Lets go a step further. Droid users would rather vote for Ron Paul where as Apple users would re-elect Barack Obama. If you're going to throw gasoline onto the fire, at least learn to do it right.

Don't worry. Most will have to ask someone what "grossly overconfident" means so few will feel the sting.

Well, Microsoft said they had "no obligation to return data" [in-other-news.com] to their customers.

You have to be both grossly overconfident and very naive to do business with that company.

And to all MS-apologists: No, it is not the case that "everybody does this". Most companies do provide guarantees to their customers, especially when it comes about handling data. What is the point of storing something in "the cloud" when the company boldly tells you that they have "no obligation" to even give you the data back? Of course there are some things where MS cannot be avoided, but it is your choice to minimize or maximize your exposure.

So when MS screws you over [xda-developers.com], don't complain. Also I just don't get the philosophy to shut down all servers when some deadline runs out - it would cost MS practically zero to just keep the servers running - but no, they must turn everything off to inflict maximum pain for their (hopefully ex-) customers.

Well, I've seen worse. It does contain some questionable questions ("There are 7 notes in an octave" is supposedly false, "Norway is the land of the midnight sun" is supposedly true even though the explanation then says that Norway is just one of several countries that bear the title, "Gout is known as the Royal Disease" is supposedly false, note the inconsistency between the Norway and Gout questions) but I e-mailed them and they immediately changed at least one of the questions. Very friendly people, and they confirm the test is still being worked on. They even thanked me for my input with a voucher for the expert test.

As long as they only include uncontestable questions, I think this idea is actually pretty good and I would expect the results to be a lot more accurate than most other personality tests (which are usually complete rubbish).

And perhaps they could give the error margins for their results. I'd put money on the error margin being a lot bigger than differences in the user group results.

It's not a study. It's an online poll. The participants self-selected

Ok, it's not strictly instead of Firefox and Chrome, but it's my default browser on my primary work machine. I'm currently running VMware Player, with a Linux machine on it, and that runs Firefox, which crashes Way Way Too Often, usually because of Flash. I do most of my web browsing there, and have NoScript, Ghostery, and AdBlock Plus, and usually a couple of other safety tools. And I keep another Virtual Machine around, with a stripped-down Linux distro with Firefox in Private Browsing Mode, which I use to read Facebook, because I don't want Facebook contaminating anything else, stealing cookies and history files, or whatever. (If there's a way to keep VMware Player paged in, using the whole 1GB I have allocated for it, instead of swapping itself out when it's not busy even though there's spare hardware RAM left, I'd appreciate pointers; I haven't found them.)

I'm also running Chrome natively, mostly for a bunch of electronics blogs like Hackaday, and occasionally for Gmail, and it's really bloated - burns almost 2GB if I have it turned on with my usual set of tabs. I'm not sure I entirely trust Google to behave themselves with Chrome, but they already know everything about my Gmail account (which I don't use for anything sensitive), and the electronics stuff doesn't get much personal information except when I'm buying equipment.

I used to run Firefox natively as my default browser, but there are a couple of problems with it - it Crashes Way Too Often, and it's also a memory hog (though better than it used to be, and not as bad as Google), and there are a couple of work applications that don't run cleanly except on IE. Until recently, it was my default browser, so if I clicked on a link in an email message, FF would either start from scratch or open another tab, spin the disk for a while while it sucked down memory, and then run, hopefully without crashing itself or crashing something else by hogging memory, and then be its usual friendly self. But I found that usually when I'm clicking on links from my work email, they're either sites I trust, or else they're work related sites like the HR website or web conference bridge that are happier running in IE, and I got tired of that.

That takes us to IE. It's IE7 because the Desktop Support department at work finally let us use IE7 instead of IE6, but is too scared to go to IE8, at least on Windows XP, and they made their saving throw against Windows Vista a couple of years ago - my next set of hardware will run Win7. And it has tabs, so it's not totally obnoxious to use, and it really doesn't crash much, so it's less obnoxious than Firefox, and it usually doesn't use a lot of memory, because I don't usually let it keep more than a couple of tabs open at a time, though it would happily be a memory pig if I let it.

(And then there's Safari and Opera, which I used to have installed - the IT department run little scanning robots that rat you out within a day if you install them, for reasons that sound more like the Software License Police rather than the IT department's normal reticence to have useful software running on our machines, and you get a call from some guy in India who's going to walk you through uninstalling them whether you like it or not. So I no longer run them.)

I suppose there's also Konqueror or other Linux-oriented browsers that I could be running in the Linux VM - are there times it's worth using them instead of Firefox?

Most of your Window users are technologically illiterate. IE is there it works why fart around with it.

Yes, but ... I'd have expected the opposite result myself - technologically illiterate people who are risk adverse would be more likely to not try installing third party software on their PCs. Overconfident types would install every browser from everywhere without consideration for where it came from and what trojans it might harbour.

Of course, it could be that this study is just as much of a hoax as the original.

The Dunning-Kruger effect is perhaps the most useful effect to use when trying to make sense of modern politics. Listen to any die-hard politico, and the more sure they are of their response, the more certain you can be of how inexperienced they are.

In today's politics, a sure, unwavering certainty is almost a sure sign of success: a "flip flopper" will get nowhere, (Mitt Romney, John Kerry take note) while idiots who never change their opinions (EG: George Bush Jr) get lots of press for "holding true" despite all the evidence to the contrary.

So, the loudest political advocates are either the idiots, or somewhat less loudly, those who actually have some idea what's going on. For those who just want to "do the right thing", without a lot of effort, it's damnably difficult to tell the difference.