Ironically while Apple executives laugh at the president at the suggestion of iphone manufacture in the states, Samsung make their chips in the US. Really its a mystery why Apple is being considered at all they are as anti American as they come. If I was cynical I would suggest its part of the deal to bring back the some imac mini manufacturing to the states...although we have seen very little actual manufacture as yet.
But continue your hater-ade. And Samsung still supplies the A6 processors to Apple. And for final irony, The SGS4. Uses the US-made Exynos processor OUTSIDE NORTH AMERICA. And uses the TSMC-fabbed Qualcomm Snapdragon INSIDE North America.
On my supposedly "archaic" x86 desktop, I download any Linux distro I feel like using and can use the exact same installer to setup a 5 year old desktop or next month's Haswell.
I bet it wouldn't work on a NEC PC-Engine. Or an original Xbox.
You see, the thing is your PC is just ONE platform. Everything about the PC has been the same standard dating all the way back to the original IBM PC. RAM is always in the same location on every PC, and even today we still have the stupid 640ki-1M memory hole (for display). I/O ports are still in the same location as they always were.
Whereas on ARM, NOTHING is standard. Some SoCs have RAM at 0x0. Others at 0x40000000. Or 0x80000000. Or 0xC0000000. Boot ROM can be at 0x0. Or 0x10000000. Or wherever else. The serial ports? Anywhere. Display? Registers are randomly here or there, at least the memory is somewhere in RAM space and usually programmable.
Pluses and minuses on both. Minus for the PC is the horrendously discontiguous RAM space (there's another RAM hole around 3GiB-4GiB for memory mapped peripherals). Pluses means one OS image will work on all platforms because the kernel knows where everything is and will not change.
ARM Linux actually has undergone huge revisions to accommodate the fact that each SoC is different - it started with the platform_device that separates I/O addresses from drivers, and proceeded to the device-tree that expands on that even more. With proper coding, it's possible to have one kernel binary be able to boot several different SoCs. Of course, getting it to work across multiple manufacturers is much harder. Including multiple OEMs.
You mean like that stupidity of charging twice for the same shopping cart serial number when the final button is pressed twice? You get this shit when you let morons design it.
You mean the brilliance of being able to ding a customer for twice their shopping cart value? Extra profit from stupid and/or impatient people.
And when they chargeback, you can provide proof and cancel their order and still keep the other payment. And tie it up with confusion because you can easily switch which payment you're talking about to cause mass confusion.
The moronic thing would be to not accept a credit card, like one site I go to always claims the payment was denied. Call the bank and show that the payment was allowed. Now that is leaving money on the table by not accepting an order from a customer. But charging a customer multiple times for the same order? Brilliant business tactic.
The radio, and associated amplifiers, will generate the majority of the heat. Just look how much longer a cell phone will operate if you disable wireless. One must also take into consideration that wireless routers operate at higher power levels.
Actually, cellphones are higher powered - I believe they top out at between 0.5-1W max transmit power. Your wireless router is typically anywhere from under 50mW to 100mW, though it's possible to get "long range" ones that do 250mW.
Of course, a cellphone dynamically adjusts its power - in urban areas, it typically is close enough to a cell tower that it can crank the transmit power way down. This, of course, is to save battery (RF level amplifiers aren't efficient at all - they waste a lot of power). If you live in a poorly covered area, you'll note your battery life is a lot lower as a result of having to crank up the power to maintain the link.
CDMA phones are interesting - the amount of power they use is proportional to usage as the more phones using it, the lower the SNR. You've hit the limit when everyone's transmitting at max power and the SNR is too low for successful correlation.
Google Glass is visible, right there up on the wearer's face. What about all those cell phones that can do video recording, and can do that video recording right there from your shirt pocket, with no visible indication? Cameras are getting pretty small these days. Someone up to something nefarious, the camera lens is going to be one of his shirt buttons.
Thing is, Glass is like spectacles - it becomes common enough, you don't know who's recording and who isn't. (And apparently, the recording indicator is a lightpipe to the display, so a properly crafted "glassware" app can simply display black to not clue others into the recording).
For cellphones, either they're in breast pockets with the camera sticking out (which I've very rarely seen people put cellphones in their breast pocket), or people are holding them out in ways it's obvious that they're recording.
With Glass, people get used to seeing them like glasses, whereas most people don't hold their cellphones in front of them at the end of their arm to use it - it's generally pretty obvious. As would a cellphone sticking out of someone's breast pocket.
One other reason is well, if you need to bend over, the breast pockets tend to empty out....
My first thought when I read the summary was that hell had frozen over: Congress is thinking about privacy!
My second thought was that *Congress is thinking about privacy*. This can only be a good thing. I think we should encourage them, saying "you're on the right track, keep going that way" rather than being derisive.
Parent is right, government surveillance/data collection is a huge privacy issue. That does not mean it's the only privacy issue. It is easier for our inherently timid Congresscritters to start by pointing the finger outward from Washington, and I'm OK with that because it at least starts the policy discussion we so desperately need.
No, what happened is that the interest of politicians and the people they're supposed to represent aligned in this one case.
You see, imagine if people were using Glass - and recording stuff around them. Let's say it captures a politician coming out of a less-than-completely-upstanding business (which could be anything someone can raise much about). That image is stored and uploaded to Google, and possibly tagged. Now any political opponent can go and claim that said politician believes in X because they just came from a store that supports it.
Think of anything mildly controversial and see how it can get blown up. Perhaps it was a store selling porn - I'm sure the family first groups will use that at any opportunity (and I'm sure it's probably a common enough event, but one that can be used as leverage).
Basically, they're worried about politicians being captured on film doing stuff. It may be normal behavior that gets twisted around like a quote out of context, or it could be someone capturing actual backroom deals taking place, etc.
And the cynical side of me says it's because the politicians don't want any recording of them doing anything "bad" like being seen with industry executives that support them, or being hypocritical, etc.
RIM certainly has SOME part of the code and as such they can give out the relevant stuff to the authorities, including the BASE KEYS.
That 'government certification' nonsense is just that.
Uh, no.
You're confusing the consumer "internet edition" with "Enterprise edition".
Internet edition blackberries are what you get when you go to your carrier and buy it on a blackberry plan and they give you email and all that. In which case, what happens is your blackberry connects to RIM's servers and gets your mail through RIM proxying to your carrier's email inbox.
BES though is different. You pair a blackberry with BES and they generate a set of keys. Your blackberry proxies its connection with RIM to reach BES. When it gets to BES, the data transferred is using the key set up during pairing. End to end, it's encrypted.
What RIM did with India was set up a RIM server in there, so internet edition phones proxy through it, and then onto the carrier email server. When a BES attached one does it, the link is still encrypted because that server does not have the key.
Basically, all BB traffic goes through RIM or a RIM server set up in the middle east or india or wherever. From there, the server is what contacts the mail server you're using. As that part is unencrypted, they can decrypt your email and such at that point.
HOWEVER, use BES, and what happens is the RIM server connects to your BES server and your BES server then communicates to your blackberry via the pre-shared key. No one can snoop on that email because its encrypted with keys only your blackberry and BES know. Even with those servers they can't examine your traffic because the server does not have the key.
Of course, the bigger question is who buys a blackberry and NOT use BES with it...
Reforming patent law would be simple, software should simply not be patentable. You can copyright it sure, but no patents.
Thing is, software is very special.
Prior to the computer age, humans generally created "stuff" or "art". Stuff like mechanical things - which are easily patentable, but not copyrightable. "art" things were copyrighted because they didn't generally serve any purpose other than aesthetic or entertainment. Of course, one could create mechanical art, but the utility of such generally wasn't there, and useful machines that also looked good generally were patented because they were stuff.
But software is neither. It's written, which implies copyright, but it can be stuff as well - like when it's a fundamental part of hardware (embedded software). Or maybe it IS hardware, when you write your RTL code for an FPGA or ASIC.
And therein lies the problem. Software is everywhere, and saying "we can't patent software' means that if I invent something that uses software in a fundamental way, that whole system of software+hardware may be unpatentable. However, if I were to implement the same software as a complex piece of hardware, it suddenly IS patentable? Like say I come up with a way to do radio transmissions more efficiently and closer to Shannon's limit than ever before. If I use it in an SDR, it would be software implemented, but if I implemented it using standard radio hardware and building blocks of mixers and detectors and other things, it would obviously be a hardware implementation. It would mean the former gets no protection, while the latter does.
The reality of life is we probably need to come up with a new form of IP protection, called, well, software. Thus it covers software, or anything written that requires hardware to perform some action. So my hardware+software thing - I can patent the hardware and protect the software inside it the same way I could with a complex assembly of chips and mechanical pieces I use to work around the need for a line of code. Likewise the RTL code would be covered under the software protection. As would Windows. or Word. Or Linux. Or whatever else.
The problem is software is very unique - it does things and sometimes it does things in clever novel ways. But at the same time, it's also something that's fixed onto a medium which means you now have two competing protections for it. Neither of which are completely adequate, either. Copyright may be good for the source, but it doesn't really handle the binary side very effectively (is the output of a compiler copyrightable? Or just the part that underwent human creativity, i.e., source code?). Patents make sense for some software (used to generate better ways of controlling some piece of hardware), but not for others (e.g., application software). It's also unique because software can be hardware.
That's because the reality is that most people don't use advanced mathematics (or, these days, hardly any mathematics at all) in their day-to-day lives. Most simple mathematical exercises in the modern world have been automated, and the complex stuff is largely the purview of engineers and other specialized pros. Academia is the only place most people ever encounter it, and very few people spend their whole lives as students (my son being a rare exception).
The article doesn't refer to advanced math, just basic arithmetic - addition, subtraction, multiplication, division. Of which is such a basic skill it's used daily in many activities, including ones you don't know exist.
A basic use is shopping - is the 12oz bottle for $4 a better deal than the 16oz bottle for $5? What is the current approximate value of your shopping cart? Including tax? (Do you have enough money?) If you're having a party and they like your cake, how much extra ingredients do you need to buy so you can bake enough for everyone?
One thing I found that helped were the "brain training" games. They fell out of favor when it turns out they can't increase your general brain health, but the footnotes all noted that they improved significantly on the areas they trained in. So doing 100 basic math problems daily did in general improve basic arithmetic ability.
Heck, even doing 100 problems shouldn't take more than a couple of minutes (you don't need much - just adding/subtracting/multiplying and dividing involving digits you find on a multiplication table is good enough - i.e. add/sub/mul single digits, and division has a dividend in the double digits while a divisor in the single digits).
Sounds like a liability. "That car accident wasn't my fault. That particular robotic bartender always manages my alcohol intake perfectly, so it is the one at fault for screwing up and letting me drink a little too much, too fast. It was programmed to cut me off one drink sooner but a bug let it give me one more drink, and made me t-bone a taxi full of nuns on vacation."
I'll take the bartender I know, who pours me a beer the moment I walk in the door and makes sure I have a ride on the rare occasion when I allow myself to get carried away. He's a better bullshitter than any robot I've met, too.
In a lot of places, that is actually true - the establishment can be found responsible for a DUI incident as well and forced to provide compensation. And yes, the establishment includes commercial public bars as well as hosts of a private party.
The general reasoning goes that the best person to judge would be the person providing the drinks. It could be the bartender (since all the drinks come fro them), the waitstaff (who carry it to you and can tell if you've perhaps had a bit too much), the hosts (it's their party, they should be encouraging responsible drinking), etc. Because alcohol impairs judgement, so expecting the user to be in control while impaired is generally not feasible.
So yeah, a robotic bartended will be held to similar standards - first, it knows who had what drinks and thus should have some basic intelligence at saying some guy should not have put back 10 shots in the past hour , and second, the establishment should be monitoring their guests even if the robot is unable to properly determine the inebriation of the patron because the patron is not in a position to properly judge.
So not only is your bartender doing you a favor by checking your ride, they're covering their ass as well because the family of the kid you ran over accidentally can sue the bar you went to for damages.
do not be worried that you're not embracing all the stuff that the masses embrace.
True, argumentum ad populum is usually a fallacy. But sometimes it isn't. Economies of scale in manufacturing is one case. As the masses have moved from "netbooks" (10" laptops) to tablets, it has become more difficult for a happy netbook user to find a new replacement for failed hardware. Communication platforms are another case. If people aren't willing to make their writing available through an open technoloby such as an Atom feed but instead prefer to lock their communication inside the closed systems of Twitter and Facebook because everyone else is doing it, one has to join what everyone else is doing in order to be able to communicate with everyone else.
The reason the masses embrace reinventions of old technology is because modern technology makes it easier to use and understand. Facebook makes it easy to get in contact with your friends - no longer do you have to understand a cryptic email address - you just shoot off a message to "J. Doe" on your list of contacts and can be reasonably sure that you're talking to the right Doe (because you can click their profile to look it up) but also not send it to the wrong Doe (your boss, say). And in one place, you can see what they were up to. At least, that's the marketing glitz that attracts people in.
Sure they could email, view blogs and such, but that's more complex and takes more time than having the information in one place, not having to deal with cryptic addresses (is this the right Doe?), having to deal with uploading photos and converting them, etc. (Remember the old jokes about people emailing 50MB worth of photos? They're true).
Basically, the reinventions are aimed at the non-technical market. It's also core to what Apple does - they repackage existing technology for the masses.
Netbooks basically were a fad because the masses wanted a cheap computer to do their facebooking and other stuff with - to passively consume content and such. But the formfactor was wrong, screens small, keyboards crappy, and all sorts of other things evident when trying to cut corners and repackage a PC to cost under $300. They killed themselves off when manufacturers, fed up at not being able to make money, started charging more and more ($400 and $500 netbooks became REALLY common. And yes, at $500, they were running into low-end laptops.).
Then Apple introduced the iPad that basically answered a lot of people's needs for a netbook (like a tablet, netbooks aren't for creating with their puny keyboards and such - and larger ones were again, costing the same as laptops).
Other manufacturers jumped into the same ring realizing they could make much more money selling a tablet rather than a netbook.
The old netbooks are still around - I still see them for $300, but they aren't movers anymore as for $100 more, you get a glitzier tablet with larger screen, multimedia, less hassle, etc.
As for technies, there's nothing wrong with the old methods - stability is good as well. But just as the masses aren't going to take up vi or emacs, it doesn't mean vi and emacs users are outdated and antiquated.
Obsolete is what happens when there are far better ways of doing what you're doing but keeping to the old methods. Like say, using a typewriter over a word processor. Using the latter means if you make a typo, you hit Backspace and fix it. You don't have to retype the entire page, or dig out the liquid paper or correction tape. Thus, the word processor obsoletes the typewriter, even though it's not entirely perfect (after all, a mechanical typewriter works with no power). But the common use case was people had power (and electric typewriters), thus making word processors far more useful for creating documents.
The old methods of shell windows, vi/emacs text editing, etc., haven't been obsoleted because there generally aren't good replacements. And vi and emacs have generally als
Both Larry and Sergei are no longer connected with reality.. I don't begrudge them anything, but they are seriously in outer space.
I would add Eric to the list as well. There must be something in the water at Mountain View. Or maybe Google has a RDF that's more powerful than Apple's. Something very bizarre is coming out of their mouths.
Perhaps it's what happens when one is a Glasshole?
That might fly for the advertising, but including a download functionality requires a deliberate effort - Microsoft is willfully including a tool with no functionality except to facilitate in the violation of Google's license agreement, and thus copyright. If this ever turns into a court case, MS would probably lose - but they could still drag it on long enough to cost both sides a few million dollars in legal fees, and get a lot of good press if they spin it right.
Given the way youtube works effectively gives you a download of the video, all Microsoft did was add the ability to save the file it downloaded.
Aren't all the open-source folks angry when some company tries to pull this stunt? After all, YouTube doesn't stream the video - it basically shoves the entire video to you. Granted, Google traffic shapes the videos now, but they don't really protect it. Hell, one could in theory use the HTML5 version, right click and save target as.
Microsoft's implementation of a download function is no different than open-source doing the same thing with many other services - hell, the FOSS community (including/.) gets defiant when some company gets miffed that someone wrote a third party client that implements these features.
If we force these companies to adapt their businesses to not do stupid things, perhaps Google should for YouTube as well. Hell, there are plenty of YouTube downloaders out there as well.
Or it could be a bunch of Google apologists willing to overlook anything Google does even though it's the same as what others have done as well , only the latter have been mocked for the very same actions.
Thunderbolt is more exciting. It's PCIE over a cable. So you can have an external graphics card and enclosure. Plug power cable into wall. Plug thunderbolt cable from video card into notebook and voila.... top end graphics power. With some variations the thunderbolt tech the cable could carry enough power to power the laptop over the thunderbolt cable.
The great thing is you still have a great portable laptop that can focus on saving power and having a great battery life but can be upgraded on the spot to a powerful gaming computer when you really need performance. The same setup can also upgrade a desktop computer in the same way so you can have a couple desktop computers and multiple notebooks and only need to buy one high end graphics card which can be quite an investment.
Sony has a laptop that does this - it's not a standard Thunderbolt connector (a rather non-standard USB one), but it has Intel graphics while mobile, and you can plug it into the "mobile dock" while at a desk which includes a Blu-Ray drive and high end graphics.
There's a third party manufacturer that's supposed to make a thunderbolt to PCIe dock for Macs so you can put in a nice graphics card in it and do just that as well. Previously they made Expresscard to PCIe docks (Expresscard is PCIe x1 + USB, while I think Thunderbolt allows PCIe 2.0 x2 or so plus displayport).
This weapon will never be used in anger by any entity authorized to use lethal force in anger: snipers would never use this, it is too expensive and is unnecessary for the average foot soldier, and too large and cumbersome to be used on anything other than a rifle that is stationary and supported, ie on a target range. This technology is clearly designed for target and hunting use only, which would completely negate the point of both activities.
You know, it's NOT that much more than a good ol' M82A1 or M107 - they are both in the 5-digit range though you can configure the M82 with iron sights to get it to just below that threshold.
And last I checked, most sniper rifles were big and heavy, as well. (The M82's about 30lbs, M107's about 10% lighter).
Now, it's true it takes the "fun" out of the shoot, but it depends what the shoot was about now. If you want to kill an animal humanely for food, it seems like an option to consider.
But I suppose the real argument boils down to the "command line vs. the GUI". People argue it takes the fun out of the sport, while others say it makes them more productive. And in this market, there's nothing that can't say we can't have both. It's like how some people wish to drive sporty cars with manual transmissions while others would pay to get a self-driving car.
Though, one does wonder how well it tracks human targets. It's not supposed to, but it's an interesting consideration some military might consider.
Guns kill. Now, that kill may be done for sport and recreation, necessity, or other reason. The ones doing the sport will probably not buy it. The ones doing it for necessity, might. Who's to say, and why can't we have the option?
These instructions are undefined; they work by taking advantage of the internal CPU architecture to execute a hybrid of other legal opcodes. A lot of other older processors have such behavior, such as the Z80. Even the 8086 had a bit of this: "pop cs" and the second encoding of "sar" come to mind. (The 8086's "pop cs" was stolen by the 286 to mean an escape to a second opcode page.)
The reason for this is simple - the instruction decoder is using a bunch of logic gates to figure out how to drive the various processing bits around, referencing a microcode ROM if necessary. The deal was that most of the opcode space was not completely used up so there are holes where no instruction is officially implemented. Using those encodings will not generate an error (most processors didn't have an illegal instruction error, nor did most microprocessors expect anything more than to crash on error). So these illegal instruction encodings merely tickled the instruction decoder in strange and wonderful ways.
A more modern processor like the Motorola 68K, however, implemented illegal instruction exceptions and thus they have no such features.
And modern processor architectures today often have plenty of gaps in the instruction space - for future instructions or whatnot, and they too generate illegal instruction exceptions. ARM actually defines an ILLEGAL opcode to cause that very exception.
What I don't quite understand is how they first explain how difficult it is to find these planets, relying on tiny changes in the star's luminosity because the planets themselves are too dim to observe directly, and then go on to describe this planet in great detail: diameter, the fact that it's tidally locked, temperature at different locations, jet stream winds,... On a planet 2000 light years away? How did they get this "stong evidence"? Did they tune in to a local alien weather station?
They're not mutually incompatible.
What you need to realize is that detecting planets is hard - because they're dim compared to their parent stars. So pointing your telescope at a random star may or may not reveal a planet.
Basically, it's a problem because space is big. Really big. And modern theory has it's filled with "dark stuff" (energy or matter - they're equivalent with E=mc^2) of unknown composition.
But once you know something's there, it's easier to analyze it - spectral analysis among other methods.
Basically, finding a planet is looking for a needle in a haystack. Once you found the needle though, you can analyze the crap out of it as you know it exists and where it is.
I have an iPod, not an iPhone. However, the AppleMobileDeviceService.exe process is running in the background. I have never seen it gobble up cycles. It normally sits at "00" CPU. When I plug in the iPod, it jumps to an incredible "01" for a very short interval and then returns to "00". So, does anyone else have this process gobbles up to 50% of the CPU?
I looked at my computer, and while it's had an uptime of probably since April's patch tuesday, that service has consumed a grand total of... 1m53s of CPU time.
He never mentions what version of iTunes he's using - perhaps it's still 10.x, which is horrible. iTunes 11 has actually fixed a LOT of stuff and is actually pretty decent and more importantly, fast. It's incredible how fast iTunes is nowadays. I'm not sure what Apple did, but damn it's fixed a lot of stutters, halts, and stalls.
Similarly, Google services don't seem to screw up Windows or Linux, and Google's MTP support for Mac (MTP is required for Nexus 4) is ridiculously minimal. It's an analogous situation. Vendors for system X don't care about system Y, news at 11.
Actually, Android's implementation of MTP is crap. It BARELY works. Hell, on Windows I can easily screw it up if I try to open a folder before it finishes enumerating.
One of our developers actually worked on the Mac trying to get Android to work better (via OSX FUSE). Besides complaining about the crappiness of libusb, he constantly complains about how crappy Android MTP is. Do something that is beyond the basic commands and it'll crash Android.
Hell, everything else I have enumerates folders and files faster than Android - it's damn sluggish because I think the Windows MTP client is being forced to use the old clunky method.
All Google did was get MTP to the point where "it works", but skipped "it works well". I'm surprised Linux MTP works - everyone I've seen had plenty of issues getting MTP and Android talking without issues. It works erratically, at best.
The s-meter in my SDR software has native Klingon support. It's one of the easter eggs. I'm imagining people finding it, then actually translating the s-meter readout by going to Bing. Having a little trouble with how they'll encode the input font, but I'm sure MS has it all figured out. Perhaps it's OCR.
However, it's rare to find something with Firefox that can't be changed via a simple plugin or even just a setting in about:config.
There's one that's been bugging me since 14 or 15 though - autocomplete no longer autocompletes deep URLs - it only goes up to the domain. Which is annoying if you have a particularly favorite long URL (like a search), or access a server on another port (no, it doesn't autocomplete ports, either). Sure, the one you want is there - just down arrow enter, but that's annoying.
Haven't found a way to revert it back to autocomplete from history yet including the full URL.
But, to be frank, apart from those, I can't tell the difference between 18, and 20. And looking at the changelog, I can't see anything that says, "I'm a major new version that breaks compatibility with previous versions".
FF20 added that horrendous download box, for starters...
Of course, you can revert it back to the more sane old download list by setting browser.download.useToolkitUI to TRUE.
It isn't that hard to miss in FF20. Not sure what UI breakage they did in 21, though.
I really don't understand why we still allow campaign contributions to specific politicians. IMO, this is the single biggest flaw in our political system currently. All contributions should be pooled and divided equally among all candidates. This should have been dealt with decades ago..
Actually, what would help is removing tax incentives for donations for political purposes. So whether it's a campaign contribution or whatever, it's paid with after tax dollars like you would buy any other thing.
In fact, the tax code is complex enough that in some cases, it's more beneficial to give $10 to a political party (or person) than it is to give that same $10 to a charity! This is direct tax benefits, not just indirect benefits, at that.
Oh, and for a time, Canada had that system - every party had a fixed amount of money which went to each party (something like $1.25 per vote). The Harper Government got rid of it for two reasons - 1) because people don't want to support something they didn't vote for, and 2) government shouldn't be wasting tax money.
1) is obviously false because they voted for them, so it's obvious they wanted to support the party. Your vote is paid for by your taxes so regardless of who you voted for, you're paying for that through your taxes. The only person who would complain is those that didn't vote, which is their fault, really.
2) Well, the real case is "keep the Conservatives in power" because they failed to bring in any other measures like well, get rid of tax incentives (would hurt them), or reduce the donation limit from $5000 per person to $1000 per person (again, would hurt them).
So really, in the end, it's seen as a political ploy.
It was just a cam, but it was paired with software that did the motion recognition (usually poorly unless you had floodlights in the room). EyeToy compatible games touted motion control with just the EyeToy and your living room. Pity it was so bad at actually detecting accurate motion. I tried mine with DDR and with the EyeToy sampler disc and I failed to get any real accuracy out of it. Then again, I paid a whopping $10 for the EyeToy yoga bundle on clearance.
The problem was EyeToy worked by using background differencing. It took two images - and then differenced them to find the motion. (It used a few more frames than that to determine what was background and what wasn't, but the idea holds).
As you can imagine, this works to a limited extent, but if you move too slowly, the differencing algorithm fails because you get low pass filtered as background. In addition, you can't differentiate between things at different depths, so if something falls behind you, it's treated as your motion.
Kinect adds depth sensing (via IR dots) to add depth, making it easier to determine background from not background, as well as do skeletal tracking and such. This gives the ability to differentiate between background movement and foreground, separate multiple people out, etc,
Slashdot Mirror
I know Americans have poor geography skills, but last I checked, Austin, TX, was in the US. And then Samsung spent $4b upgrading it.
But continue your hater-ade. And Samsung still supplies the A6 processors to Apple. And for final irony, The SGS4. Uses the US-made Exynos processor OUTSIDE NORTH AMERICA. And uses the TSMC-fabbed Qualcomm Snapdragon INSIDE North America.
I bet it wouldn't work on a NEC PC-Engine. Or an original Xbox.
You see, the thing is your PC is just ONE platform. Everything about the PC has been the same standard dating all the way back to the original IBM PC. RAM is always in the same location on every PC, and even today we still have the stupid 640ki-1M memory hole (for display). I/O ports are still in the same location as they always were.
Whereas on ARM, NOTHING is standard. Some SoCs have RAM at 0x0. Others at 0x40000000. Or 0x80000000. Or 0xC0000000. Boot ROM can be at 0x0. Or 0x10000000. Or wherever else. The serial ports? Anywhere. Display? Registers are randomly here or there, at least the memory is somewhere in RAM space and usually programmable.
Pluses and minuses on both. Minus for the PC is the horrendously discontiguous RAM space (there's another RAM hole around 3GiB-4GiB for memory mapped peripherals). Pluses means one OS image will work on all platforms because the kernel knows where everything is and will not change.
ARM Linux actually has undergone huge revisions to accommodate the fact that each SoC is different - it started with the platform_device that separates I/O addresses from drivers, and proceeded to the device-tree that expands on that even more. With proper coding, it's possible to have one kernel binary be able to boot several different SoCs. Of course, getting it to work across multiple manufacturers is much harder. Including multiple OEMs.
You mean the brilliance of being able to ding a customer for twice their shopping cart value? Extra profit from stupid and/or impatient people.
And when they chargeback, you can provide proof and cancel their order and still keep the other payment. And tie it up with confusion because you can easily switch which payment you're talking about to cause mass confusion.
The moronic thing would be to not accept a credit card, like one site I go to always claims the payment was denied. Call the bank and show that the payment was allowed. Now that is leaving money on the table by not accepting an order from a customer. But charging a customer multiple times for the same order? Brilliant business tactic.
Actually, cellphones are higher powered - I believe they top out at between 0.5-1W max transmit power. Your wireless router is typically anywhere from under 50mW to 100mW, though it's possible to get "long range" ones that do 250mW.
Of course, a cellphone dynamically adjusts its power - in urban areas, it typically is close enough to a cell tower that it can crank the transmit power way down. This, of course, is to save battery (RF level amplifiers aren't efficient at all - they waste a lot of power). If you live in a poorly covered area, you'll note your battery life is a lot lower as a result of having to crank up the power to maintain the link.
CDMA phones are interesting - the amount of power they use is proportional to usage as the more phones using it, the lower the SNR. You've hit the limit when everyone's transmitting at max power and the SNR is too low for successful correlation.
Thing is, Glass is like spectacles - it becomes common enough, you don't know who's recording and who isn't. (And apparently, the recording indicator is a lightpipe to the display, so a properly crafted "glassware" app can simply display black to not clue others into the recording).
For cellphones, either they're in breast pockets with the camera sticking out (which I've very rarely seen people put cellphones in their breast pocket), or people are holding them out in ways it's obvious that they're recording.
With Glass, people get used to seeing them like glasses, whereas most people don't hold their cellphones in front of them at the end of their arm to use it - it's generally pretty obvious. As would a cellphone sticking out of someone's breast pocket.
One other reason is well, if you need to bend over, the breast pockets tend to empty out....
No, what happened is that the interest of politicians and the people they're supposed to represent aligned in this one case.
You see, imagine if people were using Glass - and recording stuff around them. Let's say it captures a politician coming out of a less-than-completely-upstanding business (which could be anything someone can raise much about). That image is stored and uploaded to Google, and possibly tagged. Now any political opponent can go and claim that said politician believes in X because they just came from a store that supports it.
Think of anything mildly controversial and see how it can get blown up. Perhaps it was a store selling porn - I'm sure the family first groups will use that at any opportunity (and I'm sure it's probably a common enough event, but one that can be used as leverage).
Basically, they're worried about politicians being captured on film doing stuff. It may be normal behavior that gets twisted around like a quote out of context, or it could be someone capturing actual backroom deals taking place, etc.
And the cynical side of me says it's because the politicians don't want any recording of them doing anything "bad" like being seen with industry executives that support them, or being hypocritical, etc.
The only difference is whether the cargo is self-loading or not.
Uh, no.
You're confusing the consumer "internet edition" with "Enterprise edition".
Internet edition blackberries are what you get when you go to your carrier and buy it on a blackberry plan and they give you email and all that. In which case, what happens is your blackberry connects to RIM's servers and gets your mail through RIM proxying to your carrier's email inbox.
BES though is different. You pair a blackberry with BES and they generate a set of keys. Your blackberry proxies its connection with RIM to reach BES. When it gets to BES, the data transferred is using the key set up during pairing. End to end, it's encrypted.
What RIM did with India was set up a RIM server in there, so internet edition phones proxy through it, and then onto the carrier email server. When a BES attached one does it, the link is still encrypted because that server does not have the key.
Basically, all BB traffic goes through RIM or a RIM server set up in the middle east or india or wherever. From there, the server is what contacts the mail server you're using. As that part is unencrypted, they can decrypt your email and such at that point.
HOWEVER, use BES, and what happens is the RIM server connects to your BES server and your BES server then communicates to your blackberry via the pre-shared key. No one can snoop on that email because its encrypted with keys only your blackberry and BES know. Even with those servers they can't examine your traffic because the server does not have the key.
Of course, the bigger question is who buys a blackberry and NOT use BES with it...
Thing is, software is very special.
Prior to the computer age, humans generally created "stuff" or "art". Stuff like mechanical things - which are easily patentable, but not copyrightable. "art" things were copyrighted because they didn't generally serve any purpose other than aesthetic or entertainment. Of course, one could create mechanical art, but the utility of such generally wasn't there, and useful machines that also looked good generally were patented because they were stuff.
But software is neither. It's written, which implies copyright, but it can be stuff as well - like when it's a fundamental part of hardware (embedded software). Or maybe it IS hardware, when you write your RTL code for an FPGA or ASIC.
And therein lies the problem. Software is everywhere, and saying "we can't patent software' means that if I invent something that uses software in a fundamental way, that whole system of software+hardware may be unpatentable. However, if I were to implement the same software as a complex piece of hardware, it suddenly IS patentable? Like say I come up with a way to do radio transmissions more efficiently and closer to Shannon's limit than ever before. If I use it in an SDR, it would be software implemented, but if I implemented it using standard radio hardware and building blocks of mixers and detectors and other things, it would obviously be a hardware implementation. It would mean the former gets no protection, while the latter does.
The reality of life is we probably need to come up with a new form of IP protection, called, well, software. Thus it covers software, or anything written that requires hardware to perform some action. So my hardware+software thing - I can patent the hardware and protect the software inside it the same way I could with a complex assembly of chips and mechanical pieces I use to work around the need for a line of code. Likewise the RTL code would be covered under the software protection. As would Windows. or Word. Or Linux. Or whatever else.
The problem is software is very unique - it does things and sometimes it does things in clever novel ways. But at the same time, it's also something that's fixed onto a medium which means you now have two competing protections for it. Neither of which are completely adequate, either. Copyright may be good for the source, but it doesn't really handle the binary side very effectively (is the output of a compiler copyrightable? Or just the part that underwent human creativity, i.e., source code?). Patents make sense for some software (used to generate better ways of controlling some piece of hardware), but not for others (e.g., application software). It's also unique because software can be hardware.
The article doesn't refer to advanced math, just basic arithmetic - addition, subtraction, multiplication, division. Of which is such a basic skill it's used daily in many activities, including ones you don't know exist.
A basic use is shopping - is the 12oz bottle for $4 a better deal than the 16oz bottle for $5? What is the current approximate value of your shopping cart? Including tax? (Do you have enough money?) If you're having a party and they like your cake, how much extra ingredients do you need to buy so you can bake enough for everyone?
One thing I found that helped were the "brain training" games. They fell out of favor when it turns out they can't increase your general brain health, but the footnotes all noted that they improved significantly on the areas they trained in. So doing 100 basic math problems daily did in general improve basic arithmetic ability.
Heck, even doing 100 problems shouldn't take more than a couple of minutes (you don't need much - just adding/subtracting/multiplying and dividing involving digits you find on a multiplication table is good enough - i.e. add/sub/mul single digits, and division has a dividend in the double digits while a divisor in the single digits).
In a lot of places, that is actually true - the establishment can be found responsible for a DUI incident as well and forced to provide compensation. And yes, the establishment includes commercial public bars as well as hosts of a private party.
The general reasoning goes that the best person to judge would be the person providing the drinks. It could be the bartender (since all the drinks come fro them), the waitstaff (who carry it to you and can tell if you've perhaps had a bit too much), the hosts (it's their party, they should be encouraging responsible drinking), etc. Because alcohol impairs judgement, so expecting the user to be in control while impaired is generally not feasible.
So yeah, a robotic bartended will be held to similar standards - first, it knows who had what drinks and thus should have some basic intelligence at saying some guy should not have put back 10 shots in the past hour , and second, the establishment should be monitoring their guests even if the robot is unable to properly determine the inebriation of the patron because the patron is not in a position to properly judge.
So not only is your bartender doing you a favor by checking your ride, they're covering their ass as well because the family of the kid you ran over accidentally can sue the bar you went to for damages.
The reason the masses embrace reinventions of old technology is because modern technology makes it easier to use and understand. Facebook makes it easy to get in contact with your friends - no longer do you have to understand a cryptic email address - you just shoot off a message to "J. Doe" on your list of contacts and can be reasonably sure that you're talking to the right Doe (because you can click their profile to look it up) but also not send it to the wrong Doe (your boss, say). And in one place, you can see what they were up to. At least, that's the marketing glitz that attracts people in.
Sure they could email, view blogs and such, but that's more complex and takes more time than having the information in one place, not having to deal with cryptic addresses (is this the right Doe?), having to deal with uploading photos and converting them, etc. (Remember the old jokes about people emailing 50MB worth of photos? They're true).
Basically, the reinventions are aimed at the non-technical market. It's also core to what Apple does - they repackage existing technology for the masses.
Netbooks basically were a fad because the masses wanted a cheap computer to do their facebooking and other stuff with - to passively consume content and such. But the formfactor was wrong, screens small, keyboards crappy, and all sorts of other things evident when trying to cut corners and repackage a PC to cost under $300. They killed themselves off when manufacturers, fed up at not being able to make money, started charging more and more ($400 and $500 netbooks became REALLY common. And yes, at $500, they were running into low-end laptops.).
Then Apple introduced the iPad that basically answered a lot of people's needs for a netbook (like a tablet, netbooks aren't for creating with their puny keyboards and such - and larger ones were again, costing the same as laptops).
Other manufacturers jumped into the same ring realizing they could make much more money selling a tablet rather than a netbook.
The old netbooks are still around - I still see them for $300, but they aren't movers anymore as for $100 more, you get a glitzier tablet with larger screen, multimedia, less hassle, etc.
As for technies, there's nothing wrong with the old methods - stability is good as well. But just as the masses aren't going to take up vi or emacs, it doesn't mean vi and emacs users are outdated and antiquated.
Obsolete is what happens when there are far better ways of doing what you're doing but keeping to the old methods. Like say, using a typewriter over a word processor. Using the latter means if you make a typo, you hit Backspace and fix it. You don't have to retype the entire page, or dig out the liquid paper or correction tape. Thus, the word processor obsoletes the typewriter, even though it's not entirely perfect (after all, a mechanical typewriter works with no power). But the common use case was people had power (and electric typewriters), thus making word processors far more useful for creating documents.
The old methods of shell windows, vi/emacs text editing, etc., haven't been obsoleted because there generally aren't good replacements. And vi and emacs have generally als
I would add Eric to the list as well. There must be something in the water at Mountain View. Or maybe Google has a RDF that's more powerful than Apple's. Something very bizarre is coming out of their mouths.
Perhaps it's what happens when one is a Glasshole?
Given the way youtube works effectively gives you a download of the video, all Microsoft did was add the ability to save the file it downloaded.
Aren't all the open-source folks angry when some company tries to pull this stunt? After all, YouTube doesn't stream the video - it basically shoves the entire video to you. Granted, Google traffic shapes the videos now, but they don't really protect it. Hell, one could in theory use the HTML5 version, right click and save target as.
Microsoft's implementation of a download function is no different than open-source doing the same thing with many other services - hell, the FOSS community (including /.) gets defiant when some company gets miffed that someone wrote a third party client that implements these features.
If we force these companies to adapt their businesses to not do stupid things, perhaps Google should for YouTube as well. Hell, there are plenty of YouTube downloaders out there as well.
Or it could be a bunch of Google apologists willing to overlook anything Google does even though it's the same as what others have done as well , only the latter have been mocked for the very same actions.
Sony has a laptop that does this - it's not a standard Thunderbolt connector (a rather non-standard USB one), but it has Intel graphics while mobile, and you can plug it into the "mobile dock" while at a desk which includes a Blu-Ray drive and high end graphics.
There's a third party manufacturer that's supposed to make a thunderbolt to PCIe dock for Macs so you can put in a nice graphics card in it and do just that as well. Previously they made Expresscard to PCIe docks (Expresscard is PCIe x1 + USB, while I think Thunderbolt allows PCIe 2.0 x2 or so plus displayport).
You know, it's NOT that much more than a good ol' M82A1 or M107 - they are both in the 5-digit range though you can configure the M82 with iron sights to get it to just below that threshold.
And last I checked, most sniper rifles were big and heavy, as well. (The M82's about 30lbs, M107's about 10% lighter).
Now, it's true it takes the "fun" out of the shoot, but it depends what the shoot was about now. If you want to kill an animal humanely for food, it seems like an option to consider.
But I suppose the real argument boils down to the "command line vs. the GUI". People argue it takes the fun out of the sport, while others say it makes them more productive. And in this market, there's nothing that can't say we can't have both. It's like how some people wish to drive sporty cars with manual transmissions while others would pay to get a self-driving car.
Though, one does wonder how well it tracks human targets. It's not supposed to, but it's an interesting consideration some military might consider.
Guns kill. Now, that kill may be done for sport and recreation, necessity, or other reason. The ones doing the sport will probably not buy it. The ones doing it for necessity, might. Who's to say, and why can't we have the option?
The reason for this is simple - the instruction decoder is using a bunch of logic gates to figure out how to drive the various processing bits around, referencing a microcode ROM if necessary. The deal was that most of the opcode space was not completely used up so there are holes where no instruction is officially implemented. Using those encodings will not generate an error (most processors didn't have an illegal instruction error, nor did most microprocessors expect anything more than to crash on error). So these illegal instruction encodings merely tickled the instruction decoder in strange and wonderful ways.
A more modern processor like the Motorola 68K, however, implemented illegal instruction exceptions and thus they have no such features.
And modern processor architectures today often have plenty of gaps in the instruction space - for future instructions or whatnot, and they too generate illegal instruction exceptions. ARM actually defines an ILLEGAL opcode to cause that very exception.
Heck, Microsoft found illegal instructions were fastest for system call handling, much to Intel's bemusement
They're not mutually incompatible.
What you need to realize is that detecting planets is hard - because they're dim compared to their parent stars. So pointing your telescope at a random star may or may not reveal a planet.
Basically, it's a problem because space is big. Really big. And modern theory has it's filled with "dark stuff" (energy or matter - they're equivalent with E=mc^2) of unknown composition.
But once you know something's there, it's easier to analyze it - spectral analysis among other methods.
Basically, finding a planet is looking for a needle in a haystack. Once you found the needle though, you can analyze the crap out of it as you know it exists and where it is.
I looked at my computer, and while it's had an uptime of probably since April's patch tuesday, that service has consumed a grand total of... 1m53s of CPU time.
He never mentions what version of iTunes he's using - perhaps it's still 10.x, which is horrible. iTunes 11 has actually fixed a LOT of stuff and is actually pretty decent and more importantly, fast. It's incredible how fast iTunes is nowadays. I'm not sure what Apple did, but damn it's fixed a lot of stutters, halts, and stalls.
Actually, Android's implementation of MTP is crap. It BARELY works. Hell, on Windows I can easily screw it up if I try to open a folder before it finishes enumerating.
One of our developers actually worked on the Mac trying to get Android to work better (via OSX FUSE). Besides complaining about the crappiness of libusb, he constantly complains about how crappy Android MTP is. Do something that is beyond the basic commands and it'll crash Android.
Hell, everything else I have enumerates folders and files faster than Android - it's damn sluggish because I think the Windows MTP client is being forced to use the old clunky method.
All Google did was get MTP to the point where "it works", but skipped "it works well". I'm surprised Linux MTP works - everyone I've seen had plenty of issues getting MTP and Android talking without issues. It works erratically, at best.
What SDR software is that?
And no, there is a romanized way of entering pIqaD (most people use the KLI version). No OCR required, other than transliterating the characters.
There's one that's been bugging me since 14 or 15 though - autocomplete no longer autocompletes deep URLs - it only goes up to the domain. Which is annoying if you have a particularly favorite long URL (like a search), or access a server on another port (no, it doesn't autocomplete ports, either). Sure, the one you want is there - just down arrow enter, but that's annoying.
Haven't found a way to revert it back to autocomplete from history yet including the full URL.
FF20 added that horrendous download box, for starters...
Of course, you can revert it back to the more sane old download list by setting browser.download.useToolkitUI to TRUE.
It isn't that hard to miss in FF20. Not sure what UI breakage they did in 21, though.
Actually, what would help is removing tax incentives for donations for political purposes. So whether it's a campaign contribution or whatever, it's paid with after tax dollars like you would buy any other thing.
In fact, the tax code is complex enough that in some cases, it's more beneficial to give $10 to a political party (or person) than it is to give that same $10 to a charity! This is direct tax benefits, not just indirect benefits, at that.
Oh, and for a time, Canada had that system - every party had a fixed amount of money which went to each party (something like $1.25 per vote). The Harper Government got rid of it for two reasons - 1) because people don't want to support something they didn't vote for, and 2) government shouldn't be wasting tax money.
1) is obviously false because they voted for them, so it's obvious they wanted to support the party. Your vote is paid for by your taxes so regardless of who you voted for, you're paying for that through your taxes. The only person who would complain is those that didn't vote, which is their fault, really.
2) Well, the real case is "keep the Conservatives in power" because they failed to bring in any other measures like well, get rid of tax incentives (would hurt them), or reduce the donation limit from $5000 per person to $1000 per person (again, would hurt them).
So really, in the end, it's seen as a political ploy.
The problem was EyeToy worked by using background differencing. It took two images - and then differenced them to find the motion. (It used a few more frames than that to determine what was background and what wasn't, but the idea holds).
As you can imagine, this works to a limited extent, but if you move too slowly, the differencing algorithm fails because you get low pass filtered as background. In addition, you can't differentiate between things at different depths, so if something falls behind you, it's treated as your motion.
Kinect adds depth sensing (via IR dots) to add depth, making it easier to determine background from not background, as well as do skeletal tracking and such. This gives the ability to differentiate between background movement and foreground, separate multiple people out, etc,