If they'd put the diesel generators (and fuel) in different places, instead of all in a row in the same basement, the four reactors which melted down at Fukushima could've been shut down in a controlled manner despite some of the generators being swamped. Y'know, like the fifth operating reactor at Fukushima which had their diesel generators and fuel in a different location that was shut down safely. Putting redundant backup systems all in the same location just makes them vulnerable to simultaneously failing due to a single cause.

In fact, Samsung internationally hasn't been on the winning side - they've instead been stirring up shitstorms of controversy. Because what patents Samsung does assert are ones under FRAND, and it's lead to many a jurisdiction doing inquiries about asserting FRAND patents in this fashion, including the EU and Korea.

That's actually the bigger problem with everything that's been going on that supersedes Apple and Samsung. The total market value of a FRAND patent must exceed the total market value of a non-FRAND patent. The FRAND patent may have lower royalties per device, but it makes up in volume what it lacks in price. If the total value of a FRAND patent is less than a non-FRAND patent, then you've just destroyed all incentive for anyone to submit their patents under FRAND. Why in the world would anyone decrease the value of their patent by submitting it under FRAND?

Unfortunately this is precisely what was happening with Samsung losing because their patents were FRAND. The courts were getting lost in arguments over minutia and losing sight of this bigger picture. FRAND patents were being devalued relative to regular patents. The difference in pricing Apple was asking for (a few cents per device for FRAND, vs $15-$20 per device for regular patents) would've required FRAND be applicable to hundreds of times more devices than a regular patent before they'd be worth it. In other words, almost never worth it - a single licensing deal with a manufacturer with more than about 0.3% market share would've guaranteed keeping the patent non-FRAND was more profitable. It was setting us on course for a tech world where everything was proprietary and non-interoperable. That they've decided to drop all litigation before more damage was done is great news.

I am assuming that the answer is simply that they had to manually paint in the parts of the photos that were revealed when other parts were removed.

If you've used a recent version of Photoshop, their content-aware fill often does an amazing job at automatically filling in hidden backgrounds.

The Panama Canal - by virtue of being the only alternative to a trip around the tip of South America - can charge passage fees just less than the cost of a trip around South America. Consequently they make a huge profit margin off of operating it. A quick google search says it brings in about $2 billion/yr, but only costs about $600 million/yr to operate. So they've got a massive 233% profit margin.

Add a second canal, and suddenly they're not competing with a trip around South America. They're competing with each other. Unless they collude together to fix the prices so that they're essentially the same (divide traffic 50/50, which might actually be a good thing since I hear wait times at the Panama Canal can be a week or more), the price is going to drop to slightly higher than what it costs them to operate the more expensive canal. That is the nature of competition. e.g. If the profit margin drops to a still-high 50%, profit from the current level of traffic would be just $300m/yr, and it'll take them 167 years to recoup the $50b construction cost even if they were able to borrow that $50b interest-free. Since the Panama Canal is essentially paid for, the Nicaraguan canal would probably have higher costs and thus slimmer margins, and will likely take centuries to pay for its construction.

A Nicaraguan canal would have the advantage of allowing passage of larger-than-Panamax ships (ships designed so their width barely fits through the Panama Canal). But again, if they try to charge significantly more for such ships, operators will simply continue building Panamax ships. Any surcharge they add on has to be less than the money operators would save by using larger-than-Panamax ships. (Significantly more since such ships would have to be built in the first place.)

It'll be great for the rest of the world - cheaper transport costs, more capacity, faster travel. But could end up tanking both the Nicaraguan and Panamanian economies.

Why is everyone assuming this is an all-or-nothing proposition and picking the worst possible cases as a counterargument? I don't decide a spoon is useless because it does a terrible job at cutting steak. You should evaluate this idea based on whether it can improve education if applied judiciously, not focus on the cases where it won't help or could even hurt.

Being able to mix and match partial-semester courses doesn't mean every course needs to be mix and match. Certainly required sequential core subjects like calculus will be long enough that they don't need to be broken up. It was already 2 semesters when I went to undergrad, and no you weren't allowed to take Calc 2 before Calc 1 (unless you passed the AP exam, in which case you were allowed to skip Calc 1). But there were certain short topics in linear algebra I would've loved to recap (I was out for almost 3 weeks with the flu when I took the course in undergrad).

There is also no rule or law of physics says that an educational topic must be exactly one semester. Those of you who only went to undergrad may not see it because you're still mostly taking courses there which are deep enough to warrant a full semester or two. But during grad school, I've taken many courses which would've been better as half or 3/4 of a semester, but which had to be filled out with other stuff to make it fit within the semester system. One course I took covered the new material relevant to my thesis in about a month, then the remainder was an easy ride on cruise control because I had already learned it as part of another course in undergrad. I've even sat in a few classes for few weeks as a listener (i.e. I didn't get credits for it) because that was the only way to learn the part of the course material I needed without over-committing my time by taking the full course.

If you look specifically at the cases where this idea would help, I think it has a lot of merit.

Just test it with a Nexus 5. The ones you buy from Google are unlocked and there's only one version which works on AT&T, T-Mobile, and Sprint (technically the hardware is capable of mostly working on Verizon, except Verizon blacklists it). Just pop in a SIM card for the different carriers and test away.

Autonomy was audited by Deloitte. HP hired KPMG (who subcontracted Ernst & Young) to double-check Deloitte's audit. Those are three of the four biggest accounting firms in the world. All gave the green flag.

That's a nice sentiment, except this has nothing to do with Tech/IT workers. Most white collar jobs are exempt from the FLSA's overtime protections. Generally, salaried professionals and managers are exempt; in that respect Tech/IT workers are no different from other salaried professionals like doctors, lawyers, teachers, engineers, scientists, musicians. Salaried non-commissioned salespeople are one of the few white collar jobs which are non-exempt, and that's where Linkedin got in trouble - they weren't keep track of their overtime hours for some of their salespeople.

Everyone developing FOSS wants to do all the fun programming stuff. But no one wants to do the boring hard work of documentation, UI polishing, promotion/marketing, etc. That's why FOSS tends to suck in those areas compared to the commercial stuff (where they actually pay technical writers, designers, marketers, etc.)

Let me present an alternative hypothesis. There are plenty of technical writers, designers, and marketers out there interested in and using FOSS. They'd probably be more than happy to contribute to documentation, UI polish, and promotion just as programmers are happy to contribute to development. The problem is the FOSS developers won't talk to these guys, brushing them off with "if it's that important to you, figure it out yourself noob!"

The problem is a lack of respect for those who aren't programmers, including users. Commercial software developers respect users because the users are the ones who ultimately pay for the commercial developers' paycheck. One of the unintentional side-effects of the unpaid FOSS philosophy is the lack of a need to respect the users. Whether or not your FOSS project lives or dies depends on how many programmers contribute to it, not on how many users use it.

Services that people need in order to live - energy, water, medical - shouldn't be on the free market. All that stuff should be publicly owned and the goal shouldn't to be to make money but to provide critical services to the people for the cheapest amount possible.

Services that people need in order to live should start off on the free market. Once it becomes clear which method of providing the service is most efficient, then it should be transitioned to a publicly-owned service. e.g. What's the best way to provide fresh water? Wells? Desalination? Buy it from your neighbors and pipe/truck it in? The answer to that isn't at all clear and is constantly changing, so you need market forces to indicate to you the best (cheapest) method of acquiring fresh water. Distribution OTOH tends to be more static - there aren't any up and coming new ways to send water through pipes. So laying down and maintaining pipes is more amenable to a public service.

GSM is a good example of the trouble you can get yourself into if the government prematurely decides something should not be subject to market forces. The EU mandated all wireless phone carriers adhere to GSM. The U.S. did not. Consequently a different method of transmitting phone calls and data - CDMA - was also tried in the U.S. CDMA turned out to be superior to the TDMA used in GSM, particularly when it came to data services (TDMA divides bandwidth equally between users, even if they aren't using the bandwidth; CDMA bandwidth effectively gets allocated as needed as a side-effect of how the technology works). And eventually CDMA was incorporated into the GSM standard (most HSPA and HSDPA implementations use wideband CDMA - yes your GSM phone uses CDMA). If the U.S. had gone along with the EU and required GSM, data services would've been several years behind where they are now, and we'd probably still be stuck at around 1 Mbps cellular data speeds.

The distinction needs to be based on the size of the technological solution space and the uncertainty over the best solution - factors which are inherent to the technology required to provide a service. Not based on whether or not those services are necessary for life - a factor completely independent of the technology needed to provide the service. Once you realize this, you realize that other things not necessary for life should probably be publicly owned - e.g. cable TV lines. When cable TV first began, it wasn't at all obvious what was the best way to distribute high-bandwidth content to houses. But now it's pretty clear that fiber to the home is the end-game. So the government should be installing fiber to each home, then allowing multiple cable TV vendors to compete selling service over that fiber.

The money for the reactor's decommissioning comes from surcharges to electrical rates collected while the plant was in operation. This money was earmarked specifically for reactor decommissioning costs, and placed into a trust fund which currently contains about $2.7 billion (the $4.4 billion cost will be accrued over several decades, so interest on the $2.7 billion makes them more equal than the raw numbers suggest). That there is sufficient money despite the reactor shutting down only halfway through its expected lifetime means there's a huge margin for error in these nuclear decommissioning funds. Edison has said if there's any money left over, it'll be refunded to rate payers.

Logan -> Blue line -> Orange or Green Line -> Red line -> Kendall/MIT

How hard is that?

Three transfers on the T (there's also a bus you have to take from the Logan T station to your airline terminal) is hardly ideal when you're hauling around luggage with kids in tow and on a schedule. When I was traveling by myself with a single suitcase I'd do the three transfers. But outside that case, a taxi is just easier and quicker.

More succinctly, an effective free market requires participants to be rational and informed. Product behaviors like keeping a purchase window up for 15 minutes without notifying the customer deliberately try to mislead and misinform the public. That's the sort of thing the FTC is there to crack down on.

I don't see any problem with a pop-up after a purchase asking if you wish to continue to make purchases for 15 min without having to re-authorize. But to silently make it the default behavior is pretty obviously deceptive. I'd even call it a scam.

The other half is that some carriers (I'm looking at you Verizon) won't activate a phone they didn't sell, even though it's capable of operating on their network. What good is an unlocked phone if no other carriers are willing to activate it? To be effective, this needs a partner law requiring carriers to activate phones which can operate on their network, regardless of where the customer bought it.

Another benefit this would have is that manufacturers would start selling phones directly instead of only through the carriers. The carriers are already busy trying thwart the unlock requirement - most of the new carrier-branded phones I'm seeing support only the frequencies that carrier uses. Even if you could unlock it, it won't work on another carrier (or will work with degraded capability). But if the manufacturers began selling phones directly, it would be in their best interests to sell a single phone model which supported all carrier frequencies in that country. With another bonus being that it'll work in most of the rest of the world as well.

The meters on traditional cabs may sometimes be tinkered with, but that's illegal, and in the vast majority of cases they're accurate and legally binding. Whereas with the new wave of rideshare apps there's no indication of what charges you're reacking up until you arrive.

There's another way to tinker with meters besides hacking them - drive a different route. My first taxi ride from Boston Logan to MIT took what seemed like an unusually winding route through downtown Boston. A year later when I got a car and began driving around the city myself, I realized I'd been taken for a ride, literally. I mentioned this to a fellow student at my lab, and he remarked that it had taken him 3 years to figure out what the actual cost of a cab from his apartment to the airport was, because every time he'd be taken on a different, circuitous route to rack up extra miles. It was 3 years until he actually got an honest driver who took him straight home.

Modern navigation software means you can get an exact distance from start point to destination before you even step into the taxi. There's no need for a meter (other than to time the ride). A lot of airport taxis already do this for longer rides - they charge based on zones, with further zones costing more. No meter needed. The only case this doesn't handle is rerouting to avoid traffic. An alternative might be a meter which prints out your GPS route so you can see that you were taken almost straight to your destination, and not in circles.