a single system. That may be the case under certain supportive conditions, but when some of the supports fall away, as they will from time to time, don't be surprised if things unravel and you are confronted with a veritable patchwork quilt of technologies, making it next to impossible to sort out the problem. The most cheekily overhyped word is "robust." My Webster's defines it as "strongly or stoutly built." You might, then, expect to find "robust" used in describing dependably mature technologies like mainframes, Unix servers, or minicomputers such as IBM's AS/400. Yet that's hardly the case because their dependability is a given. Instead we find "robust" is used most often in describing unstable technologies. If a technology is in any way considered delicate, you can be sure the next release will be hyped and marketed as a more robust platform, operating system, application, network or whatever. What this really means is that the upgrade will likely fail less than the current version. This is why so many in the IT world are waiting impatiently for Microsoft to finally ship its long-delayed Windows 2000 version of Windows NT. Not that it's fair to single out Microsoft when so many of the technologies we depend on are flagrantly fragile. And for my money, the most delicate by far is the modem. This product incorporates so much frailty, it's difficult to decide were to begin the criticism. Outside of lab conditions, a modem almost never works at its specified speed; yet manufacturers continue to lie big about speed in ads, while covering themselves in the fine print. And if your modem does come near to working at the advertised speed, you are luckier than you might realize. I recently bought a Global Village modem that is advertised as "56K Standard-compliant V.90." Yet despite the industry standard claim, this particular make is not supported by my Internet Service Provider (ISP). That's a big problem, given it won't work properly with the equipment the ISP uses in Yokohama where I live, though it does work near its top speed if I dial their Tokyo number. When I wrote about modem troubles elsewhere, some readers told me of similar problems - even when their ISP claimed to directly support their particular make of modem. So apparently there is no guarantee you will get satisfaction, whatever make of modem you buy. And while we're at it, how many attempts does it take you to log onto your ISP? On too many occasions, it takes me two or three attempts. I thought this was just my situation, until readers told me of similar frustrations. The IT industry has achieved much over the past 50 years. It can be proud of the progress it's made since hand-assembling the first vacuum-tube monsters, to turning out today's mass-produced silicon notebook PCs that slip easily into a briefcase. But it's also apparent that a certain hubris is creeping into the minds of many who make the industry tick. Look at how some in the IT industry like to pour scorn on the their counterparts in the auto industry. Computer geeks point out that if auto engineers had kept pace with developments in computers, we would all be driving Maserati speedsters costing a couple of thousand bucks that would run for a year on a gallon of gas. Maybe so. But those same hotrods would also be breaking down every 1,000 miles; they'd require a new engine every three years - or be confined to the slow lane; and only one brand of gas and oil would work with the car. Computers may have come a long way at a fast clip. Now, though, it's time to consolidate and refine, rather than continue pushing the technology envelope until it tears. Speed or dependability? What do your prefer? Give John the buzz at boyd@gol.com.

Back to the table of contents