Rupert Goodwins' Diary

Monday 24/4/2006

It's happening already. Not only does it look like HSPDA is going to eat WiMax's lunch in Asia according to a report in today, but we've had our first 3G enabled laptop into the Reviews labs. When will we see any WiMax product? Your guess is as good as mine, and I'm guessing 2008.

I hold no particular brief for either technology. I like HSPDA because I've seen it work, and I like WiMax because it seems a sound, cost-effective way to deploy broadband where broadband is difficult to deploy. I think it's a shame that WiMax has been oversold for other tasks, and that this will result in problems for it in the near future that it really doesn't deserve to have.

What will be upsetting is if WiMax or its descendants are such a commercial failure that they lose developmental inertia. We're going to need it badly if certain moves in the US are successful, and if they then spread further abroad.

Over there, where some of the political machinations of the IT industry are terrible to behold, some of the big ISPs are lobbying to be allowed to preferentially drop traffic they don't approve of. We've already seen Verizon and friends saying "Why should we have to pay to carry Google's traffic?", a stance that merrily ignores the fact that everyone who connects to the Internet pays for their bandwidth already; now we've had the US government preparing to allow the ISPs to act on such ideas.

If unchecked, this will lead to a rapid dismantling of all the things that made the Internet work and a return to the old days of closed networks with high costs of entry. Why shoud Verizon carry Google's traffic if it thinks it can force its subscribers to use its own search engine? And how will search engines work if they can't search across the entire Internet? Hey, who cares if an ISP's customers are restricted to finding stuff hosted by that ISP? More revenues for that ISP, right?

Let nobody imagine that the big telcos are too smart to let that happen. Stupidity and greed are still not endangered species at the highest levels of corporate America. And if it does, is that it? Are we lost? Back in the clammy embrace of old-school telcos?

Not if we can deploy more networks that don't need massive infrastructures — and that means cheap, powerful, unrestricted mesh systems — exactly the sort of thing that WiMax and its fellows are designed to become. The technology's there, and the regulators — at least in Europe — are moving towards the opinion that less regulation is better, meaning we'll be able to deploy the stuff we think we need, not that someone else has decided for us.

Let's hope that if the worst comes to the worst, we've got the tools to fight back. WiMax, we'll need you. Don't lose the early fights.

Tuesday 25/4/2006

Can you guess what it is yet? Intel had been strangely circumspect about today's big announcement: the business platform to match Centrino in portables and Viiv in home systems. That's understandable; anyone who's been paying attention for the past year could probably give the PowerPoint presentation with their eyes shut. We didn't know the name — although the good old Inquirer managed to do a weekend ferret and found a site called www.vPro.com that linked to Intel's front page. So vPro it is.

That's a mildly odd name. As connoisseurs of wild media will know, vPro is the name of a Dutch public service broadcaster, Vrijzinnig Protestantse Radio Omroep or the Liberal Protestant Radio Broadcasting Company. It's a lot more liberal than Protestant these days, and specialises in top weirdness; it's a bit like Intel calling their new business platform Radio Odd Stuff.

We'll assume that this is pure coincidence.

But the strangest part of vPro is the specification. We don't know what it is. We know most of it — the VT virtualisation is thoroughly documented, the dual-coredness is as public as you like and the graphics, memory and other aspects are pretty much sorted. That's what you'd expect for a platform that's built on technologies that are all announced and out there already.

But the management aspects are less well understood. We've asked, but so far — nothing beyond marketing speil. That's a shame: unlike, say, Vista, vPro has a solid reason to exist and many of the claims Intel makes for it are liable to be both true and useful. You don't need to be told how unusual and important that is in this business.

With Intel preparing for another big internal shake-up, the worry is that it's about to change its ways about information and move to a "licence or stay ignorant" model in an attempt to keep their rivals in the dark.

Let's hope not. We're on the case.

Wednesday 26/4/2006

So there I am, typing away, minding my own business, when some internal disruption deep in the Stygian darkness of my computer bubbles to the surface like the miasma from a decaying frog.

"Rules in error" says the box. "Rule: Server Requested Client Action. Error: Unable to find destination folder. The folder may be deleted or..."

Because that sentence is too long to fit and the box cannot be resized, I have to use a horizonal scroll bar to read the rest: "... in a PST Outlook cannot open."

That's it. There is one button to click, labelled Cancel. After trying various tricks to see if there's any more information hidden away, I give in and click. The box vanishes. I check Outlook. There is no sign that anything has actually gone wrong.

If Microsoft had tried harder, they could have broken some more cardinal rules of usability — but for a first try, it was very impressive. The error contained no information I could act on. It had no context to help me diagnose the actual fault. It had no suggestions to help rectify the problem. The information was presented clumsily and in a way that actively prevented me from investigating further (the text couldn't be copied to the clipboard, for example, or into a search engine). The single action I could take was ambiguously labelled — what, exactly, was I cancelling? And I still have no idea whatsoever whether the thing was important or not. So why bother?

I don't know why, after all these years, I should still feel actual anger at Microsoft's incompetence. A healthier attitude would be to ignore it, and an even healthier one would be to move on to other software — having used Ubuntu for a month or so on a spare laptop, I'm going to make my main machine dual boot with the eventual aim of relegating Windows to occasional testbed.

Yet anger it is. It's fuelled this week by Microsoft going to the EU and blustering about innovation, about others "having a free ride" on its hard work. It's the company saying that "functional equivalence" needs people to "go far beyond interoperability" — in other words, when you say you want to work with us you really want to rip us off, so we're not going to even let you start. Isn't that what psychologists call projection? It's people like Ballmer boasting that he's "brainwashed" his kids into not having iPods or using Google: ban the competition, don't learn from it.

If Microsoft's unvarying monochromatic belief in its own divine right to do what it liked actually worked, then at least one could say that the trains ran on time. But it doesn't work. We're left with the running joke that's Vista, with the tiresome spectacle of a convicted monopolist preaching about morals to a judge, and with software like Outlook behaving in a way that'd get a first year programming student demoted to media studies.

Thursday 27/4/2006

We all know the stories about brand names gone wonky. How Coca-Cola in Chinese transliterates to "Bite the Wax Tadpole", how the Nova car meant "no go" in Spanish, how Google sounds like "the stench of a rutting warthog" in Mongolian. Some are true, some urban legend, all illustrate the pitfalls of cross-language branding.

My favourites involve bodily functions. GEC Plessey Telecommunications, an august institution used to working in a dignified fashion with important people in governments, might have thought it was merely acquiring a modern sheen when it rebranded itself to GPT. Pronounced in French, though, those three letters amount to a confession of personal gaseous emission — ah, a schoolboy howler of the first order. Still, anyone can make a mistake.

It's hard to see Nintendo's choice of name for its new console as an accident. The console previously known as Revolution — nothing wrong with that — has now been christened. It's a little difficult to imagine the meeting where senior executives gathered excitedly around their new baby while the CEO solemnly intoned — "I name this child... Wee." Sure, it's spelled Wii and it's tempting to suppose it should really be pronounced "Why?", but it's wee. Nintendo insists.

Let's get the jokes over with. Yes, it'll be hard going into a retailer and saying "I need a Wii" or "Have you got Super Monkey Wii"? One will need chutzpah to come into work on a Monday and say "I've spent all weekend playing with my Wii" or "What I like about Wii is the way the controller feels in your hand". There will be more.

Linguistically, it's interesting. There are only two previous words with "ii" — skiing, and the recently coined yet scarcely less riisible Viiv, That's pronounced "vyve", making all three examples incompatible. Perhaps Intel and Nintendo can form a group and settle on an IEEE international standard to sort it out. If it's anything like recent efforts it'll end up with 802.ii.a — "To rhyme with Eeee", 802.ii.b — "To rhyme with Aye", and 802.ii.c — "To rhyme with Being". We can then write article after article about "standards group torn by division", PRs can brief us over expensive lunches about why choice is good for the market and the other lot are just dangerous bandits, and everyone else can sigh miserably and go out to buy something completely different.

I don't suppose it matters. The gruffest Glaswegian can say "Ah fancy a wee sweetie" without feeling his manhood impugned, people take the EC Directive on Waste Electrical and Electronic Equipment seriously enough, and our octogenarian Windsor employs the Royal We at all times. Even Winny the Pooh got over it.

Now, about this vPro...

Friday 28/4/2006

"We'd love to find the dangerous criminals we've accidentally released from prison", says the Home Office, "but we can't find them on the police computer." "We'd love to MOT your car," says the garage, "but the Department of Transport system's down and nobody's saying why." "The NHS IT project is going to lose us $450m (£252m)," says Accenture, "because so much of it is late and people are going elsewhere." What, no love?

Yes! These are the people I want to run a national ID database. These are the qualifications that count, the experience that matters, the guarantee of efficiency, safety and responsibility that means the nation's biggest, most expensive and curiously least-well specified project will do us good.

I'm particularly impressed that now the law's through, Charles Clarke is confident enough to move on from the old, under-performing specification of "no medical information will be included on the database" to "medical information will be included on the database. Voluntarily." — and I have every expectation that "voluntarily" will come to mean what it does for the cards themselves: compulsory in every sense that matters.

Something that puzzles me, though. All those foreign prisoners who were let go instead of being considered for deportation — lots of them are apparently untraceable because they gave false nationalities. I know it's a shock that criminals might lie, on top of everything else, but apparently they do. So if someone claims to be a foreigner from, say, Papua New Guinea, when challenged for their ID card, what happens? Silly of me to wonder, of course, the ID card people are bound to have thought of that — and if they haven't, they can ask the Prison Service how they coped with the problem.

And with that, it's another bank holiday. This weekend, I think I'll pretend to be called Jeremy Thompson, the Irish son of two English ex-pats who moved to Cork in the 1960s. I used to work in the travel industry, but started a little IT consultancy of my own specialising in ticketing systems. You never know when you need a good back story — and it's never too soon to start.