If MHL was cheaper, then why wouldn't Apple have just gone with MHL?
...
Seriously? The answer is not only obvious, it's been pointed out a few times in this thread already. If Apple went with MHL, they would save a few cents (amortized development costs, certainly) on each iPhone, but they'd lose several dollars on every customer who just bought a cable off Monoprice or Amazon or wherever.
Of course, it would be good for the customer; they'd have more options, and could stream video for less money than Apple's ludicrous proprietary connector costs, but poor nearly-broke Apple's bottom line would suffer so, so horribly...
Besides, this way they can call it "innovation"! Just because the new way is stupid and overpriced doesn't mean it isn't new!
Most people browse the vast majority of the web via HTTP. Even leaving aside sites that don't even support HTTPS, damn near everybody will visit an HTTP page at some point. Hell, Slashdot auto-redirects from HTTPS back to HTTP. Absolutely no need to MitM SSL connections (which they'd have to get an intermediate trusted CA cert for anyhow).
I'm amazed you know that much, and yet are apparently still so ignorant. The idea was that dynamite (NOT TNT; they aren't the same thing) and the other munitions he invented would be a terrible thing to be remembered for (an obituary accidentally published prior to his death did in fact make a great deal out of his works in the field of armaments (calling him "the merchant of death"). He established the prizes, donating the vast majority of his estate to the purpose, as a way to further the causes of progress and peace.
It's been claimed that he (as many people before and after would think, of similar inventions) actually hoped dynamite would make war *too* costly, so that nobody would dare engage in it. I haven't found proper citation for that, though. So far, the only weapon to come close to that has been atomic bombs on long-range missiles, which have deterred war only between superpowers.
Trivially easy to canonicalize that to YourName@gmail.com, and since that approach is so well-known, any competent spammer (not a self-contradiction, nice though it would be; there's a lot of money to be made) will be able to strip such "custom" addresses to the real address. If you want this approach to actually work, you need to blacklist the root address (yourname@) using filters (I'm assuming Gmail filters cna handle that) and only accept mail that has the identifying tag.
Considering that Windows has had built-in dictation support for many years now as part of their accessibility stuff. It needs to be trained or it sucks horribly, but it's there, and the training doesn't take that long. I wouldn't be surprised if the default MS implementation even on consumer SKUs just used that.
We already have mandatory vehicle inspections in many, if not all, states. They're called emissions tests, and things like the odometer and vehicle ID are already logged or easily could be. It's not difficult to imagine that vehicle license registration requires an odometer tax, and do away with the fuel tax. In fact, it would be easy.
The hard part is dealing with the people who complain that they shouldn't have to pay public road taxes for vehicles which only operate on private property, but of course, they're already paying those taxes when they buy fuel for the vehicles. Big whoop.
One is also generally achieved by about the time of your legal maturity, the other requires spending years more learning things which are completely unnecessary for most jobs (sometimes called "unskilled", but really just meaning no need for extensive math/science/engineering/language skills. You don't need a ME to change oil and spark plugs in a car, you don't need an English degree to handle insurance claims, you don't need a... I'm not even sure what it would be... degree to schlep stuff around between offices.
The biggest problem, as I see it, is that it's a vicious cycle of it being *expected* that you get a real job IFF you have a degree, which leads to high schools being degraded to the point of being nothing more than college prep (when they really should be suitable prep for many careers with only minimal job training), which leads to anybody reasonably qualified needing to go to college to have a decent chance, which reinforces the impression that you need a degree to get a decent job...
The Cell processor is what, roughly 6 PPC cores at 3.something GHz? The PS4 has 8 x64 cores at (according to previous comments in this thread) 2GHz.
There's a pretty low real-world cap on the performance of dynamic recompilation. Don't forget that since we're dealing with consoles (very specific hardware target, you also face the complete loss of optimization from compilers that built for the peculiarities of the Cell (cache sizes, which instructions are fastest, etc.) when that CPU bust be emulated or the code gets dynreced. Even assuming that the two extra x64 cores will be able to do the full dynrec for the other six that *must* run in parallel, there's the question of how fast each core is. What's the real-world operations per second of a Cell core from a PS3 compared to a Jaguar core from a PS4?
Oh, and then there's the GPU emulation. This may be even trickier, because unlike on a PC (where you have no idea what GPU your code will run on, so must keep it generic), console graphics code is very, very tightly optimized for each console it runs on. It must be; there's no other way to get even vaguely decent-looking graphics (by today's standard) out of hardware that outdated. The problem for the next generation, then, is providing an absolutely perfect re-creation of all the quirks of that old GPU. If the GPU architecture has changed (as it has), that means basically a whole additional emulation or dynrec layer, with once again the associated performance hits and potential difficulty maintaining the same speed per emulated processor on new architectures whose clock speeds really haven't improved much.
Emulating the original Xbox on the 360 was relatively easy, despite the massive differences in infrastructure; the CPU was absurdly more powerful, as was the GPU. The same is true of the PS2 and PS3, although it took them a while to get around to writing the emulation layer. However, progress in computing power since then has gone wide more than deep. Clock speeds aren't that much better than they used to be, and may in fact be worse. Instructions-per-clock-per-core may be improving, but nowhere near fast enough to make up for hit imposed by switching architectures. Instead, the number of cores has increased massively. The Cell was a multi-processing beast when it came out; today, it's not that much better than the chip in a decent tablet on that metric. More parallelism doesn't necessarily make up for the performance hit of emulation, though, and the new CPU isn't vastly more parallel than the Cell was.
I doubt many (if any) HD2 purchases these days end up running the stock OS much. The HD2 is ridiculously easy to port to, from everything I've seen. It's had (unofficial) ports of WP7, Android, Maemo/Meego, Ubuntu, and Windows RT, and probably other OSes. Its hardware is dated, but not uselessly so, and it's a wonderfully hackable device and can be dual (or more, if you don't mind the storage hit) booted; a rare capability in a phone.
There were search engines before Google. Lycos was even pretty decent by pre-Google standards; other people had other preferences.
No, I don't know how well-known Python (the language) was back then, and I don't know whether ordinary people would have thought "I should do a web search on that" in those days, but I'm pretty sure the information would have been available. Hell, when was python.org registered? The copyright notice at the bottom begins with 1990.
Much though I despise the term, "cloud computing" is not meaningless or even sensationalistic, and there are some very big companies who have built their success on cloud computing. It is, perhaps, over-hyped and watered down, and it's undeniably a buzzword. It's also pretty misunderstood, as you yourself are evidence of (the basic concept is simply "hosted computing services"). However, to say you think a company lacks credibility because they speak of cloud computing is, frankly, idiotic.
The first part of your post makes sense, though. I could understand a company that wasn't deeply involved with computer software might not be aware of a programming language, even a common one, although you'd think they might make at least a trivial web search. However, for a software-focused company to be unaware of Python's importance is, frankly, unbelievable. This guy is just trying to pull whatever damage control he can.
XP is a 12-year-old OS. Why are you comparing it to modern OSes?
That said, I don't believe Win7 has the support you're referring to either. I know some MS software, such as Office, has supported exporting as PDF for some years now (and 2013 even supports importing PDFs for editing) but I don't know of OS-wide support.
Are you off your rocker? They most certainly have thrown arms dealers in prison, both for providing weapons to those who are not allowed to own them (sometimes only known felons, other times anybody who is a civilian; usually somebody who doesn't have a license to own and operate a firearm), and for providing illegal weapons at all (which a working torrent for material under copyright licensing that doesn't permit redistribution would certainly qualify as).
Bullshit. If that were true (i.e. if you could get the.torrent files from Google directly), there would be no need for sites like TPB. What you suggest works only because of TPB and its multitude of clones. It wasn't that long ago that it effectively worked *only* because of TPB. Claiming that because you can find illegal torrent using Google makes it the same as the sites hosting illegal torrents is stupid. The one takes manual human effort and intentional action to create, with the blatantly obvious intention of facilitating copyright infringement (it's in their name...). The other is an automated process that operates on a scope vastly larger than any organization like TPB, or even Google themselves, could ever manage to control.
Yes, Google could (and sometimes has) fail to return any TPB links, but that doesn't mean TPB's entire raison d'etre isn't to enable copyright infringement on a grand scale, and by now they're well enough known that using a search engine to find them is almost a pure waste of time.
Wow... never before been modded to +5, spawned a huge discussion branch, and then (almost a week after the initial post) been downmodded 5 times in the space of a few hours. I suspect Broder isn't the only one with a bias here... but, as with him, I shall refrain from speculating as to why. That's a very bizarre number of accounts getting modpoints and all being mad at me simultaneously, though.
(The accusation itself is obviously bollocks; whoever leveled it my way, enjoy your loss of moderation privileges when you get bithslapped during metamod).
A full x86 emulator written in JS. It happens to load a Linux image into it, but you could easily use a BSD one instead. Pretty cool... but not quite as cool as compiling the OS to run in JS directly.
It won't charge the battery, but it may extend the range. Consider the following scenario:
You have to drive 180 miles. You have two routes you can take. One is perfectly level, and the other is a series of moderate hills. On the level route, you get your car's standard 30 MPG, and use 6 gallons of fuel. On the hilly route, you spend equal time going up and down. Going up, you're fighting gravity and your fuel economy is reduced to 18MPG. Going down, you put the car in neutral and roll, using effectively no fuel (on some cars, using literally none as the gasoline engine will turn off). You use 5 gallons of fuel on the uphill portions (180mi/2 = 90mi, 90mi/18mpg = 5gal) and less than a gallon on the downhills. By taking the hillier route, you've saved fuel.
Note: this doesn't require any violation of the laws of physics. The thing to remember is that car engines are inefficient. On the level route, the fuel is being used to overcome the wind resistance for speed X plus the wheel friction (we'll call this force that slows the car down Y). The engine also consumes fuel to overcome or provide for the drivetrain and timing system friction, waste heat, vibration, pulling in fuel and air on the intake stroke, compressing it on the compression stroke, expelling it on the exhaust stroke, and powering the spark plugs (assuming a fairly basic 4-stroke gasoline engine). With me so far?
Now, assume that speed X is maintained the entire way, on either route (simplification of real driving, but close enough). On the downhill sections, the force of gravity is equal to the wind resistance and rolling friction at speed X (thus gravity force Y counteracts the slowing force Y from wind and friction), so the engine doesn't have to do anything at all (if idling in neutral, it will waste a small amount of fuel to keep itself running at low revs, but not much unless the idle point is set unreasonably high). On the uphill sections, this means that the car must drive against 2Y of force (gravity plus wind/friction). However, this doesn't mean it needs twice as much fuel; the other inefficiencies of the engine will not necessarily double just because the engine's power output doubles, so the total fuel usage will probably not double.
I've tested this on my car, incidentally, and for certain slopes it's true: going up the hill uses less than twice the fuel that level driving does, but coming down the hill I can coast the whole way at the same speed.
Now... for an electric, it's going to be different in the details, but that doesn't mean that the basic concept won't apply anyhow. It *may* be more efficient to speed up and then let the regenerative braking slow you down, than to maintain constant speed. It flies in the face of common sense, but it *may* be true under at least some circumstances..
Are you on crack? You can (trivially easily) run two spreadsheets, or any other desktop app, side by side. All the shortcuts from win7, both keyboard and mouse, still work. You can snap windows to half of a screen (allowing up to four apps on two screens, though you have to use the Win+Arrow keys shortcuts to use the "inner" edges of side-by-side screens), or you can maximize windows on each screen, or some combination thereof.
Yes, The Interface Formerly Known As Metro only runs on one screen at a time, but who the hell cares? TIFKAM apps are lame on the desktop anyhow (not as unusable as some people claim, but worse than the traditional alternatives). Besides, you could have a TIFKAM app on one screen and a couple of desktop apps (even two of them side-by-side, Aero Snap style) on the other, if you wanted.
This actually happens in everybody. The lens of the human eye inverts the image that comes through it. Human newborns haven't yet learned to correct for this, and their ability to follow motions with their eyes is thus impaired. By a month or so of age*, their brains have corrected and see things as we do. This is part of why it's important for babies to have moving objects to watch; it gives them things to learn eye-tracking with*.
* I'm no expert in the development of babies; this is stuff I read about years ago and I may have misremembered timescales and other details.
The tech industry severely distorts earning expectations. The median US household income in 2012 was about 45k. A household bringing in 150k annually is in the 94th percentile.
That's still a lot of people - 6.6 million households earn 150k annually or more, times an average of 3 people per household in that income range, for a total of 20 million people - but that's not exactly a mass-market-targeted product.
Many of the larger malls in my area (Seattle) already have a few charging stations for electric cars. They can't take many at once, but they are often free with the parking (i.e. the lot may require validated parking from the mall, but if you get it, there's no fee for charging your car).
Mind you, these are hardly superchargers; they are intended to top up your car so you can maybe run a few more errands and then get home, not to fill up your Model S or Roadster for a multi-hundred-mile drive. Still, it's nice to be able to go to a movie and come back to a full car at no extra charge.
If you're going to complain about a programming language on the basis of its 3-4 letter syntactic sugar, you can go right ahead, but don't expect to be taken seriously. Yeah, it's a BASIC dialect, so fucking what? Yes, I know the Djikstra quote; it's quite demonstrably wrong and, frankly, stupid. A lot of excellent programmers today and in the past learned on BASIC. Let's see...
Then / End: semantically identical to { }, but a bit less overloaded than those characters are *and* easier to understand when learning the language. Oh, but they require an extra 5 characters of typing when writing a loop! KILL IT WITH FIRE! Seriously, that's a dumb complaint. The typical Java program could probably shorten each class/interface name it uses by 5% rounded down (meaning no affect on anything less than 20 characters long) and save more.
Dim: The type-specifying version (Dim x as Int) is inelegant compared to the way C-style language do variable declaration, but aside from the somewhat archaic choice of keyword, it's clear and explicit. The VBS form (Dim x) is very straightforward, and (again, aside from the actual keyword choice) actually a hell of a lot better than, say, JavaScript (at least if you have Option Explicit on, which should be included in every VB* template and default) where the "var" as completely optional. Also, the array-declaration form is very simple in VBS (Dim a(12)) even if using parentheses for array operations just feels *wrong* to C coders. Oh, and if "Dim" realllllly offends you, leave out Option Explicit and declare all your variables on first use. That's a terrible idea, but it's also one that a great many other languages (especially scripting languages) use...
Sub: I'm sorry, you prefer "void"? Seriously, that's all a Sub[procedure] is; a method that has no return value. "Sub foo ()" is the same as "void foo () {" in C. Or perhaps "function" with no indication of whether the function returns anything or not (JS) is more your speed?
I wouldn't bother replying at all, but for some reason you've been modded up. What the hell? I haven't even touched the language since 2006 (and that was for a very small project at work) but I can still tell that your argument is idiotic. Complain about something legit like its piss-poor attempt at OOP or its highly readable but less expressive for loops or its relative dearth of libraries (not a problem for VB.NET, incidentally, since it can interact with any.NET code), or with something about the language itself, but seriously, the keywords are the stupidest of a great many stupid reasons to hate it.
Mostly irrelevant to the article, but going by both the published drop tests and the unexpected durability tests when a Surface slid off the roof of a guy's car on the highway and was run over by another car, you're likely to do a lot mode damage to your Surface by trying to service it than you are by sliding it off any reasonable-height table onto any likely floor surface. Short version to save you some searching the web: even the one that flew off the car and was driven over didn't need servicing afterward, although there was some cosmetic damage.
Your last sentence is untrue. There are numerous downsides to reparability starting with aesthetics but also covering size and weight (remember how Apple improved battery life in their MBPs with the internal-only battery? A total bitch to service that, but they made use of the space that was needed for removable casing before). Engineers like computers that can be disassembled, but ordinary people like computers that are slick and elegant and fashionable. No modern electronic device has been made more fashionable by making it easier to service.
Have you read his previous car articles? I'm not going to claim "in the pocket of Big Oil" but he shows a clear bias against electric cars - fillin your own reason for this, if you want to.
Slashdot Mirror
If MHL was cheaper, then why wouldn't Apple have just gone with MHL?
...
Seriously? The answer is not only obvious, it's been pointed out a few times in this thread already. If Apple went with MHL, they would save a few cents (amortized development costs, certainly) on each iPhone, but they'd lose several dollars on every customer who just bought a cable off Monoprice or Amazon or wherever.
Of course, it would be good for the customer; they'd have more options, and could stream video for less money than Apple's ludicrous proprietary connector costs, but poor nearly-broke Apple's bottom line would suffer so, so horribly...
Besides, this way they can call it "innovation"! Just because the new way is stupid and overpriced doesn't mean it isn't new!
Most people browse the vast majority of the web via HTTP. Even leaving aside sites that don't even support HTTPS, damn near everybody will visit an HTTP page at some point. Hell, Slashdot auto-redirects from HTTPS back to HTTP. Absolutely no need to MitM SSL connections (which they'd have to get an intermediate trusted CA cert for anyhow).
I'm amazed you know that much, and yet are apparently still so ignorant. The idea was that dynamite (NOT TNT; they aren't the same thing) and the other munitions he invented would be a terrible thing to be remembered for (an obituary accidentally published prior to his death did in fact make a great deal out of his works in the field of armaments (calling him "the merchant of death"). He established the prizes, donating the vast majority of his estate to the purpose, as a way to further the causes of progress and peace.
It's been claimed that he (as many people before and after would think, of similar inventions) actually hoped dynamite would make war *too* costly, so that nobody would dare engage in it. I haven't found proper citation for that, though. So far, the only weapon to come close to that has been atomic bombs on long-range missiles, which have deterred war only between superpowers.
Trivially easy to canonicalize that to YourName@gmail.com, and since that approach is so well-known, any competent spammer (not a self-contradiction, nice though it would be; there's a lot of money to be made) will be able to strip such "custom" addresses to the real address. If you want this approach to actually work, you need to blacklist the root address (yourname@) using filters (I'm assuming Gmail filters cna handle that) and only accept mail that has the identifying tag.
Considering that Windows has had built-in dictation support for many years now as part of their accessibility stuff. It needs to be trained or it sucks horribly, but it's there, and the training doesn't take that long. I wouldn't be surprised if the default MS implementation even on consumer SKUs just used that.
We already have mandatory vehicle inspections in many, if not all, states. They're called emissions tests, and things like the odometer and vehicle ID are already logged or easily could be. It's not difficult to imagine that vehicle license registration requires an odometer tax, and do away with the fuel tax. In fact, it would be easy.
The hard part is dealing with the people who complain that they shouldn't have to pay public road taxes for vehicles which only operate on private property, but of course, they're already paying those taxes when they buy fuel for the vehicles. Big whoop.
One is also generally achieved by about the time of your legal maturity, the other requires spending years more learning things which are completely unnecessary for most jobs (sometimes called "unskilled", but really just meaning no need for extensive math/science/engineering/language skills. You don't need a ME to change oil and spark plugs in a car, you don't need an English degree to handle insurance claims, you don't need a... I'm not even sure what it would be... degree to schlep stuff around between offices.
The biggest problem, as I see it, is that it's a vicious cycle of it being *expected* that you get a real job IFF you have a degree, which leads to high schools being degraded to the point of being nothing more than college prep (when they really should be suitable prep for many careers with only minimal job training), which leads to anybody reasonably qualified needing to go to college to have a decent chance, which reinforces the impression that you need a degree to get a decent job...
The Cell processor is what, roughly 6 PPC cores at 3.something GHz?
The PS4 has 8 x64 cores at (according to previous comments in this thread) 2GHz.
There's a pretty low real-world cap on the performance of dynamic recompilation. Don't forget that since we're dealing with consoles (very specific hardware target, you also face the complete loss of optimization from compilers that built for the peculiarities of the Cell (cache sizes, which instructions are fastest, etc.) when that CPU bust be emulated or the code gets dynreced. Even assuming that the two extra x64 cores will be able to do the full dynrec for the other six that *must* run in parallel, there's the question of how fast each core is. What's the real-world operations per second of a Cell core from a PS3 compared to a Jaguar core from a PS4?
Oh, and then there's the GPU emulation. This may be even trickier, because unlike on a PC (where you have no idea what GPU your code will run on, so must keep it generic), console graphics code is very, very tightly optimized for each console it runs on. It must be; there's no other way to get even vaguely decent-looking graphics (by today's standard) out of hardware that outdated. The problem for the next generation, then, is providing an absolutely perfect re-creation of all the quirks of that old GPU. If the GPU architecture has changed (as it has), that means basically a whole additional emulation or dynrec layer, with once again the associated performance hits and potential difficulty maintaining the same speed per emulated processor on new architectures whose clock speeds really haven't improved much.
Emulating the original Xbox on the 360 was relatively easy, despite the massive differences in infrastructure; the CPU was absurdly more powerful, as was the GPU. The same is true of the PS2 and PS3, although it took them a while to get around to writing the emulation layer. However, progress in computing power since then has gone wide more than deep. Clock speeds aren't that much better than they used to be, and may in fact be worse. Instructions-per-clock-per-core may be improving, but nowhere near fast enough to make up for hit imposed by switching architectures. Instead, the number of cores has increased massively. The Cell was a multi-processing beast when it came out; today, it's not that much better than the chip in a decent tablet on that metric. More parallelism doesn't necessarily make up for the performance hit of emulation, though, and the new CPU isn't vastly more parallel than the Cell was.
I doubt many (if any) HD2 purchases these days end up running the stock OS much. The HD2 is ridiculously easy to port to, from everything I've seen. It's had (unofficial) ports of WP7, Android, Maemo/Meego, Ubuntu, and Windows RT, and probably other OSes. Its hardware is dated, but not uselessly so, and it's a wonderfully hackable device and can be dual (or more, if you don't mind the storage hit) booted; a rare capability in a phone.
There were search engines before Google. Lycos was even pretty decent by pre-Google standards; other people had other preferences.
No, I don't know how well-known Python (the language) was back then, and I don't know whether ordinary people would have thought "I should do a web search on that" in those days, but I'm pretty sure the information would have been available. Hell, when was python.org registered? The copyright notice at the bottom begins with 1990.
Much though I despise the term, "cloud computing" is not meaningless or even sensationalistic, and there are some very big companies who have built their success on cloud computing. It is, perhaps, over-hyped and watered down, and it's undeniably a buzzword. It's also pretty misunderstood, as you yourself are evidence of (the basic concept is simply "hosted computing services"). However, to say you think a company lacks credibility because they speak of cloud computing is, frankly, idiotic.
The first part of your post makes sense, though. I could understand a company that wasn't deeply involved with computer software might not be aware of a programming language, even a common one, although you'd think they might make at least a trivial web search. However, for a software-focused company to be unaware of Python's importance is, frankly, unbelievable. This guy is just trying to pull whatever damage control he can.
XP is a 12-year-old OS. Why are you comparing it to modern OSes?
That said, I don't believe Win7 has the support you're referring to either. I know some MS software, such as Office, has supported exporting as PDF for some years now (and 2013 even supports importing PDFs for editing) but I don't know of OS-wide support.
Are you off your rocker? They most certainly have thrown arms dealers in prison, both for providing weapons to those who are not allowed to own them (sometimes only known felons, other times anybody who is a civilian; usually somebody who doesn't have a license to own and operate a firearm), and for providing illegal weapons at all (which a working torrent for material under copyright licensing that doesn't permit redistribution would certainly qualify as).
Bullshit. If that were true (i.e. if you could get the .torrent files from Google directly), there would be no need for sites like TPB. What you suggest works only because of TPB and its multitude of clones. It wasn't that long ago that it effectively worked *only* because of TPB. Claiming that because you can find illegal torrent using Google makes it the same as the sites hosting illegal torrents is stupid. The one takes manual human effort and intentional action to create, with the blatantly obvious intention of facilitating copyright infringement (it's in their name...). The other is an automated process that operates on a scope vastly larger than any organization like TPB, or even Google themselves, could ever manage to control.
Yes, Google could (and sometimes has) fail to return any TPB links, but that doesn't mean TPB's entire raison d'etre isn't to enable copyright infringement on a grand scale, and by now they're well enough known that using a search engine to find them is almost a pure waste of time.
Wow... never before been modded to +5, spawned a huge discussion branch, and then (almost a week after the initial post) been downmodded 5 times in the space of a few hours. I suspect Broder isn't the only one with a bias here... but, as with him, I shall refrain from speculating as to why. That's a very bizarre number of accounts getting modpoints and all being mad at me simultaneously, though.
(The accusation itself is obviously bollocks; whoever leveled it my way, enjoy your loss of moderation privileges when you get bithslapped during metamod).
http://bellard.org/jslinux
A full x86 emulator written in JS. It happens to load a Linux image into it, but you could easily use a BSD one instead. Pretty cool... but not quite as cool as compiling the OS to run in JS directly.
It won't charge the battery, but it may extend the range. Consider the following scenario:
You have to drive 180 miles. You have two routes you can take. One is perfectly level, and the other is a series of moderate hills.
On the level route, you get your car's standard 30 MPG, and use 6 gallons of fuel.
On the hilly route, you spend equal time going up and down. Going up, you're fighting gravity and your fuel economy is reduced to 18MPG. Going down, you put the car in neutral and roll, using effectively no fuel (on some cars, using literally none as the gasoline engine will turn off). You use 5 gallons of fuel on the uphill portions (180mi/2 = 90mi, 90mi/18mpg = 5gal) and less than a gallon on the downhills. By taking the hillier route, you've saved fuel.
Note: this doesn't require any violation of the laws of physics. The thing to remember is that car engines are inefficient. On the level route, the fuel is being used to overcome the wind resistance for speed X plus the wheel friction (we'll call this force that slows the car down Y). The engine also consumes fuel to overcome or provide for the drivetrain and timing system friction, waste heat, vibration, pulling in fuel and air on the intake stroke, compressing it on the compression stroke, expelling it on the exhaust stroke, and powering the spark plugs (assuming a fairly basic 4-stroke gasoline engine). With me so far?
Now, assume that speed X is maintained the entire way, on either route (simplification of real driving, but close enough). On the downhill sections, the force of gravity is equal to the wind resistance and rolling friction at speed X (thus gravity force Y counteracts the slowing force Y from wind and friction), so the engine doesn't have to do anything at all (if idling in neutral, it will waste a small amount of fuel to keep itself running at low revs, but not much unless the idle point is set unreasonably high). On the uphill sections, this means that the car must drive against 2Y of force (gravity plus wind/friction). However, this doesn't mean it needs twice as much fuel; the other inefficiencies of the engine will not necessarily double just because the engine's power output doubles, so the total fuel usage will probably not double.
I've tested this on my car, incidentally, and for certain slopes it's true: going up the hill uses less than twice the fuel that level driving does, but coming down the hill I can coast the whole way at the same speed.
Now... for an electric, it's going to be different in the details, but that doesn't mean that the basic concept won't apply anyhow. It *may* be more efficient to speed up and then let the regenerative braking slow you down, than to maintain constant speed. It flies in the face of common sense, but it *may* be true under at least some circumstances..
Are you on crack? You can (trivially easily) run two spreadsheets, or any other desktop app, side by side. All the shortcuts from win7, both keyboard and mouse, still work. You can snap windows to half of a screen (allowing up to four apps on two screens, though you have to use the Win+Arrow keys shortcuts to use the "inner" edges of side-by-side screens), or you can maximize windows on each screen, or some combination thereof.
Yes, The Interface Formerly Known As Metro only runs on one screen at a time, but who the hell cares? TIFKAM apps are lame on the desktop anyhow (not as unusable as some people claim, but worse than the traditional alternatives). Besides, you could have a TIFKAM app on one screen and a couple of desktop apps (even two of them side-by-side, Aero Snap style) on the other, if you wanted.
This actually happens in everybody. The lens of the human eye inverts the image that comes through it. Human newborns haven't yet learned to correct for this, and their ability to follow motions with their eyes is thus impaired. By a month or so of age*, their brains have corrected and see things as we do. This is part of why it's important for babies to have moving objects to watch; it gives them things to learn eye-tracking with*.
* I'm no expert in the development of babies; this is stuff I read about years ago and I may have misremembered timescales and other details.
The tech industry severely distorts earning expectations. The median US household income in 2012 was about 45k. A household bringing in 150k annually is in the 94th percentile.
That's still a lot of people - 6.6 million households earn 150k annually or more, times an average of 3 people per household in that income range, for a total of 20 million people - but that's not exactly a mass-market-targeted product.
Many of the larger malls in my area (Seattle) already have a few charging stations for electric cars. They can't take many at once, but they are often free with the parking (i.e. the lot may require validated parking from the mall, but if you get it, there's no fee for charging your car).
Mind you, these are hardly superchargers; they are intended to top up your car so you can maybe run a few more errands and then get home, not to fill up your Model S or Roadster for a multi-hundred-mile drive. Still, it's nice to be able to go to a movie and come back to a full car at no extra charge.
If you're going to complain about a programming language on the basis of its 3-4 letter syntactic sugar, you can go right ahead, but don't expect to be taken seriously. Yeah, it's a BASIC dialect, so fucking what? Yes, I know the Djikstra quote; it's quite demonstrably wrong and, frankly, stupid. A lot of excellent programmers today and in the past learned on BASIC. Let's see...
Then / End: semantically identical to { }, but a bit less overloaded than those characters are *and* easier to understand when learning the language. Oh, but they require an extra 5 characters of typing when writing a loop! KILL IT WITH FIRE! Seriously, that's a dumb complaint. The typical Java program could probably shorten each class/interface name it uses by 5% rounded down (meaning no affect on anything less than 20 characters long) and save more.
Dim: The type-specifying version (Dim x as Int) is inelegant compared to the way C-style language do variable declaration, but aside from the somewhat archaic choice of keyword, it's clear and explicit. The VBS form (Dim x) is very straightforward, and (again, aside from the actual keyword choice) actually a hell of a lot better than, say, JavaScript (at least if you have Option Explicit on, which should be included in every VB* template and default) where the "var" as completely optional. Also, the array-declaration form is very simple in VBS (Dim a(12)) even if using parentheses for array operations just feels *wrong* to C coders. Oh, and if "Dim" realllllly offends you, leave out Option Explicit and declare all your variables on first use. That's a terrible idea, but it's also one that a great many other languages (especially scripting languages) use...
Sub: I'm sorry, you prefer "void"? Seriously, that's all a Sub[procedure] is; a method that has no return value. "Sub foo ()" is the same as "void foo () {" in C. Or perhaps "function" with no indication of whether the function returns anything or not (JS) is more your speed?
I wouldn't bother replying at all, but for some reason you've been modded up. What the hell? I haven't even touched the language since 2006 (and that was for a very small project at work) but I can still tell that your argument is idiotic. Complain about something legit like its piss-poor attempt at OOP or its highly readable but less expressive for loops or its relative dearth of libraries (not a problem for VB.NET, incidentally, since it can interact with any .NET code), or with something about the language itself, but seriously, the keywords are the stupidest of a great many stupid reasons to hate it.
Mostly irrelevant to the article, but going by both the published drop tests and the unexpected durability tests when a Surface slid off the roof of a guy's car on the highway and was run over by another car, you're likely to do a lot mode damage to your Surface by trying to service it than you are by sliding it off any reasonable-height table onto any likely floor surface. Short version to save you some searching the web: even the one that flew off the car and was driven over didn't need servicing afterward, although there was some cosmetic damage.
Your last sentence is untrue. There are numerous downsides to reparability starting with aesthetics but also covering size and weight (remember how Apple improved battery life in their MBPs with the internal-only battery? A total bitch to service that, but they made use of the space that was needed for removable casing before). Engineers like computers that can be disassembled, but ordinary people like computers that are slick and elegant and fashionable. No modern electronic device has been made more fashionable by making it easier to service.
Have you read his previous car articles? I'm not going to claim "in the pocket of Big Oil" but he shows a clear bias against electric cars - fillin your own reason for this, if you want to.