I'm shocked — shocked that the same company that cancelled the original Star Trek doesn't understand the geek market. The best thing CBS could do for Star Trek would be to sell off its interest in the series entirely.
You are technically correct — the best kind of correct. After they wrote this crappy bill, that's true. The point where they gave in to the Republicans' every demand was before the bill even went into committee. Specifically, the Republicans insisted that they would not pass any single payer bill nor any bill that involved any government entitlement. The result was that the Democrats passed RomneyCare, which is radically different from the original plan, which was to create a modern, single-payer health system similar to what every other first-world nation has.
Security, however, should not really be a problem. You can always encapsulate the traffic on one side of the mesh network and unwrap it on the other side where it goes onto the public Internet, and so long as you trust the node at the other end, that will work fine. And even if you don't, you should be using end-to-end encryption and bidirectional authentication (e.g. TLS) on the connections anyway, so... *shrugs*.
Haven't ever worked for a business that MITM all TLS traffic? It's not hard if you own the pipes longer than any node has been connected. Plus, layer 2 attacks make things quite easy.
The way they do that is by preinstalling their own root certs on the machine. Without that, MITMed connections will fail, because the TLS certs won't be signed by any root known to the browser. You can certainly block requests (on a per-IP basis), but you cannot possibly manipulate the traffic in any useful or interesting way.
You appear to be mistaken as to how the internet works. It already is decentralized chaos. IPFS provides a similar purpose as DNS.
Well, yes and no. IPFS is interesting, but is probably completely impractical as a replacement for the web, simply because things are too decentralized and large-scale caching is usually a bad idea.
That said, the concept of a decentralized name lookup system is a good idea in theory. A better way to think of the problem, IMO, would be as a series of channels provided by some trusted authority (trusted by the user). The user could then access servers/services by name within the context of that channel. Basically, each user would have n DNS servers instead of 1, and if multiple servers provide different responses, you would get a disambiguation page consisting of a blurb from each of the possible pages along with a line that tells which channel provided that page.
Combine this with a peer-to-peer scheme for connecting to each of those trusted authorities, and you have at least the possibility of creating a secure Internet that is not practical for hostile entities (businesses, governments, etc.) to break, albeit one that is significantly slower than the current setup.
And if you aren't concerned about preventing surveillance, you can use the P2P for lookups only, and connect directly to a server if it is reachable. This reduces the performance impact by using P2P only in situations where a node is isolated from the destination. If you mandate that all nodes must be exit nodes, it should even be possible to easily build a near-optimal route in which hosts near the isolated end (or ends) of the link encapsulate traffic as needed, but the bulk of path in-between goes directly between two peers that originally did not know about each other, rather than being passed through all the peers that were used to discover the path in the first place.
Whether anyone would be willing to do such things in our sue-happy world or not is, of course, another matter.
WiFi mesh networks have been possible, and have existed, for a very long time. The problem with them is the very high latency, low security, and problems with faulty equipment causing denial of service issues. People have been reinventing the wheel for a very long time.
Mesh networks have the potential to be a great way to cover a local area reliably. They are, however, infeasible for long-haul communications, because every hop inherently adds latency, and you would need an insane number of hops.
Security, however, should not really be a problem. You can always encapsulate the traffic on one side of the mesh network and unwrap it on the other side where it goes onto the public Internet, and so long as you trust the node at the other end, that will work fine. And even if you don't, you should be using end-to-end encryption and bidirectional authentication (e.g. TLS) on the connections anyway, so... *shrugs*.
This. The original plan by the Democrats would not have had any of those problems. The version that actually got passed, in which they gave in to every single Republican demand for changes, does. Big surprise. Republicans look out for business and neuter any laws that affect businesses so that they can dodge their obligations.
The right fix, of course, is to mandate proportional coverage. It's literally a two-line change in a single spot. Just change:
In accordance with regulations promulgated by the Secretary, an employer to which this Act applies that has more than 200 full-time employees and that offers employees enrollment in 1 or more health benefits plans shall automatically enroll new full- time employees in one of the plans offered (subject to any waiting period authorized by law) and to continue the enrollment of current employees in a health benefits plan offered through the employer.
to:
In accordance with regulations promulgated by the Secretary, an employer to which this Act applies that has more than 200 full-time-equivalent employees and that offers employees enrollment in 1 or more health benefits plans shall automatically enroll all new employees in one of the plans offered (subject to any waiting period authorized by law) and to continue the enrollment of current employees in a health benefits plan offered through the employer. For non-full-time employees, the employer may divide the cost proportionally with the employee based on the number of hours worked by that employee in any given month divided by 120; employees have the right to refuse coverage, accepting the cash equivalent towards a plan offered under this act.
Or leave off that last part if you want to strongly encourage employers to hire only full-time employees. Either way.
I honestly thought you were smart-enough to not fall for the "It doesn't matter to me; so it shouldn't matter to anyone." argument. But apparently not. Case in point: When the AppleWatch 4 came out a few days ago, even the generally Apple-Hating Slashdot crowd was complimenting Apple for the INNOVATION.
Same thing with FaceID: You like TouchID better; so everyone else should, too...
You'll note that I acknowledged that both of those are innovative. That doesn't change the fact that they aren't particularly interesting. In the case of Face ID, they had something that worked well, and replaced it with something that doesn't work nearly as well for a significant percentage of users, just so that they could eliminate that pesky physical button that messed with their notion of form over function. Forbes panned the iPhone X, using words like "suck" precisely because Face ID is so much worse than what we had previously. I can guarantee that if you took a hundred random people who had never used either one and asked them to try both approaches, the overwhelming majority would prefer Touch ID, because you can unlock the device before you have to look at it. With Face ID, you have to actively pay attention to the device for several seconds *before* it is usable, which makes it an exceptionally bad user experience by comparison. So although the technology might be innovative, it should have been as a way to augment Touch ID (e.g. for added security during payments), not replace it entirely. As implemented, it was a mistake, and pretty much the only people who can't see that are Apple engineers and fanbois.
And the watch ECG is neat from a technical perspective; it just isn't enough to make me spend a few hundred bucks on a watch that really won't do much for me.
I have an iPhone 6, and although I haven't installed iOS 12 yet, I can tell you that the speed-improvements in iOS 11.4.1 are quite real, and quite noticeable. Call that "incremental improvements" all you want; it no doubt took some innovate re-design of the OS to accomplish those speedups.
I guarantee that Apple did not redesign their OS in iOS 12, much less a minor patch release of iOS 11. iOS 11.4 was actually a significant performance *regression* for many folks, and iOS 11.4.1 did not fix those problems.
iOS 12 mostly improved the algorithm they use for deciding when to run the CPU at faster speeds, resulting in some nice performance wins. But a re-design? Hardly.
And now we come to the bullshit "DongleGate" meme. There is SO much wrong with what you are saying, I really don't know where to start. But here's some thoughts:
1. NO one walks around with a Dock attached to their laptop. That's just silly.
Actually, yes, I do.
2. If you need more than one multipart USB-C dock, then NO other Laptop would be able to meet your I/O needs "out of the box".
My previous MacBook Pro met my needs until it got stolen. Most of the time, my needs are actually pretty modest: USB-A ports, an SD card slot, and HDMI. I don't know where you got the notion that I had more than one dock, but I do have one, and it basically stays attached to my laptop 24x7, mainly because I use SD cards so often that it would drive me utterly insane to have to pull out an external reader every time I would need one.
And yet, the MacBook would still have THREE more USB-C/TB3 ports available!
And yet, three years later, it is not possible to buy a retractable cord that connects a MacBook Pro to an iPhone without adapters. (One did exist very briefly, but it is no longer available.) USB-C adoption on the Mac was woefully premature, and it should have been on the iPhone *before* the Mac, not the other way around. Then, lots of things would "just work". You could buy USB-C headphones and use them on both devices. You could use USB-C retractable charging cords that are readily avail
If you have to reshoot most of the photos you attempt with a cell phone, then I don't know what to tell you - because I can't imagine a single situation in which this is the phone's fault. Either you're trying to take pictures the phone can't, or you don't know how to properly use and work within the limits of a phone camera.
You misunderstand. I mostly shoot photos in low light situations. I don't have to reshoot most of the photos I attempt. I own a real camera because I know the limits of cell phone cameras, and am not willing to limit my craft to the sorts of shooting environments where cell phones work well.
It's a poor workman that blames the tool.
You're grossly misusing that phrase. It does not mean that a good craftsman can do anything with any tool. Rather, it means that a good craftsman chooses the best tools that he or she can afford so that he or she will not be limited by them more than necessary, and then takes on only the sorts of projects that are realistic within the limits of his or her tools.
That's why I carried *two* DSLRs on my last trip, in addition to my cell phone. I used the cell phone for quick panoramas to post on Facebook, one camera for wide shots, and the other camera for close-up shots. I knew that if I didn't do that, I would probably miss shots while changing lenses. And it really paid off. At the end of a boat trip, I got some amazing shots of a bald eagle landing about 150 feet in front of me just seconds after I shot some wide-angle shots of the marina. If I had been using a cell phone with some screw-on teleconverter, either I would have missed the shot entirely while desperately trying to screw on the teleconverter, or the eagle would have been a small cluster of pixels. Instead, I said, "Whoah!" when I saw it out of the corner of my eye, flying out over the water. It took less than a second to pull up my 5D Mark IV with the attached 100-400 lens from where it was hanging around my neck and start shooting pictures that were absolutely breathtaking. Then, while other people were still futzing with their screw-on lenses, I walked to within about fifty feet of the bird and shot some even closer shots of the eagle perched atop a wooden post. Eventually, some of the cell phone users without teleconverters walked up close to the post and got some shots from under the bird, which is about the only thing you can do with a wide-angle lens. By that time, I was already back on the bus with shots in hand.
A poor craftsman blames his tools. A good craftsman replaces them when he outgrows them.
Sure. For a certain subset of users, it is a useful feature. For the majority of users, it is basically an "oh, that's neat" thing. Those are the sorts of features that Apple usually drops after about three or four years, requiring you to buy a dongle if you want to keep it.:-)
As far as image quality goes... Only four kinds of people really sweat high image quality: Clueless amateurs. Advertisers & commercial photography. Fine art photographers. High end photographers who'll be printing at poster size or larger.
You forgot four groups: Natural-light photographers, sports photographers, portrait photographers, and candid photographers, though arguably it isn't really because those groups need image quality per se so much as because they need light gathering ability, and light gathering ability is a major driver of image quality.
Natural light: The biggest difference between a cell phone camera and, for example, a modern full-frame DSLR is that the DSLR can produce tolerable IQ at crazy ISO levels that a phone can only dream about. To be fair, that comes at a significant price in terms of glass weight and $$$ spent, but for folks doing natural-light or low-light photography, it is often worth the price.
Sports: Getting serviceable shots in most arenas often requires big lenses. You have more light, but you also have more motion, which requires a shorter shutter speed. So you end up needing higher-end gear for the same reason that natural-light photography does.
Portraits: A DSLR's depth of field is narrower at any given f-stop, allowing for better foreground isolation without the sorts of multi-camera fakery that can often produce visibly distracting artifacts.
Candid photography: When trying to take candid shots, the most important thing is being able to get a shot quickly and have it always work. If the shot is blurred because your phone could not stop motion enough in the available light, or if shutter lag causes the shot to become less desirable, or if the phone's camera stalls for two seconds trying to find focus, you are now taking a posed photo instead of a candid shot, and it just doesn't give the same feel.
For me, the reason for owning a real camera is not that I can't (in theory) take any given shot with a smartphone, but rather because it makes the difference between most of the shots working and most of the shots having to be retried several times. And when you're talking about a bald eagle landing on a post, you really can't ask it to go back and circle around again. You can't ask the train or bus to stop and back up. You can't ask the folks who have now seen you to pretend the camera isn't there. And so on. If you only get one shot at a shot, the camera has to "just work" every time, and the laws of physics preclude cell phone cameras from ever really reaching equality with large-lens devices.
The big shock will be when we start to see full-frame DSLRs with global electronic shutters, so that they will be able to do the same sorts of tricks that cell phones do, taking several shots in a row and automatically selecting the one with the least motion blur. That will broaden the gap between cell phone cameras and standalone cameras back to historical levels.
Of course, cell phone cameras will still be better at shooting through wavy glass or small holes in fences, and at photographing the labels that are on the back sides of major appliances, but I digress.
You can buy a dozen "real" cameras, but unless you ram one down your pants you might not have it with you at that unexpected moment when you really want to capture an image.
Why do you want to have pictures down in your pants?
OTOH there is no real reason that high paying STEM jobs can’t be much more widely distributed across the globe and that is exactly what is going to happen.
Addressing this unrelated subject in a separate post. The main reason tech jobs are so clustered is that most people over thirty have an aversion to packing up and moving to a new city to get a job, so businesses tend to locate themselves near where they can get talent easily. Tech started out concentrated in a few places, and remains so because of that inertia, which turns out to be remarkably hard to overcome.
You missed the third choice, which is only made possible by the existence of mostly free global markets: Plant apple orchards in countries where labor is cheap, then bulldoze our apple orchards and replace them with food crops that can be mostly harvested by automation (wheat, corn, etc.). Then, import the apples while exporting grains.
More magical thinking. Agriculture simply doesn’t work that way.
Sure it does—not over the course of months, but it very much works that way. Historically, most of our fruit in the U.S. was grown here in the U.S. These days, most of it is imported. Since 2010, the amount of fruit produced in the U.S. has declined by nearly 30%, while the amount of grain increased by nearly 40%. And over the last few decades, decreased demand for tobacco has resulted in tobacco farms switching to various grains, demand for neat gadgets has replaced apple orchards with Apple office buildings, etc.
For the most part, fertile topsoil is fertile topsoil. As long as there is enough root structure to prevent runoff, you can pretty much grow whatever is needed wherever you need to grow it, unless what you're trying to grow has specific temperature limitations or sun hour requirements. You can grow apples anywhere from Alaska all the way down to the equator. You grow them differently in different places (aggressive pruning and annual leaf stripping at the equator, for example), but they will grow anywhere that humans can realistically survive. Some other fruits are more picky, like bananas and citrus (which don't like cold temperatures), but these have always been largely grown outside the U.S. anyway. And certain berries don't like heat, but there are varieties that are significantly more tolerant than others. So farmland isn't perfectly fungible, but it is pretty close.
To be fair, some older fruit orchards may be unsuitable for growing some other types of plants because of lead arsenate contamination in the soil, but that's sort of a side issue, and an artifact of poor government oversight over pesticides, rather than anything fundamental about the soil. And obviously, digging up tree roots is a pain, too, but it can be done, and indeed, that's the only way that new farmland gets created, typically, i.e. it's a common thing to do. It makes little difference whether you're felling a forest or an orchard. And, of course, most of the orchards are near the coast, where land is valuable, so a lot of it is making way for housing, rather than farmland. But that land is then being replaced by irrigation pushing agriculture out into previously unusable areas elsewhere in the country, so on the whole, we're gaining agricultural output.
Have you even BOTHERED to look at the innovation that IS there?
1. Most advanced facial recognition in a mobile device.
2. Smartwatch with built-in FDA-APPROVED ECG.
I'll give you those two, though neither is something that I care about, even slightly. For me, the fingerprint readers were much easier to use (i.e. that's a step backwards, not forwards), and smartwatches are an expensive toy with minimal actual utility.
3. 64 bit Homegrown SoCs (with Homegrown GPUs) that are hands-down best-in-class.
4. Augmented Reality that isn't a joke.
Incremental improvements to the state of the art are not innovation.
5. Laptops with the most amount of I/O expandability on the Planet.
Are we talking about the laptops that have only USB-C and dropped all other I/O, including things that a lot of pros use, like the SD card slots, forcing people to put the reliability of their machines at risk by permanently carrying them around with a dock attached? I would argue that the only reason Apple has more "expandability" is because they've dropped nearly all actual I/O, resulting in a laptop that requires piles of clumsy dongles and crap just to be able to use it in ways that the old machines could do out of the box. If that's innovation, then I'd like a much *less* "innovative" Apple, thank you very much. If they innovate much more like that, I'm going to need a second laptop bag just to carry all the dongles. (I often have to use a setup involving a Thunderbolt-2-to-Thunderbolt adapter chained to a Thunderbolt-to-FireWire-800 adapter chained to a FireWire-800-to-FireWire-400 adapter. This is *not* pleasant.)
6. A mobile OS that specifically (and markedly!) IMPROVES the performance of OLDER Devices.
Again, minor incremental improvements, not innovation.
None of that changes the fact that cell towers put out orders of magnitude more power than phones, and being near them is an associated risk that much greater despite your dithering about power levels.
Actually, it does change that "fact". These are the 5G equivalent of microcells or even picocells, which put out much less power than primary cell phone towers (2–20W for a microcell or 250mW–2W for a picocell). In fact, picocell output is comparable to the output of a cell phone (which can produce 2–3W when talking to a distant tower). Heck, even a full tower (160W max, 40W typical) is less than two orders of magnitude more than your cell phone when it is shouting at maximum volume towards a distant tower.
In fact, in short-range radio system designs, the base station power output is almost always similar to the individual device's power output. These high-density 5G cells might have slightly higher output than your phone, but not by orders of magnitude, and probably not even by an order of magnitude.
Now as you (or some other AC) said, the exposure is potentially over a greater period of time, in that the cell tower transmits when any device in the area is active, but the tower covers a much smaller area, so the number of active devices is lower proportional to the area covered. The coverage area, in turn, is proportional to the square of the radius. Its worst-case power output also increases also proportional to the square of the radius (though each cell only covers a narrow vertical and horizontal angle, so this part isn't *equal* to the square of the radius). The total exposure is proportional to the product of those numbers. Thus, the total exposure is proportional to the tower's transmission radius to the fourth power even if you're look at current picocells (omnidirectional radiators). Smaller cells that are closer to you are always going to be much safer to be around.
When you add in beamforming (which allows the tower to transmit highly directionally, thus minimizing EM exposure except in a straight line from the tower to each active device), these 5G picocells should produce much lower average exposure than existing large towers.
If you want to argue about safety, the right argument is about whether the frequency of EM used is safe. That's a bigger question. That said even if it is dangerous, you will still, on average, get less exposure if the transmission range (and thus the transmit power) is kept to a minimum by using a larger number of stations that are closer together. So in the best case, these people are making their coverage worse, and in the worst case, they are increasing their cancer risk by causing their phones to use more power. There is no form of mathematics under the sun in which using distant towers instead of MIMO picocells reduces your EM exposure, period.
Now, an example to the contrary for semi-skilled labor : truck drivers. Truck drivers are getting more and more sparse. ( example : incentives to drive trucks and dedicated trucking sites like this one : http://www.findatruckingjob.co... ). In this case you have an industry which is exploding with need, but cannot find enough resource (people) to do the job, so wages are increasing (perhaps not as fast as they should, but that's another point).
That's why this will be one of the first categories of jobs that will be replaced by automation. Right now, the logistics companies are trying to shift more towards rail, but the long-term solution is to replace long-haul trucking with self-driving vehicles. Tesla is working on it for short-haul right now. In twenty years, there won't be any truck-driving jobs left.
So it is actually a good thing that there are fewer and fewer people going into truck driving. That means fewer people will lose their jobs when inevitably makes their jobs obsolete. Yes, in the short term, it might make sense to take advantage of the wages while they are high, but if you do that, you'd better save every penny you can, knowing that the wage glut is only temporary, and that the wage collapse that will follow will be permanent.
So we still have a lot of apples to pick. The solutions are to get Americans to do it for much lower wages than they currently find acceptable, import labor that will accept the prevailing wage, or do it with robots. As you and the article note, the robot thing is just magical thinking.
You missed the third choice, which is only made possible by the existence of mostly free global markets: Plant apple orchards in countries where labor is cheap, then bulldoze our apple orchards and replace them with food crops that can be mostly harvested by automation (wheat, corn, etc.). Then, import the apples while exporting grains.
Nope, you get a much, much bigger dose from being near a tower than you do a cell phone. You're being stupid lol. It's also about cumulative exposure - nobody uses a phone 24/7, a residential community can't move away from towers.
The closer the devices are to the tower, the less power the tower uses, too. The close spacing of 5G towers means that everybody's exposure is fairly consistent, rather than having hotspots near towers where people get more exposure. The range of a 5G tower is measured in tens to hundreds of *feet*, i.e. power levels are more comparable to Wi-Fi than to existing cell towers, which means you can't get 5G service from towers that are blocks away, so moving towers farther away from people has the effect of preventing 5G service outright.
But the main reason you're wrong in your analysis is that 5G uses a MIMO design (much like 802.11n) with beamforming on its towers so that the emissions can be focused precisely on the location of the active device. Because of this design, exposure levels fall off significantly the farther you get away from the device that the tower is talking to. Because of that design, whether you are next to a tower or not, you should get significantly less exposure than you would get with existing cell tower services.
But there is one app that REALLY benefits from the pressure sensitive screen: my virtual piano app. Having a pressure sensitive screen means that I can hit my piano keys at different pressures and achieve appropriate piano key velocities for more expressive playing. SOOO much better than a non-pressure sensitive screen.
Virtual painting apps benefit, too, for pretty much the same reason. But for general app purposes, it really is an almost useless feature, IMO.
Trump's campaign rhetoric really scared the crap out of many people. And not in a "OMG, Republicans nonsense!" way. In a "Are we going to start having to hide Muslim families in our basements?" way.
This. Among other things, Trump promised a Muslim travel ban. Given that Google has multiple offices in the Middle East (Dubai, Haifa, and Istanbul), a sizable number of Google employees would have been directly affected by that, had it been implemented as promised (and some were probably affected anyway).
At this point, I think the main thing protecting everyone is the sheer incompetence and disorganization of his entire administration. Its clear now that he's far more interested Tweeting and continuing to hold those campaign rallies than in actually doing the job of President.
When he got elected, I said that this would be the truest test of checks and balances to date, but I never imagined he would be checked and balanced by his own staff. This is really turning out to be far more entertaining than I ever would have imagined.
Your unlocked phone is linked to whichever service it's initially activated with. Even an unlocked phone, you can't take out an AT&T card and swap in a Sprint card and just turn it back on and have it work.
Does unlocked mean something different over there? Because that's exactly what unlocked means, at least here in the uk, unless this is a thing specific to iphones which really wouldn't surprise me.
The issue is CDMA. That protocol doesn't use SIM cards. An unlocked CDMA iPhone can be moved to any GSM-based carrier, but can't be moved to a different CDMA-based carrier than the one it was originally activated on. And if you originally activate it on a non-CDMA carrier, the carriers generally won't let you activate it on their service even if it is capable of being activated for CDMA. (This is a cellular provider policy limitation, not a technical one.)
The main problem with Bluetooth is latency. If you're watching a TV show or something, the latency with some Bluetooth hardware is enough to be vomit-inducing, because the audio lags so far behind the video. Until Apple figures out how to get their Bluetooth stack to accurately report latency all the way up the stack to apps, Bluetooth will continue to suck.
Compounding this is the fact that Bluetooth isn't one standard, but rather a pile of competing standards. Android devices typically support aptX from Qualcomm, but Apple doesn't want to pay to license that, so they use their own, incompatible low-latency standard based on AAC (which they were already licensing for iTunes audio playback, and thus probably doesn't cost them anything). Although it is possible for Bluetooth devices to support both formats, there is a licensing cost incurred when you do so, and as a result, a decent percentage of devices support only one format or the other.
In USA you should be building with robots not laborers, and robots don't ask for wages.
Electronic devices like iPhones are already almost entirely built by robots. Robots cut the circuit boards to size, pick and place the components on the boards, and wave solder those components. Specialized testing hardware performs hardware-level verification to ensure that all this happened correctly. The only things that humans do are:
Load new component reels into the pick-and-place robot.
Stick the board in the case (*maybe* with screws).
Plug in the digitizer/screen stack.
Fasten the self-stick battery to the back.
Plug the battery in.
Snap the case together.
*Maybe* screw in a couple of screws to hold the bottom of the case together.
These are things that can't easily be done by robots. Everything that can easily be done by robots already is.
Slashdot Mirror
I'm shocked — shocked that the same company that cancelled the original Star Trek doesn't understand the geek market. The best thing CBS could do for Star Trek would be to sell off its interest in the series entirely.
I assume it logs out when you log out of gmail, too, no?
And if it turns out to be true, then I guess you could say Qualcomm's technology got Xeroxed (pun intended).
You are technically correct — the best kind of correct. After they wrote this crappy bill, that's true. The point where they gave in to the Republicans' every demand was before the bill even went into committee. Specifically, the Republicans insisted that they would not pass any single payer bill nor any bill that involved any government entitlement. The result was that the Democrats passed RomneyCare, which is radically different from the original plan, which was to create a modern, single-payer health system similar to what every other first-world nation has.
The way they do that is by preinstalling their own root certs on the machine. Without that, MITMed connections will fail, because the TLS certs won't be signed by any root known to the browser. You can certainly block requests (on a per-IP basis), but you cannot possibly manipulate the traffic in any useful or interesting way.
Well, yes and no. IPFS is interesting, but is probably completely impractical as a replacement for the web, simply because things are too decentralized and large-scale caching is usually a bad idea.
That said, the concept of a decentralized name lookup system is a good idea in theory. A better way to think of the problem, IMO, would be as a series of channels provided by some trusted authority (trusted by the user). The user could then access servers/services by name within the context of that channel. Basically, each user would have n DNS servers instead of 1, and if multiple servers provide different responses, you would get a disambiguation page consisting of a blurb from each of the possible pages along with a line that tells which channel provided that page.
Combine this with a peer-to-peer scheme for connecting to each of those trusted authorities, and you have at least the possibility of creating a secure Internet that is not practical for hostile entities (businesses, governments, etc.) to break, albeit one that is significantly slower than the current setup.
And if you aren't concerned about preventing surveillance, you can use the P2P for lookups only, and connect directly to a server if it is reachable. This reduces the performance impact by using P2P only in situations where a node is isolated from the destination. If you mandate that all nodes must be exit nodes, it should even be possible to easily build a near-optimal route in which hosts near the isolated end (or ends) of the link encapsulate traffic as needed, but the bulk of path in-between goes directly between two peers that originally did not know about each other, rather than being passed through all the peers that were used to discover the path in the first place.
Whether anyone would be willing to do such things in our sue-happy world or not is, of course, another matter.
Mesh networks have the potential to be a great way to cover a local area reliably. They are, however, infeasible for long-haul communications, because every hop inherently adds latency, and you would need an insane number of hops.
Security, however, should not really be a problem. You can always encapsulate the traffic on one side of the mesh network and unwrap it on the other side where it goes onto the public Internet, and so long as you trust the node at the other end, that will work fine. And even if you don't, you should be using end-to-end encryption and bidirectional authentication (e.g. TLS) on the connections anyway, so... *shrugs*.
This. The original plan by the Democrats would not have had any of those problems. The version that actually got passed, in which they gave in to every single Republican demand for changes, does. Big surprise. Republicans look out for business and neuter any laws that affect businesses so that they can dodge their obligations.
The right fix, of course, is to mandate proportional coverage. It's literally a two-line change in a single spot. Just change:
to:
Or leave off that last part if you want to strongly encourage employers to hire only full-time employees. Either way.
I honestly thought you were smart-enough to not fall for the "It doesn't matter to me; so it shouldn't matter to anyone." argument. But apparently not. Case in point: When the AppleWatch 4 came out a few days ago, even the generally Apple-Hating Slashdot crowd was complimenting Apple for the INNOVATION.
Same thing with FaceID: You like TouchID better; so everyone else should, too...
You'll note that I acknowledged that both of those are innovative. That doesn't change the fact that they aren't particularly interesting. In the case of Face ID, they had something that worked well, and replaced it with something that doesn't work nearly as well for a significant percentage of users, just so that they could eliminate that pesky physical button that messed with their notion of form over function. Forbes panned the iPhone X, using words like "suck" precisely because Face ID is so much worse than what we had previously. I can guarantee that if you took a hundred random people who had never used either one and asked them to try both approaches, the overwhelming majority would prefer Touch ID, because you can unlock the device before you have to look at it. With Face ID, you have to actively pay attention to the device for several seconds *before* it is usable, which makes it an exceptionally bad user experience by comparison. So although the technology might be innovative, it should have been as a way to augment Touch ID (e.g. for added security during payments), not replace it entirely. As implemented, it was a mistake, and pretty much the only people who can't see that are Apple engineers and fanbois.
And the watch ECG is neat from a technical perspective; it just isn't enough to make me spend a few hundred bucks on a watch that really won't do much for me.
I have an iPhone 6, and although I haven't installed iOS 12 yet, I can tell you that the speed-improvements in iOS 11.4.1 are quite real, and quite noticeable. Call that "incremental improvements" all you want; it no doubt took some innovate re-design of the OS to accomplish those speedups.
I guarantee that Apple did not redesign their OS in iOS 12, much less a minor patch release of iOS 11. iOS 11.4 was actually a significant performance *regression* for many folks, and iOS 11.4.1 did not fix those problems.
iOS 12 mostly improved the algorithm they use for deciding when to run the CPU at faster speeds, resulting in some nice performance wins. But a re-design? Hardly.
And now we come to the bullshit "DongleGate" meme. There is SO much wrong with what you are saying, I really don't know where to start. But here's some thoughts:
1. NO one walks around with a Dock attached to their laptop. That's just silly.
Actually, yes, I do.
2. If you need more than one multipart USB-C dock, then NO other Laptop would be able to meet your I/O needs "out of the box".
My previous MacBook Pro met my needs until it got stolen. Most of the time, my needs are actually pretty modest: USB-A ports, an SD card slot, and HDMI. I don't know where you got the notion that I had more than one dock, but I do have one, and it basically stays attached to my laptop 24x7, mainly because I use SD cards so often that it would drive me utterly insane to have to pull out an external reader every time I would need one.
And yet, the MacBook would still have THREE more USB-C/TB3 ports available!
And yet, three years later, it is not possible to buy a retractable cord that connects a MacBook Pro to an iPhone without adapters. (One did exist very briefly, but it is no longer available.) USB-C adoption on the Mac was woefully premature, and it should have been on the iPhone *before* the Mac, not the other way around. Then, lots of things would "just work". You could buy USB-C headphones and use them on both devices. You could use USB-C retractable charging cords that are readily avail
And nothing of value was lost.
You misunderstand. I mostly shoot photos in low light situations. I don't have to reshoot most of the photos I attempt. I own a real camera because I know the limits of cell phone cameras, and am not willing to limit my craft to the sorts of shooting environments where cell phones work well.
You're grossly misusing that phrase. It does not mean that a good craftsman can do anything with any tool. Rather, it means that a good craftsman chooses the best tools that he or she can afford so that he or she will not be limited by them more than necessary, and then takes on only the sorts of projects that are realistic within the limits of his or her tools.
That's why I carried *two* DSLRs on my last trip, in addition to my cell phone. I used the cell phone for quick panoramas to post on Facebook, one camera for wide shots, and the other camera for close-up shots. I knew that if I didn't do that, I would probably miss shots while changing lenses. And it really paid off. At the end of a boat trip, I got some amazing shots of a bald eagle landing about 150 feet in front of me just seconds after I shot some wide-angle shots of the marina. If I had been using a cell phone with some screw-on teleconverter, either I would have missed the shot entirely while desperately trying to screw on the teleconverter, or the eagle would have been a small cluster of pixels. Instead, I said, "Whoah!" when I saw it out of the corner of my eye, flying out over the water. It took less than a second to pull up my 5D Mark IV with the attached 100-400 lens from where it was hanging around my neck and start shooting pictures that were absolutely breathtaking. Then, while other people were still futzing with their screw-on lenses, I walked to within about fifty feet of the bird and shot some even closer shots of the eagle perched atop a wooden post. Eventually, some of the cell phone users without teleconverters walked up close to the post and got some shots from under the bird, which is about the only thing you can do with a wide-angle lens. By that time, I was already back on the bus with shots in hand.
A poor craftsman blames his tools. A good craftsman replaces them when he outgrows them.
Sure. For a certain subset of users, it is a useful feature. For the majority of users, it is basically an "oh, that's neat" thing. Those are the sorts of features that Apple usually drops after about three or four years, requiring you to buy a dongle if you want to keep it. :-)
You forgot four groups: Natural-light photographers, sports photographers, portrait photographers, and candid photographers, though arguably it isn't really because those groups need image quality per se so much as because they need light gathering ability, and light gathering ability is a major driver of image quality.
For me, the reason for owning a real camera is not that I can't (in theory) take any given shot with a smartphone, but rather because it makes the difference between most of the shots working and most of the shots having to be retried several times. And when you're talking about a bald eagle landing on a post, you really can't ask it to go back and circle around again. You can't ask the train or bus to stop and back up. You can't ask the folks who have now seen you to pretend the camera isn't there. And so on. If you only get one shot at a shot, the camera has to "just work" every time, and the laws of physics preclude cell phone cameras from ever really reaching equality with large-lens devices.
The big shock will be when we start to see full-frame DSLRs with global electronic shutters, so that they will be able to do the same sorts of tricks that cell phones do, taking several shots in a row and automatically selecting the one with the least motion blur. That will broaden the gap between cell phone cameras and standalone cameras back to historical levels.
Of course, cell phone cameras will still be better at shooting through wavy glass or small holes in fences, and at photographing the labels that are on the back sides of major appliances, but I digress.
Why do you want to have pictures down in your pants?
Addressing this unrelated subject in a separate post. The main reason tech jobs are so clustered is that most people over thirty have an aversion to packing up and moving to a new city to get a job, so businesses tend to locate themselves near where they can get talent easily. Tech started out concentrated in a few places, and remains so because of that inertia, which turns out to be remarkably hard to overcome.
Sure it does—not over the course of months, but it very much works that way. Historically, most of our fruit in the U.S. was grown here in the U.S. These days, most of it is imported. Since 2010, the amount of fruit produced in the U.S. has declined by nearly 30%, while the amount of grain increased by nearly 40%. And over the last few decades, decreased demand for tobacco has resulted in tobacco farms switching to various grains, demand for neat gadgets has replaced apple orchards with Apple office buildings, etc.
For the most part, fertile topsoil is fertile topsoil. As long as there is enough root structure to prevent runoff, you can pretty much grow whatever is needed wherever you need to grow it, unless what you're trying to grow has specific temperature limitations or sun hour requirements. You can grow apples anywhere from Alaska all the way down to the equator. You grow them differently in different places (aggressive pruning and annual leaf stripping at the equator, for example), but they will grow anywhere that humans can realistically survive. Some other fruits are more picky, like bananas and citrus (which don't like cold temperatures), but these have always been largely grown outside the U.S. anyway. And certain berries don't like heat, but there are varieties that are significantly more tolerant than others. So farmland isn't perfectly fungible, but it is pretty close.
To be fair, some older fruit orchards may be unsuitable for growing some other types of plants because of lead arsenate contamination in the soil, but that's sort of a side issue, and an artifact of poor government oversight over pesticides, rather than anything fundamental about the soil. And obviously, digging up tree roots is a pain, too, but it can be done, and indeed, that's the only way that new farmland gets created, typically, i.e. it's a common thing to do. It makes little difference whether you're felling a forest or an orchard. And, of course, most of the orchards are near the coast, where land is valuable, so a lot of it is making way for housing, rather than farmland. But that land is then being replaced by irrigation pushing agriculture out into previously unusable areas elsewhere in the country, so on the whole, we're gaining agricultural output.
Have you even BOTHERED to look at the innovation that IS there?
1. Most advanced facial recognition in a mobile device.
2. Smartwatch with built-in FDA-APPROVED ECG.
I'll give you those two, though neither is something that I care about, even slightly. For me, the fingerprint readers were much easier to use (i.e. that's a step backwards, not forwards), and smartwatches are an expensive toy with minimal actual utility.
3. 64 bit Homegrown SoCs (with Homegrown GPUs) that are hands-down best-in-class.
4. Augmented Reality that isn't a joke.
Incremental improvements to the state of the art are not innovation.
5. Laptops with the most amount of I/O expandability on the Planet.
Are we talking about the laptops that have only USB-C and dropped all other I/O, including things that a lot of pros use, like the SD card slots, forcing people to put the reliability of their machines at risk by permanently carrying them around with a dock attached? I would argue that the only reason Apple has more "expandability" is because they've dropped nearly all actual I/O, resulting in a laptop that requires piles of clumsy dongles and crap just to be able to use it in ways that the old machines could do out of the box. If that's innovation, then I'd like a much *less* "innovative" Apple, thank you very much. If they innovate much more like that, I'm going to need a second laptop bag just to carry all the dongles. (I often have to use a setup involving a Thunderbolt-2-to-Thunderbolt adapter chained to a Thunderbolt-to-FireWire-800 adapter chained to a FireWire-800-to-FireWire-400 adapter. This is *not* pleasant.)
6. A mobile OS that specifically (and markedly!) IMPROVES the performance of OLDER Devices.
Again, minor incremental improvements, not innovation.
Actually, it does change that "fact". These are the 5G equivalent of microcells or even picocells, which put out much less power than primary cell phone towers (2–20W for a microcell or 250mW–2W for a picocell). In fact, picocell output is comparable to the output of a cell phone (which can produce 2–3W when talking to a distant tower). Heck, even a full tower (160W max, 40W typical) is less than two orders of magnitude more than your cell phone when it is shouting at maximum volume towards a distant tower.
In fact, in short-range radio system designs, the base station power output is almost always similar to the individual device's power output. These high-density 5G cells might have slightly higher output than your phone, but not by orders of magnitude, and probably not even by an order of magnitude.
Now as you (or some other AC) said, the exposure is potentially over a greater period of time, in that the cell tower transmits when any device in the area is active, but the tower covers a much smaller area, so the number of active devices is lower proportional to the area covered. The coverage area, in turn, is proportional to the square of the radius. Its worst-case power output also increases also proportional to the square of the radius (though each cell only covers a narrow vertical and horizontal angle, so this part isn't *equal* to the square of the radius). The total exposure is proportional to the product of those numbers. Thus, the total exposure is proportional to the tower's transmission radius to the fourth power even if you're look at current picocells (omnidirectional radiators). Smaller cells that are closer to you are always going to be much safer to be around.
When you add in beamforming (which allows the tower to transmit highly directionally, thus minimizing EM exposure except in a straight line from the tower to each active device), these 5G picocells should produce much lower average exposure than existing large towers.
If you want to argue about safety, the right argument is about whether the frequency of EM used is safe. That's a bigger question. That said even if it is dangerous, you will still, on average, get less exposure if the transmission range (and thus the transmit power) is kept to a minimum by using a larger number of stations that are closer together. So in the best case, these people are making their coverage worse, and in the worst case, they are increasing their cancer risk by causing their phones to use more power. There is no form of mathematics under the sun in which using distant towers instead of MIMO picocells reduces your EM exposure, period.
That's why this will be one of the first categories of jobs that will be replaced by automation. Right now, the logistics companies are trying to shift more towards rail, but the long-term solution is to replace long-haul trucking with self-driving vehicles. Tesla is working on it for short-haul right now. In twenty years, there won't be any truck-driving jobs left.
So it is actually a good thing that there are fewer and fewer people going into truck driving. That means fewer people will lose their jobs when inevitably makes their jobs obsolete. Yes, in the short term, it might make sense to take advantage of the wages while they are high, but if you do that, you'd better save every penny you can, knowing that the wage glut is only temporary, and that the wage collapse that will follow will be permanent.
You missed the third choice, which is only made possible by the existence of mostly free global markets: Plant apple orchards in countries where labor is cheap, then bulldoze our apple orchards and replace them with food crops that can be mostly harvested by automation (wheat, corn, etc.). Then, import the apples while exporting grains.
The closer the devices are to the tower, the less power the tower uses, too. The close spacing of 5G towers means that everybody's exposure is fairly consistent, rather than having hotspots near towers where people get more exposure. The range of a 5G tower is measured in tens to hundreds of *feet*, i.e. power levels are more comparable to Wi-Fi than to existing cell towers, which means you can't get 5G service from towers that are blocks away, so moving towers farther away from people has the effect of preventing 5G service outright.
But the main reason you're wrong in your analysis is that 5G uses a MIMO design (much like 802.11n) with beamforming on its towers so that the emissions can be focused precisely on the location of the active device. Because of this design, exposure levels fall off significantly the farther you get away from the device that the tower is talking to. Because of that design, whether you are next to a tower or not, you should get significantly less exposure than you would get with existing cell tower services.
Virtual painting apps benefit, too, for pretty much the same reason. But for general app purposes, it really is an almost useless feature, IMO.
This. Among other things, Trump promised a Muslim travel ban. Given that Google has multiple offices in the Middle East (Dubai, Haifa, and Istanbul), a sizable number of Google employees would have been directly affected by that, had it been implemented as promised (and some were probably affected anyway).
When he got elected, I said that this would be the truest test of checks and balances to date, but I never imagined he would be checked and balanced by his own staff. This is really turning out to be far more entertaining than I ever would have imagined.
Your unlocked phone is linked to whichever service it's initially activated with. Even an unlocked phone, you can't take out an AT&T card and swap in a Sprint card and just turn it back on and have it work.
Does unlocked mean something different over there? Because that's exactly what unlocked means, at least here in the uk, unless this is a thing specific to iphones which really wouldn't surprise me.
The issue is CDMA. That protocol doesn't use SIM cards. An unlocked CDMA iPhone can be moved to any GSM-based carrier, but can't be moved to a different CDMA-based carrier than the one it was originally activated on. And if you originally activate it on a non-CDMA carrier, the carriers generally won't let you activate it on their service even if it is capable of being activated for CDMA. (This is a cellular provider policy limitation, not a technical one.)
The main problem with Bluetooth is latency. If you're watching a TV show or something, the latency with some Bluetooth hardware is enough to be vomit-inducing, because the audio lags so far behind the video. Until Apple figures out how to get their Bluetooth stack to accurately report latency all the way up the stack to apps, Bluetooth will continue to suck.
Compounding this is the fact that Bluetooth isn't one standard, but rather a pile of competing standards. Android devices typically support aptX from Qualcomm, but Apple doesn't want to pay to license that, so they use their own, incompatible low-latency standard based on AAC (which they were already licensing for iTunes audio playback, and thus probably doesn't cost them anything). Although it is possible for Bluetooth devices to support both formats, there is a licensing cost incurred when you do so, and as a result, a decent percentage of devices support only one format or the other.
Electronic devices like iPhones are already almost entirely built by robots. Robots cut the circuit boards to size, pick and place the components on the boards, and wave solder those components. Specialized testing hardware performs hardware-level verification to ensure that all this happened correctly. The only things that humans do are:
These are things that can't easily be done by robots. Everything that can easily be done by robots already is.