Some things are not politically correct to say, so lots of stats are presented in support of rash generalisations that deny those things. Often, they’re contrived to support a pre-conceived argument, rather than constructed to shed light on the matter.

That the environment is not the biggest crisis of our time is one case in point, which I’ve discussed before. You’ll see plenty charts that show, for example, what a 20-foot rise in sea levels would look like on a map. Or how hurricanes have increased in frequency or intensity. A close look at the data, however, and you’ll see signs of contrivances, such as selecting end-points for data series to support the pre-conceived argument, or presenting speculative conclusions as factual data. And how the environment manages to recover from catastrophes — such as volcanoes, meteor strikes, earthquakes, nuclear testing — when it really is a fragile system, perched precariously atop an unstable equilibrium, and sensitive to every little belch and burp of human life, is left as an exercise for the reader. How humanity came to be so prosperous and well-fed if our production methods are so stupidly unsustainable, ditto.

But let’s look at a very different example. Let’s consider Microsoft’s success. It isn’t politically correct to suggest that the monster from Redmond occasionally makes smart business decisions. How it came to be a large, successful company that made more millionaires than most is left as an exercise for the reader. How a a billion PCs came to be installed around the world in 20 years, a majority of them running Microsoft’s terrible products, ditto.

Now granted, I’m not convinced that its bid for Yahoo! was a smart business decision. It smacks of desperation. It has spent ten years trying to find a revenue stream to which the Windows-and-Office cash baton can be passed, but no convincing candidate has yet appeared. In its search, it has a habit of buying up second-best players in a market segment, and then swamping them with Microsoft branding until they’re never heard from again. My own prediction on Microsoft is that ten years hence it will be known as a very good gaming company.

But take a look at this analysis, by Max Freiert, highlighted recently by the folks over at Junk Charts, who spend their lives debunking — often in the most entertaining fashion — statistics abuse by companies, governments and the media.

It shows this chart:

Then Freiert assigns value only to new customers, which makes the $45 billion deal appear like a valuation per customer of $1 200. This is patently ridiculous. But I suspect it is designed to look ridiculous from the outset. That’s why it does not take into account that Microsoft might derive additional value from customers in the overlap area, or might offer additional value to its own customers, and that these might all lead to higher revenue for Microsoft.

If we redo the calculation, but with the (equally arbitrary, but more realistic) assumption that Yahoo! customers who also use Microsoft properties are worth half as much as new customers who didn’t use Microsoft properties before, and customers who don’t use Yahoo! at all have no additional value at all, we get a per-customer valuation of less than $200. And if you postulate that the deal might result in the ability to offer new revenue-generating value to customers that didn’t use Yahoo! properties before, at the arbitrarily-selected rate of one-third of the value of new customers from Yahoo!, you get $150 per customer, overall. Does that sound more reasonable?

But what about the new customer numbers? Does the chart above really reflect visually how many new customers Microsoft gets? How small is that top slice really? The guys at Junk Charts, instead of mentally calculating the area as a reader is meant to do, simply recast the data as a bar chart, as a statistician might do. It’s not pretty, but it does make the point: actually the growth in customer numbers is pretty darn decent.

Here’s their reworked chart:

Check out the last line of the table in Freiert’s analysis, and you’ll see that his result is the same: new customer growth will be a substantial 31%. And if you look in the next column, you’ll see a 89% projected rise in page impressions. In fact, Freiert makes a big deal of this growth in Microsoft-owned traffic later in the analysis. But if only the new customers had any value, why the massive discrepancy? This merely confirms that assigning value only to those new customers is a ludicrous assumption, apparently designed to support a pre-conceived conclusion.

The growth in customer numbers is larger than Freiert’s pie chart suggests, and the actual price paid for potential new revenue is much, much lower than the $1 200 he puts in the headline. None of this shows that the Yahoo! acquisition really was a smart acquisition, but it does show that at least part of Freiert’s neatly contrived argument for why it may have been a daft acquisition holds no water.

It was a daft acquisition for other reasons. I could create a chart to prove it, but I fear you wouldn’t believe me.

The New York Times yesterday published a damning column on the experience of some key Windows Vista users:

Here’s one story of a Vista upgrade early last year that did not go well. Jon, let’s call him, (bear with me — I’ll reveal his full identity later) upgrades two XP machines to Vista. Then he discovers that his printer, regular scanner and film scanner lack Vista drivers. He has to stick with XP on one machine just so he can continue to use the peripherals.

Did Jon simply have bad luck? Apparently not. When another person, Steven, hears about Jon’s woes, he says drivers are missing in every category — “this is the same across the whole ecosystem.”

Then there’s Mike, who buys a laptop that has a reassuring “Windows Vista Capable” logo affixed. He thinks that he will be able to run Vista in all of its glory, as well as favorite Microsoft programs like Movie Maker. His report: “I personally got burned.” His new laptop — logo or no logo — lacks the necessary graphics chip and can run neither his favorite video-editing software nor anything but a hobbled version of Vista. “I now have a $2,100 e-mail machine,” he says.

It turns out that Mike is clearly not a naïf. He’s Mike Nash, a Microsoft vice president who oversees Windows product management. And Jon, who is dismayed to learn that the drivers he needs don’t exist? That’s Jon A. Shirley, a Microsoft board member and former president and chief operating officer. And Steven, who reports that missing drivers are anything but exceptional, is in a good position to know: he’s Steven Sinofsky, the company’s senior vice president responsible for Windows.

Ouch. That must sting. Especially when it constitutes evidence in a class-action lawsuit on behalf of plaintiffs who bought PCs labelled as “Windows Vista Capable”, which turned out not to be Windows Vista capable.

Do check it out. It’s complete with PDF copies of the internal e-mails that say things like, “we set ourselves up”.

The Wall Street Journal had an interesting front-page story recently about Nicholas Negroponte’s One Laptop Per Child (OLPC) project. This project isn’t about laptops, you understand — it’s about education. But mostly, it’s about laptops. Building little green-and-white machines called “XO” that look like toys, to be sold for $100 to kids in poor countries. Problem is, the XO isn’t turning out to be very popular. So who do we blame? Microsoft, of course! And Intel!

With a subheadline one might expect from a newspaper that engages in left-wing editorialising, rather than hard business reporting, the Wall Street Journal agrees with him: “How a Computer for the Poor Got Stomped by Tech Giants”.

How, you may ask? Well, by supplying computers for the poor themselves. This, you see, is a bad thing. It appears to be less about computers for the poor than it is about who gets to wear the halo. Can’t have corporate profiteers nick Nick’s halo, now can we?

Mr. Negroponte’s ambitious plan has been derailed, in part, by the power of his idea. For-profit companies threatened by the projected $100 price tag set off at a sprint to develop their own dirt-cheap machines, plunging Mr. Negroponte into unexpected competition against well-known brands such as Intel Corp. and Microsoft Corp.’s Windows operating system.

Negroponte does say he thinks it’s a good idea for people to sell cheap laptops:

“I’m not good at selling laptops,” Mr. Negroponte has told colleagues. “I’m good at selling ideas.”

“From my point of view, if the world were to have 30 million” laptops made by competitors “in the hands of children at the end of next year, that to me would be a great success,” he said in a recent interview. “My goal is not selling laptops. OLPC is not in the laptop business. It’s in the education business.”

Good thing too, because as the article makes clear, Negroponte’s laptop actually costs $188, and prospective customers are balking, turning to alternatives such as Intel’s Classmate, which doesn’t cost much more, is backed by a large company, and comes with Microsoft Windows. Other companies are also eyeing the huge untapped markets in the developing world.

“The Intel machine is a lot better than the OLPC,” says Mohamed Bani, who chairs Libya’s technical advisory committee but doesn’t have the final say on buying laptops. “I don’t want my country to be a junkyard for these machines.” Libya has decided buy at least 150,000 Intel Classmates. The future of the One Laptop program there is now uncertain.

… Nigeria, for example, so far has failed to honor a pledge by its former president to purchase one million laptops. That’s partly because they no longer cost $100 apiece, says Tomi Davies, a Nigerian-born technology entrepreneur who helped Mr. Negroponte set up talks with Nigerian officials.

… The higher price also has made the laptop vulnerable to competition from sellers of more traditional, Windows-based machines. For many education ministries, “it’s a no-brainer you go with Microsoft,” says Mr. Davies.

But that doesn’t stop him from complaining about the competition from alternatives such as Intel’s Classmate laptop:

Mr. Negroponte says he communicated this month with Intel’s chief executive, Paul Otellini, and demanded that Intel stop selling the Classmate. Intel, which says there is room in the market for many machines, has refused, according to a spokeswoman.

… “We can’t compete,” complains Ayo Kusamotu, One Laptop’s attorney in Nigeria. “The minute we started getting some traction, they [Intel] intensified their effort.” Nigeria recently agreed to purchase 17,000 Intel Classmates.

In May, Mr. Negroponte appeared on CBS’s “60 Minutes” and blasted Intel, suggesting it was trying to drive his nonprofit out of business. Intel’s Mr. Barrett called that idea “crazy.” Two months later, Intel announced it was joining One Laptop’s board. The agreement included a “nondisparagement” clause, under which Intel and One Laptop promised not to criticize each other, according to Mr. Negroponte.

He claims, disingenuously, that the competition actually raises prices by not permitting his projects to achieve the anticipated economies of scale. What I’m seeing is that even his higher price point of $188 is sufficient to drive the prices charged by other producers down.

I’d think, if the purpose of the OLPC project is not to sell laptops, but to promote education, Negroponte would be delighted that so many private companies have picked up on his idea and realised that they can, indeed, provide cheap computers to developing-country customers. After all, a non-profit surely exists to serve a public purpose, not to compete against the private sector.

He may well dispute the arguments by customers about why the competing machines are better, and he may well have a point. I’d recommend open source software for educational computers myself. If he’s unable to deliver on his promises, when private companies can, the market is no longer failing. If he’s unable to convince prospective buyers, on price, performance or features, bitching in the press about competition strikes me as a desparate act to retain control of the market. Why try to shut up companies that are merely trying to promote their own products, unless you don’t think your own product can win on merit? Such tactics are hardly consistent with the supposedly altruistic motives of a non-profit organisation.

He’d gain a lot more respect from me if he’d stopped at declaring victory: “See? I told you it could be done!” But sadly, his slip is showing, and he just has to go on to reinforce the canard that non-profits are good and for-profits are bad. All he does is show that academics and non-profits are pathologically prejudiced against the power of free enterprise. Even when the empirical evidence in the market contradicts them.

Since I’m laid up in bed with a rather annoying flu that managed to switch off both body and mind, I figured I’d pass the time and break the silence with links to the half-dozen most popular posts on my blog, as per Google Analytics. They represent a gratifying mix of subjects, ranging from the environment to economic theory, from social networks to media freedom, from silly bureaucrats to great press photographs. In order of popularity:

1. 10 reasons to reject global warming — A summary of why I can’t accept the orthodox view that global warming is a crisis that requires large-scale government intervention. This item has not only been the most popular, despite being published only three weeks ago, but it recorded a surprisingly high average of 24 minutes spent on the page. It was a follow-up to a column published in Maverick magazine, entitled Global warming is a hoax. In some ways, the second attempt turned out to be the column I had actually set out to write.
2. Child labour: the baby dragon — This was a response to a question asked in the comments to an earlier post, which simply argued that import restrictions on Chinese goods, while protecting narrow interests, are not in the broad interests of South African consumers. “But what about child labour?” came the question. My response, namely that the description of such practices is an over-generalisation, that blanket condemnation is simplistic, and that either way, our objection can better be expressed in individual, targeted, specific boycotts rather than state-enforced punishment against an entire foreign country at the cost of local consumers, prompted a fair bit of outrage. As it would, when you see things only in black-and-white, and when every problem only has one, statist, solution.
3. This is a poke-free zone — Despite deriving some benefits from Facebook, the popular social network that attracted hundreds of thousands of South Africans in the space of just a few months, the signal-to-noise ratio had been declining, and I vowed to leave for good the day Microsoft got involved. It did. I left.
4. Info Scandal II — A cautionary tale about what happens when politicians and civil servants own media interests and try to buy out a major newspaper critical of the government. A follow-up post noted a significant difference, pointed out by Anton Harber, between the proposed buyout of Johnnic Communications (soon to be called Avusa) and the original Info Scandal of 1978.
5. The candyman can’t — Who needs to invent jokes when politically-correct bureaucrats will hand them to you on custom-printed signs?
6. The life and death of Kevin Carter — An old article about the late Pulitzer Prize-winning photographer stirred recollections of the years of South Africa’s transition to democracy. It occurred to me that many of my memories from that time aren’t memories at all. They’re Kevin Carter’s photographs.

Of these, my own favourite is the Kevin Carter piece. Like the item on William F. Deedes and the post on Isambard Kingdom Brunel (and the follow-up column it sparked), they reflect the pleasure I take in history and the great people that populate it.

The rot has begun. Scarcely has Microsoft bought its slice of Facebook (over which I left), than advertising is starting to appear. Right there in the news stream; a large, glaring advertisement. This is what it looks like:

A few things are notable. Unlike the Facebook Flyers, which appear on the left-hand-side, very visible but without wasting space needed for the news stream, and unlike Google-style text ads, which are non-intrusive and take up little space, this advertisement appears slap bang in the news stream, where interesting updates from friends vie for screen real-estate even on a very large screen. I’d guess on a typical notebook screen or a mobile device it would take up a good proportion of the visible screen space.

It is disguised as a real update, as if a friend just posted some new photos. That’s devious and offensive. Magazines (credible magazines, at least), decline advertisements that attempt to appear like regular editorial, since this hurts the integrity of the publication. I were in charge of Facebook advertising, an advert that looks like a legitimate update would be declined.

It appears to be completely untargeted. Credit reports comprise one of the biggest categories of online spam. What next? Pump-and-dump share schemes? Invitations from hot babes looking for money, honey?

It is true that some other social networks — including Orkut in particular — suffer from spam problems, especially in group discussion forums. This is something they will have to combat if they intend capitalising on the discontent created by Microsoft/Facebook deal. But at least they don’t (as far as I’m aware) condone the spam. At least they don’t place the spam themselves, where it clogs up an already-cluttered news stream.

In all ways, this particular Facebook advertisement is offensive and sends a clear message of where Facebook is going: more clutter, more noise, less signal, less usefulness. It lacks even the minimal redeeming quality — unintentional humour about Yahoo! spam filtering — of this inviting offer I recently received:

I’m super-mad at a super-poke. Was El Reg ever right. You can never leave Facebook. Wrote Chris Williams:

The funniest thing about such groups is that you can’t actually leave Facebook. Ever. The closest you can get to the escape hatch is a temporary deactivation. As Zuckerberg whispers in your ear as you grab the handle: “Even after you deactivate, your friends can still invite you to events, tag you in photos, or ask you to join groups.”

You can opt not to receive emails telling you about it, but your data cadaver is still there, waiting to be reanimated. Spooky.

Well, technically that’s true. You can indeed opt not to receive e-mails telling you about it. It’s just that nobody’s listening.

Facebook doesn’t actually heed that request. When I deleted my account, I took great care to heed the warning, and to choose not to receive any futher e-mails from Facebook. What do I need notifications for if I need a Facebook account to see what they’re about?

Within hours, I had received e-mails notifying me of a private message, that I got tagged in a note, and that someone sent me a super-poke. Even before leaving, I had told Facebook that I wasn’t interested in the damn super-poke application. So that’s twice Facebook told me to go poke myself.

The link that claims to allow me to “control” what e-mails I receive from Facebook requires… re-joining Facebook.

The spamming sods deserve Microsoft. I expect a video raspberry from Mark Zuckerman next, for falling for his phishing scam.

I’m off to update my spam filters. Facebook is now officially evil.

I promised I’d leave Facebook if it did a deal with Microsoft. I said I’d make good on this promise.

Some impatient souls (or should I say Facebook Fanboys) weren’t even prepared to grant me a few days to tie up loose ends and inform friends and group members, but insisted that I delete my Facebook account instantly.

This update is for them. My account is now gone. To be accurate, it is available for perusal only by Mark Zuckerberg and Microsoft.

I can be found on Orkut, LinkedIn and Plaxo Pulse. I’ve never liked the idea of using more than one social network site, however, so eventually I’ll choose. I quite like the update stream and look-and-feel of Plaxo, but in the end I’ll probably settle on Orkut. Especially since it’s due for a re-launch and overhaul starting next week. I can’t wait.

It’s been a good four months. I watched the South Africa network on Facebook grow from 100,000 to 400,000 members, and missed Miss South Carolina by minutes. Maybe she is on Orkut. That’s where I’m going.

Adieu, Facebook. I never was sure about the propriety of the Facebook feature that copied your blog posts as notes. But then, I never was sure about the propriety of poking random people either, or giving them growing gifts.

If the notes annoyed you, my apologies. If you enjoyed them, thanks. Soon, however, they’ll be no more. If you kept track of these posts via Facebook, you’ll have to subscribe to my RSS feed instead, or become a guinea-pig for the brand new e-mail subscription service on your right. (If it doesn’t work as expected, please do let me know.)

Microsoft is buying a teensy sliver of Facebook for a whole lotta dough, with the intention of advertising at me. Personally. It thinks it paid $300 for that right, but I’m not up for that. The deal is all over the news, and as usual, The Register has by far the best headline and funniest take on it.Shortly after I blogged about the impending deal, in which I explained my deep misgivings about the prospect of doing business with a company whose products, privacy policies and security record I don’t particularly like, and whose online services I’ve long vowed never to use again, I made a public promise. If Microsoft buys a stake in Facebook, I’m leaving.

I’m not saying that Google is any less of a privacy risk, but I sold my soul to them a long time ago, and to date, it hasn’t burnt me. There’s no turning back now, and I have no spare soul to sell to Microsoft. Call it a risk exposure minimisation strategy. Orkut was all the rage in 2004, when I last tried it, but it was a dog. It’s been groomed a little since, had its nails clipped and stuff, but still seems to enjoy some canine capriciousness. I’ll get used to it. Someone, somewhere in the Googleplex, must be paying just a little attention to Orkut, surely?

I’d better get used to it. Because in a few days, soon as I’ve informed everyone, handed over the groups I manage, and backed up whatever data I have on there, I’ll never point my browser at facebook.com again. Maybe eating the cookie will make me feel better.

Adieu, Facebook.

Update: I’m now on Orkut. Apparently, it’s up for a relaunch on 5 November, so I’m looking forward to see what the Googledroids have conjured up. If you’re there too, you can find my profile here.

“Gates war funding request increases by a third”, reads a CNN headline. The figure in question, for both Iraq and Afghanistan, is $190 billion. Another third, and Microsoft’s entire market capitalisation is on the line. Despite a huge private aircraft they keep at a federal airport near their office, Messrs. Page and Brin have a ways to go to match that kind of capital.

There’s an interesting battle brewing that may decide the fate of Facebook, the hugely popular social networking site. The country network of which I’m a member, South Africa, has tripled in size to 300 000 in just three months. I didn’t know there were that many internet connections over here.

However, there’s a dark cloud on the horizon. A very dark cloud. Microsoft is, according to the Wall Street Journal, in talks to buy a stake in the startup:

Microsoft in recent weeks approached Facebook with proposals to invest in the startup that could value the fast-growing site at $10 billion or higher, said people familiar with the matter. If those talks bear fruit, Microsoft could purchase a stake of up to 5% in the closely held startup, at a cost in the range of $300 million to $500 million, the people said.

But Microsoft must first outgun Google, which has also expressed strong interest in a Facebook stake, according to people familiar with the matter.

Microsoft’s Passport signon technology (now rebranded as Live ID) has proved to be wide open to abuse, and not only by external miscreants. When Microsoft bought Hotmail almost ten years ago, the webmail pioneer turned into a sluggish performer and a hotbed of spam. As this page documents, Microsoft itself had for years been both negligent and willfully complicit in some of the abuse. On one occasion it changed, without notification, all users’ preferences to share information with third parties, for example. On another, it tried to claim copyright on everything sent via Hotmail. It certainly has not been particularly respectful of users’ privacy, and has burned its trust relationship with its more savvy customers.

I’m sure Microsoft has tightened up its privacy policies by now. It’s appointed a Chief Privacy Officer and its PR machine makes all the right defensive noises. However, a 3 500 word policy can hide many secrets. My reading of its copyright notice suggests that it still claims an exceptionally broad licence to copy, use and sublicence anything you post on any Microsoft service, even if it is intended only for a private community.

So I vowed never to use any Microsoft-owned online service — MSN Messenger, Windows Live, Hotmail — ever again. Publications that required Passport Network registration were simply dropped from my reading list.

Facebook is already over-cluttered with applications. Some are useful, some cool, some annoying, and some just downright offensive. I don’t mean in the prurient sense; I mean in the spam hotbed sense. I usually decline to install them, but I accepted a fun one involving beer just yesterday. Contrary to explicit instructions not to, it invited a random selection of friends, some of which I really didn’t want invited. This kind of spamware can kill Facebook.

But not as quickly as Microsoft can. If Google buys Facebook, I’ll live with it. The Googleplex 0wnz me already, and I’m not even a heavy user of its services. However, it has yet to show the kind of negligence or nefarious activity that will compromise my trust. For now, the convenience of its online tools outweigh the very real privacy risks. But if Microsoft buys Facebook, I’m outta there like a shot. The Hotmail fiasco alone was enough for me to never trust Microsoft with private information of any sort again. Through negligence, incompetence and deliberate action, Microsoft has abused the trust of users too often in the past. Here’s hoping Facebook doesn’t become the latest victim.

Update: In good Facebook tradition, I’ve created a group: If Facebook sells to Microsoft, we’re leaving.

Used to be the CIA spied on the KGB, and vice versa. In the post Cold War world, the black and white hats appear to have been passed to Microsoft and Google.

It appears Microsoft Virtual Earth captured a picture of a super-secret stealth propeller design (see alongside — publish and be damned, I say) on an American Ohio-class ballistic missile submarine in the graving dock at the naval base at Bangor, Washington. A couple of months ago, Google Earth took a snap of what appeared to be China’s new nuclear submarine, so the Chinese won’t be feeling so stupid anymore.

Monster Maritime has more details, including a picture of how the Navy normally keep its secret prop under wraps. The comments section contains some more interesting places curious spies might want to check out.

Monday morning is a bad time for nostalgia. C|NET News.com did a fascinating story on the Digibarn, a private collection of curiosities. The proprietors, Bruce Damer and Alan Lundell, obtained some of their exhibits straight from some of the biggest names in the development of the personal computer.

Just about everything is there, from old mechanical devices for pilots, scientists and race car engineers, to the seminal Xerox Alto. From the popular Commodore 64, PET, and VIC20, to the IBM 5150 (aka the IBM PC). From a 1979 granddaddy of the iPod,  to evidence implicating Microsoft in work done for Apple in 1977. Successes, failures and oddities document the rise and rise of the ultimate geek toy.

Alongside is an autographed Apple II computer. I used one of those during school athletics meetings, before I decided to rewrite the software for the IBM PC. It would become my first commercial computer program, dedicated to my friend and co-designer, Frank Tintinger, who died at the age of 15 from a rare, invisible, but evidently serious condition.

Do check out the picture gallery.