Don't be too quick to assume the power drop at 400mi was out of Broder's control. The log is graphed vs. distance, and the car was stopped at that time. Every single event that occurred over that night, no matter how long they took, are compressed onto that single X coordinate on the graph.
Now, look at the cabin temperature graph. Notice anything interesting at 400mi? Broder turned the car's heater way up for at least some part of that night. Now, that may have been a momentary thing in the morning or something, followed immediately by turning it down again... but it's suspicious how well the battery drain over the course of that night matches the toll that the heater would have exerted on the batteries (eyeballing from changes to the discharge rate based on temperature setpoint) over a few hours. The equivalent of leaving a gasoline car's engine idling while parked; it's not a huge loss compared to spending the same amount of time driving, but it *is* a drain and it will impact the car's efficiency record, causing a look-ahead function based on average performance over the recent past to estimate a drastically reduced range. Indeed, around 405-410mi, after Broder turned the heat way down again, you can see that the estimated range actually *increased* slightly; to me this suggests that the heater operation had been severely impacting the car's predicted range.
Of course, It is odd that Elon Musk didn't call out that particular even in the logs. Either the actual logs (charge and range estimate over time) support Broder and Musk therefore didn't want to publish them, or he for some reason elected not to address one of the most damning parts of the review: that the car loses charge substantially when parked overnight. Indeed, if the battery was in fact intentionally discharged by using the heater and/or other power sinks over the course of the night, that completely turns Broder's review on its head; rather than the power dropping inexplicably (which could reasonably be blamed on the manufacturer), he intentionally drained the battery in order to have something to fraudulently blame the manufacturer for.
This is all hypothetical until we have more data, but I'm very interested in how long that night the heater was at 75, and what the pattern of charge over time looks like over the course of that night next to the plot of heater (and possibly other systems, like lights) over time.
According to Broder's own account, that discharge happened overnight while the car was parked. Because the charts are all plotted over distance, we don't know how long the heater was set at 75. The other times the heat was up, the car was in motion and we can look at the speed the car was traveling at that milepost and figure out how long that state was maintained. However, since the car stopped at 400mi, we don't know (from these charts - I'm sure that Tesla has other data) how long the heater was up, or over what period the charge and predicted range declined, or anything else of that nature.
However, I will say this: looking at the impact of the heater on the car's battery elsewhere, leaving the heater on for 8 or so hours while the car was stopped would probably have produced about that much drop in battery, and could easily have led the range predicter to assume that power was wasting at a precipitous rate vs distance traveled, and therefore the remaining range was very short. Looking at the predicted range/distance chart, you can see that while elsewhere the slope was -1 or a bit worse when the heater or speed were up, on that leg (to the Norwich charger), it was almost flat and even rose slightly at around 410mi, where the temp was turned way down. Smells like an attempt to manipulate the predicted range, to me...
In any case, the text of the blog post is full of references to how long was spent at each charger station and so forth, so they obviously have logs over time. We need to see those in order to have any chance of understanding what was going on at that 400mi stopover...
Well, if you want to burn off some of an electric car's battery over the course of a few hours (such as a night stop without a charger) to make a good story, what better way to do it that to turn on the heater and leave it that way? Lights too, perhaps, but those draw way less than the heater; a typical car with a few hundred Watt-hours of battery can't leave its lights on overnight, but an electric easily could. As for the difference in the battery charge percentage loss and range percentage loss, some of that is most likely the non-linear curve, but it's also probably due to a look-ahead function. If the battery has been draining heavily from a user action (such as turning on the heater up to 75 and leaving it that way) while the car is not in motion, that could affect the predicted remaining range heavily depending on how the car estimates future behavior on the driver's part.
We really need logs over time to figure this out. How long was the heater set so high? How does that time correlate with the drop in battery charge and predicted remaining range? Were any other power-consuming features running over that night, or did the charge just mysteriously decrease?
What we really need to address these discrepancies are the same logs plotted over time. For example, at the 400mi mark, where the power and predicted mileage drops, the car's heater also spikes upwards and then down again. Broder claims he left the car overnight and came back to find it with a sadly reduced battery... but going by the logs, an equally valid (and possibly more plausible, when you remember how an electric car's heater works) possibility is that he left the car for a while with the cabin heater intentionally left running to drain the battery, then came back after burning off a chunk fo the charge. Now, if only we could see the time of day when he stopped at 400mi, and for how long, and what the battery charge, predicted range, and cabin settings were during that entire time... but we can't, because the car wasn't moving, so all events over that time period take place at the same point on the X axis.
I'm pretty damn sure that Tesla *has* the logs over time as well; it would sure help if they released them...
Oddly enough, you've actually got it backwards. Two things that computers can do really well is collect data, and compare that data to existing data and solve constraint problems to map inputs to outputs.
There are, as I noted, sites where users can input their own symptoms... but yes, those things are huge problems for hypochondriacs. They also rely on fairly generic background info, unable to take factor in specific conditions that maybe only three people in a million have... yet to a personalized expert system, that's just an extra data point that it can compare against established info about that condition.
The real kicker is the data collection, though. I know fuck-all about the expected levels of various blood chemistry. I have only basic knowledge of what blood pressure means and what ranges it should be in. I'm only a little better at temperature. I can tell you things like "my back hurts when I get up in the morning" or "I have a headache" but I would have a hard time pointing out the exact vertbra or describing . A personal unit, equipped with a camera and mic, plus tools to measure common basics like pulse, respiration, blood pressure, weight, and some very basic blood work... with some good image and audio recognition software to pick up things like changes to the eyes or skin, or breathing differently... we have software capable of detecting many symptoms at least as well as a human doctor, and a computer program would able to correlate symptoms, patient medical history, family background, and current conditions (say, if you're on some supposedly unrelated drug already) with known diseases and their treatments better than almost any specialist.
As for "duplicate a doctor's thinking", strictly speaking that is of course true - we don't have true AI - but why would you want to? Doctors have unreliable memory; they may forget facts, misremember or become confused, or overlook data points. It's much harder to keep a doctor's knowledge on the state of the art up to date than it is to update a computer's database. Doctors are subject to bias in favor of, or against, certain diagnoses or treatments. Expert systems suffer none of those flaws. Of course, they're only as good as the people who program them, but it's possible for many, many doctors to provide inputs for an expert system. If a human doctor tried to consult all the sources and historical records that an expert system can call upon for each diagnoses, he or she wouldn't be able to see enough patients to stay in business.
Besides, expert systems really do exist. You say "not in my lifetime" and in a funny way, you're right; they've existed, and have been able to out-perform typical human doctors on diagnoses, since before I was born. Doctors are still required, of course - the computer can't handle something that hasn't been seen before very well, and it is good to have a second opinion available. Software isn't infallible either; it will need to be modified and updated, which will require input from doctors. Finally, administering treatment may still be beyond the typical system for anything more complex than "take these pills at these times of day"... although there's no reason you couldn't have "auto-docs" capable of things like open heart surgery or organ transplants.
Yes and no. They can, for sure, be put in neutral. My car (a Subaru) may even have a physical linkage from the shifter to its CVT.
However... that's not true of all cars. For example, some Prius models are fairly obviously a purely electronic linkage. That will rely on the computer interpreting the driver's commands correctly... soemthing which obviously was not going on here.
Hmm... did he try hitting the accelerator to see if it would break the car out of its insanity?
Yep, even cars with a CVT (such as the Prius, which also uses push-button start on many models) have a neutral "gear".
The problem is that the shifter control is electronic, not physical. I mean, there's a physical thing that you push around, but the actual control to the transmission is an electronic signal and must be interpreted by the car's computer. Given that the computer apparently was under the impression that "brake" meant "speed up", this message may not have worked.
Of course, that's weird in and of itself. On every production car I've ever seen, no matter how much electronics it has, the brake pedal *does* have a physical linkage to the calipers. You can use it when the car is completely turned off, in fact, though you have to push pretty hard once the power braking assist goes away. You won't get traction control and such in that state either, but it doesn't matter - the car will stop. The brakes should also be more powerful than the engine (especially on a car that tops out at 125MPH) so you should be able to stop even with the enging revving and in gear (this might cause the engine to stall at that speed, but that's perfectly OK in this situation even if the transmission might not like it).
Agreed; I quite like mine. No dead pixels, though pixels will "sparkle" (not sure how else to describe it) if left on for a few days; power cycling or turning it off at night avoids this problem. Otherwise, it's great; beautiful and bright, good color and contrast, good response time. 2560x1440 is a good aspect ratio and plenty of vertical pixels. Cost was $290 USD on eBay, including DHL express shipping (48 hours from warehouse in Seoul to my place in Seattle) and an included adapter for US outlets (the power supply handles 110V/60Hz just fine).
I looked at comparable US options, and at 25x16 options, and they were all vastly more expensive. I'm happy with the purchase.
I actually have one of those 27" 2560x1440 ("WQHD") IPS monitors from Korea. $290 USD, included DHL express shipping (about 48 hours after it left the warehouse in Seoul to reach me in Seattle) via eBay. It's wonderful. Bright and clear, glossy screen but bright enough that reflections and background light are no problem, good stand, and simple but functional on-screen display. The DPI is nothing amazing, but it's comparable to my old 18" 1920x1080 monitor, and that's fine by me. I don't use the 5W speakers it has built in, so I can't say how those are.
The monitor does have a minor defect where if left turned on too long (several days straight) it will start getting "sparkles" on a black screen, but this is easily fixed by power-cycling the monitor or just turning it off every night (it starts very quickly, so that's no problem). It can also get pretty warm (especially at max brightness) and has a large-ish power brick (with a plug designed for Korean outlets, though they included a USA adapter for me at no extra charge) rated for 120W output.
Contrast is good but not incredible, but the lighting is very even. The in-plane switching works great; response time is excellent and the viewing angle is superb (the ~1/4" bezel gets in the way before the screen noticeably changes color). Color and saturation look good to my eye (untrained, but an amatuer photographer); it is something I look at and check calibration on.
I plan to buy a second one... just as soon as I figure out where I'm going to put it. I may just get a 1440x900 (or similar) instead and put it in portrait mode next to the big one. Otherwise, I'll probably need wall mounts; I'm running out of desk space.
One note of caution: It requires a dual-DVI input. That means no driving two monitors off one DVI connection, and many HDMI adapters, etc. won't work.
Eh... I'd say it's awesome that, out of the box, Win8 supports taskbar (and wallpaper, if you care) spanning across monitors. That used to require a third party utility. I love that it's finally supported in-box.
And what about those who aren't "regular viewers" but instead hear from somebody that "there's a review of this really cool new car on TV?"
Unless they explicitly state (for the benefit of new watchers, if nothing else) that the "review" is scripted, it's lies and bullshit. Fiction presented as fiction is entertainment. Fiction presented as fact - even when *most* people know it's not - is grounds for libel.
Tesla "won" the lawsuit but was awarded no damages as it was determined that no financial harm had been done to them. I'm skeptical, but I wasn't following the story so maybe that's perfectly true. In any case, I have no respect for Top Gear after that incident. "But it's just for funsies" is not a legitimate excuse when you're lying to people about a product (an expensive one very important to peoples' day-to-day lives, in this case) failing to live up to its manufacturer's stated capabilities, especially by supposedly conducting a test and then ignoring the results and following the script.
You know, not that I don't agree that the whole review smells very fishy, but it would help Tesla's (and Musk's) case if logs were released showing these same numbers over *time*. For example, as was pointed out, at 400mi, the internal temperature abruptly went up a bunch, then down again, while the car lost a ton of predicted range without moving.
What happened? Did he turn on the heater and sit in a parking lot for a few hours? Did he leave the car overnight and then come back in the morning to find the charge depleted (the way he claimed) but turned on the heater briefly to warm the cabin then started out anyhow? Did he leave the heater on all night, depleting 2/3 of the car's remaining charge while it was sitting still and unoccupied?
Now, this could be easily determined... if we had logs over time. In fact, a lot of things would benefit from this analysis.
Came here to mention this. Medical "expert systems" (a class of AI) have existed, and been better than humans at some things, for literally decades. However, you almost never hear about them anymore. With the vast advances in processing power, miniaturization, and power usage, we can today build pocketable devices that tremendously exceed the capabilities of something like Mycin. With the improvements in sensor technology, we can feed those systems more and better data than ever before. With widely available wireless Internet connectivity plus incredible storage densities, we can provide these systems with all the info they could ever need. With the advances in the science of AIs over the last four decades, we can make these systems "smarter" than was possible before. Finally, with the plummeting cost of such electronics, we can make such systems affordable to middle-class individuals, rather than exclusively to large and well-funded organizations.
Yet, they almost non-existent. We have a few smartphone apps that scratch the surface of what's possible. We have dedicated machines like continuous glucose monitors for diabetics, but they have very little in the way of smarts and no versatility. We have concepts and pseudo-prototypes of "medical tricorders" and such sci-fi devices, but they aren't generally set up to make recommendations. We have online web applications where users can input symptoms and be told what they might be, but those systems have no personal background or history of the user's health, and rely on the user measuring and providing data themselves.
If it weren't for the fact that TPB expressly permits copying their content, you mean? Less irony, more "I have no idea what I'm talking about but I'll point any laugh anyhow!"
This site runs a torrent tracker for files known to be under a copyright license that doesn't permit free redistribution? Wow, I had no idea...
Or did you mean "runs a site with some CSS and a search box that does something if you enter search terms for potentially copyrighted downloadable files"? Because I don't think that's the "function" of TPB that anybody else, including the courts, were referring to.
Swap "copyright infringement" for "piracy" and so forth below if you're going to get upset over the use of terminology that has been around for over half a century.
Funny, I hadn't heard that. The version I've heard is that TPB exists to facilitate piracy (which, given their name and policies, is a bit hard to refute) and therefore must be shut down. I've never heard any blanket statements like "everyone who retrieves from TPB is necessarily a pirate" except from people like you putting words in the mouths of others. It's quite possible to have the first (and take it too far) without the second. Scratch the hyperbole; it is insulting to the intelligent.
Your second line is accurate, if scarcely worth more than an eye-roll in terms of response. TPB expressly permits copying of their stuff. There's no piracy involved there. Similarly, they don't benefit in any way from the re-use of their CSS, so it's hardly a case of aiding a site which seeks to facilitate a crime.
Your third line is predicated on your first, and a mildly generous interpretation of your first at that. Typically if somebody speaks of "retrieving" from TPB, they'd be talking about retrieving information needed to download torrents, not webpage markup information.
Yes. If only one-party consent was required, they wouldn't even tell you; their people already know and don't pick up until after the recorded message anyhow.
There are multiple sites were people are hosting.APPX files (Metro app bundles) that can be downloaded and sideloaded onto a device. Sideloading is easy, and officially documented how to do it; it costs nothing aside from a very minor annoyance (enter a Powershell command, basically).
There's also a growing amount of desktop software for RT. It requires "jailbreaking" the device in order to run it, but once you do that it's just like any other Windows machine... except it runs ARM, so native code must be recompiled first (.NET code can run unmodified, as can Java code if you're willing to put up with the speed hit of IKVM). There's a project to provide an x86 emulation layer as well, bringing full (if again somewhat slow) support for legacy apps.
The Netscape Plugin API (which is what Flash is implemented using on Firefox, Opera, Safari, and at least some versions of Chrome) doesn't really support this. I don't know of any legit reason why not, but it doesn't.
IE, which uses an ActiveX control for Flash, actually has no problem installing or updating plugins without restarting the browser. Refresh the page (which, if it's installed via the browser itself, will happen automatically) and you're good to go.
Relatedly, I believe the reason that Word is being used as the exploit vector on Windows is because it doesn't have the sandboxing of IE/Firefox/Chrome. While you could get a lot more people to run the Windows attack code if you posted it on websites, it doesn't do any good when every popular browser newer than IE6 is locked down to not be able to launch arbitrary programs or write to most of the filesystem or registry.
Good for you... but it turns out that a fair number of the most popular websites on the entire WWW play videos, typically in Flash, and they are watched by people of all ages (perhaps more among the younger set, but certainly not exclusively). Quite a few sites (perhaps not individually the most popular, but a massively common *class* of site) also serve lots of Flash video, although for legal reasons they are only supposed to be watched by adults. People also like to watch videos of events they couldn't make it to and listen to streaming music, both of which are common uses of Flash. You can do web-based video chat or even videoconferencing using Flash (Google Talk can do this, for example).
I don't like Flash, and I certainly don't trust it; I keep it tightly curtailed where it's installed at all. However, it's definitely useful in some cases. HTML5 is catching up, but not fast.
You think in 1985 a device that would be literally incapable of running any software that doesn't come pre-loaded on it (there being Apple App Store at the time) would be considered a PC? You're delusional (and/or younger than I am). In 1985, every single PC sold came with, at a bare minimum, some form of BASIC or a similar language, an editor for writing programs in that language, and at least one of a compiler or an interpreter for running them. Many came with little else. Hell, to this day, this holds true; Windows has VBScript and JScript and Windows Script Host, CMD and BAT in Command Processor, PS1 and Powershell, and you can write them in Notepad if you want to. OS X has shell scripts and probably a few other things; I don't mess with Macs that much. Linux (any PC-oriented distro, or any distro you can buy loaded on PC hardware) comes with shell scripts, perl, python, and $DEITY knows what else, including plenty of editors for them, and quite likely also has gcc (out of the box). A bare-bones PC shipped with FreeDOS can still write and run QBASIC programs the same as a MS-DOS machine in the 80s could.
Note that I'm not even counting web browser scripting environments above. You can run a game on one, but you can't write out a file. You can create a spreadsheet app that runs on the browser, but good luck programmatically opening the spreadsheet file somebody handed you on a floppy (without doing a postback to the server). The iPad can't even do those things, though, because there's no way to write a web page (with or without script) and have the browser open it built into the iPad.
In 1985, an iPad is a camera you can't get your pictures out of, an email and text messaging appliance that can't connect to anything (no way to install a network connection in one, and what the hell is this "why-fie" thing?), a useless "web browser" tool, a mapping tool with no map data and no way to add any (or any way to tell you your latitude and longitude, as GPS was not yet deployed), a music and video player with almost no music or video (and no way to add more), a lame flashlight, an electronic calendar capable of reminders, a highly advanced alarm clock with terrible battery life, and a stupendously fast electronic calculator that is capable of only very basic operations. The last three (and maybe the first, if you didn't mind never being able to send the photos to anybody) would probably be its best features. Meanwhile, PCs were being used for writing papers, creating and editing spreadsheets, sending and receiving email, accessing Usenet, building models in everything from finance to weather, and playing games. People were also using them to write software for themselves, for other PCs, and for mainframes and supercomputers.
No, in 1985, the iPad would very much not be considered a PC (for much the same reasons that it doesn't qualify today). Steve Wozniak would probably have laughed in your face if you'd called it one.
You can run an ssh server on your iPad just fine without jailbreaking? Fascinating. After you ssh into it, do you prefer to start vim or emacs for your coding? Do you compile using gcc or clang? How good are its manpages? What shell do you use? Do you prefer to elevate privileges using su or sudo? If you ssh in from an unrecognized terminal, how difficult is it to install new terminfo? How many users can ssh into your iPad at once? If you wanted to install Python for iOS on your iPad, what command would you enter over ssh to do so? After doing so, could you demonstrate for me a Python script that checks the iPad's entire storage for any software which hasn't been started in over six months and offers to uninstall it?
Or, did you completely miss the point about 'uses applications not truncated "apps"'? While I dislike the terminology, that right there really is one of the defining differences between PCs and whatever category you want to put iPads in. Software for iOS is crippled. Yes, you can run an ssh client on your iPad and consume the services of a server (or any real PC, even a Windows one). You can even use the iPad to produce stuff on that server or real PC. But that doesn't make the iPad into a PC. Consumption-based doesn't mean it can't be used to produce anything, merely that the intended uses of the OS are not focused on productivity. This is the exact opposite of PCs, which at their inception could be used for very little except productivity.
The other difference, of course, is the degree of control. A Windows machine does not come with an ssh server, but if I want to install one I can. Ubuntu doesn't allow logging in as root by default, but if I want to change this, I can. If I buy a Windows PC but then decide I'd rather have it be a Linux PC, that's as simple as re-installing the OS. If I buy a machine with Ubuntu and decide that its graphical user interface is crap, I can remove Unity and install FVWM and have it start that after login instead. None of these are possible on an iPad. Sure, you can install software... if it's Apple-approved, content to run in a sandbox with restrictions that you can't control, and doesn't require mucking with the system configuration. Even the relatively incompatible Macs of the 90s didn't actively prevent you from installing your own OS on them... marketing aside, they were PCs. iPads simply are not, though.
Nobody except you and a few other fools are debating the subject of whether they are computers. Stop being an idiot (or is it your hope that idiotic arguments will distract from the fact that you haven't addressed the question?) Paper tape and punch-card computers weren't PCs either. PCs have, for as long as the term has existed, supported writing software and executing it on the same PC. In fact, the first PCs could do precious little else (aside from load a program that somebody else had written on their PC and then used their PC to write it to a diskette).
A secretary's PC *could* run code that the secretary wrote. If a secretary chose to write some code on the PC, it would be possible to do so (and run it!) without approval from anybody else. Now, if the computer in question was actually a locked down corporate workstation, then the secretary wouldn't be able to run their own software on it... but it arguably wouldn't be a PC either, because it's not personal to the secretary.
You know what the differences are between an iPad and a corporate workstation, in terms of whether they qualify as a PC? If you reformat a corporate workstation, you can install a PC operating system on it (i.e. one that you personally control). Try that with an iPad...
Are you an idiot or a troll? The GP wasn't addressing you, and you didn't answer the question for the GPP either. Nobody in this thread was discussing what makes anything a computer. User gstoddart appears to be claiming that anything which can be owned by an individual ("personal") and can simulate a Turing machine ("computer") is a PC. The exact question was
Do tablets really count as a "PC"?
Perhaps you're the one missing the distinction between "computer" and "Personal Computer"? Or did you perhaps mean to reply to gstoddart directly, in which case the otherwise-unspecified "you" in your post would be assumed to refer to him/her?
In any case, my definition of the "Personal" in PC is much more specific than gstoddart's. Basically, if somebody doesn't *control* the computer themselves, it's not a Personal Computer. An iPad, without jailbreaking it, can only run iOS, can only run Apple-approved iOS apps, and only gives the user as much control over the device's capabilities as Apple wants them to have. The user has choices, but Apple controls what choices the user has. The nominal owner doesn't have control.
Slashdot Mirror
Don't be too quick to assume the power drop at 400mi was out of Broder's control. The log is graphed vs. distance, and the car was stopped at that time. Every single event that occurred over that night, no matter how long they took, are compressed onto that single X coordinate on the graph.
Now, look at the cabin temperature graph. Notice anything interesting at 400mi? Broder turned the car's heater way up for at least some part of that night. Now, that may have been a momentary thing in the morning or something, followed immediately by turning it down again... but it's suspicious how well the battery drain over the course of that night matches the toll that the heater would have exerted on the batteries (eyeballing from changes to the discharge rate based on temperature setpoint) over a few hours. The equivalent of leaving a gasoline car's engine idling while parked; it's not a huge loss compared to spending the same amount of time driving, but it *is* a drain and it will impact the car's efficiency record, causing a look-ahead function based on average performance over the recent past to estimate a drastically reduced range. Indeed, around 405-410mi, after Broder turned the heat way down again, you can see that the estimated range actually *increased* slightly; to me this suggests that the heater operation had been severely impacting the car's predicted range.
Of course, It is odd that Elon Musk didn't call out that particular even in the logs. Either the actual logs (charge and range estimate over time) support Broder and Musk therefore didn't want to publish them, or he for some reason elected not to address one of the most damning parts of the review: that the car loses charge substantially when parked overnight. Indeed, if the battery was in fact intentionally discharged by using the heater and/or other power sinks over the course of the night, that completely turns Broder's review on its head; rather than the power dropping inexplicably (which could reasonably be blamed on the manufacturer), he intentionally drained the battery in order to have something to fraudulently blame the manufacturer for.
This is all hypothetical until we have more data, but I'm very interested in how long that night the heater was at 75, and what the pattern of charge over time looks like over the course of that night next to the plot of heater (and possibly other systems, like lights) over time.
According to Broder's own account, that discharge happened overnight while the car was parked. Because the charts are all plotted over distance, we don't know how long the heater was set at 75. The other times the heat was up, the car was in motion and we can look at the speed the car was traveling at that milepost and figure out how long that state was maintained. However, since the car stopped at 400mi, we don't know (from these charts - I'm sure that Tesla has other data) how long the heater was up, or over what period the charge and predicted range declined, or anything else of that nature.
However, I will say this: looking at the impact of the heater on the car's battery elsewhere, leaving the heater on for 8 or so hours while the car was stopped would probably have produced about that much drop in battery, and could easily have led the range predicter to assume that power was wasting at a precipitous rate vs distance traveled, and therefore the remaining range was very short. Looking at the predicted range/distance chart, you can see that while elsewhere the slope was -1 or a bit worse when the heater or speed were up, on that leg (to the Norwich charger), it was almost flat and even rose slightly at around 410mi, where the temp was turned way down. Smells like an attempt to manipulate the predicted range, to me...
In any case, the text of the blog post is full of references to how long was spent at each charger station and so forth, so they obviously have logs over time. We need to see those in order to have any chance of understanding what was going on at that 400mi stopover...
Well, if you want to burn off some of an electric car's battery over the course of a few hours (such as a night stop without a charger) to make a good story, what better way to do it that to turn on the heater and leave it that way? Lights too, perhaps, but those draw way less than the heater; a typical car with a few hundred Watt-hours of battery can't leave its lights on overnight, but an electric easily could. As for the difference in the battery charge percentage loss and range percentage loss, some of that is most likely the non-linear curve, but it's also probably due to a look-ahead function. If the battery has been draining heavily from a user action (such as turning on the heater up to 75 and leaving it that way) while the car is not in motion, that could affect the predicted remaining range heavily depending on how the car estimates future behavior on the driver's part.
We really need logs over time to figure this out. How long was the heater set so high? How does that time correlate with the drop in battery charge and predicted remaining range? Were any other power-consuming features running over that night, or did the charge just mysteriously decrease?
What we really need to address these discrepancies are the same logs plotted over time. For example, at the 400mi mark, where the power and predicted mileage drops, the car's heater also spikes upwards and then down again. Broder claims he left the car overnight and came back to find it with a sadly reduced battery... but going by the logs, an equally valid (and possibly more plausible, when you remember how an electric car's heater works) possibility is that he left the car for a while with the cabin heater intentionally left running to drain the battery, then came back after burning off a chunk fo the charge. Now, if only we could see the time of day when he stopped at 400mi, and for how long, and what the battery charge, predicted range, and cabin settings were during that entire time... but we can't, because the car wasn't moving, so all events over that time period take place at the same point on the X axis.
I'm pretty damn sure that Tesla *has* the logs over time as well; it would sure help if they released them...
Oddly enough, you've actually got it backwards. Two things that computers can do really well is collect data, and compare that data to existing data and solve constraint problems to map inputs to outputs.
There are, as I noted, sites where users can input their own symptoms... but yes, those things are huge problems for hypochondriacs. They also rely on fairly generic background info, unable to take factor in specific conditions that maybe only three people in a million have... yet to a personalized expert system, that's just an extra data point that it can compare against established info about that condition.
The real kicker is the data collection, though. I know fuck-all about the expected levels of various blood chemistry. I have only basic knowledge of what blood pressure means and what ranges it should be in. I'm only a little better at temperature. I can tell you things like "my back hurts when I get up in the morning" or "I have a headache" but I would have a hard time pointing out the exact vertbra or describing . A personal unit, equipped with a camera and mic, plus tools to measure common basics like pulse, respiration, blood pressure, weight, and some very basic blood work... with some good image and audio recognition software to pick up things like changes to the eyes or skin, or breathing differently... we have software capable of detecting many symptoms at least as well as a human doctor, and a computer program would able to correlate symptoms, patient medical history, family background, and current conditions (say, if you're on some supposedly unrelated drug already) with known diseases and their treatments better than almost any specialist.
As for "duplicate a doctor's thinking", strictly speaking that is of course true - we don't have true AI - but why would you want to? Doctors have unreliable memory; they may forget facts, misremember or become confused, or overlook data points. It's much harder to keep a doctor's knowledge on the state of the art up to date than it is to update a computer's database. Doctors are subject to bias in favor of, or against, certain diagnoses or treatments. Expert systems suffer none of those flaws. Of course, they're only as good as the people who program them, but it's possible for many, many doctors to provide inputs for an expert system. If a human doctor tried to consult all the sources and historical records that an expert system can call upon for each diagnoses, he or she wouldn't be able to see enough patients to stay in business.
Besides, expert systems really do exist. You say "not in my lifetime" and in a funny way, you're right; they've existed, and have been able to out-perform typical human doctors on diagnoses, since before I was born. Doctors are still required, of course - the computer can't handle something that hasn't been seen before very well, and it is good to have a second opinion available. Software isn't infallible either; it will need to be modified and updated, which will require input from doctors. Finally, administering treatment may still be beyond the typical system for anything more complex than "take these pills at these times of day"... although there's no reason you couldn't have "auto-docs" capable of things like open heart surgery or organ transplants.
Yes and no. They can, for sure, be put in neutral. My car (a Subaru) may even have a physical linkage from the shifter to its CVT.
However... that's not true of all cars. For example, some Prius models are fairly obviously a purely electronic linkage. That will rely on the computer interpreting the driver's commands correctly... soemthing which obviously was not going on here.
Hmm... did he try hitting the accelerator to see if it would break the car out of its insanity?
Yep, even cars with a CVT (such as the Prius, which also uses push-button start on many models) have a neutral "gear".
The problem is that the shifter control is electronic, not physical. I mean, there's a physical thing that you push around, but the actual control to the transmission is an electronic signal and must be interpreted by the car's computer. Given that the computer apparently was under the impression that "brake" meant "speed up", this message may not have worked.
Of course, that's weird in and of itself. On every production car I've ever seen, no matter how much electronics it has, the brake pedal *does* have a physical linkage to the calipers. You can use it when the car is completely turned off, in fact, though you have to push pretty hard once the power braking assist goes away. You won't get traction control and such in that state either, but it doesn't matter - the car will stop. The brakes should also be more powerful than the engine (especially on a car that tops out at 125MPH) so you should be able to stop even with the enging revving and in gear (this might cause the engine to stall at that speed, but that's perfectly OK in this situation even if the transmission might not like it).
Agreed; I quite like mine. No dead pixels, though pixels will "sparkle" (not sure how else to describe it) if left on for a few days; power cycling or turning it off at night avoids this problem. Otherwise, it's great; beautiful and bright, good color and contrast, good response time. 2560x1440 is a good aspect ratio and plenty of vertical pixels. Cost was $290 USD on eBay, including DHL express shipping (48 hours from warehouse in Seoul to my place in Seattle) and an included adapter for US outlets (the power supply handles 110V/60Hz just fine).
I looked at comparable US options, and at 25x16 options, and they were all vastly more expensive. I'm happy with the purchase.
I actually have one of those 27" 2560x1440 ("WQHD") IPS monitors from Korea. $290 USD, included DHL express shipping (about 48 hours after it left the warehouse in Seoul to reach me in Seattle) via eBay. It's wonderful. Bright and clear, glossy screen but bright enough that reflections and background light are no problem, good stand, and simple but functional on-screen display. The DPI is nothing amazing, but it's comparable to my old 18" 1920x1080 monitor, and that's fine by me. I don't use the 5W speakers it has built in, so I can't say how those are.
The monitor does have a minor defect where if left turned on too long (several days straight) it will start getting "sparkles" on a black screen, but this is easily fixed by power-cycling the monitor or just turning it off every night (it starts very quickly, so that's no problem). It can also get pretty warm (especially at max brightness) and has a large-ish power brick (with a plug designed for Korean outlets, though they included a USA adapter for me at no extra charge) rated for 120W output.
Contrast is good but not incredible, but the lighting is very even. The in-plane switching works great; response time is excellent and the viewing angle is superb (the ~1/4" bezel gets in the way before the screen noticeably changes color). Color and saturation look good to my eye (untrained, but an amatuer photographer); it is something I look at and check calibration on.
I plan to buy a second one... just as soon as I figure out where I'm going to put it. I may just get a 1440x900 (or similar) instead and put it in portrait mode next to the big one. Otherwise, I'll probably need wall mounts; I'm running out of desk space.
One note of caution: It requires a dual-DVI input. That means no driving two monitors off one DVI connection, and many HDMI adapters, etc. won't work.
Eh... I'd say it's awesome that, out of the box, Win8 supports taskbar (and wallpaper, if you care) spanning across monitors. That used to require a third party utility. I love that it's finally supported in-box.
And what about those who aren't "regular viewers" but instead hear from somebody that "there's a review of this really cool new car on TV?"
Unless they explicitly state (for the benefit of new watchers, if nothing else) that the "review" is scripted, it's lies and bullshit. Fiction presented as fiction is entertainment. Fiction presented as fact - even when *most* people know it's not - is grounds for libel.
Tesla "won" the lawsuit but was awarded no damages as it was determined that no financial harm had been done to them. I'm skeptical, but I wasn't following the story so maybe that's perfectly true. In any case, I have no respect for Top Gear after that incident. "But it's just for funsies" is not a legitimate excuse when you're lying to people about a product (an expensive one very important to peoples' day-to-day lives, in this case) failing to live up to its manufacturer's stated capabilities, especially by supposedly conducting a test and then ignoring the results and following the script.
Lies and bullshit.
You know, not that I don't agree that the whole review smells very fishy, but it would help Tesla's (and Musk's) case if logs were released showing these same numbers over *time*. For example, as was pointed out, at 400mi, the internal temperature abruptly went up a bunch, then down again, while the car lost a ton of predicted range without moving.
What happened? Did he turn on the heater and sit in a parking lot for a few hours? Did he leave the car overnight and then come back in the morning to find the charge depleted (the way he claimed) but turned on the heater briefly to warm the cabin then started out anyhow? Did he leave the heater on all night, depleting 2/3 of the car's remaining charge while it was sitting still and unoccupied?
Now, this could be easily determined... if we had logs over time. In fact, a lot of things would benefit from this analysis.
Came here to mention this. Medical "expert systems" (a class of AI) have existed, and been better than humans at some things, for literally decades. However, you almost never hear about them anymore. With the vast advances in processing power, miniaturization, and power usage, we can today build pocketable devices that tremendously exceed the capabilities of something like Mycin. With the improvements in sensor technology, we can feed those systems more and better data than ever before. With widely available wireless Internet connectivity plus incredible storage densities, we can provide these systems with all the info they could ever need. With the advances in the science of AIs over the last four decades, we can make these systems "smarter" than was possible before. Finally, with the plummeting cost of such electronics, we can make such systems affordable to middle-class individuals, rather than exclusively to large and well-funded organizations.
Yet, they almost non-existent. We have a few smartphone apps that scratch the surface of what's possible. We have dedicated machines like continuous glucose monitors for diabetics, but they have very little in the way of smarts and no versatility. We have concepts and pseudo-prototypes of "medical tricorders" and such sci-fi devices, but they aren't generally set up to make recommendations. We have online web applications where users can input symptoms and be told what they might be, but those systems have no personal background or history of the user's health, and rely on the user measuring and providing data themselves.
Why haven't these things been combined?
If it weren't for the fact that TPB expressly permits copying their content, you mean? Less irony, more "I have no idea what I'm talking about but I'll point any laugh anyhow!"
This site runs a torrent tracker for files known to be under a copyright license that doesn't permit free redistribution? Wow, I had no idea...
Or did you mean "runs a site with some CSS and a search box that does something if you enter search terms for potentially copyrighted downloadable files"? Because I don't think that's the "function" of TPB that anybody else, including the courts, were referring to.
Swap "copyright infringement" for "piracy" and so forth below if you're going to get upset over the use of terminology that has been around for over half a century.
Funny, I hadn't heard that. The version I've heard is that TPB exists to facilitate piracy (which, given their name and policies, is a bit hard to refute) and therefore must be shut down. I've never heard any blanket statements like "everyone who retrieves from TPB is necessarily a pirate" except from people like you putting words in the mouths of others. It's quite possible to have the first (and take it too far) without the second. Scratch the hyperbole; it is insulting to the intelligent.
Your second line is accurate, if scarcely worth more than an eye-roll in terms of response. TPB expressly permits copying of their stuff. There's no piracy involved there. Similarly, they don't benefit in any way from the re-use of their CSS, so it's hardly a case of aiding a site which seeks to facilitate a crime.
Your third line is predicated on your first, and a mildly generous interpretation of your first at that. Typically if somebody speaks of "retrieving" from TPB, they'd be talking about retrieving information needed to download torrents, not webpage markup information.
Yes. If only one-party consent was required, they wouldn't even tell you; their people already know and don't pick up until after the recorded message anyhow.
There are multiple sites were people are hosting .APPX files (Metro app bundles) that can be downloaded and sideloaded onto a device. Sideloading is easy, and officially documented how to do it; it costs nothing aside from a very minor annoyance (enter a Powershell command, basically).
There's also a growing amount of desktop software for RT. It requires "jailbreaking" the device in order to run it, but once you do that it's just like any other Windows machine... except it runs ARM, so native code must be recompiled first (.NET code can run unmodified, as can Java code if you're willing to put up with the speed hit of IKVM). There's a project to provide an x86 emulation layer as well, bringing full (if again somewhat slow) support for legacy apps.
The Netscape Plugin API (which is what Flash is implemented using on Firefox, Opera, Safari, and at least some versions of Chrome) doesn't really support this. I don't know of any legit reason why not, but it doesn't.
IE, which uses an ActiveX control for Flash, actually has no problem installing or updating plugins without restarting the browser. Refresh the page (which, if it's installed via the browser itself, will happen automatically) and you're good to go.
Relatedly, I believe the reason that Word is being used as the exploit vector on Windows is because it doesn't have the sandboxing of IE/Firefox/Chrome. While you could get a lot more people to run the Windows attack code if you posted it on websites, it doesn't do any good when every popular browser newer than IE6 is locked down to not be able to launch arbitrary programs or write to most of the filesystem or registry.
Good for you... but it turns out that a fair number of the most popular websites on the entire WWW play videos, typically in Flash, and they are watched by people of all ages (perhaps more among the younger set, but certainly not exclusively). Quite a few sites (perhaps not individually the most popular, but a massively common *class* of site) also serve lots of Flash video, although for legal reasons they are only supposed to be watched by adults. People also like to watch videos of events they couldn't make it to and listen to streaming music, both of which are common uses of Flash. You can do web-based video chat or even videoconferencing using Flash (Google Talk can do this, for example).
I don't like Flash, and I certainly don't trust it; I keep it tightly curtailed where it's installed at all. However, it's definitely useful in some cases. HTML5 is catching up, but not fast.
You think in 1985 a device that would be literally incapable of running any software that doesn't come pre-loaded on it (there being Apple App Store at the time) would be considered a PC? You're delusional (and/or younger than I am). In 1985, every single PC sold came with, at a bare minimum, some form of BASIC or a similar language, an editor for writing programs in that language, and at least one of a compiler or an interpreter for running them. Many came with little else. Hell, to this day, this holds true; Windows has VBScript and JScript and Windows Script Host, CMD and BAT in Command Processor, PS1 and Powershell, and you can write them in Notepad if you want to. OS X has shell scripts and probably a few other things; I don't mess with Macs that much. Linux (any PC-oriented distro, or any distro you can buy loaded on PC hardware) comes with shell scripts, perl, python, and $DEITY knows what else, including plenty of editors for them, and quite likely also has gcc (out of the box). A bare-bones PC shipped with FreeDOS can still write and run QBASIC programs the same as a MS-DOS machine in the 80s could.
Note that I'm not even counting web browser scripting environments above. You can run a game on one, but you can't write out a file. You can create a spreadsheet app that runs on the browser, but good luck programmatically opening the spreadsheet file somebody handed you on a floppy (without doing a postback to the server). The iPad can't even do those things, though, because there's no way to write a web page (with or without script) and have the browser open it built into the iPad.
In 1985, an iPad is a camera you can't get your pictures out of, an email and text messaging appliance that can't connect to anything (no way to install a network connection in one, and what the hell is this "why-fie" thing?), a useless "web browser" tool, a mapping tool with no map data and no way to add any (or any way to tell you your latitude and longitude, as GPS was not yet deployed), a music and video player with almost no music or video (and no way to add more), a lame flashlight, an electronic calendar capable of reminders, a highly advanced alarm clock with terrible battery life, and a stupendously fast electronic calculator that is capable of only very basic operations. The last three (and maybe the first, if you didn't mind never being able to send the photos to anybody) would probably be its best features. Meanwhile, PCs were being used for writing papers, creating and editing spreadsheets, sending and receiving email, accessing Usenet, building models in everything from finance to weather, and playing games. People were also using them to write software for themselves, for other PCs, and for mainframes and supercomputers.
No, in 1985, the iPad would very much not be considered a PC (for much the same reasons that it doesn't qualify today). Steve Wozniak would probably have laughed in your face if you'd called it one.
You can run an ssh server on your iPad just fine without jailbreaking? Fascinating. After you ssh into it, do you prefer to start vim or emacs for your coding? Do you compile using gcc or clang? How good are its manpages? What shell do you use? Do you prefer to elevate privileges using su or sudo? If you ssh in from an unrecognized terminal, how difficult is it to install new terminfo? How many users can ssh into your iPad at once? If you wanted to install Python for iOS on your iPad, what command would you enter over ssh to do so? After doing so, could you demonstrate for me a Python script that checks the iPad's entire storage for any software which hasn't been started in over six months and offers to uninstall it?
Or, did you completely miss the point about 'uses applications not truncated "apps"'? While I dislike the terminology, that right there really is one of the defining differences between PCs and whatever category you want to put iPads in. Software for iOS is crippled. Yes, you can run an ssh client on your iPad and consume the services of a server (or any real PC, even a Windows one). You can even use the iPad to produce stuff on that server or real PC. But that doesn't make the iPad into a PC. Consumption-based doesn't mean it can't be used to produce anything, merely that the intended uses of the OS are not focused on productivity. This is the exact opposite of PCs, which at their inception could be used for very little except productivity.
The other difference, of course, is the degree of control. A Windows machine does not come with an ssh server, but if I want to install one I can. Ubuntu doesn't allow logging in as root by default, but if I want to change this, I can. If I buy a Windows PC but then decide I'd rather have it be a Linux PC, that's as simple as re-installing the OS. If I buy a machine with Ubuntu and decide that its graphical user interface is crap, I can remove Unity and install FVWM and have it start that after login instead. None of these are possible on an iPad. Sure, you can install software... if it's Apple-approved, content to run in a sandbox with restrictions that you can't control, and doesn't require mucking with the system configuration. Even the relatively incompatible Macs of the 90s didn't actively prevent you from installing your own OS on them... marketing aside, they were PCs. iPads simply are not, though.
Nobody except you and a few other fools are debating the subject of whether they are computers. Stop being an idiot (or is it your hope that idiotic arguments will distract from the fact that you haven't addressed the question?) Paper tape and punch-card computers weren't PCs either. PCs have, for as long as the term has existed, supported writing software and executing it on the same PC. In fact, the first PCs could do precious little else (aside from load a program that somebody else had written on their PC and then used their PC to write it to a diskette).
A secretary's PC *could* run code that the secretary wrote. If a secretary chose to write some code on the PC, it would be possible to do so (and run it!) without approval from anybody else. Now, if the computer in question was actually a locked down corporate workstation, then the secretary wouldn't be able to run their own software on it... but it arguably wouldn't be a PC either, because it's not personal to the secretary.
You know what the differences are between an iPad and a corporate workstation, in terms of whether they qualify as a PC? If you reformat a corporate workstation, you can install a PC operating system on it (i.e. one that you personally control). Try that with an iPad...
Are you an idiot or a troll? The GP wasn't addressing you, and you didn't answer the question for the GPP either. Nobody in this thread was discussing what makes anything a computer. User gstoddart appears to be claiming that anything which can be owned by an individual ("personal") and can simulate a Turing machine ("computer") is a PC. The exact question was
Perhaps you're the one missing the distinction between "computer" and "Personal Computer"? Or did you perhaps mean to reply to gstoddart directly, in which case the otherwise-unspecified "you" in your post would be assumed to refer to him/her?
In any case, my definition of the "Personal" in PC is much more specific than gstoddart's. Basically, if somebody doesn't *control* the computer themselves, it's not a Personal Computer. An iPad, without jailbreaking it, can only run iOS, can only run Apple-approved iOS apps, and only gives the user as much control over the device's capabilities as Apple wants them to have. The user has choices, but Apple controls what choices the user has. The nominal owner doesn't have control.