The problem is, inflation isn't necessarily a bad thing. Having a LOT of inflation, or unpredictable inflation, is bad... but having small, consistent amounts of inflation basically ensures that people with lots of money are forced to actively invest it, instead of just locking it away in a safe and waiting for its value to go up. Inflation isn't good, but DEFLATION is DEATH to any modern economy.
With 2% annual inflation, an investment likely to pay 4%-6% dividends is attractive. With 2% annual deflation, the economy implodes because the best low-risk investment you can make is to hoard cash and do nothing at all with it. And with greater deflation, you end up at a point where nearly anything you could conceivably invest in is almost guaranteed to lose money relative to locking it in a safe. Think back & remember what it was like to go shopping at a store in 2009... everything was cheap and on sale, but stores started to literally RUN OUT of things to sell & DIDN'T replace them. Once they were sold out, you were fucked unless you could find it online... and even finding things online merely blunted the sting and spread it out slightly. Had the recession continued for longer, even Amazon would have started running out of things to sell.
Not to mention, the apparent outright BUG in MtGox's wallet implementation that caused some large number of bitcoins to be outright LOST forever.
The last time I checked, not even a poorly-designed wallet for paper money is capable of insidiously shredding or destroying it upon insertion, and modern paper money contains enough cloth, plastic, and security threads to ensure that even paper money that's been through a washing machine is probably still good enough for a bank to exchange for new bills.
Dealing with Bitcoin today is like being in a Medieval economy where everything not involving barter has to be done with cash, but with zero modern banking infrastructure to deal with storage and transportation. A sane non-drug-dealer would never, ever, EVER walk around with $20,000 in cash, because you might as well paint a target on your chest. With Bitcoins, you're basically relying on not being the victim of a metaphorical home invasion robbery just by having them in your possession. No Bitcoin "bank" is safe the way a FDIC-insured bank is a safe place to store up to $100,000... and more importantly, the existing Bitcoin "banks" are hopelessly outgunned by thieves. If a bank with the resources of Citibank or Deutsche Bank can't consistently outgun hackers, a "mom & pop" online bank with 20 employees has ZERO long-term chances of succeeding at it. At some point, we're going to see bad guys literally taking hostages, and killing employees until they unlock wallets for them.
Banks get robbed and have money stolen from them every day... we just don't care (much) anymore, because their losses are insured, so when it happens, we can just say, "ugh, shit happens." With cryptocurrency, there IS NO insurance.
As an American, Britain's government is simultaneously fascinating & terrifying.
On one hand, it usually behaves with a professionalism & responsibility that we can only envy.
On the other hand, Britain's government seems to be TERRIFYINGLY vulnerable to the legal equivalent of a multi-stage malware attack (with Parliament passing a sequence of innocent-looking laws that lay the groundwork for the vulnerability, then triggering their viral legal payload to fend off opponents by neutralizing any laws that might be used to fight them in court).
Just to add... this is basically a somewhat new application of several older technologies:
* a wavefront scanner that maps the actual shapes of the cornea, lens, and retina
* raytracing software, to calculate the shape of a complex, 2-surface lens. The hardest part about developing this software was learning where to step back and NOT try to fix some specific higher-order aberration. Glasses will never be positioned precisely (they slip, frames bend, etc), and if you try too hard to fix HOAs precisely, you'll make matters WORSE if the lenses deviate from their ideal positions. With scleral RGPs and Lasik, you can be more aggressive because the resulting lens is more stable. The trick was finding the happy medium that makes things "sharp" without making things "weird" when the glasses slide down your nose or the frames get slightly bent.
* CAD/CAM, allowing a robot to precisely grind a complex shape (calculated from data from the above) into two sides of a lens.
* Much of the lens theory originally developed for progressive bifocals... but applied to precisely correct focus, mag/minification, and astigmatism instead of merely transition different magnification strengths.
IMHO, if you have astigmatism, freeform digital aspheric lenses (when expertly-fitted) are a huge improvement over the mass-market norm... but if you have astigmatism AND need bifocals, it's absolutely a non-negotiable requirement. The catch is, you'll probably want to experiment with a few pairs of cheaper progressive lenses first to see what progressive shape you prefer (you might hate one layout, but be ok with another), and THEN spend the cash replicating that shape (but with improved optics) into freeform lenses after determining what it is that you actually *want*. It really isn't something you can determine through research alone... it takes some live experimentation, and you'll rarely be happy with your first attempt.
I'd argue that for every person who buys ill-fitting contacts with incorrect parameters with a fake prescription, there are THOUSANDS of people with perfectly valid prescriptions risking injury by wearing old/damaged contacts because they can't afford to replace them as frequently as they should. Low-cost replacements are a GOOD thing.
The truth is, most disposable soft contacts have SO MANY engineering compromises (especially toric ones), even flawlessly-fit lenses have pretty mediocre results, so comparing the best and worst is more like "kind of mediocre, vs not great"
It's like glasses... any halfwit with a ruler, a collection of lenses, and 5-10 pages of notes on fitting theory can come up with a reasonable set of +/- sphere values that are a net improvement over "none at all". Mitigating small amounts of astigmatism when looking straight ahead isn't much harder. Most low-cost glasses (under $100 lenses) aren't much better than this anyway.
So, what *does* require a skilled optometrist with substantial gear? 2-surface freeform aspheric lenses. Normal (sphere-only) lenses are molded as-is, cut, and polished. Cheap (sphere+cylinder) lenses have the sphere molded into them + the cylinder ground into one side. In both cases, visual magnification or minification occurs, which makes new lenses hard to "adapt" to & compromises depth-perception. But if you grind curves into BOTH the front AND rear side, you can SIMULTANEOUSLY correct focus errors AND neutralize-out magnification/minification.
For astigmatism, that's still not quite good enough... to correct sphere & cylinder (while neutralizing-out magnification/minification) across the entire lens field (vs "straight ahead"), you need to grind complex curves into both surfaces that are calculated via ray tracing... AND know how to properly measure additional parameters like the angle at which the lenses are tilted & their precise distance from the pupil... not all stores selling "HD lenses" do this properly, and if they don't, the results can be WORSE than cheap molded lenses. When done correctly, glasses with 2-surface freeform aspheric lenses won't distort your peripheral vision or distort geometry (or at least, won't do it nearly as badly as cheap glasses). You'll put them on, and things will just be sharper.
2-surface raytraced freeform lenses aren't something you'd WANT to buy online... a "normal" prescription (sphere+cyl+axis) doesn't have enough information, and every pair of non-identical frames will produce slightly different measurements for things like tilt, vertex, etc.
The problem? In the US, at least, chain vision stores are fixated on promoting things like "ultra-thin" lenses and "no-line progressive bifocals". The same technology behind them can be used to make near-ideal lenses for customers with astigmatism, but most front-line sales associates at those stores have no idea what you're even *talking* about if you say "custom-raytraced 2-surface (digital/HD) freeform aspheric lenses". The OPTOMETRIST might... but s/he's not the one who'll do the half-dozen extra measurements required to complete them. Most times, it'll be done by an employee who's literally winging it & doing it for the first time in weeks/months/ever.
The moral: if you have astigmatism & want genuinely better glasses, find an opthamologist who does Lasik (ie, who has the eye scanner & above-average training/experience) with on-site store & ask about "custom digital/raytraced/HD 2-surface freeform aspheric lenses". If the opthamologist looks confused or does anything besides confidently grin with delight because he'll get the satisfaction of fitting the best glasses money can buy... go somewhere else.
They won't be cheap, but you'll never be able to stand normal cheap lenses again. In theory, an optometrist could do it... but because the scanner is so expensive, they're usually only found at places that do Lasik (and by extension, have at least one opthamologist) since you NEED one for Lasik, and they're too expensive to buy JUST for eyeglass-fitting. And chain stores tend to "simplify" the fitting process, leaving you with compromised lenses.
Japan has a MUCH bigger problem with earthquakes & tsunamis than hurricanes (Tokyo & northward is about the same climate as the Northeastern US & maritime Canada). Tsunamis bring saltwater inland (very, very bad for underground power lines). Earthquakes shear underground power lines apart (or stretch them & cause subtle, harder-to-troubleshoot flakiness).
If the current fires encountered a large expanse of concrete devoid of anything directly combustible, how wide would it have to be to actually stop the fire's spread?
As I understand it, once uncontrolled outdoor fires reach a certain size, they act kind of like weak tornadoes that lift flaming objects high into the air & hurl them out to areas that might be several thousand feet away (enabling the fire to jump over thing like freeways, canals, etc).
If a house in the middle of an affected neighborhood had reinforced concrete walls & roof, plus Miami-grade impact-glass windows, would the heat of the fire as it burned down the neighbors' houses cause the concrete house's interior to combust anyway (like food debris in a self-cleaning oven)? Would ICF construction plus roll-down steel shutters keep the interior cooler, or would the intense heat just cause the ICF styrofoam itself to melt or combust?
I know that conventional wisdom is that individual homeowners are helpless against a fire, but I remember reading about one guy in California a few years ago who put sprinklers on his roof & surrounding yard, connected them to the faucet, and left it running when he evacuated. When he got home, his home had major "baking" damage... but his neighbors' homes were literally burned down to the scorched earth. I think some local official later decided to be a dick & fined him $10,000 for violating water-conservation rules to discourage others from trying to do it in the future.
Oh, and I forgot the other detail... $9.99/mo is the 6-month "promo" rate, with $69.99 regular price, 2 year contract, and outrageous ETF. Ditto, for the $99 plan... $99 for 6 months, jumping up to $149 thereafter with 2 year commitment.
I just realized I left out a major detail from my "$9.99/mo (plus fees)" scenario -- non-partner services wouldn't be BLOCKED... that would be too blatant & would guarantee pushback. They'd just be throttled to some point that's not glacial, but not fast enough to sustain realtime high-quality HD streaming video (say, 3-5mbps).
If your choices were:
a) $9.99/mo (+ fees) for (up to) 100mbps to "partners" (with nearly every major service being a partner) and 3-5mbps to "everyone else"
b) $99.99/mo (+ fees) for (up to) 100mbps to "partners" and 10mbps to "everybody else"
c) $249/mo (+ fees) for (up to) gigabit speeds to everyone
(all with "up to" gigabit local loop down & ~20-50mbps up... throttled speeds are per-host)... you'd probably agonize long & hard between 'a' & 'b' (unless you really, really NEEDED 'c'). And you'd probably end up grudgingly choosing 'a' & hating them for charging SO MUCH MORE just to get the next step up.
ARM is only cheap & power-saving if you use it for undemanding uses on its own terms. An ARM that's beefed up to specs genuinely comparable to a desktop i7 uses as much power as (and usually costs a LOT more than) an actual i7.
Intel architecture is more efficient (behind the scenes, where it's REALLY executing RISC microcode PRETENDING to be x86) than many give it credit for. And x86/AMD64 and Windows have literally DECADES of mutual optimization for each other. It's going to be a very, very long time (if ever) before ARM's best can truly equal (let alone, outdo) Intel's best when it comes to Windows.
> Why does everyone who is pro-Windows do "heavy video" and "photo editing"?
Probably because support for hardware codec-acceleration has historically been one of Linux's major weak points. Most video codec hardware acceleration requires proprietary binaries and/or royalty payments (at least, officially).
This is now probably a bad example, but I remember that ~4 years ago, Windows Media Center could play h.264 1080p60 video without issues on my Thinkpad T61p... but the same stuttered & dropped frames if I booted Ubuntu and tried to watch the same video using VLC. 720p24 or 720p30 was pushing my luck, though 480p24/30/60 generally did OK.
Why? Windows took advantage of nVidia's hardware h.264 acceleration. Linux (in general) and VLC (specifically) didn't.
If you're importing lots of video source shot on consumer-grade cameras, the difference between "instantaneous" and "grind everything to a halt for several seconds and spike the CPU at 99%" is a big deal.
I believe recent VLC releases for Linux WILL (now) take advantage of h.264 hardware acceleration if the driver supports it... but AFAIK, the open-source nVidia drivers don't support it.
Playing videos isn't the same as editing, but editing is basically a superset of playing (importing, previewing, etc). If the platform can't play h.264-encoded videos smoothly, using it to EDIT video is going to leave you in a world of pain.
I fully support net neutrality, but one thing that seems to keep getting lost in discussions about Comcast vs Netflix is that back when Netflix live HD streaming was *new*, Netflix users were consuming *inordinate* amounts of bandwidth relative to "everyone else". On one hand, it was Comcast's fault for overselling their capacity... on the other hand, Netflix was an easy target PRECISELY because it was such a big, easy-to-see target.
The danger isn't that Comcast is going to start charging higher fees for Netflix, Hulu, Facebook, Youtube, and SlingTV... the danger is that Comcast will roll out some $9.99/month (+ thirty dollars of unavoidable fees) plan that provides "100mbps" connectivity ONLY to big, established companies willing to pony up and subsidize their cheap plan, then jack up the price of REAL 100mbps connectivity (to all internet hosts, including RDP servers and non-Facebook/Netflix/Hulu/etc) to $150/month.
It's a variant on Gresham's Law... cheap shitty service drives out good service & makes buying something even *slightly* better than "total shit" WAY more expensive.
Look at laptops... you can buy a shit netbook with 1.2ghz cpu, inadequate ram, useless keyboard that misses 1 in 20 keypresses, and hard drive that isn't even big enough to survive the next big annual Windows update for $199... but getting one that's meaningfully better (i5 or i7, 8-16gb ram, 256gb+ SSD, etc) increases the price to $1,000-2,000 because all of those shit netbooks soak up 99% of the economies of scale & make GOOD computer hardware 4-10x more expensive. And the pervasive existence of under-powered hardware induces companies like Microsoft to take away nice things like Aero Glass & replace it with ugly, awful things like Metro for the sake of making it run semi-acceptably ON shit hardware.
Disney could make a strong argument under 'equity' that bundling download codes with a physical disc as a freebie adds a small amount of value to the disc & takes away little/nothing from their potential sales, but a download code on its own has a MUCH higher value to the purchaser & probably represents a major revenue loss to the studio.
There's also a solid legal argument on licensing alone. Years ago, Microsoft won plenty of lawsuits against businesses & VARs who installed Windows using Enterprise site-licensed installation codes (using legit, purchased codes... but in a way not allowed by the license).
Hell, years ago, I knew someone who owned a video rental store in Miami. He had lots of foreign films (not released on NTSC VHS in the US). He'd buy retail tapes in Europe, make NTSC transfers himself, and rented both PAL/Secam original + NTSC transfer together as a single, inseparable unit... and STILL got sued by Sony or Disney (don't remember which one). He eventually got them to drop their lawsuit by agreeing to never do it with that studio's tapes again. Their lawyers were, shockingly, fairly open to reason & agreed that he was in a gray area and had made a good-faith (albeit unsanctioned) effort to do it legally, so they dropped the absurd multi-million dollar claim). It still hurt him, though, because it affected about a third of his foreign films.
My point is, even if you can prove the copyright owner didn't lose a cent in revenue, they STILL enjoy nearly absolute legal rights (in the US) to say 'no'. What Redbox is doing goes WAY beyond VHS-NTSC transfers of VHS-PAL/Secam tapes and results in actual lost original sales. If Redbox has a legal leg to stand on, it metaphorically has osteoporosis and polio.
The most tragic thing about Windows Mobile is that 5.x and 6.x were ugly as sin... but at their core, it was a good, stable, reasonably well-implemented (by real-world evolutionary standards, at least) OS. 7.x was going to be as pretty as Cupcake/Donut/Eclair Android.
As a practical matter, it was almost more de-facto "open" than Android is now. Sure, you can build AOSP from source, but if you want a device that can actually work as a non-crippled & dysfunctional "Android" phone, you need Google's very, very proprietary binaries. And hardware that doesn't lock you out of the bootloader. And compiled binaries from the SoC vendor... which unlike x86/AMD64 Windows, has basically zero compatibility with drivers for different kernels... change so much as a compiler-optimization flag, and existing binary kernel modules go "boom".
Microsoft *could* (and *should*) have had an official "App store", but like Android, WM imposed no mandatory restrictions on what you could download & run.
All MS *really* needed to do with WM7.0 was release it, and in WM7.5, they could have documented (properly) the way for third-parties to write things like a "phone" app (and added a C#.netCF API to allow apps like that to be written in C#.netCF).
If they'd stayed on course & done that, they would have probably gotten 80% of first-gen Android users to come back after one Android phone (because first-gen Cupcake-era Android phones WERE pretty dire compared to what WM7.0 was supposed to be). But no... Microsoft bought Danger (Sidekick OS), wasted a year porting it from Java to C#, then dumped WM for an inferior new OS that was superficially-prettier, but couldn't even do most basic BLUETOOTH-related things properly (a deficiency that pre-existed with Sidekick).
Microsoft NEEDS Bill Gates to pul a "Steve Jobs", come back and save Microsoft from itself. Pretty much every single thing Microsoft has done since Gates left MS has been bad. Windows 95's UI changes were genuinely "bold", "courageous", and a definite step in the right direction. Windows 8 (and Metro, and everything since) was NOT.
Microsoft has totally come unhinged, and they're destroying PCs with their craziness by refusing to acknowledge that the way people use high-powered desktop & laptop computers with multiple monitors is fundamentally different from the way little kids & tech-clueless adults use tablets, and trying to force a "One UI for All" policy on everything just degrades desktops & laptops into big, oversized tablets.
What Windows really needs is something akin to Android Fragments, so you can make one UI for tablets, one for real computers, then let the OS sort things out at runtime.
It has nothing to do with Linux per se, and EVERYTHING to do with the fact that KDE and Gnome (ok, Cinnamon, which as I understand it is Gnome 3 with Gnome 2 functionality restored) increasingly replicates Windows' UI better than Windows *itself* does.
Gnome and KdeWin are potentially a way to take control of Windows' UI future in a way that tells Microsoft, "We're staying with Windows... for now, at least... but piss us off even just a little, and we're bailing on you for good."
Truth be told, Windows' handling of proprietary binary drivers beats Linux's handling (or more precisely, its refusal to handle them) hands down. As companies like Qualcomm and MediaTek increasingly come to dominate hardware, we're going to *need* Windows' superior driver model (that doesn't automatically break all existing binary drivers every. goddamn. fucking. time. the kernel has even the *tiniest* change made to it, the way Linux does).
The fact is, we *need* Windows to not suck (and not self-immolate), if only to keep Linux on its toes and ensure x86/amd64 architecture has viable future. If ARM wins, ${deity} help us all, because the open(-ish) hardware architecture we know as the "PC" just doesn't exist in ARM-land. ARM shit is all proprietary, locked down, and massively empowers vendors over users. ARM doesn't even have a fucking mandatory baseline standard for "partition a hard drive in a way that allows it to boot any arbitrary OS"... literally everything is vendor-specific. If companies like Qualcomm had their way, PCs as we know them would cease to exist & we'd all be reduced to appliance-users who can do what the vendor explicitly chooses to allow, and nothing more.
The worst part is, unlike Windows 7 - vs - 8/10, Windows 10 users can't say "no thanks" and refuse the change. Eventually, Microsoft will make it a mandatory update & ram it down everyone's throat.
Unless, of course, KDE or Gnome for Windows gets taken seriously by someone like Adobe (and Gnome/KDE *themselves*), and Windows users start running KDE/Gnome Desktop in a full-screen Windows window as the penultimate 'fuck you' towards Microsoft. Once apps like Photoshop are Gnome/KDE-native, it would only take one major annoyance by Microsoft to get users to ditch Windows entirely.
Microsoft appears to have learned *nothing* from its Windows Mobile debacle (where they had the #1 mobile OS with capabilities that were *years* ahead of Apple & Android... then threw it all away and rendered themselves *irrelevant* as a mobile-device OS). Microsoft appears determined to render themselves equally-irrelevant in the non-phone realm (but will probably take most of the market for "real" PCs into their grave with them... the desktop/laptop market can't survive another "Windows 8" debacle).
Java's continued lack of an unsigned primitive 8-bit byte type is just inexcusable. It makes doing anything that involves binding with C++ (e.g., OpenGL ES) *excruciatingly* painful.
I would love to meet one person... ONE... who has EVER thought, "I'm *so* thankful Java's primitive bytes are signed instead of unsigned. That was a bold, awesome design decision."
If Java had an 'octet' type of unsigned 8-bit values, my life would have been IMMEASURABLY easier over the past 20 years, and I would have been spared weeks of troubleshooting insidious bugs that happened because I forgot a "&0xff" somewhere (or accidentally put it on the wrong side of parentheses).
The problem is, there's NO WAY for consumers to reliably know how badly LED lights flicker without buying and testing them directly. There's no "gold-star seal for bulbs with DC driver and adequate capacitor size" to distinguish them from cheaper designs that flicker at 120hz or 60hz.
Cree bulbs are reportedly among the WORST-flickering... many of their bulbs exhibit BLATANT 120hz flicker that can be seen with the naked eye... no need to test with a solar cell + oscilloscope, because you can see it directly.
Phillips is a mixed bag. Some of their bulbs have flicker-free DC drivers, but it's mostly luck of the draw... apparently, they're yet another company that thinks it's totally acceptable to change designs & components without changing the UPC or SKU, so one batch's 5-star reviews won't necessarily apply to the next batch.
IMHO, LED street lights are *worse* than high-pressure sodium. Especially in areas like rural freeway interchanges that are lit up like a stadium, but surrounded by inky darkness. Bluish LED lights destroy your night vision for a minute or two once you're back in the dark. Orange sodium lights aren't nearly as bad. And because they *aren't* as directional, there's a more gradual drop-off in brightness as you drive away.
With LEDs, it's like, "Darkness... darkness... BAM! Blinding blue-white light. BAM! Pitch black, with a few seconds of night-blindness to boot." With HPS lights, it's more gradual, and the light doesn't affect night vision as badly.
Update... I think my TV (Mitsubishi wd62627, circa 2006) can overlay content from an OTA channel from its internal tuners over component video, but not over HDMI, DVI, or VGA... and not vice-versa. It's not well-documented, and only certain permutations are allowed. In other words, let 'X' be the subset of video sources that can be full screen. Let 'Bx' be the subset of video sources that can be overlaid over each X.
So... to answer your question more fully... I've NEVER seen a TV that could do PiP with ANYTHING besides its two internal tuners... and MAYBE analog NTSC/PAL from composite-video inputs. Component video? Never. HDMI/DVI? Never. PIP appears to be implemented entirely within the tuner subsystem, so it would have to be done within an external box (regardless of whether the display is a monitor OR a TV)
AFAIK, there's one device you can buy to overlay computer-generated content onto HDMI (it uses the leaked HDCP master key if necessary). In the US, its legality is murky... almost certainly illegal (under DMCA) if the source is encrypted... possibly illegal if it's not (HDMI is a superset of DVI, but piles on additional restrictions that *might* apply if actual HDMI components -- including HDMI-to-DVI plug adapters -- are present.)
AFAIK, there is NO consumer-priced hardware (legal or otherwise) that can overlay HDMI source over another HDMI source. I'm sure somebody makes a $10k+ rack module intended for things like stadium displays that includes a 200-page license restricting the way you can use it, but a simple "HDMI+HDMI PIP box" just doesn't exist.
IMHO, HDCP's absolute control fetish goes WAY too far. The fact that its licensing body even REGARDS "unauthorized scaling, overlaying, and combining content from multiple sources" as a serious problem demanding aggressive enforcement illustrates just how far over the top and absurd they are. Apparently, they fear someone with legally-licensed content might replicate something like 1980s MTV (overlaying the video in a box while the VJ talks, then expanding it to fill the screen), or (shudder) "violate the artistic integrity of a director" (by allowing a box that bleeps out 'dirty words' based on subtitles... yeah, this was actually a real lawsuit, and Hollywood won).
Remember, DRM isn't satisfied with merely preventing COPYING... it goes 20 steps further to control every aspect of how you're allowed to VIEW content.
Security is also "unfair". You, as a conscientious developer, can do everything "right", and get totally pwn3d *anyway* because of some widespread, system/platform/framework/library vulnerability (perfect example: Heartbleed).
The only way to improve the odds is for a development team to have one or more members whose ONLY job is to be aware of every thirdparty library/platform/os used by the project & literally research every single one, every single day, to become aware of vulnerabilities as they're discovered... and for projects not under active development, maintain a working build environment at all times so *somebody* will be able to rapidly build updates to the app.
You can't expect developers to do it. We don't have time to recursively scrutinize our projects' thirdparty dependencies on a daily basis... at least, not if we need to actually get any *new* development done.
I'm not kidding about the build environment problem. I have a couple of published Android apps (developed with older versions of Android Studio) that get hopelessly mangled by recent versions of Android Studio. Google has done a *remarkably* shit job of breaking old projects with no working migration path beyond, "spend several days creating a new project from scratch & manually re-create the old project in the new release of Android Studio". I've heard similar tales of woe about Visual Studio... if you're *lucky*, an old project *might* load successfully into the *next* release of VS, but if you let your old project go *too* long, it's not likely to even be importable by future releases of VS without getting mangled into oblivion. And Visual Studio sinks its teeth *so* deeply into Windows, it's hard to keep old versions of VS around (and working) in perpetuity (without having mothballed old computers with it installed), and even harder to re-create those old installations years down the road.
Back in the happy days of Netbeans, Java, and Ant, you could generally build any old Netbeans project from the commandline using its generated Ant script, even IF a newer release of Netbeans broke it. With non-Gradle Android Studio 1.x projects... good luck... but you're probably fucked. StackOverflow has THOUSANDS of questions about "how to salvage old Android projects last built with now-obsolete IDEs (Eclipse, AS1.x, etc)".
or keep the same TV, and buy a new tuner box. ATSC 3.0 is a superset of ATSC 1.0's codecs & hardware, so any new 3.0 box will be 100% compatible with ATSC 1.0.
The big difference -- this time, a tuner box will cost about as much as a Roku, not $299-1999 like an ATSC 1.0 tuner did circa 1999-2004.
My main complaint over ATSC 3.0's omission of on-the-fly mode switching is the fact that they could have EASILY worked around the HDMI problem by having the TUNER transform everything into a specific mode (if the user desires).
IMHO, the new standard SHOULD have required that compliant tuners be able to deal with content that freely switches between modes like 720p120, 1080p60, and 2160p30 on the fly as appropriate for the content (720p120 for live-acton sports that need realtime encoding, 2160p30 for primetime TV shows that are pre-encoded in non-realtime, 1080p60 for most other stuff), then leave it up to the tuner box to transcode everything into a single mode (say, 2160p120 or 1080p60) if the TV can't handle on-the-fly mode changes.
The difference is, real 2160p120 needs a staggering amount of bitrate & processing power to ENCODE in realtime to h.265 prior to broadcast, but scaling UNCOMPRESSED bitmap frames is computationally-trivial. If the BROADCASTER has to do it, everything that wasn't 2160p120 has to be re-compressed. If the BOX does it, it can scale the uncompressed frame, then dump it straight to HDMI.
I also wish ATSC 3.0 did a better job of embedding metadata to help advanced receivers un-do the damage caused by scaling & re-encoding prior broadcast. For example, if the broadcaster took 30fps source and tweened it to 60fps or 120fps, receivers should be able to tell the difference between "real" frames & "synthesized" frames so they can throw away the synthetic frames and implement THEIR OWN (better, or at least user-configurable) algorithm for doing it. Or if 50fps-native content gets telecined to 60fps, tag the original frames so someone with a higher-end receiver and display capable of native 50fps can just buffer 5 frames, ignore the doubled sixth frame, and output the 5 frames ~100ms later at real 50fps (or locally-tweened to 100fps).
Yeah, a receiver can "sort of" sniff out the duplicate frames by examining the existing flags, but in practice this only works WELL if the content is a 24fps Hollywood movie (specifically, the movie itself... extras usually break the algorithms). For other content, the algorithms completely fall off the rails due to sloppy editing or splices made further down the pipeline. Keep the source video pristine (or at least, easily-restorable to its pristine state), and let the VIEWER's hardware do the heavy lifting.
Perfect example: "stretch-o-vision" vs pillarboxing. With ATSC 1.x, there's no easy way for a receiver to auto-detect 4:3 content stretched to 16:9 so it can de-stretch and pillarbox it instead. If the broadcaster decides to pre-stretch some (but not all) content & you HATE stretching, you're just fucked... if you're lucky, you can manually toggle de-stretching, but the TV can't do it automatically. With proper metadata documenting upstream transformations that have already taken place, the receiver end could undo those transformations automatically.
Slashdot Mirror
The problem is, inflation isn't necessarily a bad thing. Having a LOT of inflation, or unpredictable inflation, is bad... but having small, consistent amounts of inflation basically ensures that people with lots of money are forced to actively invest it, instead of just locking it away in a safe and waiting for its value to go up. Inflation isn't good, but DEFLATION is DEATH to any modern economy.
With 2% annual inflation, an investment likely to pay 4%-6% dividends is attractive. With 2% annual deflation, the economy implodes because the best low-risk investment you can make is to hoard cash and do nothing at all with it. And with greater deflation, you end up at a point where nearly anything you could conceivably invest in is almost guaranteed to lose money relative to locking it in a safe. Think back & remember what it was like to go shopping at a store in 2009... everything was cheap and on sale, but stores started to literally RUN OUT of things to sell & DIDN'T replace them. Once they were sold out, you were fucked unless you could find it online... and even finding things online merely blunted the sting and spread it out slightly. Had the recession continued for longer, even Amazon would have started running out of things to sell.
Not to mention, the apparent outright BUG in MtGox's wallet implementation that caused some large number of bitcoins to be outright LOST forever.
The last time I checked, not even a poorly-designed wallet for paper money is capable of insidiously shredding or destroying it upon insertion, and modern paper money contains enough cloth, plastic, and security threads to ensure that even paper money that's been through a washing machine is probably still good enough for a bank to exchange for new bills.
Dealing with Bitcoin today is like being in a Medieval economy where everything not involving barter has to be done with cash, but with zero modern banking infrastructure to deal with storage and transportation. A sane non-drug-dealer would never, ever, EVER walk around with $20,000 in cash, because you might as well paint a target on your chest. With Bitcoins, you're basically relying on not being the victim of a metaphorical home invasion robbery just by having them in your possession. No Bitcoin "bank" is safe the way a FDIC-insured bank is a safe place to store up to $100,000... and more importantly, the existing Bitcoin "banks" are hopelessly outgunned by thieves. If a bank with the resources of Citibank or Deutsche Bank can't consistently outgun hackers, a "mom & pop" online bank with 20 employees has ZERO long-term chances of succeeding at it. At some point, we're going to see bad guys literally taking hostages, and killing employees until they unlock wallets for them.
Banks get robbed and have money stolen from them every day... we just don't care (much) anymore, because their losses are insured, so when it happens, we can just say, "ugh, shit happens." With cryptocurrency, there IS NO insurance.
As an American, Britain's government is simultaneously fascinating & terrifying.
On one hand, it usually behaves with a professionalism & responsibility that we can only envy.
On the other hand, Britain's government seems to be TERRIFYINGLY vulnerable to the legal equivalent of a multi-stage malware attack (with Parliament passing a sequence of innocent-looking laws that lay the groundwork for the vulnerability, then triggering their viral legal payload to fend off opponents by neutralizing any laws that might be used to fight them in court).
Just to add... this is basically a somewhat new application of several older technologies:
* a wavefront scanner that maps the actual shapes of the cornea, lens, and retina
* raytracing software, to calculate the shape of a complex, 2-surface lens. The hardest part about developing this software was learning where to step back and NOT try to fix some specific higher-order aberration. Glasses will never be positioned precisely (they slip, frames bend, etc), and if you try too hard to fix HOAs precisely, you'll make matters WORSE if the lenses deviate from their ideal positions. With scleral RGPs and Lasik, you can be more aggressive because the resulting lens is more stable. The trick was finding the happy medium that makes things "sharp" without making things "weird" when the glasses slide down your nose or the frames get slightly bent.
* CAD/CAM, allowing a robot to precisely grind a complex shape (calculated from data from the above) into two sides of a lens.
* Much of the lens theory originally developed for progressive bifocals... but applied to precisely correct focus, mag/minification, and astigmatism instead of merely transition different magnification strengths.
IMHO, if you have astigmatism, freeform digital aspheric lenses (when expertly-fitted) are a huge improvement over the mass-market norm... but if you have astigmatism AND need bifocals, it's absolutely a non-negotiable requirement. The catch is, you'll probably want to experiment with a few pairs of cheaper progressive lenses first to see what progressive shape you prefer (you might hate one layout, but be ok with another), and THEN spend the cash replicating that shape (but with improved optics) into freeform lenses after determining what it is that you actually *want*. It really isn't something you can determine through research alone... it takes some live experimentation, and you'll rarely be happy with your first attempt.
I'd argue that for every person who buys ill-fitting contacts with incorrect parameters with a fake prescription, there are THOUSANDS of people with perfectly valid prescriptions risking injury by wearing old/damaged contacts because they can't afford to replace them as frequently as they should. Low-cost replacements are a GOOD thing.
The truth is, most disposable soft contacts have SO MANY engineering compromises (especially toric ones), even flawlessly-fit lenses have pretty mediocre results, so comparing the best and worst is more like "kind of mediocre, vs not great"
It's like glasses... any halfwit with a ruler, a collection of lenses, and 5-10 pages of notes on fitting theory can come up with a reasonable set of +/- sphere values that are a net improvement over "none at all". Mitigating small amounts of astigmatism when looking straight ahead isn't much harder. Most low-cost glasses (under $100 lenses) aren't much better than this anyway.
So, what *does* require a skilled optometrist with substantial gear? 2-surface freeform aspheric lenses. Normal (sphere-only) lenses are molded as-is, cut, and polished. Cheap (sphere+cylinder) lenses have the sphere molded into them + the cylinder ground into one side. In both cases, visual magnification or minification occurs, which makes new lenses hard to "adapt" to & compromises depth-perception. But if you grind curves into BOTH the front AND rear side, you can SIMULTANEOUSLY correct focus errors AND neutralize-out magnification/minification.
For astigmatism, that's still not quite good enough... to correct sphere & cylinder (while neutralizing-out magnification/minification) across the entire lens field (vs "straight ahead"), you need to grind complex curves into both surfaces that are calculated via ray tracing... AND know how to properly measure additional parameters like the angle at which the lenses are tilted & their precise distance from the pupil... not all stores selling "HD lenses" do this properly, and if they don't, the results can be WORSE than cheap molded lenses. When done correctly, glasses with 2-surface freeform aspheric lenses won't distort your peripheral vision or distort geometry (or at least, won't do it nearly as badly as cheap glasses). You'll put them on, and things will just be sharper.
2-surface raytraced freeform lenses aren't something you'd WANT to buy online... a "normal" prescription (sphere+cyl+axis) doesn't have enough information, and every pair of non-identical frames will produce slightly different measurements for things like tilt, vertex, etc.
The problem? In the US, at least, chain vision stores are fixated on promoting things like "ultra-thin" lenses and "no-line progressive bifocals". The same technology behind them can be used to make near-ideal lenses for customers with astigmatism, but most front-line sales associates at those stores have no idea what you're even *talking* about if you say "custom-raytraced 2-surface (digital/HD) freeform aspheric lenses". The OPTOMETRIST might... but s/he's not the one who'll do the half-dozen extra measurements required to complete them. Most times, it'll be done by an employee who's literally winging it & doing it for the first time in weeks/months/ever.
The moral: if you have astigmatism & want genuinely better glasses, find an opthamologist who does Lasik (ie, who has the eye scanner & above-average training/experience) with on-site store & ask about "custom digital/raytraced/HD 2-surface freeform aspheric lenses". If the opthamologist looks confused or does anything besides confidently grin with delight because he'll get the satisfaction of fitting the best glasses money can buy... go somewhere else.
They won't be cheap, but you'll never be able to stand normal cheap lenses again. In theory, an optometrist could do it... but because the scanner is so expensive, they're usually only found at places that do Lasik (and by extension, have at least one opthamologist) since you NEED one for Lasik, and they're too expensive to buy JUST for eyeglass-fitting. And chain stores tend to "simplify" the fitting process, leaving you with compromised lenses.
Japan has a MUCH bigger problem with earthquakes & tsunamis than hurricanes (Tokyo & northward is about the same climate as the Northeastern US & maritime Canada). Tsunamis bring saltwater inland (very, very bad for underground power lines). Earthquakes shear underground power lines apart (or stretch them & cause subtle, harder-to-troubleshoot flakiness).
If the current fires encountered a large expanse of concrete devoid of anything directly combustible, how wide would it have to be to actually stop the fire's spread?
As I understand it, once uncontrolled outdoor fires reach a certain size, they act kind of like weak tornadoes that lift flaming objects high into the air & hurl them out to areas that might be several thousand feet away (enabling the fire to jump over thing like freeways, canals, etc).
If a house in the middle of an affected neighborhood had reinforced concrete walls & roof, plus Miami-grade impact-glass windows, would the heat of the fire as it burned down the neighbors' houses cause the concrete house's interior to combust anyway (like food debris in a self-cleaning oven)? Would ICF construction plus roll-down steel shutters keep the interior cooler, or would the intense heat just cause the ICF styrofoam itself to melt or combust?
I know that conventional wisdom is that individual homeowners are helpless against a fire, but I remember reading about one guy in California a few years ago who put sprinklers on his roof & surrounding yard, connected them to the faucet, and left it running when he evacuated. When he got home, his home had major "baking" damage... but his neighbors' homes were literally burned down to the scorched earth. I think some local official later decided to be a dick & fined him $10,000 for violating water-conservation rules to discourage others from trying to do it in the future.
Oh, and I forgot the other detail... $9.99/mo is the 6-month "promo" rate, with $69.99 regular price, 2 year contract, and outrageous ETF. Ditto, for the $99 plan... $99 for 6 months, jumping up to $149 thereafter with 2 year commitment.
I just realized I left out a major detail from my "$9.99/mo (plus fees)" scenario -- non-partner services wouldn't be BLOCKED... that would be too blatant & would guarantee pushback. They'd just be throttled to some point that's not glacial, but not fast enough to sustain realtime high-quality HD streaming video (say, 3-5mbps).
If your choices were:
a) $9.99/mo (+ fees) for (up to) 100mbps to "partners" (with nearly every major service being a partner) and 3-5mbps to "everyone else"
b) $99.99/mo (+ fees) for (up to) 100mbps to "partners" and 10mbps to "everybody else"
c) $249/mo (+ fees) for (up to) gigabit speeds to everyone
(all with "up to" gigabit local loop down & ~20-50mbps up... throttled speeds are per-host) ... you'd probably agonize long & hard between 'a' & 'b' (unless you really, really NEEDED 'c'). And you'd probably end up grudgingly choosing 'a' & hating them for charging SO MUCH MORE just to get the next step up.
ARM is only cheap & power-saving if you use it for undemanding uses on its own terms. An ARM that's beefed up to specs genuinely comparable to a desktop i7 uses as much power as (and usually costs a LOT more than) an actual i7.
Intel architecture is more efficient (behind the scenes, where it's REALLY executing RISC microcode PRETENDING to be x86) than many give it credit for. And x86/AMD64 and Windows have literally DECADES of mutual optimization for each other. It's going to be a very, very long time (if ever) before ARM's best can truly equal (let alone, outdo) Intel's best when it comes to Windows.
> Why does everyone who is pro-Windows do "heavy video" and "photo editing"?
Probably because support for hardware codec-acceleration has historically been one of Linux's major weak points. Most video codec hardware acceleration requires proprietary binaries and/or royalty payments (at least, officially).
This is now probably a bad example, but I remember that ~4 years ago, Windows Media Center could play h.264 1080p60 video without issues on my Thinkpad T61p... but the same stuttered & dropped frames if I booted Ubuntu and tried to watch the same video using VLC. 720p24 or 720p30 was pushing my luck, though 480p24/30/60 generally did OK.
Why? Windows took advantage of nVidia's hardware h.264 acceleration. Linux (in general) and VLC (specifically) didn't.
If you're importing lots of video source shot on consumer-grade cameras, the difference between "instantaneous" and "grind everything to a halt for several seconds and spike the CPU at 99%" is a big deal.
I believe recent VLC releases for Linux WILL (now) take advantage of h.264 hardware acceleration if the driver supports it... but AFAIK, the open-source nVidia drivers don't support it.
Playing videos isn't the same as editing, but editing is basically a superset of playing (importing, previewing, etc). If the platform can't play h.264-encoded videos smoothly, using it to EDIT video is going to leave you in a world of pain.
I fully support net neutrality, but one thing that seems to keep getting lost in discussions about Comcast vs Netflix is that back when Netflix live HD streaming was *new*, Netflix users were consuming *inordinate* amounts of bandwidth relative to "everyone else". On one hand, it was Comcast's fault for overselling their capacity... on the other hand, Netflix was an easy target PRECISELY because it was such a big, easy-to-see target.
The danger isn't that Comcast is going to start charging higher fees for Netflix, Hulu, Facebook, Youtube, and SlingTV... the danger is that Comcast will roll out some $9.99/month (+ thirty dollars of unavoidable fees) plan that provides "100mbps" connectivity ONLY to big, established companies willing to pony up and subsidize their cheap plan, then jack up the price of REAL 100mbps connectivity (to all internet hosts, including RDP servers and non-Facebook/Netflix/Hulu/etc) to $150/month.
It's a variant on Gresham's Law... cheap shitty service drives out good service & makes buying something even *slightly* better than "total shit" WAY more expensive.
Look at laptops... you can buy a shit netbook with 1.2ghz cpu, inadequate ram, useless keyboard that misses 1 in 20 keypresses, and hard drive that isn't even big enough to survive the next big annual Windows update for $199... but getting one that's meaningfully better (i5 or i7, 8-16gb ram, 256gb+ SSD, etc) increases the price to $1,000-2,000 because all of those shit netbooks soak up 99% of the economies of scale & make GOOD computer hardware 4-10x more expensive. And the pervasive existence of under-powered hardware induces companies like Microsoft to take away nice things like Aero Glass & replace it with ugly, awful things like Metro for the sake of making it run semi-acceptably ON shit hardware.
Disney could make a strong argument under 'equity' that bundling download codes with a physical disc as a freebie adds a small amount of value to the disc & takes away little/nothing from their potential sales, but a download code on its own has a MUCH higher value to the purchaser & probably represents a major revenue loss to the studio.
There's also a solid legal argument on licensing alone. Years ago, Microsoft won plenty of lawsuits against businesses & VARs who installed Windows using Enterprise site-licensed installation codes (using legit, purchased codes... but in a way not allowed by the license).
Hell, years ago, I knew someone who owned a video rental store in Miami. He had lots of foreign films (not released on NTSC VHS in the US). He'd buy retail tapes in Europe, make NTSC transfers himself, and rented both PAL/Secam original + NTSC transfer together as a single, inseparable unit... and STILL got sued by Sony or Disney (don't remember which one). He eventually got them to drop their lawsuit by agreeing to never do it with that studio's tapes again. Their lawyers were, shockingly, fairly open to reason & agreed that he was in a gray area and had made a good-faith (albeit unsanctioned) effort to do it legally, so they dropped the absurd multi-million dollar claim). It still hurt him, though, because it affected about a third of his foreign films.
My point is, even if you can prove the copyright owner didn't lose a cent in revenue, they STILL enjoy nearly absolute legal rights (in the US) to say 'no'. What Redbox is doing goes WAY beyond VHS-NTSC transfers of VHS-PAL/Secam tapes and results in actual lost original sales. If Redbox has a legal leg to stand on, it metaphorically has osteoporosis and polio.
The most tragic thing about Windows Mobile is that 5.x and 6.x were ugly as sin... but at their core, it was a good, stable, reasonably well-implemented (by real-world evolutionary standards, at least) OS. 7.x was going to be as pretty as Cupcake/Donut/Eclair Android.
As a practical matter, it was almost more de-facto "open" than Android is now. Sure, you can build AOSP from source, but if you want a device that can actually work as a non-crippled & dysfunctional "Android" phone, you need Google's very, very proprietary binaries. And hardware that doesn't lock you out of the bootloader. And compiled binaries from the SoC vendor... which unlike x86/AMD64 Windows, has basically zero compatibility with drivers for different kernels... change so much as a compiler-optimization flag, and existing binary kernel modules go "boom".
Microsoft *could* (and *should*) have had an official "App store", but like Android, WM imposed no mandatory restrictions on what you could download & run.
All MS *really* needed to do with WM7.0 was release it, and in WM7.5, they could have documented (properly) the way for third-parties to write things like a "phone" app (and added a C#.netCF API to allow apps like that to be written in C#.netCF).
If they'd stayed on course & done that, they would have probably gotten 80% of first-gen Android users to come back after one Android phone (because first-gen Cupcake-era Android phones WERE pretty dire compared to what WM7.0 was supposed to be). But no... Microsoft bought Danger (Sidekick OS), wasted a year porting it from Java to C#, then dumped WM for an inferior new OS that was superficially-prettier, but couldn't even do most basic BLUETOOTH-related things properly (a deficiency that pre-existed with Sidekick).
Microsoft NEEDS Bill Gates to pul a "Steve Jobs", come back and save Microsoft from itself. Pretty much every single thing Microsoft has done since Gates left MS has been bad. Windows 95's UI changes were genuinely "bold", "courageous", and a definite step in the right direction. Windows 8 (and Metro, and everything since) was NOT.
Microsoft has totally come unhinged, and they're destroying PCs with their craziness by refusing to acknowledge that the way people use high-powered desktop & laptop computers with multiple monitors is fundamentally different from the way little kids & tech-clueless adults use tablets, and trying to force a "One UI for All" policy on everything just degrades desktops & laptops into big, oversized tablets.
What Windows really needs is something akin to Android Fragments, so you can make one UI for tablets, one for real computers, then let the OS sort things out at runtime.
It has nothing to do with Linux per se, and EVERYTHING to do with the fact that KDE and Gnome (ok, Cinnamon, which as I understand it is Gnome 3 with Gnome 2 functionality restored) increasingly replicates Windows' UI better than Windows *itself* does.
Gnome and KdeWin are potentially a way to take control of Windows' UI future in a way that tells Microsoft, "We're staying with Windows... for now, at least... but piss us off even just a little, and we're bailing on you for good."
Truth be told, Windows' handling of proprietary binary drivers beats Linux's handling (or more precisely, its refusal to handle them) hands down. As companies like Qualcomm and MediaTek increasingly come to dominate hardware, we're going to *need* Windows' superior driver model (that doesn't automatically break all existing binary drivers every. goddamn. fucking. time. the kernel has even the *tiniest* change made to it, the way Linux does).
The fact is, we *need* Windows to not suck (and not self-immolate), if only to keep Linux on its toes and ensure x86/amd64 architecture has viable future. If ARM wins, ${deity} help us all, because the open(-ish) hardware architecture we know as the "PC" just doesn't exist in ARM-land. ARM shit is all proprietary, locked down, and massively empowers vendors over users. ARM doesn't even have a fucking mandatory baseline standard for "partition a hard drive in a way that allows it to boot any arbitrary OS"... literally everything is vendor-specific. If companies like Qualcomm had their way, PCs as we know them would cease to exist & we'd all be reduced to appliance-users who can do what the vendor explicitly chooses to allow, and nothing more.
The worst part is, unlike Windows 7 - vs - 8/10, Windows 10 users can't say "no thanks" and refuse the change. Eventually, Microsoft will make it a mandatory update & ram it down everyone's throat.
Unless, of course, KDE or Gnome for Windows gets taken seriously by someone like Adobe (and Gnome/KDE *themselves*), and Windows users start running KDE/Gnome Desktop in a full-screen Windows window as the penultimate 'fuck you' towards Microsoft. Once apps like Photoshop are Gnome/KDE-native, it would only take one major annoyance by Microsoft to get users to ditch Windows entirely.
Microsoft appears to have learned *nothing* from its Windows Mobile debacle (where they had the #1 mobile OS with capabilities that were *years* ahead of Apple & Android... then threw it all away and rendered themselves *irrelevant* as a mobile-device OS). Microsoft appears determined to render themselves equally-irrelevant in the non-phone realm (but will probably take most of the market for "real" PCs into their grave with them... the desktop/laptop market can't survive another "Windows 8" debacle).
Java's continued lack of an unsigned primitive 8-bit byte type is just inexcusable. It makes doing anything that involves binding with C++ (e.g., OpenGL ES) *excruciatingly* painful.
I would love to meet one person... ONE... who has EVER thought, "I'm *so* thankful Java's primitive bytes are signed instead of unsigned. That was a bold, awesome design decision."
If Java had an 'octet' type of unsigned 8-bit values, my life would have been IMMEASURABLY easier over the past 20 years, and I would have been spared weeks of troubleshooting insidious bugs that happened because I forgot a "&0xff" somewhere (or accidentally put it on the wrong side of parentheses).
The problem is, there's NO WAY for consumers to reliably know how badly LED lights flicker without buying and testing them directly. There's no "gold-star seal for bulbs with DC driver and adequate capacitor size" to distinguish them from cheaper designs that flicker at 120hz or 60hz.
Cree bulbs are reportedly among the WORST-flickering... many of their bulbs exhibit BLATANT 120hz flicker that can be seen with the naked eye... no need to test with a solar cell + oscilloscope, because you can see it directly.
Phillips is a mixed bag. Some of their bulbs have flicker-free DC drivers, but it's mostly luck of the draw... apparently, they're yet another company that thinks it's totally acceptable to change designs & components without changing the UPC or SKU, so one batch's 5-star reviews won't necessarily apply to the next batch.
IMHO, LED street lights are *worse* than high-pressure sodium. Especially in areas like rural freeway interchanges that are lit up like a stadium, but surrounded by inky darkness. Bluish LED lights destroy your night vision for a minute or two once you're back in the dark. Orange sodium lights aren't nearly as bad. And because they *aren't* as directional, there's a more gradual drop-off in brightness as you drive away.
With LEDs, it's like, "Darkness... darkness... BAM! Blinding blue-white light. BAM! Pitch black, with a few seconds of night-blindness to boot." With HPS lights, it's more gradual, and the light doesn't affect night vision as badly.
Update... I think my TV (Mitsubishi wd62627, circa 2006) can overlay content from an OTA channel from its internal tuners over component video, but not over HDMI, DVI, or VGA... and not vice-versa. It's not well-documented, and only certain permutations are allowed. In other words, let 'X' be the subset of video sources that can be full screen. Let 'Bx' be the subset of video sources that can be overlaid over each X.
So... to answer your question more fully... I've NEVER seen a TV that could do PiP with ANYTHING besides its two internal tuners... and MAYBE analog NTSC/PAL from composite-video inputs. Component video? Never. HDMI/DVI? Never. PIP appears to be implemented entirely within the tuner subsystem, so it would have to be done within an external box (regardless of whether the display is a monitor OR a TV)
AFAIK, there's one device you can buy to overlay computer-generated content onto HDMI (it uses the leaked HDCP master key if necessary). In the US, its legality is murky... almost certainly illegal (under DMCA) if the source is encrypted... possibly illegal if it's not (HDMI is a superset of DVI, but piles on additional restrictions that *might* apply if actual HDMI components -- including HDMI-to-DVI plug adapters -- are present.)
AFAIK, there is NO consumer-priced hardware (legal or otherwise) that can overlay HDMI source over another HDMI source. I'm sure somebody makes a $10k+ rack module intended for things like stadium displays that includes a 200-page license restricting the way you can use it, but a simple "HDMI+HDMI PIP box" just doesn't exist.
IMHO, HDCP's absolute control fetish goes WAY too far. The fact that its licensing body even REGARDS "unauthorized scaling, overlaying, and combining content from multiple sources" as a serious problem demanding aggressive enforcement illustrates just how far over the top and absurd they are. Apparently, they fear someone with legally-licensed content might replicate something like 1980s MTV (overlaying the video in a box while the VJ talks, then expanding it to fill the screen), or (shudder) "violate the artistic integrity of a director" (by allowing a box that bleeps out 'dirty words' based on subtitles... yeah, this was actually a real lawsuit, and Hollywood won).
Remember, DRM isn't satisfied with merely preventing COPYING... it goes 20 steps further to control every aspect of how you're allowed to VIEW content.
Security is also "unfair". You, as a conscientious developer, can do everything "right", and get totally pwn3d *anyway* because of some widespread, system/platform/framework/library vulnerability (perfect example: Heartbleed).
The only way to improve the odds is for a development team to have one or more members whose ONLY job is to be aware of every thirdparty library/platform/os used by the project & literally research every single one, every single day, to become aware of vulnerabilities as they're discovered... and for projects not under active development, maintain a working build environment at all times so *somebody* will be able to rapidly build updates to the app.
You can't expect developers to do it. We don't have time to recursively scrutinize our projects' thirdparty dependencies on a daily basis... at least, not if we need to actually get any *new* development done.
I'm not kidding about the build environment problem. I have a couple of published Android apps (developed with older versions of Android Studio) that get hopelessly mangled by recent versions of Android Studio. Google has done a *remarkably* shit job of breaking old projects with no working migration path beyond, "spend several days creating a new project from scratch & manually re-create the old project in the new release of Android Studio". I've heard similar tales of woe about Visual Studio... if you're *lucky*, an old project *might* load successfully into the *next* release of VS, but if you let your old project go *too* long, it's not likely to even be importable by future releases of VS without getting mangled into oblivion. And Visual Studio sinks its teeth *so* deeply into Windows, it's hard to keep old versions of VS around (and working) in perpetuity (without having mothballed old computers with it installed), and even harder to re-create those old installations years down the road.
Back in the happy days of Netbeans, Java, and Ant, you could generally build any old Netbeans project from the commandline using its generated Ant script, even IF a newer release of Netbeans broke it. With non-Gradle Android Studio 1.x projects... good luck... but you're probably fucked. StackOverflow has THOUSANDS of questions about "how to salvage old Android projects last built with now-obsolete IDEs (Eclipse, AS1.x, etc)".
or keep the same TV, and buy a new tuner box. ATSC 3.0 is a superset of ATSC 1.0's codecs & hardware, so any new 3.0 box will be 100% compatible with ATSC 1.0.
The big difference -- this time, a tuner box will cost about as much as a Roku, not $299-1999 like an ATSC 1.0 tuner did circa 1999-2004.
My main complaint over ATSC 3.0's omission of on-the-fly mode switching is the fact that they could have EASILY worked around the HDMI problem by having the TUNER transform everything into a specific mode (if the user desires).
IMHO, the new standard SHOULD have required that compliant tuners be able to deal with content that freely switches between modes like 720p120, 1080p60, and 2160p30 on the fly as appropriate for the content (720p120 for live-acton sports that need realtime encoding, 2160p30 for primetime TV shows that are pre-encoded in non-realtime, 1080p60 for most other stuff), then leave it up to the tuner box to transcode everything into a single mode (say, 2160p120 or 1080p60) if the TV can't handle on-the-fly mode changes.
The difference is, real 2160p120 needs a staggering amount of bitrate & processing power to ENCODE in realtime to h.265 prior to broadcast, but scaling UNCOMPRESSED bitmap frames is computationally-trivial. If the BROADCASTER has to do it, everything that wasn't 2160p120 has to be re-compressed. If the BOX does it, it can scale the uncompressed frame, then dump it straight to HDMI.
I also wish ATSC 3.0 did a better job of embedding metadata to help advanced receivers un-do the damage caused by scaling & re-encoding prior broadcast. For example, if the broadcaster took 30fps source and tweened it to 60fps or 120fps, receivers should be able to tell the difference between "real" frames & "synthesized" frames so they can throw away the synthetic frames and implement THEIR OWN (better, or at least user-configurable) algorithm for doing it. Or if 50fps-native content gets telecined to 60fps, tag the original frames so someone with a higher-end receiver and display capable of native 50fps can just buffer 5 frames, ignore the doubled sixth frame, and output the 5 frames ~100ms later at real 50fps (or locally-tweened to 100fps).
Yeah, a receiver can "sort of" sniff out the duplicate frames by examining the existing flags, but in practice this only works WELL if the content is a 24fps Hollywood movie (specifically, the movie itself... extras usually break the algorithms). For other content, the algorithms completely fall off the rails due to sloppy editing or splices made further down the pipeline. Keep the source video pristine (or at least, easily-restorable to its pristine state), and let the VIEWER's hardware do the heavy lifting.
Perfect example: "stretch-o-vision" vs pillarboxing. With ATSC 1.x, there's no easy way for a receiver to auto-detect 4:3 content stretched to 16:9 so it can de-stretch and pillarbox it instead. If the broadcaster decides to pre-stretch some (but not all) content & you HATE stretching, you're just fucked... if you're lucky, you can manually toggle de-stretching, but the TV can't do it automatically. With proper metadata documenting upstream transformations that have already taken place, the receiver end could undo those transformations automatically.