When the Wii came out, it had fluid, well-executed 480p60 when Sony & Microsoft were lurching around with grossly-underpowered 720pWhatev3r, and true abominations like 1080i60. They finally caught up, and can now do effortless 720p60 and 1080p30 (maybe even 1080p60). Now *Nintendo* has the mediocre, underpowered HD, saddled with expensive controllers that are basically less-capable than a no-name $79 Rockchip-based Android tablet from Shenzhen. They would have been better off with a $149 720p60 big brother for the Wii.
The difference is, a Debian phone would probably be able to save camera images in every meaningful format to have existed since 1980 *EXCEPT* Jpeg unless you rebuilt the whole thing from scratch for some *unfathomably* stupid & pedantic reason (install ImageMagick under Debian sometime, then go through hours or days of pain getting Jpeg support to work, and you'll know what I'm talking about).
The next step in video: cameras with internal framerate of 300fps that capture and save 100 of them with the precise timing of 50fps and 60fps video, combined with editing software that guarantees that as long as you stick to "splice points" falling every 100ms (the 1 frame in 5 for 50fps, and the 1 frame in 6 for 60fps, that both occur at the exact same moment in time before the next 9 frames diverge), you can shoot one source copy, then use it to generate native 50fps and 60fps output copies. Or, possibly, a version with outright asynchronous framerates that basically captures 60fps video with precise timing, adds a 7th frame 50#3 exactly halfway between frames 60#3 and 60#4, then quickly grabs a reduced-detail monochrome frame a few milliseconds before 60#2 and after 60#4, so that in post-production you could do motion-vector temporal rate correction on frames 50#2 and 50#4 that used the "quicksnap" frames to determine the exact grayscale detail & calculate the motion vectors, then derived the color by applying those motion vectors to the adjacent 60fps frames.
In linear order, with some semblance of relative timing:
Then, for the next stage, keep the imaging sensor with raw 300fps capability, and grab additional frames in between the 50fps and 60fps key frames with alternating longer and shorter exposures to obtain additional dynamic range that could be retroactively applied to the adjacent 50/60fps key frames in post-production (practically rendering lighting problems for shows meant for TV irrelevant, and giving news networks an extra bit of headroom since they CAN'T go back to re-shoot some live event.
For consumer gear, they could do something similar to skip the 50+60fps dual-framerate capability, and instead capture video at double the intended framerate, where every other frame is alternatingly over- or under-exposed, and enable the extra frames to either extend the dynamic range of the "good" frames, or do motion-vector transformations on the over/under-exposed frames to replace "key" frames that are themselves too dark or light to show directly.
Or, some variant on cameras for news crews where you have one lens and 3 or more CCDs, but instead of using the different CCDs to capture red, green, and blue, you'd expose and sample one CCD with 50fps timing, one CCD with 60fps timing, and a third CCD that's lower-res & monochrome, with extended infrared sensitivity and selectable IR-cut filter. In bright light, the IR cut filter slides in, and the monochrome channel gets under-exposed. In dark light, the IR cut filter slides out, and the monochrome channel gets over-exposed. In really dark lighting, it gets over-exposed at half the framerate with tweaked 25fps timing. The idea is that given enough time in post-production, almost anything could be salvaged from bad lighting.
Add fresnel lenses to high pixel density sensors so you can go in and re-render virtually re-focused frames after the fact, and adjust things like focal depth and focal plane to your liking, and you'll end up with a camera where nearly any problem can be fixed in post-production.
The underlying technology is all here, and has been for quite a while. The only thing missing was the terabytes of storage space needed to capture multiple HD video streams simultaneously from multiple sensors capturing at different framerates, and software that's aware of it.
Honestly, I don't think there's a single state that could survive on its own if it were required to assume its share of the national debt -- payable in US dollars -- as a condition of independence. I also don't think any state would be better off, let alone thrive. Even if Texas managed to do OK for a couple of years, the first Category 5 hurricane that slammed into Galveston and Houston would either wipe both cities from the earth forever & turn them into de-facto Bangladesh, or impoverish the rest of the state trying to cover the recovery costs without FEMA, the Fed, and the nearly-infinite pockets in Washington to ultimately eat the cost.
Ditto for California. It might do OK for a few years, but eventually a horrific earthquake is going to completely lay waste to a large chunk of Los Angeles or San Francisco, and the other city would get overwhelmed by refugees within days while the destroyed city turned into a humanitarian nightmare that the rest of the US wouldn't be terribly inclined to do anything about unless there were major strings attached.
There isn't necessarily a hard, sharp line, but there's a lot of benefits that come from being a geographically-large country with a huge population, abundant resources, and reasonably-diverse economy. The United States, Russia, and Brazil are safely on the good side of it, as are China and India. It's the real reason why countries in Europe finally decided to put aside their ancient arguments and join the EU. Being big gives you the ability to absorb hits that would be individually-devastating to smaller units.
Imagine Britain, France, or Germany trying to individually deal with the aftermath of a Tunguska-sized meteorite slamming into the earth 20km from Strasbourg, Mannheim, or Liverpool. Without the EU to bail them out, their whole economies would be in ruins, the same way Florida's economy would have been destroyed by the 2004-2005 parade of hurricanes that began with Charley, continued into Ivan (twice!) & Jeanne, then had encore performances the following year with Dennis, Katrina, Rita, and Wilma. The vast Roman Empire brushed itself off after Vesuvius buried Pompeii practically a stone's throw away from its capital city. In contrast, a comparable disaster wiped out the much smaller Minoan empire within a matter of minutes, and left little besides the legends of Atlantis to show for it.
Yes, but that's because the act of actually paying off the debt would itself extinguish the Dollar's value, not because they'd be printed out of the blue.
By any chance, did you buy it on eBay? Or from anybody besides a store like Best Buy or OfficeMax? If you did, there's a VERY good chance that it actually has half the capacity it claims to have, does transparent data compression to conceal it, lies about its free space, and simply overwrites the oldest sectors with newer data if you end up totally filling it up. My dad learned the hard way that allegedly-Sandisk flash cards from China with prices that are "too good to be true" probably are, and almost lost a quarter of his vacation photos as a result. This is VERY common with cards that don't have tightly-controlled supply chains
> Right now the closest we're seeing to that are the multi hundred million dollar blockbuster movies. > If more sensible copyright law meant that they were no longer profitable to make, well, if rather have the law. > Much as I might like the spectacle, we know we can all live with lower movie budgets.
Or, more likely, the next logical step we're likely to see within a decade... big-budget Hollywood movies that film the same movie with two different sets of actors... one English-speaking, one Chinese. Same plot, same (translated) dialogue, same sets, same director, same film crews, same CGI... just different slabs of talking meat, filmed scene by scene, one after the other. Maybe even add an ethnically-Indian third cast if it's likely to make the difference between a big movie and an insane blockbuster mega-hit in India. 20-50% higher production cost, double the profits or more. Right now, I can *guarantee* that there's a bilingual Chinese & English-speaking film major who's going to school somewhere in America or China & already has the business plan mapped out.
It makes to much sense to NOT happen. China is a vast market that loves big-budget Hollywood films, but people there still kind of have a love-hate relationship with them as "foreign" films and really prefer their own big-budget blockbuster movies. Let's face it... NOBODY likes watching dubbed movies, and the only thing worse are subtitles. America has its blockbuster movie stars, but so does China. As expensive as paying two sets of actors who are "Tom Cruise" or "Wesley Snipes" level would be, it's a pittance compared to the cost of everything else that goes into making a big-budget movie. We ALREADY remake hit British TV shows into American shows and vice-versa... doing it all at once just cuts the production costs and increases the profit.
That said, I personally wouldn't mind English-dubbed movies so much if they'd just get people with the expected foreign accent to DO them (preferably, the original actors themselves). Maybe it's just an "American" preference, but to me, it seems blatantly obvious that someone who's French and speaking English in France is going to have a French accent. Ditto for German, Russian, and other languages. The moment somebody in a German film, standing in an obviously-German street opens his mouth and sounds like somebody from Ohio, the illusion is shattered, and I might very well switch the language back to German and turn on the subtitles. It's just *WRONG*. Yet, it seems like every foreign film that gets dubbed into English does the EXACT SAME STUPID THING. I've talked to British, Canadian, and Australian friends & coworkers over the years, and they've all expressed similar sentiments... dubbed English should HAVE the accent of whatever country the film is set in. Yet, I still see film after film dubbed with completely absurd and inappropriate American accents.
But anyway, if you think we have big-budget blockbuster movies now, just wait until Hollywood is cross-producing 3-megamarket movies with budgets that George Lucas would have regarded as jaw-droppingly absurd, and a movie is considered to be a flop if its first-week gross revenues don't exceed a billion dollars. When you're talking about *that kind* of money, it's *worth* filming every scene 2 or 3 times with different actors for each major market, so it can be a "native hit" in each one.
They won't, however, give you any kind of a discount for buying your own phone, and your likelihood of getting LTE to work is almost nonexistent unless the phone you bought full-price was AT&T-branded (and bootloader-locked) to begin with. The fact is, right now, if you don't spend most/all of your time in areas where T-Mobile has solid coverage and HSPA+ or LTE, you're basically fucked phone-wise if you insist on being a rebel. You'll be stuck with slow 3G data, and basically end up with the worst of all worlds.
Don't forget the ability to use flash drives via USB OTG to make up for the often-intentional omission of a microSD slot. It's like it's not bad enough that they have to gimp the phone and take away the one part likely to survive anything short of hellfire and brimstone (a microSD card is one of the few things LEFT that someone with some wire, an Arduino, and a soldering iron can try doing his own guerrilla data-recovery on if the card itself gets crushed or mangled. Just *try* mounting a crushed, waterlogged, and destroyed Android phone that can't boot as a USB flashdrive to copy your data off of it. With a microSD card, you might actually succeed, even if you have to do it one bit at a time via SPI and reconstruct the filesystem offline.
Unfortunately, in America, GETTING your bootloader unlocked almost always requires root. At least, if you want a phone that has a microSD slot, can take extended batteries, and isn't last year's hardware in a new package. Google seems determined to arbitrarily bork and cripple all their new hardware in the holy name of all that is Cloud, which leaves us in the real-world position of having to deal with bootloaders that are at least protected by the equivalent of a skeleton key hidden under a doormat that says "Monsters are under here".
It's sad, but I'm starting to feel about as non-enthusiastic about new releases of Android as I've been about new releases of Windows for the past few years... like every new release is two steps forward, and seven steps diagonally backwards towards left or right. If the new version of Google Maps is any indication of the direction Google plans to take Android, I might be living on some fork of CM10.x for a really, REALLY long time.
Let me make this easy: US national debt is Dollar-denominated, and the only limits on the number of dollars that can be printed to pay redeemed bonds are determined by the US itself. The only way the US could literally default on its debt would be as an act of deliberate grandstanding financial suicide. As long as Congress doesn't say "No", the Federal Reserve can go print a million brand new $100 bills and use them to pay $100 million in redeemed bonds.
As others have noted, if you owe me a thousand dollars and you're broke, you have a big problem. If you owe me a MILLION dollars and you're broke, I have a big problem. And if I owe you a hundred quadrillion dollars, you're either certifiably insane, or we're both playing with imaginary money that nobody ever really expects to be paid.
Contrary to popular myth, Texas does NOT have the right to secede from the United States.
It HAD the right to secede from the United States until the mid-1800s... and it exercised that right to secede, as it was entitled to do. Had it walked away from the US and never looked back, it would be entitled to be independent today (but almost certainly a northern territory of Mexico, since it would have been invaded and conquered almost instantly without anyone to defend it from invasion and annexation by Mexico). However, that's not what happened.
After the Civil War, the United States could have -- with total and complete constitutional legitimacy -- treated it like a vanquished foreign invader, stripped it bare the way the Soviet Union did to East Germany, then kicked it a few times and told Mexico, "it's yours, have fun." Instead, Texas got lucky, and was allowed to remain as a state. Some can argue about whether the federal government had the right to do that instead of making Texas re-petition the US for admission as a territory, then a state, and whether the Texans' occupation government had the right to relinquish their right to secede again... but they did, and it's political masturbation to argue otherwise.
As part of the US, Texas is a big, wealthy, relatively powerful state. As an independent nation allowed to secede only after assuming direct responsibility for its prorated share of the national debt -- without enjoying the benefits of having its own dollars accepted as a world reserve currency & the ability to write checks to itself as payment -- it would be crushed into financial ruin and bankrupt before its first official flag-raising ceremony in Austin.
> And perhaps that is because they plant millions of the same species/strain with no natural variation?
Actually, it's much worse than that. The evolution clock for Florida's orange trees stopped DECADES ago. Florida's citrus trees are pretty much ALL clones. I don't think it's even possible to grow a viable orange tree from seeds anymore.
> If we wipe out one year's corn or wheat we're going to experience starvation on a scale larger than the potatoe famine.
It depends. Corn probably has a fair amount of headroom, just because *so much* of it gets used to make ethanol & high-fructose corn syrup. The likelihood of "an entire year's crop" getting wiped out is pretty slim. The majority of corn might get grown in places like Iowa because that's where it's the cheapest and most cost-effective to grow it, but there were wholesale crop failure in places like Iowa and Nebraska one year, you can BET farms in Florida, Arizona, and California that wouldn't normally bother would seriously consider growing it for a season or two. The likelihood of some virus or insect making its way from Iowa to California or Florida within a single year is pretty slim, especially if farmers in those states were determined to keep their own fields quarrantined.
In countries like the US, the biggest fights would be between farmers and companies like Monsanto if, for example, they had a strain of corn that wasn't affected, but couldn't be supplied in full commercial quantities for a year or two. You'd have farmers cloning sprouted plants, Monsanto sending out the lawyers with C&D notices, and farmers introducing refused purchase orders in court as evidence that Monsanto suffered no actual damages since it was unable to actually fill their orders as placed.
An equally bad problem (if asymmetric-impact) would be if some virus or pest suddenly wiped out an entire generation of "Round-Up Ready" crops, forcing large-scale commercial farms to go back to more conventional strains... and try to raise, tend, and harvest them with a labor force that no longer exists, and expensive capital-intensive machines that were no longer relevant. This scenario would have almost no impact on farmers in central Africa who can't afford large-scale capital-intensive farming techniques anyway, but it would be devastating to American agribusiness.
The truth is, though, that nothing short of outright war between equally-potent nation states occurring SIMULTANEOUSLY with crop failure is likely (in the US or Europe, at least) to have the kind of devastating impact some like to predict. Fresh vegetables, fruits, and grains would become expensive overnight, but there's actually a pretty huge amount of highly-processed food already in the pipeline, stored in warehouses ready to ship to Sam's Club & McDonalds, and the actual impact for most Americans (the majority who wouldn't even know what to DO with fresh food if you put it on the kitchen counter in front of them) wouldn't be an immediate crisis so much as a long-term lopsided increase in food prices that began months after the actual crop failure, and persisted even longer after it was resolved.
Suppose tomatoes had wholesale crop failure for several years nationwide. The most immediate consequence? Restaurants would quit putting sliced and diced tomatoes on food unless you specifically asked for it. If it persisted, they might start to charge extra for them as a premium ingredient (a-la-bacon). Ketchup and spaghetti sauce? Please. They barely have statistically-meaningful amounts of actual tomato in them to start with. In the US, at least, they're basically artificial flavoring, high-fructose corn syrup, tapioca, salt, spices, and FD&C Red #40. Add hot peppers, reduce the tapioca & guar gum, and it's taco sauce. You'd see restaurants quietly deprecating chunky salsa, replacing it with melted cheese and red goo with the consistency of spaghetti sauce, and gleefully advertising their new, improved CheesyZingySalsa.
Wheat would be harder to replace... but so much wheat is grown around the world, the main consequence of wheat failure in north America would be the disappearance of wheat from African supermarkets, and American restaurants suddenly becoming VERY enthusiastic about their new "low-calorie, low-fat sandwich wraps".
Corn? Well, for starters, probably half the corn grown in the US ends up as
The fundamental problem is that cars are kind of like American Android phones... consumers are mostly powerless to do *anything* to fix vulnerabilities, and carriers won't do anything they aren't forced to do by law. And the law itself is only slightly less castrated and toothless than consumers.
If a car suddenly becomes something you can't safely use (or at least leave unattended in a non-secure location), there's no meaningful immediate recourse for consumers, and that's a real problem with real consequences for real people.
The solution isn't to contact the automakers and keep it a secret... the solution is to quietly meet with representatives from Allstate, Progressive, and State Farm (among others), and demonstrate it to THEM. They're better at getting the attention of Congress and the auto industry, and you can BET they'd be on the ball if they thought there was a vulnerability could leave them on the hook for billions of dollars worth of insurance claims.
Hmmm. Interesting. But now the million-dollar question... are signed distance fields fast enough on any current high-end Android or IOS hardware to actually render a full page of arbitrary text that includes multiple fonts and multiple styles of those fonts? All the examples I found via Google use it to render just a few glyphs at any one time.
Realistically, if you're rendering English text and you stick to pre-loading characters that are actually used or likely to be used, you're going to need about 80-90 for each font+style. For technical books, that realistically means at least 3 complete font sets in normal, bold, italic, and bold+italic -- usually, a serif font, a sans-serif font, and a monospaced "console" font with disambiguated zeroes and ones for code examples.
Ram-permitting, text has somewhat of the advantage that you can preload your font data, then reuse it from "scene" to "scene". On the OTHER hand, with a game, you can define a large scene, then just move the frustum around to change what's visible from frame to frame. With text, a new page means defining and rendering an entirely new scene from scratch... and doing it at 60fps without lag or latency is no small feat. Add page-turning visuals, and you have a task that could bring the development team from a game like "Battlefield 4" to their knees and have them begging for mercy (at least, when you tell them it has to work acceptably well on a MSM8960's Adreno 2 or a Tegra3, even if the device HAS a gig of ram and dual/quad-core 1.5GHz+ CPUs).
Text might not be sexy, but rendering print-quality text in realtime is *hard*. We've settled for less in the past, because text on a video display was always regarded as a second-rate ghetto that was good enough for nasty throw-away stuff, but was never expected to approach the visual quality and responsiveness of printed text on paper. Video displays have now largely caught up (like the iPad's retina displays), but the hardware we need to put them to proper use is still very much in its infancy. The industry used the slow update times of e-ink as an excuse for years, and when displays capable of crisp, high-resolution text display and instantaneous updates appeared, the underlying hardware wasn't even close to being ready to take advantage of it for anything besides sequential page-by-page reading of novels.
> You can use 3d acceleration to generate a scene or just draw in 2d.
Until hinted scalable fonts enter into the equation. Then everything goes straight to hell. If you tried to write pixel-shaders to define every single letter of every font, style, and weight used on a page in a way that's legible at 8 pixels tall and smooth & curved without artifacts at 100, and had a hard rule that you couldn't do anything that couldn't be fully-rendered from definition to rendering in 1/60th of a second, you'd run out of ram, time, and/or triangles LONG before you got to the end of a technical ebook's first real page.
The fundamental problem is that triangle-based acceleration is the *wrong* kind of acceleration for high-quality text rendering. Fonts need B-spline acceleration. Rendering text with triangles is like trying to emulate a Soundblaster Pro with a Gravis Ultrasound was back in 1993.Sometimes, you need hardware that's optimized for the specific task at hand, and trying to generalize it into something else just makes matters worse.
Oh, the days of using separate video cards for 2D and 3D support. It was "cool" to have a setup like that, but somehow I was never interested and held out for the TNT2.
> Oh, the days of using separate video cards for 2D and 3D support. > It was "cool" to have a setup like that, but somehow I was never interested and held out for the TNT2.
And thanks to mass-market consumers who did the same thing, we ended up with video GPUs today that are basically a pimped out 3DFX stapled onto a dumb framebuffer, with no real 2D acceleration to speak of.
Instead of getting hardware-accelerated B-splines and the ability to render subpixel-hinted scalable fonts via hardware in realtime sometime around 2006 like we were supposed to (going by ATI's roadmaps), we have Android and IOS hardware built around GPUs that couldn't render a full page of hinted dealiased text a-la-Postscript to display memory in 1/60th of a second if the future of their manufacturers' companies depended on it. Because 3D is trendy, hot, and sexy, and 2D isn't.
Joe Sixpack doesn't know what a B-spline or subpixel rendering is, but he knows that 3D is "cool", and a GPU that has "more triangles" is better (the same way he "knew" a 1.8GHz Pentium 4 was better than a 1.1GHz Pentium III Xeon, and even ran out to buy a new laptop with one).
That's why Android & IOS-based e-readers suck for interactively reading technical books that require constant page-flipping. They lack 2D spline acceleration, so they have to do everything via brute CPU force. They're too slow to render pages from scratch in realtime with "real book" aesthetics, they don't have enough memory to pre-render the whole book to ram, and they're too slow to fetch entire pre-rendered arbitrary pages from microSD in 1/60th of a second or less(*).
(*) The fastest microSD interface on any known Android phone maxes out around 25MB/s.... and a 32-bit 1280x800 bitmap weighs in around 4MB. Real-world Android phones like the S3 generally max out around 17MB/s. The fastest UHC-1 Sandisk Extreme cards have a theoretical max of 95MB/s, which STILL isn't fast enough to fetch the ~187MB/sec required for realtime brute-force 1280x800x60fps @ 24 bits. In theory, the 62MB/s required to fetch 8-bit grayscale at 60fps might be do-able on a future phone with UHC-1 microSD, but no current device can do it.
> Oracle is a rapacious, money-gobbling machine of a company.
No, they're a vast multi-level marketing scheme that happens to sell a database or two as a side venture, and recently bought a product (Java) that doesn't require an army of highly-paid consultants who have to tithe 20% of their income to them in return for mandatory certifications and licensing just to keep it running.
The big "gotcha" with Java is that Sun's license for Java was always pretty nasty, but they generally looked the other way and ignored all but the most egregious violations of it unless you were trying to use it in embedded devices. Oracle, in contrast, intends to enforce it to the letter. The only consolation prize with Java is the fact that most of Oracle's legal leverage over OpenJDK comes from patents rather than copyright, and software patents -- while questionable in most cases -- at least have the benefit of not being de-facto eternal & endlessly extended every time Steamboat Willie is at risk of becoming public domain. So in another 10-12 years, we'll be able to unambiguously do anything we like with Java, as long as it's cleanroom-engineered and called "Bali" or "Sumatra" (the islands flanking "Java").
> Bottom line, if you have N digital copies then what is the benefit of keeping the original DVD over one N+1 digital copies of the DVD?
The original DVD was pressed. Digital copies depend upon either brittle spinning hard drives that are cost-prohibitive to repair and recover data from, flash media that's the equivalent of a leaky bucket (especially MLC), organic-dye based optical media that shift over time, and manufacturing processes whose lifetimes are absolutely 100% speculation to begin with.
The more diversely-redundant your backup media, the more likely it is that you'll have enough of it survive to do high-quality archival restorations later. If you have a pressed disc, a non-LTH BD-R, and a hard drive backup, the pressed disc is the most likely out of all of them -- barring some bizarre manufacturing or design flaw -- to still work with then others have failed.
Hard drives are convenient and cheap (now), but they are NOT long-term passive storage media. You CAN NOT just throw a hard drive in a drawer for 20 years and expect it to work, assuming you can find something compatible to attach it to and software to interpret it with. Ironically, a modern version of the old SyQuest cartridges that was used only under cleanroom conditions would probably be the most robust of all.
What kills hard drives is mechanical failure combined with the near-impossibility of affordable repair. If Western Digital & Seagate did their own in-house low-cost repairs using their own parts, employees, and tools, hard drives would actually be VERY repairable, and permanent data loss would be rare barring total destruction. Unfortunately, they don't, and what we HAVE is a hard drive data-recovery industry that's basically built around reverse-engineering, parts-cannibalism, and black magic. Companies like WD could EASILY design drives to be serviceable, so that in a worst-case, you'd send in your drive for a warranty repair, they'd do some quick software tests to confirm hardware failure, remove the platters, mount them in their own recovery unit, image them to your new replacement hard drive, and ship it back to you. But no, hard drives are simultaneously engineered to be as cheap & likely to fail as possible, with no cost-effective means of data recovery when it happens.
Higher-end drives come with longer warranties, but warranties are worthless for data. The one thing the industry refuses to do is make drives that are expressly DESIGNED to fail in repairable ways. I have rather strong feelings about this matter, having gotten bitten HARD by multiple drive failures in a row that left me one drive away from lifetime data loss (well, ok, I did have older scattered backups, but that one drive was the closest thing I had left to a coherent, semi-ORGANIZED backup copy that wouldn't have left me playing data-archaeologist for 20 years).
BD-R (non-LTH) is a phase-change medium, which is pretty much the best thing we have now. Even better, the media is decoupled from the read mechanism, so failure of the read mechanism won't render the disc itself orphaned.
Another big problem that bites lots of people with CD-R specifically... CD-ROM/XA Mode 2, Form 2... and the willingness -- if not eager default -- of early software like Nero to allow users to disable error correction to gain a few more megs of storage space without making it abundantly clear to those same users what the consequences of doing it were. Lots of people went for YEARS burning discs that would develop hard errors at the slightest scratch without realizing WHY it was happening. They just attributed it to drive/media flakiness, and didn't realize that it was their burning software setting them up for disaster.
> For born-analog content, always keep the original physical copy.
And if, for whatever reason, you CAN'T retain the original copy, oversample the bejesus out of it, store it with the most lossless compression you can, and sample it multiple times, in multiple ways, with multiple scanners/digitizers/capture cards.
Case in point: videotape. If your goal is to merely capture a copy to casually watch years later, and you don't mind having a digital copy that's demonstrably worse than the analog original, just about any cheap one-click capture card will do. BUT... if you want to stop the clock and capture a copy that doesn't visibly look worse than the original, compress it to MPEG-2 it at a MINIMUM of 6,000kbps VBR, at 720x480 (or 540, if it's PAL). Yes, we all know that VHS is low-res. The problem is, it's also noisy, and noise doesn't compress well. If a 720x480 video encoded with VBR and the maximum DVD-compliant bitrate available looks as good as the original VHS, consider yourself LUCKY. Regardless, never downsample the vertical resolution. Horizontal resolution with VHS is open to debate, but your vertical resolution is known and definite -- 480 (525) if it's NTSC (or 60hz PAL-J), and 540 (576) if it's PAL.
Now, we get into my #1 beef with Blu-Ray -- its despicable lack of compliant encoding options for 480i60 and 540i50 source, so there's really no way to profoundly oversample VHS, yet retain casual "stick in the disc and hit play" usability that can be watched directly in a Blu-Ray player, but someday used for restoration efforts.
If you REALLY want to capture a videotape and preserve every possible nuance for future restoration efforts before the tape degrades further, horizontal oversampling of VHS is mandatory. If Blu-Ray supported a hypothetical 1280x480 resolution with 4:4:4 chroma (remember, with VHS, your vertical resolution is your "good" resolution... it's 480 or 540. It's your horizontal color resolution that completely sucks. Chroma-encoding variants that split color between 2 scanlines will ALWAYS visibly-degrade VHS captures. Don't do it. )
In a perfect world, there would be a near-line archival video format designed for videotape captures that directly sampled the full-bandwidth FM signal coming off the tape (preferably, at 3 slightly different trackings... spot-on, plus one slightly "off" each way). A few times, I've tried to work out the math and figure out what kind of insane sample rate would be necessary to try magnetically sampling the tape straight-on, with multiple overlapping rows of microscopic read heads, and sufficiently-high density, to literally map the flux of every flake of oxidized metal on the tape and allow future offline reconstruction by digitally simulating VHS's diagonal read path, so that even if the original tape were destroyed, a future restoration could be performed against the digital copy. I have a hunch that something like this is actually do-able now (and probably would have been do-able a few years ago, had multi-terabyte hard drives been affordable), but I've never even read speculation online about such an archival device.
The point is, if your goal is to someday do color restoration, resolution-enhancement, or whatever, taking advantage of analog's tendency to keep surprising us with ghostly hints of lost detail that we'd always assumed were gone, you need a profoundly-oversampled copy with basically NO lossy compression (or oversampling so profound, you can downsample away the artifacts). Why do you need so much oversampling depth and detail? True hardcore restoration and after-the-fact re-synthesis is going to depend upon what are known as "higher-order artifacts", and those are precisely what get mangled by any kind of lossy compression. Remember, lossy compression is designed to throw away bits in areas where your eyes won't notice detail. An algorithm looking for subtle field-ripples as evidence that a pixel nearby is supposed to be intensely red instead of brownish-orange won't see them if some earlier compression algor
> I have burned CDs from 1995 that still work perfectly fine.
If they're redbook audio CDs, and your definition of "work perfectly fine" is "I can stick the disk in, hit play, it spins up at 1X, music comes out, and the player doesn't totally gag", you might be right. Now try ripping the disc using software that can monitor the realtime bit error rate. You'll probably be *horrified* to see how high it is.
Redbook audio CDs are very robust, even when their bits are rotting all over the place. They were designed in an era when hardware couldn't do much in realtime, so they bent over backwards to make sure they had a "plan B" to make sure the show would go on after the disc got scratched, dirty, or whatever else happened to it. They were designed so the audio data is interleaved in a way that when a read error occurs, the left and right channels get merged for 1 sample. A redbook audio CD has to be nearly *destroyed* (cracked, melted, fried, whatever) before it literally won't play, as long as the player is able to find the lead-in and sync up to the spiral track.
It'll start to sound "rough" and lose channel separation, but things have to be pretty bad before it will LITERALLY stop playing. At least, as long as the player itself is faithfully following the original redbook audio specs, and isn't trying to realtime-rip the audio to a ram buffer and play it back from there (which is what some, if not most, new optical-disc media players do TODAY). I have plenty of CDs that new players choke on and refuse to even try playing, but yet my 25 year old antique CD player that cost something outrageous like $600 or $800 when new, can play just fine. Apparently, it's because first-generation CD players were precision hardware that could blindly track a CD spiral as long as the disc itself was 100% within spec, whereas new players depend upon realtime error-analysis to stumble and wobble around, and make up for the fact that discs no longer spin precisely, and worm-gear optical assemblies no longer track with precision measured in microns.
That said, my experience has ALSO been that CD-R discs manufactured in THIS century are less likely to rot and become unplayable in new drives, but are more likely to have major problems with old players. The old players were precision hardware, and assumed the discs themselves were manufactured to precision specs. The first-gen CD-R media had dye that deteriorated over time, but their spiral tracks were spot-on, just like pressed discs. As drives got better at handling sloppy tracking, the discs themselves became sloppier.
Net effect: first-gen redbook audio CD-R media is likely to play with acceptable audio quality on an old CD player from the 80s or early 90s, but be unplayable on many modern drives & be un-rippable on most drives (some will allow you to spin down to 1X & emulate the playback mode of a legacy player if you're running a sophisticated ripping app). Newer discs that are still old will probably skip and have problems playing on an old player, but might still be equally bad on a new one. When today's bargain-bin CD-R media is 10 years old, it will probably be unplayable on anything, the same way my old VHS tapes from the 80s still play fine, but VHS tapes recorded after ~1998 are largely unplayable on anything I can find.
TLDR point: the storage life of "last-gen" CD-R media is likely to be better than first-gen CD-R media was at the same age, but enormously WORSE than that of the best "turn of the century" CD-R media (the golden era when quality standards were still high, and the worst faults of the first-gen media were addressed. Any box of CD-R media you buy TODAY is probably shit of the worst kind. The best media you can buy TODAY for long storage life? Non-LTH BD-R single-layer discs. But MAKE SURE they aren't LTH... most manufacturers don't go out of their way to scream, "These discs are LTH garbage!"
In South Florida, even the urban-ish parts? Water moccasins. Even in central Miami, they swim up the river, the Tamiami Canal, or large drainage ditches from the Everglades & *somehow* climb out into back yards. The lights also help deter break-ins.
Slashdot Mirror
Yes... and no.
When the Wii came out, it had fluid, well-executed 480p60 when Sony & Microsoft were lurching around with grossly-underpowered 720pWhatev3r, and true abominations like 1080i60. They finally caught up, and can now do effortless 720p60 and 1080p30 (maybe even 1080p60). Now *Nintendo* has the mediocre, underpowered HD, saddled with expensive controllers that are basically less-capable than a no-name $79 Rockchip-based Android tablet from Shenzhen. They would have been better off with a $149 720p60 big brother for the Wii.
The difference is, a Debian phone would probably be able to save camera images in every meaningful format to have existed since 1980 *EXCEPT* Jpeg unless you rebuilt the whole thing from scratch for some *unfathomably* stupid & pedantic reason (install ImageMagick under Debian sometime, then go through hours or days of pain getting Jpeg support to work, and you'll know what I'm talking about).
The next step in video: cameras with internal framerate of 300fps that capture and save 100 of them with the precise timing of 50fps and 60fps video, combined with editing software that guarantees that as long as you stick to "splice points" falling every 100ms (the 1 frame in 5 for 50fps, and the 1 frame in 6 for 60fps, that both occur at the exact same moment in time before the next 9 frames diverge), you can shoot one source copy, then use it to generate native 50fps and 60fps output copies. Or, possibly, a version with outright asynchronous framerates that basically captures 60fps video with precise timing, adds a 7th frame 50#3 exactly halfway between frames 60#3 and 60#4, then quickly grabs a reduced-detail monochrome frame a few milliseconds before 60#2 and after 60#4, so that in post-production you could do motion-vector temporal rate correction on frames 50#2 and 50#4 that used the "quicksnap" frames to determine the exact grayscale detail & calculate the motion vectors, then derived the color by applying those motion vectors to the adjacent 60fps frames.
In linear order, with some semblance of relative timing:
50/60.1 --- 60.2 -- 50.2 - 60.3 - 50.3 - 60.4 -- 50.5 --- 60.6
Then, for the next stage, keep the imaging sensor with raw 300fps capability, and grab additional frames in between the 50fps and 60fps key frames with alternating longer and shorter exposures to obtain additional dynamic range that could be retroactively applied to the adjacent 50/60fps key frames in post-production (practically rendering lighting problems for shows meant for TV irrelevant, and giving news networks an extra bit of headroom since they CAN'T go back to re-shoot some live event.
For consumer gear, they could do something similar to skip the 50+60fps dual-framerate capability, and instead capture video at double the intended framerate, where every other frame is alternatingly over- or under-exposed, and enable the extra frames to either extend the dynamic range of the "good" frames, or do motion-vector transformations on the over/under-exposed frames to replace "key" frames that are themselves too dark or light to show directly.
Or, some variant on cameras for news crews where you have one lens and 3 or more CCDs, but instead of using the different CCDs to capture red, green, and blue, you'd expose and sample one CCD with 50fps timing, one CCD with 60fps timing, and a third CCD that's lower-res & monochrome, with extended infrared sensitivity and selectable IR-cut filter. In bright light, the IR cut filter slides in, and the monochrome channel gets under-exposed. In dark light, the IR cut filter slides out, and the monochrome channel gets over-exposed. In really dark lighting, it gets over-exposed at half the framerate with tweaked 25fps timing. The idea is that given enough time in post-production, almost anything could be salvaged from bad lighting.
Add fresnel lenses to high pixel density sensors so you can go in and re-render virtually re-focused frames after the fact, and adjust things like focal depth and focal plane to your liking, and you'll end up with a camera where nearly any problem can be fixed in post-production.
The underlying technology is all here, and has been for quite a while. The only thing missing was the terabytes of storage space needed to capture multiple HD video streams simultaneously from multiple sensors capturing at different framerates, and software that's aware of it.
Honestly, I don't think there's a single state that could survive on its own if it were required to assume its share of the national debt -- payable in US dollars -- as a condition of independence. I also don't think any state would be better off, let alone thrive. Even if Texas managed to do OK for a couple of years, the first Category 5 hurricane that slammed into Galveston and Houston would either wipe both cities from the earth forever & turn them into de-facto Bangladesh, or impoverish the rest of the state trying to cover the recovery costs without FEMA, the Fed, and the nearly-infinite pockets in Washington to ultimately eat the cost.
Ditto for California. It might do OK for a few years, but eventually a horrific earthquake is going to completely lay waste to a large chunk of Los Angeles or San Francisco, and the other city would get overwhelmed by refugees within days while the destroyed city turned into a humanitarian nightmare that the rest of the US wouldn't be terribly inclined to do anything about unless there were major strings attached.
There isn't necessarily a hard, sharp line, but there's a lot of benefits that come from being a geographically-large country with a huge population, abundant resources, and reasonably-diverse economy. The United States, Russia, and Brazil are safely on the good side of it, as are China and India. It's the real reason why countries in Europe finally decided to put aside their ancient arguments and join the EU. Being big gives you the ability to absorb hits that would be individually-devastating to smaller units.
Imagine Britain, France, or Germany trying to individually deal with the aftermath of a Tunguska-sized meteorite slamming into the earth 20km from Strasbourg, Mannheim, or Liverpool. Without the EU to bail them out, their whole economies would be in ruins, the same way Florida's economy would have been destroyed by the 2004-2005 parade of hurricanes that began with Charley, continued into Ivan (twice!) & Jeanne, then had encore performances the following year with Dennis, Katrina, Rita, and Wilma. The vast Roman Empire brushed itself off after Vesuvius buried Pompeii practically a stone's throw away from its capital city. In contrast, a comparable disaster wiped out the much smaller Minoan empire within a matter of minutes, and left little besides the legends of Atlantis to show for it.
Yes, but that's because the act of actually paying off the debt would itself extinguish the Dollar's value, not because they'd be printed out of the blue.
By any chance, did you buy it on eBay? Or from anybody besides a store like Best Buy or OfficeMax? If you did, there's a VERY good chance that it actually has half the capacity it claims to have, does transparent data compression to conceal it, lies about its free space, and simply overwrites the oldest sectors with newer data if you end up totally filling it up. My dad learned the hard way that allegedly-Sandisk flash cards from China with prices that are "too good to be true" probably are, and almost lost a quarter of his vacation photos as a result. This is VERY common with cards that don't have tightly-controlled supply chains
> Right now the closest we're seeing to that are the multi hundred million dollar blockbuster movies.
> If more sensible copyright law meant that they were no longer profitable to make, well, if rather have the law.
> Much as I might like the spectacle, we know we can all live with lower movie budgets.
Or, more likely, the next logical step we're likely to see within a decade... big-budget Hollywood movies that film the same movie with two different sets of actors... one English-speaking, one Chinese. Same plot, same (translated) dialogue, same sets, same director, same film crews, same CGI... just different slabs of talking meat, filmed scene by scene, one after the other. Maybe even add an ethnically-Indian third cast if it's likely to make the difference between a big movie and an insane blockbuster mega-hit in India. 20-50% higher production cost, double the profits or more. Right now, I can *guarantee* that there's a bilingual Chinese & English-speaking film major who's going to school somewhere in America or China & already has the business plan mapped out.
It makes to much sense to NOT happen. China is a vast market that loves big-budget Hollywood films, but people there still kind of have a love-hate relationship with them as "foreign" films and really prefer their own big-budget blockbuster movies. Let's face it... NOBODY likes watching dubbed movies, and the only thing worse are subtitles. America has its blockbuster movie stars, but so does China. As expensive as paying two sets of actors who are "Tom Cruise" or "Wesley Snipes" level would be, it's a pittance compared to the cost of everything else that goes into making a big-budget movie. We ALREADY remake hit British TV shows into American shows and vice-versa... doing it all at once just cuts the production costs and increases the profit.
That said, I personally wouldn't mind English-dubbed movies so much if they'd just get people with the expected foreign accent to DO them (preferably, the original actors themselves). Maybe it's just an "American" preference, but to me, it seems blatantly obvious that someone who's French and speaking English in France is going to have a French accent. Ditto for German, Russian, and other languages. The moment somebody in a German film, standing in an obviously-German street opens his mouth and sounds like somebody from Ohio, the illusion is shattered, and I might very well switch the language back to German and turn on the subtitles. It's just *WRONG*. Yet, it seems like every foreign film that gets dubbed into English does the EXACT SAME STUPID THING. I've talked to British, Canadian, and Australian friends & coworkers over the years, and they've all expressed similar sentiments... dubbed English should HAVE the accent of whatever country the film is set in. Yet, I still see film after film dubbed with completely absurd and inappropriate American accents.
But anyway, if you think we have big-budget blockbuster movies now, just wait until Hollywood is cross-producing 3-megamarket movies with budgets that George Lucas would have regarded as jaw-droppingly absurd, and a movie is considered to be a flop if its first-week gross revenues don't exceed a billion dollars. When you're talking about *that kind* of money, it's *worth* filming every scene 2 or 3 times with different actors for each major market, so it can be a "native hit" in each one.
> AT&T does though,
They won't, however, give you any kind of a discount for buying your own phone, and your likelihood of getting LTE to work is almost nonexistent unless the phone you bought full-price was AT&T-branded (and bootloader-locked) to begin with. The fact is, right now, if you don't spend most/all of your time in areas where T-Mobile has solid coverage and HSPA+ or LTE, you're basically fucked phone-wise if you insist on being a rebel. You'll be stuck with slow 3G data, and basically end up with the worst of all worlds.
Don't forget the ability to use flash drives via USB OTG to make up for the often-intentional omission of a microSD slot. It's like it's not bad enough that they have to gimp the phone and take away the one part likely to survive anything short of hellfire and brimstone (a microSD card is one of the few things LEFT that someone with some wire, an Arduino, and a soldering iron can try doing his own guerrilla data-recovery on if the card itself gets crushed or mangled. Just *try* mounting a crushed, waterlogged, and destroyed Android phone that can't boot as a USB flashdrive to copy your data off of it. With a microSD card, you might actually succeed, even if you have to do it one bit at a time via SPI and reconstruct the filesystem offline.
Unfortunately, in America, GETTING your bootloader unlocked almost always requires root. At least, if you want a phone that has a microSD slot, can take extended batteries, and isn't last year's hardware in a new package. Google seems determined to arbitrarily bork and cripple all their new hardware in the holy name of all that is Cloud, which leaves us in the real-world position of having to deal with bootloaders that are at least protected by the equivalent of a skeleton key hidden under a doormat that says "Monsters are under here".
It's sad, but I'm starting to feel about as non-enthusiastic about new releases of Android as I've been about new releases of Windows for the past few years... like every new release is two steps forward, and seven steps diagonally backwards towards left or right. If the new version of Google Maps is any indication of the direction Google plans to take Android, I might be living on some fork of CM10.x for a really, REALLY long time.
Let me make this easy: US national debt is Dollar-denominated, and the only limits on the number of dollars that can be printed to pay redeemed bonds are determined by the US itself. The only way the US could literally default on its debt would be as an act of deliberate grandstanding financial suicide. As long as Congress doesn't say "No", the Federal Reserve can go print a million brand new $100 bills and use them to pay $100 million in redeemed bonds.
As others have noted, if you owe me a thousand dollars and you're broke, you have a big problem. If you owe me a MILLION dollars and you're broke, I have a big problem. And if I owe you a hundred quadrillion dollars, you're either certifiably insane, or we're both playing with imaginary money that nobody ever really expects to be paid.
Contrary to popular myth, Texas does NOT have the right to secede from the United States.
It HAD the right to secede from the United States until the mid-1800s... and it exercised that right to secede, as it was entitled to do. Had it walked away from the US and never looked back, it would be entitled to be independent today (but almost certainly a northern territory of Mexico, since it would have been invaded and conquered almost instantly without anyone to defend it from invasion and annexation by Mexico). However, that's not what happened.
After the Civil War, the United States could have -- with total and complete constitutional legitimacy -- treated it like a vanquished foreign invader, stripped it bare the way the Soviet Union did to East Germany, then kicked it a few times and told Mexico, "it's yours, have fun." Instead, Texas got lucky, and was allowed to remain as a state. Some can argue about whether the federal government had the right to do that instead of making Texas re-petition the US for admission as a territory, then a state, and whether the Texans' occupation government had the right to relinquish their right to secede again... but they did, and it's political masturbation to argue otherwise.
As part of the US, Texas is a big, wealthy, relatively powerful state. As an independent nation allowed to secede only after assuming direct responsibility for its prorated share of the national debt -- without enjoying the benefits of having its own dollars accepted as a world reserve currency & the ability to write checks to itself as payment -- it would be crushed into financial ruin and bankrupt before its first official flag-raising ceremony in Austin.
> And perhaps that is because they plant millions of the same species/strain with no natural variation?
Actually, it's much worse than that. The evolution clock for Florida's orange trees stopped DECADES ago. Florida's citrus trees are pretty much ALL clones. I don't think it's even possible to grow a viable orange tree from seeds anymore.
> If we wipe out one year's corn or wheat we're going to experience starvation on a scale larger than the potatoe famine.
It depends. Corn probably has a fair amount of headroom, just because *so much* of it gets used to make ethanol & high-fructose corn syrup. The likelihood of "an entire year's crop" getting wiped out is pretty slim. The majority of corn might get grown in places like Iowa because that's where it's the cheapest and most cost-effective to grow it, but there were wholesale crop failure in places like Iowa and Nebraska one year, you can BET farms in Florida, Arizona, and California that wouldn't normally bother would seriously consider growing it for a season or two. The likelihood of some virus or insect making its way from Iowa to California or Florida within a single year is pretty slim, especially if farmers in those states were determined to keep their own fields quarrantined.
In countries like the US, the biggest fights would be between farmers and companies like Monsanto if, for example, they had a strain of corn that wasn't affected, but couldn't be supplied in full commercial quantities for a year or two. You'd have farmers cloning sprouted plants, Monsanto sending out the lawyers with C&D notices, and farmers introducing refused purchase orders in court as evidence that Monsanto suffered no actual damages since it was unable to actually fill their orders as placed.
An equally bad problem (if asymmetric-impact) would be if some virus or pest suddenly wiped out an entire generation of "Round-Up Ready" crops, forcing large-scale commercial farms to go back to more conventional strains... and try to raise, tend, and harvest them with a labor force that no longer exists, and expensive capital-intensive machines that were no longer relevant. This scenario would have almost no impact on farmers in central Africa who can't afford large-scale capital-intensive farming techniques anyway, but it would be devastating to American agribusiness.
The truth is, though, that nothing short of outright war between equally-potent nation states occurring SIMULTANEOUSLY with crop failure is likely (in the US or Europe, at least) to have the kind of devastating impact some like to predict. Fresh vegetables, fruits, and grains would become expensive overnight, but there's actually a pretty huge amount of highly-processed food already in the pipeline, stored in warehouses ready to ship to Sam's Club & McDonalds, and the actual impact for most Americans (the majority who wouldn't even know what to DO with fresh food if you put it on the kitchen counter in front of them) wouldn't be an immediate crisis so much as a long-term lopsided increase in food prices that began months after the actual crop failure, and persisted even longer after it was resolved.
Suppose tomatoes had wholesale crop failure for several years nationwide. The most immediate consequence? Restaurants would quit putting sliced and diced tomatoes on food unless you specifically asked for it. If it persisted, they might start to charge extra for them as a premium ingredient (a-la-bacon). Ketchup and spaghetti sauce? Please. They barely have statistically-meaningful amounts of actual tomato in them to start with. In the US, at least, they're basically artificial flavoring, high-fructose corn syrup, tapioca, salt, spices, and FD&C Red #40. Add hot peppers, reduce the tapioca & guar gum, and it's taco sauce. You'd see restaurants quietly deprecating chunky salsa, replacing it with melted cheese and red goo with the consistency of spaghetti sauce, and gleefully advertising their new, improved CheesyZingySalsa.
Wheat would be harder to replace... but so much wheat is grown around the world, the main consequence of wheat failure in north America would be the disappearance of wheat from African supermarkets, and American restaurants suddenly becoming VERY enthusiastic about their new "low-calorie, low-fat sandwich wraps".
Corn? Well, for starters, probably half the corn grown in the US ends up as
The fundamental problem is that cars are kind of like American Android phones... consumers are mostly powerless to do *anything* to fix vulnerabilities, and carriers won't do anything they aren't forced to do by law. And the law itself is only slightly less castrated and toothless than consumers.
If a car suddenly becomes something you can't safely use (or at least leave unattended in a non-secure location), there's no meaningful immediate recourse for consumers, and that's a real problem with real consequences for real people.
The solution isn't to contact the automakers and keep it a secret... the solution is to quietly meet with representatives from Allstate, Progressive, and State Farm (among others), and demonstrate it to THEM. They're better at getting the attention of Congress and the auto industry, and you can BET they'd be on the ball if they thought there was a vulnerability could leave them on the hook for billions of dollars worth of insurance claims.
Hmmm. Interesting. But now the million-dollar question... are signed distance fields fast enough on any current high-end Android or IOS hardware to actually render a full page of arbitrary text that includes multiple fonts and multiple styles of those fonts? All the examples I found via Google use it to render just a few glyphs at any one time.
Realistically, if you're rendering English text and you stick to pre-loading characters that are actually used or likely to be used, you're going to need about 80-90 for each font+style. For technical books, that realistically means at least 3 complete font sets in normal, bold, italic, and bold+italic -- usually, a serif font, a sans-serif font, and a monospaced "console" font with disambiguated zeroes and ones for code examples.
Ram-permitting, text has somewhat of the advantage that you can preload your font data, then reuse it from "scene" to "scene". On the OTHER hand, with a game, you can define a large scene, then just move the frustum around to change what's visible from frame to frame. With text, a new page means defining and rendering an entirely new scene from scratch... and doing it at 60fps without lag or latency is no small feat. Add page-turning visuals, and you have a task that could bring the development team from a game like "Battlefield 4" to their knees and have them begging for mercy (at least, when you tell them it has to work acceptably well on a MSM8960's Adreno 2 or a Tegra3, even if the device HAS a gig of ram and dual/quad-core 1.5GHz+ CPUs).
Text might not be sexy, but rendering print-quality text in realtime is *hard*. We've settled for less in the past, because text on a video display was always regarded as a second-rate ghetto that was good enough for nasty throw-away stuff, but was never expected to approach the visual quality and responsiveness of printed text on paper. Video displays have now largely caught up (like the iPad's retina displays), but the hardware we need to put them to proper use is still very much in its infancy. The industry used the slow update times of e-ink as an excuse for years, and when displays capable of crisp, high-resolution text display and instantaneous updates appeared, the underlying hardware wasn't even close to being ready to take advantage of it for anything besides sequential page-by-page reading of novels.
> You can use 3d acceleration to generate a scene or just draw in 2d.
Until hinted scalable fonts enter into the equation. Then everything goes straight to hell. If you tried to write pixel-shaders to define every single letter of every font, style, and weight used on a page in a way that's legible at 8 pixels tall and smooth & curved without artifacts at 100, and had a hard rule that you couldn't do anything that couldn't be fully-rendered from definition to rendering in 1/60th of a second, you'd run out of ram, time, and/or triangles LONG before you got to the end of a technical ebook's first real page.
The fundamental problem is that triangle-based acceleration is the *wrong* kind of acceleration for high-quality text rendering. Fonts need B-spline acceleration. Rendering text with triangles is like trying to emulate a Soundblaster Pro with a Gravis Ultrasound was back in 1993.Sometimes, you need hardware that's optimized for the specific task at hand, and trying to generalize it into something else just makes matters worse.
Oh, the days of using separate video cards for 2D and 3D support. It was "cool" to have a setup like that, but somehow I was never interested and held out for the TNT2.
> Oh, the days of using separate video cards for 2D and 3D support.
> It was "cool" to have a setup like that, but somehow I was never interested and held out for the TNT2.
And thanks to mass-market consumers who did the same thing, we ended up with video GPUs today that are basically a pimped out 3DFX stapled onto a dumb framebuffer, with no real 2D acceleration to speak of.
Instead of getting hardware-accelerated B-splines and the ability to render subpixel-hinted scalable fonts via hardware in realtime sometime around 2006 like we were supposed to (going by ATI's roadmaps), we have Android and IOS hardware built around GPUs that couldn't render a full page of hinted dealiased text a-la-Postscript to display memory in 1/60th of a second if the future of their manufacturers' companies depended on it. Because 3D is trendy, hot, and sexy, and 2D isn't.
Joe Sixpack doesn't know what a B-spline or subpixel rendering is, but he knows that 3D is "cool", and a GPU that has "more triangles" is better (the same way he "knew" a 1.8GHz Pentium 4 was better than a 1.1GHz Pentium III Xeon, and even ran out to buy a new laptop with one).
That's why Android & IOS-based e-readers suck for interactively reading technical books that require constant page-flipping. They lack 2D spline acceleration, so they have to do everything via brute CPU force. They're too slow to render pages from scratch in realtime with "real book" aesthetics, they don't have enough memory to pre-render the whole book to ram, and they're too slow to fetch entire pre-rendered arbitrary pages from microSD in 1/60th of a second or less(*).
(*) The fastest microSD interface on any known Android phone maxes out around 25MB/s.... and a 32-bit 1280x800 bitmap weighs in around 4MB. Real-world Android phones like the S3 generally max out around 17MB/s. The fastest UHC-1 Sandisk Extreme cards have a theoretical max of 95MB/s, which STILL isn't fast enough to fetch the ~187MB/sec required for realtime brute-force 1280x800x60fps @ 24 bits. In theory, the 62MB/s required to fetch 8-bit grayscale at 60fps might be do-able on a future phone with UHC-1 microSD, but no current device can do it.
> Oracle is a rapacious, money-gobbling machine of a company.
No, they're a vast multi-level marketing scheme that happens to sell a database or two as a side venture, and recently bought a product (Java) that doesn't require an army of highly-paid consultants who have to tithe 20% of their income to them in return for mandatory certifications and licensing just to keep it running.
The big "gotcha" with Java is that Sun's license for Java was always pretty nasty, but they generally looked the other way and ignored all but the most egregious violations of it unless you were trying to use it in embedded devices. Oracle, in contrast, intends to enforce it to the letter. The only consolation prize with Java is the fact that most of Oracle's legal leverage over OpenJDK comes from patents rather than copyright, and software patents -- while questionable in most cases -- at least have the benefit of not being de-facto eternal & endlessly extended every time Steamboat Willie is at risk of becoming public domain. So in another 10-12 years, we'll be able to unambiguously do anything we like with Java, as long as it's cleanroom-engineered and called "Bali" or "Sumatra" (the islands flanking "Java").
> Bottom line, if you have N digital copies then what is the benefit of keeping the original DVD over one N+1 digital copies of the DVD?
The original DVD was pressed. Digital copies depend upon either brittle spinning hard drives that are cost-prohibitive to repair and recover data from, flash media that's the equivalent of a leaky bucket (especially MLC), organic-dye based optical media that shift over time, and manufacturing processes whose lifetimes are absolutely 100% speculation to begin with.
The more diversely-redundant your backup media, the more likely it is that you'll have enough of it survive to do high-quality archival restorations later. If you have a pressed disc, a non-LTH BD-R, and a hard drive backup, the pressed disc is the most likely out of all of them -- barring some bizarre manufacturing or design flaw -- to still work with then others have failed.
Hard drives are convenient and cheap (now), but they are NOT long-term passive storage media. You CAN NOT just throw a hard drive in a drawer for 20 years and expect it to work, assuming you can find something compatible to attach it to and software to interpret it with. Ironically, a modern version of the old SyQuest cartridges that was used only under cleanroom conditions would probably be the most robust of all.
What kills hard drives is mechanical failure combined with the near-impossibility of affordable repair. If Western Digital & Seagate did their own in-house low-cost repairs using their own parts, employees, and tools, hard drives would actually be VERY repairable, and permanent data loss would be rare barring total destruction. Unfortunately, they don't, and what we HAVE is a hard drive data-recovery industry that's basically built around reverse-engineering, parts-cannibalism, and black magic. Companies like WD could EASILY design drives to be serviceable, so that in a worst-case, you'd send in your drive for a warranty repair, they'd do some quick software tests to confirm hardware failure, remove the platters, mount them in their own recovery unit, image them to your new replacement hard drive, and ship it back to you. But no, hard drives are simultaneously engineered to be as cheap & likely to fail as possible, with no cost-effective means of data recovery when it happens.
Higher-end drives come with longer warranties, but warranties are worthless for data. The one thing the industry refuses to do is make drives that are expressly DESIGNED to fail in repairable ways. I have rather strong feelings about this matter, having gotten bitten HARD by multiple drive failures in a row that left me one drive away from lifetime data loss (well, ok, I did have older scattered backups, but that one drive was the closest thing I had left to a coherent, semi-ORGANIZED backup copy that wouldn't have left me playing data-archaeologist for 20 years).
BD-R (non-LTH) is a phase-change medium, which is pretty much the best thing we have now. Even better, the media is decoupled from the read mechanism, so failure of the read mechanism won't render the disc itself orphaned.
Another big problem that bites lots of people with CD-R specifically... CD-ROM/XA Mode 2, Form 2... and the willingness -- if not eager default -- of early software like Nero to allow users to disable error correction to gain a few more megs of storage space without making it abundantly clear to those same users what the consequences of doing it were. Lots of people went for YEARS burning discs that would develop hard errors at the slightest scratch without realizing WHY it was happening. They just attributed it to drive/media flakiness, and didn't realize that it was their burning software setting them up for disaster.
Auxiliary sound channel:
http://www.youtube.com/watch?v=_iTXuwJpeV0
> For born-analog content, always keep the original physical copy.
And if, for whatever reason, you CAN'T retain the original copy, oversample the bejesus out of it, store it with the most lossless compression you can, and sample it multiple times, in multiple ways, with multiple scanners/digitizers/capture cards.
Case in point: videotape. If your goal is to merely capture a copy to casually watch years later, and you don't mind having a digital copy that's demonstrably worse than the analog original, just about any cheap one-click capture card will do. BUT... if you want to stop the clock and capture a copy that doesn't visibly look worse than the original, compress it to MPEG-2 it at a MINIMUM of 6,000kbps VBR, at 720x480 (or 540, if it's PAL). Yes, we all know that VHS is low-res. The problem is, it's also noisy, and noise doesn't compress well. If a 720x480 video encoded with VBR and the maximum DVD-compliant bitrate available looks as good as the original VHS, consider yourself LUCKY. Regardless, never downsample the vertical resolution. Horizontal resolution with VHS is open to debate, but your vertical resolution is known and definite -- 480 (525) if it's NTSC (or 60hz PAL-J), and 540 (576) if it's PAL.
Now, we get into my #1 beef with Blu-Ray -- its despicable lack of compliant encoding options for 480i60 and 540i50 source, so there's really no way to profoundly oversample VHS, yet retain casual "stick in the disc and hit play" usability that can be watched directly in a Blu-Ray player, but someday used for restoration efforts.
If you REALLY want to capture a videotape and preserve every possible nuance for future restoration efforts before the tape degrades further, horizontal oversampling of VHS is mandatory. If Blu-Ray supported a hypothetical 1280x480 resolution with 4:4:4 chroma (remember, with VHS, your vertical resolution is your "good" resolution... it's 480 or 540. It's your horizontal color resolution that completely sucks. Chroma-encoding variants that split color between 2 scanlines will ALWAYS visibly-degrade VHS captures. Don't do it. )
In a perfect world, there would be a near-line archival video format designed for videotape captures that directly sampled the full-bandwidth FM signal coming off the tape (preferably, at 3 slightly different trackings... spot-on, plus one slightly "off" each way). A few times, I've tried to work out the math and figure out what kind of insane sample rate would be necessary to try magnetically sampling the tape straight-on, with multiple overlapping rows of microscopic read heads, and sufficiently-high density, to literally map the flux of every flake of oxidized metal on the tape and allow future offline reconstruction by digitally simulating VHS's diagonal read path, so that even if the original tape were destroyed, a future restoration could be performed against the digital copy. I have a hunch that something like this is actually do-able now (and probably would have been do-able a few years ago, had multi-terabyte hard drives been affordable), but I've never even read speculation online about such an archival device.
The point is, if your goal is to someday do color restoration, resolution-enhancement, or whatever, taking advantage of analog's tendency to keep surprising us with ghostly hints of lost detail that we'd always assumed were gone, you need a profoundly-oversampled copy with basically NO lossy compression (or oversampling so profound, you can downsample away the artifacts). Why do you need so much oversampling depth and detail? True hardcore restoration and after-the-fact re-synthesis is going to depend upon what are known as "higher-order artifacts", and those are precisely what get mangled by any kind of lossy compression. Remember, lossy compression is designed to throw away bits in areas where your eyes won't notice detail. An algorithm looking for subtle field-ripples as evidence that a pixel nearby is supposed to be intensely red instead of brownish-orange won't see them if some earlier compression algor
> I have burned CDs from 1995 that still work perfectly fine.
If they're redbook audio CDs, and your definition of "work perfectly fine" is "I can stick the disk in, hit play, it spins up at 1X, music comes out, and the player doesn't totally gag", you might be right. Now try ripping the disc using software that can monitor the realtime bit error rate. You'll probably be *horrified* to see how high it is.
Redbook audio CDs are very robust, even when their bits are rotting all over the place. They were designed in an era when hardware couldn't do much in realtime, so they bent over backwards to make sure they had a "plan B" to make sure the show would go on after the disc got scratched, dirty, or whatever else happened to it. They were designed so the audio data is interleaved in a way that when a read error occurs, the left and right channels get merged for 1 sample. A redbook audio CD has to be nearly *destroyed* (cracked, melted, fried, whatever) before it literally won't play, as long as the player is able to find the lead-in and sync up to the spiral track.
It'll start to sound "rough" and lose channel separation, but things have to be pretty bad before it will LITERALLY stop playing. At least, as long as the player itself is faithfully following the original redbook audio specs, and isn't trying to realtime-rip the audio to a ram buffer and play it back from there (which is what some, if not most, new optical-disc media players do TODAY). I have plenty of CDs that new players choke on and refuse to even try playing, but yet my 25 year old antique CD player that cost something outrageous like $600 or $800 when new, can play just fine. Apparently, it's because first-generation CD players were precision hardware that could blindly track a CD spiral as long as the disc itself was 100% within spec, whereas new players depend upon realtime error-analysis to stumble and wobble around, and make up for the fact that discs no longer spin precisely, and worm-gear optical assemblies no longer track with precision measured in microns.
That said, my experience has ALSO been that CD-R discs manufactured in THIS century are less likely to rot and become unplayable in new drives, but are more likely to have major problems with old players. The old players were precision hardware, and assumed the discs themselves were manufactured to precision specs. The first-gen CD-R media had dye that deteriorated over time, but their spiral tracks were spot-on, just like pressed discs. As drives got better at handling sloppy tracking, the discs themselves became sloppier.
Net effect: first-gen redbook audio CD-R media is likely to play with acceptable audio quality on an old CD player from the 80s or early 90s, but be unplayable on many modern drives & be un-rippable on most drives (some will allow you to spin down to 1X & emulate the playback mode of a legacy player if you're running a sophisticated ripping app). Newer discs that are still old will probably skip and have problems playing on an old player, but might still be equally bad on a new one. When today's bargain-bin CD-R media is 10 years old, it will probably be unplayable on anything, the same way my old VHS tapes from the 80s still play fine, but VHS tapes recorded after ~1998 are largely unplayable on anything I can find.
TLDR point: the storage life of "last-gen" CD-R media is likely to be better than first-gen CD-R media was at the same age, but enormously WORSE than that of the best "turn of the century" CD-R media (the golden era when quality standards were still high, and the worst faults of the first-gen media were addressed. Any box of CD-R media you buy TODAY is probably shit of the worst kind. The best media you can buy TODAY for long storage life? Non-LTH BD-R single-layer discs. But MAKE SURE they aren't LTH... most manufacturers don't go out of their way to scream, "These discs are LTH garbage!"
In South Florida, even the urban-ish parts? Water moccasins. Even in central Miami, they swim up the river, the Tamiami Canal, or large drainage ditches from the Everglades & *somehow* climb out into back yards. The lights also help deter break-ins.