Dialling phone numbers is much less common than it used to be, thanks to electronic address books, skype, and so on. The risks of the internet are better solved by making systems easy to use, not by sitting around and wishing that users were better educated.

The car analogy is just stupid. We make people take a driving test because half a ton of steel at 70 mph can kill a lot of people very quickly without you doing anything obviously stupid. That's not a risk on the internet. My mother doesn't understand a URL, but the internet is still hugely useful to her (email, skype, online shopping, weather forecasts...). It's not your internet, and we're not kicking out 90% of the world when they're using it very productively.

And no, URLs are not "easy to understand". You start at the beginning, then jump into the middle and go backwards for a while, then jump back to the middle and read forwards. The obvious way to read them is left to right, so everything is a subdomain of www. They often contain odd codes ("sid=2004418"?). To properly explain it, you have to talk about servers (for many people, websites aren't on a server, they're just somehow "on the internet"), and static versus dynamic content. And now you can have internationalised domain names in various alphabets. To be fair, some browsers are now making it easier by highlighting the domain name in the URL bar (that is, improving the usability, not berating people for their ignorance).

Yes, Google make money. They provide a service that people use. That's capitalism for you. Wikia tried to make a non-profit search engine, but it didn't work out. I don't particularly like the idea of one company profiting from running such a key function of the web, but I don't think I could do it better...

Why do you think that people should type in URLs? If you accidentally type in paypall.com instead of paypal.com, you could end up getting your money stolen (in fact, it's a domain squatter - but the risk is there). If you type paypall into Google, you get pointed to paypal. Like IP addresses, URLs are becoming a technical detail that users don't generally think about (except perhaps for sharing links). It's not the end of the world.

There is a form that tells them to really delete it. They don't draw attention to it, and it's separate from the "deactivate account" option, but a search in the help centre does bring it up.

Well, if you have a smartphone, you could install the key generator app, which, since it doesn't require a signal, presumably doesn't report your phone number to Google. If you want to check that, you could always look at the source (for Android & Blackberry, at least).

I very much doubt they'll ever make it compulsory. It's just too much hassle for most people. If they really want to push it, they might show a banner ("Find out how to make your account more secure..."), but I bet it'll just sit on the settings page, only to be used by those who know they want it.

Do we really have nothing better to do than complain? It's an entirely optional way to add some security, and they do seem to have given it at least a modicum of thought (several methods to get the tokens, limited backup tokens if you lose your device...). What else should they provide? Free ponies?

I read it before it was /.ed. It wasn't obviously libel: he more or less approved of the atmosphere, said two dishes were alright, nothing great, then laid into them on the quality of a couple more dishes. Of course, I have no way of knowing if he was telling the truth, or being paid by a competitor, but it read like a negative review, not a hate-filled diatribe.

You miss his point. If you understand what's going on, you can happily switch over to LibreOffice and carry on without worrying. When you're management, and you don't follow open source software in your spare time, you hear "Most of the developers are jumping ship, and attempting to create a separate codebase in competition with the company that owns OpenOffice", and you think "that's not something we want to rely on for now. Let's just upgrade to the next version of Office. Maybe we'll give it another look in a few years." If you've got a particularly non-techy manager, they might even be wondering if this "forking" is entirely legal.

There's a lot in a name, and Microsoft know that. That's why they still release Windows and Office, even when they undergo a real change as dramatic as Office 2007. Libreoffice is a new name, so it will be considered as a new thing. And in a sense it is: the codebase isn't new, and many of the developers aren't new, but the process of writing it and releasing it is.

You know, I think any computer UI is likely to be a new one for many of the children they're targetting. They've got a rare chance to design an interface for people who don't already have expectations of how to use a computer. I know I'd take that opportunity to see if I could work out a better model.

Well, I have. And I've had pretty much no problems, no hoops to jump through. The worst response I've got was an 'unreferenced' message put on a new page I created. Which was quite accurate. I added one reference, and the page has seen no further problems.

Of course, the fact that it's still working pretty well isn't nearly as interesting as complaining about idiots reverting your changes, so carry on.

And other people like it. It's just that haters shout louder.

Actually, I prefer it. Extra space to display pages. I'm not necessarily sure that their solution to show URLs is better than the pop-up bar in Google Chrome, but I prefer either to the fixed status bar.

In other words (I see someone else agrees), you're just blindly assuming that everyone agrees with you. Has it escaped you that Mozilla might actually have done some UI testing? Or that people can have legitimate opinions that aren't yours?

I thought they fed back results, so that after an unclear word was 'confirmed' by enough users, it was reused as the known word. Which would still leave you trying to OCR words that hadn't been OCRed before.

...thoggles.

When you get down to it, there are several multi-billion dollar industries entirely based around things we don't really need, and many that have been around for a good while. Music, film, drugs, sports, perfume, computer games... In fact, in first world countries, stuff people don't need probably accounts for the majority of economic activity*. Facebook is hardly unique in that respect.

*Disclaimer: this claim is a wild guess based on no actual statistics, and what people "need" is arguable anyway.

And here's why that's not a problem. Even if the market is basically targetting that 95%, the hardware that's produced for it can be repurposed easily enough for what we want. Most likely the companies making it will sell some with traditional operating systems, but even if they don't, it'll be hacked. So far, even the most concerted attempts to prevent users getting full control have fallen (PS3, anyone?). And not all companies will make it very challenging to hack. Apart from anything else, the geeky 5% (or even 1%) of the market is still worth having.

Perhaps the more serious concern is that, if computers become appliances, and only tinkerers have full access to their systems, where do the next generation of tinkerers come from? Will people be able to start prodding things, and finding things out, and slowly developing an interest? My guess is that they will, but it will mostly be at a higher level. Just like writing in assembler or directly controlling tape drives are niche interests now, I guess that compiling software on your own machine is going to become increasingly niche, as more tinkerers play around with the web, or code running in scripting environments.

Little to do with the *code* security, yes. But it's got a lot to do with real-users-not-getting-viruses security.

Seriously, everyone. I know it's sponsored by Microsoft, and I wouldn't be surprised if there's some dodgy selection of test URLs behind the scenes. But if these results are even in the right ballpark, then it's something that Google (and Mozilla, and Opera) really need to pay more attention to. Stop finding excuses to ignore it just because we don't like what it says. Go and try to find the methodology, and see how it's dodgy. Or even do your own tests.