In the UK, independent schools and state-run schools have different requirements for the number of teaching days per year. For independent schools, the requirements is about two weeks less, and yet independent schools (most of which are only open for the required minimum) consistently do better at university admissions than state schools. This doesn't mean that reducing the number of teaching days increases student success, but it does show that this single factor is not always the dominant feature. For example, independent schools typically set more homework and have smaller class sizes, which balance out the reduced teaching time. They also have better paid teachers who spend a higher proportion of their time teaching, rather than doing admin work.

It's not normally the same companies. Mostly defence contractors get the benefits, which sort-of makes sense as much of the US military depends on the products from these industries, and so if you squint enough it looks like a national security concern.

Why would they need to stage a coup? US elections are almost invariably won by the candidate who spends the most on advertising (one dollar, one vote, although I think at the last election is was closer to $10/vote, but that's inflation for you) and the Supreme Court recently removed the checks that made it easy to block foreign funding of candidates. It's still technically illegal, but it's now basically unenforceable. Given the amount of money the US hands out around the world, buying an election would probably be quite a good investment. The $5-10bn of up-front investment can easily be recouped with a couple of favourable trade deals.

Because that's 0.75 for the play store, on top of what they charge for the other Google apps.

How do the revenue numbers look when you add in apps? The last numbers I saw were about a year ago so things may have changed, but back then iOS users were a lot more likely to buy apps, but Android had a much bigger share of downloads for ad-supported apps. The revenue was about the same for both platforms, because you'd get less from the ad sales in Android apps, but you'd get more downloads.

Is that a serious question? Take a look at the proceedings from any security conference in the last 2 years and you can find a very long list. The latest trick is for individuals who release small apps for free or a token amount to be offered money to sell their app, especially if the app already asks for more permissions than it really needs (great incentives there...). The buyers then release a new version bundled with malware. The new version is installed automatically if it doesn't need any more permissions, and since most manufacturers don't ship software updates for Android phones in a timely fashion there are typically a few nice root vulnerabilities lying around on a significant fraction of the installed base. From there, the attacker can do what they want (attack mobile banking apps, harvest passwords, send premium-rate SMS, or just proxy all network traffic and inject their own ads, the last being the most common).

I know a couple of people who have turned down money to sell their (free, with only a few thousand users) apps for this purpose.

They're not charging me. They're charging the OEM, who will pass on the cost to me without providing me an option of opting out, for something that they will then use to harvest personal information about me to sell to advertisers. Give me a phone for $2 less without the Google crap and I'll happily take it in preference.

I'm just leaving the Bay Area, and most of my US-based friends use an app that you give a destination to, it automatically knows your location from GPS, calls a taxi, takes you to the destination, your phones both agree on the distance, and the charge is taken from your account. It's very smooth and convenient (and removes that awkward thing in the US of working out how much you're meant to tip the taxi driver, one of the tipping situations that appears to make no sense because having someone drive you is the service you are paying for). Add in self-driving cars and you remove the cost of the driver. The ancillary infrastructure is there already, waiting for the self-driving cars...

Not an alternate universe, but a planet closer to the sun. With current panels, you get around 100-200W per square metre of sunlight. The theoretical maximum efficiency for solar panels is somewhere around 40%, bringing it up to 400W. Over 8 hours a day of useable sunlight (that's the output with the sun directly overhead, it slowly drops off over the day, giving an average of around 8 hours, assuming good weather). So that gives a total of around 11.5MJ per day. One litre of petrol releases around 34MJ when burned, so to generate the equivalent energy of one litre of petrol per day, you need three square metres of solar panels, assuming magic future panel technology and losses equivalent to a petrol engine.

With current technology, you'll need closer to nine or ten square metres, or more if you don't have a very efficient charging system. If you're living out in the countryside, this is quite possible (if you can afford the massive up-front investment for the panels, but let's assume that the price will come down quickly), but for anyone in a city it's quite unlikely. Add to that, you don't (depending on where you live, of course) get bright sunlight every day, so you're most likely going to need to store energy over the winter in fairly large amounts. Why not make that more efficient, by centralising it? You could lay a set of power lines to people's houses and they could send their unused power back to your storage plant. And, once those wires are there, you can probably build a centralised power generation facility and sell them power more cheaply (and reliably) than they can generate it themselves, if you factor in the capital and maintenance costs (after all, solar panels need cleaning, replacing, and so on). Such a system would be like a computational grid, but for electricity. You could call it an electricity grid...

The UK has a first-past-the-post system, and we currently have a coalition government, formed by one of the two major parties and the third party. Going back to the middle of the last century, the third party was one of the dominant two and was displaced (and reformed after merging with another small party) by one of the current two. FPTP electoral systems do tend towards two-party systems, but they're not stable and can be periodically upset (at which point they'll again start tending towards a two-party system. They trick is to upset them frequently).

No. But they won't win either. The point of voting for a third party is to build a group of the electorate who aren't voting for either of the two big parties. Once that happens, either the two major parties will start to make changes to their policies to try to win back those voters, or candidates from a third party will actually stand a chance and so you're likely to see an increase in candidates you might actually want (as well some some crazy fringe parties that you almost certainly don't).

And yet they went with a set of first-past-the-post elections that pretty much guarantee this outcome. If they'd written their constitution a hundred or so years later, then they'd have had the mathematical tools available to study and understand this, rather than just some vague disquiet.

NT4 got DirectX support in later releases, although a lot of games didn't work because they did things like try to modify Program Files or the local machine register key. It had OpenGL support from the start, so it ran GLQuake fine, which was most of what I cared about at the time. 2K was a lot better, although there were some issues with IPX not being quite compatible with 9x, and a lot of Windows 95-era games only supported IPX for multiplayer.

I opted for option 2 about 7 years ago, when I realised how cheap projectors and competent 5.1 surround sound systems had got. It spent around £250 on a projector and a set of speakers, which I drove from a DVD player. My local cinemas all had really bad equalisation in their sound (far too much base, no midrange, so you got too-loud explosions and talking was hard to hear) and had so much dust in their projector lenses that I got a better quality experience at home and could sit in comfy chairs, drink beer, and pause the movie whenever I wanted. Having friends over and getting them to bring food and beer still ended up cheaper than the cinema and was more fun. Even including power and movie rental, my cost per film has been significantly less than going to the cinema, although I have watched a lot more films on the setup than I would have gone to see.

As long as you've got a room with a spare wall, it's quite easy to make something that is both cheaper and better quality than a low-end cinema. You won't be able to make an IMAX competitor (unless you've got a really huge living room), but I don't live near an IMAX so that wasn't the competitor.

I recently replaced the bulb in my projector, after 3,000 hours of life. It cost £50 for a new one, which works out at around 2p/film for bulb costs. With a newer LED projector, it's even cheaper (although the up-front costs are higher). With HD projectors coming down in price, you can get even better quality.

The MS student OS pricing had 95/NT4 and ME/2K at the same price, so anyone who qualified for that discount could have picked the non-sucky one...