There are two schools of thought in the arcane tradition of naming wireless networking standards. One tends towards the cryptic number group, as in 802.11b, harking back to the days of five-figure Enigma codes crackling their way through war-swept skies. Cool. The other eschews this in favour of mysterious figures of myth and history -- Bluetooth, the ancient Scandinavian king, is a case in point. The latest wireless standard is called ZigBee. I have no idea to whom, what or where this refers, but I'm trying to find out. (In an attempt to confuse the issue, it's also called 802.15.4: I'm ignoring that.) I'm also trying to find out how it works, what it's for and how it'll fit in with everything else. That's OK, that's my job -- what's not OK is that one of the primary ZigBee protagonists is Philips. Philips is unique for a couple of reasons. It is Europe's only truly huge consumer and professional electronics company, and it is the hardest company to get any information out of. Ever. Sony comes close, but for sheer ineptitude at marketing, sales and PR, Philips takes the gold. Let me illustrate. With a sinking heart, for I knew of old what was to come, I dug up the main Philips UK switchboard number and prepared myself. There was absolutely no point in asking about wireless networking -- let alone ZigBee -- at this point, but that's OK. I know the score. "Hello, Philips". Ah, hello. Can I talk to your press office, please? "Hold on..."
Curious goings-on over at The Register, our comrades in cyber-news and fellow seekers after truth, fairness and highly-paid advertising. Not content with merely reporting the news, they're taking over the means of production and setting up their own ISP. Now, these are challenging times and no source of revenue is too infra dig to overlook -- just this week we went out and bought another online news company, Silicon, and we're about to launch our fab newsletter, ZDNet Week -- but doing an ISP? When you spend your time writing about the things? Not that the Reg will in any way let this new commercial interest change their attitude to other ISPs, which will remain, I'm quite sure, one of unabashed cynicism with a garnish of full and frank suspicion. And I'm delighted to give them a spot of free publicity: if it works -- and I feel there may have been better times to launch such a thing -- then it'll open up new media models for the rest of us. I for one fancy starting up a wireless ISP, turning all of central London in a giant 802.11b/a, Bluetooth (and what the heck, ZigBee) hotspot by means of a giant antenna stretched between the pointy bits of Tower Bridge. Alternatively, now that crows have been shown to be cleverer than chimps in designing and using tools, we could train the ravens of the Tower itself to do simple IT management and configuration skills -- they might have a bit of a problem with the mouse, so we'll stick to text-mode systems. They're good at hunt-and-peck typing, and they won't run up a fortune in cab fares rushing to sort out clients' problems. Tell me, expert IT corvid, will processors continue to double in power every eighteen months? Quoth the raven: "Never Moore!". Wednesday 7/08/2002
Dijkstra, father of structured programming, is dead. A sad day. On my first programming job, on practically the first day, my new fellow coders were curious about my experience (very limited) and skills (patchy). "You know about Dijkstra?" I was asked, and when I replied in the negative that was all they needed to know. Emergency remedial programming books were procured, and I obviously absorbed enough of them quickly enough to be allowed to carry on. One of Dijkstra's many skills was phrase coining. A favourite of mine was 'deadly embrace', where two processes running on a multi-tasking system end up waiting for each other to produce a result before continuing. When, later in my career, I was working on a 386-based networking OS, I perpetrated at least two of those (resource contention, don'tcha know), and I was delighted to learn more about it. Today, I get asked to write a quick obituary for the chap, and I look up deadly embrace to make sure I've got my facts right. This leads me to the Hacker's Dictionary, a fine online document covering huge amounts of informal computer history... including, I discover, the smoot. A smoot is a unit of length, created in 1958 at MIT. An undergraduate, the estimable Oliver R Smoot, Jr, was laid on Harvard Bridge -- no, not like that -- and used to measure out the length. It was 364.4 smoots long, and the markings are regularly renewed to this day. They do that sort of thing a lot at MIT, you know. Curious as to the whereabouts of Oliver R Smoot, Jr. -- it's not a name, one feels, that's too common, even in America -- I ask the mighty Google. It turns out that Mr Smoot is now the chairman of the American National Standards Institute, the group responsible for measuring and defining more things than Adam. Predestination sometimes seems the only explanation... obviously, a life made to measure. Thursday 8/08/2002
A hundred years ago today, Paul Dirac was born. A physicist arguably on the same plane as Einstein -- but painfully shy and averse to publicity -- his contributions to 20th century science were enormous. Most famously, he predicted the existence of anti-matter and described the nature of electron spin, two aspects of his work in quantum physics. His Nobel prize -- shared with cat-fancier Erwin Shrodinger -- came his way at 31, the year after he'd been appointed Lucasian Professor of Mathematics at Cambridge. All through his life he continued working on some of the great problems in physics, and it's arguable that nobody has contributed as much to the field. This week, though, shows that his ideas still have their finest hour ahead of them. Researchers at the University of Wisconsin in Madison have published details of how they plan to build a quantum computer -- built out of single-electron devices -- using standard silicon chip construction techniques. This depends on the spin of those single electrons being modified -- spintronics, in other words, which until now has been a strictly lab-based technology needing many white coats per working device. So far, spintronic systems have been limited to one or two working examples: these guys say that a quantum computer with a million devices is perfectly possible with today's tools. This could be the first quantum computer that can do mathematics that can't be done any other way, according to the researchers. It's a good birthday present. Friday 9/08/2002
Poor old Graham 'Don't Call Me Norton' Wearden, our broadband reporter and scourge of the dodgy telco. BT offered him ADSL free of charge for six months, so he could find out just how wonderful it is, and think warmly of BT the while. Our man G, not being one to shirk on his research, happily acquiesced to this offer, And lo, it came to pass that a man came out, stapled bits to Wearden's walls, and there he was. Broadbanded. And then a couple of months later, off it went. Not a bit to be seen. No flashing lights, no half-meg a sec of business critical information flooding in (that was what you were downloading, Graham, wasn't it?), nothing. He phoned BT and was told "Well, the bill hasn't been paid". Graham, being firmly of the opinion that this was the whole idea, patiently explained. That was around a month ago. It would be nice to say that BT has sorted it all out by now, but it would be a lie. There's only one thing worse than not having broadband, and that's having it only to see it taken away again. And there's only one thing worse than having a journalist suspect you're a company unable to find your RS-232 lead with both hands, and that's demonstrating it beyond doubt. I would be with BT myself, but the company told me that I couldn't keep my line when I moved from one DSL enabled exchange to another, a mile down the road. Well, they told me eventually, after making and losing a couple of appointments... nice to see that in these turbulent telco times, some things never change.