the Kyrgyz have cultivated the jew's harp, or kobuz. This instrument has only one vibrating element, and though it can produce only a single tone, the performer can create a variety of sounds through changing the contours of his mouth and lips. It's a humble instrument but so endless.

These instruments and overtone singing take advantage of something most people know nothing about - how our perception of sound works. When you hear, say, a violin, you aren't hearing a "violin sound". You're hearing a combination of sinusoidal waves - a fundamental combined with multiple overtones. Those of you well-versed in math may recognize this as a Fourier series. Our ears are designed to perform a Fourier transform on incoming sounds, and separate out the individual frequencies. Individually these sinusoids are no different from the sinusoids generated by any other instrument (or any sound source for that matter). The stimulation a violin gives your ear at 440 Hz is exactly the same as what a horn gives, a voice gives, or even a jet engine gives at that one frequency - a sinusoid.

It's the ratio of the magnitudes of different frequencies (fundamental + overtones) which make a violin sound like a violin. A horn has a different ratio of magnitudes at the different overtones. Same with a voice (each individual voice - that's how you recognize someone's voice). Same with a jet engine. The raw data our brain gets from the ear is just a bunch of separate frequencies and and their magnitudes. It then compares the ratios of potential fundamentals + overtones and says, "oh, that's a violin". It always amazed me that you listen to an orchestra and pick out the individual instruments - all their frequencies and magnitudes are overlapping into a mishmash of sound. When I later learned about spread spectrum multiplexing, that's when I understood your brain is doing the equivalent of orthgonal frequency division multiplexing (which is the technology used for LTE). Even though the frequencies generated by the instruments in an orchestra overlap into a mishmash of sound, the ratios of their overtones are orthogonal enough that if you listen carefully you can pick out the individual instruments. Exactly how your cell phone picks up the signal intended for it from the mishmash of signals the tower puts out for all phones.

In many ways, a Jew's harp and overtone singing are akin to analog sound synthesis. If you thought the high pitch she was singing sounded vaguely flute- or recorder-like, it's because those instruments have very close to a pure sinusoidal waveform (fundamental, very little overtones). The individual overtone she is amplifying is sinusoidal in waveform, so your brain interprets it as flute- or recorder-like. It's very similar to how the early synthesizers worked (before computer memory became cheap enough to sample everything). The name comes from frequency modulation synthesis - you modulate the magnitudes of the different frequencies (overtones) to alter what instrument the sound sounds like. Match the pattern of violin overtones and it sounds like a violin. Match the pattern of horn overtones and it sounds like a horn. The overtone singer has only mastered modulating the fundamental and one overtone, which limits her synthesis capabilities. But if she's capable of modulating a second or third overtone, she could start making her voice sound like different instruments. Just like an old analog FM synthesizer.

Google is following Apple and Microsoft and moving away from widescreen tablets. Good riddance, I say; 4:3 or 3:2 is much better for showing a 'page' of information.

Most people who think this forget about margins and compare to the entire page size. 4:3 is actually the worst aspect ratio. The aspect ratio of a tablet only refers to the screen size. Your tablet already has bezels which act an awful lot like margins. Why do you want to waste valuable screen space on displaying blank margins?

A trade paperback is typically 6"x9". Margins are asymmetrical, typically .75" and .875" on the sides (larger margin for the center gutter), .75" and .5" for the top and bottom (larger margin for the page number). That leaves a printed area of 4.375" x 7.75", which is a 1.77 aspect ratio. Almost exactly 16:9 (1.78). If you go with smaller .5" and .75" margins on the sides, .5" and .5" margins on top and bottom, you get a 1.68 aspect ratio - between 16:10 and 16:9.

For a regular paperback that's 5"x8", these margins give a 2.0 and 1.87 aspect ratio respectively. For a pocket paperback (4.18"x6.88"), the aspect ratios are 2.2 and 2.0. So for something the size of a phablet or 7" tablet, 16:9 is pretty close to ideal.

"But what about 10" tablets?" The printed area of an A4-sized sheet of paper with 25 cm margins is 1.54:1. Right in between 3:2 and 16:10. A letter-sized sheet of paper with 1 inch margins is 1.38, right between 3:2 and 4:3. However, if you look at anything published on A4-sized or letter-sized paper, the text is nearly always arranged in two columns. So 4:3 and even 3:2 is really too wide for displaying scrollable text. That's why nearly all websites have switched to a format with menus on the left, a narrow column of text, and misc links on the right. The main reason a "page" is this wide is so you can include wider pictures which span both columns. This becomes unnecessary when you can zoom into the picture like on a tablet, or rotate it to landscape mode and have the picture automatically flip to fill the longer width of the screen.

(Also note that the printed area of A4 and letter size paper is actually between 11"-13". Tablets are only 10" because of cost and weight. Assuming the publishing industry knew what they were doing if after centuries of printing they standardized on A4 and letter sizes, 10" tablets are eventually going to be phased out for 11", 12", and even 13" models as technology improves and they become lighter and cheaper.)

Competition in the last-mile is infeasable

Competition in the last mile was necessary (past tense). When Cable Internet was first rolling out, it wasn't at all obvious what was the best way to do it. Cable TV is easy - you transmit a bunch of signals over a wire, and each house taps into it. It's basically broadcast over wires. But with cable Internet, you have to be able to transmit different signals to each house, and each house also needs to be able to send signals back along the same wire. You've now got a network problem akin to 10base5 or Token Ring with no obvious best solution (both algorithmically and in terms of equipment). Competition is a great way to solve these kinds of nebulous open-ended problems.

By now though, pretty much every cable company is using the same technique to partition their cable networks (the methodology has become so standardized you can buy your own DOCSIS cable modem now instead of each cable company having their own proprietary modem). And it's pretty clear the equipment is going to transition to fiber to the home. When everyone is implementing the same last mile solution, competition is no longer necessary and you should transition it to a utility.

The Japanese Shinkansen is now running over 7 times times as fast as the average U.S. express passenger train.

What could possess someone to think it's ever valid to compare a maximum to an average?

Compare a maximum to a maximum (500 kph for this Shinkansen vs 241 kph for Acela). Or an average to an average (261 kph for newer Shinkansen vs. 129 kph for Acela). So the difference is only 2:1, and mostly has to do with (1) established rail routes in the U.S. being much, much older so as not conducive to high speed, and (2) travel distances being much greater in the U.S. resulting in air travel being more economical/time-efficient.

The only difference that buying direct links in meant was that they got to skip the congestion in the peering points. Comcast has alot more bandwidth internally and once traffic makes it into the network, congestion is not usually a problem (things do break, redundant links become saturated, etc. It's a big network, but in normal operation mode, congestion doesn't exist).

That's just it. Comcast didn't need to buy a direct link. Netflix offers a CDN and caching hardware for free to ISPs to help alleviate the peering congestion you're describing. Comcast (and Verizon, TW, etc) refused to accept Netflix's free offer. Instead they had the gall to charge Netflix money before they'd accept it.

Plus, as fuel efficiency has gotten better and Americans have started driving less, the tax has naturally raised less revenue anyway.

That's only true if you compare to gasoline consumption during the economic bubble from 2003-2008. If you look over a longer period, gas tax revenue is the highest it's been since before 2003 in nominal dollars, and is roughly the average it's been from 1990-2014 in inflation-adjusted dollars. The tax is due for an increase to counter inflation, not because of the reasons TFS cites.

Federal gas tax revenue (fig 6-2) has never been enough to cover highway construction and maintenance expenses (figure 6-3). The gap has always been made up by state fuel taxes and other revenue.

For the weight of an RTG they could have dropped several different probes onto the comet, all of which could have very large solar panels.

Why do people think RTGs are the huge things the Soviets used to power their remote lighthouses? The RTGs used for space missions with the same power requirements as Philae have been about 12-15 kg. About the same as the solar panels, regulation electronics, and batteries aboard Philae.

The science payload is ~25 kg. The lander itself (including frame, power systems, communications array, electronics, stabilizers, descent systems, etc) is about 100 kg.

An RTG is heavy. Solar panels are much lighter, and the comet is currently on a sun approach.

According to Wikipedia, Philae's power system weighs 12.2 kg and generates 32 Watts @ 3 AU (approximately halfway between perihelion and aphelion).

A SNAP-19 (1970s-era RTG) weighs 13-15 kg and generates a constant 40+ Watts electrical.

The comet's perihelion is 1.2 AU, aphelion 5.7 AU. Generally, Mars (~1.5 AU) is about the point where solar ceases to be cost-effective. Orbiters sent to Mars are solar powered. But landers (which have to deal with longer nights) have used RTGs when possible (Viking landers, Curiosity rover), with solar powered landers having a life expectancy of weeks to years.

Given they were landing on a tumbling comet ((the comet has a 12.4h rotational period so the lander would experience a relatively lengthy "night"), and the perihelion being somewhere between Earth and Mars, this was probably a good candidate for a RTG. I suspect they weren't expecting the lander to survive past perihelion however (13 Aug 2015), which could have tipped it in favor of solar.

The real danger is that Sony is also a content producer (Columbia TriStar). Them fiddling around with a distribution method means there's a risk they will pull their film library from other distributors and put it exclusively on their own network. Want to stream movies like Lawrence of Arabia and The Shawshank Redemption? I'm sorry, they're only available on Playstation Vue. (Netflix is no exception - they produce House of Cards and it's available exclusively via Netflix)

This sort of vertical integration really distorts the market. Microsoft refusing for years to release Office for Android and iOS because they wanted to protect their Windows franchise. Verizon playing hardball with MVNOs so they can retain claim to the title of best mobile network (imagine if the companies who owned the cellular towers were different from the companies which provided the service), Cable companies trying to control and charge extra for the data that goes through their exclusive pipes, etc. I can understand the integration if nobody else is willing to offer the product/service to your satisfaction. But if there are lots of competitors out there providing the service, it should be illegal to blatantly withhold your product from all other services except yours.

More succinctly, if you're developing a game and you know it's good, why would you turn down free publicity before release?

The Peter Principle is a concept in management theory in which the selection of a candidate for a position is based on the candidate's performance in his or her current role rather than on abilities relevant to the intended role. Thus, employees only stop being promoted once they can no longer perform effectively, and "managers rise to the level of their incompetence."

The solution (assuming you're already in a state with incompetent managers) is to allow incompetent managers to be demoted back into a position they're competent in. Unfortunately, society has a huge bias against workplace demotion.

Android is a walled garden too - just with different types of walls. There are countless Android devices that are locked by the manufacturer to older versions of Android, loaded with crapware which cannot be removed and otherwise effectively turned into a walled garden. Google does little to prevent this from happening and in fact largely facilitates this abuse of users via indifference.

Aside from Amazon devices (which use a forked version of Android), pretty much all Android devices are not a walled garden. Yes many are locked to a carrier, or have preinstalled apps you can't delete. But on the vast majority of them you can simply go to the settings, check the option to "allow installation from unknown sources," and you are outside the walls.

Breaking up the cable companies probably wouldn't do much without a new technology introduction. Break up of AT&T worked in retrospect because of advances in cell phone transmission, a leapfrog technology. Otherwise the Baby Bells would have still owned the local cable (like Fairpoint in New England).

Cell phones had nothing to do with helping the breakup of AT&T. The breakup was about long distance phone service. Bell Systems owned the termination points (the switching equipment directing calls over "last mile" wires to people's phones) so could effectively shut out competing long distance carriers who came up with technologies to bypass Bell's long distance wires. That's what MCI was - Microwave Communications Incorporated. They bypassed Bell's long distance wires by transmitting long distance phone calls via microwave towers. You called a local MCI number, they forwarded your call via microwave to a local phone in the area code you were calling, and that phone called the phone number you wished to reach. Bell handled the two local legs, MCI handled the long distance leg. When Bell tried to shut them down by refusing to connect their microwave calls to the phone number being called, they sued and precipitated what ended up with the Bell breakup.

The Baby Bells still own the local cable. Bell chose to satisfy the U.S. government's anti-trust investigation by spinning off the local phone companies as separate entities. Bell (now AT&T) would remain a long distance carrier (back then long distance phone charges were the most profitable part of the phone business). This eliminated the conflict of interest when terminating calls from competing long distance carriers. Verizon is a Baby Bell. SBC (now AT&T) is a Baby Bell. BellSouth (now AT&T) is a Baby Bell. Qwest (now CenturyLink) is a Baby Bell. The breakup worked because local service became independent of long distance service (and the local services are regulated as utilities so cannot discriminate against any particular long distance service - basically what net neutrality seeks to do).

The Bell breakup is really a perfect model for Internet service. It makes sense to only have one set of wires for the "last mile". So the DSL/cable/fiber company can be a monopoly, but should be regulated as a utility. The internet service itself should be wide open to competition. Anyone who wishes to start an ISP company (analogous to long distance carriers) should be allowed to, and they can pay the local DSL/cable/fiber utility (analogous to a Baby Bell) a fixed rate for "leasing" that last mile. Network packets get sent over that last mile from the home to the ISP. If the customer chooses a different ISP, the packets just get routed via a different ISP.

The *only* reason to offer some sort of price promotion later instead of just giving you the discount up front and making you sign a contract (even for three months of service) is so that they can roll the dice on customers signing up after being enticed by the promotion, and then not claiming the promotion. Rebates work the same way. Why is it that we can pass a thousand consumer protection laws about credit card interest rates and privacy disclosures, but can't just ban the blatant rip-off of rebate promotions?.

Actually, rebates serve two different purposes. It allows the manufacturer to pass along a price cut to the final buyer without affecting its contract pricing with distributors and retailers. And it allows manufacturers to make price cuts (sometimes rather large ones) without having all of the product bought up by middlemen - e.g. eBay resellers who quickly snap up all of the available inventory at the lower price and list it on eBay at close to the original price (rebates are typically limited to a certain number per household address).

The rebate companies hired by manufacturers to run the rebate programs work as you say. They get paid a lump sum by the manufacturer, and get to keep anything left over after paying out the rebates. So they have an incentive to stall, lose, and deny your rebates. But the reason rebates exist are legitimate and have nothing to do with ripping off consumers. In fact the second reason actually helps consumers (by cutting out flippers who drive up the price).