Google PageRank is a term that's bandied about a lot in search-optimisation circles, and to an extent it's like referring to the Roswell UFO. Some people think they know exactly what it's all about, others are more doubtful, and the story given out by those actually in a position to know is incomplete and open to question.

There are some places on the web where you can check the PageRank of your site, and their results seem to tally with common sense. Whatever shadowy source they are using, both PRchecker.info and Page-Rank-Lookup.com concur that www.computing.co.uk scores 7/10 and that sister site vnunet.com scores 6/10. Our much less important domains achieve suitably lower figures. A holding page at www.businesscomputerworld.co.uk scores a lowly 2/10, for example.

But these results make me question whether the bald PageRank score is as important as the SEO people would have us believe. I have accurate traffic figures for all our domains and www.vnunet.com gets quite a bit more views than www.computing.co.uk, and indeed gets more visitors from Google search referrals. So there is more to traffic generation than meets the PageRank-obsessed eye.

It's also peculiar that defunct sites seem to hang on to their PageRank scores for a long time after they've expired. We ceased publishing Network IT Week and Network News years ago, for example, yet both of their home page URLs continue to notch up scores of 7/10 - better scores than the highly active and regularly updated vnunet.com.

It's all very odd. Spooky, even.

One of the many duties of the modern editor is that of lexicographer. You may have to decide, for example, if your publication will adopt e-mail in preference to email, you may be called upon to rule on when to accept a neologism such as "phish" without explanation for the uninitiated reader, and you will certainly have to decide what to do about potentially transient or contentious terms like web 2.0 (or Web 2.0, if you believe this much-despised term qualifies as a proper noun).

Historically, editors may have ruled on this stuff but they have rarely presided over it. That is typically the job of the chief sub editor, managing editor or production editor. To a sub worth his salt, consistency is an attainable goal. Know the rules of house style, apply them diligently, and you can reflect on a job well done.

But I wonder if this phase is coming to a close. For example, recently I was asked if a lengthy expository document ought to be called a white paper, white-paper or whitepaper. In years gone by I might have considered prevalence, precedence and reader confusion. Which alternative is most often used out in the world? Will one usage be clearer than another? Where is the term on the much-travelled road from two words to portmanteau? Today, by contrast, I'm just as likely to ask, what are people searching for?

Indeed, there's now a certain attraction in literary inconsistency, from a search-engine-optimisation point of view. SEO techniques depend on matching your means of expression to the method of deduction used by your target audience. In other words, you should use the same words your desired readers will choose when they stick a phrase in Google. But if your audience hasn't settled on consistent terminology, you might be wise to spread your bets.

So there may well be an SEO benefit in ensuring that our site lists white papers under white-papers, white papers and whitepapers.

Chaos beckons.

I'm relieved to learn that one of the banks I currently use, HBSC, is taking steps to beef up its online security. For a while I've been uncomfortable with the log-on process, which requires that you enter your 10-digit internet banking number (issued by the bank on a card, hard to remember, and so often written down elsewhere) plus a date of birth and a user-chosen security number. The security number is supposed to be six to 10 digits, but I'm guessing that most people (unlike me) will pick six digits and indeed will choose the birthday of a spouse, child or other relative.

I think this system is weak because (a) it's entirely numerical, greatly reducing its combinatorial complexity and (b) two of the fields are likely to be even further constrained to dates within the last 80-odd years.

If a thief were able to swipe your HSBC internet banking number and social-engineer your date of birth out of you, I don't think the remaining memorable number would last long against a determined assault. I don't know how many failed log-on attempts the front-end will allow: I'm reluctant to experiment with my own account and don't want to lock out anyone else.

Anyway, as with most security systems there is often an open window to try before breaking out the lock-picks. As I've been learning...

The HSBC is currently paying a lamentable 1.8% AER on its standard Cash ISA accounts, the rate having fallen recently in line with base rates. When it first dropped I asked if it might be possible to convert my ISA into one of the same bank's new e-ISAs, which currently pay 3.25% AER.

The answer can be summarised down to a short word that begins with the letter N. I was surprised, as I had thought that low rates on old accounts were there to snare the complacent, not to punish the loyal. Clearly the HSBC feels no need to keep my business, so I set about finding a better rate elsewhere.

I chose Abbey's Direct ISA, which doesn't have the absolute best rate but at least is a bank with which I have another account, making dealing with it somewhat easier. I filled out the necessary forms online, then downloaded them and filled them out again in pen when Abbey sent the wrong paperwork in the post, sent them off by Royal Mail and waited for the bulk of my worldly wealth to worm its way from HSBC to Abbey.

So far, two surprising things have happened.

Firstly, my HSBC ISA has disappeared. When I log on to my HSBC home page, the account is simply no longer there - no messages in my inbox or explanation on the page, it's just gone. This feels odd because the HSBC will often contact me by phone to check that it really is me paying for dinner or a pair of jeans with my credit card, but evidently it is happy to make changes to my largest savings account without a call, or indeed any credentials except the paperwork that's arrived via Abbey. I'm not sure I like this. It's not beyond the realm of possibility that someone might find the details needed to fill out a fraudulent transfer application and it must surely be worth the cost of a call to check.

Secondly, my new Abbey ISA has been created - but with a balance of £0.00.

At the start of the process, Abbey warned me that the transfer of funds can take up to 30 days, so I'm not panicking yet. But overall, my confidence in our financial service providers, low as it was to begin with, is currently plumbing a new and deeper depth.

The pen-written signature feels as archaic as a sword and scabbard in our digital age, but it unfortunately remains the standard means of committing to a contract. So it's as well to be cautious about who one gives a sample signature to. With identity theft on the rise it is probably no longer such a great idea to pay for things by cheque, for example. Never mind the hand-written forgery, it's a doddle for me to scan your signature and print it onto my fraudulent credit card application.

It's frighteningly easy for an identity thief to get a digital copy of your signature, too. Yesterday I sent a letter by recorded delivery, and the Royal Mail's online tracking service handily provided me with the digitally captured signature of the recipient, ready for me to download, as proof of delivery. So I can now anonymously and untraceably capture the signature of a targeted individual I've never met, simply by sending them a recorded-delivery letter (at a cost of £1.08). ID thieves will be pleased.

As I've written before, when asked for personal information there is often no need to provide the real thing. In future I think I will write my name in block capitals when signing for post.

I wouldn't normally blog on a Saturday night, nor indeed blog about BBC dance-show Strictly Come Dancing, but I have to vent frustration at the lack of mathematical ability in evidence. Never mind learning to dance, does nobody at the BBC know how to count to four anymore?

This evening was supposed to be the semi-final of Strictly, with three couples whittled down to the final two, ready for next week's finale. But at the end of tonight's results show, it was announced in confusing fashion that all three couples would go through, with phone votes carried through to the final.

No explanation was given on air and pundits clutching for an explanation have leaped on John Sergeant's early exit from the contest leaving numbers short. But while numbers were clearly the problem, the number of warm bodies was not it.

Strictly's voting works on a 50:50 basis - the on-set judges' votes count for the same amount as the viewing public's calls. With three couples dancing, this would mean a maximum of 6 points for topping both votes, with 3 coming from the public and 3 from the judges. The worst possible outcome would have been 2 points - one for coming bottom in the studio and one for coming last on the phones.

The spanner in the works was that on the night the two female contestants, Rachel and Lisa, tied in first place, earning 3 points each, leaving Tom in third on just a single point.

At that juncture, the show's producers ought to have jotted down some numbers on a pad and worked out that they should not have opened the phone vote for Tom. His maximum possible score, even if he won the public vote, would be 4. The two girls, meanwhile, could end on a score of 4, 5 or 6, but neither could score lower than 4.

In other words, while one of the girls could have avoided the dance-off, Tom could not, so his fate would have been decided by the studio judges. There was no point whatsoever in voting for him by phone this evening.

Fortunately for the BBC, someone must have worked out the numbers before the final cut. Sending all three couples through to the final was clearly deemed to be a preferable outcome to refunding the cost of all the calls received in futile votes for Tom. It was no doubt also seen as preferable to raising the ire of Ofcom again after last year's voting scandals.

I'm trying to imagine what the functional requirements for a password recovery system ought to be, never having thought much about the topic before.

Obviously you want the recovery process to be at least as secure as the normal password validation process. Recovery is, after all, effectively a back door and it ought to be at least as sturdy as the front entrance.

In general a security system relies on one or more so-called "factors" relating to the user - something that they know, something that they possess, or some biometric or other. To be really secure you need a two- or even three-factor system. But on the whole, on the web, for most purposes, validation typically boils down to a single-factor system based solely on what you know or can remember. It would be impractical, expensive and unwelcome if every web site you signed up for sent you a SecurID password generating keyfob, say.

So you hopefully know your password and username. If you don't know those, you should know your email address. And you ought to be able to consistently answer personal questions when challenged.

To my mind, the best web-site password recovery systems are those that let the user set both the challenge and the response to more than one personal question, perhaps with a few prompts for those left wondering what sort of query might suit. That way I don't have to provide information I might like to keep private, like my mother's maiden name, or my place and date of birth. Much better to let me decide if I want to ask myself what I paid for my first car, the exact colour of its rusting bodywork, what major item first malfunctioned, and what kind of immovable object brought it to an abrupt end.

But try as I might, I can't imagine what was going on in the heads of the people who set the validation requirements for Moveable Type, the blogging system as deployed by my employer.

For your everyday login it requires a username and password. So far, so normal. And if you forget the password, why there's the normal, helpful link saying, "Forgot your password?"

Clicking that helpful-seeming link is an education in security by obscurity, however. It takes you to a page where you are again asked for your username, plus your "Password recovery word/phrase".

So in other words, if you forget the password you normally use, you can recover it by remembering a password you never use. And if you forget your username you're out of luck.

I deduce that this system must have been dreamed up by a helpdesk administrator worried about the security of his or her job.

I'm quite pleased with this new site, which I built myself over the last few days as an experiment - the official technical team being somewhat busy with a yet-to-be-revealed wholesale redesign of another Incisive brand.

I've learned more about site coding in this past week than in the last decade.

The page provides a single place to sign up for the wide variety of email newsletters you can get from the Business Technology Group at Incisive Media. I'm the editor-in-chief of this group.

My old BlackBerry fell to bits, and I've been issued with a replacement. In terms of user experience, it's been a little like jumping from a Model T into a Mondeo - or perhaps the other way around. As you may or may not know, a Model T Ford has three foot pedals - although none of them is an accelerator or clutch. So it is that my fingers are having trouble unlearning the motions used to operate my old 7730, now that I have a shiny new BlackBerry Curve 8320.

It may be shiny, but is it any good? The 7730's side-mounted click wheel is gone, replaced with a "pearl" trackball, that has an action that makes me shiver in the manner of nails on blackboards. Alt-Return no longer locks the keypad. I no longer have a desktop cradle. Worse, the battery life is not as good, meaning a stone-dead BlackBerry on Monday morning where the old faithful 7730 had enough juice left to flick through a few emails on the Tube.

The 8320 is smaller and lighter, with a better screen, but the keypad is less good for typing. Bearing in mind it lives in my bag not my pocket and I also carry a phone, this newer BlackBerry is not much of an improvement. I think it's more likely to get nicked, too.

A case of two steps forward, three steps back.

On a related topic, in today's IT Week, Daniel Robinson laments the passing of Windows XP, now that Microsoft has withdrawn the operating system from the new-PC channel. "It is difficult to think of a single area where Vista is better than XP," Dan writes.

That may be fair comment but I'm still not 100 per cent sure I agree with Dan's prognosis. I recently shelled out some of my own meagre funds for a new laptop, and decided to buy with Vista preinstalled, rather than going for one of the many XP offerings still in stock.

This was not because I thought Vista was particularly great, it's just that I've conservatively picked the wrong OS twice before. I specified Windows 98 when a still wet Windows 2000 was generally considered power-hungry and pointless. And then I plumped for Windows 2000 on my next laptop when the gawky Windows XP was said by those in the know to have nothing to offer a business user. Both times I ended up regretting my decision as the newer platform became more established, proved more capable, and became more likely to work with new peripherals.

So twice bitten, thrice shy, as they don't say. This time I went for Vista.

I had a budget of £500 (not enough for a MacBook) and ended up mired in indecision, unable to decide among the incredible variety of notebooks on offer at this price point. In the end I plumped for a Samsung R20 Plus, although had I spent ten minutes' more or less time pondering I could easily have bought an Acer, HP, Sony, Dell or Toshiba for the same money.

What the R20 offers is a reasonable compromise between weight, size, and apparent sturdiness, a 14.1-inch 1280x800 screen, a keyboard with nice tactile feel for this budget, a 1.8GHz Intel Core 2 Duo processor, 2GB of RAM - and Vista Home Premium.

The hardware offers enough horsepower to run Vista well, complete with fancy graphical transparency courtesy of the ATI Radeon 1250M graphics subsystem, which is the weakest part of the whole box according to Vista's built-in slothometer.

As Daniel says, there's not much to write home about in terms of usability improvements. I like the miniature, real-time screen-shot that hovers into view when you mouseover items in the Windows bar at the bottom of the screen and, erm, that's about it.

I haven't had any problems with peripherals, and I also haven't experienced any issues with audio capture, processing and playback, both areas that have caused endless problems for early adopters like our own Tim Anderson.

Vista runs quickly enough, boots up rapidly compared to my XP work machine, awakens from hibernation in the blink of an eye, and hasn't crashed yet. So far the only annoyance I've spotted is that when my BT Broadband connection drops (as it tends to whenever the wind changes) the R20 sometimes can't find the internet after the ADSL router comes back online. This happens about one drop in 50, and can be cured with a reboot.

No doubt the Samsung's capable hardware would make XP absolutely fly, but it won't get the chance. It's stuck with Vista.

I wonder if I will be happy with my decision in a year's time - or will I turn out to have made three bad OS choices on the trot?

It's time for me to buy a new laptop, the old one having gone well past the point of pension.

There was a time when I would simply have bought the lightest Toshiba I could afford, but things have become so much more confusing since I last pried open my wallet, dodged the moths, and bought a Portégé.

As my colleague Daniel Robinson points out in his latest video review, the lightweight end of the notebook market is currently undergoing a schism. While some ultraportables I'd like to own are the wrong side of two grand, you can now also find low-spec subnotebooks with very low prices. These can even be had under a kilo and under £350, in the shape of the latest Asus Eee PC 900.

Interest in Asus's wee beastie is acute - at our recent Channel Expo show, quite a few of the attending resellers were huddled in the Asus stand and keen to see the Eee in action with its new and bigger screen. It's not often that an 8.9-inch screen is considered bigger, such was the pixie-sized nature of the older 7-inch Eee PC 701.

Despite its low RRP, there is still an attractive margin to be made on the 900 model. Asus has clearly worked miracles in its own supply chain to keep the wholesale price so low.

Lovely as the Eee PC is, I found the keyboard too small for my not-very-fat fingers. I couldn't touch-type reliably, which rules it out, sadly.

I got on better with HP's 2133 Mini-Note, which boasts a slightly larger keyboard. This is about the smallest keyboard on which I can comfortably type without errors, although it would no doubt be bad for the wrists. In the office I use a split ergonomic keyboard, but the Mini-Note would be OK for short bursts.

But while my wrists might get by, I'm not sure my eyes would. The Mini-Note's 1280x768 screen is crisp and colourful, but each pixel seems about the size of an ant's toenail. System fonts show up in hair-fine lines. It's just about useable for my ageing eyes, but I would prefer something with the same physical screen size and a lower pixel count - the 1024x600 screen on the Asus, in fact, would be ideal.

I can't hang around waiting for someone to marry the Mini-Note keyboard with the Eee PC's screen, so I guess I will have to buy something completely different.

Having said that, I've just sent back to Acer a 17-inch, desktop replacement behemoth that I've enjoyed using on long-term test. I loved the screen and keyboard on that. Now, if only someone could offer a 17-inch screen with a lightweight chassis that doesn't cost an arm and a leg...

Can it really be 10 years? Hard to believe but oddly true - 10 years ago today we published the first edition of IT Week.

I'm now the last person on the launch team still involved in the publication - then as software editor, now as editor in chief of a group comprising IT Week, CRN, BusinessGreen.com and Computing.

But while my role has changed out of all recognition, the spirit of IT Week is still the same. In 1998 we set out to create a business weekly that would be really worth reading, that would provide no-nonsense information for senior IT professionals - and that's what IT Week still provides. When we surveyed a representative sample IT Week's audience recently, 86 percent rated the print weekly as good or excellent, with only 13 percent ranking it average and only two percent giving it the thumbs down. With almost a quarter of readers choosing "excellent", we are happy to know that we are still providing a worthwhile service.

Looking back, I actually remember issue two rather better than issue number one. The first issue had a long lead-up, but number two was the first we actually wrote in a week (with hangovers from the launch party, too), so it was of course a lot more intense.

For issue two I wrote a news story about software giant Oracle that then-news-editor Martin Veitch put on the cover, about Oracle's questionable handling of the impending Millennium Bug problem.

Oracle didn't like it, not a bit of it, and Oracle's PR sent me to Coventry as result. For what must have been a year afterwards, calls for comment went unanswered and my name fell off the invite list for press conferences and events.

This cold-shoulder treatment made my job as software editor a bit tricky in those early months, but as IT Week's reputation grew it probably did Oracle more of a disservice than it did me. That kind of overreaction is, thankfully, a lot rarer among spin doctors these days. Blogging, which was of course unheard of a decade ago, has changed the perception of criticism quite a bit. Most companies have become both thicker skinned and more nimble and professional about addressing their own failings.

Talking of which, I'd be very happy to hear your thoughts on IT Week - good and bad. Tell me what we're doing right and what we're doing wrong, and we'll do our best to take appropriate action.