Forget about Moore's Law faltering at the edge of physics, put aside all worries about complexity defeating our security and management needs. The abyss into which we stare is deeper than either. We are running out of acronyms.
For years, engineers and marketing types alike rapaciously mined the English language. At first, the pickings were plentiful. Your display is composed of single bits of information that make up the picture elements? Call them pixels. Randomly accessible memory? RAM. The language of information technology glistened with newly cut jewels: modems, IO, VDU, hex, DMA, CPU -- some might be less than beautiful, but all filled a need, were quick to say and brooked no ambiguity. It was all so easy.
Too easy. As the seams close to the surface were worked out, finding new ways to describe new ideas became harder and the results less satisfactory. Where once the simple LAN did full service for your local area network, now the 802.11a/b/g high speed DSS/FSS wireless LAN has to be coupled to the incomprehensible Wi-Fi -- wireless fidelity, whatever that means. That in turn mutates to WiMax which some marketeer would probably say means Wireless to the Max, if you let them. Some people have given up altogether: Zigbee might be a fine name for a children's cartoon character, but it has no cognitive connection whatsoever with the 802.15.4 low-power personal area networking standard Might as well call your next CPU Spongebob Squarepantium. At least Pentium sounded like a real word.
Yet those who continue to try and stick to the rules do no better. Ultrawideband wireless takes the standard inflection to UWB, which is fair enough -- were it not for the fact it has five syllables, enough for the second line of a haiku and one more than the word it seeks to replace. There is a move afoot to call the technology uweeba, but that has gained no ground, probably because any word with 'wee' in the middle has overtones of childishness and tweeness. Or take the well-meaning folks behind OpenOffice.org. That's easy enough to say but looks as ugly as a bullfrog in print: you can't call it Open Office because that's not its name, and attempts to render it as OOo resemble a hieroglyph for a grazing cow.
There is even the suspicion that some may be using this global shortage for their own ends. Take the Microsoft trusted computing initiative called Palladium. That was a splendid name. Not only is it a tarnish-resistant metal, but in classical times it was the name for a sacred object that safeguarded a city or state. Yet as controversy mounted, Palladium was ditched in favour of the Next Generation Secure Computing Base -- NGSCB. Try saying that, remembering what it stands for or understanding what it means -- and watch it drift off the radar like a stealth bomber.
We have no time to lose. French speakers have the right idea: wherever you have Francophones, you have tongue troopers who tirelessly repel foreign invaders and promote the one true language. L'Académie Française should be our model - a committed collection of conservative experts who apply strict standards to any linguistic innovation. Before any new technology is allowed out of the labs, it and the proposed name -- with any acronyms or abbreviations -- should be submitted for inspection. The ugly, the unpronounceable and the frankly absurd can thus be stoppered before birth, and our desperately limited supply of untapped resources be put to the best use possible. Where a new idea simply cannot be given a snappy, obvious and informative name, it should be quietly forgotten. Old, outdated or unused acronyms can be reclaimed and refurbished: this will put an end to such problems as ATM standing for Asynchronous Transfer Mode, Automatic Teller Machine and Adobe Type Manager. We may even get back to a world where IP make sense.
I realise that this move alone will cause the greatest difficulties to the many teams of management consultants working in IT whose well-padded existence depends on inventing new and incomprehensible terms with which to wrap up old ideas that nobody's thought about for a couple of years. They shouldn't worry: the Mondial Institute for Naming Technology (MINT -- see how easy it is?) will need to do much research and issue many papers on how it is to carry out its essential tasks - and nothing is as opaque to the general reader as a linguist talking about language. A little retraining and everyone will be kept profitably busy, as before.
For now, I call for a moratorium on new coinings and a more careful use of the old.
Rupert Goodwins is the technology editor of ZDNet UK