Look, we've had a good technology that produced fairly precise readings for years. It's called (...drumroll...) "resistive touchscreens". Unfortunately, Apple made capacitive touchscreens trendy & cool, and left those among us who need higher-resolution input handicapped and crippled.
Why, for the love of God, Xenu, and the Flying Spaghetti Monster, can't someone like Otterbox just laminate a resistive touchscreen onto the Defender's screen guard, add some active electronics & an independent small battery, leech USB power when charging to charge itself too, and let us have a damn resistive touchscreen that pairs via Bluetooth (or as a USB OTG HID device) and coexists in parallel with the official touchscreen? No need to screw with the mass-market-dominated plans of Samsung and HTC... by moving the resistive touchscreen to the case & making it bluetooth, it becomes something end users can acquire and add THEMSELVES without having to screw around with official manufacturer support.
Or, if not Otterbox, one of the thousands of companies in China that would otherwise make Otterbox knock-offs... something like this would be trivial to engineer, design, and make functional with Cyanogen (even if it took years to ever make it into stock Android), and it would transform the manufacturer from "another nameless company making shoddy cases for ebay sellers in Shenzhen" to "the company that makes a case that makes Asian-language input easy and functional again".
It's a shame Ben Franklin and Thomas Jefferson couldn't have talked the French into a compromise & gotten them to define 1mm as being exactly 1/24th of an imperial inch. It would have totally solved the problem we have now of common fractional-inch measurements that can't be expressed as rational -- let alone nice -- millimeter values. Even ugly fractions like 19/32 of an inch would have ended up as a nice, binary-friendly value like 14.25mm. You'd have to get down to 64ths of an inch before you had to deal with eighths of a millimeter, and COMMON fractional values (like eighths of an inch) would resolve to 1/8" = 3mm.
Would it have been arbitrary? Sure. But ultimately, SI units ended up being as arbitrary as imperial units anyway, because the original measurements upon which the units were based ended up being wrong (when they estimated the Earth's circumference to derive the meter, they didn't realize the planet is squashed at the poles and bulges slightly at the equator). At the end of the day, a system of measurement whose base unit began as 1/24 inch is no less arbitrary than a system whose base unit is supposed to be one ten-millionth the incorrectly-estimated distance between the equator and North Pole.
Ultimately, there's a reason why traditional units persist -- they tend to have evolved over time into useful quantities with useful demarcations. For answering the question, "how cold (or hot) is it outside", it's hard to imagine a better scale than Fahrenheit. It's not correlated to freezing or boiling, but in real life, water has impurities, and hardly anyone literally lives precisely at sea level. Plus, Fahrenheit (for climate temperature) conveys additional information... subzero isn't just bitterly cold... 0F is roughly the point where it goes from being unpleasantly cold to "frostbite is a very, very real possibility". Ditto for 100F. You can split hairs about how hot 95F is, but when it's 100F (wet bulb), you have to actively protect yourself or you'll risk heatstroke -- even if you're a young athlete in top physical condition.
When you get down to it, Fahrenheit AND Celsius are totally arbitrary scales, and the only absolute in BOTH is "absolute zero". Imperial units might have rods, furlongs, drams, slugs, horsepower, and light years, but those units mainly exist so people who deal with niche quantities all day can have nicely-distributed units that tend towards whole values, and units of meaningful size. A "14 stone" man is more meaningfully overweight compared to one who's "13 stone". In contrast, the difference between 203lb and 207lb is little more than an abstract number.
Assuming, of course, that they don't get FAP'ed within 12 minutes of takeoff, and end up with an entire plane sharing the equivalent of dialup 9600 baud.
The videos had forced interactivity (hence, their dependence upon Flash features that Android didn't support). It would stream, then pause and force you to make a choice, like "Which delicious menu item should our hero Jose order?", then made you watch more, then asked a final question you had to get right to prove you watched the video, or it would make you watch it again. (I heard somebody with a laptop at another table angrily complaining about it)
No, that's the point. The way they implemented it caused Android apps to break on multiple levels. If you were tunneled through a PPTP VPN (to encrypt all of your traffic, and prevent a badly-written app that used http from leaking information about you to someone running Wireshark in the vicinity), it broke everything after ~10 minutes when it decided to block you for not watching an ad. Attempting to launch the web browser at that point didn't help, because the web browser's traffic was tunneled through the ipsec VPN too, and never got its request intercepted and redirected to the ad. Even after you killed the VPN, there was ANOTHER bug that prevented that second ad (and beyond) from running properly, even after you closed the VPN session and tried to relaunch the browser. It tried to display an embedded flash video ad, then either crashed the browser trying to display a partial mis-zooned chunk of the video, or it would run to completion, but not credit you for actually watching the ad.
Basically, its developers tested it with desktop browsers running desktop Flash, and failed to realize that IOS devices can't do Flash at all, and Android devices can only handle a subset of Flash. Judging by the artwork style, dysfunctional operation, Flash-focused design, and restaurant chain, I'm guessing it was probably written by the same team who went on to create the steaming mountain of poo known as "Pizza Hut's Android App" (don't get me started... that app was SO BAD, it cost Pizza Hut my business for MONTHS because I literally couldn't complete an order without the app crashing. It's the only one-star rating I've ever given to an Android app.)
^^^ Oh, I forgot... it also broke non-http-based apps (including ipsec VPNs and SSH), because it would periodically decide to make you watch an ad, and start blocking all traffic from your IP address until you watched one. Except in the meantime, your app is freaking out because it's supposedly connected, but has no apparent connectivity. Oh... and the best part... whatever they were using to serve the ads had a bug that caused the Flash-based ad host to crash when you tried watching ads beyond the first, so once the initial session ran out of time and it decided to make you watch another ad, there was nothing you could do to reconnect and make it work again short of spoofing a different MAC address.
Maybe this is something IETF needs to address, so "free-as-in-no-cash-trading-hands" wifi can at least communicate to OTHER applications that they need to make you watch an ad to avoid having the connection go away. In the meantime, though, I officially regard "free-if-you-watch-the-ads" wifi as a plague that does nothing except cause misery and render the service completely useless.
Please, for the love of God, Xenu, and the Flying Spaghetti Monster, no, not in-flight internet that screws with the stream and inserts its own ads into it, or intercepts random http requests and redirects them to interstitial ads. Taco Bell in South Florida tried that a few months ago, and it broke SO FUCKING MANY Android apps it isn't funny (because the access point's stupid software couldn't tell the difference between a http request for a web page, and a http request made to some web service whose client app is just going to crash and burn if it gets a 302 redirect in a context where the real app would never, ever return one).
The problem in America isn't density or population, it's stupid laws that force HSR into absurd "all or nothing" scenarios where they have to either build brand new dedicated corridor every last inch of the way, or make the trains capable of surviving a head-on collision with a mile-long freight train at the highest speed they travel ANYWHERE along their route... not just the highest speed they'd run at along the shared segment of track.
In Germany, it would be entirely legal to build 200km of brand new HSR track, then run trains 200km at low speed along tracks shared with non-HSR trains and freight, merge onto the new HSR tracks and run at 186mph, then move back onto shared tracks and run another 100km at low speed. If the trains only run at 100km/hour on tracks shared with freight trains, but 300km/hour on exclusive HSR tracks, they only have to be rated for a 100km/hour collision with a freight train. In the US, our stupid laws would require that the passenger train be capable of surviving a 300km/hour (186mph) head-on collision with an 80km/hour (~60mph) freight train EVEN IF the passenger train would never, ever run faster than 100km/hour (80mph) when sharing tracks with freight trains.
The law wasn't written with passenger safety in mind... it was written at the behest of freight railroads in the US to make it nearly impossible for high-speed passenger trains to share tracks with freight trains, period, because freight railroads don't want passenger trains sharing their tracks... ever. Not at high speed OR low speed.
The first attempt was in 2006 (ATTiny2323s, each controlling 4 RGB LEDs, wired together with an unholy amount of wire-wrap wire, with the controller chips soldered "dead-bug style" and wrapped in heatshrink.
The first decent-looking prototype was 2007. The first music-choreographed video was from 2008, with another one from 2009. I haven't made any new music videos since then, but the lights still work fine. My goal for next year is to assemble a bunch more, and redo the controller to use an Android tablet (or old Android phone) and Arduino ADK for the controller.
The design is pretty simple... Atmel ATTiny25, 4-resistor array, RGB LED, and a linear regulator capable of delivering 5v@~100mA from 9-12v. The code is 100% assembly, the serial protocol is bitbanged and vaguely inspired by the way infrared remote controls work. The onboard voltage regulator is so I can use three thin (AWG22) wires instead of doing something brittle/dangerous, like rectified 120v in series, or be forced to use thick wires to keep the voltage drop down.
I learned a lot about serial bus design while developing it. The original design (and in fact, the circuit boards) was for daisy-chained serial (with each module regenerating it), but it ended up working well enough at ~2kbps with everything wired in parallel to do it that way. Daisy-chaining and regenerating bit-by-bit caused bit-stretching and distorted the timing too badly, and trying to regenerate it byte-by-byte caused delays when the string got to be too big (because each light module delayed the 9-bit "byte" for one full byte before passing it along). The other catch with daisy-chaining is the fact that if one light died, everything downstream from it would have died, too. In an ideal world, it would be implemented with multidrop RS422 (485?), but AFAIK, neither Atmel nor anybody else makes a MCU that can output balanced serial without additional driver chips (though I DID contemplate trying to power the chip with "ground" = -2.5v, and "Vcc" = +2.5v, on the theory that it would still act like 0 and 5v, and let me transmit the serial with -2.5v and +2.5v by bitbanging two pins with opposite values; not sure whether it would actually work, though. I don't particularly understand analog electronics very well, least of all things like negative DC acting like "ground" relative to Vcc).
The light modules themselves are fully autonomous, and have their own mostly-complete semi-assembly language (kind of like the quasi-assembly used by the Cosmac Elf/Studio 2, or the TI-99/4A, or other systems of the 8-bit era). Most of the opcodes translate loosely to "transition [now|slowly|quickly] to color $x, then [continue|pause|delay|stall]", but I also have opcodes to implement for/next, subroutines, if/then, random numbers, etc. The lights are addressed by individual ID (192 max without resorting to segmentation) and by row/column (7 rows, 7 columns). In other words, I can direct commands to "All lights in row 6", "all lights in column 5", "All lights in row 3, column 4", or "light number 97". The row/column addressability compensates for the slow serial protocol (2kbps) and long datagrams (9, 18, or 27 bits), as does the fact that I can send opcode-arg-address, then modify it for other lights by just sending arg-address if opcode remains unchanged.
Anyway, enjoy. By the way, the treegee.com domain name has long-since lapsed and is probably a porn site now, so don't bother trying to go there.
Actually, there IS a bigger problem -- they can't (or won't) even tell us the precise algorithm they use to COUNT the number of bytes.
Go ahead, AT&T. Prove me wrong. Tell me how many bytes I'm getting billed for as of this second, allow me to send a single-byte UDP datagram to some arbitrary IP address, then tell me how many bytes my new usage reports.
I can almost guarantee that it's not going to increase by a single byte. It's not even going to increase by 9 bytes (8-byte UDP header + 1-byte payload).
(what follows is a hypothetical conversation with an AT&T employee who has electrodes attached to his genitals that zap him automatically whenever he's caught lying).
(Zap!) What? I'm getting charged for retries? Try again. UDP doesn't do retries. (Zap!) You're counting my single-byte datagram as ~1,500 bytes because that's the size of an ATM MTU? Then why the fuck did you lie and say the cap was "3 gigabytes", instead of "2 million ATM packets of 1,500 byte max length"? Except I know you're still lying, because I did another experiment and sent 10 widely-spaced single-byte UDP datagrams. Even if you counted them as 10 ATM packets, I shouldn't have gotten charged for more than 10 x 1500 bytes (10.5 kilobytes). (ZAP! ZAP! ZAP!) You listed my use as "12 megabytes". WTF? You round up to the nearest megabyte? Every hour? Then re-round everything up to the nearest whole multiple of 10 megabytes every day? So if I send even a single-byte ping at some point during the day, you're going to count it as 10mb of use?
(grabs remote, angrily presses the button over and over until his dick is charred and he's about to pass out from pain)
The point is, if you look at the info that leaks out about how carriers bill for data use, it's blatantly obvious that they're inflating usage on a scale that would get bank officials sent to prison for fraud if a credit card company tried to calculate interest charges the same way. Actually, the way they DO calculate interest isn't far from that, but at least they're required by law to meticulously disclose their precise formula for calculating it. They'd NEVER be allowed to get away with saying shit like, "the way we calculate interest charges is too technical and hard for average consumers to understand, so we just calculate the charges & that's what they are."
Now, admittedly, I wouldn't get so hot and bothered about AT&T's counting methodology IF overage gigabytes either carried over indefinitely, or were prorated at the end of the month... but they aren't. The nanosecond AT&T decides I've used 3,000,000,001 bytes of data, I get hit for another $10, even if there are only 17 seconds left until the month ends. Frankly, I don't mind being charged $10/gig. I do, however, mind getting charged $10 for 32 megabytes of data. Carriers now are allowed to play us both ways... pretending they have no idea how much data we've actually used, and endlessly mumbling about how use is "just an estimate", yet billing us with surgical precision for a huge additional chunk of data use the NANOSECOND they think we've gone over the cap by even a single byte.
All I ask is for honest accounting, and additional use charges that are either prorated or carry over. If carriers tried to say, "You get N minutes of voice airtime, and if you use N+1 minutes, but less than 2N minutes, you'll get billed an additional $100 for excessive voice calls during the month" back when voice minutes actually mattered, they would have been metaphorically hanged from a tree on Capitol Hill. Because, as the GP pointed out, most carriers charge MORE per gigabyte for use beyond the limit than they charge for the bytes under the limit.
For those who don't know what a FPGA is, think of it as being like an ephemeral homebrew ASIC. It's the integrated circuit equivalent of a Star Trek "replicator". FPGAs are usually too expensive to use as compoents in finished commercial projects (at least, mass-market ones), but for prototyping, they're the greatest thing that's ever existed.
Let me put it this way -- Freescale might have discontinued the last of their "real" m68k CPUs and non-Coldfire MCUs, but thanks to the magic of FPGAs, you can synthesize your own whenever you want to ( http://opencores.org/project,tg68 ). With a little tweaking, you can synth a hypothetical 50mhz 68000 that can address 4 gigs of ram, use DDR, and execute "move SR, <ea>" without going down in flames.
With FPGAs, you can make an Amiga, turn it into a Commodore 64, resynth it as an Atari ST with two SID chips hypothetically grafted on, then repurpose it to overlay your own bitmap data on top of HDCP-protected HDMI content (as long as you know how to use Google well enough to find a copy of the leaked HDCP master key).
However, I'd recommend the Xgamestation Mach64 CPLD kit as a first step, instead ( http://www.xgamestation.com/view_product.php?id=40 ). It comes with everything you need, includes a good manual that's well-written and easy to understand, and works flawlessly under Vista & Win7 (including 64-bit) as long as the computer has a real (non-USB) parallel port. AFAIK, it's the ONLY dev board that's targeted towards users who understand programming and have some background in electronics, but nevertheless have some huge gaps in their background knowledge that an EE wouldn't have.
Note: I'm not affiliated with them. Just a really happy and satisfied customer:-)
Basically, programming convenience and bug-avoidance. This is a particularly big deal with languages like PHP that embed server-side code into HTML content -- by the time your code realizes that it needs to store something in session context, the header has already been committed (and probably sent). It's not *completely* impossible to work around, but it's a major pain. It's much easier to just set the cookie and establish the session so it'll be there and ready to use when and if you end up needing it.
In any case, the tantrums some people throw about cookies and "storage space" border upon absurd. The most low-end and ghetto throwaway Android phone money can buy has more ram and storage space than a high-end workstation did ~10 years ago. In some cases, it takes more storage space to store the cookie's filename than it takes to store the cookie's payload. Getting mad about being wantonly-tracked everywhere is one thing, but getting mad because a site is consuming a block of hard drive space or a few cells of flash (whose marginal cost is probably a fraction of what it cost to climate-control the air you just inhaled a moment ago) is silly.
Check out the Mach64 CPLD kit at xgamestation.com.
It requires a computer with a "real" parallel port (non-USB), but pcmcia ports work fine (be careful with ExpressCard... they're probably pci-express, but COULD be usb in disguise), as will docking stations that implement them via pci or pass through the "raw" (hidden) signal lines from the southbridge (like Thinkpads).
It even works with Vista, Win7, and/or 64-bit Windows. I've personally confirmed and tested it (I didn't want to risk having Christmas ruined by a nonworking new favorite toy) on my T61 running Vista/32 with generic cardbus parallel port bought on eBay. It installed & flashed the board without a hitch:-)
It is. It's called "1080i60". 120/240fps TVs use their higher framerate to simulate interlace flicker so they can fade & blend adjacent fields without comb artifacts and BOB blur.
120/240hz (and 100/200hz) TVs were developed as an alternative way to reconcile the curse and pain of interlaced content with the inherently progressive-scan reality of modern TVs. Interpolated higher framerates (for "Soap Opera Effect") was a tacked-on afterthought.
Trivia: a 120hz plasma TV is roughly equivalent to a 240hz LCD, because plasma has its own fading effect and doesn't literally have to bit-bang the whole thing from start to finish the way LCD does.
^^^ Not quite. Kensington's is just a power supply with a few USB-shaped 5v receptacles. Lenovo's is a real, honest-to-god powered USB hub. HUGE difference.
Well, ok... you have a point. I'm kicking myself right now for forgetting the best example of all -- the near-impossibility of getting host/AP/master mode to work properly with ANY USB-based wifi adapter due to timing requirements. You have to either move 99% of the active AP logic to the "interface" end of the USB cable, or wantonly abuse the way USB is supposed to work by firehosing a nonstop stream of nonstop isochronous USB data (with error-correction) in both directions (regardless of actual activity) to eliminate the delay you'd otherwise get if you tried to just poll the interface for activity and react to it (and in the process, render the host PC almost unusable for anything BESIDES monitoring and responding to the network interface).
That said, perhaps a pair of internal mini-PCI slots, and a healthy assortment of CardBUS and ExpressCard slots might be more useful. Then you'd have direct access to the PCI/PCI Express bus, and could implement "real" interfaces of whatever kind you desire (including legacy parallel ports... handy for realtime bitbanged i/o).
How can Instagram casually assume that the uploader even HAS the right to assign republishing rights to them? OK, fine... the TOS requires that uploaders have the rights. We all know that a certain percentage won't comply. How many times does Instagram really want to spin the roulette wheel and risk getting nailed by a lawsuit from someone who owns the copyright on a wrongly-uploaded photo... in a strict-liability jurisdiction with joint and several liability? In English, that means Jim might, under Instagram TOS, be 100% liable for infringement if he uploads a photo and gets Instagram sued when they republish it, but at the end of the day, Jim isn't going to pay that million-dollar lawsuit... Instagram will, because Jim is likely to be judgment-proof, and any halfway-competent attorney could get the judgment to adhere to Instagram regardless of what they might claim.
Not to mention, model releases. If Jim posts pictures taken at a birthday party his child attends, Instagram would legally need releases from every person (or their legal guardian) recognizable in the picture (with a few exceptions, but it's still a minefield).
Did I mention the legal suicide mission of using pics that have anything to DO with kids from Europe? I think in Germany, it's not even legal to use kids in an advertisement for anything, period... consent from fame-whoring parents or not. Or for that matter, the fact that fucked up French copyright law allows you to copyright the image of buildings and structures, even structures that dominate the horizon and are visible from literally miles away (like the Eiffel Tower and the Millau Viaduct), and (in legal theory, at least) make it almost impossible to publish photos taken almost anywhere in Paris (due to the large number of "historically and/or architecturally-significant structures") if they show a complete building facade of one or more buildings in the background? Granted, the French situation is slightly unique, and is used mainly by the French government as a tool for censorship of unflattering and politically-sensitive images, but that's just one country out of hundreds.
There's a reason why big corporations get all of their public photos from companies like Getty Images -- it lets their management and lawyers sleep at night knowing that the copyright clearances and model releases have all been taken care of, and the image vendor itself is big enough to pay any lawsuit that might arise from the photo's licensed use. It's also why some people have had so much fun showing the same clip-art models really getting around, and showing up in everything from ads to "happy employee" photos to patients at STD clinics.
Dual ethernet jacks are kind of silly, just because USB ethernet jacks now cost about $10, and are now almost smaller than a PCMCIA ethernet jack dongle ALONE used to be 15 years ago.
On the other hand, a real sliding switch to physically cut the connection to the speakers, so you can safely boot up someplace where you CAN NOT have it making noise... well, that's another matter entirely, because that's NOT something an end user can go graft onto the system himself after purchase (at least, not without completely destroying the laptop's resale value).
Or better yet, stick a powered USB hub into the power brick like Lenovo does with one of their power supplies (http://www.amazon.com/Lenovo-57Y4600-ThinkPad-65W-Adapter/dp/B0044KR91U) . Sigh. Dammit Lenovo, can you PLEASE make this in 95-watt size? (for those who don't know, 65W is enough to run a Thinkpad OR charge its battery, but not both at once. To charge AND run simultaneously, you need 95w).
Combine a USB3 hub with beefy power brick big enough to supply the laptop itself with 95w so we can use it to power a bright second travel monitor, and there'd be just two words to describe it: Flawless Perfection.
> Most people can't see a difference in rates above 30fps and pretty much nobody can distinguish fps over 60 fps.
Bullshit. You most certainly CAN see a difference, particularly when there's high-resolution, high-contrast detail with fast movement across the screen. In fact, high-framerate video has its own "uncanny valley" problem (above a certain framerate, generally in the neighborhood of ~300fps, hyperfluid 2-dimensional video becomes disorienting and vertigo-inducing, because your brain can't reconcile the seemingly-lifelike motion with its lack of depth).
That said, AFAIK, nothing today actually USES 112, and no vertical service codes BEGIN with 112, so it could technically be used as an emergency number. Nevertheless, I see one of three things happening:
Scenario 1: ITU declares that 112 and 911 are emergency numbers everywhere, except in countries where it would screw up the phone system. The US yawns and says, "OK, we'll make 112 work here as an alias for 911".
Scenario 2: ITU declares that 112 must be the One and Only emergency number worldwide, and that countries must stop using 911 entirely. The US tells the ITU to go to hell. Canadians quietly do the same in less heated terms, but implement 112 as a fallback second emergency number anyway. The FCC plans to quietly do the same, then some halfwit elected official gets the stupid idea of making it the nationwide "patriot hotline" number to report suspicious un-American activity to DHS, and the whole thing goes down in flames.
Scenario 3: ITU declares that 112 is mandatory and 911 is optional. The US grudgingly agrees, asks carriers to implement it, and sets a compliance deadline of 2025.
And what's worse, they've replicated that goddamn failure every single time they go back to make another try at $1 coins, because they don't want vending machine owners to have to deal with two different "$1" coins.
Replace dollar bills with $1 coins that look like Canada's $2 coins (bi-metallic coin-in-a-coin). Make them TOTALLY unlike the nickel, dime, and quarter. Say, a coin the size of a Kennedy 50-cent piece, consisting of nickel alloy surrounding a copper disc. A copper disc that's the same thickness as the outer nickel ring, but not quite flush, so it sticks out on one side, and sinks into the other (nice for stacking and identifying them by feel). Stamp Washington's head into the copper disc on one side, Lincoln's head into the copper disc on the other, and embed the "ONE"-related text and stuff into the nickel rings.
OK, having created a totally cool new coin, there's an easy way to abolish pennies and dollar bills from circulation almost overnight:
1. Pass a law making it unambiguously legal, when giving change in actual US currency, to round.01 and.02 down to.00, and.03 and.04 up to.05. Electronic transactions would still have to be calculated to the nearest cent. Declare that pennies are still legal tender, but REQUIRE rounding to the nearest 5 cents EVEN WHEN payment or change includes pennies.
2. Have the Federal Reserve charge banks 1.11 cents per penny (55 cents per 50), and $1.10 per dollar bill, and allow them to pass along the higher charge to customers -- but ONLY if the customer DEMANDS payment/change in pennies/dollar bills, and ONLY when handing out crisp, virgin, brand new dollar bills and pennies newly-acquired from the mint. In other words, only the "first purchaser" of the dollar bill or penny is legally allowed to pass along the surcharge (otherwise, they'd recirculate tattered old bills forever to wring the extra fee out of customers forever).
By making pennies and dollar bills slightly more expensive than their nominal value, it will eliminate them both as viable units of currency almost overnight. Some people will hoard them, most will decline to pay extra for them, but either way, they'll be dead as active currency. Don't believe me? OK, when's the last time YOU paid for something with a $2 bill, or a 50-cent piece? Or even SAW a $2 bill or 50-cent piece being used as payment for something... as opposed to a gift, or in a jar, or anywhere besides somebody's wallet? The moment people think a bill or coin is worth more than its face value, they quit using it, because now it's "too valuable" to spend. Even $2 bills, whose acquisition price at the bank RIGHT now is... $2 (not sure about 50-cent pieces).
Rather than deal with "old" and "new" dollar bills, banks would quit bothering with "old" $1 bills -- they'd just send them straight back to the Fed, and sell brand new ones for $1.10 apiece so they wouldn't have to deal with two categories of dollar bills. Businesses wouldn't buy them, because (unlike banks), they WOULDN'T be allowed to pass along the surcharge, so it would be money straight out of their pockets. And consumers would be in a frenzy to sell each other $1 bills for more than $1, but less than $1.10, until they ended up in the hands of people who'd think they were as valuable as gold & hoard them forever. Either way, 99.99999% of dollar bills would be out of active circulation within a matter of weeks.
> You can knock out a lot of your gen ed *before* going to the better, more expensive four-year college where you intend to complete your degree.
Assuming that you, unlike roughly 97% of your peers, have the *slightest* idea what you actually want to do for the next 40 years when you're 19 or 20. That's the whole problem with the "get your general education credits out of the way" plan of community colleges... by packing all of your "major" courses into two years, and by extension DEFERRING nearly all of them until years 3 and 4, you've raised the stakes considerably, and made changing your mind about your major a much, much more disruptive and expensive process.
If you wait until the fall semester of your third year to take your first real courses in your major, then discover you don't actually LIKE your major after all, you've just *incinerated" at least one semester... maybe two. In contrast, if you've taken the first 4 courses in your major by the middle of your sophomore year, then discover you don't really like it after all, you've only REALLY wasted one or two of those classes, because the others ended up satisfying your general-ed requiremends anyway.
That's why most private colleges and universities encourage you to spread out your general-ed classes, and to begin taking your "major" classes early and often, and why they encourage you to satisfy many of your "general ed" classes with classes that do double-duty as the "intro/survey/101" courses for other majors. They have every incentive to help you graduate in 4 years... they're expensive, they know it, and they know there's a nontrivial chance you might not graduate at all if they seriously derail you. They know that 70% of their students change their majors at least once before year 3, and most of them have had more than a hundred years to refine the formula and get it right.
The generic community-college scenario only really works for two groups of students... those whose only goal is "a degree", regardless of what it might be in, and those for whom community college is a second chance to shine, catch up, and redeem themselves. A student who's already at the top of his high school class and a shoo-in at just about any university is basically just wasting his time, and is actually INCREASING his odds of stumbling and losing his way before graduation.
The fact is, the "2+2" formula just doesn't work for the majority of students.
Galaxy S3 is a bit more complicated and expensive due to its 11-pin semi(?)-proprietary MHL connectors ("SaMHL"). The cheapest 11-pin SaMHL cable I'm aware of is ~$15, and the word at XDA is that it's totally luck of the draw... about half work fine, and half fall somewhere between "occasionally-flaky" and "just plain doesn't work". The cheapest cables known to work at least as reliably as Samsung's official ones start at around $25-30.
Yes, you can spend a few bucks on an 11-pin to 5-pin MHL adapter and use a cheap generic 5-pin MHL cable with reasonable success, but THEN you're putting a HUGE amount of lever-induced stress on the S3's USB port, and just begging for a cracked mainboard or weakened solder joint. Upgrade the adapter to one with at least an inch or two of flexible cable between the plug and the socket, and you're back up into the $25+ range and might as well get a good non-Samsung cable made for 11-pin SaMHL from the start.
The situation will obviously improve in a year or two since all Samsung phones (and probably others) will be using 11-pin MHL going forward... but right now, there's still a 100% or more price premium for S3-compatible MHL cables.
Slashdot Mirror
Look, we've had a good technology that produced fairly precise readings for years. It's called (...drumroll...) "resistive touchscreens". Unfortunately, Apple made capacitive touchscreens trendy & cool, and left those among us who need higher-resolution input handicapped and crippled.
Why, for the love of God, Xenu, and the Flying Spaghetti Monster, can't someone like Otterbox just laminate a resistive touchscreen onto the Defender's screen guard, add some active electronics & an independent small battery, leech USB power when charging to charge itself too, and let us have a damn resistive touchscreen that pairs via Bluetooth (or as a USB OTG HID device) and coexists in parallel with the official touchscreen? No need to screw with the mass-market-dominated plans of Samsung and HTC... by moving the resistive touchscreen to the case & making it bluetooth, it becomes something end users can acquire and add THEMSELVES without having to screw around with official manufacturer support.
Or, if not Otterbox, one of the thousands of companies in China that would otherwise make Otterbox knock-offs... something like this would be trivial to engineer, design, and make functional with Cyanogen (even if it took years to ever make it into stock Android), and it would transform the manufacturer from "another nameless company making shoddy cases for ebay sellers in Shenzhen" to "the company that makes a case that makes Asian-language input easy and functional again".
It's a shame Ben Franklin and Thomas Jefferson couldn't have talked the French into a compromise & gotten them to define 1mm as being exactly 1/24th of an imperial inch. It would have totally solved the problem we have now of common fractional-inch measurements that can't be expressed as rational -- let alone nice -- millimeter values. Even ugly fractions like 19/32 of an inch would have ended up as a nice, binary-friendly value like 14.25mm. You'd have to get down to 64ths of an inch before you had to deal with eighths of a millimeter, and COMMON fractional values (like eighths of an inch) would resolve to 1/8" = 3mm.
Would it have been arbitrary? Sure. But ultimately, SI units ended up being as arbitrary as imperial units anyway, because the original measurements upon which the units were based ended up being wrong (when they estimated the Earth's circumference to derive the meter, they didn't realize the planet is squashed at the poles and bulges slightly at the equator). At the end of the day, a system of measurement whose base unit began as 1/24 inch is no less arbitrary than a system whose base unit is supposed to be one ten-millionth the incorrectly-estimated distance between the equator and North Pole.
Ultimately, there's a reason why traditional units persist -- they tend to have evolved over time into useful quantities with useful demarcations. For answering the question, "how cold (or hot) is it outside", it's hard to imagine a better scale than Fahrenheit. It's not correlated to freezing or boiling, but in real life, water has impurities, and hardly anyone literally lives precisely at sea level. Plus, Fahrenheit (for climate temperature) conveys additional information... subzero isn't just bitterly cold... 0F is roughly the point where it goes from being unpleasantly cold to "frostbite is a very, very real possibility". Ditto for 100F. You can split hairs about how hot 95F is, but when it's 100F (wet bulb), you have to actively protect yourself or you'll risk heatstroke -- even if you're a young athlete in top physical condition.
When you get down to it, Fahrenheit AND Celsius are totally arbitrary scales, and the only absolute in BOTH is "absolute zero". Imperial units might have rods, furlongs, drams, slugs, horsepower, and light years, but those units mainly exist so people who deal with niche quantities all day can have nicely-distributed units that tend towards whole values, and units of meaningful size. A "14 stone" man is more meaningfully overweight compared to one who's "13 stone". In contrast, the difference between 203lb and 207lb is little more than an abstract number.
Assuming, of course, that they don't get FAP'ed within 12 minutes of takeoff, and end up with an entire plane sharing the equivalent of dialup 9600 baud.
The videos had forced interactivity (hence, their dependence upon Flash features that Android didn't support). It would stream, then pause and force you to make a choice, like "Which delicious menu item should our hero Jose order?", then made you watch more, then asked a final question you had to get right to prove you watched the video, or it would make you watch it again. (I heard somebody with a laptop at another table angrily complaining about it)
> HTTPS for everything. Problem solved.
No, that's the point. The way they implemented it caused Android apps to break on multiple levels. If you were tunneled through a PPTP VPN (to encrypt all of your traffic, and prevent a badly-written app that used http from leaking information about you to someone running Wireshark in the vicinity), it broke everything after ~10 minutes when it decided to block you for not watching an ad. Attempting to launch the web browser at that point didn't help, because the web browser's traffic was tunneled through the ipsec VPN too, and never got its request intercepted and redirected to the ad. Even after you killed the VPN, there was ANOTHER bug that prevented that second ad (and beyond) from running properly, even after you closed the VPN session and tried to relaunch the browser. It tried to display an embedded flash video ad, then either crashed the browser trying to display a partial mis-zooned chunk of the video, or it would run to completion, but not credit you for actually watching the ad.
Basically, its developers tested it with desktop browsers running desktop Flash, and failed to realize that IOS devices can't do Flash at all, and Android devices can only handle a subset of Flash. Judging by the artwork style, dysfunctional operation, Flash-focused design, and restaurant chain, I'm guessing it was probably written by the same team who went on to create the steaming mountain of poo known as "Pizza Hut's Android App" (don't get me started... that app was SO BAD, it cost Pizza Hut my business for MONTHS because I literally couldn't complete an order without the app crashing. It's the only one-star rating I've ever given to an Android app.)
^^^ Oh, I forgot... it also broke non-http-based apps (including ipsec VPNs and SSH), because it would periodically decide to make you watch an ad, and start blocking all traffic from your IP address until you watched one. Except in the meantime, your app is freaking out because it's supposedly connected, but has no apparent connectivity. Oh... and the best part... whatever they were using to serve the ads had a bug that caused the Flash-based ad host to crash when you tried watching ads beyond the first, so once the initial session ran out of time and it decided to make you watch another ad, there was nothing you could do to reconnect and make it work again short of spoofing a different MAC address.
Maybe this is something IETF needs to address, so "free-as-in-no-cash-trading-hands" wifi can at least communicate to OTHER applications that they need to make you watch an ad to avoid having the connection go away. In the meantime, though, I officially regard "free-if-you-watch-the-ads" wifi as a plague that does nothing except cause misery and render the service completely useless.
Please, for the love of God, Xenu, and the Flying Spaghetti Monster, no, not in-flight internet that screws with the stream and inserts its own ads into it, or intercepts random http requests and redirects them to interstitial ads. Taco Bell in South Florida tried that a few months ago, and it broke SO FUCKING MANY Android apps it isn't funny (because the access point's stupid software couldn't tell the difference between a http request for a web page, and a http request made to some web service whose client app is just going to crash and burn if it gets a 302 redirect in a context where the real app would never, ever return one).
The problem in America isn't density or population, it's stupid laws that force HSR into absurd "all or nothing" scenarios where they have to either build brand new dedicated corridor every last inch of the way, or make the trains capable of surviving a head-on collision with a mile-long freight train at the highest speed they travel ANYWHERE along their route... not just the highest speed they'd run at along the shared segment of track.
In Germany, it would be entirely legal to build 200km of brand new HSR track, then run trains 200km at low speed along tracks shared with non-HSR trains and freight, merge onto the new HSR tracks and run at 186mph, then move back onto shared tracks and run another 100km at low speed. If the trains only run at 100km/hour on tracks shared with freight trains, but 300km/hour on exclusive HSR tracks, they only have to be rated for a 100km/hour collision with a freight train. In the US, our stupid laws would require that the passenger train be capable of surviving a 300km/hour (186mph) head-on collision with an 80km/hour (~60mph) freight train EVEN IF the passenger train would never, ever run faster than 100km/hour (80mph) when sharing tracks with freight trains.
The law wasn't written with passenger safety in mind... it was written at the behest of freight railroads in the US to make it nearly impossible for high-speed passenger trains to share tracks with freight trains, period, because freight railroads don't want passenger trains sharing their tracks... ever. Not at high speed OR low speed.
My string of networked RGB Christmas lights: http://www.youtube.com/treegeergb
The first attempt was in 2006 (ATTiny2323s, each controlling 4 RGB LEDs, wired together with an unholy amount of wire-wrap wire, with the controller chips soldered "dead-bug style" and wrapped in heatshrink.
The first decent-looking prototype was 2007. The first music-choreographed video was from 2008, with another one from 2009. I haven't made any new music videos since then, but the lights still work fine. My goal for next year is to assemble a bunch more, and redo the controller to use an Android tablet (or old Android phone) and Arduino ADK for the controller.
The design is pretty simple... Atmel ATTiny25, 4-resistor array, RGB LED, and a linear regulator capable of delivering 5v@~100mA from 9-12v. The code is 100% assembly, the serial protocol is bitbanged and vaguely inspired by the way infrared remote controls work. The onboard voltage regulator is so I can use three thin (AWG22) wires instead of doing something brittle/dangerous, like rectified 120v in series, or be forced to use thick wires to keep the voltage drop down.
I learned a lot about serial bus design while developing it. The original design (and in fact, the circuit boards) was for daisy-chained serial (with each module regenerating it), but it ended up working well enough at ~2kbps with everything wired in parallel to do it that way. Daisy-chaining and regenerating bit-by-bit caused bit-stretching and distorted the timing too badly, and trying to regenerate it byte-by-byte caused delays when the string got to be too big (because each light module delayed the 9-bit "byte" for one full byte before passing it along). The other catch with daisy-chaining is the fact that if one light died, everything downstream from it would have died, too. In an ideal world, it would be implemented with multidrop RS422 (485?), but AFAIK, neither Atmel nor anybody else makes a MCU that can output balanced serial without additional driver chips (though I DID contemplate trying to power the chip with "ground" = -2.5v, and "Vcc" = +2.5v, on the theory that it would still act like 0 and 5v, and let me transmit the serial with -2.5v and +2.5v by bitbanging two pins with opposite values; not sure whether it would actually work, though. I don't particularly understand analog electronics very well, least of all things like negative DC acting like "ground" relative to Vcc).
The light modules themselves are fully autonomous, and have their own mostly-complete semi-assembly language (kind of like the quasi-assembly used by the Cosmac Elf/Studio 2, or the TI-99/4A, or other systems of the 8-bit era). Most of the opcodes translate loosely to "transition [now|slowly|quickly] to color $x, then [continue|pause|delay|stall]", but I also have opcodes to implement for/next, subroutines, if/then, random numbers, etc. The lights are addressed by individual ID (192 max without resorting to segmentation) and by row/column (7 rows, 7 columns). In other words, I can direct commands to "All lights in row 6", "all lights in column 5", "All lights in row 3, column 4", or "light number 97". The row/column addressability compensates for the slow serial protocol (2kbps) and long datagrams (9, 18, or 27 bits), as does the fact that I can send opcode-arg-address, then modify it for other lights by just sending arg-address if opcode remains unchanged.
Anyway, enjoy. By the way, the treegee.com domain name has long-since lapsed and is probably a porn site now, so don't bother trying to go there.
Actually, there IS a bigger problem -- they can't (or won't) even tell us the precise algorithm they use to COUNT the number of bytes.
Go ahead, AT&T. Prove me wrong. Tell me how many bytes I'm getting billed for as of this second, allow me to send a single-byte UDP datagram to some arbitrary IP address, then tell me how many bytes my new usage reports.
I can almost guarantee that it's not going to increase by a single byte. It's not even going to increase by 9 bytes (8-byte UDP header + 1-byte payload).
(what follows is a hypothetical conversation with an AT&T employee who has electrodes attached to his genitals that zap him automatically whenever he's caught lying).
(Zap!) What? I'm getting charged for retries? Try again. UDP doesn't do retries.
(Zap!) You're counting my single-byte datagram as ~1,500 bytes because that's the size of an ATM MTU? Then why the fuck did you lie and say the cap was "3 gigabytes", instead of "2 million ATM packets of 1,500 byte max length"? Except I know you're still lying, because I did another experiment and sent 10 widely-spaced single-byte UDP datagrams. Even if you counted them as 10 ATM packets, I shouldn't have gotten charged for more than 10 x 1500 bytes (10.5 kilobytes).
(ZAP! ZAP! ZAP!) You listed my use as "12 megabytes". WTF? You round up to the nearest megabyte? Every hour? Then re-round everything up to the nearest whole multiple of 10 megabytes every day? So if I send even a single-byte ping at some point during the day, you're going to count it as 10mb of use?
(grabs remote, angrily presses the button over and over until his dick is charred and he's about to pass out from pain)
The point is, if you look at the info that leaks out about how carriers bill for data use, it's blatantly obvious that they're inflating usage on a scale that would get bank officials sent to prison for fraud if a credit card company tried to calculate interest charges the same way. Actually, the way they DO calculate interest isn't far from that, but at least they're required by law to meticulously disclose their precise formula for calculating it. They'd NEVER be allowed to get away with saying shit like, "the way we calculate interest charges is too technical and hard for average consumers to understand, so we just calculate the charges & that's what they are."
Now, admittedly, I wouldn't get so hot and bothered about AT&T's counting methodology IF overage gigabytes either carried over indefinitely, or were prorated at the end of the month... but they aren't. The nanosecond AT&T decides I've used 3,000,000,001 bytes of data, I get hit for another $10, even if there are only 17 seconds left until the month ends. Frankly, I don't mind being charged $10/gig. I do, however, mind getting charged $10 for 32 megabytes of data. Carriers now are allowed to play us both ways... pretending they have no idea how much data we've actually used, and endlessly mumbling about how use is "just an estimate", yet billing us with surgical precision for a huge additional chunk of data use the NANOSECOND they think we've gone over the cap by even a single byte.
All I ask is for honest accounting, and additional use charges that are either prorated or carry over. If carriers tried to say, "You get N minutes of voice airtime, and if you use N+1 minutes, but less than 2N minutes, you'll get billed an additional $100 for excessive voice calls during the month" back when voice minutes actually mattered, they would have been metaphorically hanged from a tree on Capitol Hill. Because, as the GP pointed out, most carriers charge MORE per gigabyte for use beyond the limit than they charge for the bytes under the limit.
^^^ Yes.
For those who don't know what a FPGA is, think of it as being like an ephemeral homebrew ASIC. It's the integrated circuit equivalent of a Star Trek "replicator". FPGAs are usually too expensive to use as compoents in finished commercial projects (at least, mass-market ones), but for prototyping, they're the greatest thing that's ever existed.
Let me put it this way -- Freescale might have discontinued the last of their "real" m68k CPUs and non-Coldfire MCUs, but thanks to the magic of FPGAs, you can synthesize your own whenever you want to ( http://opencores.org/project,tg68 ). With a little tweaking, you can synth a hypothetical 50mhz 68000 that can address 4 gigs of ram, use DDR, and execute "move SR, <ea>" without going down in flames.
With FPGAs, you can make an Amiga, turn it into a Commodore 64, resynth it as an Atari ST with two SID chips hypothetically grafted on, then repurpose it to overlay your own bitmap data on top of HDCP-protected HDMI content (as long as you know how to use Google well enough to find a copy of the leaked HDCP master key).
However, I'd recommend the Xgamestation Mach64 CPLD kit as a first step, instead ( http://www.xgamestation.com/view_product.php?id=40 ). It comes with everything you need, includes a good manual that's well-written and easy to understand, and works flawlessly under Vista & Win7 (including 64-bit) as long as the computer has a real (non-USB) parallel port. AFAIK, it's the ONLY dev board that's targeted towards users who understand programming and have some background in electronics, but nevertheless have some huge gaps in their background knowledge that an EE wouldn't have.
Note: I'm not affiliated with them. Just a really happy and satisfied customer :-)
Wait until Saturday, when somebody goes up to the 2k12 apocalypse-anticipators and says, "Hey, don't feel bad... it's not the end of the world"
(...roflmao... "not the end of the world" *cough*, *cough*).
Basically, programming convenience and bug-avoidance. This is a particularly big deal with languages like PHP that embed server-side code into HTML content -- by the time your code realizes that it needs to store something in session context, the header has already been committed (and probably sent). It's not *completely* impossible to work around, but it's a major pain. It's much easier to just set the cookie and establish the session so it'll be there and ready to use when and if you end up needing it.
In any case, the tantrums some people throw about cookies and "storage space" border upon absurd. The most low-end and ghetto throwaway Android phone money can buy has more ram and storage space than a high-end workstation did ~10 years ago. In some cases, it takes more storage space to store the cookie's filename than it takes to store the cookie's payload. Getting mad about being wantonly-tracked everywhere is one thing, but getting mad because a site is consuming a block of hard drive space or a few cells of flash (whose marginal cost is probably a fraction of what it cost to climate-control the air you just inhaled a moment ago) is silly.
Check out the Mach64 CPLD kit at xgamestation.com.
It requires a computer with a "real" parallel port (non-USB), but pcmcia ports work fine (be careful with ExpressCard... they're probably pci-express, but COULD be usb in disguise), as will docking stations that implement them via pci or pass through the "raw" (hidden) signal lines from the southbridge (like Thinkpads).
It even works with Vista, Win7, and/or 64-bit Windows. I've personally confirmed and tested it (I didn't want to risk having Christmas ruined by a nonworking new favorite toy) on my T61 running Vista/32 with generic cardbus parallel port bought on eBay. It installed & flashed the board without a hitch :-)
> AND all the programming is also in that format
It is. It's called "1080i60". 120/240fps TVs use their higher framerate to simulate interlace flicker so they can fade & blend adjacent fields without comb artifacts and BOB blur.
120/240hz (and 100/200hz) TVs were developed as an alternative way to reconcile the curse and pain of interlaced content with the inherently progressive-scan reality of modern TVs. Interpolated higher framerates (for "Soap Opera Effect") was a tacked-on afterthought.
Trivia: a 120hz plasma TV is roughly equivalent to a 240hz LCD, because plasma has its own fading effect and doesn't literally have to bit-bang the whole thing from start to finish the way LCD does.
^^^ Not quite. Kensington's is just a power supply with a few USB-shaped 5v receptacles. Lenovo's is a real, honest-to-god powered USB hub. HUGE difference.
Well, ok... you have a point. I'm kicking myself right now for forgetting the best example of all -- the near-impossibility of getting host/AP/master mode to work properly with ANY USB-based wifi adapter due to timing requirements. You have to either move 99% of the active AP logic to the "interface" end of the USB cable, or wantonly abuse the way USB is supposed to work by firehosing a nonstop stream of nonstop isochronous USB data (with error-correction) in both directions (regardless of actual activity) to eliminate the delay you'd otherwise get if you tried to just poll the interface for activity and react to it (and in the process, render the host PC almost unusable for anything BESIDES monitoring and responding to the network interface).
That said, perhaps a pair of internal mini-PCI slots, and a healthy assortment of CardBUS and ExpressCard slots might be more useful. Then you'd have direct access to the PCI/PCI Express bus, and could implement "real" interfaces of whatever kind you desire (including legacy parallel ports... handy for realtime bitbanged i/o).
How can Instagram casually assume that the uploader even HAS the right to assign republishing rights to them? OK, fine... the TOS requires that uploaders have the rights. We all know that a certain percentage won't comply. How many times does Instagram really want to spin the roulette wheel and risk getting nailed by a lawsuit from someone who owns the copyright on a wrongly-uploaded photo... in a strict-liability jurisdiction with joint and several liability? In English, that means Jim might, under Instagram TOS, be 100% liable for infringement if he uploads a photo and gets Instagram sued when they republish it, but at the end of the day, Jim isn't going to pay that million-dollar lawsuit... Instagram will, because Jim is likely to be judgment-proof, and any halfway-competent attorney could get the judgment to adhere to Instagram regardless of what they might claim.
Not to mention, model releases. If Jim posts pictures taken at a birthday party his child attends, Instagram would legally need releases from every person (or their legal guardian) recognizable in the picture (with a few exceptions, but it's still a minefield).
Did I mention the legal suicide mission of using pics that have anything to DO with kids from Europe? I think in Germany, it's not even legal to use kids in an advertisement for anything, period... consent from fame-whoring parents or not. Or for that matter, the fact that fucked up French copyright law allows you to copyright the image of buildings and structures, even structures that dominate the horizon and are visible from literally miles away (like the Eiffel Tower and the Millau Viaduct), and (in legal theory, at least) make it almost impossible to publish photos taken almost anywhere in Paris (due to the large number of "historically and/or architecturally-significant structures") if they show a complete building facade of one or more buildings in the background? Granted, the French situation is slightly unique, and is used mainly by the French government as a tool for censorship of unflattering and politically-sensitive images, but that's just one country out of hundreds.
There's a reason why big corporations get all of their public photos from companies like Getty Images -- it lets their management and lawyers sleep at night knowing that the copyright clearances and model releases have all been taken care of, and the image vendor itself is big enough to pay any lawsuit that might arise from the photo's licensed use. It's also why some people have had so much fun showing the same clip-art models really getting around, and showing up in everything from ads to "happy employee" photos to patients at STD clinics.
Dual ethernet jacks are kind of silly, just because USB ethernet jacks now cost about $10, and are now almost smaller than a PCMCIA ethernet jack dongle ALONE used to be 15 years ago.
On the other hand, a real sliding switch to physically cut the connection to the speakers, so you can safely boot up someplace where you CAN NOT have it making noise... well, that's another matter entirely, because that's NOT something an end user can go graft onto the system himself after purchase (at least, not without completely destroying the laptop's resale value).
Or better yet, stick a powered USB hub into the power brick like Lenovo does with one of their power supplies (http://www.amazon.com/Lenovo-57Y4600-ThinkPad-65W-Adapter/dp/B0044KR91U) . Sigh. Dammit Lenovo, can you PLEASE make this in 95-watt size? (for those who don't know, 65W is enough to run a Thinkpad OR charge its battery, but not both at once. To charge AND run simultaneously, you need 95w).
Combine a USB3 hub with beefy power brick big enough to supply the laptop itself with 95w so we can use it to power a bright second travel monitor, and there'd be just two words to describe it: Flawless Perfection.
> Most people can't see a difference in rates above 30fps and pretty much nobody can distinguish fps over 60 fps.
Bullshit. You most certainly CAN see a difference, particularly when there's high-resolution, high-contrast detail with fast movement across the screen. In fact, high-framerate video has its own "uncanny valley" problem (above a certain framerate, generally in the neighborhood of ~300fps, hyperfluid 2-dimensional video becomes disorienting and vertigo-inducing, because your brain can't reconcile the seemingly-lifelike motion with its lack of depth).
> the only time you'd start with a 1 is when dialing the NANP country code followed by an area code,
> or when dialing the emergency services.
Not quite. You forgot about 11-prefixed vertical service codes ( http://en.wikipedia.org/wiki/Vertical_service_code ).
That said, AFAIK, nothing today actually USES 112, and no vertical service codes BEGIN with 112, so it could technically be used as an emergency number. Nevertheless, I see one of three things happening:
Scenario 1: ITU declares that 112 and 911 are emergency numbers everywhere, except in countries where it would screw up the phone system. The US yawns and says, "OK, we'll make 112 work here as an alias for 911".
Scenario 2: ITU declares that 112 must be the One and Only emergency number worldwide, and that countries must stop using 911 entirely. The US tells the ITU to go to hell. Canadians quietly do the same in less heated terms, but implement 112 as a fallback second emergency number anyway. The FCC plans to quietly do the same, then some halfwit elected official gets the stupid idea of making it the nationwide "patriot hotline" number to report suspicious un-American activity to DHS, and the whole thing goes down in flames.
Scenario 3: ITU declares that 112 is mandatory and 911 is optional. The US grudgingly agrees, asks carriers to implement it, and sets a compliance deadline of 2025.
And what's worse, they've replicated that goddamn failure every single time they go back to make another try at $1 coins, because they don't want vending machine owners to have to deal with two different "$1" coins.
Replace dollar bills with $1 coins that look like Canada's $2 coins (bi-metallic coin-in-a-coin). Make them TOTALLY unlike the nickel, dime, and quarter. Say, a coin the size of a Kennedy 50-cent piece, consisting of nickel alloy surrounding a copper disc. A copper disc that's the same thickness as the outer nickel ring, but not quite flush, so it sticks out on one side, and sinks into the other (nice for stacking and identifying them by feel). Stamp Washington's head into the copper disc on one side, Lincoln's head into the copper disc on the other, and embed the "ONE"-related text and stuff into the nickel rings.
OK, having created a totally cool new coin, there's an easy way to abolish pennies and dollar bills from circulation almost overnight:
1. Pass a law making it unambiguously legal, when giving change in actual US currency, to round .01 and .02 down to .00, and .03 and .04 up to .05. Electronic transactions would still have to be calculated to the nearest cent. Declare that pennies are still legal tender, but REQUIRE rounding to the nearest 5 cents EVEN WHEN payment or change includes pennies.
2. Have the Federal Reserve charge banks 1.11 cents per penny (55 cents per 50), and $1.10 per dollar bill, and allow them to pass along the higher charge to customers -- but ONLY if the customer DEMANDS payment/change in pennies/dollar bills, and ONLY when handing out crisp, virgin, brand new dollar bills and pennies newly-acquired from the mint. In other words, only the "first purchaser" of the dollar bill or penny is legally allowed to pass along the surcharge (otherwise, they'd recirculate tattered old bills forever to wring the extra fee out of customers forever).
By making pennies and dollar bills slightly more expensive than their nominal value, it will eliminate them both as viable units of currency almost overnight. Some people will hoard them, most will decline to pay extra for them, but either way, they'll be dead as active currency. Don't believe me? OK, when's the last time YOU paid for something with a $2 bill, or a 50-cent piece? Or even SAW a $2 bill or 50-cent piece being used as payment for something... as opposed to a gift, or in a jar, or anywhere besides somebody's wallet? The moment people think a bill or coin is worth more than its face value, they quit using it, because now it's "too valuable" to spend. Even $2 bills, whose acquisition price at the bank RIGHT now is... $2 (not sure about 50-cent pieces).
Rather than deal with "old" and "new" dollar bills, banks would quit bothering with "old" $1 bills -- they'd just send them straight back to the Fed, and sell brand new ones for $1.10 apiece so they wouldn't have to deal with two categories of dollar bills. Businesses wouldn't buy them, because (unlike banks), they WOULDN'T be allowed to pass along the surcharge, so it would be money straight out of their pockets. And consumers would be in a frenzy to sell each other $1 bills for more than $1, but less than $1.10, until they ended up in the hands of people who'd think they were as valuable as gold & hoard them forever. Either way, 99.99999% of dollar bills would be out of active circulation within a matter of weeks.
> You can knock out a lot of your gen ed *before* going to the better, more expensive four-year college where you intend to complete your degree.
Assuming that you, unlike roughly 97% of your peers, have the *slightest* idea what you actually want to do for the next 40 years when you're 19 or 20. That's the whole problem with the "get your general education credits out of the way" plan of community colleges... by packing all of your "major" courses into two years, and by extension DEFERRING nearly all of them until years 3 and 4, you've raised the stakes considerably, and made changing your mind about your major a much, much more disruptive and expensive process.
If you wait until the fall semester of your third year to take your first real courses in your major, then discover you don't actually LIKE your major after all, you've just *incinerated" at least one semester... maybe two. In contrast, if you've taken the first 4 courses in your major by the middle of your sophomore year, then discover you don't really like it after all, you've only REALLY wasted one or two of those classes, because the others ended up satisfying your general-ed requiremends anyway.
That's why most private colleges and universities encourage you to spread out your general-ed classes, and to begin taking your "major" classes early and often, and why they encourage you to satisfy many of your "general ed" classes with classes that do double-duty as the "intro/survey/101" courses for other majors. They have every incentive to help you graduate in 4 years... they're expensive, they know it, and they know there's a nontrivial chance you might not graduate at all if they seriously derail you. They know that 70% of their students change their majors at least once before year 3, and most of them have had more than a hundred years to refine the formula and get it right.
The generic community-college scenario only really works for two groups of students... those whose only goal is "a degree", regardless of what it might be in, and those for whom community college is a second chance to shine, catch up, and redeem themselves. A student who's already at the top of his high school class and a shoo-in at just about any university is basically just wasting his time, and is actually INCREASING his odds of stumbling and losing his way before graduation.
The fact is, the "2+2" formula just doesn't work for the majority of students.
Galaxy S3 is a bit more complicated and expensive due to its 11-pin semi(?)-proprietary MHL connectors ("SaMHL"). The cheapest 11-pin SaMHL cable I'm aware of is ~$15, and the word at XDA is that it's totally luck of the draw... about half work fine, and half fall somewhere between "occasionally-flaky" and "just plain doesn't work". The cheapest cables known to work at least as reliably as Samsung's official ones start at around $25-30.
Yes, you can spend a few bucks on an 11-pin to 5-pin MHL adapter and use a cheap generic 5-pin MHL cable with reasonable success, but THEN you're putting a HUGE amount of lever-induced stress on the S3's USB port, and just begging for a cracked mainboard or weakened solder joint. Upgrade the adapter to one with at least an inch or two of flexible cable between the plug and the socket, and you're back up into the $25+ range and might as well get a good non-Samsung cable made for 11-pin SaMHL from the start.
The situation will obviously improve in a year or two since all Samsung phones (and probably others) will be using 11-pin MHL going forward... but right now, there's still a 100% or more price premium for S3-compatible MHL cables.