To put another way, there's exactly no reason to allow remote control over refrigeration like this. The refrigerator is designed to keep the contents at a particular target temperature, and no amount of remote intervention is likely to fix it if it fails to do so.

The only reason you would want to change the temperature would be if you changed what products were going into that refrigerator, and that means you're physically at the unit and can change the temperature manually just as easily.

So yeah, just stick some temperature monitoring units in each refrigerator, and don't connect the refrigerator to the Internet (or even an intranet). There's just no good reason for that. For that matter, more complex the control circuitry on your fridge, the more likely you are to lose everything in it, so don't buy anything that even *can* be connected, if at all possible.

Assuming a roughly linear degradation, if it is 70% at 450,000 miles, it will have lost 60% of its range at roughly 900,000 miles. How many cars do you know that aren't rust buckets at 900,000 miles?

Sorry, misread. :-)

Why do you think that? Most people don't live alone, and most households have more than one car. People tend to take long trips as a family, not as individuals. That means second and third vehicles in a household, statistically speaking, almost *never* get driven on long trips again unless the cars subsequently get traded in and become some other household's primary vehicle. So in practice, those vehicles are almost never driven more than about 60 miles. Even in the rare situations where this is not the case, such as a beater car that gets driven on a long daily commute, it is exceptionally rare for that number to significantly exceed one-hundred miles or thereabouts. So for a vehicle that starts with a 300-mile range, losing 25% of the range is just not a big deal for second and third automobiles. This is just common sense, and anyone arguing that this is wrong needs to show some data, because such claims are prima facie quite shocking.

Yeah, sometimes hardware fails. There are many, many complaints of fuel pumps failing, of transmissions failing, etc., too. ICE cars have a whole lot more high-failure-rate components. Transmissions, in particular are the bane of ICE cars' existence, with the average lifespan being probably 150k miles or so, but some models routinely requiring multiple rebuilds within the first 100k miles, particularly if you do a lot of driving in hilly areas. That's one huge win with EVs; I'm pretty sure Porsche is the only company that puts any kind of transmission into their EVs. Everything else just has a fixed gearbox that will pretty much last forever.

What I am doing is pointing out that for most vehicles, it doesn't matter. Some hard numbers:

• 33% of households have one car.
• 37% of households have two cars.
• 22% have three or more.

(source) For a very conservative estimate, if we treat "3 or more" as exactly three, we can compute the number of cars proportional to the number of primary cars by multiplying 33 * 1, 37 * 2, and 22 * 3, and adding these numbers together to get 173 cars per 92 households. When you divide 33 (the single-car households) by 173 (the total cars), you find that only 19% of all cars on the road are a family's primary vehicle, and 81% are a household's second or third car!

Even if you assume that everybody who owns a car will periodically take a long road trip, that still means that more than four out of every five cars on the road will likely never drive more than a double-digit-mile round trip.

So again, losing 25% of the car's range for those cars does not matter. At all. Most of them could lose 75% of their range and still be good enough for what they are used for.

But even better, long-distance travel by car tends to not be all that long-distance. The median round-trip distance for long-distances car trips is just 194 miles (source). So unless you're driving around a lot while you're at your destination, for a vehicle with a 300-mile-range, even a whopping 33% reduction range would not require even a single additional charging stop for *half* of all long-distance trips taken by car.

In fact, only about 10% of all long-distance trips are more than 500 miles, which means you might need one extra charging stop if your range dropped by 25%. Only 5% are more than 1000 miles, which means you might need two or more extra charging stops. And I can't find the statistic right now, but I saw a stat that something like 5% of city drivers take 90% or more of the long-distance trips.

Factor that in, and that means 95% of the 33 vehicles will rarely or never take a long-distance trip. That means that for a whopping ninety-nine percent of cars, losing 25% of their range does not matter. At all. Far from all drivers caring, if they actually look at their usage patterns and evaluate them rationally, you should find that 5% of drivers care, and they would care about that range loss for only about 1% of vehicles.

Your assumptions about how cars are used simply do not match up with objective reality. In fact, they're basically the exact opposite of what the actual numbers show. So again, you can feel free to disagree, but you'd better be prepared to support your opinions with numbers, or we're just going to write your opinion off as baseless.

That's not realistic. By the time an EV loses 60% of its range, the battery has probably caught fire from dendrites, not to mention that the rest of the car will have succumbed to rust twenty years earlier.

That's not even remotely accurate. Most Tesla vehicles that are at 200k miles have around 85% of their original range, not 70% as you claim. You're literally doubling the amount of degradation compared with what happens in the real world. And given that you usually lose the first 5% within the first year, losing an additional 10% range over a decade and a half is really not that interesting.

A 25% range loss really isn't a problem for more than maybe 1% of drivers. For daily commuting, most EV owners charge their cars at night every night, and add maybe 60 miles of range each time, so for a vehicle that is just used for commuting (the vast majority of cars, and ~100% of second vehicles in a household), the range loss has zero impact on them whatsoever.

The only situations where this really matters in practice is for tow vehicles, where that might genuinely prevent you from making it to the nearest charging stop. For all other vehicles, realistically speaking, that just means maybe one extra charging stop per day on long cross-country trips. It might be a little annoying, but it's hardly earth-shattering. And given that most people don't do those sorts of trips very often, the extent to which it is annoying is pretty limited — certainly not worth spending tens of thousands of dollars on a new battery.

So contrary to your breathless assertion, you would not typically get an EV fixed for that, because in practice, it really isn't important for 99% of use cases. For the remaining 1%, they'll just buy a new car and contribute their old car to the used car market, where that car will be more than good enough for somebody else to use for another two decades.

Realistically, no. Most EV batteries outlast most ICE cars. Of course, it depends on the size of the battery. Smaller batteries result in more charge cycles per 100k miles, which means more battery wear, which means earlier replacement. But at least Tesla batteries have an expected lifespan of 1500 charge cycles, so if your battery has a range of 300 miles per charge, you should get 450,000 miles out of it. With an average ICE car being replace after 200,000 miles or so, you'd expect the battery on any EV with more than 133 miles of range per charge to outlast an average ICE car. So basically, pretty much every EV ever made except for the Rav4 EV and plug-in hybrids. :-)

Wait, you're using the Nissan Leaf as your EV example, and the Camry as your ICE example? The early Leaf cars are well known to be a disaster, and are an extreme outlier in terms of low life expectancy, which brings that number way down. The Camry is well known to be one of the most reliable ICE cars out there. You're comparing one the worst EVs to one of the best ICE cars. You can't do that. You have to compare averages if you want your comparison to be meaningful.

Or, if you want to compare the exemplar of each class, compare a Model S (300k to 500k typical) to a Camry (200k to 300k typical), in which case you'd conclude that the ICE car lasts from 60 to 66% as long as the EV.

I stand corrected. I'm assuming the majority of the push for this is from folks doing their own low-volume manufacturing of hardware? Oh, yeah, I see the supported device list. :-D

The point remains that to within a small margin of error, operating system projects on mobile devices exist largely to support hardware vendors, and that real-world end-user interest in installing a different OS on their phones is likely to be close to zero (except for users of those specific hardware vendors) for the same reason that the number of end users who install Linux on their computers for non-server purposes is remarkably close to zero, for the same reason that BeOS, GeoWorks, Visi On, TopView, GEM, OS/2, etc. all failed, etc.

And this is why operating systems have to be designed to be resilient and secure by default. If you can run any app you want, but you have to explicitly allow an app to access any content from any other source app on a per-source basis, to access passwords on a per-password basis, etc., then there's approximately zero danger in running the app, because it literally can't do anything that the user doesn't let it do. And if the user lets it do something that it shouldn't, it's the user's responsibility for cleaning that up, same as if they gave somebody else their password.

There should be exceptions, of course, such as allowing device-wide access for an app designed for backup purposes, but those exceptions should be very few and far between, and should require reasonable certainty that the companies that created them are legitimate before granting the exceptions, and that is a useful gatekeeping behavior for the device/OS manufacturer to wield, provided that they do so in a fashion that does not effectively prevent competition against their own offerings. If they do, they should be forced to divest their own competing offerings as punishment for doing so.

In fact, that should be the default punishment for antitrust violations by gatekeepers. If, for example, Apple were found guilty of violating antitrust law by unlawfully harming competition against Apple Music, they should be forced to sell off Apple Music. Same for iCloud Backup if any company is refused permission to create a system-wide backup tool. And so on. Fix the problem permanently by fixing the conflict of interest.

They still arguably take it way too far, just slightly less way too far.

Although in theory, it might be nice to be able to run a different OS on a phone, in practice, this would likely just involve people rooting their Android phones or running Android on iPhones, which people have done. Sort of. Realistically, nobody is going to create a second open source phone operating system, now that one exists, because there just aren't not enough users who care. Maybe you'd have some Android forks that turn off certain security features or whatever, but it will still probably be Android.

And in practice, nobody wants to run a different phone OS. They want to not be limited in what apps they can run on their device. Running a different OS means that all the app developers have to support that other OS, and they're not going to do that, so unless you then run Android or iOS in a VM (and end up with a slower and even-more-hard-to-use device because of the confusing VM UI), running a different non-Android OS would mean being even *more* limited in what apps you can run on your device. Almost exactly nobody other than perhaps Richard Stallman wants this. This is why Linux on the desktop has never gone anywhere and, realistically speaking, almost certainly never will.

So while the idea of being able to run an alternate OS has some merit from the perspective of being able to keep old devices out of the landfill (i.e. being able to flash stock Android on the thing when the hardware manufacturer decides to provide only a couple of years of security updates and zero major OS version updates), it's mostly a solution for crappy hardware companies that provide inadequate operating system support, not a realistic solution for operating systems that limit what you can do.

More than that, companies that do this have a very real risk of being in violation of a number of federal laws. If the person they are calling didn't previously explicitly consent to be robocalled, then it is outright illegal to call a cell phone in this way. And if they're on the do-not-call registry, this is also likely illegal.

More to the point, the more companies do this, the more people will sign up for the do-not-call registry again to make them stop, and you'll end up ruining it for everybody.

Only if you have an Apple Watch. The iPhone app can track steps, but it has no UI to let you switch modes to tell it that you are cycling unless either A. you are doing a Fitness+ workout (which costs money for the subscription) or you have an Apple Watch (which costs $$$).

So it laughably says that I walked 6,477 steps over 2.26 miles and used 195 calories, when in fact, I biked 20.6 miles, mostly with a headwind, and used probably O(800) calories. How it can think I'm walking at those speeds, I have no idea.

Not sure now you got that from my post. There are lots of factors that go into the price of goods, and cost of production is only one of them. But what is certain is that cost represents a floor for the price of goods. The price will never fall below the cost of production and stay there, with the possible exception of the occasional loss leader. That's why anything that fundamentally increases the cost of goods — particularly while reducing the potential for competition — is inherently harmful to a functioning free market.

Over time, markets tend towards increasing efficiency, in the absence of other factors that reduce it. Ignoring that reality is far greater madness, IMO.

Yes, and in those narrowly tailored circumstances, they are advantageous. For example, when countries flood the market with products made below actual cost (e.g. Chinese solar panels a few years back), they can destroy the proper functioning of the free market by causing all other competitors to go under, which is followed by them raising prices once they have no more competition. That sort of market manipulation should never be allowed, and tariffs are a reasonable tool for dealing with that. The same dumping has occurred with raw materials from China, and the same solution is at least potentially reasonable.

Sorry, I should have been more clear. Tariffs are never a reasonable tool for fixing a general trade imbalance. Surgical use of tariffs can of course be useful in very limited situations.

I mean, you're not wrong there. Forcing minimum working conditions, environmental standards, etc. is a reasonable use of tariffs as well. But such tariffs should be designed in such a way that products that meet those standards are exempt, rather than in such a way that all products from a country that often doesn't meet those standards are always penalized. Otherwise, you're creating a perverse incentive to keep polluting or paying people substandard wages. After all, a company in one of those countries that chooses to not pollute and chooses to pay workers a living wage would bear the cost of doing do *and* their products would get hit with the tariff. So they're better off not changing the way they do business unless and until their country's government forces them to do so across the board (which can happen, but is not guaranteed).

Taxes should generally be the same for a given product made in a similar way regardless of the country of origin. Using a tariff to force compliance is great, but only if products that comply are treated equally no matter where they are produced (ostensibly even including products that are not imported).

A tariff is a tax, but not all taxes are tariffs. A VAT is not a tariff, because it is not specific to imported goods.

Tariffs are generally understood by economists to be worse than sales taxes, because they distort the market in consumer-harmful ways, such as reducing competition and reducing the benefits of economies of scale.

Consider two companies that each build silicon chips. One has a manufacturing line that produces CPUs and a line that produces RAM chips. The CPU line is in one country, the RAM line is in another country. Both lines operate at 100% capacity and produce the number of components needed for both countries.

The second company has a similar setup, but because of high tariffs, determines that they need to build RAM and CPUs in both countries. Now you have redundant manufacturing capacity, and that capacity is being used at half capacity. The fixed costs of manufacturing are now double. It is still less than the tariffs would cost, but it does substantially increase the price of goods.

In both of these scenarios, the same amount of manufacturing work is being done in both countries. No country is "winning" with these tariffs. But because there's redundant capacity that is not being used efficiently, the costs for consumers are significantly higher.

This is what tariffs do. Tariffs are madness. They are using a nuclear weapon when what you need is a sniper rifle.

My favorite was a roughly thirty-second section of near absolute silence with someone walking up to the lectern during a church service. Apparently, one of those companies (I forget which) thought that nothing was violating their copyright.

Meanwhile, Sony and Warner-Chappell constantly file false music match claims against orchestras performing public domain music, claiming that the recordings are theirs.

So UMG might be the worst offender, but none of the major record companies are innocent of using false or questionable copyright claims in a manner that the law never intended.

The only thing that's going to stop it is a class action lawyer filing a multi-billion-dollar class action on behalf of tens of thousands of content creators, naming a terrifying number of codefendants, and taking it all the way to SCOTUS.